5.3 The Single-Bath Method
149
approach is to include a pause after the deposition of the Ni layer so that the concentration of the Cu
2+ ions can increase (while possibly some degree of cementation
may also take place), then the Cu deposition pulse can be adjusted in accord with
the achievement of the desired sample property [44, 59]. Instead of a pause with
no regulation of the electrode potential, a potentiostatic pulse was also applied with
more positive potential than the onset of the Cu deposition [60]. While cementation
in this case is not possible, it was speculated that this pulse may lead to the formation
of a monoatomic oxide layer on Ni that facilitates the Cu deposition later, but this
explanation has not obtained clear evidence.
If the dissolution of the LN metal takes place at more negative potentials than
the onset of the deposition of the MN metal (curves 1a and 2a in Fig. 5.3), a pause
after the deposition of the MN metals leads to displacement. However, the application of a third pulse after the deposition of the MN metal layer was assumed
to be suitable for the fine-tuning of the thickness of the MN layer. This idea was
tested for Co/Cu multilayers where the accomplishment of the Cu coverage at small
nominal Cu thicknesses cannot be achieved with the conventional two-pulse plating.
Therefore, after the deposition of a sufficiently thick Cu layer, a dissolution pulse
was applied with a smaller current density than that of the diffusion-limited one of
the Cu deposition, leading to a G/P/G pulse sequence [61]. The result was that the
samples were relatively rich in Cu as compared to their nominal composition, which
may stem from two effects: on the one hand, the dissolution of the Co layer during
the anodic pulse; and, on the other hand, the excess Cu deposition with Co at the
high-current pulse, which was made possible by the accumulation of the Cu
2+ ions
near the cathode during the dissolution pulse. As a whole, the G/P/G pulse sequence
with anodic dissolution after the Cu deposition did not lead to a breakthrough in
either the continuity of the layers or in their magnetic properties. It is to be noted
that similar reverse pulse techniques are occasionally applied in the preparation of
other nanolaminated coatings (like Ni x W 1−x /Ni y W 1−y [62]), although the reasoning
was rather empirical and the reverse pulse was rationalized by the improved surface
morphology without offering a detailed mechanistic picture.
Concerning the four-pulse method, it gained importance for magnetic/nonmagnetic multilayers only where either the thickness or the composition of the
magnetic layers varied, leading to two types of magnetic layers for odd and even
number magnetic layers, respectively. This multilayer type is called the pseudo-spinvalve structure. In the four-pulse method, a cathodic current is applied, regardless of
the polarization mode (G or P). The majority of the works relied on the Co/Cu system
[63–65]. Here, the thickness of the magnetic layers was modulated while the current
density during the deposition of the Co-rich layer was high and its Cu content was
kept below a few atomic per cent. However, when the magnetic components of the
LN layer were Ni and Fe, the modulation of the current density was of high importance since it regulated the ratio of the magnetic elements, hence leading to different
coercivity [66]. More complicated waveforms were also tested for multiplying one
of the types of the magnetic layer, while they were always separated with Cu layers.
However, the desired sample properties were mostly obtained for a small number of
repeat periods where the roughness increase with thickness did not counteract the
desired magnetic effect.
149
approach is to include a pause after the deposition of the Ni layer so that the concentration of the Cu
2+ ions can increase (while possibly some degree of cementation
may also take place), then the Cu deposition pulse can be adjusted in accord with
the achievement of the desired sample property [44, 59]. Instead of a pause with
no regulation of the electrode potential, a potentiostatic pulse was also applied with
more positive potential than the onset of the Cu deposition [60]. While cementation
in this case is not possible, it was speculated that this pulse may lead to the formation
of a monoatomic oxide layer on Ni that facilitates the Cu deposition later, but this
explanation has not obtained clear evidence.
If the dissolution of the LN metal takes place at more negative potentials than
the onset of the deposition of the MN metal (curves 1a and 2a in Fig. 5.3), a pause
after the deposition of the MN metals leads to displacement. However, the application of a third pulse after the deposition of the MN metal layer was assumed
to be suitable for the fine-tuning of the thickness of the MN layer. This idea was
tested for Co/Cu multilayers where the accomplishment of the Cu coverage at small
nominal Cu thicknesses cannot be achieved with the conventional two-pulse plating.
Therefore, after the deposition of a sufficiently thick Cu layer, a dissolution pulse
was applied with a smaller current density than that of the diffusion-limited one of
the Cu deposition, leading to a G/P/G pulse sequence [61]. The result was that the
samples were relatively rich in Cu as compared to their nominal composition, which
may stem from two effects: on the one hand, the dissolution of the Co layer during
the anodic pulse; and, on the other hand, the excess Cu deposition with Co at the
high-current pulse, which was made possible by the accumulation of the Cu
2+ ions
near the cathode during the dissolution pulse. As a whole, the G/P/G pulse sequence
with anodic dissolution after the Cu deposition did not lead to a breakthrough in
either the continuity of the layers or in their magnetic properties. It is to be noted
that similar reverse pulse techniques are occasionally applied in the preparation of
other nanolaminated coatings (like Ni x W 1−x /Ni y W 1−y [62]), although the reasoning
was rather empirical and the reverse pulse was rationalized by the improved surface
morphology without offering a detailed mechanistic picture.
Concerning the four-pulse method, it gained importance for magnetic/nonmagnetic multilayers only where either the thickness or the composition of the
magnetic layers varied, leading to two types of magnetic layers for odd and even
number magnetic layers, respectively. This multilayer type is called the pseudo-spinvalve structure. In the four-pulse method, a cathodic current is applied, regardless of
the polarization mode (G or P). The majority of the works relied on the Co/Cu system
[63–65]. Here, the thickness of the magnetic layers was modulated while the current
density during the deposition of the Co-rich layer was high and its Cu content was
kept below a few atomic per cent. However, when the magnetic components of the
LN layer were Ni and Fe, the modulation of the current density was of high importance since it regulated the ratio of the magnetic elements, hence leading to different
coercivity [66]. More complicated waveforms were also tested for multiplying one
of the types of the magnetic layer, while they were always separated with Cu layers.
However, the desired sample properties were mostly obtained for a small number of
repeat periods where the roughness increase with thickness did not counteract the
desired magnetic effect.