Journal of Energy Chemistry 25 (2016) 251–257
Contents lists available at ScienceDirect
Journal of Energy Chemistry
journal homepage: www.elsevier.com/locate/jechem
http://www.journals.elsevier.com/
journal-of-energy-chemistry/
Electrocatalytic hydrogen peroxide formation on mesoporous
non-metal nitrogen-doped carbon catalyst✩
Frédéric Hasché a,∗, Mehtap Oezaslan a,c, Peter Strasser a, Tim-Patrick Fellinger b
a
Technische Universität Berlin, Department of Chemistry, 10623 Berlin, Germany
Max Planck Institut für Kolloid- und Grenzflächenforschung, Department of Colloids, 14424 Potsdam, Germany
c
Carl von Ossietzky University of Oldenburg, Department of Chemistry, Electrocatalysis, 26111 Oldenburg, Germany
b
a r t i c l e
i n f o
Article history:
Received 2 November 2015
Revised 21 January 2016
Accepted 23 January 2016
Available online 3 February 2016
Keywords:
Electrochemical hydrogen peroxide
formation
Selectivity
Mesoporous nitrogen-doped carbon
Green synthesis
Electrocatalysis
Metal-free catalysis
a b s t r a c t
Direct electrochemical formation of hydrogen peroxide (H2 O2 ) from pure O2 and H2 on cheap metal-free
earth abundant catalysts has emerged as the highest atom-efficient and environmentally friendly reaction
pathway and is therefore of great interest from an academic and industrial point of view. Very recently,
novel metal-free mesoporous nitrogen-doped carbon catalysts have attracted large attention due to the
unique reactivity and selectivity for the electrochemical hydrogen peroxide formation [1–3]. In this work,
we provide deeper insights into the electrocatalytic activity, selectivity and durability of novel metal-free
mesoporous nitrogen-doped carbon catalyst for the peroxide formation with a particular emphasis on the
influence of experimental reaction parameters such as pH value and electrode potential for three different electrolytes. We used two independent approaches for the investigation of electrochemical hydrogen
peroxide formation, namely rotating ring-disk electrode (RRDE) technique and photometric UV–VIS technique. Our electrochemical and photometric results clearly revealed a considerable peroxide formation
activity as well as high catalyst durability for the metal-free nitrogen-doped carbon catalyst material in
both acidic as well as neutral medium at the same electrode potential under ambient temperature and
pressure. In addition, the obtained electrochemical reactivity and selectivity indicate that the mechanisms
for the electrochemical formation and decomposition of peroxide are strongly dependent on the pH value
and electrode potential.
© 2016 Science Press and Dalian Institute of Chemical Physics. All rights reserved.
1. Introduction
Direct synthesis of chemicals from its pure elements or bimolecules under ambient conditions often represents the most
atomically-economic reaction route. The understanding of the elemental processes as well as the identification of all relevant experimental and intrinsic parameters and steps for many (electro)
catalytic reactions are, however, highly complex and still poorly
understood to date. The electrochemical direct conversion of pure
hydrogen and oxygen to hydrogen peroxide is such a prominent
example. Hydrogen peroxide is among one of the 100 most important chemicals in the world [4], because it is used in a lot of different fields of chemical and chemistry-related industries e.g. for
pulp and paper bleaching, wastewater treatment or as “green” detergents and in different chemical syntheses [5]. Today, hydrogen
peroxide is industrially produced via the anthraquinone process in
✩
This work was supported by the Technische Universität Berlin, the Max Planck
Society and the Cluster of Excellence “Unifying Concepts in Catalysis (UniCat)”.
∗
Corresponding author.
E-mail address: [email protected] (F. Hasché).
large-scale plants [6,7]. Alternative and “on the place of use” located production methods are also discussed in the literature e.g.
membrane, fuel cell or plasma reactors [8–10].
In general, the reaction of pure hydrogen and oxygen leads to
the formation of two main products, namely water and/or hydrogen peroxide. Since the direct conversion of hydrogen and oxygen is strongly exothermic, electrochemical membrane reactors,
e.g. fuel cells, are typically employed to separate the two half-cell
reactions; that are the hydrogen oxidation reaction [11,12] on the
anode and the oxygen reduction reaction (ORR) on the cathode
side [8,13,14]. From the perspective of the oxygen electroreduction,
three reaction pathways are possible: (a) the direct two-electron
process to generate H2 O2 , (b) the overall four-electron process consisting of H2 O2 formation followed by consecutive reductive decomposition of H2 O2 to H2 O and finally (c) the direct four-electron
process for the formation of H2 O [15–17].
In fuel cell research, different cathode electrocatalyst concepts
are discussed in terms of activity and stability for the efficient
electrochemical conversion of oxygen. These concepts largely involve the utilization of metal nanoparticles dispersed on high surface area carbon support materials. The metal nanoparticles, such
http://dx.doi.org/10.1016/j.jechem.2016.01.024
2095-4956/© 2016 Science Press and Dalian Institute of Chemical Physics. All rights reserved.
