- Email: [email protected]
Spectroelectrochemistry of salicylaldehyde oxidation
Electrochimica Acta 125 (2014) 133–140
Contents lists available at ScienceDirect
Electrochimica Acta journal homepage: www.elsevier.com/locate/electacta
Spectroelectrochemistry of salicylaldehyde oxidation Yan Wang, Huan Jiang, Jing-Jing Tian, Jian-Bo He ∗ Anhui Key Lab of Controllable Chemical Reaction & Material Chemical Engineering, School of Chemical Engineering, Hefei University of Technology, Hefei 230009, China
a r t i c l e
i n f o
Article history: Received 26 August 2013 Received in revised form 9 November 2013 Accepted 13 January 2014 Available online 28 January 2014 Keywords: Thin-layer electrolysis Spectroelectrochemistry Cyclic voltabsorptometry Oxidation mechanism Salicylaldehyde
a b s t r a c t The advances in knowledge of the oxidation mechanism of salicylaldehyde are important in understanding its role and conversion in the involved oxidative degradation, synthesis and aerobic metabolism processes. The electrochemical oxidation of salicylaldehyde was investigated in different pH media using cyclic voltammetry, in situ UV–vis spectroscopy and cyclic voltabsorptometry based on a long opticalpath thin-layer electrochemical cell. ATR-FTIR spectroscopy was used for characterization of the oxidation products deposited on the electrode surfaces. Time-derivative cyclic voltabsorptograms were obtained at the characteristic wavelengths of salicylaldehyde and the soluble oxidized salicylaldehyde, for comparative discussion with the corresponding cyclic voltammograms. Two couples of redox peaks, subsequent to the main oxidation peak of salicylaldehyde, were observed on the voltabsorptograms but nearly indistinguishable on the voltammograms. Salicylaldehyde was initially oxidized to reactive phenoxyl radicals, followed by a series of transformation steps leading to different final products. A parallel-consecutive reaction mechanism was proposed for the pH-dependent formation of a deposited polyester and two trace amounts of soluble quinoid products. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Salicylaldehyde (o-hydroxybenzaldehyde) is a common highlyfunctionalized arene that has often been used as analysis reagent, spices and gasoline additive. It is also a key precursor in organic synthesis of a variety of chelating agents and other chemicals. Because of its widespread application, salicylaldehyde becomes an organic pollutant that finds its way into the water system through pharmaceutical medications, cosmetic products, and agricultural chemicals [1]. Also, salicylaldehyde naturally occurs in grape, tomato, cinnamon, milk and milk products, beer, coffee and tea. It has been found to be an intermediate formed during the aerobic metabolism of some polycyclic aromatic hydrocarbons [2–4]. Salicylic acid is one of the possible oxidation products of salicylaldehyde. Electrochemical oxidation of salicylic acid [1,5–12] and acetylsalicylic acid [13–15] has been widely investigated focusing on the electrochemical removal and degradation [1,5–7], electrochemical detection [8–11,13,14], electro-polymerization [7,12] and electro-oxidation mechanism [9,15]. Salicylic acid and its derivatives have also been used as an electrode modifier for the determination of trace copper(II) in water [16], the amperometric
∗ Corresponding author. Tel.: +86 551 6290 4653; fax: +86 551 6290 1450. E-mail address: [email protected] (J.-B. He). 0013-4686/$ – see front matter © 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.electacta.2014.01.087
nonenzymatic determination of glucose free of interference [17], and for the visible light photoelectrocatalytic degradation of pnitrophenol [18]. However, only a few reports are available on electrochemical oxidation of salicylaldehyde [1,5]. Chen et al. studied the kinetics of electrochemical oxidation of salicylic acid and salicylaldehyde in 0.5 M H2 SO4 at a Ti/IrO2 -SnO2 -Sb2 O5 electrode [1] and of their photoelectrochemical oxidation on titanium dioxide nanotube arrays [5] using in situ UV-vis spectroscopy. They revealed the impact of temperature, initial concentration, current density and supporting electrolyte etc. on the degradation kinetics. The electro-oxidation pathway of salicylaldehyde is currently unclear and deserves further study for better understanding its role and conversion in the involved oxidative degradation, synthesis and aerobic metabolism processes. A combination of cyclic voltammetry (CV) and cyclic voltabsorptometry (CVA) allows simultaneously measuring the potential-dependent current and absorbance [19,20]. The light-absorbing reactant, intermediates and products involved in the electrochemical processes can be followed photometrically by setting the corresponding characteristic wavelengths. The time-derivative cyclic voltabsorptometry (DCVA) presents the data curves in a shape resembling that of the corresponding CV curves, providing direct comparison between the CV and DCVA peaks. Unlike the CV peaks, the DCVA peaks may be caused not only by electrochemical reaction but also by chemical conversion. Moreover, no charging/discharging background signals are
134
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
recorded on the DCVA curves, which may improve the response to the electrochemical and chemical processes. The CVA has generally been used along with an optically transparent electrode, in a light incident mode perpendicular to the electrode surface. In the present work, the CVA technique was used with a long path-length thin-layer electrochemical cell to investigate the electrochemical oxidation of salicylaldehyde in acidic, physiological and alkaline pH media. This thin-layer cell, in a light incident mode parallel to and close to the electrode surface, allows monitoring soluble intermediates and products generating from an opaque electrode. The use of long path-length greatly improves the sensitivity of detection of the analytes present in the thinlayer solutions. Enhanced responses are therefore obtained to the potential- and pH-dependent formation and reduction of the soluble products. 2. Experimental
Fig. 1. CVs of 100 M salicylaldehyde in different pH buffers. pH (1→7): 1.8, 3.2, 5.8, 7.4, 8.0, 9.2 and 11.9. Scan rate 100 mV s−1 . Inset: pH dependence of the anodic peak potential.
2.1. Chemicals and solutions 2.3. Procedures Reagent grade salicylaldehyde (98%+ pure) from Guangfu (Tianjin, China) was used as received. Spectrograde graphite powder (320 mesh) and spectrograde paraffin wax (solidification point 62–65 ◦ C) from Shanghai Chemical Works were used for preparing the graphite paste electrode. Graphite sheet was from Aidahengsheng (Tianjin, China). Other chemicals were of analytical grade from China-Reagent group. All solutions were prepared with doubly-distilled water from an all-glass distillation apparatus. Salicylaldehyde stock solution was prepared with ethanol and water in a volume ratio of 1:4 by means of ultrasonic agitation, and then kept at 4 ◦ C in a refrigerator. The supporting electrolytes with various pH values were a mixture of 0.2 M Britton-Robinson (BR) buffered solution and 0.5 M KCl. High pure N2 was used for solution deaeration. 2.2. Apparatus, electrodes and cells Cyclic voltammetry and spectroelectrochemistry were carried out on a CHI660 C electrochemical workstation (Chenhua, Shanghai, China). UV-vis spectroscopic and photometric measurements were carried out on an UV-2500 spectrophotometer (Shimadzu, Japan) to monitor the soluble reactants and products under potentiostatic and potentiodynamic conditions. Attenuated total reflectance (ATR)-FTIR spectra were recorded on a Vertex 80 infrared spectrometer coupled to a Hyperion 2000 IR microscope (Bruker Optics). The three-electrode system was composed of a graphite working electrode, a KCl-saturated Ag/AgCl reference electrode (0.195 V vs. SHE, self-made) and a platinum wire counter electrode. The graphite working electrode used for the conventional voltammetric measurements was a disk solid graphite paste electrode (GPE) with a smaller geometrical area of 0.049 cm2 , whereas the electrode for the thin-layer electrochemical experiments was a quadrate graphite sheet electrode with a larger area of 0.77 cm2 . The graphite paste electrode was selected because of its advantages of low background currents, low noise and fast base line stabilization, and its preparation was described previously [21,22]. A conventional single-compartment cell was used for the voltammetric measurements. A thin-layer spectroelectrochemical cell was self-made, using a standard quartz photometric cell with 10 mm optical path length as the cell body. The schematic view of the thin-layer cell can be found in literature [23]. The incident light beam parallels to the working electrode and goes through the thin-layer electrolyte solution (0.2 mm thick) on the electrode surface.
