CARBON 6 0 ( 2 0 1 3 ) 7 6 –8 3

Available at www.sciencedirect.com

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

The use of cyclic voltammetry to assess the activity of carbon

materials for hydrogen peroxide decomposition
a,*
C.M. Domı́nguez , A. Quintanilla a, P. Ocón b, J.A. Casas a, J.J. Rodriguez a

a
Sección de Ingenierı́a Quı́mica, Universidad Autónoma de Madrid, Crta.de Colmenar km 15, 28049 Madrid, Spain
b
Departamento de Quı́mica-Fı́sica Aplicada, Universidad Autónoma de Madrid, Crta.de Colmenar km 15, 28049 Madrid, Spain

A R T I C L E I N F O A B S T R A C T

Article history: It is known that carbon materials catalyze hydrogen peroxide decomposition in aqueous
Received 10 January 2013 media. However, the catalytic activity of a particular carbon is dependent on various cou-
Accepted 27 March 2013 pled structural, textural and chemical characteristics of the material, such that, formerly,
Available online 6 April 2013 the prediction of activity has not been possible. Here, the application of cyclic voltammetry
(CV) is introduced as a rapid and conclusive technique in this respect. Three classes of car-
bon materials have been investigated: activated carbons, carbon blacks, and graphites,
including some selected acid-washed samples which were used to examine the roles of
mineral matter and surface oxygen. Characterization by electrochemical capacity mea-
surements with CV, together with catalytic activity tests for hydrogen peroxide decompo-
sition, reveal that the exchange current is directly proportional to the catalytic activity for
hydrogen peroxide decomposition. That is, a linear dependence was found between this
variable and the apparent first order catalytic decomposition rate constant. CV measure-
ments with modified carbons also allow the elucidation of the effects of physicochemical
characteristics of carbon materials on the rate of hydrogen peroxide decomposition.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction content [9–13], the microstructure and porosity [3,7,8] and

chemical composition, specifically the quantity and nature
Hydrogen peroxide decomposition over carbon materials is of of surface functional groups [1,4–10,12,14]. For this reason,
interest in the field of wastewater treatment because it leads the studies dealing with the application of carbon catalysts
to the formation of radicals (OH and OOH), highly oxidizing in hydrogen peroxide decomposition, or the processes in
species allowing the breakdown of organic pollutants in aque- which this reaction is involved viz. wastewater treatments
ous solution and/or adsorbed onto the carbon surface [1]. by wet peroxide oxidation [6,7,10,12,15], usually include a
That decomposition reaction can take place due to the do- number of tedious, time-consuming and expensive character-
nor–acceptor properties of the carbon surface (C), through ization techniques (e.g., Raman spectra, Fourier transform
an electron-transfer mechanism similar to the Fenton reac- infrared spectroscopy (FTIR), diffuse reflectance infrared
tion (Fig. 1), acting C and C+ as the reduced (1) and oxidized Fourier transform spectroscopy (DRIFTS), temperature pro-
(2) carbon states, respectively [2–7]. grammed desorption (TPD), X-ray photoelectron spectroscopy
This redox cycle is highly dependent on the structural, tex- (XPS), thermal gravimetric analysis (TGA), elemental analysis
tural and chemical characteristics of carbon materials [7,8]. (EA), X-ray diffraction (XRD), total reflection X-ray fluores-
The most important features affecting the electron-transfer cence (TXRF), induced coupled plasma (ICP), N2 adsorption/
reactivity and, consequently, the carbon activity are the ash desorption isotherms). In spite of this, achieving conclusive

* Corresponding author: Fax: +34 914973516.

E-mail address: [email protected] (C.M. Domı́nguez).
0008-6223/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.carbon.2013.03.058
 CARBON 6 0 (2 0 13 ) 7 6–83 77

