Effect of bimodal mesoporous carbon as PtRu catalyst support for direct methanol fuel cells

Article abstract of DOI:10.1039/d0ra05676f

Mesoporous carbons (MCs) with different pore sizes were synthesized and evaluated as a catalyst support for fuel cells. The MCs were obtained from resorcinol-formaldehyde precursors, polymerized in the presence of polydiallyldimethylammonium chloride (cationic polyelectrolyte) as a structuring agent and commercial silica (Sipernat or Aerosil) as the hard template. The MC obtained with Aerosil shows a broad pore size distribution with a maximum at 21 nm. On the other hand, the MCs with Sipernat show a bimodal pore size distribution, with a narrow peak centered at 5 nm and a broad peak with a maximum ca. 30 nm. All MCs present a high specific surface area (800-1000 m2 g-1) and total pore volume ranging from 1.36 to 1.69 cm3 g-1. PtRu nanoparticles were deposited onto the MC support by an impregnation-reduction method with NaBH4 at 80 °C in basic media. The electrochemical characterization reveals improved electrocatalysis towards the methanol oxidation for the catalyst deposited over the carbon with the highest total pore volume. This catalyst also presented the highest CO2 conversion efficiency, ca. 80%, for the methanol oxidation as determined by differential electrochemical mass spectroscopy analysis. Moreover, the catalyst as a fuel cell anode showed the best performance, reaching a power density of 125 mW cm-2 at 90 °C with methanol as fuel and dry O2.

Full text of DOI:10.1039/d0ra05676f

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Effect of bimodal mesoporous carbon as PtRu
catalyst support for direct methanol fuel cells†
Cite this: RSC Adv., 2020, 10, 30631
abc
Eduardo Fuentes-Quezada, Mariano M. Bruno,*d
bc
Gonzalo Montiel,
bc
bc
Horacio R. Corti and Federico A. Viva
*
Mesoporous carbons (MCs) with different pore sizes were synthesized and evaluated as a catalyst support
for fuel cells. The MCs were obtained from resorcinol–formaldehyde precursors, polymerized in the
presence of polydiallyldimethylammonium chloride (cationic polyelectrolyte) as a structuring agent and
commercial silica (Sipernat® or Aerosil®) as the hard template. The MC obtained with Aerosil® shows
a broad pore size distribution with a maximum at 21 nm. On the other hand, the MCs with Sipernat®
show a bimodal pore size distribution, with a narrow peak centered at 5 nm and a broad peak with
2
ꢀ1
a maximum ca. 30 nm. All MCs present a high specific surface area (800–1000 m
g ) and total pore
3
ꢀ1
volume ranging from 1.36 to 1.69 cm
g
. PtRu nanoparticles were deposited onto the MC support by
ꢁ
an impregnation–reduction method with NaBH
4
at 80 C in basic media. The electrochemical
characterization reveals improved electrocatalysis towards the methanol oxidation for the catalyst
deposited over the carbon with the highest total pore volume. This catalyst also presented the highest
Received 29th June 2020
Accepted 9th August 2020
CO
2
conversion efficiency, ca. 80%, for the methanol oxidation as determined by differential
DOI: 10.1039/d0ra05676f
electrochemical mass spectroscopy analysis. Moreover, the catalyst as a fuel cell anode showed the best
ꢀ2
ꢁ
rsc.li/rsc-advances
performance, reaching a power density of 125 mW cm at 90 C with methanol as fuel and dry O .
2
5,6
the proton exchange membrane. The most commonly used
catalysts in DMFCs are PtRu on the anode side and Pt on the
1
Introduction
7
–9
Improving the performance of components that constitute cathode side, where the high cost of the noble metals is
a fuel cell (FC) is of great interest as it holds the promise of considered to be one of the main obstacles to
1
,2
10
a convenient energy source for various types of applications.
FCs are still ideal candidates to provide electrical energy with
commercialization.
The metal catalysts in the CL are employed as nanoparticles
low to no environmental impact. Among the FCs operating at dispersed over a conductive support. The most common
3
11–14
low temperature, the direct methanol fuel cell (DMFC) has support for fuel cell catalysts are carbon-based materials.
attracted a great deal of attention, partly because of methanol's The structure of the catalyst support can determine the catalyst
favorable properties such as high volumetric energy density and nanoparticles stability and activity toward methanol oxida-
1
,2
15,16
ease of handling as a liquid under ambient conditions.
Moreover, this type of fuel cell are ideal candidate for portable well as the electroactive catalyst area, while improves the mass
tion.
