5.3 The Single-Bath Method
143
LN metal. In this case, the increase in the cathodic current density is faster than that
expected from the exponential rate law until the previously deposited MN metal is
fully covered with the LN one.
The anodic behaviour of the system after reversing the sweep direction is not
displayed because it substantially depends on a number of factors such as (i) the
relative position of the anodic partial polarization curves of the two metals, (ii) the
thickness of the MN metal on the deposit which may form when the deposition of
the LN metal stops in the anodic-going sweep. Below the trends concerning the
above-mentioned factors will be given.
In many cases, the onset of the dissolution potential of the LN metal is more
negative than that of the MN metal. This is the case for systems such as Co/Cu and
Co/Pd (in sulphate baths), Fe/Ag and Co/Ag (in perchlorate baths) and Ni/Cu (in
chloride baths). However, the onset of the dissolution of the LN metal can be more
positive than that of the MN metal if the LN metal is relatively inert; this is very
typical for Ni/Cu and Ni/Ag systems when halogenide ions are not present (i.e.,
sulphate and perchlorate baths, respectively) and also occur for the Co/Pb system (in
an acetate bath).
It cannot be stressed enough that when cyclic voltammetry is used for the characterization of the deposition and dissolution process, the elucidation of the anodicgoing sweeps is much more difficult than that of the cathodic-going ones. The reasons
for the difficulties are diverse and range much beyond the relative positions of the
partial polarization curves of the constituents [35]. When the anodic-going sweep
(usually the 2nd segment of the CV curve) is run, the potential reaches again a value
where the MN metal is deposited alone. Depending on how easily the MN metal
can cover the previously deposited layer of graded composition (where composition
grading comes from the variety of electrode potentials scanned before), the dissolution of the LN metal may or may not remain hidden. The “covering ability” of the
MN metal is generally the larger, the slower is the sweep, the less negative is the
cathodic limit of the polarization, the larger is the concentration of the MN metal
and the stronger is the structural coherence of the deposit. As it was shown for the
deposition of Co/Cu multilayers, the cyclic voltammetry is often insufficient to find
the right deposition potential of the MN metal of multilayered samples [35]. Similar
results were achieved when the quartz crystal microbalance was used to minimize
the unwanted dissolution of the LN metal [36].
It is very important to clarify whether the current or the potential control (i.e.,
galvanostatic /G/ or potentiostatic /P/ mode, respectively) is expected to yield a better
result during multilayer deposition. Let us start the analysis with the deposition of
the LN metal layer. Here, the deposition can be optimized for either of the electrical
control modes. In the G mode, the layer composition can be easily calculated from
the current density, as it was described for Fig. 5.3. When the LN layer is deposited
in the P mode, at least two additional factors are to be taken into account. First,
the current may vary in time as the concentration of the LN metal ions in diffusion
layer near the cathode decreases. This results in a graded composition, which may be
undesired. Secondly, the potential control requires a well-defined cell geometry. The
position of the reference electrode is particularly important due to the Ohmic drop
143
LN metal. In this case, the increase in the cathodic current density is faster than that
expected from the exponential rate law until the previously deposited MN metal is
fully covered with the LN one.
The anodic behaviour of the system after reversing the sweep direction is not
displayed because it substantially depends on a number of factors such as (i) the
relative position of the anodic partial polarization curves of the two metals, (ii) the
thickness of the MN metal on the deposit which may form when the deposition of
the LN metal stops in the anodic-going sweep. Below the trends concerning the
above-mentioned factors will be given.
In many cases, the onset of the dissolution potential of the LN metal is more
negative than that of the MN metal. This is the case for systems such as Co/Cu and
Co/Pd (in sulphate baths), Fe/Ag and Co/Ag (in perchlorate baths) and Ni/Cu (in
chloride baths). However, the onset of the dissolution of the LN metal can be more
positive than that of the MN metal if the LN metal is relatively inert; this is very
typical for Ni/Cu and Ni/Ag systems when halogenide ions are not present (i.e.,
sulphate and perchlorate baths, respectively) and also occur for the Co/Pb system (in
an acetate bath).
It cannot be stressed enough that when cyclic voltammetry is used for the characterization of the deposition and dissolution process, the elucidation of the anodicgoing sweeps is much more difficult than that of the cathodic-going ones. The reasons
for the difficulties are diverse and range much beyond the relative positions of the
partial polarization curves of the constituents [35]. When the anodic-going sweep
(usually the 2nd segment of the CV curve) is run, the potential reaches again a value
where the MN metal is deposited alone. Depending on how easily the MN metal
can cover the previously deposited layer of graded composition (where composition
grading comes from the variety of electrode potentials scanned before), the dissolution of the LN metal may or may not remain hidden. The “covering ability” of the
MN metal is generally the larger, the slower is the sweep, the less negative is the
cathodic limit of the polarization, the larger is the concentration of the MN metal
and the stronger is the structural coherence of the deposit. As it was shown for the
deposition of Co/Cu multilayers, the cyclic voltammetry is often insufficient to find
the right deposition potential of the MN metal of multilayered samples [35]. Similar
results were achieved when the quartz crystal microbalance was used to minimize
the unwanted dissolution of the LN metal [36].
It is very important to clarify whether the current or the potential control (i.e.,
galvanostatic /G/ or potentiostatic /P/ mode, respectively) is expected to yield a better
result during multilayer deposition. Let us start the analysis with the deposition of
the LN metal layer. Here, the deposition can be optimized for either of the electrical
control modes. In the G mode, the layer composition can be easily calculated from
the current density, as it was described for Fig. 5.3. When the LN layer is deposited
in the P mode, at least two additional factors are to be taken into account. First,
the current may vary in time as the concentration of the LN metal ions in diffusion
layer near the cathode decreases. This results in a graded composition, which may be
undesired. Secondly, the potential control requires a well-defined cell geometry. The
position of the reference electrode is particularly important due to the Ohmic drop