Chapter 6

The Electrochemistry of Low Doped Diamond Films

 

6.0       Outline

 

·          The chapter describes the electrochemistry of low doped diamond films. Cyclic Voltammogams were recorded with a number of well known redox couples to show the behaviour of the films.

 

·          The effect of surface termination of the electrode was investigated.

 

·          The effect of the concentration of the electroactive species was studied.

 

·          The results of the electrochemical experiments are explained in terms of the theory outlined in chapters 4 and 7.

 

 

6.1       Experimental Set-up

 

The apparatus and procedures used were as outlined in section 5.1.

 

 

6.2       Surface Termination

 

To obtain Ohmic contacts to the diamond working electrodes, three layer metal contacts (3LM) and titanium underlayer contacts (TiUL) were used. The fabrication process for the formation of the 3LM contacts necessitated that the surface of the diamond was oxidised. TiUL did not require any special treatment of the diamond samples after they had been grown, so experiments could be performed on both native, hydrogen terminated specimens and oxidised specimens.

 

6.3       Cyclic Voltammetry of oxygen-terminated low-doped diamond

 

6.3.1    Cyclic Voltammetry of Ferrous Sulphate and Ferric Sulphate

 

Sample B131a was grown in a hydrogen/methane/diborane atmosphere with a [B]/[C] ratio of 37 p.p.m. in the gas phase (corresponding to a boron doping level of 7 ´ 1018 cm-1 if it is assumed that carbon and boron are incorporated into the diamond film in quantities proportional to their gas phase concentrations).

 

A 3LM contact was then applied to the surface of the diamond sample following the procedure outlined in table 3.1.

 

Figures 6.1 and 6.2 show cyclic voltammograms of sample B131a recorded in a aqueous solution containing 0.5 mol dm-3 H2SO4, 0.01 mol dm-1 FeSO4 and 0.005 mol dm-1 Fe2(SO4)3.

 

The oxygen-terminated low-doped diamond sample exhibited irreversible behaviour with the Fe2+/3+ redox couple. A distinct forward, oxidative peak can be seen but there is no clear reverse, reductive peak. These experimental results fit with the theory presented in the next chapter. At low doping levels, semiconductor behaviour is to be expected.

 

Figure 6.3 shows a cyclic voltammogram recorded with a platinum (Pt) wire working electrode in the same solution. Reversible electrochemistry is exhibited by the Fe2+/3+ redox couple at this metal electrode. Figure 6.4 shows a plot for diamond sample B131a superimposed on the scan for the platinum working electrode.

Figure 6.1 – CV of sample B131a in 0.5 mol dm-3 H2SO4, 0.01 mol dm-3 FeSO4
and 0.005 mol dm-3 Fe2(SO4)
2,  oxygen terminated surface, various scan rates, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.2 – CV of sample B131a in 0.5 mol dm-3 H2SO4, 0.01 mol dm-3 FeSO4 and 0.005 mol dm-3 Fe2(SO4)2,  oxygen terminated surface, scan rate 5 mV/s, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.3 – CV of a Pt wire WE  in 0.5 mol dm-3 H2SO4, 0.01 mol dm-3 FeSO4 and 0.005 mol dm-3 Fe2(SO4)2,  scan rate 50 mV/s

 

 

Figure 6.4 – CVs of diamond sample B131a & a Pt wire WE in 0.5 mol dm-3 H2SO4, 0.01 mol dm-3 FeSO4 and 0.005 mol dm-3 Fe2(SO4)2,  scan rate 50 mV/s

 

6.3.2    Cyclic Voltammetry of Potassium Ferrocyanide and Potassium Ferricyanide

 

Figures 6.5 and 6.6 show cyclic voltammograms of sample B131a recorded in a aqueous solution containing 1 mol dm-3 KCl, 3 ´ 10-4 mol dm-1 K3Fe(CN)6 and 3 ´ 10-4 mol dm-1 K4Fe(CN)6.

