Size Effects in the Surface Properties of Electroplated Alloys between Iron Group Metals and Tungsten

Factors leading to size-dependent surface properties of electroplated coatings obtained by induced codeposition of iron group metals and tungsten are investigated. The size effect in microhardness described in earlier studies and the size effect in the corrosion resistance revealed in this work are shown to stem from a common origin, which is the formation of surface oxides. Removing surface oxides by abrasive processing leads to a higher corrosion rate and cancels the size effect in microhardness. Factors contributing to the formation of surface oxide layers during induced codeposition of considered alloys are studied.


INTRODUCTION
Research in the area of electrodeposition of alloys between iron group metals and tungsten in the form of bulk materials, thin films, and quasi-one-dimensional structures (e.g., nanowire, nanotube, and nanoribbon ensembles) is continuously expanding, because new mechanical, magnetic, and catalytic properties of these materials continue to emerge [1][2][3]. Manipulation of the composition and properties of such coatings depends on specific features of their electrodeposition mechanism, which, on the detail level, remains open to discussion. The process enabling us to obtain this type of coatings has been termed "induced codeposition" [4], because tungsten cannot be deposited from aqueous solutions. But alloys with a tungsten content up to 50 wt % (and more) can be deposited in the presence of certain complexes of iron group metals added to the electrolyte. Put differently, a complex of an iron group metal induces electrodeposition of tungsten to form a corresponding alloy. Because the deposition mechanism for this type of alloys has not been proposed yet, a number of "anomalies" are observed in their deposition. First, as was remarked in [1], the very fact of their deposition can be considered as an anomaly, because the classic electrochemical approach to controlling the composition of an electrodeposited alloy, which makes use of polarization curves for individual components, is not applicable to induced codeposition.
The macroscopic size effect in microhardness observed for such coatings should also be viewed as an anomaly [5][6][7][8][9][10][11]. The effect consists in that the microhardness of a freshly deposited layer depends not only on the composition, pH, and temperature of the plating bath, and current density (or potential) applied in electrodeposition but also the electrode surface area. Moreover, for a definite deposition current density and electrode surface area, the property (i.e., microhardness) depends on the electrolyte (bath) volume [9]. This necessitated us to introduce the volume current density (VCD), defined as the ratio of current to electrolyte volume (I/V, mA/L), a parameter which is not typically used in electrochemical materials science for controlling processes.
The macroscopic size effect was shown to occur in electrodeposition of Fe-W, Co-W, and Ni-W alloys from citrate-and gluconate-containing electrolytes [5][6][7][8][9][10][11]. In electrodeposition of Fe-W alloys from the citrate bath, the effect under discussion was shown to be caused by cathode surface oxidation due to chemical interaction between the intermediate species FeOH ads and water, which is a pathway alternative to reduction of this species to elemental iron [11]. The effect of VCD was shown to be related to a rapid change in the concentration of deposition-inducing metal, which is Fe(II) citrate for the case at hand: (1) or, substituting the current density of electrodeposition i = I/S into the above expression, we get where ΔС = С i -C x is the difference between concentrations of the deposition-inducing metal complex, with С i and С x being its initial and current concentrations, respectively; τ is the time; E q is the electrochemical equivalent; l is the current efficiency; I is the current; V is the electrolyte volume; and S is the surface area.
As can be seen from Eq. (2), the rate of change in the concentration of deposition-inducing metal increases with increasing the current density and the surface area (which is the cause of the mentioned size effect) and decreases with increasing the electrolyte volume at a definite deposition current density. From the above relations, which are based on the Faraday law, we can see that it is the rate of change in the concentration of deposition-inducing metal that contributes to a rise in VCD; favors the reduction of the indicated intermediate species by water, which results in oxidation of the surface [11]; and causes the influence of surface area on the microhardness (i.e., the size effect). Removing surface oxide aided in an increase in the microhardness, and the increase was greater when the VCD was higher. In addition, the size effect was not observed after the microhardness changed as a result of removal of surface film [11].
The deposition-inducing metal complex was shown [11] to be consumed in reduction reactions leading to the formation of elemental iron via a step that gives rise to the FeOH ads intermediate and oxidation on the anode, and it also catalyzes the reduction of tungsten species to elemental tungsten incorporated in the alloy. As was established earlier (e.g., in [10]), this was particularly the cause for the microhardness of Co-W layers deposited from a gluconate bath to display a strong dependence on the anode material; the smallest microhardness values were observed when using a platinum anode, because of the high rate of oxidation of the Fe(II) complex on platinum. Evidently, the size effect must diminish or disappear altogether if the deposition-inducing metal complex is not consumed or consumed at a lower rate.
Surface layer oxidation must manifest not only as the influence on microhardness but it will also determine, e.g., corrosion properties. We previously studied the size effect in corrosion rate in electrodeposition on nano-and microscopic surfaces [12,13]. In contrast, the aim of this work was to detect and investigate macroscopic size effects for the surfaces of alloys between tungsten and iron group metals by the example of deposition of Fe-W coatings from a citrate bath, including their corrosion rate.
It seems obvious that any anomalous manifestations in properties of considered materials and methods for their manipulation will be considered as such until the mechanism of the process is elucidated. We can therefore anticipate that studies of observed effects must give us deeper insights into the mechanism of induced codeposition.