Salicylaldehyde solutions were bubbled with high pure N2 for about 15 min to remove dissolved oxygen before being put into the electrochemical cells. The working electrode was polished carefully with 800 grit emery paper. Before each run, the working electrode was cleaned and activated by repetitive cyclic scans between 0.0 and 1.6 V in 0.1 M NaHCO3 solution until only the background current remained. Considering the adsorption of salicylaldehyde on graphite electrode, a pre-accumulation step was always performed in open circuit for 60 s. All the experiments were carried out at room temperature (22 ± 1 ◦ C). The UV–vis absorption spectra were recorded while the thinlayer solution was electrolyzed at a constant potential. Blank BR buffers were used for the spectral baseline correction. Cyclic voltabsorptometry was performed at a certain wavelength to follow the absorbance changes of species in the thin layer solution. The ATRFTIR spectra were recorded ex situ for characterizing oxidation products deposited on the electrode surfaces. The spectra were collected with 64 scans and at a resolution of 4 cm−1 . 3. Results and discussion 3.1. Cyclic voltammetry Cyclic voltammetric behavior of salicylaldehyde was examined in a wide pH range of 1.8–11.9. The cyclic voltammograms recorded using the GPE shows one anodic peak (Fig. 1), corresponding to the oxidation of salicylaldehyde. The anodic peak current density (jp,A ) first decreased and then increased with increasing pH, showing a minimum peak current at pH 8.0. The anodic peak potential (Ep,A ) shifted negatively with increasing pH until reaching pH 9.2, following a linear relationship Ep,A /V = 1.3008–0.0571 pH (R = –0.99837) (the inset of Fig. 1). The equation slope of –57 mV per pH unit supports a reaction mechanism involving an equal number of electrons and protons. At the pH values greater than 9.2, the peak potential did not depend on the pH of media, suggesting the absence of proton transfer in these pH conditions. The pKa values of salicylaldehyde, phenol and benzaldehyde are about 8.37, 9.99 and 14.90 at 25 ◦ C [24]. Accordingly, the initial electro-oxidation should occur at the phenolic group of salicylaldehyde, through one-electron one-proton transfer for the undissociated molecules but only one-electron transfer for the deprotonated phenolate anions. The corresponding cathodic peaks were extremely small (Fig. 1), indicating the poor reversibility of salicylaldehyde oxidation. The
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
species reducible at the cathodic peaks were not the principal products arising from the oxidation of salicylaldehyde. The amount of the reducible species further decreased as the pH of the media increased, as reflected by the decrease in the cathodic peak current with increasing pH. A variable scan rate (v) experiment was performed over a scan rate range of 1–1000 mV s−1 , and a linear relationship of log(jp,A ) versus log(v) (data not shown) was observed for 100 M salicylaldehyde, with slopes of 0.42, 0.40 and 0.40 at pH 1.8, 7.4 and 11.9, respectively. The slopes near 0.5 indicate a diffusion-controlled [25] oxidation process of salicylaldehyde.
3.2. In situ thin-layer UV–vis spectroscopy Fig. 2 shows the time-dependent spectral absorbance of the thin-layer electrolyte taken at 62 s intervals before and during the controlled potential oxidation of salicylaldehyde in different pH BR buffers. In the pH range of 1.8 − 7.4, salicylaldehyde exhibits three absorption bands locating around 216 (band I), 256 (band II) and 326 nm (band III) before electrolysis (the first curves in Fig. 2A−D), the former two corresponding to the →* transition of the benzene ring and the latter to the weak n→* transition of the C = O group. A new absorption band arose around 380 nm (band III , see the first curve in Fig. 2E) at pH 8.0, a value close to the pKa of salicylaldehyde (≈ 8.37 [24]). This band showed an increase in intensity as the pH increased to 9.2 (Fig. 2F), at the expense of the disappearance of the band III around 326 nm. Accordingly, the bands III and III can be assigned to the undissociated molecule and deprotonated phenolate anion of salicylaldehyde, respectively. When the hydroxyl group is not deprotonated, a resonance structure can be stabilized through an intramolecular hydrogen bond between the hydroxyl hydrogen and the carbonyl oxygen. The electrochemical oxidation of salicylaldehyde was carried out at the anodic peak potential, which is dependent on the buffer pH. At all the pH values tested, the oxidation of salicylaldehyde resulted in a large decrease in intensity of the bands I, II, III and III , especially at the alkaline pH values (Fig. 2). Five isosbestic points are observed at the pH 1.8 and 3.2, while less isosbestic points are present at the higher pH values. A new absorption band with a low intensity appeared around 408 nm (band IV) attributed to a soluble oxidation product of salicylaldehyde. This emerging band decreased in maximum intensity with increasing pH, and became vanish at pH 8.0 and 9.2. This indicates that the soluble product with a characteristic wavelength of about 408 nm can not be obtained in the media at the highest pHs. After the exhaustive electrolysis, the low absorption intensity over the whole tested wavelength range suggests that only very small amounts of soluble light-absorbing species were present in the thin-layer. Therefore, the principal oxidation products of salicylaldehyde should be deposited on the electrode surface, especially at the higher pH values. This was also evidenced by a cyclic voltammetry experiment in a common-volume cell, which showed a fast decrease of current in the following cycles (1 mM salicylaldehyde, scan rate 20 mV s−1 , data not shown), especially at the pH not less than 7.4. The resulting film on the electrode surface is non-conductive, which obstructed further oxidation of salicylaldehyde from the solution in the subsequent cycles. One possible oxidation product of salicylaldehyde is salicylic acid, which has an absorption band centered at about 300 nm [1–4,6]. Fig. 2 shows the absence of absorption around this wavelength after electrolysis, thereby excluding the formation of salicylic acid. The band IV emerging around 408 nm supports the formation of an o-quinone structure [26–28] in the thin-layer solution.
135
3.3. Thin-layer cyclic voltabsorptometry 3.3.1. pH 1.8 Multi-cycle cyclic voltabsorptometry was performed in the thin-layer cell at the characteristic absorption wavelengths of salicylaldehyde and a product (Fig. 3). A slow scan rate (2.0 mV s−1 ) was used, so that the diffusion layer thickness was larger than the thin layer thickness. Fig. 3A and C show two CVA curves recorded at 256 and 408 nm, respectively, for the pH 1.8, 100 M salicylaldehyde. During the cyclic scan, the variation of absorbance at 256 and 408 nm (A256 and A408 ) with potential reflects the changes in concentrations of salicylaldehyde and the o-quinone product, respectively. The first scan in the positive direction caused a slow decrease in A256 until 0.98 V due to the potential-driven adsorption of salicylaldehyde, followed by a sharp decline in A256 due to the anodic oxidation of salicylaldehyde (Fig. 3A). During the negative scan, only very small change in A256 was observed, indicating the irreversibility of salicylaldehyde oxidation. Meanwhile, the fast rise in A408 during the positive scan corresponds to the formation of an o-quinone structure (Fig. 3C), and the decline in A408 during the negative scan is consistent with the reducibility of o-quinone. The derivative absorbance signal (dA /dt) is proportional to the reaction rate of the monitored species. A well-shaped DCVA peak A1 appeared on both the DCVA curves, corresponding to the principal oxidation peak on the thin-layer CV curve (also marked as A1, see the inset of Fig. 3B). The DCVA peak A1 obtained at the wavelengths 256 and 408 nm was centered at 1.08 and 1.14 V, respectively, with a potential (or time) difference due to the diffusion lag between the reactant and the generated product in the thin-layer. The counterpart of peak A1 marked as peak C1 was seen at 408 nm (Fig. 3D) but not at 256 nm (Fig. 3B), which can be attributed to chemical conversion of an intermediate with absorption at 408 nm (possibly a semiquinone radical) to a more stable product without or with less strong absorption at this wavelength, but not to the parent salicylaldehyde. This final product may be an insoluble species deposited on the electrode as discussed above, without causing light-absorption in the thin-layer. On the other hand, the corresponding cathodic voltammetric peak C1 in the inset of Fig. 3B was extremely small, which should be associated with a reversed reduction of a highly unstable intermediate formed in initial oxidation of salicylaldehyde, e.g. phenoxy radicals, which are key intermediates in the anodic oxidation of phenols. Interestingly, two subsequent pairs of DCVA peaks, C2/A2 and C3/A3, were definitely detected following the first couple A1/C1 (Fig. 3B and D). In contrast, the corresponding voltammetric pairs in the inset of Fig. 3B were very small and nearly indistinguishable. Both the peak pairs on the DCVA curves showed a peak-topeak separation of about 60 mV, indicating two quasi-reversible redox couples, most likely o-quinone/o-diphenol and p-quinone/pdiphenol, present in the thin-layer aqueous-alcoholic solution. A similar pair of redox peaks following a main anodic peak has also been observed in the second cycle of cyclic voltammetric scans of salicylic acid at multiwalled carbon nanotube electrode [9] and graphite-epoxy-composed solid electrode [10] as well as of phenol on polyethylenedioxythiophene deposited stainless steel substrate [29], all of which were attributed to the quasi-reversible behavior of benzoquinone/hydroquinone. The peak pair C2/A2 was obtained at 408 nm, with a midpoint potential of 0.795 V (vs. Ag/AgCl/KClsat ) at pH 1.8. The 408 nm band has been assigned to an o-quinone structure. o-Quinone/catechol redox couple without a–CHO substituent has a redox potential of 0.494 V (vs. Ag/AgCl/KClsat ) at pH 1.8 (derived from the standard redox potential of 0.795 V vs. SHE [30]), lower by about 0.3 V than the midpoint potential of the peak pair C2/A2. This potential difference is due to the presence of formyl group on the carbon ring as a
136
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
Fig. 2. In situ thin layer UV–vis spectra of 100 M salicylaldehyde in different pH buffers during electro-oxidation. (A) pH = 1.8, E = 1.12 V; (B) pH = 3.2, E = 1.06 V; (C) pH = 5.8, E = 0.91 V; (D) pH = 7.4, E = 0.82 V; (E) pH = 8.0, E = 0.81 V; (F) pH = 9.2, E = 0.75 V. Spectral tracing was repeated every 62 s after the potential was applied. The first curve in every panel was recorded before electrolysis.
