1
The search for stable and efficient sonoelectrocatalysts for oxygen reduction and hydrogen peroxide formation: azobenzene and derivatives

2
We report the electrochemical reduction of oxygen using glassy carbon electrodes modified with azobenzene, hydroazobenzene, or fast black K salt (2,5-dimethoxy-4-[(4-nitrophenyl)azo] benzenediazonium tetrachlorozincate).

3
The performance of the electrodes under ultrasound was explored.

4
At the highest intensity (87 W cm−2) with one hour continuous insonation, voltammetric signals for azobenzene decreased by 7%, hydroazobenzene by 9% and fast black K by 18%.

5
The catalytic rate constants of the immobilised species towards oxygen reduction were assessed via cyclic, rotating disc and sono-voltammetry.

6
The rate constants were found to be 6.1 × 103 M−1 s−1 for azobenzene, 7.4 × 103 M−1 s−1 for hydroazobenzene and 10.4 × 103 M−1 s−1 for fast black K. These values suggest the use of these modified electrodes as practical hydrogen peroxide generators.

Introduction

7
The electrochemical reduction of oxygen to form hydrogen peroxide is of wide interest.1–4

8
The usual approach is through the immobilisation of catalytic materials to reduce the typically large over-potentials which are required for the direct reduction of oxygen at most electrode substrates.5

9
Electrode modification for electro-catalysis of oxygen reduction is popular since through immobilisation onto electrode surfaces via physically or covalent attachment electrocatalytic activities can be significantly improved.6

10
Many different materials have been proposed as electrocatalysts.

11
These include carbon nanotubes,7 Co(ii)-macrocycle complexes,8 poly(naphthoquinone) films,9 manganese dioxide,10 copper-nickel alloys,11 platinum modified gold electrodes,12 cobaloxime complexes,6 transition metal phthalocyanines,13 anthraquinone polymers14 and gold nanoparticles15 and anthraquinones.5,16–20

12
In the last example, immobilization of anthraquinone species onto the surface of electrodes leads to hydrogen peroxide production via the following reaction:5,6,18–21Q(ads) + 2e + 2H+(aq) → H2Q(ads)where Q(ads) is the adsorbed anthraquinone moiety.

13
A large number of synthetic chemists seeking greener syntheses are turning toward hydrogen peroxide based oxidations, since water is the sole by-product of these oxidations.22

14
Moreover expensive catalysts and solvents may be avoided while the need for costly organic peroxides or chlorine based oxidations can be eliminated.22

15
Noyori has pioneered ‘green’ oxidation chemistry with aqueous hydrogen peroxide as oxidant23 showing that cyclohexene can be oxidized to adipic acid,24 benzylic alcohols to benzaldehydes and benzoic acids.25

16
Also possible is the epoxidation of terminal olefins26 and the oxidation of alcohols.27

17
Recently trans-1,2-diols have been synthesized using olefins and hydrogen peroxide with high yield and selectivity.22

18
Hydrogen peroxide is made commercially via anthraquinone reduction known as the ‘AO-process’.28

19
This involves the hydrogenation, for example, of 2-alkyl-9,10-anthraquinone (with a catalyst such as nickel or palladium) forming the corresponding hydroquinone and oxidation with oxygen (usually air) to yield hydrogen peroxide and reforming the starting anthraquinone.

20
Commonly the anthraquinone is dissolved in a solvent or solvent mixture for hydrogenation, oxidation and extraction.

21
Due to the differing solubility of the quinone and hydroquinone species, solvent mixtures are commonly employed.28

22
The benefits of ultrasound in electrochemistry have recently been highlighted.29–31

23
In this context sustained high rates of mass transport are achieved from a combination of acoustic streaming and interfacial cavitational activity.

24
This is extremely useful in sono-electrosynthesis where the sustained transport can very significantly reduce electrolysis times whilst cavitational activity can keep the electrode ‘active’ under circumstances where passivation otherwise takes place.