252
F. Hasché et al. / Journal of Energy Chemistry 25 (2016) 251–257
as pure platinum or platinum alloys with e.g. Cu, Co, Ni largely
are the catalytically active reaction centers for the oxygen electroreduction. Improved ORR activities have been reported for dealloyed core-shell nanoparticle catalysts [18–23], Pt skin catalysts
[24,25], Pt monolayer catalysts [26,27] or non-noble metal catalysts, e.g. Fe/Co/N/C or nitrogen-doped carbons [28–33]. Recently,
pure nitrogen-doped carbon catalysts have attracted great attention due to their high performance for the electrochemical O2
reduction toward H2 O without usage of costly precious metals
[34,35]. On the other side, only a handful of electrocatalyst concepts for the electrocatalytic hydrogen peroxide production are
reported to date. Most of the H2 O2 electrocatalysts involve complexes (e.g. N-ligands like porphyrin or chlorin) containing metals
like Fe, Pd or Co [36,37] and supported metal (e.g. Pd, Au-Pd, PtHg, Pd-Hg) nanoparticles [2,36,38–44]. In comparison, typical Pt
nanoparticle fuel cell catalysts, polycrystalline Pt and single crystals of Pt show a H2 O2 yield of less than 5% in acidic environment
[22,45–47].
The long-term performance of electrocatalysts for the peroxide
formation strongly suffers from the reaction conditions, such as
aggressive H2 O2 -containing electrolyte, high potentials as well as
from heat and pressure. For instance, the decomposition of peroxide leads to the release of OH· and OOH· radicals, resulting in
an accelerated catalyst aging. A further challenge for the successful design of electrocatalysts is the selectivity toward H2 O2 at high
yields. In particular, the series reaction of H2 O2 toward H2 O via
two-electron transfer should be largely eliminated.
Recent results of theoretical studies suggest, that hydrogen peroxide only shows physisorption to N-doped graphene and that the
presence of protons facilitates O–O bond cleavage, i.e. peroxide
reduction [48]. Inspired by investigations on the electron transfer process of mesoporous metal-free nitrogen-doped carbon catalyst using ionic liquid N-butyl-3-methylpyridinium dicyanamide
(referred to as meso-BMP) in acidic and electrochemical environments, we have performed a comprehensive study on highly selective and active meso-BMP catalyst to highlight the influence of
pH values and nature of electrolyte (acid–HClO4 , alkaline–KOH and
neutral–KClO4 ) for the electrochemical formation of H2 O2 using
the rotating ring disk electrode (RRDE) technique. To support our
electrochemical results in terms of activity, selectivity and durability for H2 O2 formation, we used additionally an independent
photometric method to monitor the H2 O2 formation rate over the
reaction time at a certain applied potential. We show that the
H2 O2 formation rate is strongly influenced by the supporting electrolyte, pH values and applied voltage range. The observed difference in the reactivity-selectivity characteristics investigated on the
meso-BMP catalyst is largely related to the pH dependent mechanisms and kinetics of the electrochemical H2 O2 formation. More
importantly, the metal-free catalyst showed considerable stability
during the long-term RRDE experiments, highlighting a high resistance of the active catalyst in the presence of preformed peroxide.
2. Experimental
All chemicals
purification.
were
used
as
received
without
further
2.1. Synthesis of mesoporous nitrogen-doped carbon catalyst
Mesoporous nitrogen-doped carbon catalyst, referred to as
meso-BMP, was synthesized using ionic liquids like N-butyl3-methylpyridinium dicyanamide (BMP-dca, Ionic Liquide
Technologies GmbH, Germany) followed by hard-templating
using commercial silica nanoparticles (Ludox HS40).
2.2. Electrochemical measurement
All electrochemical measurements were conducted in a homemade three compartment electrochemical glass cell with a commercially available bipotentiostat (VSP-5, BioLogic, France) and
a rotator (PINE Instruments, USA). The three electrode arrangement consisted of a Pt mesh as counter electrode, a reversible
hydrogen electrode (Hydroflex HREF, Gaskatel, Germany) or
mercury/mercury sulfate (AMETEK GmbH, Germany) as reference
electrode and a glassy carbon electrode (PINE Instruments, USA,
diameter 5 mm, 0.196 cm2 geo. ) with a Pt ring (11 mm2 ) as working
electrode. For thin catalyst film RRDE preparation, ∼15 mg of catalyst powder was suspended in a mixture of 1.99 mL pure water
(18 MOhm cm at room temperature, Satorious), 0.5 mL 2-propanol
(Sigma-Aldrich) and 10 μL ionomer (Nafion solution, 5 wt%, SigmaAldrich). After horn sonification (Branson Sonifier 150D), an aliquot
of 10 μL was pipetted onto the mirror-like polished and cleaned
surface of a glassy carbon electrode and subsequently dried in air
for 10 min at 60 °C. All measurements were conducted at room
temperature.
2.3. Photometric peroxide measurement
The photometric peroxide investigations took place at 500 nm
by utilization of a peroxide test kit (Merck, #118789) as described
from the supplier. The voltage of the working electrode was 0.1 V
versus RHE with 1600 rotation per minute in a continuously O2
purged (∼250 mL/min) electrolyte solution. The electrolyte was
0.1 M KOH (prepared by dissolving of solid KOH pastilles, Sigma–
Aldrich), KClO4 (prepared by dissolving of solid KClO4 powder,
Sigma–Aldrich) or HClO4 (prepared by diluting from a 70% stock
solution, Sigma–Aldrich) solution. The photometric investigation is
described in the following: A small volume of electrolyte from
the reaction solution was taken and neutralized. Subsequently,
reagents as instructed for the commercial peroxide test (Merck,
#118789) were added. The color of the solution changed from colorless and clear into green/yellow corresponding to the peroxide
concentration. The colored sample was measured at 500 nm. According to the Lambert−Beer´s law, which describes a linear behavior of concentration and absorption, a fresh calibration curve
was established for each series of electrolyte measurements. The
peroxide concentrations were subsequently determined based on
the calibration curve.