The support can maximize the particle dispersion as
4
3,6
applications. The core of a polymer electrolyte membrane fuel transfer of reactants and products. Due to the mentioned cost
cell (PEMFC), is the membrane electrode assembly (MEA) that of noble metals, the DMFC anode and cathode catalyst loadings
ꢀ2
includes the catalytic layers (CL), gas diffusion layers (GDL) and must drop below 1.0 mg cm from the present 2.0–4.0 mg
ꢀ
2
12
cm
while maintaining the cell performance. Recently,
a
different routes for the preparation of advanced nanostructured
carbon materials have emerged, providing extra ne-tuning of
Departamento de F ´ı sica de la Materia Condensada, Comisi ´o n Nacional de Energ ´ı a the supported catalyst electroactivity.
Instituto Nacional de Tecnolog ´ı a Industrial, Av. General Paz 5445 San Mart ´ı n, Buenos
Aires, Argentina
b
12,13,17–19
At ´o mica, Av. General Paz 1499, San Mart ´ı n, Buenos Aires, Argentina. E-mail: viva@
tandar.cnea.gov.ar
An ideal carbon support should allow the preparation of
highly dispersed catalytic nanoparticles, whereas the porosity
should ensure the ionomer penetration for proton transport
while allowing a facile access path for reactants and by-prod-
c
Instituto de Nanociencia
y
Nanotecnolog ´ı a, CNEA-CONICETCentro At ´o mico
Constituyentes, Av Gral Paz 1499, San Martin, Buenos Aires, Argentina
d
Instituto de Investigaciones en Tecnolog ´ı as Energ ´e ticas y Materiales Avanzados
17,20
ucts.
It was shown that pores over 20 nm can ensure the
(
IITEMA), Universidad Nacional de R ´ı o Cuarto, Facultad de Cs. Exactas F ´ı sico
Qu ´ı mica y Naturales, Departamento de Qu ´ı mica, 5800 R ´ı o Cuarto, Argentina. formation of the triple phase boundary (TPB) region by allowing
E-mail: mbruno@exa.unrc.edu.ar
an optimal contact between the catalyst and the ionomer
†
Electronic supplementary information (ESI) available. See DOI:
0.1039/d0ra05676f
21,22
(Naon).
Additionally, the carbon surface defects, surcial
1
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 30631–30639 | 30631

View Article Online
RSC Advances
Paper
groups and the microporosity have been identied as anchoring (trihydrate PA, Ciccarelli) were dissolved in 50 mL of milli-Q
sites for metal nanoparticle, improving particle dispersion and water. In the solution B, the HT was completely dispersed in
23,24
catalytic activity.
The porosity can also inuence the resi- 4 g of methanol (Biopack 99.8% wt), 5 g of glycerol (Biopack
dence time of the reactant and the by-products near of the 99.5% wt), and 50 mL of milli-Q water. Both solutions were
12,25–27
catalyst nanoparticles.
In this sense, a carbon material mixed and heated in a reux system with magnetic stirring. The
with hierarchical pore size distribution could offer a plausible polymerization began by adding 1.4 g of a formaldehyde solu-
way to guarantee the mentioned features. In previous works, we tion (F) (37% wt, Sigma-Aldrich) to the dispersion. Aer 45 min,
have demonstrated that carbon with micropores/mesopores another 3 g of F were added. The heating and magnetic stirring
ꢁ
obtained by using polyelectrolyte as a structuring agent can was maintained for 20 minutes and cooled it to 25 C. Glycerol
17,22,23,28
improve the catalyst performance in a fuel cell.
More and methanol were added to the aqueous polymerization media
recently, we have presented a new method for the synthesis of as dispersant and wetting agents, respectively, rendering the
carbon with dual mesopores size which can satisfy the polymer in a powder form. By this synthesis procedure, three
29
requirements of an appropriate porous carbon support. This different MC were obtained; MC A05 was synthesized by using
study analyzes the effect of the pore size distribution and the 0.50 g of Aerosil® 200, while MC S15 and MC S30 were prepared
pore volume fraction of the carbon support on the fuel cell by using 1.50 g and 3.00 g of Sipernat® 50, respectively. The
performance. The strategy was to produce carbons with composite resin was vacuum ltered from the solution, dried in
ꢁ
a different pore size distribution and mesopore volume, a vacuum oven at 100 C overnight and then carbonized under
ꢀ
1
preserving the rest of the textural properties. The set of prop- a N stream of 1 L min in a tubular furnace (Indef model T-
2
ꢁ
ꢁ
ꢁ
ꢀ1
erties were tuned by adding pore forming agents in the resor- 150) from 20 C to 1000 C at a heating rate of 3 C min
29–31
ꢁ
cinol formaldehyde polymerization media.
and nally held at 1000 C for 120 minutes. The HT from the
In the present work, we describe the preparation and char- carbon was removed by etching with 3 M NaOH solution at
ꢁ
acterization of PtRu catalyst, synthesized by the impregnation– 60 C for 24 h under constant stirring. MCs were washed with
reduction method, supported on three different MCs. The milli-Q water in Soxhlet apparatus until a neutral pH was ob-
ꢁ
carbon supports were obtained by carbonization of a resor- tained and nally dried in a vacuum oven at 110 C overnight.
