4.1 Layer Preparation Methods Based Solely on UPD Processes
83
4.1.3 Optimization of the EC-ALD Processes
It is essential that the first atomic layer exhibits a UPD process on the substrate.
It may occur that the components forming the alternating atomic layers undergo
UPD process on each other, but only one of them can form a surface-area-limited
monoatomic layer on the substrate. This was demonstrated for indium telluride deposition onto GaAs [19]. In such cases, it is essential that the first component to be
deposited can cover the substrate in a surface-area-limited manner, which has a
consequence on the right order of the layers to be deposited. The same situation
occurs when NiS is deposited onto silver substrate [20].
The primary optimization of the EC-ALD process involves the establishment of
the potential regions of the UPD process. When a single-crystalline substrate is used
as working electrode, a monoatomic layer of one of the components (A) is deposited
onto the working electrode. Then, the solution is exchanged under controlled working
electrode potential for introducing the precursor compound of the other component
(B), and cyclic voltammetry is used to establish at which potential this component
forms a UPD layer with the desired coverage. This preliminary screening of the suitable deposition potential of both components on the other component as substrate
has to be completed (“A” on “Su/B” and “B” on “Su/A”, Su being the substrate)
[7]. Figure 4.2 presents an example for the preliminary experiments and the establishment of the suitable deposition potentials. In the example given in Fig. 4.2, both
components are deposited with a reductive UPD process from an oxidized precursor
material.
The graphs in Figs. 4.2a, b, compare the reduction of Te(IV) on bare Pt and Sbcovered Pt surface, respectively. It can be seen that the peak intensities change, but
the shift in the peak positions is fairly small. Since the UPD coverage of Pt with Sb
is less than 1, UPD of Te on the Sb-modified Pt surface is at least partly due to the
Te–Pt interaction. The situation is very similar to the other pair of curves; i.e., Sb on
either Pt or Te-modified Pt. The diagnostic criterion for the appropriate deposition
potential in the initial EC-ALD cycles is as follows. If the cathodic-going sweep is
reversed at the right potential, the formation of the UPD layer is already complete,
but the forthcoming anodic-going sweep does not exhibit the stripping peak of the
bulk deposit (which is indicated as A3 in all graphs in Fig. 4.2). Since the deposition
of semiconductor components is seldom electrochemically reversible (i.e., the onset
of the dissolution potential of the bulk phase is more positive than the onset of the
formation of the same phase), it is advised that a number of cyclic voltammograms
are run with varying cathodic limit to establish the onset of the formation of the
bulk phase. This approach is particularly important when there is also a current
contribution due to hydrogen evolution.
The reactants for the preliminary study of an EC-ALD process shown in Fig. 4.2
were in the oxidized state, Te(IV) and Sb(III). Hence, the study of the reactant on
Pt could be performed by running cyclic voltammograms starting at the positive
potential limit. However, when the substrate is pre-covered with the UPD layer of
one of these components, it is important to keep an accurate potential control when
83
4.1.3 Optimization of the EC-ALD Processes
It is essential that the first atomic layer exhibits a UPD process on the substrate.
It may occur that the components forming the alternating atomic layers undergo
UPD process on each other, but only one of them can form a surface-area-limited
monoatomic layer on the substrate. This was demonstrated for indium telluride deposition onto GaAs [19]. In such cases, it is essential that the first component to be
deposited can cover the substrate in a surface-area-limited manner, which has a
consequence on the right order of the layers to be deposited. The same situation
occurs when NiS is deposited onto silver substrate [20].
The primary optimization of the EC-ALD process involves the establishment of
the potential regions of the UPD process. When a single-crystalline substrate is used
as working electrode, a monoatomic layer of one of the components (A) is deposited
onto the working electrode. Then, the solution is exchanged under controlled working
electrode potential for introducing the precursor compound of the other component
(B), and cyclic voltammetry is used to establish at which potential this component
forms a UPD layer with the desired coverage. This preliminary screening of the suitable deposition potential of both components on the other component as substrate
has to be completed (“A” on “Su/B” and “B” on “Su/A”, Su being the substrate)
[7]. Figure 4.2 presents an example for the preliminary experiments and the establishment of the suitable deposition potentials. In the example given in Fig. 4.2, both
components are deposited with a reductive UPD process from an oxidized precursor
material.
The graphs in Figs. 4.2a, b, compare the reduction of Te(IV) on bare Pt and Sbcovered Pt surface, respectively. It can be seen that the peak intensities change, but
the shift in the peak positions is fairly small. Since the UPD coverage of Pt with Sb
is less than 1, UPD of Te on the Sb-modified Pt surface is at least partly due to the
Te–Pt interaction. The situation is very similar to the other pair of curves; i.e., Sb on
either Pt or Te-modified Pt. The diagnostic criterion for the appropriate deposition
potential in the initial EC-ALD cycles is as follows. If the cathodic-going sweep is
reversed at the right potential, the formation of the UPD layer is already complete,
but the forthcoming anodic-going sweep does not exhibit the stripping peak of the
bulk deposit (which is indicated as A3 in all graphs in Fig. 4.2). Since the deposition
of semiconductor components is seldom electrochemically reversible (i.e., the onset
of the dissolution potential of the bulk phase is more positive than the onset of the
formation of the same phase), it is advised that a number of cyclic voltammograms
are run with varying cathodic limit to establish the onset of the formation of the
bulk phase. This approach is particularly important when there is also a current
contribution due to hydrogen evolution.
The reactants for the preliminary study of an EC-ALD process shown in Fig. 4.2
were in the oxidized state, Te(IV) and Sb(III). Hence, the study of the reactant on
Pt could be performed by running cyclic voltammograms starting at the positive
potential limit. However, when the substrate is pre-covered with the UPD layer of
one of these components, it is important to keep an accurate potential control when