3. Results and discussion
3.1. Metal-free mesoporous nitrogen-doped carbon catalyst
The metal-free mesoporous nitrogen-doped carbon (meso-BMP)
catalyst was synthesized by pyrolysis of ionic liquids and contains a relatively high nitrogen content of around 17 wt% obtained
from the element analysis. The nitrogen adsorption BET (BrunauerEmmett-Teller) surface area of the meso-BMP is around 320 m2 /g.
The total pore volume derived from Non Localized Density Functional Theory (NLDFT) calculation is 0.749 cm3 /g, with a mesopore
volume of 0.09 cm3 /g. It is noted that no metal was used or added
during the synthesis. Element analysis, X-ray photoelectron spectroscopy (XPS) and cyclic voltammograms (Fig. S1) do not hint on
any metals inside the meso-BMP material.
Fig. 1(a) shows a XPS spectrum of the meso-BMP catalyst.
Based on the multi-peak fit analysis, the nitrogen N 1 s signal
indicates the presence of pyridinic (44% at 398.4 eV), pyrrolic
(17% at 399.8 eV), graphitic (33% at 401.0 eV) and a few N-oxidic
(6% at 402.9 eV) nitrogens. The theoretically predicted order of
reactivity is pyridinic > quaternary > pristine graphene for the
F. Hasché et al. / Journal of Energy Chemistry 25 (2016) 251–257
253
Fig. 1. Structural characterization of metal-free nitrogen-doped carbon (meso-BMP) catalyst. (a) X-ray photoelectron spectroscopy (XPS) measurement (blue line) and the
corresponding multi-peak fit analysis for various nitrogen-based components and (b) high-resolution TEM micrograph with scale bar of 10 nm.
reduction of hydrogen peroxide to water via a two electron reaction mechanism [48]. Compared to the majority of literature our
nitrogen-doped carbon shows a relatively high nitrogen content,
with pyridinic nitrogen being the main species, although theoretically even more nitrogen-rich compounds such as carbon nitrides
are thermodynamically stable up to 800 °C [49]. In Fig. 1(b), the
TEM micrograph shows the long-range ordered structure of the
meso-BMP. For synthesis of nitrogen-doped carbon, ionic liquid
N-butyl-3-methylpyridinium dicyanamide (BMP-dca) was used as
direct precursor. The high meso- and macroporosity of this material was introduced via a hard-templating (nanocasting) synthesis
strategy using commercial colloidal silica nanoparticles (Ludox
HS-40). More details about the synthesis procedure, structural
and chemical investigations of the meso-BMP catalyst are given in
reference [1,50,51].
3.2. Influence of pH value and nature of electrolyte on the
electrocatalytic hydrogen peroxide formation
The electrochemical experiments were carried out in a three
electrode setup at room temperature at ambient pressure in
0.1 M electrolyte concentration of HClO4 (pH 1), KOH (pH
13) and KClO4 (pH 7), respectively, to evaluate the influence
of the pH values and nature of electrolyte on the peroxide
formation and solvent-based stabilization. The investigations of pH
on the peroxide formation is of great interest because in particular theoretical prediction suggests that the presence of protons enhances the O–O bond cleavage, which should lead to reduced activities/efficiencies in the peroxide synthesis [48]. Therefore, rotating
ring disk electrode (RRDE) experiments were conducted to determine the rate of peroxide formation and its (electro) chemical stability over the reaction time at a constant voltage (chronoamperometry) as well as during a linear sweep voltammetry. The preparation of the working electrode with a uniform thin catalyst film
for electrochemical measurements was conducted as follows: The
catalyst dispersion was prepared by mixing of catalyst powder, ultrapure water, 2-propanol and ionomer. An aliquot of the dispersion was deposited onto a previously polished and cleaned surface
of a glassy carbon electrode disk used as a RRDE electrode. After drying under ambient conditions, a thin uniform catalyst film
was formed onto the surface of the glassy carbon electrode. A Pt
ring electrode as a second working electrode was centered on the
catalyst-coated disk electrode. The ring voltage was set to 1.2 V ver-
sus RHE, while the voltage of the thin-catalyst-film-disk electrode
was linearly varied with a constant scan rate in a range of 0.06–
1.00 V versus RHE. The utilization of a second working electrode
(Pt ring) allows us detecting the released peroxide formed from the
catalyst film on the disk electrode by the electrochemical conversion of O2 . Based on the used geometric arrangement of the electrode disk and ring, the collection efficiency was 38%. The collection efficiencies for all three used electrolytes were determined in
independent electrochemical experiments at the same electrolytes
used in this work and at different rotating speeds for Fe+2 /Fe+3
systems. For the electrochemical detection of peroxide the Pt ring
was held at a constant voltage of 1.2 V versus RHE to electrochemically reduce H2 O2 to H2 O and hence results in a reductive current. Fig. 2 shows the detected current for peroxide formation on
the Pt ring at constant voltage of 1.2 V versus RHE (top) and the
geometric-based current density during the linear voltage scan on
the catalyst-coated disk electrode (down) in different electrolytes
and at different pH values.