strong electron-withdrawing substituent, which makes the reduction easier but more difficult to be oxidized. This explanation is supported by the fact that, at the same pH, the oxidation potential of phenol [29] was lower by about 0.25 V than that of salicylaldehyde (Fig. 1). At pH 1.8, 3-formylcatechol has not absorption around 408 nm [31], thus the reduction of 3-formyl-o-benzoquinone to 3formylcatechol led to a decrease in A408 and then the appearance of DCVA peak C2. Inversely, the oxidation of 3-formylcatechol to 3formyl-o-benzoquinone resulted in the peak A2 in the next positive scan. The pair C3/A3 was observed at 256 nm, with a midpoint potential of 0.40 V, which fits well with the redox potential of pbenzoquinone/hydroquinone couple, 0.398 V (vs. Ag/AgCl/KClsat ) at pH 1.8 (derived from Ref. [32]). The electron-withdrawing effect of–CHO on the redox potential was not observed, which suggests that the–CHO group might be removed from the pquinone ring during the oxidation. At 256 nm, p-benzoquinone has an absorption much stronger than hydroquinone [33],
therefore the reduction of p-quinone to hydroquinone led to the decrease in A256 and then the appearance of DCVA peak C3. The reverse oxidation resulted in the peak A3 in the next positive scan. 3.3.2. pH 7.4 and 9.2 Fig. 4 shows the CVA and DCVA curves of salicylaldehyde at pH 7.4, recorded at the same two wavelengths as the previous study. A significant difference with the results obtained at pH 1.8 was that the peaks C3 and A3 did not appear on the 256 nm DCVA curve (Fig. 4B), indicating that the p-benzoquinone was not formed at pH 7.4 during the oxidation of salicylaldehyde. Another important feature is that the absorbance at 408 nm started decreasing even before the scan was reversed (Fig. 4C), thus the DCVA peak C1 occurred with a peak potential more positive than that of the anodic peak A1 (Fig. 4D). This means that the conversion of an intermediate at the DCVA peak C1 was not an electrochemical reduction process and occurred more
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
137
Fig. 3. Multi-cycle thin-layer CVA (A, C) and DCVA (B, D) curves of 100 M salicylaldehyde in pH 1.8 buffer. Scan rate 2.0 mV s−1 , cycle number 3, wavelength 256 (A, B) and 408 nm (C, D). Inset in panel (B): the corresponding thin-layer CV curve.
Fig. 4. Multi-cycle thin-layer CVA (A, C) and DCVA (B, D) curves of 100 M salicylaldehyde in pH 7.4 buffer. Scan rate 2.0 mV s−1 , cycle number 3, wavelength 256 (A, B) and 408 nm (C, D). Inset in panel (B): the corresponding thin-layer CV curve.
138
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
Fig. 5. Multi-cycle thin-layer CVA (A) and DCVA (B) curves of 100 M salicylaldehyde in pH 11.9 buffer. Scan rate 2.0 mV s−1 , wavelength 227 nm, cycle number 3. Inset in panel (B): the corresponding thin-layer CV.
easily than at pH 1.8. The DCVA peaks C2 and A2, attributed to the 3-formyl-o-benzoquinone/3-formylcatechol redox couple, were still observed on the 408 nm DCVA curve (Fig. 4D). This peak couple showed a midpoint potential of 0.480 V at pH 7.4, lower than the value at pH 1.8, with a slope of −56 mV per pH unit which supports the two-electron two-proton quinone/catechol redox reaction. At the characteristic wavelength of salicylaldehyde, similar CVA and DCVA curves were obtained at pH 11.9 (Fig. 5). However, no changes in absorbance were recorded at 408 nm (data not shown). The chemical conversion of the intermediate was too fast to form the soluble o-quinone product at such a high pH value.
3.4. ATR-FTIR spectroscopy The solid oxidation products of salicylaldehyde deposited on graphite sheet electrode were characterized by ATR-FTIR (Fig. 6a). The absorptions in the range of 1600 − 1454 cm−1 correspond to the aromatic C = C ring stretching vibrations. The two absorptions at 2920 and 2850 cm−1 are known to be the characteristic bands of CH2 groups in asymmetric and symmetric stretching modes, respectively. The CH2 groups may come from ethanol and/or acetic acid in the background electrolyte solution. The phenolic hydroxyl group of salicylaldehyde was no longer present in the deposits, as indicated by the vanishing of the broad O-H stretching band
Scheme 1. A parallel-consecutive reaction mechanism proposed for the electro-oxidation of salicylaldehyde in the pH range of 1.8–11.9.
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
139
latter two quinoid products show reversible redox behaviors, which can be detected at two corresponding wavelengths by thin-layer cyclic voltabsorptometry. This work provides an enhanced observation of trace amounts of redox couples by cyclic voltabsorptometry, and hence enriches the knowledge about the oxidation pathway of salicylaldehyde.
Acknowledgment The authors gratefully acknowledge the financial support from the National Nature Science Foundation of China (No. 51203040, 20972038).
References Fig. 6. (a) ATR-FTIR spectrum for the oxidized species on a graphite electrode after electrolysis at 0.81 V in pH 8.0 1.0 mM salicylaldehyde. (b) FTIR of salicylaldehyde in KBr pellet shown for comparison. The absorption intensities of spectrum (a) were 100-times magnified for clearer presentation.
between 3500 and 3000 cm−1 and the strong C-O stretching band at about 1276 cm−1 (compare Fig. 6a with b). Possible formation of aryl ether by dimerization of phenoxy radicals is ruled out by the lack of absorption bands around 1230 and 1100 cm−1 (the asymmetric and symmetric stretching of aryl ether group [34–36]). The formyl groups of salicylaldehyde showed two C-H stretching vibrations at 2846 and 2750 cm−1 and a C-H out-of-plane deformation vibration at 883 cm−1 (Fig. 6b), all of which disappeared after electrolysis. This indicates the conversion of –CHO groups during the electrolysis. A new broad band arose around 1366 cm−1 which can be assigned to COO stretching vibration (a broad doublet for symmetric and asymmetric COO stretching modes) [37,38]. In the IR spectra of sodium salicylate there is a very strong band at 1377 cm−1 which was also assigned to COO stretching vibration [39]. However, oxidation of the formyl group to carboxyl group seems not to occur, because the soluble salicylic acid was not detected by the in situ UV–vis spectroelectrochemistry. Therefore it is reasonable to conclude that the –CHO group of salicylaldehyde was converted to ester group, by a nucleophilic attack of phenoxy radical intermediate on the carbon atom of the formyl group. On the basis of all the above observations, we propose a parallelconsecutive reaction mechanism for electrochemical oxidation of salicylaldehyde in acidic, physiological and alkaline pH buffers (Scheme 1). The phenoxyl radicals were initially formed through one-electron one-proton oxidation, followed by an irreversible dimerization reaction to form phenyl benzoate and further polymerization to polysalicylate. This polyester was deposited on the electrode surface and was the principal oxidation product of salicylaldehyde, especially in alkaline media. Meanwhile, the phenoxyl radicals underwent tautomerism to form two semiquinone radicals, followed by the second oxidation to form trace amounts of 3-formyl-o-benzoquinone and p-benzoquinone mainly in acidic media. 4. Conclusion The oxidation of salicylaldehyde is pH-dependent. It was initially oxidized to reactive phenoxyl radicals via a one-electron one-proton transfer reaction, followed by a series of chemical and electrochemical steps leading to different final products. Polysalicylate was deposited on the electrode surface as the principal oxidation product of salicylaldehyde, especially in alkaline media. The phenoxyl radical was chemically more stable in the lower pH media, allowing further oxidation to form trace amounts of 3formyl-o-benzoquinone and p-benzoquinone in acidic media. The