 

The oxygen-terminated low-doped diamond sample exhibited irreversible behaviour with the Fe(CN)63-/4- redox couple. A distinct forward, oxidative peak can be seen but there is no clear reverse, reductive peak.

 

Figure 6.7 shows a cyclic voltammogram recorded with a platinum (Pt) wire working electrode in the same solution. Reversible electrochemistry is exhibited by the Fe(CN)63-/4- redox couple at this metal electrode. Figure 6.8 shows a plot for diamond sample B131a superimposed on the scan for the platinum working electrode.

Figure 6.5 – CV of sample B131a in 1 mol dm-3 KCl, 3 ´ 10-4 mol dm-1 K3Fe(CN)6 and 3 ´ 10-4 mol dm-1 K4Fe(CN)6,  oxygen terminated surface, various scan rates, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.6 – CV of sample B131a in 1 mol dm-3 KCl, 3 ´ 10-4 mol dm-1 K3Fe(CN)6 and 3 ´ 10-4 mol dm-1 K4Fe(CN)6,  oxygen terminated surface, scan rate 5 mV/s, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.7 – CV of a Pt wire WE in 1 mol dm-3 KCl, 3 ´ 10-4 mol dm-1 K3Fe(CN)6 and 3 ´ 10-4 mol dm-1 K4Fe(CN)6,  scan rate 50 mV/s geometric area of working electrode = 7 mm2

Figure 6.8 – CVs of diamond sample B131a & a Pt wire WE in 1 mol dm-3 KCl,
´ 10-4 mol dm-1 K3Fe(CN)6 and 3 ´ 10-4 mol dm-1 K4Fe(CN)6,  scan rate 50 mV/s

 

 

Sample B129b was grown in a hydrogen/methane/diborane atmosphere with a [B]/[C] ratio of 50 p.p.m. in the gas phase (corresponding to a boron doping level of 9 ´ 1018 cm-1 assuming the solid phase concentration is proportional to the concentration in the gas phase).

 

A 3LM contact was then applied to the surface of the diamond sample following the procedure outlined in table 3.1.

 

Figures 6.9 and 6.10 show cyclic voltammograms of sample B129b recorded in a aqueous solution containing 1 mol dm-3 KCl, 3 ´ 10-4 mol dm-1 K3Fe(CN)6 and 3 ´ 10-4 mol dm-1 K4Fe(CN)6.

 

The oxygen-terminated low-doped diamond sample exhibited irreversible behaviour with the Fe(CN)63-/4- redox couple. A distinct forward, oxidative peak can be seen but there is no clear reverse, reductive peak.

 

 

6.3.3    Cyclic Voltammetry of Other Redox Couples

 

Attempts were made to study other well know redox couples. Experiments were performed with europium sulphate ( Eu2(SO4)3 ) and cobalt sulphate (CoSO4). Neither of these species were detected at the diamond electrode.

 

Figures 6.11 and 6.12 show cyclic voltammograms of sample B134b recorded in a aqueous solution containing 0.5 mol dm-3 H2SO4, 1 ´ 10-5 mol dm-1 Eu2(SO4)3. The scans show a small forward peak which was due to contamination by Fe(CN)64- ions. There is no evidence of europium electrochemistry at the electrode.

 

Contamination peaks could be removed by thorough cleaning of the apparatus.

 

No hypothesis has yet formed to explain the lack of response of Eu3+ and Co2+ ions at diamond electrodes.