cinol–formaldehyde polymer combining commercial silica as
hard template and polydiallyldimethylammonium chloride as
2.2 Supported catalyst preparation
structuring agent. The MCs were characterized by N adsorp-
2
The preparation of PtRu nanoparticles deposited over the MCs
tion–desorption isotherms. The supported catalysts were char-
acterized by powder X-ray diffraction (PXRD), transmission
electron microscopy (TEM), and energy-dispersive X-ray spec-
troscopy (EDS). Stripping of CO was used for the determination
of the catalyst electrochemical surface area (ECSA), whereas the
electrocatalytic activity was determined by cyclic voltammetry
17,33
was carried out following a procedure previously described.
The metal precursors solutions, H PtCl $6H O (tetrahedron)
and RuCl $XH O (Aldrich), were added to a slurry of the MCs
2
6
2
3
2
while stirring for 30 min. The pH was adjusted to 8.0 with 1 M
ꢁ
NaOH (PA, Merck) aqueous solution and heated to 80 C. Once
the temperature was reached, NaBH4 (granular 98%, Sigma-
Aldrich) was added in a molar ratio of 5 : 1 (NaBH to metal
4
(CV), chronoamperometry, and potentiodynamic differential
electrochemical mass spectrometry (DEMS) measurements.
Finally, the performance of the MEA with the prepared mate-
rials as anode catalyst were evaluated with methanol as fuel.
salt) to the suspension. Heating was maintained for 2 h, fol-
lowed by stirring for 24 h at room temperature. The powder
obtained was washed with milli-Q water in a Soxhlet apparatus
ꢁ
and nally dried in a vacuum oven at 80 C for 24 h.
2
Experimental
2
.1 Mesoporous carbon synthesis
2.3 Mesoporous carbon and supported catalyst surface
characterization
Mesoporous carbons were obtained by carbonization of
a
resorcinol–formaldehyde (RF) resin. The polymer was An ASAP 2020 (Micrometrics) instrument was used to measure
ꢁ
prepared in the presence of sodium acetate as catalyst, the nitrogen adsorption/desorption isotherms at ꢀ196 C. The
a cationic polyelectrolyte as structuring agent (SA) and porous Brunauer–Emmett–Teller (BET) equation was used for deter-
silica as hard template (HT). The SA employed was poly- mining the specic surface area (SBET) while the volume of
diallyldimethylammonium chloride (pDADMAC, 20% wt in micropores (pores size <2 nm) was determined by applying the
ꢀ
1
H
2
O, average M
w
¼ 100 000–200 000 g mol , Sigma-Aldrich), Dubinin–Radushkevich (DR) equation. The mesopore volume
while the commercial silica powders Aerosil® 200 and and pore size distribution (PSD) were obtained with the Barrett–
Sipernat® 50 (EVONIK) were employed as the HT. The Aerosil® Joyner–Halenda (BJH) method. The fractions of mesopore
2
00 is a nonporous fumed silica with a specic surface area volumes were determined from the adsorption branch of the
2
ꢀ1
between 50 and 500 m g , and a particles with sizes between 5 isotherm, whereas mesopore size distribution was calculated
and 50 nm. The Sipernat® 50 are porous silica particles of 70 from the desorption branch. The total volume was calculated at
mm
4
32
0
34,35
in
75 m g
diameter
and
a
surface
area
of a relative pressure P/P ¼ 0.99.
2
ꢀ1 29
.
Briey, two solutions were prepared, the solution
PXRD pattern of the catalysts were obtained using a Siemens
A containing 2 g of resorcinol (R) (99.0% ACS, Sigma-Aldrich), D5000 diffractometer with a Cu Ka source operating at 40 kV
ꢁ
1
g of the SA solution, and 0.25 g of sodium acetate and 30 mA. The angle extended from 20 to 100 with a step size
30632 | RSC Adv., 2020, 10, 30631–30639
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
ꢁ
of 0.02 and a counting time of 2 s. TEM images were acquired vacuum chamber was pumped with a rotary vane pump (DUO 5,
with JEOL 100 CX II, meanwhile EDS was performed using an Pfeiffer) equipped with liquid N trap. The secondary chamber
2
ꢀ1
SEM Philips 505 with EDAX detector to quantify the atomic ratio was pumped by a 60 L s turbomolecular pump backed by a dry
of Pt and Ru in the catalysts. Thermogravimetric analyses (TGA) diaphragm pump (turbo drag pumping station TSH/U 071 E,
were performed with a Q600 SDT Thermal Analyzer from TA Pfeiffer). A motorized gas-dosing valve (EVR 116, Pfeiffer)
Instruments controlled by Q Series soware. The experiments regulates the gas ow from the cell to the detector. The quad-
were carried out using approximately 10 mg of sample in rupole mass spectrometer, equipped with a continuous dynode
alumina pans under air atmosphere (Praxair A10.0XD-T secondary electron multiplier/Faraday cup detector having
ꢁ
ꢀ1
9
9.99%), with a heating rate of 10 C min and a gas ow a sensitivity of 200 A per mbar (QMS 200 M1, Prisma), was
ꢀ
1
rate of 50 mL min . The metal content on the supported connected to the analysis chamber (secondary chamber).
catalysts were calculated from the difference between the initial
and nal weights.
