Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Journal of The Electrochemical Society, 166 (15) H783-H790 (2019) H783

0013-4651/2019/166(15)/H783/8/$38.00 © The Electrochemical Society

A Facile Route to Synthesize a TiNT-RuO2 Electrocatalyst for


Electro-Generated Active Chlorine Production
D. Elizarragaz,1,= J. E. Carrera-Crespo,2,= J. Vazquez-Arenas, 3,z
I. Romero-Ibarra,4
E. J. Ruiz-Ruiz, 1,z and I. Chairez2
1 Universidad Autónoma de Nuevo León, UANL, Facultad de Ciencias Químicas, Cd. Universitaria, San Nicolás de los
Garza, Nuevo León, C.P. 66455, México
2 Unidad Profesional Interdisciplinaria de Biotecnología, Instituto Politécnico Nacional, La Laguna Ticomán,
07340 Ciudad de México, México
3 Centro Mexicano para la Producción más Limpia, Instituto Politécnico Nacional, Col. La Laguna Ticomán,
07340 Ciudad de México, México
4 Unidad Profesional Interdisciplinaria en Ingeniería y Tecnologías Avanzadas-Instituto Politécnico Nacional. Gustavo
A. Madero, C.P. 07340 Ciudad de México, México

This study aimed to fabricate a novel TiNT-RuO2 electrocatalyst to produce active chlorine with the capacity to oxidize recalcitrant
organic matter. Titanium oxide nanotubes (TiNTs) were first grown via Ti foil anodization in potentiostatic mode in an ethylene glycol
solution containing H2 O and NH4 F and then annealed at 450°C in air. RuO2 nanoparticles were deposited on the TiNTs using a hybrid
strategy that included pulsed electrodeposition of the Ru nanoparticles, followed by thermal treatment. X-ray diffraction (XRD) only
revealed peaks related to metallic Ti and anatase (TiO2 ) for the TiNTs, whereas additional peaks associated with the tetragonal
RuO2 structure were shown for TiNT-RuO2 . Scanning electron microscopy showed a well-organized structure for the TiNTs with an
average inner diameter of approximately 185 nm for the NTs, while RuO2 nanoparticles (size <10 nm) were homogenously deposited
on the walls and around the openings of the TiNTs. TiNT-RuO2 preferred the active chlorine formation over oxygen evolution owing
to the presence of the RuO2 nanoparticles. Rapid degradation testing revealed complete elimination of picloram (PCL) in 2 h when
10 mA cm−2 was imposed on the TiNT-RuO2 , whereas 48% of the 2,4-dichlorophenoxyacetic acid (2,4-D) was removed under the
same conditions using the electro-generated active chlorine on the TiNTs.
© 2019 The Electrochemical Society. [DOI: 10.1149/2.0071915jes]

Manuscript submitted June 7, 2019; revised manuscript received September 29, 2019. Published October 25, 2019.

The formation of electro-generated active chlorine on anodes has 2 to 7.10 Additionally, this oxidant is generated in a large amount and
gained renewed interest as an indirect oxidant of recalcitrant organic with adequate oxidation potential for environmental applications, as
compounds (e.g., dyes, pesticides, and pharmaceutical compounds) described in a recent study.11 Of note, in this pH range, Cl2 has not
in chloride media.1–4 The chlorine mechanism (CM) typically relies been detected on the electrode surface by differential electrochemical
on the strength of three oxidizing species (Cl2 /Cl − E°° = 1.36 V vs. mass spectroscopy (DEMS).
SHE, HClO/Cl − E° = 1.49 V vs. SHE, and ClO− /Cl − E ° = 0.89 V In general, the electroactive phase of DSA is typically deposited
vs. SHE), predominating at different pH values. Importantly, the oxi- on a Ti substrate, which has been preferred over other surfaces owing
dizing power of electro-generated chlorine species (i.e., adsorbed on to its excellent mechanical, thermal, and stability features,12,13 which
the anode surface) contrasts against typical aqueous chlorine species, have been derived from multiple nano- and micro-structures with a
which are very stable in solution. Confirming this idea, the destruc- rich diversity of textural and electronic properties.14,15 Specifically,
tion of crystal violet was considerably higher when active chlorine was titanium oxide nanotubes (TiNTs) have stood out for electrocatalytic
electro-generated on an electrode surface than when directly incorpo- applications owing to their large specific surface area, semi-conductive
rated as sodium hypochlorite into the solution at similar concentration features, adsorption capacity, adherence, and dispersion of different
conditions.5 This drew attention since similar activities could be ex- active phases, among other reasons.14,16 Accordingly, they have be-
pected for these chlorine species, presenting analogous structures and come a typical substrate for the design and development of elec-
thus suggesting an essential distinction between soluble and electro- trocatalysts, including Au,17 Pt,18,19 Pd,20,21 Sb-doped SnO2 ,22,23 Ni
generated (adsorbed) active chlorine species. In this direction, De Bat- nanoparticles,24 Ni nanoparticles with Cu,25 and β-PbO2 /α-PbO2 /Sb–
tisti group6 theoretically proposed a variation of the mechanism for SnO2 .26 However, RuO2 coatings on TiNTs have been barely evaluated
electrochemical incineration of organics proposed by Comninellis,7 for the destruction of organic-persistent compounds via the electro-
which is applicable to active chlorine in which chloro and oxychloro catalytic CM. A recent study carried out thermal decomposition to
radicals “co-adsorbed” (HOClads ) on the anode surface play impor- deposit RuO2 on anodized TiNTs, and the electrode was only tested
tant roles in the pollutant mineralization, rather than soluble chlorine for electrochemical chlorine generation.27 Nevertheless, this synthesis
species, along with the co-electrosorption of hydroxyl radicals. method favors the formation of a RuO2 coating to obstruct the mouth of
The discovery of RuO2 -based dimensionally stable anodes (DSAs) the TiO2 NTs, as confirmed by scanning electron microscopy (SEM)
for chlor-alkali cells triggered research on low cost and durable mate- images. A more suitable method to considerably increase the number
rials, emphasizing the role of the anode phase for this electrocatalytic of electroactive sites formed by RuO2 would be the electrocrystalliza-
application.8,9 DSAs are typically constituted by a thin film made of a tion of nanoparticles, which could benefit from the large specific sur-
conductive oxide (i.e., RuO2 ) deposited on Ti, and their shape does not face area of TiNTs and the adhesion and compatibility between these
change when subjected to severe anodic attacks during electrolysis.9 two phases because TiO2 and RuO2 crystals share similar tetragonal
This technology has somehow consolidated over the years; thus, the structures and have close lattice parameters.28 This method could also
exploration of new DSAs started to focus on investigating electrocat- be more controllable, and this electrocatalyst could be quickly fabri-
alytic aspects concerning the active chlorine generation for pollutant cated on an industrial scale by using a DC power supply. In addition,
destruction. Although this mechanism could involve any of the three the oxidation ability of chlorine species produced from TiNT-RuO2
aforementioned oxidants depending on the pH, attention has been par- needs to be assessed against recalcitrant compounds, which has not
ticularly paid to the formation of HClOads species in the pH range of been conducted thus far. Therefore, we fabricated for the first time
a TiNT-RuO2 electrocatalyst using only electrochemical methods to
yield active chlorine species with the capacity to oxidize recalcitrant
=
These authors contributed equally to this work. organic matter. To avoid clogging formation over the mouth of the
z
E-mail: jorge_gva@hotmail.com; edgar.ruizrz@uanl.edu.mx TiO2 NTs, the RuO2 nanoparticles were synthesized using a hybrid