Very interestingly, four different regimes for the voltagedependent peroxide formation and its lifetime were observed during the potential scan, signifying the influence of the nature of
electrolyte and pH value (see Fig. 2). Regime I (at lower voltage
range of 0.06–0.20 V versus RHE) exhibits the strong effects of the
pH and nature of electrolyte on the catalyst reactivity and selectivity for the formation of peroxide by electrochemical reduction
of oxygen. Based on the experimental results, the reactivity and
the selectivity toward peroxide on meso-BMP decrease in the order
of KClO4 > HClO4 > KOH in the regime I. Contrary, in regime III (in
the higher voltage range of 0.40–0.80 V versus RHE) the increasing
formation of peroxides is according to KOH > HClO4 > KClO4 . However, it seems that the potential range of 0.2 and 0.4 V versus RHE
represents a mixed transfer regime (regime II). Finally, no peroxide
formation was observed above 0.80 V versus RHE (regime IV). In
the presence of KOH, the current-potential behavior and the corresponding H2 O2 response clearly differs from those measured in
KClO4 and HClO4 . A maximum of the ring current was detected at
0.5 V versus RHE unlike to the electrochemical responses in acidic
and neutral media. In comparison, in 0.1 M HClO4 (pH 1) and KClO4
(pH 7) at similar potential range, the absence of a current maximum is visible in Fig. 2. In the case of 0.1 M HClO4 (pH 1) and
KClO4 (pH 7), a gradual decrease of the ring current on the Pt ring
electrode was observed with increasing disk potential, signifying a
strong decrease of peroxide content with sufficient lifetime at the
254
F. Hasché et al. / Journal of Energy Chemistry 25 (2016) 251–257
Fig. 3. Percentage peroxide formation detected on the Pt ring electrode in various
0.1 M electrolyte solutions at different pH values in a potential range from 0.06 to
0.20 V versus RHE, established from RRDE measurements shown in Fig. 2.
hydrogen electrode (SHE) for the hydrogen peroxide formation by
electroreduction of oxygen is
O2 (g )+2H+ (aq ) + 2e− → H2 O2 (aq ) E0 = 0.695 V versus SHE
Fig. 2. Rotating ring disk electrode (RRDE) measurements at room temperature
with 1600 rounds per minute (rpm) for the mesoporous nitrogen-doped carbon catalyst in dependence of the nature of electrolyte and pH values. The ring current
(top) was detected on the pure Pt ring at a constant voltage (1.2 V versus RHE) and
the geometric-based current density (bottom) was measured on the catalyst-coated
glassy carbon disk (0.196 cm2 , loading 310 μgcat. /cm2 geo. ) during the linear voltage
scan with 5 mV/s. Romaic I to IV and the dashed lines indicate the different observed regimes for the electrochemical peroxide formation and its lifetime in each
electrolyte.
corresponding anodic potential and electrolyte solution. The observed different current-voltage behaviors in all three electrolytes
used in this work clearly show that the formation mechanisms
of peroxide are strongly dependent on the pH value and applied
potential. Non-metal nitrogen doped carbons show low sensitivity to effects of ion adsorption such as crossover effects, so that
the influence of the ClO4 − anion can be neglected. Considering the
Pourbaix diagram of water decomposition it is more likely that
the difference in the chemical composition of the peroxide species
gives rise to the different profiles of the respective polarization
curves [52]. At pH higher than 11.63 the main species is hydroperoxide HO2 − instead of hydrogen peroxide H2 O2 and this obviously
reduces the overpotential for the formation of peroxides, but also
for their further reduction to water as we can observe from the
reduced ring currents at lower potentials (in regions I and II).
Moreover, theoretic calculations and experimental observations
are contradictory. The experimentally found trends are opposite to
the expected role of protons, facilitating the decomposition of peroxides [48]. Also the amount of pyridinic nitrogen species, which
is expected to be the most active nitrogen site for the reduction of
hydrogen peroxide cannot be supported by our experimental result, indicating that other effects, such as Fermi levels of the electrocatalysts may be more promising descriptors for the catalytic
activity [53,54].
As shown in Fig. 2, in the presence of oxygen the open circuit
potential decreased in the order of KOH (0.86 V versus RHE)
> HClO4 (0.63 V versus RHE) > KClO4 (0.51 V versus RHE). As
described in the literature [55] and shown in Eq. 1, the standard
(1)
and for the consecutive reaction to water via another 2 electron
transfer (Eq. 2)
H2 O2 (aq )+2H+ (aq )+2e− → 2H2 O(aq ) E0 = 1.763 V versus SHE
(2)
Unlike to SHE, the reversible hydrogen electrode (RHE) is independent from the pH value and the relation can be described as
SHE = RHE – 59 mV ∗ pH. Throughout the paper, all electrode potentials were converted and reported in the RHE scale. Due to the
simplicity, the RHE scale allows us a direct comparison between
the potentials obtained from the different electrochemical experiments at a broad pH range without any additional conversion. It is
noted that at pH larger than 11.63 the HO2 – species is thermodynamically preferred compared to H2 O2 [56]. Thus, the formed peroxide exists as alkaline peroxide (e.g. K+ – O2 H) in strongly alkaline
environments.
Based on our experimental results shown in Fig. 2, the selective formation Xperoxide (in % H2 O2 ) of stable (detectable) peroxide species was calculated by establishing the ratios of the faradaic
currents measured on the disk electrode and ring electrode in a
potential range between 0.06 and 0.20 V versus RHE (see Fig. 3).
Eq. 3 shows the relation between the measured faradaic currents
on the Pt ring and meso-BMP coated disk electrode with a collection efficiency N of 0.38 ± 0.01 for the peroxide formation. As
mentioned above, the value of collection efficiency was determined
by independent electrochemical experiments for the all three used
electrolytes as described in the literature [45].