studied: two activated carbons (AC), supplied by Merck (AC-

M, Ref.: 102514) and Panreac (AC-P, Ref.: 3108L); two carbon
blacks (CB), supplied by Chemviron (CB-C, Ref.: 2156090) and
Vulcan (CB-V, Ref.: CC72R) and two graphites (G), supplied
by Sigma–Aldrich (G-S, Ref.: 282863) and Fluka (G-F, Ref.:
1249167). All samples were provided in powder form except
for the AC-M, which was sieved to 80–100 lm before use.
In addition, the AC-M and G-S samples were treated with
concentrated HCl (37%) solution, at room temperature for
4 h (2 g sample/25 mL HCl solution), and then washed with
distilled water until constant pH, filtered, and dried overnight
at 60 °C. The resulting carbons were identified as AC-M-HCl
Fig. 1 – Hydrogen peroxide reduction (1) and oxidation (2) and G-S-HCl, respectively. A similar treatment was performed
over carbon materials. on AC-P but using HNO3 (65%) yielding the AC-P-HNO3 sam-
ple. Table 1 summarizes the nomenclature of all the carbon
materials tested.
results about the characteristics of carbon materials, directly
or indirectly related to the activity, is not always possible. The 2.2. Characterization
reason for this is that the different issues cannot be analyzed
independently, due to the difficulty of tailoring a single car- The structural parameters were determined by X-ray diffrac-
bon characteristic upon the preparation process, without tion. XRD patterns were obtained using a Siemens Model
affecting the rest. Therefore, the occurrence of coupled ef- D5000 X-ray diffractometer, Cu Ka (8.04 keV) radiation, and a
fects is common and makes impracticable a reasonable pre- step of 0.02°/s for 2h=5–100°. XRD data were analyzed with
diction of the catalytic activity of carbon materials as a PDF 2000 (JCPDS-ICDD) software.
function of their structural and physico-chemical The BET surface area (SBET) of the initial and modified car-
characteristics. bons was obtained from the 77 K N2 adsorption/desorption
In this paper, we present cyclic voltammetry as a simple isotherms using Micromeritics Tristar apparatus, outgassing
and direct technique which is able to provide conclusive the samples overnight at 150 °C to a residual pressure of
information about the catalytic activity of carbon materials <103 torr. The external or non-microporous surface area
in hydrogen peroxide decomposition reaction. We demon- (Aext) was calculated by the t-method.
strate by using different types of carbon materials viz. acti- Elemental analyses of the carbon materials were per-
vated carbons, carbon blacks and graphites, that the formed in a LECO Model CHNS-932 apparatus. The identifica-
electrochemical behaviour of the carbon for some particular tion and quantification of the elements was determined by
structural and physico-chemical characteristics given, gov- TXRF (Extra-II Rich & Seifert spectrometer).
erns the activity. This technique has occasionally been used The surface oxygen groups were analyzed by temperature-
to characterize metallic catalysts [16] and to predict their programmed desorption under N2 (TPD). A 100 mg sample of
activity [17,18] but to the best of our knowledge it has not carbon was placed in a quartz tube and heated at 10 °C/min
yet been reported for carbon materials. from room temperature up to 900 °C under a continuous N2
flow of 1 L/min. The CO2 and CO evolved were measured in
2. Experimental a gas analyzer (SIEMENS mod. Utramat 22).

2.1. Carbon materials 2.3. Cyclic voltammetry experiments

Three types of commercial carbon materials with different Electrochemical measurements were carried out in a
structural, textural and chemical characteristics have been conventional three-electrode electrochemical cell, using a

Table 1 – Structural, textural and chemical characteristics of the carbon materials tested.
Sample d002 (Å) Lc (Å) La (Å) SBET (m2/g) Aext (m2/g) Ashes (%) Fe (%) CO2 (lmol/g) CO (lmol/g)

AC-M 3.614 17.7 38.0 1019 175 4.00 0.04 30 320

AC-M-HCl – – – 843 128 1.38 0.03 74 562
AC-P 3.793 16.2 36.0 931 472 1.00 0.01 140 662
AC-P-HNO3 – – – 894 354 0.35 0.01 669 1866
CB-C 3.476 70.8 65.7 75 75 0.00 0.00 83 160
CB-V 3.618 38.7 38.9 233 110 1.00 0.00 148 106
G-S 3.383 440.9 487.0 12 12 0.50 0.44 105 36
G-S-HCl – – – 11 11 0.08 0.02 – –
G-F 3.367 466.3 488.7 7 5 0.20 0.00 128 79
78 CARBON 6 0 ( 2 0 1 3 ) 7 6 –8 3