Downloaded on 2019-10-25 to IP 187.189.114.208 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
H784 Journal of The Electrochemical Society, 166 (15) H783-H790 (2019)

strategy consisting of Ru nanoparticle nucleation on a TiO2 support voltage of 5.0 kV and equipped with energy dispersive X-ray spec-
via pulse potential techniques, followed by thermal treatment to ob- troscopy (EDS) capabilities for elemental analysis. The crystal struc-
tain the RuO2 nanoparticles. The idea is to provide a rapid alternative ture of the electrodes was determined using an X-ray diffractometer
to the typical chemical synthesis of these catalysts using a sol-gel or (Bruker D-8 Advance with Cu-Kα radiation) in the grazing incidence
polymeric precursor (Pechini), which generally requires days to fab- X-Ray diffraction (GIXRD) geometry by fixing the incident beam an-
ricate a single electrode. Additionally, RuO2 is efficiently deposited gle at 5.0°. X-ray data was analyzed using HighScore Plus software
using a very low load of this metal, which reduces costs compared V.3.0.5 (PANalytical B.V., Almelo, Netherlands). The electrochemical
to that of traditional synthesis methods (e.g., sol-gel, hydrothermal, characterization was carried out in a VMP3 (Bio-Logic Science In-
and Pechini). TiNTs were used as the substrate rather than a typi- strument) potentiostat/galvanostat controlled by EC-Lab software that
cal flat TiO2 surface during electrocatalysis to enable Ru deposition was equipped with a frequency response analyzer (FRA) by employ-
and provide mechanical integrity to the dimensionally stable anode. ing a three-electrode cell (200 mL) and using TiNTs or TiNT-RuO2 as
Although Ti oxides are typically not electrocatalytic for most electro- the working electrode and a graphite rod (30.0-cm length × 1.016-cm
chemical reactions, they can alter the dielectric properties of the films diameter, Alfa Aesar, 99.997%) and Ag/AgCl (0.197 V vs. SHE) as
deposited on them.29 Then, the electrocatalyst was micro-structurally the counter and reference electrodes, respectively. Cyclic voltamme-
characterized by XRD and texturally characterized using SEM. An try (CV) was conducted in 0.05 M NaCl at 20 mV s−1 to evaluate
electrochemical characterization was conducted in 0.05 M NaCl to the electrochemical activity of the electrocatalysts toward the chlo-
evaluate the eventual production of active chlorine and to perform rine evolution reaction (CER) mechanism. The O2 (air) dissolved in
rapid testing concerning the degradation of a commercial formulation the electrolyte was not removed. Electrochemical impedance spec-
of Tordon 101 comprised of a mixture of the herbicides 2,4-D (39.2% troscopy (EIS) characterization of the electrodes was performed in a
w/v) and PCL (10.2% w/v) (molecule test), and excipients (50.2% w/v, 0.05-M Na2 SO4 solution at 1.2 V vs. Ag/AgCl with an AC perturbation
reported as water, ethanol, and surfactants). The concentration decay of ± 10 mV in the frequency range of 100 kHz to 0.1 Hz.
of the initial molecules and the intermediate formation arising from The concentration of active chlorine species was tracked by iodo-
the degradation process were tracked using high-performance liquid metric methods according to Refs. 1,2,11. Aliquots sampled from the
chromatography (HPLC). electrochemical cell were added to a quartz cell containing potassium
iodide (0.1 M) and ammonium heptamolybdate (0.01 M). The ab-
sorbance was recorded after 2 min of reaction at 350 nm using a Cary
Materials and Methods
5000 UV-Vis-NIR spectrophotometer.
Chemicals.—PCL and 2,4-D analytical standards were supplied by
Sigma Aldrich, while the commercial formulation of these herbicides
(Tordon 101) was supplied by Dow AgroSciences. Ruthenium chlo- Electrochemical degradation via active chlorine.—The produc-
ride (RuCl3 ), ammonium fluoride (NH4 F), ethylene glycol, ethanol, tion of active chlorine on the TiNT-RuO2 electrode was tested for the
sodium chloride (NaCl), potassium perchlorate (KClO4 ), sodium sul- degradation of Tordon 101, which is a commercial formulation con-
fate (Na2 SO4 ), potassium iodide (KI), and ammonium heptamolybdate taining the herbicides 2,4-D (39.2% w/v) and PCL (10.2% w/v) and
were analytical grade reagents and used as received. All solutions were excipients reported as isopropanol (5.0% w/v), alkylphenol alkoxylate
prepared with ultrapure water (Milli-Q Direct 8 system, Millipore, (5.2% w/v), triisopropanolamine (1.3% w/v), and other related com-
18 M·cm). A titanium foil (Alfa Aesar, 0.25-mm thickness, 99.5% pounds (39.1% w/v).30 The degradation was carried out via in situ
purity) was used to fabricate the electrodes. electrochemical generation of the active chlorine species in a reactor
containing 150 mL of 0.05 M NaCl (pH 5.8) and an initial concen-
Preparation of electrodes.—TiNTs and TiNT-RuO2 were fabri- tration of Tordon 101 that was equivalent to 30 and 144 mg L–1 of
cated with an exposed area of 3.0 cm2 using potentiostatic anodization PCL and 2,4-D, respectively. The imposed current density was tested
of Ti foils at 70 V for 50 min in a two-electrode cell by employing at 10 mA cm–2 for 120 min, and sampling was conducted every 20 min
a Pt wire (Alfa Aesar, 99.99%) as the counter electrode in an ethy- to analyze the concentration decay of the pesticides and evolution of
lene glycol solution containing H2 O (15 wt%) and NH4 F (0.3 M). The subproducts (i.e., formic and oxalic acids). To evaluate the degradation
TiNTs formed were washed with ultrapure water and ethanol and dried performance of electro-generated active chlorine (i.e., that formed on
for 30 min at room temperature; then, they were annealed at 450°C the electrocatalyst), a control degradation test was conducted using
for 30 min in air using a ramp temperature of 5°C min−1 . To de- 1 M hypochlorous acid (i.e., soluble chlorine) for Tordon 101 under
posit RuO2 nanoparticles on the TiNTs using a hybrid strategy, TiNTs the same conditions as established above (reactor volume of 150 mL,
were immersed in a 70:30 (% v/v) H2 O:ethanol KClO4 saturated so- initial concentrations of 30 and 144 mg L–1 for PCL and 2,4-D, re-
lution that contained 0.005 M RuCl3 . The pulsed electrodeposition of spectively). The pH of the initial solution was adjusted to 5.8 with
the Ru nanoparticles on the TiNTs was conducted in a conventional HCl.
three-electrode cell with TiNTs as the working electrode and a glassy The decay of PCL and 2,4-D concentrations was monitored by
carbon plate and Ag/AgCl electrode, as the counter and reference elec- HPLC using a Hyperclone C-18 reversed phase as the stationary phase
trodes, respectively, using the following conditions: Eon = −3 V vs. (250 mm × 4.6 mm, 5-μm particle size, from Phenomenex), a mixture
Ag/AgCl, ton = 0.3 s, Eoff = OCPinitial , toff = 1 s for 300 cycles. Subse- of acidified H2 O with H3 PO4 at pH 2.4, and acetonitrile (60:40 V/V) in
quently, the TiNTs-Ru were washed with ultrapure water and dried at isocratic elution mode as the mobile phase at 0.8 mL min−1 . The PCL
room temperature for 30 min and then annealed in an air atmosphere and 2,4-D compounds presented maximum absorbance at 224 nm;
at 450°C for 30 min using a ramp temperature of 5°C min−1 to fi- therefore, this wavelength was used to detect these compounds at dif-
nally obtain the TiNT-RuO2 electrocatalyst. The total time required ferent retention times. The injection volume and temperature of the
for the electrocatalyst synthesis was approximately 18 h. Of note, the column were 20 μL and 25°C, respectively.
above procedure guarantees the homogeneity and good dispersion of The evolution of organic acids was monitored by HPLC using
the RuO2 phase, unlike depositing Ru on amorphous TiNTs (with- a Prevail organic acids column (150 mm × 4.6 mm, 5-μm particle
out calcination) and then applying the thermal treatment to both. This size, from Alltech-Grace), and phosphate buffer (0.25 M at pH 2.5) as
pathway would generate a heterogeneous deposition of Ru owing to the mobile phase in isocratic elution mode at 0.8 mL min–1 . Formic
the absence of crystallinity (i.e., poor electronic conductivity) in the (Carbon oxidation state, +II) and oxalic (Carbon oxidation state, +III)
NTs. acids were detected at 200 nm. The formation of these intermediates
during the degradation of PCL and 2,4-D indicated that the oxidation
Structural and electrochemical characterization of anodes.— of these contaminants was close to full mineralization (CO2 and H2 O).
SEM images were obtained for the TiNT and TiNT-RuO2 anodes with The injection volume and temperature of the column were 30 μL and
a high-field emission microscope (JMS-7600F) using an accelerating 25°C, respectively.