Figure 6.9 – CV of sample B129b in 1 mol dm-3 KCl, 3 ´ 10-4 mol dm-1 K3Fe(CN)6 and 3 ´ 10-4 mol dm-1 K4Fe(CN)6,  oxygen terminated surface, various scan rates, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.10 – CV of sample B129b in 1 mol dm-3 KCl, 3 ´ 10-4 mol dm-1 K3Fe(CN)6 and 3 ´ 10-4 mol dm-1 K4Fe(CN)6,  oxygen terminated surface, scan rate 25 mV/s, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.11 – CV of sample B129b in 0.5 mol dm-3 H2SO4, 1 ´ 10-5 mol dm-1 Eu2(SO4)3 oxygen terminated surface, various scan rates, 3LM contact, Fe(CN)64- contamination geometric area of working electrode = 7 mm2

 

Figure 6.12 – CV of sample B129b in 0.5 mol dm-3 H2SO4, 1 ´ 10-5 mol dm-1 Eu2(SO4)3 oxygen terminated surface, scan rate 5 mV/s, 3LM contact, Fe(CN)64- contamination geometric area of working electrode = 7 mm2

 

6.4       Cyclic Voltammetry of hydrogen-terminated low-doped diamond

 

6.4.1    Cyclic Voltammetry of Potassium Ferrocyanide and Potassium Ferricyanide

 

Sample B144a was grown in a hydrogen/methane/diborane atmosphere with a [B]/[C] ratio of 50 p.p.m. in the gas phase (corresponding to a boron doping level of 9 ´ 1018 cm-1).

 

A TiUL contact was formed as outlined in table 3.2.

 

Figures 6.13 and 6.14 show cyclic voltammograms of sample B144a recorded in a aqueous solution containing 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6.

 

The hydrogen-terminated low-doped diamond sample exhibited considerably more reversible behaviour than was seen for the oxygen terminated samples. The scans are approximately symmetrical with clearly defined forward and reverse peaks.

 

If the electrode was not exposed to oxidising environments then the reversible behaviour could be maintained.  Figures 6.15 and 6.16 show a repeat experiment with sample B144a. The scans were performed two days after those shown in figures 6.13 and 6.14.  To maintain the unoxidised surface, high potentials and basic environments were avoided.

 

It should be noted that iR compensation was not used in these experiments. There was a potential drop across the electrolyte and so reversible voltammograms exhibited a peak separation of greater than 59 mV.

Figure 6.13 – CV of sample B144a in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  hydrogen terminated surface, various scan rates, TiUL contact geometric area of working electrode = 7 mm2

 

 

Figure 6.14 – CV of sample B144a in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6
and 0.01 mol dm-1 K4Fe(CN)6,  hydrogen terminated surface, scan rates 5 mV/s, TiUL contact geometric area of working electrode = 7 mm2

 

Figure 6.15 – CV of sample B144a in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  hydrogen terminated surface, various scan rates, TiUL contact geometric area of working electrode = 7 mm2

 

 

Figure 6.16 – CV of sample B144a in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  hydrogen terminated surface, scan rates 5 mV/s, TiUL contact geometric area of working electrode = 7 mm2

 

6.5       Cyclic Voltammetry of low-doped diamond with indeterminate surface termination

 

The results presented in section 6.3 showed electrochemistry at a well oxidised surface (diamond samples B129b and B131a).

 

Section 6.4 showed the results for a hydrogen terminated diamond surface (sample B144a).

 

In this section, results are presented for a series of experiments which were performed on a sample that had not been fully oxidised.

 

Sample B134b was grown in a hydrogen/methane/diborane atmosphere with a [B]/[C] ratio of 37 p.p.m. in the gas phase (corresponding to a boron doping level of 7 ´ 1018 cm-1).

 

A 3LM contact was applied to the surface of the diamond film.  Only the surface of the film that was to be placed under the metal contact was exposed to the full chromic acid treatment.  This left the surface of the diamond that was to be used as electrode unoxidised. However, it should be noted that the other post-treatments required to fabricate a 3LM contact may have altered the surface termination of the diamond.