The experiments were performed in a ow electrochemical
17,41
cell designed for the DEMS.
The WE consisted of a glassy
carbon disk built ad hoc (6 mm in diameter) with a 1 mm
diameter hole in the center through which the reactant owed
during the measurement. A suspension of the supported cata-
2.4 Electrochemical measurements
A suspension of the supported catalyst prepared in milli-Q lyst as described in Section 2.4 was spread over the working
water, isopropyl alcohol (Biopack) and Naon ionomer (5% wt electrode (WE). The inlet port, from the cell to the DEMS pres-
of Naon dispersion in isopropyl alcohol, Aldrich) in a 0.15 : 1 sure chamber, is through a stainless steel frit. The electro-
ratio of Naon to catalyst was deposited with a micropipette(1– chemical cell opening is separated from the frit by a porous
1
0 mL, Rontaig) over the working electrode (WE), consisting of Teon membrane (0.02 mm pore, 50 mm thickness, 50% porosity
a glassy carbon disk (5 mm diam.) mounted on a Teon rod. by Gore). The WE lies on top of the Teon® membrane, sepa-
The WE was previously cleaned with isopropyl alcohol in an rated from it by a 100 mm Teon® gasket, allowing the forma-
ultrasonic bath for 10 minutes and aer deposition of the tion of a thin liquid reacting layer. Volatile product species
ꢁ
catalyst suspension, was dried in a vacuum oven at 80 C for 10 diffuse through the membrane to reach the mass spectrometer.
ꢀ
1
minutes. CV, and chronoamperometry experiments were per- The electrolyte ow rate (0.12 mL min ) was controlled by
formed in a three-electrode cell. The counter electrode (CE) a syringe pump (PC11U, Apema). Calibration by CO stripping
consisted of a coiled Pt wire 0.5 mm in diameter and 30 cm and further quantication of the conversion efficiency was
17,42
length, whereas a Ag/AgCl (sat. KCl) electrode was used as carried out as previously reported.
a reference electrode (RE). All potentials were converted against
the standard hydrogen electrode (SHE). The upper potential
2
.5 MEA preparation and fuel cell testing
limit for the voltammetric determinations was set to 0.8 V vs.
SHE to avoid the formation of irreversible ruthenium oxides or The synthesized PtRu/MCs were used as anode catalyst for the
36
Ru dissolution. The electrochemical surface area (ECSA) was preparation of MEAs, while Pt supported over Vulcan carbon
measured by CO stripping voltammetry (ESI-Fig. 1†). The cell (Pt/VC) 60% (Fuel Cell Store) was used as cathode catalyst. A
was lled with 0.5 M H SO (95–97%, Merck) solution and MEA with commercial PtRu catalyst over Vulcan carbon (PtRu/
2 4
saturated with CO (RG, Indura) for 45 min while the WE VC) 60% (Fuel Cell Store) as anode catalyst was also assem-
potential was maintained at 0.2 V vs. SHE. Aer the time bled for comparison. The catalyst suspension was prepared by
elapsed, and while maintaining the potential, the solution was mixing the catalyst with milli-Q water and Naon ionomer
purged with N (RG, Indura) for 15 min to remove the unad- solution (5% wt low MW alcohols, Ion Power) in a 1 : 10 : 7 mass
2
2
sorbed CO, and immediately two scans between 0.05 and 0.8 V proportion, respectively, and spread on one side of a 5 cm
vs. SHE at a scan rate of 1 mV s were performed. The catalysts Toray C paper TGP-H 60 10% PTFE coated (Fuel Cell Technol-
ECSA was calculated based on the mass deposited onto the WE ogies), for a nal electrode loading ca. 3 mg cm . A Naon 212
ꢀ
1
ꢀ2
and the CV peak integral using the reference charge value of 420 membrane (Ion Power) was placed in between the electrodes
ꢀ
2
37
ꢁ
mC cm for the oxidation of a CO monolayer, and employed to and hot pressed at 150 C and 40 bar for 25 min. The Naon
convert the measured current (i) to current density (j). The membrane was previously treated by boiling in H 3% wt
catalysts CVs were performed on a 1 M methanol solution in (H O 30% wt, Bio-pack) followed by H SO 3% wt (95–97%wt,
2 2
O
2
2
2
4
0
.5 M H SO by sweeping the potential between 0.05 and 0.8 V Merck). The MEAs were mounted in a standard single cell
2 4
ꢀ1
vs. SHE at 2, 5, 10 and 20 mV s (ESI-Fig. 2†). The chro- housing with serpentine ow elds (Fuel Cell Technologies,
noamperometry were performed in the same electrolyte solu- Inc.). Teon gasket lms (50–150 mm) were employed as seal
tion at 0.5 V vs. SHE for 1 h. These measures were employed to and the cell uniformly bolted with a torque of 2.3 Nm. Aer
obtain the poisoning rate and the turn over frequency (TOF) as assembly of the cell, the MEA was re-humidied by circulating
38–40
ꢁ
was calculated in previous reports.