[1] N. Matyasovszky, M. Tian, A.C. Chen, Kinetic study of the electrochemical oxidation of salicylic acid and salicylaldehyde using UV/vis spectroscopy and multivariate calibration, J. Phys. Chem. A 113 (2009) 9348–9353. [2] D. Ghosal, J. Chakraborty, P. Khara, T.K. Dutta, Degradation of phenanthrene via meta-cleavage of 2-hydroxy-1-naphthoic acid by Ochrobactrum sp. strain PWTJD, FEMS Microbiol. Lett. 313 (2010) 103–110. [3] S. Mallick, S. Chatterjee, T.K. Dutta, A novel degradation pathway in the assimilation of phenanthrene by Staphylococcus sp. strain PN/Y via meta-cleavage of 2-hydroxy-1-naphthoic acid: formation of trans-2,3-dioxo5-(2’-hydroxyphenyl)-pent-4-enoic acid, Microbiology 153 (2007) 2104–2115. [4] R.W. Eaton, P.J. Chapman, Bacterial metabolism of naphthalene: construction and use of recombinant bacteria to study ring cleavage of 1,2-dihydroxynaphthalene and subsequent reactionst, J Bacteriol. 174 (1992) 7542–7554. [5] M. Tian, B. Adams, J.L. Wen, R.M. Asmussen, A.C. Chen, Photoelectrochemical oxidation of salicylic acid and salicylaldehyde on titanium dioxide nanotube arrays, Electrochim. Acta 54 (2009) 3799–3805. [6] W.-L. Chou, C.-T. Wang, K.-Y. Huang, T.-C. Liu, Electrochemical removal of salicylic acid from aqueous solutions using aluminum electrodes, Desalination 271 (2011) 55–61. ´ J. Klíma, J. Ludvík, A spectroelec[7] K. Kratochvilová, I. Hoskovcová, J. Jirkovsky, trochemical study of chemisorption, anodic polymerization and degradation of salicylic acid on conductor and TiO2 surfaces, Electrochim. Acta 40 (1995) 2603–2609. [8] I. Gualandi, E. Scavetta, S. Zappoli, D. Tonelli, Electrocatalytic oxidation of salicylic acid by a cobalt hydrotalcite-like compound modified Pt electrode, Biosens. Bioelectron. 26 (2011) 3200–3206. [9] W.D. Zhang, B. Xu, Y.X. Hong, Y.X. Yu, J.S. Ye, J.Q. Zhang, Electrochemical oxidation of salicylic acid at well-aligned multiwalled carbon nanotube electrode and its detection, J.Solid State Electrochem. 14 (2010) 1713–1718. [10] Y.C. Zhu, X.Y. Guan, H.G. Ji, Electrochemical solid phase micro-extraction and determination of salicylic acid from blood samples by cyclic voltammetry and differential pulse voltammetry, J.Solid State Electrochem. 13 (2009) 1417–1423. [11] A.A.J. Torriero, J.M. Luco, L. Sereno, J. Raba, Voltammetric determination of salicylic acid in pharmaceuticals formulations of acetylsalicylic acid, Talanta 62 (2004) 247–254. [12] C. Lárez, O.P. Márquez, J. Márquez, Anodic oxidation and electropolimerization of salicylic acid, Rev. Téc. Ing. Univ. Zulia 29 (2006) 3–13. [13] V. Supalkova, J. Petrek, L. Havel, S. Krizkova, J. Petrlova, V. Adam, D. Potesil, P. Babula, M. Beklova, A. Horna, R. Kizek, Electrochemical sensors for detection of acetylsalicylic acid, Sensors 6 (2006) 1483–1497. [14] T.L. Lu, Y.C. Tsai, Electrocatalytic oxidation of acetylsalicylic acid at multiwalled carbon nanotube-alumina-coated silica nanocomposite modified glassy carbon electrodes, Sensor. Actuators B: Chem. 148 (2010) 590–594. [15] E. Wudarska, E. Chrzescijanska, E. Kusmierek, J. Rynkowski, Voltammetric studies of acetylsalicylic acid electrooxidation at platinum electrode, Electrochim. Acta 93 (2013) 189–194. [16] J. Xu, X. Zhuang, Poly-salicylic acid modified glassy-carbon electrode and its application, Talanta 38 (1991) 1191–1195. [17] J. Li, J. Yu, Q. Lin, Amperometric nonenzymatic determination of glucose free of interference based on poly (sulfosalicylic acid) modified nickel microelectrode, Anal. Lett. 43 (2010) 631–643. [18] X. Wang, H.M. Zhao, X. Quan, Y.Z. Zhao, S. Chen, Visible light photoelectrocatalysis with salicylic acid-modified TiO2 nanotube array electrode for p-nitrophenol degradation, J. Hazard. Mater. 166 (2009) 547–552. [19] E.E. Bancroft, J.S. Sidwell, H.N. Blount, Derivative linear sweep and derivative cyclic voltabsorptometry, Anal. Chem. 53 (1981) 1390–1394. [20] J.-B. He, Y. Zhou, F.-S. Meng, Time-derivative cyclic voltabsorptometry for voltammetric characterization of catechin film on a carbon-paste electrode: one voltammogram becomes four, J.Solid State Electrochem. 13 (2009) 679–685. [21] J.-B. He, S.-J. Yuan, J.-Q. Du, X.-R. Hu, Y. Wang, Voltammetric and spectral characterization of two flavonols for assay-dependent antioxidant capacity, Bioelectrochemistry 75 (2009) 110–116.
140
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
[22] X.-R. Hu, J.-B. He, Y. Wang, Y.-W. Zhu, J.-J. Tian, Oxidative spectroelectrochemistry of two representative coumarins, Electrochim. Acta 56 (2011) 2919–2925. [23] J.-B. He, Y. Wang, N. Deng, X.-Q. Lin, Study of the adsorption and oxidation of antioxidant rutin by cyclic voltammetry-voltabsorptometry, Bioelectrochemistry 71 (2007) 157–163. [24] D.R. Lide, CRC Handbook of Chemistry and Physics, Internet Version 2005 ed., CRC Press, Boca Raton, FL, 2005. [25] Y. Shih, J.-M. Zen, A.S. Kumar, P.-Y. Chen, Flow injection analysis of zinc pyrithione in hair care products on a cobalt phthalocyanine modified screenprinted carbon electrode, Talanta 62 (2004) 912–917. [26] F.V. Acholla, K.B. Mertes, Catalytic oxidation of 3,5-di-tertbutyl catechol (3,5DTBC) and 2,6-di-tertbutyl phenol (2,6-DTBP) by copper(II) and cobalt(II) complexes of a binucleating tetrapyrrole macrocycle, Bull, Chem. Soc. Ethiop. 3 (1989) 17–24. [27] W.M.A. El-Rahim, H. Moawad, Testing the performance of small scale bioremediation unit designed for bioremoval/enzymatic biodegradation of textile azo dyes residues, New York Sci. J. 3 (2010) 77–92. [28] G. Albarran, W. Boggess, V. Rassolov, R.H. Schuler, Absorption spectrum, mass spectrometric properties, and electronic structure of 1,2-benzoquinone, J Phys. Chem. A 114 (2010) 7470–7478. [29] E.S. Gil, R.O. Couto, Flavonoid electrochemistry: A review on the electroanalytical applications, Braz. J. Pharmacogn. 23 (2013) 542–558. [30] L. Horner, E. Geyer, Zur kenntnis der o-chinone, XXVII: Redoxpotentiale von brenzcatechin-derivaten, Chem. Berichte 98 (1965) 2016–2045.
[31] C. Queirós, A.M.G. Silva, S.C. Lopes, G. Ivanova, P. Gameiro, M. Rangel, A novel fluorescein-based dye containing a catechol chelating unit to sense iron(III), Dyes Pigments 93 (2012) 1447–1455. [32] O.T. Can, M. Bayramoglu, The effect of process conditions on the treatment of benzoquinone solution by electrocoagulation, J. Hazard. Mater. 173 (2010) 731–736. [33] M.I. Sirajuddin, A. Bhanger, A. Niaz, A. Rauf Shah, Ultra-trace level determination of hydroquinone in waste photographic solutions by UV-vis spectrophotometry, Talanta 72 (2007) 546–553. [34] M. Obi, S.y. Morino, K. Ichimura, Factors affecting photoalignment of liquid crystals induced by polymethacrylates with coumarin side chains, Chem. Mater. 11 (1999) 656–664. [35] R.T.S.M. Lakshmi, J. Meier-Haack, K. Schlenstedt, H. Komber, V. Choudhary, I.K. Varma, Synthesis, characterisation and membrane properties of sulphonated poly(aryl ether sulphone) copolymers, React. Funct. Polym. 66 (2006) 634–644. [36] M. Nechifor, Synthesis and properties of some aromatic polyamides with coumarin chromophores, React. Funct. Polym. 69 (2009) 27–35. [37] S.C. Pillai, S.W. Boland, S.M. Haile, Low-temperature crystallization of sol-gel processed Pb0.5 Ba0.5 TiO3 : Powders and oriented thin films, J. Am. Ceram. Soc. 87 (2004) 1388–1391. [38] M. Ibrahim, A. Nada, D.E. Kamal, Density functional theory and FTIR spectroscopic study of carboxyl group, Indian J. Pure Appl. Phys. 43 (2005) 911–917. [39] D. Philip, A., John, C.Y., Panicker, H.T., Varghese, FT-Raman, FT-IR and surface enhanced Raman scattering spectra of sodium salicylate, Spectrochim. Acta A 57 (2001) 1561–1566.