computerized potentiostat (Autolab PGSTAT 302, Eco Chemie) catalyst was separated by filtration (0.45 lm Nylon filter)
controlled by GPES software. A glassy carbon rotating disk and oven-dried at 60 °C.
electrode (3 mm diameter) was used as substrate for the car-
bon samples (working electrode). Prior to each test, the elec-
trode was polished to 3 lm with alumina powder and rinsed 3. Results and discussion
with Milli-Q ultrapure water. A gold electrode was used as
the counter electrode and a saturated Ag/AgCl, KCl electrode 3.1. Characterization study
served as the reference. The electrochemical experiments
were carried out at room temperature and ambient pressure Table 1 summarizes the structural characteristics of the car-
using an HCl (pH = 3.5) solution as the electrolyte. bons obtained from XRD, the BET and external surface area,
The carbons under study were deposited on the working ash and Fe content, and the amount of CO2 and CO evolved
electrode as an ink. Typically, 6 mg of sample was dispersed upon TPD analysis.
in Milli-Q water (730 lL) with ultrasonic agitation (Hyelscher, The values of the structural parameters in Table 1, such as
UP50H), to obtain a homogeneous suspension; 10 lL of this the interlayer spacing (d002), average crystalline height (Lc),
suspension was carefully dropped on a glassy carbon elec- and average crystalline diameter (La), were calculated from
trode and allowed to dry at room temperature for 60 min to the XRD patterns (shown in Fig. S1 of the Supplementary
obtain a uniform film. Material). The calculated d002 values confirm the amorphous
In a typical experiment, the working electrode was sub- character of activated carbons, AC-M and AC-P. Conversely,
jected to cyclic voltammetry (CV) measurements in a back- the G-S and G-F samples, present a highly ordered structure
ground solution (HCl, pH = 3.5) at a sweep rate (v) of 100 mV/s as indicated by the d002 close to the value of graphite
and the potential range from  0.6 to 1.0 V. Once unchanging (3.354 Å). While the carbon blacks, CB-C and CB-V, exhibit
cyclic voltammograms were obtained for several cycles (varia- intermediate values in the XRD parameters. The lower d002
tions in the coulumbic charge lower than 2%), the sweep rate and higher Lc and La values of CB-C indicate a more ordered
was reduced to a value of 10 mV/s, and data acquisition started. structure compared with CB-V.
First, the coulombic charge (CC) for the carbon was calculated The SBET values calculated from the N2 adsorption/desorp-
from the background voltammogram. Then, the precise vol- tion isotherms at 77 K (Fig. S2 of the Supplementary Material)
ume of hydrogen peroxide (30%) was added to the electrolyte demonstrate that activated carbons, AC-M and AC-P, present
solution and the open circuit potential (Eocp) registered. Finally, a highly developed porosity whereas the carbon blacks, CB-
a new CV was recorded to determine the electrochemical C and CB-V, and even more so, the graphites, G-S and G-F,
behavior of the carbon materials in the presence of hydrogen show substantially lower values of surface area without
peroxide. The absence of diffusion limitations in the inter- microporosity contribution, except in the case of the CB-V
phase for the overall electrochemical process should be noted carbon black. Differences in the distribution of porosity can
because identical results were obtained within the rotation be observed between the activated carbons, showing AC-P a
velocity range from 0 to 3000 rpm, as was expected for a sur- substantially higher contribution of mesoporosity.
face-controlled process. All measurements were carried out The ash content of the carbon materials tested is, in gen-
under nitrogen atmosphere (Air Liquid) by bubbling 1 mL/min eral, quite low, except in the case of the as-received activated
into the cell and all experiments were duplicated. carbon AC-M (4%). It is important to know the content of iron
By comparing the CVs obtained in the absence and pres- of the ashes because of its catalytic activity for hydrogen per-
ence of hydrogen peroxide, the onset potential (Eonset) for first oxide decomposition enhancing the ability of activated car-
reduction and oxidation was obtained. Also, the Tafel slope for bon in this reaction [3,10–13,20]. For this reason, the iron
both processes, called as b, and the current exchange (io) were content was quantified (Table 1). Unexpectedly, G-S had the
calculated from the overpotential (E  Eocp) vs. log (I) plots for highest amount of Fe (0.44%) whereas the presence of this
each test using the following equation: metal in the other materials was negligible. Regarding the sur-
face composition, the amount of CO2 and CO evolved, as ob-
ðE  Eocp Þ ¼ m þ 2:30  b  logðIÞ ð1Þ
tained from the integration of the TPD curves, (Fig. S3 of the
Supplementary Material) are shown in Table 1. As expected,
2.4. Hydrogen peroxide decomposition with carbon the two activated carbons show the highest content of sur-
catalysts face oxygen groups (SOG) as a result of the activation process.
Summarizing, three different types of carbons have been
The reactions were carried out batch-wise in a magnetically studied. Activated carbons, which are amorphous materials
stirred three-necked glass reactor equipped with a reflux con- with a developed porous structure and significant amounts
denser. The operating conditions were selected according to a of surface oxygen groups; graphites, which have the opposite
previous work [19]. In a typical experiment, 45 mL of distilled characteristics because they are almost crystalline materials,
water at pH = 3.5 (HCl, Sigma–Aldrich) was placed in the flask with a very low specific surface area and scarce surface oxy-
together with 0.125 g powdered catalyst. The content was gen groups; carbon blacks, with structural, textural and
heated to 80 °C (IKA RCT basic). Once this temperature had chemical features in between the two.
been reached, 5 mL of 250 g/L hydrogen peroxide (Sigma–Al- The effect of those characteristics on the catalytic behav-
drich) was injected and the content stirred at 1200 rpm. After ior of activated carbons for hydrogen peroxide decomposition
a reaction time of 2 h the heating was switched-off and the has been widely studied in the literature [1,4–6,8,10–14]. Some
flask cooled to room temperature in cold water. Finally, the of these studies conclude that disordered structures with less
 CARBON 6 0 (2 0 13 ) 7 6–83 79