Downloaded on 2019-10-25 to IP 187.189.114.208 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
Journal of The Electrochemical Society, 166 (15) H783-H790 (2019) H785

general, the anatase phase was favored when the calcination temper-
a) TiNTs b) TiNTs-RuO2
ature in the air atmosphere was below 600°C.31,32 These signals were
also detected in the GIXRD pattern obtained for TiNT-RuO2 with
similar intensities, revealing the non-agglomerated coating of the ac-
tive phase. Thus, the peaks associated with the anatase phase dimin-
ished when the GIXRD geometry was used owing to the presence
of large agglomerates of the deposited phases (i.e., CdS) on top of
TiO2 nanotubes.33 Likewise, the crystal structure of TiO2 (pdf 00-
021-1272) in the TiNT-RuO2 was predominantly the anatase phase as
c) TiNTs d) TiNTs-RuO2 previously discussed, indicating that RuO2 loading does not change
the crystal structure of the TiNTs.34 Thus, three small peaks at 2θ =
28.0°, 35°, and 54° were measured, corresponding to (110), (011),
and (211) diffraction planes of RuO2 , respectively (tetragonal struc-
ture) (pdf 00-040-1290).35 The samples were analyzed using GIXRD
owing to the low amount of RuO2 , which hampered its detection in
the diffractograms obtained via Bragg–Brentano arrangement, where
only the peaks associated with metallic Ti and TiO2 were observed.
Figure 1. SEM images of synthesized TiNTs and TiNTs-RuO2 collected at No other peaks related to an additional phase were observed in this
different magnifications: a, b) 50.0 and c, d) 500.0 K. study, confirming the synthesis of the TiNT-RuO2 electrocatalyst via
the facile route to decorate RuO2 nanoparticles on the TiNTs. Note that
the crystalline structure (i.e., tetragonal) of RuO2 similarly adopted the
structure of the TiO2 anatase phase (pdf 00-021-1272) that was used
Results and Discussion as the substrate. This showed the clear effect of the substrate on the
Electrode characterization.—Figure 1 shows the surface of the catalysis of the electroactive phase.
as-synthesized TiNT (a, c) and RuO2 -TiNT (b, d) anodes at different
SEM magnifications. In particular, Figs. 1a and 1c show the well- Electrochemical characterization.—The voltammetry characteri-
organized structure of the TiNTs with inner diameters ranging from zation was initially performed with the TiNTs and TiNT-RuO2 electro-
180 to 190 nm. Figures 1b and 1d reveal the nanoparticles of RuO2 catalysts to analyze the oxidation processes occurring on the surfaces.
(sizes of <10 nm) deposited on the walls and around the mouths of the Figure 4a shows the voltammograms measured from the initial OCP
TiNTs without blocking them, unlike a recent study.27 It is of primary (0.026 V vs. Ag/AgCl) to the positive direction for bare TiNTs in
concern to achieve this distribution during plating to guarantee the 0.05 M NaCl and in 0.05 M Na2 SO4 solutions under stagnant condi-
mechanical and structural integrity of the nanotubes during use and to tions, where it was clear that significant faradaic contributions (i.e.,
increase the electroactive area of the catalyst by decreasing the particle charge transfer) were not observed in any of these electrolytes within
size, assuming some RuO2 nanoparticles are deposited in the inner and the evaluated potential window. In fact, when the potential scan was in-
outer walls of the TiNTs. Figure 2 shows the EDS analysis conducted verted in the negative direction, no peaks associated with the reduction
of the TiNT-RuO2 , which resulted in the following average elemental of recently oxidized species were detected in the curves. Most likely,
composition of the surface: oxygen (O K 40.2 wt%), titanium (Ti this behavior was due to the semi-conductive features of TiNTs be-
K 54.6 wt%), and ruthenium (Ru L 3.1 wt%). Fluor (F K 1.4 wt%) cause they generally are poor electronic conductors. However, TiNTs
was also found as a synthesis residue because the TiNTs were grown play an important role as substrates that modify the dielectric prop-
using an electrolyte containing NH4 F. This compositional analysis erties of the top film (i.e., RuO2 ), as revealed by an in-depth analysis
confirmed the elements contained in the catalyst, and in particular, the of the composition gradients in Ti/SnO2 -Sb2 O5 anodes composed of
good dispersion and very low load of Ru, as displayed by its elemental multiple layers of Sn and TiO2 .29 Additionally, this substrate confers
mapping and low-intensity peaks in the EDS spectrum, respectively mechanical and electrical stabilities for producing a dimensionally
(Fig. 2). stable anode. However, when the potential sweep was initiated in the
The grazing incidence X-ray diffraction (GIXRD) patterns of the positive direction over TiNT-RuO2 in sulfate media, an increase in
TiNTs and TiNT-RuO2 are displayed in Fig. 3. As observed for bare current density was observed at approximately 0.8 V (vs. Ag/AgCl),
TiNTs, only peaks associated with metallic titanium and the anatase which has been associated with H2 O oxidation to form O2 , because
phase of TiO2 (pdf 00-021-1272) were detected. No peaks related there are no other species to overcome oxidation in this potential range.
to rutile (TiO2 ) were found because the calcination temperature was Under this condition, it is not expected that the RuO2 phase will mod-
450°C, where the anatase phase (TiO2 ) preferentially predominates. In ify the electronic conductivity of the TiNTs; rather, it acts as an active

Figure 2. a) EDS analysis and b) mapping for Ru element in the synthetized TiNTs-RuO2 .