Contents lists available at ScienceDirect
Electrochimica Acta journal homepage: www.elsevier.com/locate/electacta
Spectroelectrochemistry of salicylaldehyde oxidation Yan Wang, Huan Jiang, Jing-Jing Tian, Jian-Bo He ∗ Anhui Key Lab of Controllable Chemical Reaction & Material Chemical Engineering, School of Chemical Engineering, Hefei University of Technology, Hefei 230009, China
a r t i c l e
i n f o
Article history: Received 26 August 2013 Received in revised form 9 November 2013 Accepted 13 January 2014 Available online 28 January 2014 Keywords: Thin-layer electrolysis Spectroelectrochemistry Cyclic voltabsorptometry Oxidation mechanism Salicylaldehyde
a b s t r a c t The advances in knowledge of the oxidation mechanism of salicylaldehyde are important in understanding its role and conversion in the involved oxidative degradation, synthesis and aerobic metabolism processes. The electrochemical oxidation of salicylaldehyde was investigated in different pH media using cyclic voltammetry, in situ UV–vis spectroscopy and cyclic voltabsorptometry based on a long opticalpath thin-layer electrochemical cell. ATR-FTIR spectroscopy was used for characterization of the oxidation products deposited on the electrode surfaces. Time-derivative cyclic voltabsorptograms were obtained at the characteristic wavelengths of salicylaldehyde and the soluble oxidized salicylaldehyde, for comparative discussion with the corresponding cyclic voltammograms. Two couples of redox peaks, subsequent to the main oxidation peak of salicylaldehyde, were observed on the voltabsorptograms but nearly indistinguishable on the voltammograms. Salicylaldehyde was initially oxidized to reactive phenoxyl radicals, followed by a series of transformation steps leading to different final products. A parallel-consecutive reaction mechanism was proposed for the pH-dependent formation of a deposited polyester and two trace amounts of soluble quinoid products. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Salicylaldehyde (o-hydroxybenzaldehyde) is a common highlyfunctionalized arene that has often been used as analysis reagent, spices and gasoline additive. It is also a key precursor in organic synthesis of a variety of chelating agents and other chemicals. Because of its widespread application, salicylaldehyde becomes an organic pollutant that finds its way into the water system through pharmaceutical medications, cosmetic products, and agricultural chemicals [1]. Also, salicylaldehyde naturally occurs in grape, tomato, cinnamon, milk and milk products, beer, coffee and tea. It has been found to be an intermediate formed during the aerobic metabolism of some polycyclic aromatic hydrocarbons [2–4]. Salicylic acid is one of the possible oxidation products of salicylaldehyde. Electrochemical oxidation of salicylic acid [1,5–12] and acetylsalicylic acid [13–15] has been widely investigated focusing on the electrochemical removal and degradation [1,5–7], electrochemical detection [8–11,13,14], electro-polymerization [7,12] and electro-oxidation mechanism [9,15]. Salicylic acid and its derivatives have also been used as an electrode modifier for the determination of trace copper(II) in water [16], the amperometric
∗ Corresponding author. Tel.: +86 551 6290 4653; fax: +86 551 6290 1450. E-mail address: [email protected] (J.-B. He). 0013-4686/$ – see front matter © 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.electacta.2014.01.087
nonenzymatic determination of glucose free of interference [17], and for the visible light photoelectrocatalytic degradation of pnitrophenol [18]. However, only a few reports are available on electrochemical oxidation of salicylaldehyde [1,5]. Chen et al. studied the kinetics of electrochemical oxidation of salicylic acid and salicylaldehyde in 0.5 M H2 SO4 at a Ti/IrO2 -SnO2 -Sb2 O5 electrode [1] and of their photoelectrochemical oxidation on titanium dioxide nanotube arrays [5] using in situ UV-vis spectroscopy. They revealed the impact of temperature, initial concentration, current density and supporting electrolyte etc. on the degradation kinetics. The electro-oxidation pathway of salicylaldehyde is currently unclear and deserves further study for better understanding its role and conversion in the involved oxidative degradation, synthesis and aerobic metabolism processes. A combination of cyclic voltammetry (CV) and cyclic voltabsorptometry (CVA) allows simultaneously measuring the potential-dependent current and absorbance [19,20]. The light-absorbing reactant, intermediates and products involved in the electrochemical processes can be followed photometrically by setting the corresponding characteristic wavelengths. The time-derivative cyclic voltabsorptometry (DCVA) presents the data curves in a shape resembling that of the corresponding CV curves, providing direct comparison between the CV and DCVA peaks. Unlike the CV peaks, the DCVA peaks may be caused not only by electrochemical reaction but also by chemical conversion. Moreover, no charging/discharging background signals are
134
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
recorded on the DCVA curves, which may improve the response to the electrochemical and chemical processes. The CVA has generally been used along with an optically transparent electrode, in a light incident mode perpendicular to the electrode surface. In the present work, the CVA technique was used with a long path-length thin-layer electrochemical cell to investigate the electrochemical oxidation of salicylaldehyde in acidic, physiological and alkaline pH media. This thin-layer cell, in a light incident mode parallel to and close to the electrode surface, allows monitoring soluble intermediates and products generating from an opaque electrode. The use of long path-length greatly improves the sensitivity of detection of the analytes present in the thinlayer solutions. Enhanced responses are therefore obtained to the potential- and pH-dependent formation and reduction of the soluble products. 2. Experimental
Fig. 1. CVs of 100 M salicylaldehyde in different pH buffers. pH (1→7): 1.8, 3.2, 5.8, 7.4, 8.0, 9.2 and 11.9. Scan rate 100 mV s−1 . Inset: pH dependence of the anodic peak potential.
2.1. Chemicals and solutions 2.3. Procedures Reagent grade salicylaldehyde (98%+ pure) from Guangfu (Tianjin, China) was used as received. Spectrograde graphite powder (320 mesh) and spectrograde paraffin wax (solidification point 62–65 ◦ C) from Shanghai Chemical Works were used for preparing the graphite paste electrode. Graphite sheet was from Aidahengsheng (Tianjin, China). Other chemicals were of analytical grade from China-Reagent group. All solutions were prepared with doubly-distilled water from an all-glass distillation apparatus. Salicylaldehyde stock solution was prepared with ethanol and water in a volume ratio of 1:4 by means of ultrasonic agitation, and then kept at 4 ◦ C in a refrigerator. The supporting electrolytes with various pH values were a mixture of 0.2 M Britton-Robinson (BR) buffered solution and 0.5 M KCl. High pure N2 was used for solution deaeration. 2.2. Apparatus, electrodes and cells Cyclic voltammetry and spectroelectrochemistry were carried out on a CHI660 C electrochemical workstation (Chenhua, Shanghai, China). UV-vis spectroscopic and photometric measurements were carried out on an UV-2500 spectrophotometer (Shimadzu, Japan) to monitor the soluble reactants and products under potentiostatic and potentiodynamic conditions. Attenuated total reflectance (ATR)-FTIR spectra were recorded on a Vertex 80 infrared spectrometer coupled to a Hyperion 2000 IR microscope (Bruker Optics). The three-electrode system was composed of a graphite working electrode, a KCl-saturated Ag/AgCl reference electrode (0.195 V vs. SHE, self-made) and a platinum wire counter electrode. The graphite working electrode used for the conventional voltammetric measurements was a disk solid graphite paste electrode (GPE) with a smaller geometrical area of 0.049 cm2 , whereas the electrode for the thin-layer electrochemical experiments was a quadrate graphite sheet electrode with a larger area of 0.77 cm2 . The graphite paste electrode was selected because of its advantages of low background currents, low noise and fast base line stabilization, and its preparation was described previously [21,22]. A conventional single-compartment cell was used for the voltammetric measurements. A thin-layer spectroelectrochemical cell was self-made, using a standard quartz photometric cell with 10 mm optical path length as the cell body. The schematic view of the thin-layer cell can be found in literature [23]. The incident light beam parallels to the working electrode and goes through the thin-layer electrolyte solution (0.2 mm thick) on the electrode surface.