developed graphitic layers have a higher concentration of 3.2. Carbon voltammetry experiments
electron-rich centers promoting hydrogen peroxide decompo-
sition [3,7]. However, the more disordered the structure, the Fig. 2 depicts the unchanging cyclic voltammograms in the
more surface oxygen groups. In general, acid SOG reduce background solution (pH = 3.5, HCl) for each type of carbon.
the reactivity of the activated carbons due to their electron When the applied potential changes to a negative direction
withdrawal capacity [1,5–7,12,14] whereas basic SOG are al- (reduction scan), a gradually increasing negative current
ways beneficial [4,8]. On the one hand, the diffusion of the re- was measured; in the same way, if the potential went in a po-
agents toward the active sites is favored by the hydrophilic sitive direction, the positive current increases and the oxida-
character of the carbon material, a feature controlled by the tion process takes place. None of the cyclic voltammograms
acid SOG [21], in the presence of a sufficiently developed mes- recorded were totally symmetrical. This fact led us to believe
oporosity [22]. By contrast, carbon materials with a highly or- that the carbon materials include not only purely capacitive,
dered structure such as graphites have electronic and but also electro-active species (viz. surface oxygen groups,
conductive properties which can favor the electron transfer heteroatoms or ashes) with redox properties.
between these materials and hydrogen peroxide [12]. The coulombic charge (CC), calculated from the back-
In general, carbon materials have a wide diversity of char- ground voltammograms, varied according to the specific sur-
acteristics which would hinder a predictive approach of their face area of the materials: AC > CB > G and a linear
activity for hydrogen peroxide decomposition as coupled ef- relationship between this parameter (CC) and SBET has been
fects frequently exist. In this context our proposal is that vol- found as can be seen in Fig. 3. The higher the surface area,
tammetry measurements are a useful tool for the assessment the more active sites (electron-rich locations on the carbon
of the activity of carbon materials in the aforementioned surface), and so, an increased coulombic charge can be ex-
reaction. In that respect, the starting and modified carbon pected. Eventually, cyclic voltammetry could be used for spe-
materials collected in Table 1 provide a series of widely differ- cific surface determination of carbon materials.
ent samples for testing. Additional cyclic voltammetry experiments were carried
out in presence of hydrogen peroxide with the purpose of
analyzing the electrochemical behavior of carbons when in
contact with that electroactive reactant. The Eocp values re-
corded before the potential application are given in Table 2
Activated carbons and the voltammograms are shown in Fig. 4. The Eocp values,
0.04
ranged from 0.413 to 0.499 V, being within the thermody-
AC-M
AC-P namic reduction/oxidation potentials of hydrogen peroxide
0.02
(0.905 and 1.566 V, respectively, Ag/AgCl) and correspond
I (mA)

with a mixed potential. The oxidation and reduction current

0.00
intensities measured in the presence of hydrogen peroxide
(Fig. 4) are clearly higher than those in the absence of it
-0.02
(Fig. 2). The carbon that exhibits the maximum ability for
the reduction step also shows the maximum for the oxidation
0.02
step. It is noteworthy the crossover loops observed in the for-
Carbon Blacks
mer. These hysteresis loops observed in both oxidation and
0.01 CB-C
reduction directions, indicate that the carbon surface is irre-
CB-V
versibly modified upon cyclic voltammetry when hydrogen
I (mA)