Xperoxide = 2 jRing /N / jDisk + jRing /N
(3)
As shown in Fig. 3, the highest detectable formation rates of
stable peroxide was observed in KClO4 and HClO4 electrolyte solutions at low potentials. This observation is in agreement with the
literature, that peroxide molecules are stable in acid, while in alkaline solutions peroxide tends to decompose [6]. At a certain potential of 0.10 V versus RHE the detected peroxide formation (Fig. 3)
follows the order: KClO4 (30%) > HClO4 (22%) > KOH (5%). This
F. Hasché et al. / Journal of Energy Chemistry 25 (2016) 251–257
255
result highlights that meso-BMP showed a strong pH-dependent
reactivity and selectivity toward peroxide. The highest peroxide selectivity was reached at neutral pH value compared to those at low
or high pH values at an applied potential of 0.10 V versus RHE. In
other words, in neutral media (KClO4 , pH 7) the peroxide selectivity is 6 times higher and in acidic media (HClO4 , pH 1) 4 times
higher than that in alkaline conditions (KOH, pH 13) at 0.1 V versus RHE. To demonstrate the effects of the nitrogen doping, we
note that the peroxide formation of pure carbon catalysts in acidic
medium (HClO4 , pH 1) ranges around 10% and thus by adding N
atoms into the carbon framework we can generate twice as much
peroxide for each mole of converted oxygen. Considering all RRDE
experimental data, the mesoporous nitrogen-doped carbon catalyst
shows excellent peroxide selectivity and high catalytic activity in
neutral and acidic media. The relatively low activity of meso-BMP
observed in alkaline media is likely also related to the chemical
decomposition of preformed and released peroxide.
3.3. Photometric determination of peroxide formation over the
reaction time
To corroborate the RRDE based peroxide formation data, an
independent photometric method was carried out to establish
the content of solved peroxide in the bulk electrolyte over the
reaction time. In certain reaction time intervals, small volumes
of electrolyte were probed from the electrochemical cell. Subsequently, a commercially available quantitative peroxide test solution (purchased from Merck KGaA) was rapidly added to the sample. In a very short time, a color reaction appeared and allowed
a colorimetric determination of the peroxide content in the bulk
electrolyte. Solutions turned from transparent and colorless to
yellow–orange (dependent on the existing content of peroxide in
the solution). The photometric UV–VIS measurements were then
performed at 500 nm to establish the peroxide concentration. The
time evolution of the peroxide formation in the three different
electrolytes at respective pH values was monitored over 4 h at an
applied potential of 0.1 V versus RHE. As shown above (Fig. 2 and
Fig. 3) we observed that at 0.1 V versus RHE the highest content of
peroxide was detected using the RRDE technique in acidic and neutral electrolytes. Although in alkaline media the peroxide content
was also significantly high at 0.1 V versus RHE, the highest selectivity toward peroxide was formed in the potential range of 0.4 and
0.6 V versus RHE. Since the peroxide formation in acidic and neutral electrolytes for this potential range (0.4–0.6 versus RHE) was
not detectable using photometry, we used a constant potential of
0.1 V versus RHE for all photometric experiments. As a result, the
dependence of the peroxide concentration over the reaction time
for meso-BMP at different pH and constant voltage (0.1 V versus
RHE) are shown in Fig. 4. In accordance to the electrochemically
observed results, the meso-BMP catalyst showed the highest activity at a potential of 0.1 V versus RHE in neutral and acidic environments. The peroxide formation by water oxidation at 0.1 V versus
RHE can be excluded due to kinetic and thermodynamic aspects
(see Eq. 2).
Fig. 4 shows a relationship between peroxide formation over
the reaction time and reaction media (electrolyte, pH) at a certain potential and thus the resulting reactivity and selectivity of
the meso-BMP in different electrolytes and at different pH values. It is obvious that the pH has a strong effect on the formation and lifetime of peroxide. The almost linear courses for
the time-resolved formation of peroxide in three different electrolytes/pH values clearly showed different slopes, indicating, that
the formation of peroxide proceeds on different reaction mechanisms and kinetics. It is evident that the pH plays an important
role on the formation and lifetime of peroxide. The highest concentration and therefore the highest formation rate of peroxide
Fig. 4. Experimental time evolution of peroxide concentration in three different electrolytes and pH determined by photometry. Conditions: 1600 rotation per
minute, ∼307 μgcat. /cm2 geo. , room temperature, Econst . = 0.1 V versus RHE, continuously O2 -purged electrolyte solution.
were observed in neutral media. This result is in good agreement
with the RRDE results as shown in Fig. 3, where the highest selectivity was detected at 0.1 V versus RHE. In contrast, the lowest
activity for meso-BMP was observed at the same potential in alkaline solution. This observation is likely related to the fact that
for the electrochemical peroxide formation the optimum potential range is between 0.4 and 0.6 V versus RHE. After 4 h of reaction time at 0.1 V versus RHE, the detected peroxide concentrations
behave like KClO4 (44 mgperoxide /L) > HClO4 (13 mgperoxide /L) > KOH
(3 mgperoxide /L). Reasons for the different catalytic behaviors in the
different pH values and electrolytes are likely due to the different formation and decomposition mechanisms of peroxide to
H2 O induced by the catalyst or by the electrolyte/pH in oxidizing
environments. Mukerjee et al. discussed that in acidic medium
FeNx /C catalysts show increased peroxide formation due to outer
sphere oxygen reduction and the resulting stable peroxide species,
whereas in alkaline medium an inner sphere mechanism enhances
both oxygen and peroxide reduction reaction [57]. Importantly the
improved onset potential for the ORR in alkaline medium as compared to the acidic medium is hereby explained by the high activity toward peroxide reduction. In case of non-metal N-doped carbons accordingly the ORR seems to follow an outer sphere mechanism independent of the pH, however the onset potential is nevertheless up-shifted in alkaline medium without peroxide reduction
activity worth mentioning. We emphasized that apart from the
peroxide formation process an increased decomposition process at
extreme pH conditions certainly contributes. In the case of alkaline
and acidic media, the detectable formation rate grew slightly during the reaction time compared to neutral media, indicating that
the preformed peroxide is more stable in this media. It is noted,
that we used a high catalyst loading of ∼307 μgcat /cm2 geo on the
disk electrode as working electrode. In the fuel cell research, it is
well known that with increasing catalyst loading (typical loadings
are 5–50 μgPt /cm2 geo. , results in a similar carbon loading used in
this study) the H2 O2 rate decreases [58]. It can be explained by readsorption of H2 O2 on neighboring catalyst surface followed by the
consecutive decomposition of H2 O2 to H2 O on the Pt catalyst [59].