Salicylaldehyde solutions were bubbled with high pure N2 for about 15 min to remove dissolved oxygen before being put into the electrochemical cells. The working electrode was polished carefully with 800 grit emery paper. Before each run, the working electrode was cleaned and activated by repetitive cyclic scans between 0.0 and 1.6 V in 0.1 M NaHCO3 solution until only the background current remained. Considering the adsorption of salicylaldehyde on graphite electrode, a pre-accumulation step was always performed in open circuit for 60 s. All the experiments were carried out at room temperature (22 ± 1 ◦ C). The UV–vis absorption spectra were recorded while the thinlayer solution was electrolyzed at a constant potential. Blank BR buffers were used for the spectral baseline correction. Cyclic voltabsorptometry was performed at a certain wavelength to follow the absorbance changes of species in the thin layer solution. The ATRFTIR spectra were recorded ex situ for characterizing oxidation products deposited on the electrode surfaces. The spectra were collected with 64 scans and at a resolution of 4 cm−1 . 3. Results and discussion 3.1. Cyclic voltammetry Cyclic voltammetric behavior of salicylaldehyde was examined in a wide pH range of 1.8–11.9. The cyclic voltammograms recorded using the GPE shows one anodic peak (Fig. 1), corresponding to the oxidation of salicylaldehyde. The anodic peak current density (jp,A ) first decreased and then increased with increasing pH, showing a minimum peak current at pH 8.0. The anodic peak potential (Ep,A ) shifted negatively with increasing pH until reaching pH 9.2, following a linear relationship Ep,A /V = 1.3008–0.0571 pH (R = –0.99837) (the inset of Fig. 1). The equation slope of –57 mV per pH unit supports a reaction mechanism involving an equal number of electrons and protons. At the pH values greater than 9.2, the peak potential did not depend on the pH of media, suggesting the absence of proton transfer in these pH conditions. The pKa values of salicylaldehyde, phenol and benzaldehyde are about 8.37, 9.99 and 14.90 at 25 ◦ C [24]. Accordingly, the initial electro-oxidation should occur at the phenolic group of salicylaldehyde, through one-electron one-proton transfer for the undissociated molecules but only one-electron transfer for the deprotonated phenolate anions. The corresponding cathodic peaks were extremely small (Fig. 1), indicating the poor reversibility of salicylaldehyde oxidation. The
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
species reducible at the cathodic peaks were not the principal products arising from the oxidation of salicylaldehyde. The amount of the reducible species further decreased as the pH of the media increased, as reflected by the decrease in the cathodic peak current with increasing pH. A variable scan rate (v) experiment was performed over a scan rate range of 1–1000 mV s−1 , and a linear relationship of log(jp,A ) versus log(v) (data not shown) was observed for 100 M salicylaldehyde, with slopes of 0.42, 0.40 and 0.40 at pH 1.8, 7.4 and 11.9, respectively. The slopes near 0.5 indicate a diffusion-controlled [25] oxidation process of salicylaldehyde.
3.2. In situ thin-layer UV–vis spectroscopy Fig. 2 shows the time-dependent spectral absorbance of the thin-layer electrolyte taken at 62 s intervals before and during the controlled potential oxidation of salicylaldehyde in different pH BR buffers. In the pH range of 1.8 − 7.4, salicylaldehyde exhibits three absorption bands locating around 216 (band I), 256 (band II) and 326 nm (band III) before electrolysis (the first curves in Fig. 2A−D), the former two corresponding to the →* transition of the benzene ring and the latter to the weak n→* transition of the C = O group. A new absorption band arose around 380 nm (band III , see the first curve in Fig. 2E) at pH 8.0, a value close to the pKa of salicylaldehyde (≈ 8.37 [24]). This band showed an increase in intensity as the pH increased to 9.2 (Fig. 2F), at the expense of the disappearance of the band III around 326 nm. Accordingly, the bands III and III can be assigned to the undissociated molecule and deprotonated phenolate anion of salicylaldehyde, respectively. When the hydroxyl group is not deprotonated, a resonance structure can be stabilized through an intramolecular hydrogen bond between the hydroxyl hydrogen and the carbonyl oxygen. The electrochemical oxidation of salicylaldehyde was carried out at the anodic peak potential, which is dependent on the buffer pH. At all the pH values tested, the oxidation of salicylaldehyde resulted in a large decrease in intensity of the bands I, II, III and III , especially at the alkaline pH values (Fig. 2). Five isosbestic points are observed at the pH 1.8 and 3.2, while less isosbestic points are present at the higher pH values. A new absorption band with a low intensity appeared around 408 nm (band IV) attributed to a soluble oxidation product of salicylaldehyde. This emerging band decreased in maximum intensity with increasing pH, and became vanish at pH 8.0 and 9.2. This indicates that the soluble product with a characteristic wavelength of about 408 nm can not be obtained in the media at the highest pHs. After the exhaustive electrolysis, the low absorption intensity over the whole tested wavelength range suggests that only very small amounts of soluble light-absorbing species were present in the thin-layer. Therefore, the principal oxidation products of salicylaldehyde should be deposited on the electrode surface, especially at the higher pH values. This was also evidenced by a cyclic voltammetry experiment in a common-volume cell, which showed a fast decrease of current in the following cycles (1 mM salicylaldehyde, scan rate 20 mV s−1 , data not shown), especially at the pH not less than 7.4. The resulting film on the electrode surface is non-conductive, which obstructed further oxidation of salicylaldehyde from the solution in the subsequent cycles. One possible oxidation product of salicylaldehyde is salicylic acid, which has an absorption band centered at about 300 nm [1–4,6]. Fig. 2 shows the absence of absorption around this wavelength after electrolysis, thereby excluding the formation of salicylic acid. The band IV emerging around 408 nm supports the formation of an o-quinone structure [26–28] in the thin-layer solution.
135
3.3. Thin-layer cyclic voltabsorptometry 3.3.1. pH 1.8 Multi-cycle cyclic voltabsorptometry was performed in the thin-layer cell at the characteristic absorption wavelengths of salicylaldehyde and a product (Fig. 3). A slow scan rate (2.0 mV s−1 ) was used, so that the diffusion layer thickness was larger than the thin layer thickness. Fig. 3A and C show two CVA curves recorded at 256 and 408 nm, respectively, for the pH 1.8, 100 M salicylaldehyde. During the cyclic scan, the variation of absorbance at 256 and 408 nm (A256 and A408 ) with potential reflects the changes in concentrations of salicylaldehyde and the o-quinone product, respectively. The first scan in the positive direction caused a slow decrease in A256 until 0.98 V due to the potential-driven adsorption of salicylaldehyde, followed by a sharp decline in A256 due to the anodic oxidation of salicylaldehyde (Fig. 3A). During the negative scan, only very small change in A256 was observed, indicating the irreversibility of salicylaldehyde oxidation. Meanwhile, the fast rise in A408 during the positive scan corresponds to the formation of an o-quinone structure (Fig. 3C), and the decline in A408 during the negative scan is consistent with the reducibility of o-quinone. The derivative absorbance signal (dA /dt) is proportional to the reaction rate of the monitored species. A well-shaped DCVA peak A1 appeared on both the DCVA curves, corresponding to the principal oxidation peak on the thin-layer CV curve (also marked as A1, see the inset of Fig. 3B). The DCVA peak A1 obtained at the wavelengths 256 and 408 nm was centered at 1.08 and 1.14 V, respectively, with a potential (or time) difference due to the diffusion lag between the reactant and the generated product in the thin-layer. The counterpart of peak A1 marked as peak C1 was seen at 408 nm (Fig. 3D) but not at 256 nm (Fig. 3B), which can be attributed to chemical conversion of an intermediate with absorption at 408 nm (possibly a semiquinone radical) to a more stable product without or with less strong absorption at this wavelength, but not to the parent salicylaldehyde. This final product may be an insoluble species deposited on the electrode as discussed above, without causing light-absorption in the thin-layer. On the other hand, the corresponding cathodic voltammetric peak C1 in the inset of Fig. 3B was extremely small, which should be associated with a reversed reduction of a highly unstable intermediate formed in initial oxidation of salicylaldehyde, e.g. phenoxy radicals, which are key intermediates in the anodic oxidation of phenols. Interestingly, two subsequent pairs of DCVA peaks, C2/A2 and C3/A3, were definitely detected following the first couple A1/C1 (Fig. 3B and D). In contrast, the corresponding voltammetric pairs in the inset of Fig. 3B were very small and nearly indistinguishable. Both the peak pairs on the DCVA curves showed a peak-topeak separation of about 60 mV, indicating two quasi-reversible redox couples, most likely o-quinone/o-diphenol and p-quinone/pdiphenol, present in the thin-layer aqueous-alcoholic solution. A similar pair of redox peaks following a main anodic peak has also been observed in the second cycle of cyclic voltammetric scans of salicylic acid at multiwalled carbon nanotube electrode [9] and graphite-epoxy-composed solid electrode [10] as well as of phenol on polyethylenedioxythiophene deposited stainless steel substrate [29], all of which were attributed to the quasi-reversible behavior of benzoquinone/hydroquinone. The peak pair C2/A2 was obtained at 408 nm, with a midpoint potential of 0.795 V (vs. Ag/AgCl/KClsat ) at pH 1.8. The 408 nm band has been assigned to an o-quinone structure. o-Quinone/catechol redox couple without a–CHO substituent has a redox potential of 0.494 V (vs. Ag/AgCl/KClsat ) at pH 1.8 (derived from the standard redox potential of 0.795 V vs. SHE [30]), lower by about 0.3 V than the midpoint potential of the peak pair C2/A2. This potential difference is due to the presence of formyl group on the carbon ring as a
136
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
Fig. 2. In situ thin layer UV–vis spectra of 100 M salicylaldehyde in different pH buffers during electro-oxidation. (A) pH = 1.8, E = 1.12 V; (B) pH = 3.2, E = 1.06 V; (C) pH = 5.8, E = 0.91 V; (D) pH = 7.4, E = 0.82 V; (E) pH = 8.0, E = 0.81 V; (F) pH = 9.2, E = 0.75 V. Spectral tracing was repeated every 62 s after the potential was applied. The first curve in every panel was recorded before electrolysis.
