Applied Surface Science 534 (2020) 147517

Contents lists available at ScienceDirect

Applied Surface Science

journal homepage: www.elsevier.com/locate/apsusc

Electrochemical oxidation of toluene with controlled selectivity: The effect T

of carbon anode
Bora Seoa, Woong Hee Leea, Young Jin Sab, Ung Leea, Hyung-Suk Oha, Hyunjoo Leea,
⁎

a
Clean Energy Research Center, Korea Institute of Science and Technology (KIST), Seoul 02792, Republic of Korea
b
Department of Chemistry, Kwangwoon University, Seoul 01897, Republic of Korea

ARTICLE INFO ABSTRACT

Keywords: Benzaldehyde is much easier to be oxidized than toluene, which makes the selective oxidation of toluene to
Direct electrochemical oxidation benzaldehyde highly challenging. Here, we investigated the electrochemical oxidation of toluene to benzalde-
Toluene hyde via direct electron transfer at a carbon electrode. Selectivity to benzaldehyde depended on the current and
Benzaldehyde electrode type, being highest (92%) for a graphite rod. The higher selectivity to benzaldehyde at graphite rod
Selectivity control
attribute to its high electrochemically active surface area and the large number of defect sites, which favorably
Carbon electrode
absorb benzaldehyde, thereby preventing further oxidation to benzoic acid. Tafel analysis revealed the rate-
determining step in the electrochemical toluene oxidation reaction, which has been underexplored previously.
The developed method was successfully applied to other para-substituted toluene and was therefore well suited
for the sustainable production of benzaldehyde derivatives.

1. Introduction performed at ambient temperature and pressure; (iii) the applied po-
tential can be adjusted to improve chemoselectivity, and (iv) electro-
The selective oxidation of toluene to benzaldehyde or benzoic acid chemical reactions such as the CO2 reduction reaction and the hydrogen
is highly important for the production of versatile intermediates and evolution reaction (HER) powered by renewable energies can be in-
raw materials for the manufacture of perfumes, pharmaceuticals, tegrated to significantly increase the worth of the electro-oxidation of
plasticizers, food preservatives, and resin modifiers [1–3], as high- organics [13].
lighted by the steadily growing global market for these two compounds Recently, several research groups reported the electrochemical
[4]. As benzaldehyde is much easier (by several orders of magnitude) to oxidation of sp3 C–H bonds in aromatic alkanes to produce value-added
be oxidized than toluene, the selective oxidation of the latter to the aldehydes [14–17]. In these studies, mediated electrochemical oxida-
former, i.e., the minimization of further oxidation to benzoic acid, re- tion was used to drive the reaction at low potentials of ~0.5–1.2 V vs.
mains a challenge [5,6]. Ag/AgCl. However, the use of organic mediators and/or redox metal
Benzaldehyde is commonly produced by thermo-oxidation pro- catalysts lacks economic viability because of difficult separation and
cesses such as electrophilic formylation [7], haloarene carbonylation recovery. In addition, radicals generated during organic mediator oxi-
[8], acid chloride reduction [9], and metal oxide–mediated auto-oxi- dation can react with organic substrates to produce undesired by-pro-
dation [10], most of which require the use of pressurized O2 or air at ducts [18,19], while redox metal catalysts can also chelate organic
high temperature (> 100 °C), rely on metal (Pd, Zn, Mn, or Al) cata- products [19].
lysts, produce abundant by-products, and consume much energy. Re- To address the above challenges, we herein employed direct elec-
cently, toluene-to-benzaldehyde oxidation has been achieved under tron transfer between toluene and the anode in the absence of organic
mild conditions via the introduction of active radical species [1,11,12], mediators and redox metal catalysts under aerobic conditions. As the
which increase selectivity toward benzaldehyde but often compromise anode, carbon materials were used because of their high economic
yield. Consequently, electrochemical oxidation of methylarenes has feasibility (low price and abundancy) and high stability under oxidative
emerged as a promising alternative to conventional oxidation routes, treatment in acidic media [20–22]. In contrast, most metal electrode
featuring numerous advantages: (i) no harmful oxidizers are required, has been known to exhibit transient dissolution during oxidation po-
as air or water is employed as an oxygen donor; (ii) oxidation is tential larger than 1.5 V (vs. RHE) [23]. Previous work showed that a

Corresponding author.
⁎

E-mail addresses: [email protected] (B. Seo), [email protected] (W.H. Lee), [email protected] (Y.J. Sa), [email protected] (U. Lee),
[email protected] (H.-S. Oh), [email protected] (H. Lee).

https://doi.org/10.1016/j.apsusc.2020.147517
Received 29 April 2020; Received in revised form 24 July 2020; Accepted 11 August 2020
Available online 19 August 2020
0169-4332/ © 2020 Published by Elsevier B.V.
B. Seo, et al. Applied Surface Science 534 (2020) 147517

