Cu was electrodeposited on AZ91 D Mg alloy with zincate pretreatment in an alkaline plating bath, and the effect of zincate pretreatment on adhesion strength of the Cu layer was investigated. AZ91-type alloy is a MgAl binary phase system composed of two main phases: a Mg-rich -phase and an Al-rich -phase, Mg17Al12. From SEM observation of zincated AZ91 D alloy, a dense and compact Zn layer covered the -phase surface, but the -phase surface was covered with the corrosion film. Fig. 1 shows a schematic presentation of this non-uniform Zn deposition process. Initially, the thin Mg(OH)2 film reacted with pyrophosphate to form a soluble complex and thus the Mg substrate was exposed. Both -phase and -phase were dissolved and a corrosion film was formed on the substrate. Dissolution of Mg on the -phase surface caused enrichment of the surface concentration of Al and thus an Al hydroxide film was formed, obstructing electron transfer necessary for Zn deposition. On the other hand, the corrosion film formed on the -phase surface was mainly composed of Mg hydroxide with rather low density, which was gradually dissolved by pyrophosphate. This allowed Zn ions to reach the substrate through the corrosion film and to be reduced to form Zn deposits at the interface between the corrosion film and -phase surface. The area of Zn deposits increased rapidly as the corrosion film was detached from the -phase surface. Mg dissolution reaction was slowed down when the surface was covered with Zn deposits, and finally the -phase surface of the Mg alloy was almost entirely covered with a Zn layer.
As discussed above, Al rich corrosion film was not peeled off from the -phase surface, and thus disturbed Zn deposition. In order to improve the quality of the Zn layer, therefore, ultrasonic irradiation was applied to the initial stage of the zincate pretreatment for 240 s. Fig. 2 shows SEM images of Zn-coated samples prepared with (b and d) or without (a and c) ultrasonic agitation. The non-agitated sample shows Zn deposition on the -phase surface but not on the -phase surface. Cracks were observed on the -phase surface, and these cracks were probably formed in the Al hydroxide layer due to the drying process at the SEM observation. Observation of the samples prepared with ultrasonic agitation showed that most of the -phase surface was covered with a porous Zn layer. Such a porous structure was expected to improve adhesion strength of the successive electroplated Cu layer to the substrate.
After zincate pretreatment Cu was electrodeposited on AZ91 D samples, and adhesion strength of the plating film to the substrate was evaluated by the pull-off test. The results shown in Table 1 confirms that the adhesion strength was improved by the ultrasonic agitation. Fig. 3 shows SEM images of surface of the substrate (Fig. 3a) and the dolly used in the pull-off test (Fig. 3b). From EDS analysis, Zn and Cu were detected on the -phase surface of the substrate (#1 in Fig. 3a) and at the location on the dolly surface corresponding to the -phase surface before peeling off (#1 in Fig. 3b). On the other hand, signal of Zn or Cu was low on the -phase surface of the substrate (#3 in Fig. 3a), and Zn or Cu was mainly detected on the dolly surface at the location corresponding to the -phase surface (#3 and 2 in Fig. 3b). A part of the Zn film on the -phase surface near the -phase was also peeled off (#2 in Fig. 3a). Based on these results, it can be concluded that the plating layer on the -phase surface peeled off easily and transferred to the dolly side. On the other hand, adhesion of the plating layer on the -phase surface is relatively strong. Since the deposits remaining on the -phase surface and the deposits transferred to the dolly showed signals of Zn and Cu, peeling off seems to occur by cleavage of a mixed layer of Zn and Cu deposits. The peeling-off process is schematically presented in Fig. 4. A mixed layer of Zn and Cu was formed between the Zn layer and Cu layer because Cu was deposited rapidly on the substrate not only through the Cu electrodeposition process but also through the substitution reaction of Zn dissolution and Cu deposition. This process may have started immediately after the specimen had been immersed in the Cu plating bath and then Zn was electrodeposited within the Cu layer again. Results of the pull-off adhesion test indicated the quality of this mixed layer determines the overall adhesion strength.