25
Last it is worth noting that when ultrasound is applied to an aqueous media, cavitational collapse can produce low levels of in-situ formed hydrogen peroxide.30,32,33

26
Recently we have investigated the electro-catalytic reduction of oxygen at an Alizarin physically modified glassy carbon under ultrasound where catalytic currents were significantly enhanced under the high mass-transport conditions of ultrasound and quantitatively consistent with a modified Koutecky–Levich equation.20

27
Furthermore we have explored 9,10-phenanthraquinone and 1,2-naphthoquinone physically absorbed onto glassy carbon working electrodes and found them to be suitable as sonoelectrocatalysts for oxygen reduction and hydrogen peroxide formation.16

28
Such sono-electrodes offer scope for a green source of hydrogen peroxide in which oxygen (air) can be reduced with a high efficiency at an insonated electrode to provide a cheap and clean supply of hydrogen peroxide at the point of usage.

29
In this paper we report the use of azobenzene, hydroazobenzene and fast black K salt (2,5-dimethoxy-4-[(4-nitrophenyl)azo] benzenediazonium chloride) for the catalytic reduction of oxygen.

30
Azobenzene can be adsorbed either in its oxidised form, or in its two electron reduced form, hydroazobenzene: C6H5–N=N–C6H5 + 2e + 2H+ ⇌ C6H5–NH–NH–C6H5All three species were physically adsorbed onto glassy carbon electrodes with the catalytic rate constants assessed using cyclic, rotating disc and sono-voltammetry.

31
The stability of the modified glassy carbon electrodes were investigated under ultrasound and were found to be extremely stable providing a basis for the sono-electrochemical production of hydrogen peroxide.

Experimental

Reagents and chemicals

32
All chemicals used in this work were of analytical grade and used as received without further purification.

33
These were: azobenzene (Aldrich, 96%), fast black K salt (2,5-dimethoxy-4-[(4-nitrophenyl)azo] benzenediazonium tetrachlorozincate, Sigma,+98%), hydrazobenzene (Aldrich, 99%) and acetonitrile (Fischer Chemicals, synthetic grade).

34
The buffers used were as follows: acetate buffers for pH 3.0 and 5.0, phosphate buffers for pH 2.0, 7.0 and 8.0 and boric buffer for pH 10.0.

35
All solutions were prepared with deionised water of resistivity not less than 18.2 MΩ cm (Vivendi Water Systems, UK).

36
Before commencing experiments, nitrogen (BOC, Guildford, Surrey, UK) was used for deaeration of solutions.

37
Conversely, for oxygen measurements, oxygen (BOC, Guildford, Surrey, UK) was bubbled for at least 10 min to saturate the solution.

Instrumentation

38
Voltammetric measurements were carried out on a μ-Autolab (ECO-Chemie, Utrecht, The Netherlands) potentiostat.

39
All measurements were conducted in a thermostatted (22 °C) three-electrode cell with a solution volume of 40 cm3.

40
The working electrodes were either an edge plane pyrolytic graphite electrode (Le Carbone Ltd, Sussex, UK, 5 mm diameter) or a glassy-carbon electrode (GC, 3 mm diameter, BAS, Indiana, USA).

41
A different glassy carbon disc (3 mm diameter, mounted in teflon) was used for rotating disc experiments.

42
The counter electrode was a bright platinum wire, with a saturated calomel electrode (Radiometer, Copenhagen, Denmark) completing the circuit.

43
Sonoelectrochemical studies were performed using an ultrasonic horn, model CV 26 (Sonics and Materials Inc. USA), operating at a frequency of 25 kHz and fitted with a 3 mm diameter titanium alloy microtip (Jencons, Leyton Buzzard, UK).

44
The intensity of the ultrasound was determined calorimetrically34–36 and it was found that a 20% amplitude corresponded to 87 W cm−2.

45
The working electrode was placed in a face-on arrangement to the ultrasonic horn.

46
The horn was immersed beyond the shoulder of the stepped tip to ensure that ultrasound is efficiently applied into the solution.