EXPERIMENTAL
Considering the results of previous studies, we can state that the size effects are caused by surface oxidation that occurs in the conditions of induced codeposition of iron group metals and tungsten; i.e., these effects must not be observed or be considerably diminished upon (1) removal of the surface layer or (2) lowering the concentration of tungsten salt in the bath. Here, they are studied for Fe-W alloys electroplated from the citrate bath by measuring their microhardness and the corrosion rate before and after removing the surface layer that formed under the condition of constant current density (1) on cathodes with different surface areas and (2) at different VCD values (i.e., I/V, mA/L). The VCD was adjusted by both varying the cathode surface area and changing the electrolyte volume while keeping the surface area at a definite value.
The electroplating bath used in this work contained 0.2 M FeSO 4 , 0.17 M citric acid, and 0.33 M sodium citrate and the concentration of sodium tungstate was varied in the range of 0 to 0.5 M. The bath pH was 6.9 and electroplating was carried out at 80°С. The current density was constant at 20 mA/cm 2 in all experiments. Copper substrates used for alloy electrodeposition were preplated with a nickel layer (thickness 0.5 μm) deposited from a bath containing 240 g/L of NiCl 2 ·6H 2 O in diluted HCl (four parts of HCl (conc.) and one part of water); the current density was 30 mA/cm 2 , and the electrodeposition duration was 60 s.
Coatings used for microhardness measurements were deposited on the surfaces with an area of ~1 cm 2 in a 0.5-L cell using a consumable iron anode. VCD was varied in the range of 40-50 mA/L. In experiments for evaluating the corrosion resistance, electrodeposition was performed at the same current density and using a platinum anode; and the cathode surface area was varied from 0.1 to 2.5 cm 2 . These experiments were also carried out for a constant 0.5 cm 2 electrode surface area while varying VCD from 10 to 200 mA/L and the electrolyte volume from 50 mL to 1 L.
The chemical composition and morphology of prepared coatings were studied on a Hitachi TM360 scanning electron microscope (SEM) equipped with an attachment for energy-dispersive X-ray spectroscopy (EDS). Microhardness was measured at a load of 100 g using a PMT-3 hardness tester, and values reported here are averages with corresponding standard deviation.
The corrosion rate was measured for both freshly deposited layers and after removing part of the deposited surface layer with sandpaper for 30 s (type 64S, M-28 grit size, polishing rate of 0.006 g/min). Polarization curves were measured using an Autolab potentiostat (Metrohm) controlled with NOVA 2.1.4 software. Polarization curve measurements were carried out at room temperature in a three-electrode cell using an Ag/AgCl (sat.) reference electrode and a platinum counter electrode. Potential scans were performed in the anodic direction starting from the cathodic region, i.e., from -0.7 to +0.8 V, at a rate of 1 mV/s using 10 -3 M H 2 SO 4 and HCl solutions as corrosion media. The corrosion current and current density were determined by the polarization resistance method in accordance with the ASTMG59 Standard, and extrapolation of the initial regions of polarization curves to determine the corrosion potential was performed manually in some cases.