0.00

-0.01 0.006
AC-M
-0.02 0.005
Coulombic Charge (C)

Graphites AC-P
0.01 0.004
G-S
G-F
0.003
I (mA)

0.00
0.002
CB-V
-0.01 0.001
CB-C
G-S
G-F
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 0.000
E (V) 0 200 400 600 800 1000
2
SBET (m /g)
Fig. 2 – Cyclic voltammograms of carbon materials in the
absence of hydrogen peroxide (background). Operating Fig. 3 – Relationship between SBET and coulombic charge.
conditions: v = 10 mV/s, T = 25 °C, pH = 3.5 (HCl). Operating conditions: v = 10 mV/s, T = 25 °C, pH = 3.5 (HCl).
80 CARBON 6 0 ( 2 0 1 3 ) 7 6 –8 3

Table 2 – Electrochemical properties of the carbon materials tested.

Sample Eocp Reduction process Oxidation process io
Eonset b Eonset b

(V) (V) (V/decade) (V) (V/decade) (A/g)

AC-M 0.460 0.094 0.437 0.513 0.276 0.232

AC-M-HCl 0.489 0.022 0.393 0.575 0.227 0.125
AC-P 0.413 0.057 0.434 0.690 0.307 0.147
AC-P-HNO3 0.461 0.050 0.443 0.686 0.294 0.121
CB-C 0.493 0.050 0.398 0.559 0.219 0.060
CB-V 0.430 0.024 0.426 0.679 0.344 0.137
G-S 0.492 0.086 0.454 0.520 0.264 0.172
G-S-HCl 0.499 0.045 0.342 0.629 0.261 0.099
G-F 0.420 0.009 0.375 0.587 0.211 0.052

0.2 tammetry also disfavors the reduction of hydrogen peroxide.

Activated carbons As a consequence, a lower reduction current is obtained in
0.1 the reverse scan, from 0.1 to 0.6 V, and the second hyster-
AC-M
AC-P esis loop observed. Therefore, these findings suggest that the
I (mA)

0.0 regeneration of the carbon active sites (step 2 in Fig. 1) is ex-

pected to be the limiting reaction.
-0.1 Several electrochemical parameters viz. Eonset, Tafel slopes
and io were calculated from the voltammograms (see Section
-0.2 2). The values are summarized in Table 2.
The Eonset values for hydrogen peroxide reduction were
Carbon Blacks close to 0 mV for all the carbons tested, while for oxidation
0.1 more variable values were obtained, ranging from 0.513 to
CB-C
CB-V 0.690 V. The Tafel slope values corresponding to hydrogen per-
I (mA)

oxide reduction were between 0.375 and 0.454 V/decade,

0.0 while for oxidation, the range covered from 0.211 to 0.344 V/
decade (Table 2). These relatively high values indicate that
hydrogen peroxide decomposition over carbon materials is a
-0.1 complex process, which in addition to the charge transfer,
other phenomena, like adsorption/desorption, must be in-
Graphites volved. Furthermore, the fact that the Tafel slopes were quite
0.1 similar for all the carbon materials tested, suggests that the
G-S
G-F reaction mechanism does not vary from one material to
I (mA)

another.
0.0 According to the current exchange values, io (the most rep-
resentative parameter in electrochemical reactions, Table 2),
the activity of carbon materials in the cyclic voltammetry
-0.1 experiments follows the order: AC-M > G-S > AC-P > CB-
V > CB-C > G-F. The lowest activity of G-F is consistent with
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 the characterization results in Table 1 which show that this
E (V) carbon material has the lowest surface area and no iron con-
tent. Nevertheless, the voltammetry results indicate that AC-
Fig. 4 – Cyclic voltammograms of carbon materials in M, and unexpectedly, G-S, are electrochemically more active
presence of hydrogen peroxide. Operating conditions: than AC-P. In the case of G-S, the explanation could be that
v = 10 mV/s, T = 25 °C, pH = 3.5 (HCl), CH2 O2 ¼ 25 g=L. the effect of the iron content prevails over that of surface area
(or its equivalent electrochemical parameter, the coulombic
peroxide is in the media. At 1 V potential value, the carbon charge, according to Fig. 3). To confirm the important effect
surface has been deeply oxidized by the effect of the going of iron on the electrochemical carbon capacity, the cyclic vol-
sweep and the created SOGs contribute to a more effective tammetry response of G-S-HCl sample was compared to that
hydrogen peroxide oxidation upon the reverse scan. Then, of G-S in Fig. 5 (Fe content equal to 0.02 and 0.44 wt%, respec-
currents registered are higher than those values measured tively, see Table 1). The decrease in the reduction Eonset and the
in the going sweep, resulting in the first hysteresis loop. The increase in the oxidation one of G-S-HCl with respect to the
irreversible oxidation of the carbon surface upon cyclic vol- starting G-S (see values in Table 2) indicate a diminished
 CARBON 6 0 (2 0 13 ) 7 6–83 81