47
The horn-to-electrode distance was fixed at 10 mm throughout all measurements.

48
During experiments, the temperature was kept constant at 22 °C (±2 °C) by the use of a stainless steel cooling coil.

49
It is important to note that for sono-Koutecky–Levich measurements, the concentration of oxygen under insonation has already been investigated where it was observed that ultrasound can be applied for up to 2 min before 10% of the oxygen is lost20.

Electrode modification

50
GC electrodes were carefully pre-treated before modification since it has been reported37 that the electrode history has effects on both the modification and the electrochemical behaviour of the electrode.

51
Prior to each modification, GC electrodes were polished with diamond lapping compounds (Kelmet, UK) and then carefully rinsed.

52
The oxidative pre-treatment of GC electrodes was the same as the one reported by Salimi and co-workers:20 the polished GC electrodes were subjected to cycling between −0.5 and +2.0 V at a scan rate 100 mV s−1 in 0.1 M sulfuric acid for 10 min.

53
The electrode was then poised at +1.8 V in the same solution for 3 min.

54
The pre-treated electrode was modified with azobenzene by pippetting 20 μL of 1 mM solution of azobenzene in acetonitrile on to the electrode surface.

55
The solvent was allowed to evaporate off at room temperature leaving azobenzene physically adsorbed on to the GC electrode.

56
Note that, as discussed below, microcrystals of azobenzene are also formed by this procedure so that the electrode surface immediately after modification has both physically adsorbed azobenzene and microcrystals on it.

57
The latter can be removed by potential cycling (vide infra).

58
Glassy carbon electrodes modified with hydrazobenzene or fast black K salt were prepared by the same procedure.

59
Note that if the electrode is simply placed into a 1 mM solution of azobenzene, left for 5 min, removed and rinsed and then placed into a pH 2 buffer solution, no voltammetric waves are observed.

Results and discussion

Characterisation of azobenzene modified electrodes

60
The voltammetry of azobenzene at an edge plane pyrolytic graphite electrode was first explored.

61
The electrode was modified as described in section 2.3 and placed into an oxygen free pH 2 phosphate buffer solution with cyclic voltammograms recorded over various scan rates.

62
As shown in Fig. 1, a reduction peak is visible at ca. −0.60 V (vs. SCE) and this corresponds to the reduction of azobenzene to hydroazobenzene as described by eqn. (3)38,39 with the corresponding oxidation peak observed at ca. +0.40 V. Next a new electrode was prepared and the voltammetric window extended to more negative potentials.

63
The first scan is shown in Fig. 2, curve A, where the reduction peak of azobenzene to hydroazobenzene is first observed with the second wave at ca. −1.2 V (vs. SCE) corresponding to the irreversible reduction of hydroazobenzene to aniline.38

64
[C6H5–HN–NH–C6H5 + 2e + 2H+→ 2C6H5–NH2]On the second scan (Fig. 2, curve B) a single reduction wave is only seen corresponding to azobenzene reduction, with continual scanning resulting in a decrease of the voltammetric response, consistent with literature reports that hydroazobenzene is reduced irreversibly to aniline.38,39

65
A polished edge plane electrode was left in a 2 mM acetonitrile solution of azobenzene for 12 h, removed, carefully rinsed and put in fresh, oxygen-free pH 2 phosphate buffer.

66
It is interesting to note that cyclic voltammograms recorded after this treatment show a small reduction wave at ca. −0.6 V (vs. SCE), which is ca. seven times smaller than that seen compared to the electrode modified via evaporation.

67
We next turn to investigating the response of a modified glassy carbon (GC) electrode.

68
The electrochemical behaviour of the modified GC electrode was examined in an oxygen free pH 2 phosphate buffer solution.

69
The first scan is shown in Fig. 3A which shows two reduction waves at −0.11 (±0.01) V and −0.45 (±0.01) V (vs. SCE) with a single oxidation wave on the cathodic sweep.