functionalized graphite rod (GR) anode results in higher production working and reference electrodes, respectively, while Pt wire was used
yield (44.6% at 90 °C) compared to that for Pt anode (32%) in the as the counter electrode (cathode). Anode compartments was filled with
electro-oxidation of toluene into benzoic acid [24]. Wendt et al. tried to 0.1 M LiClO4 in a mixed solution of CH3CN (45 mL) and hexafluoro-2-
minimize further oxidation of benzaldehyde to benzoic acid by forming propanol (HFIP, 5 mL). Cathode compartment was charged with 0.5 M
dimethyl acetal intermediate in methanol and ethanol media at a H2SO4 (50 mL). Working and reference electrodes were immersed in the
carbon anode [25,26]. However, it produced numerous unidentified by- anode electrolyte. Counter electrode was immersed in the cathode
products simultaneously due to product functionalization with methoxy electrolyte. Before measurements, electrochemical impedance spectra
and ethoxy groups, and toluene polymerization. were recorded at 0.5 V (vs. Ag/AgCl) and an AC potential amplitude of
In this study, we advanced the electrochemical conditions by 10 mV from 10 kHz to 1 Hz. Constant-current electrolysis was con-
modifying cell configuration, electrolyte, solvent, electrode material, ducted at applied currents of 5, 10, 15, 20, 25, or 30 mA for various
and applied current for precise selectivity control, revealing that the reaction times. Long-term stability tests were conducted by 24-h con-
selectivity of the toluene oxidation reaction (TOR) strongly depends on stant-potential electrolysis at 2 V (vs. RHE).
the applied current/potential and anode material. After optimizing the
current, three kinds of carbon electrodes, GR, glassy carbon plate 2.4. Cdl measurement
(GCP), and carbon paper (CP) were tested for the TOR. GR showed the
higher selectivity to benzaldehyde, while CP produced benzoic acid as a Electrochemically active surface area can be estimated from the
major product. In addition, GR showed good long-term drivability electrochemical double-layer capacitance (Cdl) of the catalytic surface
under oxidation potential (~2 V vs. RHE) in acidic media. The reason [28]. The Cdl was determined from cyclic voltammograms (CVs) re-
for its high selectivity to benzaldehyde and stability was scrutinized corded at different scan rates in the non-faradaic potential region where
through near-edge X-ray absorption fine structure, X-ray photoelectron, all measured currents were assumed to be associated with double-layer
Fourier transform infrared, 1H nuclear magnetic resonance, and Raman charging. The potential range was typically centered at the open-circuit
spectroscopy analyses, which has been underexplored previously. In potential (OCP) with a potential window of 0.1 V. CVs were measured
addition, the rate-determining step in the direct electrochemical TOR in a potential range of (OCP − 0.05 V to OCP + 0.05 V) at scan rates of
over the GR electrode was first gained by Tafel analysis. Finally, the v = 10, 20, 40, 60, 80, 100, 150, and 200 mV s−1. The measured
developed method was successfully extended to the oxidation of other charging current (ic) was equal to the product of v and Cdl (ic = vCdl),
para-substituted toluenes, namely p-xylene, 4-chlorotoluene, and 4- and the slope of the ic vs. v plot therefore equal to Cdl.
methylanisole, and was concluded to offer a promising way of con-
trolling selectivity and simplifying the preparation of high-value alde- 2.5. Product analysis
hyde derivatives.
Toluene and its oxidation products were quantified by high-per-
2. Experimental formance liquid chromatography (HPLC) measurements. Toluene con-
version yields (%), product yields (%), selectivities (%), and Faradaic
2.1. Physicochemical characterization efficiencies were calculated as follows.
mol of toluene consumed
Scanning electron microscopy (SEM) imaging and quantitative en- Conversion yield (%) = × 100%
mol of initial toluene
ergy dispersive X-ray spectroscopy analysis with elemental mapping
were performed using a field emission-SEM (FE-SEM) microscope mol of product
Production yield (%) = × 100%
(Teneo VS, FEI). X-ray photoelectron spectroscopy measurements were mol of initial toluene
performed on an PHI 5000 VersaProbe instrument (Ulvac-PHI, Japan)
equipped with a monochromatic Al Kα X-ray source (1486.6 eV). O 1s mol of BzH or BzOH
Selectivity (%) = × 100%
and C 1s spectra were deconvoluted using the mixed (Gaussian 70, mol of BzH + mol of BzOH
Lorentzian 30)-function after Shirley-type background subtraction. The mol of BzH formed
full-width-at-half-maximum (FWHM) values for the major peaks were Faradaic efficiency for BzH (%) =
total charge passed ÷ (F × 4)
constrained to 1.6 eV. The Raman spectra of carbon anodes were re-
corded on a Renishaw In Via Raman microscope using a 50-mW 532-nm × 100%
laser for excitation. mol of BzOH formed
Faradaic efficiency for BzOH (%) =
total charge passed ÷ (F × 6)
2.2. NEXAFS analysis
× 100%
Near-edge X-ray absorption fine structure (NEXAFS) analysis was Here, F is Faradaic constant. Selectivity was calculated considering
performed at the 10D beamline of the Pohang Accelerator Laboratory only the mainly produced benzaldehyde (BzH) and benzoic acid (BzOH)
(PAL) in the Republic of Korea using an electron beam energy and dissolved in the reaction medium, as other (minor) aromatic com-
current of 3.0 GeV and 300 mA, respectively. A Si(1–1–1) double- pounds were poorly detectable by GC and LC. To calculate toluene
crystal monochromator was used to filter incident photon energy. O and conversion yields, 10-µL solution aliquots were sampled from the anode
C K-edge NEXAFS experiments for carbon anodes were conducted in compartment before and after the TOR and diluted with 490 µL of
fluorescence detection mode. Background removal and normalization acetonitrile for HPLC analysis (YL9100 HPLC system containing an
of the spectra were carried out using the IFEFFIT (Athena) software ultraviolet–visible detector set at 265 nm). Acetonitrile (100%) was
[27]. used as the mobile phase in isocratic mode with a flow rate of
0.5 mL min−1 at 35 °C. To calculate BzH and BzOH production yields,
2.3. Electrochemical characterization 10-µL solution aliquots were sampled from the anode compartment
before and after the TOR and diluted with 490 µL of a mixed solution
Electrochemical experiments were performed using an IviumStat (acetonitrile : 5 mM potassium phosphate pH 3.0 in water with phos-
electrochemical analyzer at room temperature (25 °C) and atmospheric phoric acid) for HPLC analysis (YL9100 HPLC system containing an
pressure in a three-electrode configuration using a H-cell separated ultraviolet–visible detector set at 235 nm). 5 mM potassium phosphate
with a Nafion membrane (Nafion® 117). A carbon electrode (GR, GCP, pH 3.0 in water with phosphoric acid (A) and acetonitrile (B) were used
or CP) and Ag/AgCl (3 M NaCl filling solution) were used as the as mobile phase in gradient mode: 40 to 80% B in 9 min; hold at 80% B

