9.1 SEM images of the Cu replica at low (i) and high (ii) magnification

A general remark after imaging the whole sample with both microscopy techniques is that the Cu features are randomly spread throughout the Cu replica;

their average width is ca. 200 nm. The height of the Cu forms is of ca. 400 ± 100 nm, while a small proportion of them (less than 1 %) reaches 900 nm in height.

The existence of Cu features having different shapes and orientations indicates that diffusion of the Cu²⁺ species is not hindered by the pores inner shape.

Consequently, electrodeposition of other metals within the pores of the TiO2 film might be feasible, although an small fraction of them is expected to effectively cross the microporous TiO2 film.

Fig. 9.1 SEM images of the Cu replica at low (i) and high (ii) magnification. Before imaging the sample was tilted by 45°. Features having finger shape (a, b), bridge and tree shape (c) are observed.

(ai) (bi) 700 nm

(ci)

(bii) (cii) 200 nm

283 nm

200 nm

283 nm 990 nm

700 nm

990 nm

700 nm

990 nm

(aii) 200 nm

283 nm

(a)

(c) 400 nm

564 nm 934 nm

1320 nm

(d) 600 nm

846 nm

600 nm

846 nm

(c) (b) Protective layer

4 µm

Cu Protective layer

Cu

Cu Cu

Fig. 9.2 FIB cross-sectional images of the Cu replica at low (a) and higher magnifications (b, c). Before imaging the sample was tilted by 45°. The contrast in the Cu layer corresponds to different plane orientations. The strips within the crystals are twinned regions.

Fig. 9.4 Superficial FIB images of a standard Ti/TiO₂/Pt specimen. The sample was tilted by 0° (a-c) and 45° (d-f). The holes in the surface represent unfilled micropores.

1.41 µm 1 µm

2 µm 2 µm

1

1µmµm 1 1 µµmm

500 nm Pt

T TiiOO22

(a) (b)

(c) (d)

(e) (f)

1010 µµmm PrProotteeccttiivve lelaayeyerr

(a)

1 µm

1.41 µm PrProotteeccttiivvee llaayyeerr

TTiiOO22 TTii

Pt (bi)

Pt

TiTiOO22

TiTi

(c)

500 nm

705 nm

(bii)

1 µm

1.41 µm p

pooreress

ppororeess

Fig. 9.5 Cross-sectional FIB images of a standard Ti/TiO2/Pt specimen at low (a) and higher (b-d) magnifications. Here (i) and (ii) correspond to the same Ti/TiO2/Pt interface before and after a thin slide has been ion milled. The small voids in the titania layer correspond to unfilled micropores oriented in different directions. The location of two vertical Pt-filled pores is marked with a white solid line.

Fig. 9.5 (Continuation).

1.41 µm 1 µm

(bi)

1.41 µm 1 µm

T TiiOO22 T Tii

Pt T

TiiOO22 TTii

PPrrooteteccttiivvee llaayyerer Pt

(d) (bii)

500 nm

705 nm p

pororeess

To explore the morphology of the Ti/TiO2/Pt system further, cross-sectional FIB images were recorded (Fig. 9.5). These images relate to a situation where the Pt-filled pores are oriented perpendicular to the sample surface. For clarity, the Pt deposits are marked with a white solid line. It is possible to identify circular holes in the titania film, corresponding to unfilled pores in accordance with cross- sectional TEM observations (section 8.4). The thickness of the anodic oxide film for this particular cross-section is 571 ± 25 nm, see Table A.4 in Appendix V for more details.

For the Pt-filled pore in Fig. 9.5c it is evident that the pore does not reach the underlying Ti substrate. More exactly, there is a considerable TiO2 "barrier layer" of ca. 100nm between the Ti substrate and the deposited Pt. Fig. 9.5d shows the formation of a Pt deposit from the filling of two closely spaced pores. A general observation after analysing several FIB cross-section images from the same sample is that only the pores, which traversed straight across the film, with a large diameter and a thin barrier layer (~ 0 to ~ 100 nm), acted as a nucleation centre for Pt. This can be easily explained, as the diffusional flux of the metallic species into wider and straighter pores, for subsequent reduction at the Ti or Ti/TiO2 interface, is expected to be less restricted compared to the smaller, more complex geometry pores. The typical mushroom shape of the Pt deposits could be easily explained in terms of ionic diffusion and current distribution (as discussed in section 4.2). The circular contour of the metallic deposits in the region near the TiO2 surface can be rationalised by hemispherical diffusion of the metallic cations and current crowding at the upper edge of the pores. Additionally, the upper surface of the Pt bumps, lightly flatter than a circular contour, can arise due to planar diffusion of Pt(IV) ions and planar current distribution in this region [140,182].