Downloaded on 2019-10-25 to IP 187.189.114.208 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
H786 Journal of The Electrochemical Society, 166 (15) H783-H790 (2019)

Figure 3. Grazing incidence X-Ray diffraction patterns of synthesized TiNTs and TiNTs-RuO2 , obtained by fixing the incident beam angle at 5.0°.

site for electrochemical reactions. The occurrence of the oxygen evo- inversion potential owing to the facile desorption of chlorine species
lution reaction (OER) on the RuO2 phases in sulfate electrolytes have from the TiNT-RuO2 surface. Accordingly, a voltammogram was
been previously reported in the literature.1–3,11,13 When the potential recorded under the same experimental conditions but with constant
is reversed toward the cathodic direction, reduction waves are not de- stirring. Surprisingly, the cathodic wave disappeared when the volt-
tected because the TiNT-RuO2 phase is not catalytic to reduce O2 age was inverted at 1.7 V (Fig. 4b - stirred) owing to the convection
(i.e., dissolved into the electrolyte from air). When this material is applied to the electrode, confirming the facile desorption of the elec-
exposed to 0.05 M NaCl, the onset voltage of the oxidation process trogenerated active chlorine species from the electrocatalyst. Note that
is displaced to more positive values (∼1 V) compared to that of the some of these adsorbed species can be in gaseous form, whereby their
OER in sulfate media (refer to Fig. 4a). A comparison with a Ti/RuO2 separation from the electrode surface can readily occur. Additionally,
anode synthesized via the polymeric precursor (Pechini) method re- the TiNT-RuO2 electrode remained stable after undergoing different
vealed that the onset potential for chloride oxidation was positively CV cycles (not shown), which demonstrated its stability and main
displaced by approximately 100 mV more positive (∼1.1 V, Fig. 4b in denomination as a “dimensionally stable anode”.
Ref. 2). However, the current density significantly surpassed the one To understand the effect of RuO2 on the electrical properties of
for our electrode material owing to the considerably larger loading of TiNTs, EIS measurements were performed in 0.05 M Na2 SO4 under
RuO2 , as observed by the very intense peaks for this metal oxide in the the potentiostatic mode at 1.2 V (vs. Ag/AgCl) for both TiNT-based
diffractogram of the Ti/RuO2 Pechini sample obtained via the Bragg– electrodes and at 0.6 V (vs. Ag/AgCl) only for the TiNT-RuO2 elec-
Brentano configuration.2 The occurrence of the anodic processes de- trode. The first potential was selected because it enables the charac-
scribed in Fig. 4a suggests that the main reaction mechanism occurring terization of the H2 O oxidation without the production of a significant
on the electrode surface was modified; in fact, the slope of the voltam- amount of bubbles from O2 evolution. Figure 4c shows the Nyquist
mogram was steeper than the one obtained for Na2 SO4 . Under this plots recorded for bare TiNTs and TiNT-RuO2 electrodes at 1.2 V
premise, active chlorine formation presumably occurs once the cur- (vs. Ag/AgCl), where the arc radius of the TiNTs abruptly decreased
rent flows on the catalyst/electrolyte interface, which was confirmed when the nanotubes were decorated with RuO2 nanoparticles; this in-
when the potential was inverted in the negative direction (Fig. 4a), dicated that this phase significantly improved the charge transfer to
where a peak related to the reduction of the active chlorine species perform the OER, as previously shown in Fig. 4a. The experimental
was recorded. According to the bulk pH of the electrolyte (5.8), this EIS spectra were fitted using the ZView software using the equiv-
peak could be attributed to reduction of HClOads , according to recent alent electric circuit (eec) shown in the inset of Fig. 4c. This eec
studies of the CER mechanism carried out with DEMS11 and the frac- has been previously reported for the analysis of electrodes based on
tion diagram presented in Fig. 8. The OER began at some point during TiO2 nanostructures,33,36–38 where Rs is the solution resistance, Rf is
the oxidation scan as a parasitic reaction. The reaction mechanism for the charge transport resistance in the internal film (TiNTs or TiNT-
the occurrence of this chlorine species will be discussed in Degrada- RuO2 ), and Rct is the charge transfer resistance between the electrode
tion of PCL and 2,4-D and evolution of its byproducts section. surface and the electrolyte. Qsc and Qdl are constant phase elements
To confirm the occurrence of the cathodic peak in Fig. 4a, cyclic (CPE) associated with the space charge capacitance and the double
voltammograms were recorded in chloride media at different anodic layer capacitance, respectively. As observed in Table I, when TiNTs
inversion potentials of 1.2, 1.6, and 1.7 V (vs. Ag/AgCl) under stag- were coupled with RuO2 nanoparticles, the Rct decreased four orders
nant conditions (Fig. 4b). As observed, the cathodic wave was only of magnitude compared to the values obtained for the bare TiNTs,
detected in the last two inversion potentials, and it grew considerably confirming that the RuO2 noticeably improved the electrocatalyst per-
by increasing this value from 1.6 to 1.7 V (vs. Ag/AgCl), indicat- formance of the NTs. Moreover, the Rf and Rct significantly decreased
ing a higher generation of chlorine species on the electrode surface. when the potential was changed from 0.6 V to 1.2 V (vs. Ag/AgCl)
Probably, this cathodic wave was not observed at the least positive for the TiNT-RuO2 electrode, showing that a more positive potential

