4.1 Layer Preparation Methods Based Solely on UPD Processes
85
potential shift if the peak position on the pre-coated substrate is compared to that of
the bare substrate. However, the UPD layer of one of the components also modifies
the work functions, hence giving rise to a contact resistance. If the latter effect is
dominating, the peak shift is negative. This was found for Bi deposition on Se-covered
Pt [23] and also for Bi deposition on Se-covered Au [22].
The deposition of the metallic component of an EC-ALD deposit (mostly Cu,
Zn, Ga, Cd, In, Sn, Hg, Pb, Bi) is a reduction process with no exception since the
precursor material of these components contain the cation (i.e., an oxidized form) of
these elements. By running a potential sweep from positive to negative potentials,
working regimes of ‘no deposit’ → ‘UPD layer(s)’ → ‘bulk deposit’ follow each
other, which dissolve electrochemically in the reverse order during an anodic sweep.
For some chalcogenide elements, the deposition in a reductive pulse is also possible
when the precursor compound contains an oxoanion of these elements in the solution
(H x SeO 3
(2−x)− , H x TeO 3
(2−x)− and H x AsO 3
(3−x)− from SeO 2 , TeO 2 and As 2 O 3 as
solute, respectively). However, the atomic layer of a chalcogenide element can also
be deposited with an oxidative pulse if the precursor material is the hydride of these
elements such as H 2 S (or Na 2 S) as demonstrated first for CdS by Colletti et al. [24],
and the oxidative S deposition became predominant later. In the case of oxidative
deposition, the order of the potential regimes of the surface-area-limited and bulk
deposition region is the opposite than for cations.
The selection of the UPD peaks is often carried out by varying the concentration of
the precursor material when cyclic voltammograms are recorded [11, 25]. However,
as the concentration of the precursor material is in the sub-millimolar range, the
position of the UPD peaks as observed in the cyclic voltammograms may vary,
especially at a high sweep rate [26]. Since the position of the stripping peaks does
not change with the solute concentration, this shift of the UPD peak can be attributed
to a sluggish kinetics of the UPD layer formation, to mass transport effects or both.
Apart from the concentration change, a shift in the cathodic end potential of the
cyclic voltammogram is often employed, which leads to a constant stripping charge
for UPD peaks and a growing bulk deposition peak as the negative potential limit
increases.
The diagnostic method for selecting the suitable deposition potential of component
“A” on an electrode pre-covered with a monolayer of the component “B” is the
stripping of component “A” under circumstances that leave the layer of “B” intact
(or after the stripping of component “B”). If we take into account one bilayer only, the
total charge corresponding to the stripping peak of the topmost component usually
shows a plateau region as a function of the deposition potential of this component. The
independence of the stripping charge from the deposition potential of the component
is the indication of the surface-area-limited nature of the process. A deviation from
the constant charge indicates potential-dependent coverage because of either the
formation of a sub-monolayer coating or that of a bulk deposit. This optimization
method of the deposition potential of both components was well exemplified by
Loglio et al. [27] whose work will be used below for reference. In this work, CdSe
was deposited by applying reduction to produce the monolayer of Cd and Se from
solutions containing Cd
2+ and SeO 3
2− ions, respectively. The Cd layer formation
85
potential shift if the peak position on the pre-coated substrate is compared to that of
the bare substrate. However, the UPD layer of one of the components also modifies
the work functions, hence giving rise to a contact resistance. If the latter effect is
dominating, the peak shift is negative. This was found for Bi deposition on Se-covered
Pt [23] and also for Bi deposition on Se-covered Au [22].
The deposition of the metallic component of an EC-ALD deposit (mostly Cu,
Zn, Ga, Cd, In, Sn, Hg, Pb, Bi) is a reduction process with no exception since the
precursor material of these components contain the cation (i.e., an oxidized form) of
these elements. By running a potential sweep from positive to negative potentials,
working regimes of ‘no deposit’ → ‘UPD layer(s)’ → ‘bulk deposit’ follow each
other, which dissolve electrochemically in the reverse order during an anodic sweep.
For some chalcogenide elements, the deposition in a reductive pulse is also possible
when the precursor compound contains an oxoanion of these elements in the solution
(H x SeO 3
(2−x)− , H x TeO 3
(2−x)− and H x AsO 3
(3−x)− from SeO 2 , TeO 2 and As 2 O 3 as
solute, respectively). However, the atomic layer of a chalcogenide element can also
be deposited with an oxidative pulse if the precursor material is the hydride of these
elements such as H 2 S (or Na 2 S) as demonstrated first for CdS by Colletti et al. [24],
and the oxidative S deposition became predominant later. In the case of oxidative
deposition, the order of the potential regimes of the surface-area-limited and bulk
deposition region is the opposite than for cations.
The selection of the UPD peaks is often carried out by varying the concentration of
the precursor material when cyclic voltammograms are recorded [11, 25]. However,
as the concentration of the precursor material is in the sub-millimolar range, the
position of the UPD peaks as observed in the cyclic voltammograms may vary,
especially at a high sweep rate [26]. Since the position of the stripping peaks does
not change with the solute concentration, this shift of the UPD peak can be attributed
to a sluggish kinetics of the UPD layer formation, to mass transport effects or both.
Apart from the concentration change, a shift in the cathodic end potential of the
cyclic voltammogram is often employed, which leads to a constant stripping charge
for UPD peaks and a growing bulk deposition peak as the negative potential limit
increases.
The diagnostic method for selecting the suitable deposition potential of component
“A” on an electrode pre-covered with a monolayer of the component “B” is the
stripping of component “A” under circumstances that leave the layer of “B” intact
(or after the stripping of component “B”). If we take into account one bilayer only, the
total charge corresponding to the stripping peak of the topmost component usually
shows a plateau region as a function of the deposition potential of this component. The
independence of the stripping charge from the deposition potential of the component
is the indication of the surface-area-limited nature of the process. A deviation from
the constant charge indicates potential-dependent coverage because of either the
formation of a sub-monolayer coating or that of a bulk deposit. This optimization
method of the deposition potential of both components was well exemplified by
Loglio et al. [27] whose work will be used below for reference. In this work, CdSe
was deposited by applying reduction to produce the monolayer of Cd and Se from
solutions containing Cd
2+ and SeO 3
2− ions, respectively. The Cd layer formation