RESULTS AND DISCUSSION
The microhardness of Fe-W coatings deposited using different concentrations of sodium tungstate in the electrolyte was measured before and after mechanical removal of part of their surface layer by polishing, and corresponding dependences are shown in Fig. 1a. The dependences shown are for coatings deposited at a constant 20 mA/cm 2 current density and similar VCD values (40 and 50 mA/L). We highlight several specific features of these dependences: (1) in the absence of the tungstate from electrolyte, the coating microhardness did not change after removal of the surface layer, and (2) raising the WO 4 2concentration contributes to larger differences between HV values measured before and after removal of the surface film by sanding. As can be observed from Fig. 1b, raising the tungstate concentration (also to a certain limit) leads not only to an increase in surface hardness but also to a higher current efficiency (here, the value for electrochemical equivalent was taken to be 1.09 g/(A h), as in [11]). Both the increase in current efficiency and the growing difference in microhardness values before and after removal of the surface layer demonstrate that the rate of change in the concentration of deposition-inducing metal and conse-quently the extent of surface oxidation increase with raising the tungstate concentration in the electrolyte. Qualitatively, this conclusion follows from the difference in microhardness values measured before and after the surface was sanded.
Quantitatively, this is confirmed by the results of SEM studies and EDS analysis (Fig. 2). We can see that, after removing the surface layer, the oxygen was absent from the coating (Fig. 2a), while its concentration reached ~10 wt % in intact regions (i.e., where sanding was not done; Fig. 2b). We can draw another important conclusion from the electron microscopy studies: sanding removed only part of oxidized layer, not the entire layer (Fig. 2).
The cause of this lies in that the microthrowing power of the citrate electrolyte for electrodeposition of Fe-W coatings is negative, i.e., deposition occurs predominantly on asperities of natural inhomogeneities of the surface (a more detailed discussion will follow). This type of microscale distribution of deposition rates is a consequence of locally increased current density on the asperities. It is the asperities where elevated content of oxides in coatings is to be observed, because the rate of change in the concentration of depositioninducing metal grows with increases in the current density (Eq. (2)). But the oxide was also found in grooves after sanding (Fig. 2).
Thus, the results of our experiments qualitatively corroborate the phenomenological model introduced in [11], which describes the nature of size effects in properties of surfaces formed in induced codeposition of Fe and W during deposition of Fe-W coatings from the citrate electrolyte.

Size Effect in the Corrosion Properties of Fe-W Coatings
Potentiodynamic anodic-cathodic polarization curves (Fig. 3) and values for corrosion current densities calculated from them (Tables 1, 2) suggest that the corrosion properties also exhibit a size-dependent effect, because values for the corrosion current density registered when using Н 2 SO 4 and HCl as the corrosion media depend on the electrode surface area. Moreover, as can be observed from the calculated values for i corr , the removal of the surface layer led to an increase in the corrosion current density, i.e., the corrosion rate increased. Thus, the microhardness was lower in the presence of oxide in the surface layer, but, at the same time, the corrosion resistance was better.
The most salient manifestation of this is seen on the dependences of corrosion current density, registered in different corrosion media, on the electrode surface area for which the upper layer was removed before electrodeposition (Fig. 4). We emphasize that the difference in values measured for corrosion current densities in different corrosion media is substantially smaller than the differences due to the electrode surface area, and, therefore, data shown in Fig. 4 are the average values obtained for the different media. In addition, we can see that diminution of corrosion current density related to enlarging the surface area is observed not only before but also after sanding the surface (Tables 1, 2).
We also note the presence of slight but persistent anodic shift in the corrosion potential after the surface film was removed (Tables 1, 2), with the maximum diminution observed for the smallest surface areas. Evidently, simultaneous satisfaction of conditions for growth of corrosion rate after removing the surface film and the anodic shift in corrosion potential results from the fact that the corrosion rate is determined by the presence of passivating surface oxide film.
Because the deposition current density remained constant (20 mA/cm 2 ) in the experiments under discussion, and the electrolyte volume used in electrode-  However, this was not the only reason. As we will show below, the corrosion current density changed bỹ 50% upon changing the VCD by more than an order of magnitude while keeping the surface area at the same level, whereas the measured corrosion rate differed severalfold with a similar change in the cathode surface area (Fig. 4). The causes of observed specific behavior of the dependences of corrosion rate on surface area (i.e., the presence of size effect in the corrosion rate) will be discussed below.

Effect of VCD on the Corrosion Current Density
In contrast to experiments described in the preceding section, all electrodeposition experiments presented below were carried out not only at a constant current density (20 mA/cm 2 ) but also at a constant surface area (0.5 cm 2 ), and the VCD was adjusted by varying the volume of plating bath in the range of 50 mL to 1 L. The results are presented in Tables 3 and  4 and in Figs. 5 and 6. As can be seen, despite scatter-  ing of experimental data, the corrosion current density decreases when increasing the VCD, which demonstrates that the corrosion resistance of the coatings improved. As was the case in the preceding section, the data shown in Tables 3, 4 and Fig. 6 were obtained for different corrosion media before and after removing the surface layer. We note that the removal of surface film in experiments in which an electrode with constant surface area was used for electrodeposition also caused the corrosion potential to shift in the anodic direction (Tables 3 and 4), which is an indirect indication that the passivating film was removed.
We highlight that these results can be considered only as semi-quantitative, because it is difficult to quantitatively evaluate the extent of surface layer removal.