Downloaded on 2019-10-25 to IP 187.189.114.208 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
Journal of The Electrochemical Society, 166 (15) H783-H790 (2019) H787

Table I. Parameters obtained by fitting of the experimental EIS spectra using the equivalent electric circuit shown in the inset of Fig. 4c.

Electrode E (V vs. Ag/AgCl) Rs (Ω cm2 ) Rf (Ω cm2 ) Qdl (F sn cm−2 ) n Rct (Ω cm2 )

TiNTs 1.2 48.9 6940.8 0.4 × 10−5 0.9 4.5 × 106


TiNTs-RuO2 0.6 46.6 1.3 × 109 2.1 × 10−5 0.9 4.4 × 106
1.2 53.7 16.4 2.7 × 10−5 0.9 717.5

reduces the energy barriers of this electrode for electronic conduction 10 mA cm–2 to track the formation of oxidants in these media. Fig-
and charge transfer to perform chloride oxidation and the OER. ure 5 shows the evolution of the oxidants over 240 min, where it is
Estimation of the active chlorine production by chemical analy- evident that its concentration (absorbance) increased up to 180 min,
sis is crucial to confirm its occurrence during chloride oxidation over achieving a quasi-steady behavior after that time. This oxidant de-
TiNT-RuO2 , as shown in Fig. 4. Therefore, iodometric titration was tected in the curve was mainly associated with HClO desorbed from
performed during electrolysis conducted in 0.05 M NaCl and 0.05 M the electrocatalyst surface (i.e., HClOads ) according to the discussion
Na2 SO4 over TiNT-RuO2 by applying a constant current density of above regarding the voltammetry study. This was because oxidants
were not observed when a similar iodometric test was conducted in
Na2 SO4 , where O2 was the main oxidant (Fig. 5).

Degradation of PCL and 2,4-D and evolution of its byproducts.—


We conducted rapid testing of the electrode capacity to form active
chlorine, and its potential utilization as an oxidant for the degrada-
tion of recalcitrant compounds (i.e., pesticides) was examined. Thus,
a more detailed analysis (i.e., HPLC-MS) of the chemical composi-
tion associated with the intermediate products resulting from the PCL
and 2,4-D elimination will be the motivation of a forthcoming study.
Figure 6a shows the relative percentage decay of PCL ([PCL]initial :
30 mg L–1 ) and 2,4-D ([2,4-D]initial : 144 mg L–1 ) as a function of the
electrocatalysis time using TiNT-RuO2 as the anode and applying dif-
ferent current densities (5 and 10 mA cm–2 ). Therefore, the decrease
in the pesticide concentration as a function of the electrolysis time (t)
results from the in situ electrogeneration of active chlorine species.
These curves are displayed with their relative values (with respect to
the initial concentration), i.e., [PCL]rel = [PCL]t / [PCL]initial , [2,4-D]rel
= [2,4-D]t / [2,4-D]initial . As observed in Fig. 6a, complete degradation
of PCL was achieved in 2 h of electrocatalysis when 10 mA cm–2 was
applied on the electrode, while 90% of degradation was reached in
the same time when 5 mA cm–2 was used. However, the concentration
decays of 2,4-D were 42 and 48% when 5 and 10 mA cm–2 were ap-
plied, respectively. Under this condition, the initial degradation rates
of PCL and 2,4-D seemed to slightly depend on the current density,
although higher degradation values were somewhat observed when
j increased. These rate differences can be more clearly observed in
Fig. 6b, assuming that the degradation of PCL and 2,4-D follows a
pseudo-first order kinetics, suggesting typical mass transfer control
that is presumably associated with the transport of contaminants to
become in contact with the oxidants. The apparent constant (k) and
coefficient of determination (R2 ) for PCL were 1.8 × 10–2 min–1 (R2
= 0.9934) and 2.6 × 10–2 min–1 (R2 = 0.9795) at 5 and 10 mA cm–2 ,

Figure 4. Cyclic Voltammetric characterization conducted at 20 mV s−1 for:


a) bare TiNTs and TiNTs-RuO2 in the electrolytes indicated in the figure, and
b) TiNTs-RuO2 in an electrolyte containing 0.05 M NaCl at different inversion
potentials (Einv ). c) Nyquist plots obtained in 0.05 M Na2 SO4 at 1.2 V vs.
Ag/AgCl for bare TiNTs and TiNTs-RuO2 . Inset: zoom of Nyquist spectra Figure 5. Evolution of active chlorine species conducted in 0.05 M NaCl (blue
displaying fitted curves calculated using the equivalent electric circuit shown circles) and 0.05 M Na2 SO4 (red squares) on TiNTs-RuO2 at 10 mA cm−2 (pH
in this figure. 5.8 and 25°C), and measured by iodometric titration methods.