70
Fig. 3B shows the second scan (curve A), fifth (curve B) and the final stabilised scan (curve C).

71
Comparison of the first scan (Fig. 3A) and second scan (Fig. 3B, curve A) reveals a large decrease in the peak current.

72
This suggests that when the electrode is modified a mixture of azobenzene microcrystals and a monolayer (or less) exists such that as the potential is scanned to negative values, the first reduction peak at −0.11 V is the reduction of azobenzene to hydroazobenzene (eqn. (3)) with both species in the form of a monolayer while the large broader peak at −0.45 V is the reduction of azobenzene in the form of crystals which dissolve from the electrode surface when reduced, such that on the second scan the large broad peak is absent suggesting that only the monolayer (or less) response is left.

73
We note that Komorsky-Lovrić has studied the voltammetry of azobenzene microcrystals abrasively attached to a basal plane pyrolytic graphite electrode, observing a reduction peak at ca. −0.40 V at pH .238

74
The effect of pH on the response of azobenzene was explored (after the signal had stabilised).

75
It was observed that redox peak was pH-dependent and shifted towards more negative potentials upon increasing the solution pH.

76
The potential of the reduction peak shifts by 57 mV (pH unit)−1 consistent with a two-proton two-electron process.

77
A plot of peak current vs. scan rate was found to be linear, as expected for surface confined species with the surface concentration of azobenzene physically adsorbed on glassy carbon electrode (after stabilisation) was calculated using the following relationship:Γ = Q/nFAwhere Q is charge under voltammetric peak, n is the number of electrons transferred, F is the Faraday constant and A, the electrode area.

78
The surface coverage was found to be 68.9 (±0.6) × 10−10 mol cm−2.

79
The reduction of oxygen on glassy carbon electrodes modified with monolayer physically adsorbed azobenzene was next studied.

Oxygen reduction at azobenzene (AB) modified GC electrodes

80
The reduction of oxygen using an azobenzene modified glassy carbon electrode was next studied.

81
First, cyclic voltammograms were run in an oxygen-saturated solution.

82
Shown in Fig. 4 are the responses obtained for a bare glassy carbon electrode (curve C) and for a modified glassy carbon electrode modified with azobenzene in a pH 2 phosphate buffer solution saturated with oxygen (curve B).

83
For comparison the modified electrode in the absence of oxygen (curve A) is shown.

84
At a bare electrode, the oxygen reduction occurs with a peak potential at ca. −0.53 (±0.01) V (vs. SCE).

85
For the modified electrode a large reduction wave is observed with the corresponding oxidation peak disappearing on the reverse scan suggesting excellent catalytic behaviour.

86
Comparison of oxygen reduction at the modified and bare glassy carbon electrodes reveals that the oxygen wave in the latter case shifts to less negative potentials from −0.53 (±0.01) V to −0.37 (±0.01) V.

87
Cyclic voltammograms were next recorded in various pH buffer solutions saturated with oxygen with the ‘best’ pH for oxygen reduction investigated; this criterion is met when the oxygen reduction wave using the modified electrode is at its most positive potential.

88
This was found to correspond to pH 2.

89
Consequently all further electrochemical studies using azobenzene were carried out at pH 2.0.

90
The electrocatalytic rate constant for oxygen reduction on a GC electrode modified with azobenzene was determined from cyclic voltammetry data using Andrieux and Savéant theory:40Ip = yFACbulkD1/2(Fv/RT)1/2where F is the Faraday constant, A is the electrode area, Cbulk is the bulk concentration of oxygen, D is the diffusion coefficient of oxygen, ν is the scan rate.

91
Experimentally, the modified electrode for the voltammetric reduction of oxygen in an oxygen-saturated solution is scanned at slow scan rates.

92
The gradient of a plot of peak current versus square root of scan rate produces y.

93
This value is used via a theoretically derived curve to determine x, where x is:40where the symbols have their usual meanings.