9.3 Ti/TiO2/Ir electrodes

9.3.1 Preparation from an Ir(III) solution

With the aim of coating Ir on Ti/TiO2 specimens and recovering a previously existing Ir(III) salt, a solution containing IrCl3·3H2O and HCl was employed. Reducing the metallic species from this solution was problematic due to the fact that this happened together with an important amount of hydrogen evolution, and consequently an extremely low deposition yield. Therefore long- term experiences were needed to eventually obtain a reasonable amount of Ir. This was determined indirectly by employing a special diaphragm cell, after H2

collection from the cathodic compartment as described in section 6.4.3 and exemplified in Appendix VI.

Experimental conditions leading to Ir deposition were characterised by a highly negative potential (in the range -0.6 to -1.5 V/NHE) and a change in the colour of both anolyte and catholyte solutions. In fact, the characteristic light yellow colour of the catholyte, corresponding to the [IrCl6]^3- species [183], gradually turned into dark brown. This might result from the mixture of [IrCl6]^2- species (generated at the anode), having a characteristic blue/purple colour [184], with [IrCl6]^3- species (see Fig. 9.6). In the same way, the anolyte, made of a white HCl solution, switched gradually to dark brown.

catholyte anolyte catholyte anolyte catholyte anolyte

(i) (ii) (iii)

diaphragm

Fig. 9.6 Changes in colour in the anolyte and the catholyte during the Ir electrodeposition process (from an [IrCl6]^3- solution); beginning (i), after ca. 5 mn (ii) and after ca. 15 mn (iii).

This is without doubt due to the diaphragm separating anolyte and catholyte, which allowed diffusion of ions in both directions.

It is worth mentioning that no Ir deposition was obtained without hydrogen evolution and when the colour of the solution remained unchanged, i.e., at potentials positive of ~ 0 V/NHE. As a consequence, Ir deposition only takes place after the [IrCl6]^3- species has been oxidised to [IrCl6]^2- at the anode. This is consistent with Mayne's work [183]. This author highlighted that Ir plating results only after the reduction of Ir(IV) to Ir(I) and then to Ir(0).

In the present work, the best results were obtained at a constant Q^* of 216 C cm^-2 using constant current densities in the range -60 to -180 mA cm^-2 leading to superficial concentrations comprised between 0.2 and 1.2 g m^-2. The resulting deposition yields are extremely low as displayed in Table 9.1.

Table 9.1 Average^* Ir electroplating yield as determined after H2 collection as a function of the experimental conditions employed.

Conditions yield superficial concentration

j (mA cm^-2) t (s) (%) (g m^-2)

60 3600 0.05 1.0

120 1800 0.02 0.5

180 1200 0.01 0.3

*In each case the average of three experimental data is displayed.

The best condition, in terms of a high deposition yield, was a deposition at 60 mA cm^-2 for 3600 s leading to a potential of ca. - 0.75 V/NHE as depicted in Fig. 9.7. Hereafter, the sample obtained in this fashion is referred to as Ti/TiO2/Ir⁽ⁱ⁾.

It is evident that the high amount of Q^* (216 C cm^-2) required for Ir plating compared to theoretical predictions (0.43 C cm^-2) is due to the extremely low current efficiency of the deposition process. This results from the high amount of H2 evolved and the competition between the reduction of Ir(IV) into Ir(I) and the reduction of Ir(IV) into Ir(III) [183].

-1500 -1000 -500 0 500 1000

-1000 0 1000 2000 3000 4000

V (V/NHE)

t (s) 1.000

0.500

- 1.000 - 1.500

V/(V/NHE)

0.000 - 0.500

Fig. 9.7 Typical voltage vs. time depence for Ir plating on a standard Ti/TiO2 in Cl^- medium.

9.3.2 Characterisation

Fig. 9.8 shows typical FIB images of a Ti/TiO2/Ir⁽ⁱ⁾ specimen at low (a) and high (b) magnification. In a similar fashion as for FIB images of the Ti/TiO2/Pt samples, the small dark zones in the TiO2 film are presumed to be non-conductive areas. Ir deposition appears to take place preferentially on the surface of the TiO2

as evidenced by FIB surface (Fig. 9.8) and cross sectional (Fig. 9.9) imaging. It is worth noting that the thickness of the TiO2 layer varies dramatically as can be observed in Fig. 9.9b, and has an average value of 483 ± 29 nm (see Table A.5 in Appendix V for more details).

CHAPTER 9: ELECTROPLATING 170

Fig. 9.8 FIB images of a Ti/TiO2/Ir⁽ⁱ⁾ sample at low (a) and high (b) resolution. The small holes in the surface correspond to unfilled micropores

1 µm

2 µm

Ir (a)

1 µm (b)

Fig. 9.9 Cross-sectional FIB images of a standard Ti/TiO2/Ir⁽ⁱ⁾ specimen at low (a) and higher (b-d) magnifications. The small voids in the titania layer correspond to unfilled micropores. The location of the superficial Ir deposit is indicated with a white solid line.