Downloaded on 2019-10-25 to IP 187.189.114.208 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
H788 Journal of The Electrochemical Society, 166 (15) H783-H790 (2019)

Figure 6. a) Relative decay of 2,4-D and PCL concentrations (C0 = 144 and 30 mg L−1 , respectively) during destruction with electrogenerated active chlorine
in 0.05 M NaCl (pH 5.8 and 25°C) on TiNTs-RuO2 at 5 (red line) 10 mA cm−2 (blue line); and b) pseudo-first order degradation kinetics of contaminants: PCL
(squares) and 2,4-D (triangles).

respectively; and for 2,4-D, they were 3.0 × 10–3 min–1 (R2 = 0.9310) acids as the main subproducts of PCL destruction using commercial
and 4.8 × 10–3 min–1 (R2 = 0.9927) at 5 and 10 mA cm–2 , respec- DSA (Ti/Ru0.3 Ti0.7 O2 ), where the maximum concentrations were de-
tively. From these results, the increase in the degradation rate could tected for both acids in 60 min of reaction, and their almost complete
result from the yield of the electro-generated active chlorine species, degradation was reached after 120 min of electrocatalysis. Contrary
which was mostly HOCl, according to the initial pH of 5.8.11 A similar to these results and presumably due to the 2,4-D presence, formic and
complete PCL degradation was obtained by Pereira et al.39 in chloride oxalic acids were the main subproducts of the PCL and 2,4-D degra-
media ([PCL]initial : 100 mg L–1 , NaCl 25 mM) after 150 min using dations in this study conducted on the TiNT-RuO2 electrocatalyst, al-
an initial pH = 6 and 30 mA cm–2 , although the toxic β-PbO2 was though it is important to note the differences in degradation techniques
used as the anode material. Of note, these experiments were similarly and electrocatalyst phase compared with the previous studies.
conducted in 0.05 M Na2 SO4 (plots not shown) without observing the According to the results shown above in terms of the active chlo-
PCL or 2,4-D degradations, which corroborated the absence of direct rine mechanism occurring between pH 5.8 (initial) and 4.6 (final) and
oxidation of these compounds on the electrode material. the recent studies conducted by our research group using DEMS,11
The evolutions of formic and oxalic acids were detected at 2.07 it is possible to propose the following reaction pathway over TiNT-
and 2.58 min of retention time, respectively, in the analysis of the RuO2 when HOCl predominates in solution (Fig. 8):
short chain carboxylic acids (refer to Electrochemical degradation via
active chlorine section). These intermediates are considered poten-
tially mineralizable compounds because the oxidation states of carbon RuOx + (H2 O)l → RuOx (H2 O)ads [1]
 
species (+II and +III, respectively) are close to CO2 , which should RuOx (H2 O)ads → RuOx e− − H O• ads + H + [2]
be achieved during the complete mineralization of PCL and 2,4-D.
• − −
This evaluation led to a slight change in the pH to more acidic con- RuOx − H O ads + Cl(aq) → RuOx (H OCl )ads + e [3a]
ditions (4.6 after 2 h) without influencing the HClO predominance
in the pH range from 3 to 6.10 Figure 7 shows the rapid increases in RuOx (HOCl)ads → RuOx + H OCl(aq) [4a]
both acids during the first 30 min of active chlorine production, fol- H OCl(aq) ↔ +ClO−
(aq)
+
pKa = 7.54
H(aq) [5a]
lowed by attenuated increases up to 120 min reaching concentrations  −  
of 58.04 and 24.84 mg L–1 , respectively. A. Ozcan et al. described RuOx e − H O• ads + H2 O → RuOx e− − O• ads + H + [3b]
the evolution of oxamic and glyoxylic acids in the early stages of PCL    
degradation by the electro-Fenton process and subsequent transforma- RuOx e− − O• ads + H2 O → RuOx e− − H OO• ads + H + [4b]
tions of the first two into oxalic and glycolic acids.40 However, D.A.C.    
Coledman et al.41 reported the formation of dichloroacetic and oxamic RuOx e− − H OO• ads → RuOx e− − O2ads + H + [5b]

Figure 7. Evolution of oxalic (squares) and formic (circles) acids concentra-


tion as a function of the treatment time with electrogenerated active chlorine Figure 8. Fraction of chloride species as a function of pH using a total con-
in 0.05 M NaCl (pH 5.8 and 25°C) on TiNTs-RuO2 at 10 mA cm−2 . centration of 0.05 M chloride.

Downloaded on 2019-10-25 to IP 187.189.114.208 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
Journal of The Electrochemical Society, 166 (15) H783-H790 (2019) H789