94
Shown in the electronic supplementary information (ESI) is a plot of peak current versus square root of scan rate, which is linear.

95
The rate constant from this plot using the Andrieux and Savéant theory was found to be 6.1 (±0.1) × 103 M−1s−1 for the surface coverage of azobenzene of 69 (±0.3) × 10−10 mol cm−2 using the oxygen diffusion coefficient of 1.6 × 10−5 cm2 s−1 and a value of 1.2 mM as the concentration of an oxygen saturated aqueous solution.18

96
For comparison, the kinetic parameters for the oxygen reduction reaction on a GC electrode modified with azobenzene was next evaluated using the rotating disc electrode (RDE) method via the Koutecky–Levich equation:41where k is catalytic rate constant, Cbulk the bulk concentration of oxygen, ν is the hydrodynamic viscosity, ω is the rotation speed, while all other parameters have usual meaning.

97
First however, a glassy carbon rotating disc electrode was modified with azobenzene by the same procedure as used for the GC electrode as described in section 2.3.

98
Using cyclic voltammetry the coverage of azobenzene on the RDGCE was found to be 66.4 (±0.5) × 10−10 mol cm−2.

99
Rotating disc voltammograms at a range of rotation speeds were recorded in oxygen saturated solution.

100
Using a potential of −0.37 V (vs. SCE), a Koutecky–Levich plot was constructed.

101
The rate constant was evaluated from the intercept of Koutecky–Levich plot.

102
Using this plot, the catalytic rate constant was found to be 5.9 (±0.2) × 103 M−1 s−1, which is in excellent agreement with the value obtained via cyclic voltammetry data.

103
Next, we turn to oxygen reduction at a glassy carbon electrode modified with azobenzene using sonovoltammetry.

104
First we consider the behaviour and stability under ultrasound of the azobenzene modified glassy carbon electrode: it is recognised that for some adsorbates sustained insonation can lead to significant ablation42 Ultrasound was applied to the modified electrode at different intensities for various time periods at a constant horn-to-electrode distance of 10 mm.

105
Fig. 5 shows cyclic voltammograms recorded at different time intervals, with a plot of the peak current versus applied time of insonation.

106
Applying a maximum ultrasound intensity of 87 W cm−2 resulted in practically no ablation; the value of initial peak current before application of ultrasound was 21 μA, while after 10 min of insonation the peak current decreased to 20 μA, corresponding to a 5% decrease.

107
After applying ultrasound for 1 h, a 7% decrease in the signal was observed.

108
Conversely applying ultrasound intensity of 36 W cm−2, the peak current was observed to decrease by 4% after 10 min and 5% after 1 h of continuous insonation from its original value.

109
These results show that the azobenzene-modified electrodes are very stable under ultrasound and therefore potentially suitable for use as sonoelectrocatalysts for hydrogen peroxide generation.

110
Next, the catalytic rate constant was measured via sono-voltammetry.

111
First the surface coverage of azobenzene on the GC electrode was determined via cyclic voltammetry and found to be 57.5 (±0.4) × 10−10 mol cm−2.

112
The modified GC electrode was placed into a face-on arrangement in an oxygen-saturated pH 2 phosphate buffer solution at a fixed horn-to-electrode distance of 10 mm.

113
As shown in Fig. 6, comparison of the response of the modified electrode in the absence and presence of an acoustic field reveals a significant increase in current arising from the acoustic streaming.30

114
The rate constant for oxygen reduction was calculated using the following equation:20where n is the number of electrons transferred, F, is the Faraday constant, A is the electrode area, D is the diffusion coefficient, Cox is the bulk concentration of oxygen, and Γ is the surface coverage of the adsorbate.

115
The rate constant was found to be 7.3 (±0.2) 103 M−1 s−1 if the value of surface coverage before insonation was used (57.5 (±0.4) × 10−10 mol cm−2).