strong electron-withdrawing substituent, which makes the reduction easier but more difficult to be oxidized. This explanation is supported by the fact that, at the same pH, the oxidation potential of phenol [29] was lower by about 0.25 V than that of salicylaldehyde (Fig. 1). At pH 1.8, 3-formylcatechol has not absorption around 408 nm [31], thus the reduction of 3-formyl-o-benzoquinone to 3formylcatechol led to a decrease in A408 and then the appearance of DCVA peak C2. Inversely, the oxidation of 3-formylcatechol to 3formyl-o-benzoquinone resulted in the peak A2 in the next positive scan. The pair C3/A3 was observed at 256 nm, with a midpoint potential of 0.40 V, which fits well with the redox potential of pbenzoquinone/hydroquinone couple, 0.398 V (vs. Ag/AgCl/KClsat ) at pH 1.8 (derived from Ref. [32]). The electron-withdrawing effect of–CHO on the redox potential was not observed, which suggests that the–CHO group might be removed from the pquinone ring during the oxidation. At 256 nm, p-benzoquinone has an absorption much stronger than hydroquinone [33],
therefore the reduction of p-quinone to hydroquinone led to the decrease in A256 and then the appearance of DCVA peak C3. The reverse oxidation resulted in the peak A3 in the next positive scan. 3.3.2. pH 7.4 and 9.2 Fig. 4 shows the CVA and DCVA curves of salicylaldehyde at pH 7.4, recorded at the same two wavelengths as the previous study. A significant difference with the results obtained at pH 1.8 was that the peaks C3 and A3 did not appear on the 256 nm DCVA curve (Fig. 4B), indicating that the p-benzoquinone was not formed at pH 7.4 during the oxidation of salicylaldehyde. Another important feature is that the absorbance at 408 nm started decreasing even before the scan was reversed (Fig. 4C), thus the DCVA peak C1 occurred with a peak potential more positive than that of the anodic peak A1 (Fig. 4D). This means that the conversion of an intermediate at the DCVA peak C1 was not an electrochemical reduction process and occurred more
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
137
Fig. 3. Multi-cycle thin-layer CVA (A, C) and DCVA (B, D) curves of 100 M salicylaldehyde in pH 1.8 buffer. Scan rate 2.0 mV s−1 , cycle number 3, wavelength 256 (A, B) and 408 nm (C, D). Inset in panel (B): the corresponding thin-layer CV curve.
Fig. 4. Multi-cycle thin-layer CVA (A, C) and DCVA (B, D) curves of 100 M salicylaldehyde in pH 7.4 buffer. Scan rate 2.0 mV s−1 , cycle number 3, wavelength 256 (A, B) and 408 nm (C, D). Inset in panel (B): the corresponding thin-layer CV curve.
138
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
Fig. 5. Multi-cycle thin-layer CVA (A) and DCVA (B) curves of 100 M salicylaldehyde in pH 11.9 buffer. Scan rate 2.0 mV s−1 , wavelength 227 nm, cycle number 3. Inset in panel (B): the corresponding thin-layer CV.
easily than at pH 1.8. The DCVA peaks C2 and A2, attributed to the 3-formyl-o-benzoquinone/3-formylcatechol redox couple, were still observed on the 408 nm DCVA curve (Fig. 4D). This peak couple showed a midpoint potential of 0.480 V at pH 7.4, lower than the value at pH 1.8, with a slope of −56 mV per pH unit which supports the two-electron two-proton quinone/catechol redox reaction. At the characteristic wavelength of salicylaldehyde, similar CVA and DCVA curves were obtained at pH 11.9 (Fig. 5). However, no changes in absorbance were recorded at 408 nm (data not shown). The chemical conversion of the intermediate was too fast to form the soluble o-quinone product at such a high pH value.
3.4. ATR-FTIR spectroscopy The solid oxidation products of salicylaldehyde deposited on graphite sheet electrode were characterized by ATR-FTIR (Fig. 6a). The absorptions in the range of 1600 − 1454 cm−1 correspond to the aromatic C = C ring stretching vibrations. The two absorptions at 2920 and 2850 cm−1 are known to be the characteristic bands of CH2 groups in asymmetric and symmetric stretching modes, respectively. The CH2 groups may come from ethanol and/or acetic acid in the background electrolyte solution. The phenolic hydroxyl group of salicylaldehyde was no longer present in the deposits, as indicated by the vanishing of the broad O-H stretching band
Scheme 1. A parallel-consecutive reaction mechanism proposed for the electro-oxidation of salicylaldehyde in the pH range of 1.8–11.9.
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
139
latter two quinoid products show reversible redox behaviors, which can be detected at two corresponding wavelengths by thin-layer cyclic voltabsorptometry. This work provides an enhanced observation of trace amounts of redox couples by cyclic voltabsorptometry, and hence enriches the knowledge about the oxidation pathway of salicylaldehyde.
Acknowledgment The authors gratefully acknowledge the financial support from the National Nature Science Foundation of China (No. 51203040, 20972038).
References Fig. 6. (a) ATR-FTIR spectrum for the oxidized species on a graphite electrode after electrolysis at 0.81 V in pH 8.0 1.0 mM salicylaldehyde. (b) FTIR of salicylaldehyde in KBr pellet shown for comparison. The absorption intensities of spectrum (a) were 100-times magnified for clearer presentation.
between 3500 and 3000 cm−1 and the strong C-O stretching band at about 1276 cm−1 (compare Fig. 6a with b). Possible formation of aryl ether by dimerization of phenoxy radicals is ruled out by the lack of absorption bands around 1230 and 1100 cm−1 (the asymmetric and symmetric stretching of aryl ether group [34–36]). The formyl groups of salicylaldehyde showed two C-H stretching vibrations at 2846 and 2750 cm−1 and a C-H out-of-plane deformation vibration at 883 cm−1 (Fig. 6b), all of which disappeared after electrolysis. This indicates the conversion of –CHO groups during the electrolysis. A new broad band arose around 1366 cm−1 which can be assigned to COO stretching vibration (a broad doublet for symmetric and asymmetric COO stretching modes) [37,38]. In the IR spectra of sodium salicylate there is a very strong band at 1377 cm−1 which was also assigned to COO stretching vibration [39]. However, oxidation of the formyl group to carboxyl group seems not to occur, because the soluble salicylic acid was not detected by the in situ UV–vis spectroelectrochemistry. Therefore it is reasonable to conclude that the –CHO group of salicylaldehyde was converted to ester group, by a nucleophilic attack of phenoxy radical intermediate on the carbon atom of the formyl group. On the basis of all the above observations, we propose a parallelconsecutive reaction mechanism for electrochemical oxidation of salicylaldehyde in acidic, physiological and alkaline pH buffers (Scheme 1). The phenoxyl radicals were initially formed through one-electron one-proton oxidation, followed by an irreversible dimerization reaction to form phenyl benzoate and further polymerization to polysalicylate. This polyester was deposited on the electrode surface and was the principal oxidation product of salicylaldehyde, especially in alkaline media. Meanwhile, the phenoxyl radicals underwent tautomerism to form two semiquinone radicals, followed by the second oxidation to form trace amounts of 3-formyl-o-benzoquinone and p-benzoquinone mainly in acidic media. 4. Conclusion The oxidation of salicylaldehyde is pH-dependent. It was initially oxidized to reactive phenoxyl radicals via a one-electron one-proton transfer reaction, followed by a series of chemical and electrochemical steps leading to different final products. Polysalicylate was deposited on the electrode surface as the principal oxidation product of salicylaldehyde, especially in alkaline media. The phenoxyl radical was chemically more stable in the lower pH media, allowing further oxidation to form trace amounts of 3formyl-o-benzoquinone and p-benzoquinone in acidic media. The