Nature of Macroscopic Size Effects in the Properties of Surfaces Obtained by Induced Codeposition
The results of our studies show that the macroscopic size effect manifests not only in the microhardness but also in the corrosion resistance in electrodeposition of iron group metals and tungsten (i.e., induced codeposition). Taking as example the electrodeposition of Fe-W coatings at a definite current density from the citrate bath, we show that the corrosion current density decreases severalfold as a result of enlarging the electrode surface area. We also found that the observed effect is caused by the oxide film present in the surface layer, and removing it contributes to reduction in the corrosion resistance. In contrast, the formation of surface oxide layer follows from the mechanism proposed in [11] (Fig. 7), which is based on conclusions made in studies [9][10][11][14][15][16][17][18]. According to the reaction pathways presented in Fig. 7, the oxidation of cathode surface occurs by reaction VI which, in turn, is due to an elevated rate of reduction in the concentration of deposition-inducing metal (Eqs. (1)-(2)). Admittedly, as can be observed from the results of present work and study [11], the reduction in microhardness [11] or enhancement of the corrosion resistance (this work) occur with increasing the current density, surface area, and electrolyte volume (Eq. (2)).
However, admittedly, this is not the only cause involved. As was mentioned above, due to this factor, the change in current density does not exceed 50% (Fig. 6), whereas a severalfold difference is observed when varying the surface area (Fig. 4).
The microthrowing power of electrolyte, which is negative in the case of codeposition of iron group metals with tungsten [19], is another factor causing the size effect (Fig. 8). In turn, the negative microthrowing power (i.e., deposition occurs predominantly on asperities) is caused by two factors: diffusion control of the electrodeposition process and the increasing    [11,20].
In this case, the distribution of local current density and, consequently, the deposition rate and oxygen content of coating will take the form as shown in Fig. 9. Roughness width a, defined here as the distance between valleys on the surface of a film in which surface oxides remained after abrasive treatment (Fig. 9), is ~40 μm (Fig. 10). It is absolutely clear that the number of roughness widths depends on the total surface area, and this results in an increase in the oxide content of the surface layer, which determines its properties (both microhardness and corrosion rate), and the presence of macroscopic size effect. According to the mechanism schematized in Fig. 7, the increase in oxygen content of the coating, which occurs when local current density and the surface area grow, while the average current density is maintained at the same level, must be accompanied by hydrogen absorption by the surface layer. Findings of study [21] provide indirect evidence for this possibility.
As a result, we have two interdependent mechanisms that determine the presence of macroscopic size effects in the surface properties of alloys obtained by induced codeposition. They are related to the development of oxide film (and perhaps hydrogen absorption) in the surface layer. The removal of film contributes to an increase in microhardness and worsening of the corrosion properties of the coating. With mechanical removal of the surface layer after electrodeposition, the size effect in microhardness disappears. The presence of this type of size effect complements the size effects observed earlier for other electrochemical systems [22].
It seems obvious that this type of effect must be crucial in scaling up laboratory experiments to industrial production, and this concerns both an increase in surface area and transition to microprocessing.   Fig. 8. Surface morphology of Co-W coatings electroplated at a current density of (a) 10 and (b) 50 mA/cm 2 from a citrate bath [19].

CONCLUSIONS
We showed that size effects observed under the conditions of induced codeposition of iron group metals with tungsten involved not only microhardness but also the corrosion rate. The observed effects stem from the formation of oxide layers in the surface layer of deposited coatings, with oxide content growing as the electrode surface area increases. The presence of surface oxide film results in diminution of microhardness, while the corrosion resistance improves. The formation of surface film, in turn, is caused by a fast decrease in the concentration of deposition-inducing metal complex. The high rate of consumption of this complex favors the chemical reaction between the FeOH ads intermediate and the solvent (i.e., water), which results in the formation of surface oxides and (possibly) hydrogen absorption by the surface. The removal of surface oxides by abrasive treatment can-cels the size effect in microhardness and diminishes the corrosion resistance of the surface.