This way we want to point out the high selectivity and activity
of the mesoporous nitrogen-doped carbon for the electrochemical
peroxide formation as a materials property rather than a property
of electrode design.
256
F. Hasché et al. / Journal of Energy Chemistry 25 (2016) 251–257
meso-BMP catalyst showed a high electrochemical stability and
selectivity toward peroxide at 0.1 V versus RHE in 0.1 M KClO4 .
4. Conclusions
Fig. 5. Long-term rotating ring disk electrode (RRDE) measurements for the mesoBMP catalyst in 0.1 M KClO4 electrolyte at room temperature. Percentage of peroxide formation (top) (Pt ring at constant 1.2 V versus RHE) and catalyst-coated
disk (glassy carbon, 0.196 cm2 geo. ) current density (down) at constant voltage
(0.1 V versus RHE) over the time. Conditions: 1600 rounds per minute, loading
∼307 μgcat. /cm2 geo. , room temperature, continuously O2 -purged electrolyte solution
with a flow of ∼250 mL/min.
3.4. Stability–selectivity relation for the hydrogen peroxide formation
in 0.1 M KClO4
As shown in Figs. 2 and 3, in neutral electrolyte solution
consisting of weakly adsorbed perchlorate anions the meso-BMP
showed the highest initial activity and selectivity at 0.1 V versus RHE. To examine the activity–selectivity–stability relations, a
long-term RRDE experiment was performed on meso-BMP-coated
disk electrode at constant voltage of 0.1 V versus RHE for 6 h (see
Fig. 5). The selectivity toward peroxide over reaction time was
measured on the Pt ring holding the potential of 1.2 V versus RHE.
A constant geometric-based current density on the disk electrode (jdisk ) was reached after around 30 min and was almost
unchanged during a reaction time of 6 h, indicating a high
(electro) chemical resistance of the catalyst against the applied
voltage, pH value, electrolyte and, more important, preformed peroxide. By measuring the current on the Pt ring electrode over
the same reaction time of 6 h the course of the selectivity toward peroxide could be investigated. During the 6 h the peroxide formation clearly grew from 35% (75 μA ring current) to
81% (182 μA ring current). Based on the course behavior of the
ring current, an extension of the reaction time might lead to
a continuous increase of the electrochemical peroxide formation.
It is noted that the origin of the growth of the peroxide content over the time is likely related to the continuous formation of peroxide on the meso-BMP electrocatalyst as well as
the additional accumulation of preformed peroxide in the bulk
electrolyte. At the same time and in contrast to peroxide accumulation in the bulk electrolyte solution, peroxide decomposition can occur during long-term measurements as well. The
long-term RRDE experiments highlights that the noble metal-free
A metal-free mesoporous nitrogen-doped carbon catalyst
showed a high electrocatalytic activity, durability and selectivity
toward peroxide by electrochemical converting of O2 in a noncorrosive neutral as well as in acidic reaction medium. For the first
time, the effects of the pH on the electrochemical formation of
peroxide were demonstrated by using two different independent
experimental methods (RRDE and UV–VIS). We showed that the
H2 O2 selectivity and formation rates strongly depend on the supporting electrolyte i.e. pH value and nature of electrolyte, and increases according to: neutral (KClO4 ) > acidic (HClO4 ) > alkaline
(KOH). The observed experimental results evidences that the electrochemical formation mechanisms for peroxide are strongly dependent on the pH and the kind of electrolyte, going along with
the respective change of the peroxide species from hydrogen peroxide to hydroperoxide. The results additionally indicate that increased ORR activity in alkaline medium as compared to acidic
medium for N-doped carbons is not explained by an upshifted
onset potential due to preferred reduction of intermediate peroxide species. This mechanism was recently suggested by Mukerjee
et al. for FeNx catalysts [57]. Further experiments and studies are
in progress to evaluate the full potential of this type of metalfree catalyst as in such a way an electrochemical on-demand/onside production of peroxides might be possible avoiding hazardous
large scale production by means of the anthraquinone process and
transportation of concentrated hydrogen peroxide solutions [5].
The characterization and identification of the active center and catalyst structure using e.g. in-situ methods like ambient-pressure Xray photoemission spectroscopy for topmost surface or (N)EXAFS
for near surface and bulk analysis as well as the theoretical understanding and calculation for the possible reaction pathways and
steps in various electrolytes / pH values are helpful to get deeper
insights in the reaction mechanism. Furthermore, the transfer of
the catalyst in reactor concepts [8,9,60] e.g. fuel cell reactor type
which should produce peroxide directly in a continuous mode with
a minimum amount of liquid electrolyte, are required.