Conclusions
A TiNT-RuO2 electrocatalyst was fabricated using a facile route in-
tended to catalyze the production of active chlorine to oxidize recalci-
trant organic compounds. During the first stage, Ti foils were anodized
at 70 V and subsequently annealed at 450°C in air to obtain crystalline
TiO2 NTs. A second stage was used to deposit a low content of Ru
on the TiO2 NTs via pulsed electrodeposition, followed by thermal
treatment. Metallic Ti and anatase were detected via XRD analysis of
the bare TiNTs. Similar signals were observed in the XRD pattern for
the TiNT-RuO2 with similar intensities, indicating a non-agglomerated
and well-dispersed coating of RuO2 nanoparticles. Additionally, peaks
associated with the tetragonal RuO2 structure were found for this cat-
alyst, which showed the structure of the anatase phase used as the
substrate. SEM images revealed the well-organized structure of the
TiNTs with an average inner diameter of approximately 185 nm, and
the RuO2 nanoparticles (size <10 nm) were regularly distributed on the
Figure 9. Chemical oxidation of 2,4-D and PCL (C0 = 144 and 30 mg L−1 , walls and around the mouths of the TiNTs. EDS analysis confirmed the
respectively) performed with 1 mol L−1 of hypochlorous acid. The experiment
was conducted at pH 5.8 and 25°C.
elements in the catalyst. Thus, this fabrication procedure renders a suit-
able method to consistently spread RuO2 nanoparticles on TiNTs with-
out clogging formation over their openings, while delivering mechan-
ical and structural integrity to the electrocatalyst via the support. Ad-
ditionally, it could also be more manageable to fabricate a TiNT-RuO2
1 electrocatalyst for industrial applications by using a DC power supply.
MOx+1 → MOx + O2 [6b] The voltammograms recorded for the TiNTs in 0.05 M NaCl and
2
0.05 M Na2 SO4 showed that the pure NTs were not active for chloride
and H2 O oxidations, respectively, presumably because this material is
a poor electronic conductor. Activities toward these reactions were ob-
where RuOx represents an active site on an oxygen atom of the
tained on the TiNT-RuO2 electrocatalyst, where the CER mechanism
TiNT-RuO2 catalyst, (HOCl)ads is hypochlorous acid, (H O• ads ) is the
preferentially occurred over the OER on the active sites connected
hydroxyl radical, (O• ads ) is the oxyl, and (H OO• ads ) is the peroxyl rad-
to RuO2 . EIS measurements corroborated these results because the
ical. The subscript “ads” denotes an adsorbed intermediate species on
Nyquist spectra and the parameters obtained by fitting the experimen-
the anode surface. Additionally, the oxygen evolution reaction (OER,
tal data showed a noticeable decay in the charge transfer resistance of
Reactions 3b–6b) concomitantly arises with the occurrence of active
the TiNTs when coupled with the RuO2 nanoparticles. The estimation
chlorine (Reactions 3a–5a). The above mechanism presents two re-
of active chlorine production over this catalyst was confirmed during
markable connotations: pH dependence and electro-generated active
electrolysis in NaCl via chemical analysis using iodometric titration.
chlorine. The pH dictates the predominant active chlorine species on
Complete destruction of PCL was achieved within 2 h over TiNT-
the electrode surface and in the bulk solution, which is HOClads for
RuO2 using 10 mA cm–2 in 0.05 M NaCl (pH 5.8), whereas 48% of 2,4-
the range of pH (5.8 initially to 4.6 after 2 h of reaction) operating
D was abated under the same conditions. The generation of formic and
during chloride electrolysis. This consideration is also based on the
oxalic acids during the electrocatalytic oxidation confirmed the con-
thermodynamic diagram (fraction-pH) shown in Fig. 8 for the chlo-
version of the contaminants into potentially mineralizable compounds,
ride concentrations we considered and the DEMS analysis reported
which were compared with the results from other reports where the de-
for this pH range.11 However, the activity of electro-generated active
tection of similar byproducts was described. The control degradation
chlorine (HOClads ) producing the PCL and 2,4-D degradations in Tor-
test carried out under the same conditions as above using 1 mol L–1 of
don 101 needs to be contrasted against the activity of soluble HOCl to
hypochlorous acid (soluble chlorine at pH = 5.8) showed that the elim-
confirm in which state (adsorbed or soluble) the chlorine species is a
ination of the contaminants did not exceed 30% for each case without
better oxidant. The first evidence is provided in Fig. 4, where no ox-
forming organic acids (null mineralization), confirming the higher ox-
idants are shown to be produced in the absence of chloride oxidation
idizing power of the active chlorine species formed over TiNT-RuO2 .
on the anode (Na2 SO4 ). However, Fig. 9 shows the control degrada-
tion tests conducted in 1 M of hypochlorous acid (pH = 5.8) used to
Acknowledgments
evaluate the elimination degree of the above compounds. Note that at
this pH, HOCl thermodynamically predominates in solution accord- D. Elizarragaz and J.E. Carrera-Crespo thank CONACYT for
ing to the diagram in Fig. 8. As observed, the elimination of these the economic support to pursue the Ph.D. and postdoctoral studies,
contaminants using soluble HOCl was not significant (i.e., lower than respectively. The authors acknowledge the facilities at UAM-A
30% for both pesticides) compared to the concentration decays shown for performing the XRD (Dr. Jorge Flores) and SEM (Dr. Rosa
in Fig. 6 using electro-generated active chlorine (HOClads ), confirm- Luna) analyses. Financial support was provided by SECITI-CDMX-
ing the higher oxidizing power of the chlorine species formed over MEXICO (grant CM 8995c19 SECITI/2019), and CONACyT,
TiNT-RuO2 . Additionally, no organic acids were detected during the “Investigación Científica Básica” 2017-2018 grant No. A1-S-21608.
control test, indicating null mineralization in the presence of solu- E.J. Ruiz received financial support from the PAICYT program of the
ble HOCl. Similar results were shown during the abatement of crys- Universidad Autónoma de Nuevo León.
tal violet, which were considerably higher when active chlorine was
electro-generated on the electrode surface compared to those of the
ORCID
addition of sodium hypochlorite to the solution with similar chloride
concentrations.5 The superior oxidizing activity found for the electro- J. Vazquez-Arenas https://orcid.org/0000-0001-8298-328X
generated active chlorine (Fig. 6) against hypochlorous acid (Fig. 9) E. J. Ruiz-Ruiz https://orcid.org/0000-0001-9524-7927
during the degradations of PCL and 2,4-D confirmed the theoretical
proposal of the above reaction mechanism,6 wherein chloro and oxy-
chloro radicals “co-adsorbed” (HOClads ) on the anode surface play References
important roles in the pollutant mineralization rather than the soluble 1. R. Palma-Goyes, J. Vazquez-Arenas, R. Torres-Palma, C. Ostos, F. Ferraro, and
chlorine species. I. González, Electrochimica Acta, 174, 735 (2015).