0.10 AC-P
0.10 G-S
AC-P-HNO3
G-S-HCl
0.05
0.05

0.00

I (mA)
I (mA)

0.00
-0.05

-0.05 -0.10

-0.10
-0.15

-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0

V (E)
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
Fig. 7 – Effect of surface oxygen groups content on the cyclic
V (E)
voltammetry response of AC-P carbon. Operating
Fig. 5 – Effect of Fe content on the cyclic voltammetry conditions: v = 10 mV/s, T = 25 °C, pH = 3.5 (HCl),
response of G–S carbon. Operating conditions: v = 10 mV/s, CH2 O2 ¼ 25 g=L.
T = 25 °C, pH = 3.5 (HCl), CH2 O2 ¼ 25 g=L.

content was reduced by more than half after HCl treatment

(see AC-M and AC-M-HCl in Table 1). The corresponding vol-
tammogram obtained in presence of hydrogen peroxide is gi-
AC-M ven in Fig. 6 and the results (see io values in Table 2) show a
0.15
AC-M-HCl decrease in the electrochemical capacity of AC-M after reduc-
0.10 tion of the ash content.
To investigate the influence of surface oxygen groups, the
0.05 AC-P carbon was oxidized with HNO3 and a voltammogram of
the resulting modified carbon (AC-P-HNO3) obtained (Fig. 7).
HNO3 oxidation substantially increased the amount of SOG
I (mA)

0.00
(see Table 1) and the voltammetry results (Fig. 7 and Table 2)
-0.05 indicate a decrease in the electrochemical capacity of the car-
bon. According to the shape of the voltammograms (Fig. 7),
-0.10 the decline is more pronounced in the reduction reaction
than in the oxidation reaction, because the former is more af-
-0.15
fected by the decrease of the electronic density of the basal
planes, caused by the presence of the acid oxygen groups.
-0.20
In conclusion, AC-M showed a higher electrochemical
activity than AC-P because of its higher ash content and lower
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 amount of acid SOG.
V (E)
3.3. Catalyst screening
Fig. 6 – Effect of ash-content reduction on the cyclic
voltammetry response of AC-M carbon. Operating
The decomposition of hydrogen peroxide is assumed to fol-
conditions: v = 10 mV/s, T = 25 °C, pH = 3.5 (HCl),
low pseudo-first-order kinetics [5,7,8,14]:
CH2 O2 ¼ 25 g=L.
dCH2 O2
 ¼ kd  CH2 O2 ð2Þ
dt
electrochemical capacity of the former, supporting the initial where kd represents an apparent kinetic constant which
hypothesis. includes several parameters such as temperature, catalyst
In the case of AC-M and AC-P, the iron content must have concentration, and pH. Fig. 8 shows the plot of (ln CH2 O2 ) vs.
only a negligible effect because it is very low in both cases. reaction time for the carbon materials tested. The values of
However, other features such as the ash content and the the apparent kinetic constant are given in Table 3 together
amount of surface oxygen groups could explain their differ- with the corresponding correlation coefficients which con-
ent electrochemical behavior. AC-M is the carbon tested with firm the validity of the simple rate equation used. According
highest ash content by far (see Table 1). Among the elements to these results, the activity of the carbons tested follows
identified by TXRF were Fe, Ti, Ca and Cu. The ash the same order as that previously found using voltammetry
82 CARBON 6 0 ( 2 0 1 3 ) 7 6 –8 3

10.0 10.0

9.5 9.5

Ln CH2O2 (C in mg/L)
Ln CH2O2 (C in mg/L)