All electrochemical water at 80 C overnight. Galvanodynamic polarization test was
measurements were performed with an Autolab PGSTAT302N performed with a test station (University Test Station model
ꢁ
potentiostat.
from Fuel Cell Technologies) at 90 C, from the open circuit
2.4.1 DEMS setup and cell conguration. The DEMS setup voltage (OCV) to a voltage close of short circuit (0.05 V) while
consisted of two differentially pumped chambers and a quad- circulating 1 M methanol (Merk, HPLC grade) through the
rupole mass spectrometer (100 amu, Pfeiffer). The primary anode and dry O
2
(RG 4.8, Indura) through the cathode. A
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 30631–30639 | 30633

View Article Online
RSC Advances
Paper
Gilson Minipuls 3 peristaltic pump was used to circulate the S30 and MC S15. Contrarily, the volume of the small mesopores
methanol solution and a digital mass ow meter (MC 200 from (V ) is higher for MC S30 followed by MC S15 and MC A05.
s
Alicat Scientic) to control the O ows. For all the measure-
ments the methanol ow was set to 2.0 mL min while the O2 synthesis method yield samples that present similar textural
The aforementioned results suggest that the specic
2
ꢀ1

ow was 200 SCCM.
properties despite the different mesopore distribution, mono-
modal for the MC A05 and bimodal for the MC S15. Addition-
ally, bimodal samples can be prepared with different ratios of
small and large mesopores.
3
Results and discussion
3.1 Mesoporous carbons characterization
3.2 Catalyst surface characterization
The textural properties of the MCs were analyzed by the
nitrogen adsorption–desorption isotherms and the PSD as The size and distribution of metal nanoparticles deposited on the
shown in Fig. 1(A and B), respectively. The isotherms of the different carbon supports were analyzed by TEM images, as shown
three carbons show a similar prole at low relative pressure. in Fig. 2(A, B and C). Low-magnication TEM images show a good
o
The sharp rise at low P/P is indicative of the micropore pres- dispersion of catalyst particles for the three MCs supports. The
o
ence. At higher P/P , MC A05 shows a hysteresis loop in the catalyst particle size diameter distribution was obtained by
adsorption–desorption isotherm, while two loops can be measuring the diameter of 100 randomly selected particles with
appreciated for MC S15 and MC S30. The single hysteresis loop the soware ImageJ. The Fig. 2 insets show the bar chart with the
for MC A05 and the double hysteresis loop for MC S15 and MC overlapping Gaussian distribution. The mean particle size ob-
S30 indicates the monomodal and bimodal population of the tained from the images were 4.5 nm for PtRu/MC A05, 4.2 nm for
mesopore size distribution, respectively. The MC S30 shows the PtRu/MC S15 and 3.9 nm for PtRu/MC S30. As mentioned above,
highest adsorption volume. Fig. 1B shows the PSD for the three supports with small mesopores and/or micropores can serve as
MCs. The plot for MC A05 presents a broad monomodal anchoring sites, decreasing the particle size and improving their
distribution with a maximum centered on 21 nm, whereas MC dispersion. The PtRu nanoparticle size decreases as the small
S15 and MC S30 show a narrow peak centered at 5 nm and mesopore volume increases (Table 1). While the difference in the
s
a broad peak with a maximum ca. 30 nm. The textural proper- mean particle size is not as signicant as the difference in V , the
ties obtained from the isotherms are summarized in Table 1. particle size distribution observed in the bar charts (Fig. 2 Insets)
Based in the plots observed in Fig. 1B, two different mesopores is more representative of the tendency observed for Vs.
sizes were arbitrarily dened: the small mesopores with pore
diameter (d) between 2 and 7 nm and the large mesopores with shown in Fig. 3. The diffraction peaks at 2q angles ca. 40, 46, 67,
d between 7 and 50 nm. All the MCs show high values of SBET and 81 are due to face-centered cubic (fcc) crystalline Pt,
around 800 m g for MC A05 and MC S15, and slightly higher assigned to the planes (111), (200), (220), and (311), respectively,
The PXRD diffractograms of the synthesized catalyst are
,
2
ꢀ1
2
ꢀ1
for the MC S30 (1000 m g ) due to the larger fraction of small with a slight shiing to higher 2q values due to the presence of
43–45
mesopores and micropores. The micropore volume (V
m
) and the Ru.