 

6.5.1    Cyclic Voltammetry of Ferrous Sulphate and Ferric Sulphate

 

Figures 6.17 and 6.18 show cyclic voltammograms of sample B131a recorded in a aqueous solution containing 0.5 mol dm-3 H2SO4, 0.01 mol dm-1 FeSO4 and 0.005 mol dm-1 Fe2(SO4)3. A forward peak can be seen but no reverse peak can be seen.

Figure 6.17 – CVs of diamond sample B134b 0.5 mol dm-3 H2SO4,
0.01 mol dm-3 FeSO4 and 0.005 mol dm-3 Fe2(SO4)
2,  various scan rates,  3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.18 – CVs of diamond sample B134b 0.5 mol dm-3 H2SO4, 0.01 mol dm-3 FeSO4 and 0.005 mol dm-3 Fe2(SO4)2, scan rates 100 mV/s,  3LM contact geometric area of working electrode = 7 mm2

 

6.5.2    Cyclic Voltammetry of Potassium Ferrocyanide and Potassium Ferricyanide in an Aqueous Solution of Potassium Chloride

 

Figures 6.19 and 6.20 show cyclic voltammograms of sample B134b recorded in a aqueous solution containing 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6. The scans showed considerable peak separation.  Both forward and reverse peaks could be seen.

 

The electrochemical response changed with time and figure 6.21 compares two cyclic voltammograms recorded at a scan rate 200 mV/s. The first plot shows an early scan where forward and reverse peaks can both be seen. The second plot shows a later scan where the surface has become oxidised and the reverse peak is no longer visible. This change may be due to the oxidation of the diamond surface at high potentials (up to 2 V vs. Ag|AgCl (3 M Cl-) ).

 

The electrochemical cell was adjusted to expose a fresh area of the diamond sample to the electrolyte solution. The reverse peak could again be seen, as shown in figures 6.22 and 6.23.

 

Figure 6.19 – CV of sample B134b in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  various scan rates, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.20 – CV of sample B134b in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  scan rates 25 mV/s, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.21 – CV of sample B134b in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  scan rate 100 mV/s, 3LM contact,  before and after exposure geometric area of working electrode = 7 mm2

Figure 6.22 – CV of sample B134b in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  various scan rates, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.23 – CV of sample B134b in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6
and 0.01 mol dm-1 K4Fe(CN)6,  scan rates 5 mV/s, 3LM contact geometric area of working electrode = 7 mm2

 

6.5.3    Cyclic Voltammetry of Potassium Ferrocyanide and Potassium Ferricyanide in an Aqueous Solution of Potassium Hydroxide

 

Figures 6.24 and 6.25 show cyclic voltammograms of sample B134b recorded in a aqueous solution containing 1 mol dm-3 KOH, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6.

 

The presence of hydroxide ions in the electrolyte increased the rate of oxidation of the films and the electrochemical response rapidly became irreversible. Figures 6.26 and 6.27 show cyclic voltammograms recorded in a aqueous solution containing 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6.  The two figures compare the response before and after exposure to the basic solution. A reverse, reductive peak can be seen prior to the exposure to a basic medium. In the cyclic votammograms recorded after exposure, this reverse peak is not well defined.

 

Figures 6.28 to 6.32 show cyclic voltammograms recorded with a platinum (Pt) wire working electrode.

 

Figure 6.24 – CV of sample B134b in 1 mol dm-3 KOH, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  various scan rates, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.25 – CV of sample B134b in 1 mol dm-3 KOH, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  scan rates 5 mV/s, 3LM contact geometric area of working electrode = 7 mm2

 

Figure 6.26 – CV of sample B134b in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  scan rates 200 mV/s, 3LM contact, before and after hydroxide exposure geometric area of working electrode = 7 mm2

Figure 6.27 – CV of sample B134b in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  scan rates 100 mV/s, 3LM contact, before and after hydroxide exposure geometric area of working electrode = 7 mm2

 

Figure 6.28 – CV of a Pt wire WE in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6

 

Figure 6.29 – CV of a Pt wire WE in 1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  scan rates 50 mV/s

