M.M. Hossain, L. Aldous / Electrochemistry Communications 69 (2016) 32–35
33
Polyoxometalates (POMs) are polyatomic ions which contain transi-
tion metals joined into a 3D framework by multiple oxygen atoms [19].
POMs are widely available with different dimensions, charge density,
shape, reactivity and multiple redox features [19]. POMs have been
utilised as stoichiometric chemical oxidants for different organic com-
pounds, including phenolic compounds [19]. POMs have also been
utilised as electrocatalytic mediators [20], although not in conjunction
with phenols.
In this study, we have demonstrated that the conventional electrode-
fouling voltammetry for phenolic compounds can alter to ideal voltamm-
etry upon addition of a suitable POM. Electrode fouling is suppressed and
exhaustive bulk electrolysis can even be performed, allowing isolation of
the solution-phase (electro)oxidised phenol product.
second and successive scans, the oxidation peak decreased dramatically,
and no corresponding reduction features were observed. This is
indicative of electrode fouling by the oxidised vanillin, due to the one
electron formation of a cationic phenoxyl radical [7,25]. Enache and
Oliveira-Brett have shown that the peak current for the oxidation of
phenol in aqueous H2SO4 does not scale with concentration due to elec-
trode fouling [10]; the Randles–Sevcik equation for the system in Fig. 1
also predicts a peak current value ca. 10-fold higher (vide infra). This
indicates significant passivation early within the first scan. Assuming
vanillin is a sphere with ca. 1 nm diameter, a densely packed monolayer
would consist of ca. 1.3 × 1014 molecules cm−2. In the 40 mM vanillin
solution, this quantity of vanillin is present within the first ca. 50 μm
directly adjacent to the electrode surface; diffusion layers larger than
100 μm are expected for CVs at 100 mV s−1 in aqueous electrolytes.
The voltammetry of PTA was also investigated (Fig. 1(a)). Scanning
oxidatively from the open circuit potential, no Faradaic features were
observed, indicating that the POM was already present in its most
2. Experimental
Milli-Q water (18.2 MΩ cm−1) was used to prepare all the
aqueous solutions. Vanillin, para-cresol, dopamine hydrochloride
and 3,4-dihydroxybenzaldehyde (all from Sigma-Aldrich), guaiacol
(Chem Supply) and 2,4-dihydroxybenzaldehyde (Lancaster) were
used as received.
The POMs STA (H4[SiW12O40)], Sigma Aldrich), PMA (H3[PMo12O40],
Sigma-Aldrich) and PTA (H3[PW12O40], Fluka) were purchased
and used as received. The POMs PVM (H5[PV2Mo10O40]) [21],
PVW (K4[PVW11O40]) [22], SiVW (K5[SiVW11O40]) [23] and BVW
(K7[BVW11O40]) [24] were prepared according to the reported
literature methods.
All cyclic voltammetry (CV) experiments were performed using a
glassy carbon working electrode (GC, 3 mm diameter), Pt counter and
Ag|AgCl reference electrode (all BASi Analytical, USA) using an Autolab
PGSTAT101 (Ecochemie, the Netherlands). A 100 mV s−1 scan rate was
used for all the experiments. Bulk electrolysis experiments used a GC
bulk electrode (SIGRADUR G, HTW, Germany), using a separate elec-
trode compartment for the counter electrode (BASi Analytical, USA).
Electrochemical simulation was performed using DigiElch7 software.
oxidised state ([PW12O40]
3−). On the reverse scan, two distinct
reversible one-electron processes were observed (Fig. 1(a)), followed
by a two-electron proton-coupled reduction and then multiple electron
reductions (not shown, but consistent with the previously reported
voltammetry for PTA) [20]. In 1 M H2SO4 PTA is fully dissociated, and
the first two reductions do not include protonation [20].
The combination of 40 mM vanillin and 10 mM PTA in the same
solution resulted in a significant change in the vanillin voltammetry.
As shown in Fig. 1(c), an oxidative peak with ca. 10-fold higher peak
current was observed for vanillin, and no evidence of any electrode
passivation, even after multiple cycles. The oxidation of vanillin
became chemically reversible, with an associated reduction feature at
ca. +0.45 V. Lower concentrations of PTA resulted in lower current
and some electrode fouling; more than 10 mM PTA did not further in-
crease the peak current, demonstrating that the interaction between
PTA and vanillin was catalytic (with respect to PTA) rather than stoi-
chiometric. POMs can physisorb at GC surfaces from H2SO4 [20], but
physisorption of PTA onto the GC electrode before transferring into a
vanillin solution demonstrated vanillin voltammetry consistent with
the absence of PTA. This demonstrates that the process is likely
solution-based, rather than a layer of PTA at the GC acting as either
mediator or physical anti-fouling layer.
3. Result and discussion
As our results are exemplified by the case of vanillin (phenolic
Homogeneous two electron oxidation of vanillin will result in a
carbocation via the one-electron phenoxyl radical intermediate
[26]. The ortho-methoxy group can subsequently hydrolyse to yield
1,2-benzoquinone-4-carbaldehyde [26]. A similar process has already
been established for the electro-oxidation of similar ortho-methoxy
phenolic species, such as capsaicin oxidation [27,28]. Therefore the
well-defined voltammetry in Fig. 1(b) was digitally simulated,
compound) and phosphotungstic acid (PTA, H3[PW12O40],
a
polyoxometalate), this system will be discussed in detail. A wider
range of phenolics and polyoxometalates are briefly described at the
end.
Fig. 1(a) displays a cyclic voltammogram (CV) of vanillin at a glassy
carbon (GC) electrode. A broad peak is present at ca. +0.95 V (vs
Ag|AgCl) on the first scan, due to the oxidation of vanillin. For the
Fig. 1. Cyclic voltammograms in 1 M aqueous H2SO4 for (a) 40 mM vanillin for successive 5 scans (
) or 10 mM PTA (−); (b) 40 mM vanillin without (
) and with (
) 10 mM
PTA; overlaid ( ) is simulation for an EE mechanism; (c) 40 mM vanillin and 10 mM PTA before bulk electrolysis and after passing one e− and two e− per vanillin.