116
If the value of 52.6 (±0.3) × 10−10 mol cm−2 obtained for surface concentration after applying ultrasound was used, the rate constant was calculated to be 8.0 × 103 M−1 s−1.

117
The average value of 7.6 × 103 M−1 s−1 is in reasonable agreement with the values obtained by cyclic voltammetry and RDE analysis.

118
The chemical rate parameter, defined as the catalytic rate constant multiplied by the surface coverage, was calculated for all three methods and found to be 41.9 × 10−3 cm s−1 for cyclic voltammetry measurements, 39.2 × 10−3 cm s−1 for the RDE method and 41.9 × 10−3 cm s−1via sonovoltammetry.

Oxygen reduction at hydrazobenzene (HAB) modified GC electrodes

119
First the voltammetry of hydroazobenzene was explored at an edge plane pyrolytic graphite modified electrode.

120
The cyclic voltammogram was started at −0.3 V (vs. SCE), scanned positive to +0.5 V and reversed back to −0.6 V. This was repeated several times.

121
As shown in Fig. 7 an oxidation wave corresponding to the oxidation of hydroazobenzene to azobenzene is clearly seen at ca. +0.2 V, and on the reverse scan a reduction peak of azobenzene back to hydroazobenzene is observed at −0.3 V which is comparable to that seen for the electrochemical reduction of azobenzene on the edge plane electrode.

122
Next a modified glassy carbon electrode was investigated in a pH 2 buffer solution.

123
Shown in Fig. 8 are the cyclic voltammograms for the first (curve A), second (curve B) and tenth (curve C) scans.

124
On the first scan, two reduction peaks are observed at ca. −0.13 V and −0.4 V (vs. SCE), which again can be assigned to the reduction of azobenzene as two different forms from the electrode surface.

125
A peak is observed on the oxidation scan corresponding to the oxidation of hydroazobenzene to azobenzene.

126
On the second scan the latter reduction peak has virtually disappeared suggesting that hydroazobenzene in the form of crystals is dissolved from the electrode surface with a single reduction wave due to the reduction of azobenzene to hydroazobenzene.

127
The scan rate was varied and a plot of peak current vs. scan rate was found to be linear, consistent with a surface bound species.

128
The surface concentration of hydrazobenzene on a glassy carbon electrode (after stabilisation, ten scans) was determined using cyclic voltammogram, and was found to be 54.7 (±0.4) × 10−10 mol cm−2.

129
Cyclic voltammograms for glassy carbon electrode modified with hydrazobenzene were recorded in various buffer solutions in pH range from 2 to 10.

130
A plot of peak potential against pH was linear with a gradient of 57 mV (pH unit)−1 which is in excellent agreement with expected value for a two-electron, two-proton process.

131
Next, oxygen reduction at a glassy carbon electrode modified with hydrazobenzene was investigated.

132
Buffer solutions were saturated with cyclic voltammograms run.

133
Comparison of oxygen reduction at the modified (Fig. 9, curve B) and bare glassy carbon electrode (Fig. 9, curve C) reveals that the reduction waves shift to a more positive potential with a significantly large increase in the peak current suggesting good electro-activity.

134
The largest enhancement was observed for pH 2.

135
All further measurements with hydroazobenzene were carried out in pH 2 phosphate buffer solution.

136
The catalytic rate constant analysing cyclic voltammetry data via Andrieux and Savéant theory40 was 7.4 × 103 M−1 s−1 for a surface coverage of 54.7 (±0.3) × 10−10 mol cm−2.

137
Kinetic parameters for the reduction of oxygen on a glassy carbon electrode modified with hydrazobenzene were evaluated via the RDE method (eqn. (7)).

138
A modified rotating disc GC electrode was placed into an oxygen saturated solution and run at various rotation speeds.

139
A plot of reciprocal of limiting current versus square root of rotation speed was constructed.

140
Using the Koutecky–Levich equation with a surface coverage of hydroazobenzene, which was found to be 47.7 (±0.4) × 10−10 mol cm−2 the catalytic rate constant of oxygen reduction was determined to be 7.3 × 103 M−1s−1 which is in excellent agreement with that obtained using cyclic voltammetry.