2
B. Seo, et al. Applied Surface Science 534 (2020) 147517

Fig. 1. (a) Reaction scheme for non-mediated direct electro-oxidation of toluene. (b) Chronopotentiometric curves measured at applied currents of 5–30 mA. (c)
Conversion yields (black scatters), production yields (vertical bars), and selectivities for BzH (blue scatters) after 4-h TOR at different applied currents. Red and green
bars indicate BzH and BzOH production yields, respectively. (d) FE obtained after 4-h TOR at different applied currents. Black and blue scatters indicate FE for BzH
and BzOH, respectively.

for 2 min; to 40% B in 0.1 min; hold at 40% B for 6.9 min with a flow First of all, the applied current was screened to optimize selectivity
rate of 0.5 mL min−1 at 30 °C. Aliquots of the diluted samples (10 µL) toward BzH by conducting chronopotentiometric measurements at a
were injected into a 4.6 mm × 150 mm C18 column (5 µm). Product constant current of 5, 10, 15, 20, 25, or 30 mA, which corresponded to
identities and concentrations were determined from calibration curves oxidation potentials (E) of ~2.0, 2.1, 2.2, 2.26, 2.32, or 2.35 V (vs.
obtained using standard solutions of known concentrations. The pro- RHE), respectively (Fig. 1b). Quantitative analysis of products by high-
ducts were further identified by GC–MS (HP-6890 GC coupled with an performance liquid chromatography (HPLC) performed after 4 h con-
Agilent 5973 mass spectrometer equipped with an HP-5 column stant-current chronopotentiometry shows the resulting conversion,
(30 m × 250 μm × 0.25 μm). He (99.9997%) was used as the carrier production yields, and selectivities toward BzH (Fig. 1c). The toluene
gas. conversion was linearly increased from 5.5% to 23.8% according to
current increase from 5 mA to 30 mA. Production yield for BzH in-
creases from 5.1% to 12% as the current increases from 5 mA to 15 mA,
3. Results and discussion
however, from 15 mA to 30 mA, the yield slightly decreases to around
10.7% and the yield for BzOH increases from 1.0% to 13.1%. Gas
3.1. Electrochemical toluene oxidation
chromatography-mass spectrometry (GC–MS) analysis showed that BzH
was the dominant product of TOR at 15 mA for 4 h, in agreement with
The electrochemical toluene oxidation reaction (TOR) was per-
HPLC results (Fig. S1). Noticeable side products were not observed in
formed using an H-type electrochemical cell equipped with Nafion
both HPLC and GC–MS chromatograms.
membrane as a separator, and graphite rod (GR) and Ag/AgCl were
Fig. 1d shows corresponding faradaic efficiency (FE) for the BzH
used as the anode and reference electrode, respectively. As an electro-
and BzOH. The FE for BzH decreases from 80.1% (at 5 mA) to 26.9% (at
lyte solvent, the mixture of acetonitrile/HFIP was used, where HFIP was
30 mA), while, the FE toward BzOH increases with increasing applied
known to stabilize benzaldehyde from over oxidation to benzoic acid by
current from 9.4% (at 5 mA) to 49.3% (at 30 mA). Low current of 5 mA
forming hydrogen bond (Fig. 1a) [12]. To allow for an efficient HER
can guarantee high selectivity and FE but slow conversion rate.
counter reaction, Pt wire and 0.5 M H2SO4 were used as the cathode
Therefore, subsequent experiments were conducted using a mild ap-
and catholyte, respectively. Overall, toluene (C6H5CH3) was oxidized to
plied current of 10 mA to maintain high selectivity toward BzH.
benzaldehyde (BzH) or benzoic acid (BzOH) at the anode, and hydrogen
evolved at the cathode.
Anode reaction (TOR): 3.2. Effect of carbon electrode and reaction time
+ −
C6H5CH3 + H2O → C6H5CHO + 4H + 4e Similar to the applied current, the anode material also strongly af-
Cathode reaction: fected charge transfer at the electrode-toluene interface. Three kinds of
commercial carbon graphite rod (GR), glassy carbon plate (GCP), and
4H+ + 4e− → 2H2 (g), HER carbon paper (CP) with different surface morphologies were tested as
anode materials [29]. Scanning electron microscopy (SEM) images
Full-cell four-electron reaction:
(Fig. 2a) showed that GR had a sheet-like morphology, while GCP had a
C6H5CH3 + H2O → C6H5CHO + 2H2↑ flat surface, and CP featured entangled carbon fibers and a smooth