9.0 9.0

8.5 CB-V 8.5

AC-M CB-C
AC-M-HCl G-S
AC-P G-S-HCl
8.0 AC-P-HNO3 G-F 8.0

0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
t (h) t (h)

Fig. 8 – First-order plots of hydrogen peroxide decomposition with the carbon materials. Operating conditions: T = 80 °C,
pH = 3.5 (HCl), Ccarbon = 5 g/L CH2 O2 ¼ 25 g=L.

measurements. Also, the activity of the modified carbons is

Table 3 – Values of the apparent first-order rate constant for lower than those of their corresponding starting ones
hydrogen peroxide decomposition with the carbons tested.
(Table 3).
Sample kd (h1) r2 A linear relationship has been found between the exchange
current (io) (Table 2) and the decomposition rate constants (kd)
AC-M 1.005 0.993
AC-M-HCl 0.637 0.986 (Table 3) as shown in Fig. 9. As can be seen, this relationship
AC-P 0.571 0.983 also applies to the modified carbons. Therefore, the electro-
AC-P-HNO3 0.370 0.995 chemical capacity of carbons, which is a measurable property
CB-C 0.032 0.999 in cyclic voltammetry, can be used to predict the activity of
CB-V 0.424 0.978 the carbon materials in hydrogen peroxide decomposition.
According to these results, and further investigations being
G-S 0.869 0.997
carried out in this laboratory, the outlook for the potential
G-S-HCl 0.203 0.995
G-F 0.049 0.966 application of this technique to evaluate the activity of carbon
materials in redox reactions in general is quite promising.

4. Conclusions
1.2
It has been proved that cyclic voltammetry provides a rapid,
1.0 AC-M direct, and useful technique for predicting the catalytic activ-
G-S
ity of carbon materials for hydrogen peroxide decomposition.
0.8 The exchange current, determined by the voltammograms, is
an overall variable governing carbon activity. A linear depen-
kd (h )
-1

AC-M- HCl
0.6 dence has been obtained between both.
AC-P
Cyclic voltammetry also enables a greater insight into the cat-
0.4 CB-V alytic behaviour because it becomes clear that the limiting step in
AC-P-HNO3 hydrogen peroxide decomposition with carbon materials, is the
0.2 G-S-HCl regeneration of the active sites upon carbon reduction.
Applying the electrochemical characterization to a num-
G-F
0.0
CB-C ber of carbon materials of widely different physico-chemical
0.00 0.05 0.10 0.15 0.20 0.25 0.30 characteristics provided an understanding of the effects of
io (A/g) those features on the catalytic activity for hydrogen peroxide
decomposition. The results indicate that the most important
Fig. 9 – Relationship between the exchange current (io) and factor in the catalytic activity is the content of metals, in par-
apparent first-order rate constant (kd) for hydrogen peroxide ticular iron, this is followed by the specific surface area and
decomposition with the carbons tested. finally the content of surface oxygen groups.
 CARBON 6 0 (2 0 13 ) 7 6–83 83