141
Next, oxygen reduction under an acoustic field was evaluated.

142
First however, the stability under ultrasound of the glassy carbon electrode modified with physically adsorbed hydrazobenzene was addressed.

143
The modified electrode was first exposed to ultrasound of 87 W cm−2 at a horn-to-electrode distance of 10 mm.

144
Note that this is the highest intensity used in analysis for the catalytic rate constant via use of the modified Koutecky–Levich equation.

145
Shown in Fig. 10 are the cyclic voltammograms recorded at different time intervals, along with a plot showing the peak current of the hydroazobenzene as a function of time applying an ultrasound intensity of 87 W cm−2.

146
After 10 min of insonation, the initial peak current before application of ultrasound decreased only ca. 10%, from 19 μA to 17 μA.

147
For 1 h of insonation a 16% decrease was observed.

148
Furthermore, 10 min insonation at 36 W cm−2 was found to decrease the voltammetric signal from 23 μA to 21 μA corresponding to a 9% decrease while for 1 h a 10% decrease was observed suggesting that the hydrazobenzene is extremely stable under ultrasound and therefore appropriate as a sonoelectrocatalyst for oxygen reduction and hydrogen peroxide producer.

149
As in the case of azobenzene modified glassy carbon electrode, the catalytic rate constant of oxygen reduction at glassy carbon electrode modified with hydrazobenzene was calculated from the intercept of the sono-Koutecky–Levich plot.20

150
The surface coverage of hydrazobenzene before applying ultrasound was found to be 51.7 (±0.3) × 10−10 mol cm−2, while after application of acoustic field it was found to decrease to 48.9 (±0.4) × 10−10 mol cm−2.

151
Using an average value for the surface coverage, the catalytic rate constant was evaluated to be 8.1 × 103 M−1 s−1.

152
The combined chemical rate constant obtained was 40.7 × 10−3 cm s−1 which is the same as that found from cyclic voltammetry, differing slightly from the one obtained for RDE method (34.8 × 10−3 cm s−1).

Oxygen reduction at fast black K salt (FBK) modified GC electrode

153
The electro-reduction of fast black K salt (2,5-dimethoxy-4-[(4-nitrophenyl)azo] benzenediazonium tetrachlorozincate) was first explored.

154
The structure of fast black K is shown in Scheme 1.

155
As shown in Fig. 11a three reduction waves are observed at +0.15 V, −0.30 V and −0.41 V with an oxidation peak at +0.4 V (vs. SCE).

156
The two reduction peaks at −0.30 and −0.41 V can be attributed to the reduction of the nitro group of the FBK salt as two different forms on the GC electrode, specifically as a monolayer and as microcrystals, while the first at +0.15 V can be assigned to the reduction of the azo linkages of the fast black K compound with the corresponding oxidation wave at +0.4 V.43

157
Fig. 11b shows the second scan which clearly shows two reduction waves at ca. +0.15 V and −0.41 V.

158
The electrochemical behaviour of FBK/GC electrode was examined over a range of potentials at various pHs.

159
The variation of the (first) peak potential as a function of pH was found to be linear with a slope of 51 mV (pH unit)−1 consistent with a two-proton, two-electron process.

160
Next an oxygen-saturated solution was prepared with the cyclic voltammetric response recorded.

161
As shown in Fig. 12, the peak current corresponding to the reduction of oxygen was increased with the peak potential shifting to a more positive potential compared to the unmodified GC electrode.

162
The biggest increase in peak current was observed in pH 2 solution.

163
Therefore subsequent measurements were performed at this pH.

164
A new modified electrode was prepared and found to have a surface coverage of 48.1 (±0.3) × 10−10 mol cm−2.

165
The catalytic rate constant was evaluated using cyclic voltammetry data via the Andrieux and Savéant theory40 which was found to be 10.4 (±0.1) × 103 M−1 s−1.