3
B. Seo, et al. Applied Surface Science 534 (2020) 147517

Fig. 3 shows the TOR performances of three carbon anodes mea-

sured by conducting electrolysis at 10 mA. After 4-h electrolysis, CP
showed the highest toluene conversion of 10.6%, while those of GR and
GCP were 7.9% and 4.4%, respectively. Benzaldehyde (BzH) was ob-
tained as the major product at the three electrodes; the yields were 7.3,
3.0, 8.7% with selectivities of 92.4, 68.2, and 82.1% for GR, GCP, and
CP, respectively (Fig. 3b). Because the high selectivity to BzH at GR
electrode is highly encouraging, by the longer reaction time, higher
yield of BzH was expected.
When the electrolysis was maintained during 12 h, the electrodes
showed notable TOR performance differences in terms of production
yield and selectivity (Fig. 3c,d). GCP showed a negligible increase in
conversion and production yields compared to those observed after 4-h
electrolysis, while increases in toluene conversion were observed for GR
and CP up to 35% (Fig. 3c). GR produced BzH with a yield of 17.2%,
while CP mainly produced benzoic acid (BzOH) with 26.3% yield.
For GR, a large gap between conversion and production yields
(hatched area in Fig. 3c) was observed. Prior to discussing the structural
Fig. 2. (a) SEM images and (b) Cdl measurements of used carbon electrodes.
factors responsible for high selectivity toward BzH in detail, we scru-
tinized the origin of the gap between conversion and production yields
surface. Their different structural properties were further identified and found the crossover of BzH through Nafion membrane was one of
with double-layer capacitance (Cdl) measurements (Fig. 2b). They the reason. When analyzing the solution in cathode compartment after
showed different electrochemically active surface area (ECSA), which is 12-h TOR, around 6% of synthesized benzaldehyde was delivered to
proportional to the Cdl measured in the non-faradaic region [28]. Cdl cathode part through the Nafion membrane. More direct reason of the
decreased in the order of GR (1.4 mF) > GCP (0.054 mF) > CP gap could be found from the adsorption of BzH on the GR electrode,
(0.0037 mF) (Fig. 2b). The high ECSA of GR was expected to provide which was verified from the various spectroscopic analyses of GR
high TOR performance, but CP was the electrode that showed the electrode after the reaction.
highest toluene conversion after 4-h electrolysis. Fig. 4 shows the changes of GR electrode before and after TOR at the

Fig. 3. (a–d) TOR performances of different carbon electrodes at 10 mA after (a,b) 4-h and (c,d) 12-h electrolysis; (a,c) conversion and production yields, and (b,d)
selectivities. Black scatters represent toluene conversion yield. Red and green boxes indicate BzH and BzOH production yields, respectively.

4
B. Seo, et al. Applied Surface Science 534 (2020) 147517

Fig. 4. (a) O K-edge and (b) C K-edge NEXAFS spectra of GR before and after TOR. (c) O 1 s and (d) C 1 s X-ray photoelectron spectra of GR before and after TOR.

near-edge X-ray absorption fine structure (NEXAFS) spectroscopy and Nuclear magnetic resonance (NMR) spectroscopy further confirmed
X-ray photoelectron spectroscopy (XPS). The O K-edge NEXAFS spectra that BzH was adsorbed on the GR electrode along with a small amount
of GR before TOR featured three peaks at 536.6, 538.0, and 547.1 eV of other unknown aromatic compounds and polymeric compounds
corresponding to CeO (π*), OeH (π*), and C]O (σ*) functionalities, which might be formed from the absorbed BzH (Fig. S4). While, char-
respectively (Fig. 4a) [30,31]. After TOR, C]O (π*) at 533.8 eV and acteristic peak of BzOH (at ~12 ppm) was not observed.
C]O (σ*) at 547.1 eV gained intensity, possibly because of the ad- The absorption of BzH was only observed on GR during electrolysis.
sorption of carbonyl compounds on the O-functionalized carbon surface When carbon electrodes were immersed in BzH solution in the presence
[20–22]. In C K-edge NEXAFS spectra, signals of aromatic/olefinic C]C and absence of an applied current for 12 h. In the presence of an applied
(285.0 eV) and C]O (287.9 eV) functionalities lost and gained in- current, 58% of added BzH was adsorbed onto GR (Fig. S5a), while the
tensity after the TOR, respectively (Fig. 4b) [30–32]. XPS analysis adsorbed amount was negligible for GCP and CP (Fig. S5b,c). In the
further supported the NEXAFS results (Fig. 4c,d). Detailed peak para- absence of an applied current, BzH was hardly adsorbed on any of the
meters are described in Table S1. The deconvoluted O 1 s spectra of GR carbon electrodes (Fig. S5d–f), suggesting that the surface of GR was
showed three peaks at 531.6, 532.6, and 533.5 eV corresponding to C] modified during electrolysis to facilitate BzH adsorption [20–22]. Ad-
O, CeOeC/OeC]O, and CeO(H) moieties, respectively (Fig. 4c) ditionally, we confirmed that toluene was not adsorbed on the carbon
[33–35]. The deconvoluted C 1 s spectra of GR showed five peaks electrodes in the absence of an applied current (Fig. S6).
corresponding to C]C (284.6 eV), CeO(H) (285.4 eV), C]N When the amount of adsorbed BzH was considered, the selectivity
(286.7 eV), C]O (288.5 eV), and eOeC]O (290.9 eV) species toward BzH decreased in the order of GR (83%) > GCP (68%) > CP
(Fig. 4d) [34–38]. The C]O peak gained intensity after the TOR, while (23%) (Fig. 3d), which was explained by different TOR oxidation po-
the intensity of the eOeC]O peak remained unchanged, which was tentials. In the case of GR, a potential of ~2.1 V was required to drive a
indicative of negligible benzoate formation on the carbon electrode. current of 10 mA, while CP needed a larger potential of ~2.3 V to drive
Similarly, the X-ray photoelectron spectra of GCP and CP demonstrated the same current. The higher oxidative potential required in the latter
that the content of C]O groups on the carbon surface increased after case could be the reason of further oxidation to BzOH. In the case of
the TOR (Fig. S2). GCP, the much larger oxidation potential of ~3.5 V required to main-
Fourier transform infrared spectra acquired before and after TOR tain the current of 10 mA induced over-oxidation of BzOH to unknown
further demonstrated the adsorption of BzH on GR. Peaks at 1728 and products poorly detectable by LC and GC.
1580 cm−1 were attributed to the stretching vibrations of C]O bonds
in BzH and C]C bonds in GR, respectively (Fig. S3) [39,40]. Notably,
no BzOH peak (~1754 cm−1) was observed for GR after TOR [40].