Downloaded on 2019-10-25 to IP 187.189.114.208 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
H790 Journal of The Electrochemical Society, 166 (15) H783-H790 (2019)

2. R. E. Palma-Goyes, J. Vazquez-Arenas, C. Ostos, F. Ferraro, R. A. Torres-Palma, and 23. G. Zhao, X. Cui, M. Liu, P. Li, Y. Zhang, T. Cao, H. Li, Y. Lei, L. Liu, and D. Li,
I. Gonzalez, Electrochimica Acta, 213, 740 (2016). Environmental Science & Technology, 43, 1480 (2009).
3. R. E. Palma-Goyes, J. Silva-Agredo, J. Vazquez-Arenas, I. Romero-Ibarra, and 24. M. Hosseini, M. Momeni, and M. Faraji, Electroanalysis, 22, 2620 (2010).
R. A. Torres-Palma, Journal of Environmental Chemical Engineering, 6, 3010 (2018). 25. H. Cao, Z. Fan, G. Hou, Y. Tang, and G. Zheng, Electrochimica Acta, 125, 275
4. A. Perea, R. E. Palma-Goyes, J. Vazquez-Arenas, I. Romero-Ibarra, C. Ostos, and (2014).
R. A. Torres-Palma, Science of The Total Environment, 648, 377 (2019). 26. J. Wu, H. Xu, and W. Yan, Rsc Advances, 5, 19284 (2015).
5. F. L. Guzmán-Duque, R. E. Palma-Goyes, I. González, G. Peñuela, and 27. J. Kim, C. Kim, S. Kim, and J. Yoon, Journal of Industrial and Engineering Chemistry,
R. A. Torres-Palma, Journal of Hazardous Materials, 278, 221 (2014). 66, 478 (2018).
6. F. Bonfatti, S. Ferro, F. Lavezzo, M. Malacarne, G. Lodi, and A. De Battisti, Journal 28. R. Jaimes, J. Vazquez-Arenas, I. González, and M. Galván, Electrochimica Acta, 229,
of The Electrochemical Society, 147, 592 (2000). 345 (2017).
7. C. Comninellis, Electrochimica Acta, 39, 1857 (1994). 29. Q. Ni, D. Kirk, and S. Thorpe, Journal of The Electrochemical Society, 162, H40
8. H. Beer, British Pat., 1, 10 (1969). (2015).
9. S. Trasatti, Electrochimica Acta, 45, 2377 (2000). 30. http://msdssearch.dow.com/PublishedLiteratureDAS/dh_09a3/0901b803809a3c67.
10. M. Deborde and U. Von Gunten, Water Research, 42, 13 (2008). pdf?filepath=/pdfs/noreg/010-21993&fromPage=GetDoc.
11. R. Palma-Goyes, J. Vazquez-Arenas, C. Ostos, A. Manzo-Robledo, I. Romero-Ibarra, 31. P. Acevedo-Peña, J. E. Carrera-Crespo, F. González, and I. González, Electrochimica
J. Calderón, and I. González, Electrochimica Acta, 275, 265 (2018). Acta, 140, 564 (2014).
12. D. García-Osorio, R. Jaimes, J. Vazquez-Arenas, R. Lara, and J. Alvarez-Ramirez, 32. D. A. Hanaor and C. C. Sorrell, Journal of Materials Science, 46, 855 (2011).
Journal of The Electrochemical Society, 164, E3321 (2017). 33. J. Carrera-Crespo, M. E. Rincón, F. González, E. Barrera, and I. González, Journal
13. D. García-Osorio, J. Vazquez-Arenas, and R. Jaimes, Journal of The Electrochemical of Solid State Electrochemistry, 20, 2713 (2016).
Society, 165, J3101 (2018). 34. Z. Wang, B. Liu, Z. Xie, Y. Li, and Z.-Y. Shen, Ceramics International, 42, 13664
14. X. Chen and S. S. Mao, Chemical Reviews, 107, 2891 (2007). (2016).
15. P. Acevedo-Peña, J. Vazquez-Arenas, R. Cabrera-Sierra, L. Lartundo-Rojas, and 35. M. T. Uddin, Y. Nicolas, C. Olivier, T. Toupance, M. M. Müller, H.-J. Kleebe,
I. González, Journal of The Electrochemical Society, 160, C277 (2013). K. Rachut, J. Ziegler, A. Klein, and W. Jaegermann, The Journal of Physical Chem-
16. T. Kasuga, Thin Solid Films, 496, 141 (2006). istry C, 117, 22098 (2013).
17. M. Hosseini and M. M. Momeni, Journal of Solid State Electrochemistry, 14, 1109 36. M. Sun, X. Ma, X. Chen, Y. Sun, X. Cui, and Y. Lin, Rsc Advances, 4, 1120 (2014).
(2010). 37. M. A. Mahadik, P. S. Shinde, M. Cho, and J. S. Jang, Journal of Materials Chemistry
18. L. Xing, J. Jia, Y. Wang, B. Zhang, and S. Dong, International Journal of Hydrogen A, 3, 23597 (2015).
Energy, 35, 12169 (2010). 38. J. Carrera-Crespo, J. Ghilane, H. Randriamahazaka, S. Ammar, and I. González,
19. B. Abida, L. Chirchi, S. Baranton, T. W. Napporn, H. Kochkar, J.-M. Léger, and Journal of The Electrochemical Society, 164, H286 (2017).
A. Ghorbel, Applied Catalysis B: Environmental, 106, 609 (2011). 39. G. F. Pereira, R. C. Rocha-Filho, N. Bocchi, and S. R. Biaggio, Electrochimica Acta,
20. M. Wang, D.-J. Guo, and H.-L. Li, Journal of Solid State Chemistry, 178, 1996 (2005). 179, 588 (2015).
21. W. Xie, S. Yuan, X. Mao, W. Hu, P. Liao, M. Tong, and A. N. Alshawabkeh, Water 40. A. Özcan, Y. Şahin, A. S. Koparal, and M. A. Oturan, Journal of Hazardous Materials,
Research, 47, 3573 (2013). 153, 718 (2008).
22. S. Chai, G. Zhao, P. Li, Y. Lei, Y.-N. Zhang, and D. Li, The Journal of Physical 41. D. A. C. Coledam, I. Sánchez-Montes, B. F. Silva, and J. M. Aquino, Applied Catal-
Chemistry C, 115, 18261 (2011). ysis B: Environmental, 227, 170 (2018).

Downloaded on 2019-10-25 to IP 187.189.114.208 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

You might also like