166
For comparison, oxygen reduction at a FBK modified GC electrode was determined by RDE method.

167
A RDGC electrode was modified as described in section 2.3 and placed into a pH 2 phosphate buffer solution, which was saturated with oxygen.

168
Using the Koutecky – Levich equation,41 the catalytic rate constant for oxygen reduction was calculated to be 9.9 (±0.1) × 103 M−1 s−1 using a surface coverage of 59.0 (±0.3) × 10−10 mol cm−2.

169
This value is identical to the one obtained from cyclic voltammetry.

170
The combined chemical rate parameter was found to be 58.4 × 10−3 cm s−1 for RDE method compared to 50.0 × 10−3 cm s−1 obtained from cyclic voltammetry.

171
Next, the reduction of oxygen using the FBK modified electrode was explored using sonovoltammetry.

172
First the stability under ultrasound of the modified glassy carbon electrode was investigated.

173
The modified electrode was exposed to ultrasound of 87 W cm−2 at fixed horn-to-electrode distance of 10 mm.

174
Shown in Fig. 13 are the cyclic voltammograms recorded at different time intervals, along with a plot showing the peak current of the FBK as a function of time applying an ultrasound intensity of 87 W cm−2.

175
After 10 min of insonation, the initial peak current before application of ultrasound decreased only 13.5%, from 12.6 μA to 10.9 μA with an 1 h of insonation resulting in an 18.2% decrease.

176
When 36 W cm−2 ultrasound intensity was used, 10 min of insonation was found to decrease the voltammetric signal from 9.4 μA to 8.5 μA corresponding to a 9.9% decrease while 1 h insonation produced a 15% decrease.

177
Clearly the fast black K modified electrode is extremely stable under ultrasound and therefore appropriate as a sonoelectrocatalyst.

178
The catalytic rate constant of oxygen reduction at a glassy carbon electrode modified with FBK was calculated via the sono-Koutecky–Levich equation.20

179
The surface coverage of FBK before applying ultrasound was found to be 47.2 × 10−10 mol cm−2, while after application of acoustic field it was found to decrease to 43.7 × 10−10 mol cm−2.

180
Using an average value for the surface coverage, the catalytic rate constant was evaluated to be 10.9 (±0.1) × 103 M−1 s−1.

181
The combined chemical rate parameter obtained was 49.5 × 10−3cm s−1 which is almost the same as that found from cyclic voltammetry, 50.0 × 10−3cm s−1.

Comparison of electro-catalytic activity of the modified electrodes for oxygen reduction

182
Table 1 shows the deduced electrocatalytic rate constant and combined chemical rate parameter values for the oxygen reduction on glassy carbon electrodes modified with azobenzene, hydroazobenzene and fast black K salt.

183
Also included are the values for anthraquinone species previously reported.16

184
Note that all the data refers to pH 2; faster rates can be found at different pH’s.21

185
Comparisons of the combined chemical rate parameters for azobenzene and derivatives with the anthraquinone species, reveal that the former is ca. five times faster.

186
Note that this is the first time these have been investigated for the electrochemical reduction of oxygen.

187
Furthermore the modified electrodes investigated in this work were more stable under ultrasound and exhibited a more remarkable electro-catalytic activity toward oxygen reduction.

Conclusions

188
The electrochemical reduction of oxygen using glassy carbon electrodes modified with azobenzene, hydroazobenzene, or fast black K salt (2,5-dimethoxy-4-[(4-nitrophenyl)azo] benzenediazonium tetrachlorozincate) at glassy carbon electrodes have been explored.

189
The performance of the modified electrodes under ultrasound was found to be highly stable; the superiority of these modified electrodes under insonation over their anthraquinone counterparts19,44 is noted.

190
The catalytic rate constants of the immobilised species towards oxygen reduction were assessed via cyclic, rotating disc and sono- voltammetry with the values obtained suggesting their use as practical hydrogen peroxide generators.