5
B. Seo, et al. Applied Surface Science 534 (2020) 147517

3.3. Stability of carbon electrodes indicated by the fact that no BzH absorption was monitored in short
reaction less than 5 h (Fig. 3a). For GCP (Fig. S7b), decrease of current
Although 4-h TOR shows an attractive result for the synthesis of after 5 h seems due to the destruction of carbon materials. This result
benzaldehyde (BzH), the result of 12-h reaction revealed structural agrees well with TOR experiments; no further reaction was proceeded
modification at the anode material has happened during the electro- after 4-h reaction at GCP-electrode system (Fig. 3a,c). For CP, there was
lysis. It was reported that the carbon materials are oxidized under the no significant change in current profile (Fig. S7c), indicating the
highly oxidative potential and degraded to CO2 or other organic com- moderate stability of CP at the oxidation potential. However, because of
pound [41]. Furthermore, as the reaction proceeded, the pH of anode the small portion of defect sites in CP (ID/IG = 0.17), the selectivity to
electrolyte decreased due to the crossover of acid solution from the benzaldehyde is relatively low compared to GR (Fig. 3d).
cathode side, which can facilitate the destruction of carbon anode. To As mentioned above, crossover of catholyte (0.5 M H2SO4) to anode
investigate the stability of the carbon electrode under the reaction en- side can affect TOR as well as electrode corrosion. During the electro-
vironment, long-term electrolysis was conducted in the absence of to- lysis, we found water in anode reactor increased to 10 wt% after 4-h
luene in pH 2 condition at an oxidation potential of ~2 V (vs. RHE) reaction, which could act as an oxygen source not only for TOR
(Fig. S7a–c). Fig. S7a shows, at GR, the currents began to increase (C6H5CH3 + H2O → C6H5CHO + 4H+ + 4e−) and also for oxygen
gradually after 5-h steady state, indicating the oxidation of carbon evolution reaction (2H2O → O2 + 4H+ + 4e–). However, adding a
material started after 5-h electrolysis. The increase of current was also larger amount of water (10 mmol, 4 equiv.) to the anode part didn’t
observed at GCP from the start of the reaction, but it decreased after affect the reaction as shown in Fig. S8; the changes in conversion yield
5 h. While, at CP, no distinctive increase of current was observed during and selectivity were negligible.
24-h reaction. According to previous reports, an anodic treatment in
acidic media fractures graphitic carbon surfaces, creating a porous and 3.4. Rate-determining step
rough surfaces and providing benefit to improve the absorptivity and
reaction efficiency [20–22]. Among the tested electrodes, only the GR To gain insights into the TOR mechanism, we analyzed Tafel plots
electrode is continuously activated and durable under the high oxida- obtained for different carbon anodes (Fig. S9) by performing constant-
tion potential in an acidic environment. current electrolysis at a series of applied currents (5–25 mA) (Fig.
This stability of carbon electrode can be elucidated with D-to-G S9a–c). The Tafel slope, extracted from the linear portion of the Tafel
band intensity ratio (ID/IG) observed in Raman spectra (Fig. 5). It was plot, increased in the order of GR (75 mV dec−1) < GCP (88 mV
reported that defects in the graphitic lattice with A1g symmetry (re- dec−1) < CP (360 mV dec−1) (Fig. S9d–f).
ferred to as the D-band) appeared at 1348 cm−1, while in-plane E2g Next, we identified the rate-determining step (rds) for benzaldehyde
mode of single-crystalline graphitic carbon atoms in the honeycomb and benzoic acid production on the GR electrode by extracting the
lattice (referred to as the G-band) appeared at 1580 cm−1 [42]. The ID/ partial current for benzaldehyde (iBzH) and benzoic acid (iBzOH) pro-
IG is regarded as an indicator of disorder in carbon materials. Fig. 5 duction from the total current while considering current-dependent
shows the ID/IG of three carbon electrodes decreased in the order of faradaic efficiency (Fig. 6a,b). The similar Tafel slopes for benzalde-
GCP (1.78) > GR (1.01) > CP (0.17). As discussed above, oxidative hyde (72 mV dec−1) and benzoic acid (68 mV dec−1) indicated that the
treatment in acidic media could modulate graphitic carbon surface, corresponding reaction kinetics were similar. The high selectivity to-
resulting in increase of defect sites on the graphite surface [20–22]. In ward benzaldehyde mainly originated from the low onset potential
the TOR, the defect site formed on GR seems acts not only as active site required to drive the current for benzaldehyde compared with that
for toluene oxidation but also as a benzaldehyde absorption site as required for benzoic acid. According to Tafel plots (Fig. 6a,b), poten-
tials of 1.90 and 1.98 V were required to drive a partial current of 1 mA
for benzaldehyde and benzoic acid, respectively.
From the Tafel slopes, rds was inferred using the simplified equation
reported by Nagy and Bockris [43]:
1
E n s RT
(Tafel slope) = r
Logia v F