[1] N. Matyasovszky, M. Tian, A.C. Chen, Kinetic study of the electrochemical oxidation of salicylic acid and salicylaldehyde using UV/vis spectroscopy and multivariate calibration, J. Phys. Chem. A 113 (2009) 9348–9353. [2] D. Ghosal, J. Chakraborty, P. Khara, T.K. Dutta, Degradation of phenanthrene via meta-cleavage of 2-hydroxy-1-naphthoic acid by Ochrobactrum sp. strain PWTJD, FEMS Microbiol. Lett. 313 (2010) 103–110. [3] S. Mallick, S. Chatterjee, T.K. Dutta, A novel degradation pathway in the assimilation of phenanthrene by Staphylococcus sp. strain PN/Y via meta-cleavage of 2-hydroxy-1-naphthoic acid: formation of trans-2,3-dioxo5-(2’-hydroxyphenyl)-pent-4-enoic acid, Microbiology 153 (2007) 2104–2115. [4] R.W. Eaton, P.J. Chapman, Bacterial metabolism of naphthalene: construction and use of recombinant bacteria to study ring cleavage of 1,2-dihydroxynaphthalene and subsequent reactionst, J Bacteriol. 174 (1992) 7542–7554. [5] M. Tian, B. Adams, J.L. Wen, R.M. Asmussen, A.C. Chen, Photoelectrochemical oxidation of salicylic acid and salicylaldehyde on titanium dioxide nanotube arrays, Electrochim. Acta 54 (2009) 3799–3805. [6] W.-L. Chou, C.-T. Wang, K.-Y. Huang, T.-C. Liu, Electrochemical removal of salicylic acid from aqueous solutions using aluminum electrodes, Desalination 271 (2011) 55–61. ´ J. Klíma, J. Ludvík, A spectroelec[7] K. Kratochvilová, I. Hoskovcová, J. Jirkovsky, trochemical study of chemisorption, anodic polymerization and degradation of salicylic acid on conductor and TiO2 surfaces, Electrochim. Acta 40 (1995) 2603–2609. [8] I. Gualandi, E. Scavetta, S. Zappoli, D. Tonelli, Electrocatalytic oxidation of salicylic acid by a cobalt hydrotalcite-like compound modified Pt electrode, Biosens. Bioelectron. 26 (2011) 3200–3206. [9] W.D. Zhang, B. Xu, Y.X. Hong, Y.X. Yu, J.S. Ye, J.Q. Zhang, Electrochemical oxidation of salicylic acid at well-aligned multiwalled carbon nanotube electrode and its detection, J.Solid State Electrochem. 14 (2010) 1713–1718. [10] Y.C. Zhu, X.Y. Guan, H.G. Ji, Electrochemical solid phase micro-extraction and determination of salicylic acid from blood samples by cyclic voltammetry and differential pulse voltammetry, J.Solid State Electrochem. 13 (2009) 1417–1423. [11] A.A.J. Torriero, J.M. Luco, L. Sereno, J. Raba, Voltammetric determination of salicylic acid in pharmaceuticals formulations of acetylsalicylic acid, Talanta 62 (2004) 247–254. [12] C. Lárez, O.P. Márquez, J. Márquez, Anodic oxidation and electropolimerization of salicylic acid, Rev. Téc. Ing. Univ. Zulia 29 (2006) 3–13. [13] V. Supalkova, J. Petrek, L. Havel, S. Krizkova, J. Petrlova, V. Adam, D. Potesil, P. Babula, M. Beklova, A. Horna, R. Kizek, Electrochemical sensors for detection of acetylsalicylic acid, Sensors 6 (2006) 1483–1497. [14] T.L. Lu, Y.C. Tsai, Electrocatalytic oxidation of acetylsalicylic acid at multiwalled carbon nanotube-alumina-coated silica nanocomposite modified glassy carbon electrodes, Sensor. Actuators B: Chem. 148 (2010) 590–594. [15] E. Wudarska, E. Chrzescijanska, E. Kusmierek, J. Rynkowski, Voltammetric studies of acetylsalicylic acid electrooxidation at platinum electrode, Electrochim. Acta 93 (2013) 189–194. [16] J. Xu, X. Zhuang, Poly-salicylic acid modified glassy-carbon electrode and its application, Talanta 38 (1991) 1191–1195. [17] J. Li, J. Yu, Q. Lin, Amperometric nonenzymatic determination of glucose free of interference based on poly (sulfosalicylic acid) modified nickel microelectrode, Anal. Lett. 43 (2010) 631–643. [18] X. Wang, H.M. Zhao, X. Quan, Y.Z. Zhao, S. Chen, Visible light photoelectrocatalysis with salicylic acid-modified TiO2 nanotube array electrode for p-nitrophenol degradation, J. Hazard. Mater. 166 (2009) 547–552. [19] E.E. Bancroft, J.S. Sidwell, H.N. Blount, Derivative linear sweep and derivative cyclic voltabsorptometry, Anal. Chem. 53 (1981) 1390–1394. [20] J.-B. He, Y. Zhou, F.-S. Meng, Time-derivative cyclic voltabsorptometry for voltammetric characterization of catechin film on a carbon-paste electrode: one voltammogram becomes four, J.Solid State Electrochem. 13 (2009) 679–685. [21] J.-B. He, S.-J. Yuan, J.-Q. Du, X.-R. Hu, Y. Wang, Voltammetric and spectral characterization of two flavonols for assay-dependent antioxidant capacity, Bioelectrochemistry 75 (2009) 110–116.
140
Y. Wang et al. / Electrochimica Acta 125 (2014) 133–140
[22] X.-R. Hu, J.-B. He, Y. Wang, Y.-W. Zhu, J.-J. Tian, Oxidative spectroelectrochemistry of two representative coumarins, Electrochim. Acta 56 (2011) 2919–2925. [23] J.-B. He, Y. Wang, N. Deng, X.-Q. Lin, Study of the adsorption and oxidation of antioxidant rutin by cyclic voltammetry-voltabsorptometry, Bioelectrochemistry 71 (2007) 157–163. [24] D.R. Lide, CRC Handbook of Chemistry and Physics, Internet Version 2005 ed., CRC Press, Boca Raton, FL, 2005. [25] Y. Shih, J.-M. Zen, A.S. Kumar, P.-Y. Chen, Flow injection analysis of zinc pyrithione in hair care products on a cobalt phthalocyanine modified screenprinted carbon electrode, Talanta 62 (2004) 912–917. [26] F.V. Acholla, K.B. Mertes, Catalytic oxidation of 3,5-di-tertbutyl catechol (3,5DTBC) and 2,6-di-tertbutyl phenol (2,6-DTBP) by copper(II) and cobalt(II) complexes of a binucleating tetrapyrrole macrocycle, Bull, Chem. Soc. Ethiop. 3 (1989) 17–24. [27] W.M.A. El-Rahim, H. Moawad, Testing the performance of small scale bioremediation unit designed for bioremoval/enzymatic biodegradation of textile azo dyes residues, New York Sci. J. 3 (2010) 77–92. [28] G. Albarran, W. Boggess, V. Rassolov, R.H. Schuler, Absorption spectrum, mass spectrometric properties, and electronic structure of 1,2-benzoquinone, J Phys. Chem. A 114 (2010) 7470–7478. [29] E.S. Gil, R.O. Couto, Flavonoid electrochemistry: A review on the electroanalytical applications, Braz. J. Pharmacogn. 23 (2013) 542–558. [30] L. Horner, E. Geyer, Zur kenntnis der o-chinone, XXVII: Redoxpotentiale von brenzcatechin-derivaten, Chem. Berichte 98 (1965) 2016–2045.
[31] C. Queirós, A.M.G. Silva, S.C. Lopes, G. Ivanova, P. Gameiro, M. Rangel, A novel fluorescein-based dye containing a catechol chelating unit to sense iron(III), Dyes Pigments 93 (2012) 1447–1455. [32] O.T. Can, M. Bayramoglu, The effect of process conditions on the treatment of benzoquinone solution by electrocoagulation, J. Hazard. Mater. 173 (2010) 731–736. [33] M.I. Sirajuddin, A. Bhanger, A. Niaz, A. Rauf Shah, Ultra-trace level determination of hydroquinone in waste photographic solutions by UV-vis spectrophotometry, Talanta 72 (2007) 546–553. [34] M. Obi, S.y. Morino, K. Ichimura, Factors affecting photoalignment of liquid crystals induced by polymethacrylates with coumarin side chains, Chem. Mater. 11 (1999) 656–664. [35] R.T.S.M. Lakshmi, J. Meier-Haack, K. Schlenstedt, H. Komber, V. Choudhary, I.K. Varma, Synthesis, characterisation and membrane properties of sulphonated poly(aryl ether sulphone) copolymers, React. Funct. Polym. 66 (2006) 634–644. [36] M. Nechifor, Synthesis and properties of some aromatic polyamides with coumarin chromophores, React. Funct. Polym. 69 (2009) 27–35. [37] S.C. Pillai, S.W. Boland, S.M. Haile, Low-temperature crystallization of sol-gel processed Pb0.5 Ba0.5 TiO3 : Powders and oriented thin films, J. Am. Ceram. Soc. 87 (2004) 1388–1391. [38] M. Ibrahim, A. Nada, D.E. Kamal, Density functional theory and FTIR spectroscopic study of carboxyl group, Indian J. Pure Appl. Phys. 43 (2005) 911–917. [39] D. Philip, A., John, C.Y., Panicker, H.T., Varghese, FT-Raman, FT-IR and surface enhanced Raman scattering spectra of sodium salicylate, Spectrochim. Acta A 57 (2001) 1561–1566.