3.1. Deposition by CV
The CV curves of quercetin were compared to those of an equimolar blend of catechol and resorcinol at different potential sweep rates (
Figure 1 and
Figure S1 in the Supplementary Materials). The CVs of the catechol + resorcinol blends were also compared to those of catechol and resorcinol alone (at the same concentration as in the blend) and to the CV calculated by the sum of the CVs of the two used molecules (
Figure 2).
First, it appears that quercetin is characterized by the presence of 3 oxidation peaks, among them a first one (green arrows in
Figure 1A,B) at around +0.2 V vs. Ag/AgCl, which is not observed in catechol, resorcinol, and the CVs obtained from their blends (
Figure 1C,D). However, the two oxidation peaks of quercetin detected at around +0.5–0.6 V and +0.7–0.8 V are close to the oxidation peaks detected in the blend. Those peaks will now be called Peak 1 and Peak 2. Inspection of
Figure 2 allows us to attribute Peak 1 and Peak 2 to catechol and resorcinol, respectively. However, these two peaks in quercetin do not exactly occur at the same potential as in the catechol + resorcinol blend: they are shifted to slightly lower potentials (about 0.1 V, as will be detailed later), suggesting that the oxidation of the catechol and resorcinol moieties in quercetin are facilitated in the fused quercetin structure. The first peak (labeled with a green arrow in the CVs of quercetin may well originate from the oxidation of the OH group in the central B ring of quercetin (
Scheme 1). No further focus will be given to this peak since it rapidly vanished at all the investigated potential sweep rates (
Figure 1A,B and
Figure S1A in the Supplementary Materials). The intensity of Peaks 1 and 2 also decreases with the number of performed CV cycles but differently with quercetin and the catechol + resorcinol blend and in a potential sweep-rate-dependent manner.
The presence of a supplementary oxidation peak in quercetin, as well as the different evolutions of Peak 1 and Peak 2, is sufficient to demonstrate that the electrochemical behavior of catechol and resorcinol fused together in a flavonoid is not simply the sum of the catechol and resorcinol behavior which was more or less expected but not yet demonstrated. Note also that the first CV cycle of the catechol + resorcinol blend is also not exactly the sum of the CVs of catechol and resorcinol, particularly in the potential window between Peak 1 and Peak 2, where the theoretical CV obtained from the summation of the CV of the two selected molecules (red line in
Figure 2) yields to a higher current than the experimental one (black full line in
Figure 2). This means that when catechol starts to be oxidized and to undergo non electrochemical crosslinking at the electrode surface, it hinders the further deposition of resorcinol occurring at higher potentials. Note that the electrodeposition of catechol-based films [
11,
12] and resorcinol-based films [
13] on amorphous carbon electrodes has been investigated previously.
When considering the evolution of the positions of Peak 1 and Peak 2 in quercetin and in the catechol + resorcinol blend, it appears visually that those peaks are anodically shifted upon faster potential sweep rates as expected for irreversible electrochemical processes (i.e., processes where the reduction current is different from the oxidation current due to the non-electrochemical transformation of the oxidation products) [
18]. Note, however, that even if Peak 2 (affiliated to the oxidation of resorcinol) is of irreversible nature, this is not exactly true for Peak 1 (affiliated to the oxidation of catechol), which appears partially reversible, the more so, the higher the potential sweep rate (compare
Figure 1C with
Figure 1D and with
Figure S1B in the Supplementary Materials). When plotting the oxidation peak potentials during the first CV cycle as a function of the logarithm of the potential sweep rate (
v), the value of the electron transfer coefficients,
a, are obtained according to [
18]:
where
T is the absolute temperature, and
F is the Faraday constant.
If Equation (1) is applicable, the plots of the two peak potentials against lnv should yield a straight line. This is indeed the case (
Figure 3) with fairly good linear regression coefficients. The slope of this straight line is inversely proportional to the electron transfer coefficient and the obtained values are gathered in
Table 1. The electron transfer coefficients attributed to the two peaks of quercetin are higher than those attributed to the two peaks in the catechol + resorcinol blend, suggesting that the central B ring of quercetin promotes the electron transfer of the catechol and resorcinol moiety to the amorphous carbon electrode.
The electron transfer coefficients obtained herein are higher than those reported by Nady et al. [
19] for phenol, resorcinol and pyrogallol (between 0.146 and 0.431), suggesting a slower electron transfer than in these three systems. However, the experiments were performed in different conditions, such as a Pt working electrode at pH = 3.0.
According to
Figure 1, the oxidation peaks and the oxidation peak currents undergo some shifts and some changes in intensity from cycle to cycle and in a potential sweep-rate-dependent manner. The data taken from
Figure 1 are given in
Figure 4 for the oxidation peak potentials and in
Figure 5 for the oxidation peak currents. Quercetin is distinguished from the catechol + resorcinol blend in the sense that the two oxidation peak potentials only moderately increase from cycle to cycle (by about 20 mV between the first and the tenth cycles) when the CVs are performed at 10 and 100 mV
−1. Surprisingly, at 1000 mV·s
−1, the positions of Peaks 1 and 2 decrease upon an increase in the number of CV cycles, suggesting a facilitated oxidation process (
Figure 4A,B). However, the peak positions of the catechol + resorcinol blend increase by about 300 mV for peak 1 for CVs performed at 10 mV·s
−1 and by about 150–200 mV for CVs performed at 100 and 1000 mV·s
−1 (
Figure 4C). From
Figure 4D, it can also be seen that the peak position of resorcinol (peak 2) saturates at 1.2 V for all the CV cycles. This comes from the fact that the potential was scanned only up to +1.2 V vs. Ag/AgCl because this corresponds to the limit of the electrochemical stability window of water. The resorcinol peak also totally disappears (as also seen in
Figure 1C) after the first CV cycle performed at 10 mV·s
−1 in agreement with previous findings [
13].