Here, n is the total number of transferred electrons for a single re-

action, s is the total number of electrons transferred before the rds, v is
the number of times rds has to occur for the overall reaction to occur
once (stoichiometric number of the reaction mechanism), r is the
number of electrons transferred in the rds (r = 0 and r = 1 indicate
chemical and electron-transfer rds, respectively), and = 0.5 is the
symmetry factor. The Tafel slopes of close to 60 mV dec−1 observed for
both benzaldehyde (n = 4) and benzoic acid (n = 6) corresponded to
integer combinations of s = 3, v = 1, r = 0 and s = 5, v = 1, r = 0,
respectively (at 298 K). These numbers suggest that the rds in the
electrochemical TOR corresponds to chemical rearrangement prior to
final electron transfer (Fig. 6c,d) [43].

3.5. Direct oxidation of other methylarenes

The direct electro-oxidation method was applied to different me-

thylarenes (Table 1). Under optimized conditions (applied cur-
rent = 10 mA, GR anode), para-substituted methylarenes were suc-
cessfully oxidized to the corresponding aldehydes, as confirmed by
HPLC and GC–MS analyses (Fig. S10 and S11). The selectivity of 4-
Fig. 5. Raman spectra of three carbon electrodes. methoxybenzaldehyde production (97%) exceeded that of 4-

6
B. Seo, et al. Applied Surface Science 534 (2020) 147517

Fig. 6. Tafel plots of TOR at GR: iR-corrected potential vs. logarithmic partial current graphs for (a) BzH and (b) BzOH. Mechanism of toluene oxidation to (c) BzH
and (d) BzOH. The rds is depicted by a green dashed arrow, corresponding to chemical rearrangement before final electron transfer.

Table 1 xylene proceeded with intermediate selectivity for p-tolualdehyde

Conversion yields (%), production yields (%), and selectivities (%) of sub- (62%). Notably, ortho-substituted methylarenes were not converted to
stituted toluene electro-oxidation. the corresponding aldehydes, possibly because functional groups in the
Materials Conversion Production Selectivity (%) ortho position hindered effective charge transfer between methyl groups
yield (%) yield (%) and the carbon anode. The observed conversion/production yields and
selectivities are summarized in Table 1. A comparison of state-of-the-art
16.7 10.4 62
records for benzylic CeH bond oxidation clearly showed that the de-
veloped direct electrochemical TOR showed promising benzaldehyde
11.3 5.7 50
yield and selectivity (Table S2). Best of all, this approach doesn’t use
any mediators and metal redox catalysts, thus reducing separation
costs.
4.1 4.0 97

4. Conclusion

No conversion! Toluene can be electrochemically oxidized to benzaldehyde with

high selectivity in the absence of organic mediators and metal redox
catalysts. The selectivity of electrochemical TOR was strongly affected
by applied current and carbon anode material. In an optimized current
of 10 mA for 4-h electrolysis, toluene was oxidized to benzaldehyde
with a high selectivity of 92.4% and 82.1% over graphite rod and
carbon paper electrode, respectively. Notably, longer reaction time
resulted in the formation of defect site on graphite rod, where the
chlorobenzaldehyde production (50%), which was ascribed to the ac-
synthesized benzaldehyde could be favorably adsorbed. The high se-
tivation/deactivation of toluene toward anodic attack by electron-do-
lectivity of the graphite rod electrode to benzaldehyde attributes to the
nating/-withdrawing groups, respectively [25,26]. Oxidation of p-
large number of defect sites responsible for the relatively low oxidation