The lattice parameters of the metal nanoparticles were
total pore volume (Vtotal) are in the same order of magnitude for calculated by indexing the rst three peaks yielding 3.900 ꢂ
˚
˚
the three MCs. However, the main structural difference between 0.002 A for PtRu/MC A05, 3.895 ꢂ 0.002 A for PtRu/MC S15, and
˚
the supports resides in the mesopore volumes. The volume of 3.896 ꢂ 0.002 A for PtRu/MC S30. Using the reference value of
˚
large size mesopores (V ), is higher for MC A05 followed by MC 3.923 A, the lattice parameter was used to estimate the Ru
l
ꢁ
Fig. 1 (A) N
2
adsorption–desorption isotherms of the MCs measured at ꢀ196 C. Closed symbols: adsorption branch, open symbols: desorption
branch. (B) Pore size distribution obtained by applying the BJH method.
30634 | RSC Adv., 2020, 10, 30631–30639
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Table 1 BET Surface area, volume of micro and mesopores, and total pore volume of the MCs
a
2
ꢀ1
b
m
3
ꢀ1
c
s
3
ꢀ1
d
3
ꢀ1
e
3
ꢀ1
Sample
HT
HT/R
S
BET (m g
)
V
(cm g
)
V
(cm g
)
V
l
(cm g
)
V
total (cm g )
MC A05
MC S15
MC S30
Aerosil
Sipernat
Sipernat
0.25
0.37
0.75
796
787
1000
0.33
0.32
0.40
0.11
0.34
0.59
0.77
0.54
0.66
1.50
1.36
1.69
a
b
c
d
Specic surface area using the BET method. Micropore volume from DR equation. Volume of small mesopores (2 < d < 7 nm). Volume of large
e
ꢁ
mesopores (7 < d < 50 nm). Total pore volume at P/P ¼ 0.99.
46
atomic fraction alloyed with Pt. The values obtained were 19%
for PtRu/MC A05, 23% for PtRu/MC S15, and 22% for PtRu/MC
S30, with an error of ꢂ2%. The alloying Ru percentages, which
are of the same order of magnitude for the three catalyst, shows
that most of the Ru is present in an amorphous phase. None-
theless, the percentage values are high for a catalyst that has not
been subjected to a thermal treatment aer deposition of the
17,47
metal nanoparticles.
The EDS results obtained for the catalyst indicated a Pt : Ru
atomic ratio of 52 : 48 for PtRu/MC A05, 49 : 51 for PtRu MC
S15, and 51 : 49 for PtRu/MC S30 close to the 1 : 1 atomic
nominal ratio. The thermogravimetric analysis of the metal
catalyst indicated metal loadings of 49% for PtRu/MC A05, 61%
for PtRu MC S15, and 62% for PtRu/MC S30. The metal catalyst
percentage over MC S15 and S30 are close to the intended value
Fig. 3 XRD diffractograms of the PtRu catalysts supported on MC A05,
MC S15 and MC S30.
Fig. 2 TEM images of the different supported catalysts (A) PtRu/MC A05, (B) PtRu/MC S15, (C) PtRu/MC S30. Inset: corresponding particle size
distributions.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 30631–30639 | 30635

View Article Online
RSC Advances
Paper
of 60% while is slightly lower over MC A05. As discussed above,
the lower fraction of small size mesopores provides the least
favorable support for particle anchoring.
3.3 Electrochemical catalyst characterizations
The ECSA values obtained from the CO striping voltammo-
grams for PtRu/MC A05, PtRu/MC S15 and PtRu/MC S30 were 37
2
ꢀ1
2
ꢀ1
2
ꢀ1
m g , 41 m g , and 42 m g , respectively. As in the case for
the mean particle size, the difference between the three cata-
lysts is small, nonetheless, the ECSA follows the expected trend
were the lowest particle size correspond to the highest ECSA
(PtRu/MC A05 < PtRu/MC S15 < PtRu/MC S30).
The catalytic activity of the prepared catalysts was assessed
Fig. 5 Chronoamperometry determination of the three different
catalysts in 1 M methanol + 0.5 M H SO at 0.5 V vs. SHE.