Acknowledgments
The authors thank Annette Wittebrock for laboratory support
(TU Berlin, Germany), Sören Selve for TEM measurement (TU
Berlin, Zentraleinrichtung Elektronenmikroskopie, Germany) and
Carmen Serra for XPS measurement (Universidad de Vigo, Spain).
Supplementary Materials
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.jechem.2016.01.024.
References
[1] T.-P. Fellinger, F. Hasché, P. Strasser, M. Antonietti, J. Am. Chem. Soc. 134 (2012)
4072–4075.
[2] S. Siahrostami, A. Verdaguer-Casadevall, M. Karamad, D. Deiana, P. Malacrida,
B. Wickman, M. Escudero-Escribano, E.A. Paoli, R. Frydendal, T.W. Hansen,
I. Chorkendorff, I.E.L. Stephens, J. Rossmeisl, Nat. Mater. 12 (2013) 1137–1143.
[3] A. Verdaguer-Casadevall, D. Deiana, M. Karamad, S. Siahrostami, P. Malacrida,
T.W. Hansen, J. Rossmeisl, I. Chorkendorff, I.E.L. Stephens, Nano Lett. 14 (2014)
1603–1608.
[4] R.L. Myers, The 100 Most Important Chemical Compounds: A Reference Guide,
Greenwood Publishing Group, 2007.
[5] G. Goor, J. Glenneberg, S. Jacobi, Hydrogen Peroxide, Ullmann’s Encyclopedia
of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, 20 0 0.
[6] J.M. Campos-Martin, G. Blanco-Brieva, J.L.G. Fierro, Angewandte Chem. Int. Ed.
45 (2006) 6962–6984.
F. Hasché et al. / Journal of Energy Chemistry 25 (2016) 251–257
[7] G. Goor, J. Glenneberg, S. Jacobi, Hydrogen Peroxide, Wiley-VCH Verlag GmbH
& Co. KGaA, 20 0 0.
[8] I. Yamanaka, T. Onisawa, T. Hashimoto, T. Murayama, ChemSusChem 4 (2011)
494–501.
[9] T. Inoue, Y. Tanaka, D.A. Pacheco Tanaka, T.M. Suzuki, K. Sato, M. Nishioka,
S. Hamakawa, F. Mizukami, Chem. Eng. Sci. 65 (2010) 436–440.
[10] Y. Yi, J. Zhou, H. Guo, J. Zhao, J. Su, L. Wang, X. Wang, W. Gong, Angewandte
Chem. Int. Ed. 125 (2013) 8604–8607.
[11] J. Durst, C. Simon, A. Siebel, P.J. Rheinländer, T. Schuler, M. Hanzlik, J. Herranz,
F. Hasché, H.A. Gasteiger, ECS Trans. 64 (2014) 1069–1080.
[12] J. Durst, A. Siebel, C. Simon, F. Hasché, J. Herranz, H.A. Gasteiger, Energy Environ. Sci. 7 (2014) 2255–2260.
[13] K. Otsuka, I. Yamanaka, Electrochimica Acta 35 (1990) 319–322.
[14] I. Yamanaka, T. Onizawa, S. Takenaka, K. Otsuka, Angewandte Chem. Int. Ed. 42
(2003) 3653–3655.
[15] I. Katsounaros, W.B. Schneider, J.C. Meier, U. Benedikt, P.U. Biedermann,
A.A. Auer, K.J.J. Mayrhofer, Phys. Chem. Chem. Phys. 14 (2012) 7384–7391.
[16] I. Katsounaros, K.J.J. Mayrhofer, Chem. Commun. 48 (2012) 6660–6662.
[17] T. Jacob, Fuel Cells 6 (2006) 159–181.
[18] F. Hasché, M. Oezaslan, P. Strasser, ChemCatChem 3 (2011) 1805–1813.
[19] F. Hasché, M. Oezaslan, P. Strasser, J. Electrochem. Soc. 159 (2012) B25–B34.
[20] M. Oezaslan, F. Hasché, P. Strasser, J. Electrochem. Soc. 159 (2012) B394–B405.
[21] M. Oezaslan, F. Hasché, P. Strasser, J. Electrochem. Soc. 159 (2012) B444–B454.
[22] M. Oezaslan, P. Strasser, J. Power Sour. 196 (2011) 5240–5249.
[23] T. Toda, H. Igarashi, H. Uchida, M. Watanabe, J. Electrochem. Soc. 146 (1999)
3750–3756.
[24] V.R. Stamenkovic, B.S. Mun, K.J.J. Mayrhofer, P.N. Ross, N.M. Markovic, J. Am.
Chem. Soc. 128 (2006) 8813–8819.
[25] V. Stamenkovic, T.J. Schmidt, P.N. Ross, N.M. Markovic, J. Phys. Chem. B 106
(2002) 11970–11979.
[26] J. Zhang, Y. Mo, M.B. Vukmirovic, R. Klie, K. Sasaki, R.R. Adzic, J. Phys. Chem. B
108 (2004) 10955–10964.
[27] K. Sasaki, H. Naohara, Y. Cai, Y.M. Choi, P. Liu, M.B. Vukmirovic, J.X. Wang,
R.R. Adzic, Angewandte Chem. Int. Ed. 49 (2010) 8602–8607.