7
B. Seo, et al. Applied Surface Science 534 (2020) 147517

potential required to drive current for benzaldehyde production com- electrochemical oxidation of toluene to benzaldehyde using Ce3+/Ce4+ as med-
pared to that needed for benzoic acid production. In addition, the non- iator, J. Appl. Electrochem. 18 (1988) 298–303.
[15] E.J. Horn, B.R. Rosen, Y. Chen, J. Tang, K. Chen, M.D. Eastgate, P.S. Baran, Scalable
mediated direct electron transfer method was successfully extended to and sustainable electrochemical allylic C-H oxidation, Nature 533 (2016) 77–81.
other para-substituted methylarenes. The obtained results allow one to [16] Y. Kawamata, M. Yan, Z. Liu, D.-H. Bao, J. Chen, J.T. Starr, P.S. Baran, Scalable,
simplify the preparation of high-value aldehyde derivatives and avoid electrochemical oxidation of unactivated C-H bonds, J. Am. Chem. Soc. 139 (2017)
7448–7451.
complex synthetic procedures. [17] D.P. Hruszkewycz, K.C. Miles, O.R. Thiel, S.S. Stahl, Co/NHPI-mediated aerobic
oxygenation of benzylic C-H bonds in pharmaceutically relevant molecules, Chem.
CRediT authorship contribution statement Sci. 8 (2017) 1282–1287.
[18] M. Rafiee, F. Wang, D.P. Hruszkewycz, S.S. Stahl, N-hydroxyphthalimide-mediated
electrochemical iodination of methylarenes and comparison to electron-transfer-
Bora Seo: Conceptualization, Investigation, Writing - original draft. initiated C−H functionalization, J. Am. Chem. Soc. 140 (2018) 22–25.
Woong Hee Lee: Investigation, Writing - review & editing. Young Jin [19] J.E. Nutting, M. Rafiee, S.S. Stahl, Tetramethylpiperidine N-oxyl (TEMPO), phtha-
limide N-oxyl (PINO), and related N-oxyl species: electrochemical properties and
Sa: Investigation, Writing - review & editing. Ung Lee: Writing - review
their use in electrocatalytic reactions, Chem. Rev. 118 (2018) 4834–4885.
& editing. Hyung-Suk Oh: Writing - review & editing. Hyunjoo Lee: [20] R. Bowling, R.T. Packard, R.L. McCreery, Mechanism of electrochemical activation
Supervision. of carbon electrodes: role of graphite lattice defects, Langmuir 5 (1989) 683–688.
[21] H. Xu, X. Fan, Y. Lu, L. Zhong, X. Kong, J. Wang, Preparation of an electro-
chemically modified graphite electrode and its electrochemical performance for
Declaration of Competing Interest pseudo-capacitors in a sulfuric acid electrolyte, Carbon 48 (2010) 3300–3303.
[22] M. Rueffer, D. Bejan, N.J. Bunce, Graphite: an active or inactive anode?
Electrochim. Acta 56 (2011) 2246–2253.
The authors declare that they have no known competing financial [23] S. Cherevko, A.R. Zeradjanin, A.A. Topalov, N. Kulyk, I. Katsounaros,
interests or personal relationships that could have appeared to influ- K.J.J. Mayrhofer, Dissolution of noble metals during oxygen evolution in acidic
ence the work reported in this paper. media, ChemCatChem 6 (2014) 2219–2223.
[24] Y. Zhu, B. Wang, H. Wang, X. Liu, S. Licht, Towards efficient solar STEP synthesis of
benzoic acid: role of graphite electrode, Sol. Energy 113 (2015) 303–312.
Acknowledgements [25] H. Wendt, S. Bitterlich, Anodic synthesis of benzaldehydes—I. voltammetry of the
anodic oxidation of toluenes in non-aqueous solutions, Electrochim. Acta 37 (1992)
1951–1958.
This work was supported by the program of the Korea Institute of
[26] H. Wendt, S. Bitterlich, E. Lodowicks, Z. Liu, Anodic synthesis of benzaldehydes—II.
Science and Technology (KIST) (2E29482) and C1 Gas Refinery optimization of the direct anodic oxidation of toluenes in methanol and ethanol,
Program through the National Research Foundation (NRF) of Korea Electrochim. Acta 37 (1992) 1959–1969.
[27] B. Ravel, M. Newville, ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray
(NRF-2015M3D3A1A01065435). The XAS experiments performed at
absorption spectroscopy using IFEFFIT, J. Synchrotron Radiat. 12 (2005) 537–541.
Beamline 10D of the Pohang Accelerator Laboratory (PAL) were sup- [28] C.C.L. McCrory, S. Jung, J.C. Peters, T.F. Jaramillo, Benchmarking heterogeneous
ported in part by the Ministry of Education, Pohang University of electrocatalysts for the oxygen evolution reaction, J. Am. Chem. Soc. 135 (2013)
Science and Technology (POSTECH). 16977–16987.
[29] A.G. Pandolfo, A.F. Hollenkamp, Carbon properties and their role in super-
capacitors, J. Power Sources 157 (2006) 11–27.
Appendix A. Supplementary material [30] S.G. Urquhart, H. Ade, Trends in the carbonyl core (C 1s, O 1s) → π*C=O transition
in the near-edge X-ray absorption fine structure spectra of organic molecules, J.
Phys. Chem. B 106 (2002) 8531–8538.
Supplementary data to this article can be found online at https:// [31] K.G. Latham, M.I. Simone, W.M. Dose, J.A. Allen, S.W. Donne, Synchrotron based
doi.org/10.1016/j.apsusc.2020.147517. NEXAFS study on nitrogen doped hydrothermal carbon: insights into surface