Acknowledgments [9] Pakula M, Biniak S, Swiatkowski A. Chemical and

electrochemical studies of interactions between iron(III) ions
The authors wish to thank the Spanish MICINN for the finan- and an activated carbon surface. Langmuir 1998;14:3082–9.
[10] Rey A, Faraldos M, Bahamonde A, Casas JA, Zazo JA,
cial support for the projects CTQ2008-03988/PPQ, CTQ2010-
Rodriguez JJ. Role of the activated carbon surface on catalytic
14807 and S2009/AMB-1588. Our gratitude to the anonymous
wet peroxide oxidation. Ind Eng Chem Res 2008;47:8166–74.
reviewer for his/her helpful comments and recommendations [11] Aguinaco A, Pablo Pocostales J, Garcia-Araya JF, Beltran FJ.
which have significantly contributed to improve the quality of Decomposition of hydrogen peroxide in the presence of
the paper. activated carbons with different characteristics. J Chem
Technol Biotechnol 2011;86:595–600.
[12] Taran O, Polyanskaya E, Ogorodnikova O, Kuznetsov V,
Appendix A. Supplementary data Parmon V, Besson M, et al. Influence of the morphology and
the surface chemistry of carbons on their catalytic
Supplementary data associated with this article can be found, performances in the catalytic wet peroxide oxidation of
in the online version, at http://dx.doi.org/10.1016/j.carbon. organic contaminants. Appl Catal, A Gen 2010;387:55–66.
2013.03.058. [13] Voitko KV, Whitby RLD, Gun’ko VM, Bakalinska OM, Kartel
MT, Laszlo K, et al. Morphological and chemical features of
nano and macroscale carbons affecting hydrogen peroxide
decomposition in aqueous media. J Colloid Interface Sci
R E F E R E N C E S
2011;361:129–36.
[14] Huang HH, Lu MC, Chen JN, Lee CT. Catalytic decomposition
of hydrogen peroxide and 4-chlorophenol in the presence of
[1] Gomes HT, Miranda SM, Sampaio MJ, Silva AMT, Faria JL. modified activated carbons. Chemosphere 2003;51:935–43.
Activated carbons treated with sulphuric acid: catalysts for [15] Quintanilla A, Casas JA, Rodriguez JJ. Hydrogen peroxide-
catalytic wet peroxide oxidation. Catal Today 2010;151:153–8. promoted-CWAO of phenol with activated carbon. Appl
[2] Kimura M, Miyamoto I. Discovery of the activated-carbon Catal, B Environ 2010;93:339–45.
radical AC+ and the novel oxidation-reactions comprising the [16] Pozio A, De Francesco M, Cemmi A, Cardellini F, Giorgi L.
AC/AC+ cycle as a catalyst in an aqueous-solution. Bull Chem Comparison of high surface Pt/C catalysts by cyclic
Soc Jpn 1994;67:2357–60. voltammetry. J Power Sources 2002;105:13–9.
[3] Lücking F, Köser H, Jank M, Ritter A. Iron powder, graphite [17] Hutchings GJ, Carrettin S, McMorn P, et al. Cyclic
and activated carbon as catalysts for the oxidation of 4- voltammetry as a potential predictive method for supported
chlorophenol with hydrogen peroxide in aqueous solution. nanocrystalline gold catalysts for oxidation in aqueous
Water Res 1998;32:2607–14. media. In: Bozell JJ, Patel MK, editors. Renewables for the
[4] Oliveira LCA, Silva CN, Yoshida MI, Lago RM. The effect of H2 Production of Chemicals and Materials. ACS Symposium
treatment on the activity of activated carbon for the Series, vol. 921. 2006. pp. 82–98.
oxidation of organic contaminants in water and the H2O2 [18] Zhu J, Yang X, Xu X, Wei K. Cyclic voltammetry for predicting
decomposition. Carbon 2004;42:2279–84. oxidation process in heterogeneous catalysis. Z Phys Chem
[5] Georgi A, Kopinke FD. Interaction of adsorption and catalytic 2006;220:1589–94.
reactions in water decontamination processes. Part I. [19] Quintanilla A, Garcia-Rodriguez S, Dominguez CM, Blasco S,
Oxidation of organic contaminants with hydrogen peroxide Casas JA, Rodriguez JJ. Supported gold nanoparticle catalysts
catalyzed by activated carbon. Appl Catal, B Environ for wet peroxide oxidation. Appl Catal, B Environ
2005;58:9–18. 2012;111:81–9.
[6] Santos VP, Pereira MFR, Faria PCC, Orfao JJM. Decolourisation [20] Zazo JA, Casas JA, Mohedano AF, Rodriguez JJ. Catalytic wet
of dye solutions by oxidation with H2O2 in the presence peroxide oxidation of phenol with a Fe/active carbon
of modified activated carbons. J Hazard Mater catalyst. Appl Catal, B Environ 2006;65:261–8.
2009;162:736–42. [21] Hernandez-Fernandez P, Baranton S, Rojas S, Ocon P, Leger J,
[7] Rey A, Zazo JA, Casas JA, Bahamonde A, Rodriguez JJ. Fierro JLG. Insights into the effects of functional groups on
Influence of the structural and surface characteristics of carbon nanotubes for the electrooxidation of methanol.
activated carbon on the catalytic decomposition of hydrogen Langmuir 2011;27:9621–9.
peroxide. Appl Catal, A Gen 2011;402:146–55. [22] Zazo JA, Bedia J, Fierro CM, Pliego G, Casas JA, Rodriguez JJ.
[8] Khalil LB, Girgis BS, Tawfik TAM. Decomposition of H2O2 on Highly stable Fe on activated carbon catalysts for CWPO upon
activated carbon obtained from olive stones. J Chem Technol FeCl3 activation of lignin from black liquors. Catal Today
Biotechnol 2001;76:1132–40. 2012;187:115–21.