2 4
2 4
in a 1 M methanol + 0.5 M H SO aqueous solution. Fig. 4 shows
ꢀ
1
the voltammograms for the three catalysts at 2 mV s . The peak
current density for the methanol oxidation increases in the
order PtRu/MC S30 > PtRu/MC S15 > PtRu/MC A05, while the
onset potential from the CVs are 0.49 V for the MC A05, 0.39 V
for the MC S15, and 0.37 for the MC S30, following the same
order. This suggest a more facile methanol oxidation on PtRu/
MC S30.
and the nanoparticle size and therefore the relation with the
support porosity. The poisoning rate shows similar results for
the three catalysts. As the d is related to the adsorption of
intermediate species, the results would indicate that either the
intermediates leave rapidly the catalyst surface, or the methanol
oxidation produces a low amount of intermediates. The DEMS
analysis, vide infra, shows that the catalyst conversion efficiency
is high. Previous reported values of d and TOF parameters
suggest that mesoporous support allows the intermediates to
The chronoamperometry measurements of the catalysts at
.5 V vs. SHE for a 30 min period are shown in Fig. 5. The rapid
0
current decay shows a more restricted diffusion for PtRu/MC
A05 than for PtRu/MC S30. The slope of the current transient
in the chronoamperograms were employed for the determina-
17,48–51
39,40
rapidly leave the catalyst surface.
tion of the catalyst poisoning rate (d),
while the steady-state
current density was used for the calculation of the turnover
1
7,38,40
frequency (TOF).
00 and 1800 s has been attributed to the catalyst poisoning by
CO, whereas at longer times the decay is related to anion
The chronoamperometry slope between 3.3.1 Catalyst characterization by DEMS
5
DEMS experiments were carried out to measure the conversion
efficiency from methanol to CO for the synthesized catalysts.
2
17
adsorption. The values obtained for the poisoning rate and
TOF, respectively, are presented in Table 2. The PtRu/MC S30
presents the highest number of reacting molecules per site,
since TOF is directly proportional to the steady state current
density. The TOF for the supported catalyst increases as the
nanoparticle size decreases, i.e. as the support V increases. The
results show the relationship between the turn over frequency
The electrochemical technique has been used to identify and
quantify electrocatalysts reactions products for the methanol
oxidation. The products directly identied by DEMS are CO
2
(
(
m/z ¼ 44) (complete oxidation product) and methyl formate
40
m/z ¼ 60), formed between the methanol present in the
s
electrolyte and the formic acid generated during the electro-
38,52,53
chemical oxidation.
formaldehyde (m/z ¼ 30), CO (m/z ¼ 28) and formic acid (m/z ¼
6) exhibit signals that overlap with the fragments of methanol
m/z ¼ 32) and CO present in the mass spectra, which
complicates their direct determination.
Other oxidation products such as
4
(
2
3
8,52
Fig. 6 shows the
results obtained, which include the electrochemical CVs and
the mass signal m/z ¼ 44 as a function of the applied potential.
The electrochemical response acquired on the DEMS cell
presented the expected shape for the CVs (Fig. 4) considering
Table 2 Poisoning rate (d) and TOF of the three catalysts obtained
from the chronoamperometry experiments
ꢀ1
Catalyst sample
d (% s
)
TOF (molecules per site$per s)
PtRu/MC A05
PtRu/MC S15
0.0125
0.0083
0.0118
0.018
0.023
0.034
Fig. 4 Cyclic voltammograms for the three different catalysts in 1 M PtRu/MC S30
ꢀ1
2 4
methanol + 0.5 M H SO at 2 mV s .
30636 | RSC Adv., 2020, 10, 30631–30639
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
2 4
Fig. 6 Potentiodynamic DEMS measurement of PtRu/MC A05, PtRu/MC S15 and PtRu/MC S30 in 1 M methanol + 0.5 M H SO with the signal
corresponding to m/z ¼ 44 (CO
2
).
that the measures were carried out in a cell with a owing thin catalyst for comparison. The cells with PtRu/MC A05 and PtRu/
1
7,42,54
layer of electrolyte.
Moreover, the CVs shape in the DEMS MC S15 as well as the one with PtRu/VC show an OCV of ca.
cell allow to conrm that there is an adequate ow of elec- 0.65 V while the one with PtRu/MC S30 presents a slightly lower
trolyte, avoiding any starvation of reactants during the OCV. The polarization curves show better performance for the
measurements. From the mass signal (m/z ¼ 44), the average cells with PtRu/MC S15 and PtRu/MC S30 as anode catalysts, i.e.