functionalities and formation mechanisms, Carbon 114 (2017) 566–578.
[32] H. Yabuta, M. Uesugi, H. Naraoka, M. Ito, A.L.D. Kilcoyne, S.A. Sandford,
References F. Kitajima, H. Mita, Y. Takano, T. Yada, Y. Karouji, Y. Ishibashi, T. Okada, M. Abe,
X-ray absorption near edge structure spectroscopic study of Hayabusa category 3
[1] Y. Ishii, S. Sakaguchi, T. Iwahama, T. Innovation of hydrocarbon oxidation with carbonaceous particles, Earth, Planets Space 66 (2014) 156.
molecular oxygen and related reactions, Adv. Synth. Catal. 343 (2001) 393–427. [33] S. Wu, G. Wen, R. Schlögl, D.S. Su, Carbon nanotubes oxidized by a green method as
[2] R.A. Sheldon, H. Van Bekkum, Fine Chemicals through Heterogeneous Catalysis, efficient metal-free catalysts for nitroarene reduction, Phys. Chem. Chem. Phys. 17
Wiley, Hoboken, 2008. (2015) 1567–1571.
[3] F. Recupero, C. Punta, Free radical functionalization of organic compounds cata- [34] X. Li, Z. Zhao, C. Pan, Electrochemical exfoliation of carbon dots with the narrowest
lyzed by N-hydroxyphthalimide, Chem. Rev. 107 (2007) 3800–3842. full width at half maximum in their fluorescence spectra in the ultraviolet region
[4] K. Wallia, Benzaldehyde market | Industry overview report 2018–2025. Market, using only water as electrolyte, Chem. Commun. 52 (2016) 9406–9409.
Res. Reports (2019). [35] R. Hu, T. Furukawa, X. Wang, M. Nagatsu, Morphological study of graphite-en-
[5] W. Partenheimer, The high yield synthesis of benzaldehydes from benzylic alcohols capsulated iron composite nanoparticles fabricated by a one-step arc discharge
using homogeneously catalyzed aerobic oxidation in acetic acid, Adv. Synth. Catal. method, Appl. Surf. Sci. 416 (2017) 731–741.
348 (2006) 559–568. [36] I.-Y. Jeon, H.-J. Choi, M.J. Ju, I.T. Choi, K. Lim, J. Ko, H.K. Kim, J.C. Kim, J.-J. Lee,
[6] R. Sheldon, Metal-Catalyzed Oxidations of Organic Compounds: Mechanistic D. Shin, S.-M. Jung, J.-M. Seo, M.-J. Kim, N. Park, L. Dai, J.-B. Baek, Direct nitrogen
Principles and Synthetic Methodology Including Biochemical Processes, Elsevier, fixation at the edges of graphene nanoplatelets as efficient electrocatalysts for en-
Amsterdam, 2012. ergy conversion, Sci. Rep. 3 (2013) 2260.
[7] W.E. Truce, The Gattermann synthesis of aldehydes, Org. React. 9 (1957) 37–72. [37] M. Coroş, F. Pogăcean, M.-C. Roşu, C. Socaci, G. Borodi, L. Mageruşan, A.R. Biriş,
[8] A. Schoenberg, R.F. Heck, Palladium-catalyzed formylation of aryl, heterocyclic, S. Pruneanu, Simple and cost-effective synthesis of graphene by electrochemical
and vinylic halides, J. Am. Chem. Soc. 96 (1974) 7761–7764. exfoliation of graphite rods, RSC Adv. 6 (2016) 2651–2661.
[9] D.V. Gutsulyak, G.I. Nikonov, Chemoselective ruthenium-catalyzed reduction of [38] L. Zeng, X. Li, S. Fan, J. Li, J. Mu, M. Qin, L. Wang, G. Gan, M. Tadé, S. Liu, The
acid chlorides to aldehydes with dimethylphenylsilane, Adv. Synth. Catal. 354 bioelectrochemical synthesis of high-quality carbon dots with strengthened elec-
(2012) 607–611. tricity output and excellent catalytic performance, Nanoscale 11 (2019)
[10] B. Nammalwar, C. Fortenberry, R.A. Bunce, S.K. Lageshetty, K.D. Ausman, Efficient 4428–4437.
oxidation of arylmethylene compounds using nano-MnO2, Tetrahedron Lett. 54 [39] Z. Wang, J. Shi, D. Wang, Y. Pu, J.-X. Wang, J.-F. Chen, Metal-free catalytic oxi-
(2013) 2010–2013. dation of benzylic alcohols for benzaldehyde, React. Chem. Eng. 4 (2019) 507–515.
[11] Y. Yoshino, Y. Hayashi, T. Iwahama, S. Sakaguchi, Y. Ishii, Catalytic oxidation of [40] Z. Jia, X. Wang, F. Thevenet, A. Rousseau, Dynamic probing of plasma-catalytic
alkylbenzenes with molecular oxygen under normal pressure and temperature by N- surface processes: oxidation of toluene on CeO2, Plasma Process Polym. 14 (2017)
hydroxyphthalimide combined with Co(OAc)2, J. Org. Chem. 62 (1997) 1600114.
6810–6813. [41] W. Luc, J. Rosen, F. Jiao, An Ir-based anode for a practical CO2 electrolyzer, Catal.
[12] E. Gaster, S. Kozuch, D. Pappo, Selective aerobic oxidation of methylarenes to Today 288 (2017) 79–84.
benzaldehydes catalyzed by N-hydroxyphthalimide and cobalt(II) acetate in hexa- [42] M. Praveen Kumar, T. Kesavan, G. Kalita, P. Ragupathy, T.N. Narayanan,
fluoropropan-2-ol, Angew. Chem. Int. Ed. 56 (2017) 5912–5915. D.K. Pattanayak, On the large capacitance of nitrogen doped graphene derived by a
[13] Y. Zhu, B. Wang, X. Liu, H. Wang, H. Wu, S. Licht, STEP Organic Synthesis: An facile route, RSC Adv. 4 (2014) 38689–38697.
efficient solar, electrochemical process for the synthesis of benzoic acid, Green [43] J. O’M Bockris, Z. Nagy, Symmetry factor and transfer coefficient. A source of
Chem. 16 (2014) 4758–4766. confusion in electrode kinetics, J. Chem. Educ. 50 (1973) 839–843.
[14] J.-J. Jow, T.-C. Chou, Catalytic effects of the silver ion on the indirect