[28] F. Jaouen, J. Herranz, M. Lefèvre, J.-P. Dodelet, U.I. Kramm, I. Herrmann, P. Bogdanoff, J. Maruyama, T. Nagaoka, A. Garsuch, J.R. Dahn, T. Olson, S. Pylypenko,
P. Atanassov, E.A. Ustinov, ACS Appl. Mater. Interfaces 1 (2009) 1623–1639.
[29] M. Lefevre, E. Proietti, F. Jaouen, J.-P. Dodelet, Science 324 (2009) 71–74.
[30] R. Bashyam, P. Zelenay, Nature 443 (2006) 63–66.
[31] G. Wu, K.L. More, C.M. Johnston, P. Zelenay, Science 332 (2011) 443–447.
[32] X. Yuan, X. Zeng, H.-J. Zhang, Z.-F. Ma, C.-Y. Wang, J. Am. Chem. Soc. 132 (2010)
1754–1755.
[33] W. Yang, T.-P. Fellinger, M. Antonietti, J. Am. Chem. Soc. 133 (2011) 206–209.
[34] K. Gong, F. Du, Z. Xia, M. Durstock, L. Dai, Science 323 (2009) 760–764.
[35] K.-H. Wu, D.-W. Wang, D.-S. Su, I.R. Gentle, ChemSusChem 8 (2015) 2772–
2788.
257
[36] K. Mase, K. Ohkubo, S. Fukuzumi, J. Am. Chem. Soc. 135 (2013) 2800–2808.
[37] I. Yamanaka, S. Tazawa, T. Murayama, T. Iwasaki, S. Takenaka, ChemSusChem 3
(2010) 59–62.
[38] J.K. Edwards, A. Thomas, B.E. Solsona, P. Landon, A.F. Carley, G.J. Hutchings,
Catal. Today 122 (2007) 397–402.
[39] J.K. Edwards, J. Pritchard, L. Lu, M. Piccinini, G. Shaw, A.F. Carley, D.J. Morgan,
C.J. Kiely, G.J. Hutchings, Angewandte Chem. Int. Ed. 53 (2014) 2381–2384.
[40] A. Verdaguer-Casadevall, D. Deiana, M. Karamad, S. Siahrostami, P. Malacrida,
T.W. Hansen, J. Rossmeisl, I. Chorkendorff, I.E.L. Stephens, Nano Lett. (2014).
[41] J.S. Jirkovský, I. Panas, E. Ahlberg, M. Halasa, S. Romani, D.J. Schiffrin, J. Am.
Chem. Soc. 133 (2011) 19432–19441.
[42] C.H. Choi, H.C. Kwon, S. Yook, H. Shin, H. Kim, M. Choi, J. Phys. Chem. C 118
(2014) 30 063–30 070.
[43] V. Perazzolo, C. Durante, R. Pilot, A. Paduano, J. Zheng, G.A. Rizzi, A. Martucci,
G. Granozzi, A. Gennaro, Carbon 95 (2015) 949–963.
[44] Y. Liu, X. Quan, X. Fan, H. Wang, S. Chen, Angewandte Chem. Int. Ed. 54 (2015)
6837–6841.
[45] U.A. Paulus, T.J. Schmidt, H.A. Gasteiger, R.J. Behm, J Electroanal. Chem. 495
(2001) 134–145.
[46] O. Antoine, R. Durand, J. Appl. Electrochem. 30 (20 0 0) 839–844.
[47] N.M. Markovic, P.N. Ross Jr, Surface Sci. Rep. 45 (2002) 117–229.
[48] P. Wu, P. Du, H. Zhang, C. Cai, Phys. Chem. Chem. Phys. 15 (2013) 6920–6928.
[49] S. Zhang, S. Tsuzuki, K. Ueno, K. Dokko, M. Watanabe, Angewandte Chem. Int.
Ed. 54 (2015) 1302–1306.
[50] F. Hasché, T.-P. Fellinger, M. Oezaslan, J.P. Paraknowitsch, M. Antonietti,
P. Strasser, ChemCatChem 4 (2012) 479–483.
[51] J.P. Paraknowitsch, J. Zhang, D. Su, A. Thomas, M. Antonietti, Adv. Mater. 22
(2010) 87–92.
[52] M. Pourbaix, Lectures on Electrochemical Corrosion, Springer US, Boston, MA,
1973.
[53] J.Y. Cheon, J.H. Kim, J.H. Kim, K.C. Goddeti, J.Y. Park, S.H. Joo, J. Am. Chem. Soc.
136 (2014) 8875–8878.
[54] Y. Jiao, Y. Zheng, M. Jaroniec, S.Z. Qiao, J. Am. Chem. Soc. 136 (2014) 4394–
4403.
[55] F. Tian, R. Jinnouchi, A.B. Anderson, J. Phys. Chem. C 113 (2009) 17484–17492.
[56] M. Pourbaix, N. De Zoubov, Atlas of Electrochemical Equilibria in Aqueous
Solutions, National Association of Corrosion Engineers, Houston, Tex., 1974,
pp. 106–111.
[57] N. Ramaswamy, U. Tylus, Q. Jia, S. Mukerjee, J. Am. Chem. Soc. 135 (2013)
15443–15449.
[58] M. Inaba, H. Yamada, J. Tokunaga, A. Tasaka, Electrochem. Solid-State Lett. 7
(2004) A474–A476.
[59] A. Bonakdarpour, T.R. Dahn, R.T. Atanasoski, M.K. Debe, J.R. Dahn, Electrochem.
Solid-State Lett. 11 (2008) B208–B211.
[60] C. Oloman, A.P. Watkinson, J. Appl. Electrochem. 9 (1979) 117–123.