 

Figure 6.30 – CV of a Pt wire WE in 1 mol dm-3 KOH, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6

 

Figure 6.31 – CV of a Pt wire WE in 1 mol dm-3 KOH, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  scan rates 50 mV/s

 

Figure 6.32 – CV of a Pt wire WE in 1 mol dm-3 KOH/KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6,  scan rates 50 mV/s

 

 

6.6       Concentration Effects

 

The theory presented in chapter 7 suggests that at low concentrations of electrocactive species, there should be a switch from semiconductor behaviour to metal behaviour as the rate determining step changes from the electron transfer across the space charge region to the electron transfer across the Helmholtz layer.

 

Diamond sample B146b was grown in a hydrogen/methane/diborane atmosphere with a [B]/[C] ratio of 50 p.p.m. in the gas phase (corresponding to a boron doping level of 9 ´ 1018 cm-1).

 

A TiUL contact was formed as outlined in table 3.2.

 

The sample was given to an undergraduate project student who performed a series of experiments to investigate the effects of concentration on the electrochemistry of the film.

 

Figure 6.33 shows the results of the experiments.

 

The inset graph shows a cyclic voltammogram of sample B146b recorded in a aqueous solution containing 0.1 mol dm-3 KCl, 0.01 mol dm-1 K3Fe(CN)6 and 0.01 mol dm-1 K4Fe(CN)6 with a scan rate of 100 mV/s.  The scan showed a similar reversible response to that shown in section 6.4.

 

After the sample had been oxidised by exposure to various electrolytes, the reverse peak became less well defined. A series of different solutions were then used. The concentration of the electroactive species was varied from 1.5 ´ 10-3 mol dm-3 to 0.094 ´ 10-3 mol dm-3.

 

No clear switch to reversible behaviour was seen. However, the presence of impurities in the solution became significant at low concentrations. This may have masked any effect.

Figure 6.33 – CV of a diamond sample B146b in 0.1 mol dm-3 KCl, equimolar concentrations of K3Fe(CN)6 and K4Fe(CN)6,  scan rate 100 mV/s, concns (´ 10-3 mol dm-3): 10 (inset), 1.5, 0.75, 0.38, 0.19, 0.09 geometric area of working electrode = 7 mm2

 

6.7       Mott-Schottky Plots

 

Diamond sample B134b was grown in a hydrogen/methane/diborane atmosphere with a [B]/[C] ratio of 50 p.p.m. in the gas phase (corresponding to a boron doping level of 9 ´ 1018 cm-1).

 

A 3LM contact was then applied to the surface of the diamond sample following the procedure outlined in table 3.1.

 

Figure 6.34 shows Mott-Schottky plots, recorded at different frequencies, for sample B134b immersed in an indifferent electrolyte (1 mol dm-3 KCl). The amplitude of the AC potential modulation was 5 mV.

 

The plots are linear over a potential range of ‑0.2 V to 0.2 V and have a common intercept at 1.7 V. These results are comparable to those for B107 as presented in section 5.7.

Figure 6.34 – Mott-Schottky plots recorded at various frequencies in an aqueous solution of 1 mol dm‑3 KCl,  3LM contacts

 

6.8       Summary

 

At low doping levels, boron doped polycrystalline diamond films exhibited semiconductor behaviour.

 

At low doping levels, the electrochemistry of boron doped polycrystalline diamond films was dependant on surface termination.

 

Titanium underlayer (TiUL) contacts allowed Ohmic contacts to be made to hydrogen terminated diamond electrodes.

 

Oxygen terminated diamond samples exhibited irreversible behaviour with an absence of a reverse reductive peak.

 

Hydrogen terminated samples were more reversible. Both forward and reverse peaks were visible in cyclic voltammograms.

 

Mott-Schottky plots for oxidised diamond samples showed a flatband potential corresponding to a surface state at an energy level at Ev = 1.7 eV.