Concerning the peak currents, a potential sweep rate dependent decrease is observed (
Figure 5) but is particularly marked in the case of the catechol + resorcinol blend when the CVs are performed at 10 mV·s
−1 (
Figure 5C,D) as was already apparent from
Figure 1C. This current decrease is due to electrode passivation by a deposited film. In this case, the coatings should progressively become impermeable to a redox probe like potassium hexacyanoferrate.
3.2. Permeability and Film Properties of the Quercetin Based Films
The coatings made from resorcinol and catechol alone have been investigated for their permeability to potassium hexacyanoferrate in previous investigations: the coatings made from the electrodeposition of catechol become impermeable to this negatively charged redox probe after a few CV cycles, the faster this occurs, the slower the potential sweep rate during the CV cycles is [
11]. However, the coating made from resorcinol becomes conformal and impermeable after a very small number of CV cycles, namely from 1 to 3 cycles [
13], reflecting the efficient film-forming properties of this molecule. The blends behave similarly to resorcinol (
Figure S2 in the Supplementary Materials), whereas quercetin-based coatings become impermeable to the redox probe (after 10 CV cycles) only at low potential sweep rates of 10 mV·s
−1 (
Figure 6A) some important residual permeability remaining when the films are deposited at higher potential sweep rates (
Figure 6B).
Hence, the film-forming ability of quercetin is efficient in the sense of getting impermeable films when the deposition by CV is performed at sufficiently low potential sweep rates. In this sense, quercetin behaves as other polyphenols, like catechol, for which the film-forming ability is the combination of electrode-triggered oxidation followed by subsequent non-electrochemical reactions of adsorbed and oxidized species, probably a kind of radical crosslinking as suggested from electrodeposition experiments performed at different pH values [
11].
Since the film-forming ability of quercetin on amorphous carbon is more efficient at 10 mV·s
−1 than at higher potential sweep rates, we also performed CV deposition on gold working electrodes in these conditions (
Figure S3 in the Supplementary Materials) in order to characterize the film morphology by means of AFM (
Figure 7). When compared to the pristine gold electrode, the quercetin-based deposit (10 CV cycles at 10 mV·s
−1) appears conformal and grainy with a root mean squared roughness of 3.3 nm for an average film thickness of 9.5 ± 0.5 nm as obtained by scanning the AFM tip through a needle-induced scratch on the film. The AFM surface topographies of catechol and resorcinol-based films deposited under the same conditions (10 CV cycles performed at 10 mV·s
−1) on gold electrodes are displayed in
Figure S4 in the Supplementary Materials. They are much rougher than the quercetin-based film, with RMS roughness values of 31 and 14 nm for the catechol and resorcinol-based coatings. In addition, the catechol-based coating contains large grains or domains in the µm size domain in addition to smaller particles, whereas the resorcinol-based film is more homogeneous, displaying only particles in the 100 nm range or less, as on the quercetin-based coating. Hence, the surface morphology of the quercetin-based electrodeposited films seems to be similar to that of the resorcinol-based films. Note, however, that the AFM topographies displayed in
Figure 7 and in
Figure S4 of the Supplementary Materials were not acquired in the same size range: 10 µm × 10 µm for quercetin film and 4 µm × 4 µm for the films made from the catechol and resorcinol building blocks. This should, however, not change the global interpretation proposed here.
The solubility of the quercetin-based films deposited on gold electrodes was investigated by immerging them for 1 h in organic solvents of various polarity, namely DMSO, ethanol, dichloromethane, N,N-dimethylformamide. No apparent solubilization was observed in all cases because the gold-coated electrode kept its brownish appearance.
In addition, the film deposited after 10 CV cycles on a 2 × 2 cm2 gold electrode (10 mV·s−1) was immersed in 1 mL of a DPPH solution at 10−4 mol·L−1. After 40 min of reaction, the absorbance of the DPPH solution was measured against the original DPPH solution (present in the reference cuvette of the spectrophotometer) and was found to be decreased by 0.18 absorbance units. This value was used to calculate the equivalent concentration of quercetin, leading to the same quenching of DPPH. To that aim, a calibration curve of DPPH quenching by ethanol-dissolved quercetin was used (A = −9 × 10−4 − 65.7C, r2 = 0.995, 6 points in the calibration curve, C being the quercetin concentration in mg·mL−1) and yielded an equivalent antioxidant concentration of 2.8 × 10−3 mg·mL−1 or 9.3 × 10−6 mol·L−1 on the electrodeposited quercetin film.