(a) TiO2 Ti

Protective layer Ir

(b)

(ci) (di)

1.41 µm 1 µm

705 nm 500 nm

1.41 µm 1 µm

705 nm 500 nm

(cii) (dii)

pores

( (aa))

Ir PrProotteeccttiivvee llaayyeerr

1100µµmm

1.41 µm 1 µm

1.41 µm 1 µm

705 nm 500 nm

(dii)

9.3.3 Preparation from an Ir(IV) species

Different factors affect the Ir plating efficiency:

(i) the reduction of Ir(IV) into Ir(0) requires the production of an intermediary species, the Ir(I), which is very unstable and easily undergoes oxidation into Ir(III) [183] as pointed out previously.

(ii) the Ir(IV) complexes have an extremely high binding energy, which enables hydrogen evolution before metal deposition [183,185,186].

(iii)hydration of Ir(IV) complexes improves the deposition yield to a certain extent. For example, the relative performance of the iridium chloride complex species, in terms of deposition yield, is as follows [187,184]:

[IrCl5(H2O)]^- >> [IrCl6]^2- > [IrCl4(H2O)2] > [IrCl3(H2O)3]⁺.

Consequently, in order to improve the Ir deposition yield a different procedure was followed:

(i) a solution containing [IrBr6]^2- rather than [IrCl6]^2- species was employed due to the fact that bromide is less electronegative than chloride, thus a lower binding energy is expected between the ligand (Br^-)and the metallic ion (Ir(IV)).

(ii) the electrolyte was prepared with a low bromide concentration to obtain Ir(IV) complexes with a high plating efficiency [188], likely being monohydrated Ir(IV) molecules.

(iii)during the electrodeposition, N2 was bubbled in the solution in order to avoid any dissolved O2 from oxidising the Ir(I) species as described in [189].

Experiments done at a high Q^* of 5.4 C cm^-2 gave enhanced deposition yields compared to previous results. The Ir plating efficiency ranged between 9 and 54 %. The optimum condition, in terms of a high current efficiency, was a constant current density of -1.5 mA cm^-2 for 3600 s leading to an average yield 50

% (see Table 9.4).

Fig. 9.11 FIB images of a Ti/TiO2/Ir⁽ⁱⁱ⁾ sample at low (a) and high (b) magnification. The sample has been tilted by 45° for imaging.

1 µm (a)

500 nm (b)

1.41 µm 1 µm

Ir

705 nm 500 nm

(a)

(b)

(c)

2 µm PrProotteeccttiivvee llaayyeerr

TiO2 Ti

Protective layer

Fig. 9.12 Cross-sectional FIB images of a standard Ti/TiO2/Ir⁽ⁱⁱ⁾specimen at low (a) and higher (b-d) magnifications. The small voids in the titania layer correspond to unfilled micropores. The location of the Ir deposit is marked with a solid line.

itip/i∞ -30

-20 -10

0 10

20

-30 -20 -10 0 10 20

00.10.20.30.40.50.60.70.80.91 0.10

0.30.2 0.50.4 0.70.6 0.8 0.91

-30 -20

-10 0

10 20

-30 -20 -10 0 10 20

0 0.2

0.4 0.6 0.8 1

1.2

0 0.2 0.4 0.6 0.8 1 1.2

-30 -20

-10 0

10 20

-30 -20 -10 0 10 20

0 0.2

0.4 0.6 0.8

1

0 0.2 0.4 0.6 0.8 1

(ai)

-30 -20 -10 0 10 20 30

-30 -20 -10 0 10 20 30

-30 -20 -10 0 10 20 30

-30 -20 -10 0 10 20 30

-30 -20 -10 0 10 20 30

-30 -20 -10 0 10 20 30

(aii)

(bi) (bii)

(ci) (cii)

y (µm)

0.00 0.92 itip/i∞

0.93 1.04 itip/i∞

0.00 0.08

x (µm) y (µm)

itip/i∞

itip/i∞

x (µm) y (µm)

x (µm)

Fig. 10.14 Typical SECM SG-TC x-y current maps for the diffusion-controlled tip detection (atip = 1 µm) of MV^+• (a) and [IrCl6]^2-(b) and (c), generated at an underlying Ti/TiO2/Pt substrate, from the diffusion–controlled reduction or oxidation of either 2 × 10^-3 mol dm^-3 MV²⁺ or 2 × 10^-3 mol dm^-3 [IrCl6]^3-.

itip/i∞

0 400 nm

0 0.2

C C B

B

B B

AA itip/i∞

CC B B

BB

AA

AA (a)

(b)

A A

Fig. 10.15 Simultaneously recorded fixed height current (a) and topography (b) maps for the diffusion-controlled tip detection of [IrCl6]^2- generated from transport-limited oxidation of [IrCl6]^3- at an underlying Ti/TiO2/Pt substrate. The area scanned was 10 x 10 µm (line rate of 1 Hz) in a solution containing 10 x 10^-3 mol dm^-3 [IrCl6]^3- and 0.5 mol dm^-3 KNO3. The AFM-SECM tip was characterised by an effective atip = 0.5 µm. Electrochemical data was acquired with the tip imaging at a fixed distance from the Pt deposits of 0.25 µm. The arrow denotes the unidirectional scan direction.