1
7
current efficiency (h) for the formation of CO from methanol those with catalyst over bimodal carbon, than the cell with
2
was calculated. The obtained values for the catalyst were 72% PtRu/MC A05. The peak power density observed is 126 mW
ꢀ
2
ꢀ2
for PtRu/MC A05, 74% for PtRu/MC S15 and 84% for PtRu/MC cm for PtRu/MC S30, 115 mW cm for PtRu/MC S15 and 109
ꢀ
2
S30. The results show that the catalysts supported over the mW cm for PtRu/MC A05. For the commercial catalyst, the
ꢀ
2
bimodal carbons present a higher conversion of methanol to peak power density is 90 mW cm , which is an optimum value
CO than the monomodal carbon. Moreover, for the PtRu/MC for the MEA composition and the experimental settings
S30 catalyst, the signal for the m/z ¼ 44 presents an onset at employed (temperature, O ow, methanol concentration) as
a lower potential (ca. 0.3 V) than the observed for the other two compared with previously reported values.
catalysts, indicating an improved conversion to CO . As The polarization plots show a better performance of the
2
2
12,13,20,57,58
2
described in the previous sections, the three catalysts have catalyst deposited over the bimodal carbons compared to the
a similar electrochemical surface area, while the main differ-
ence resides in the PSD and their contribution to the carbon
total pore volume. The conversion efficiency results would
indicate that the catalyst support affects the methanol oxida-
tion, which can proceed via a parallel path mechanism: (1) via
8
oxidation of adsorbed CO and (2) via dissolved intermediates.
Under the ow conditions of the DEMS cell, any intermediate
should be able to escape easily from the catalytic layer.
However, it can be seen an increasing trend in catalyst
performance and CO efficiency by increasing the mesopore
2
content. The results are in agreement with the fact that
methanol oxidation on PtRu proceeds via the direct pathway
5
5,56
through adsorbed CO.
3
.4 Fuel cell performance characterizations
ꢁ
Fig. 7 Polarization and power curves at 90 C of the synthesized and
The fuel cell polarization and power curves are shown in Fig. 7.
The plot presents the results of the MEAs prepared with the
catalysts synthesized over the MCs and with a commercial
ꢀ1
commercial catalysts, measured with 1 M methanol (2.0 mL min flow
rate) at the anode and dry O (200 SCCM flow rate) at the cathode.
2
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 30631–30639 | 30637

View Article Online
RSC Advances
Paper
monomodal support. The PtRu/MC S30, which has the highest
amount of small mesopores (Fig. 1B and Table 1), shows the
highest power density followed by PtRu/MC S15. The PtRu/MC
A05 presents the lowest performance of the prepared catalysts
in this report, but still it performs better than a previously re-
References
1 D. Edlund, Methanol Fuel Cell Systems: Advancing Towards
Commercialization, Pan Stanford Publishing, Singapore, 1st
edn, 2011.
2 H. R. Corti and E. R. Gonzalez, Direct alcohol fuel cells:
Materials, performance, durability and applications, Springer
Netherlands, Dordrecht, 1st edn, 2014.
3 N. Abdullah, S. K. Kamarudin and L. K. Shyuan, Nanoscale
Res. Lett., 2018, 13, 90.
4 E. A. Franceschini and H. R. Corti, in Direct Alcohol Fuel Cells:
Materials, Performance, Durability and Applications, ed. H. R.
Corti and E. R. Gonzalez, Springer Netherlands, Dordrecht,
1st edn, 2014, pp. 321–355.
17,22
ported PtRu catalyst deposited over a monomodal MC.
The
reported MC, synthesized by a process without hard template,
exhibited a PSD with a narrow peak centered at 20 nm and
3
ꢀ1
a total pore volume of ca 1 cm g . The commercial catalyst,
with Vulcan® carbon as support, displays the lowest perfor-
mance. Vulcan® is a carbon black with a low surface area (252
2
ꢀ1
3
m g ) without mesoporosity or microporisity (V ¼ 0.63 cm
T
ꢀ1
14
g
). The measured fuel cell polarizations indicates that the
support's mesoporosity, particularly the small mesopores, have
a positive inuence on the methanol oxidation, improving the
overall cell performance.
5 S. S. Munjewar, S. B. Thombre and R. K. Mallick, Ionics, 2017,
23, 1–18.
6
D. S. Falc ˜a o, V. B. Oliveira, C. M. Rangel and A. M. F. R. Pinto,
Renewable Sustainable Energy Rev., 2014, 34, 58–70.
R. Parsons and T. VanderNoot, J. Electroanal. Chem.
Interfacial Electrochem., 1988, 257, 9–45.
7
4
Conclusions
In the present work, anode catalysts for DMFC formed by PtRu
nanoparticles deposited on porous carbon with different pore
8 N. Markovic and P. N. Ross, Surf. Sci. Rep., 2002, 45, 117–229.
9 M. Mansor, S. N. Timmiati, K. L. Lim, W. Y. Wong,
S. K. Kamarudin and N. H. Nazirah Kamarudin, Int. J.
Hydrogen Energy, 2019, 44, 14744–14769.
size distribution were synthesized.
A synthesis method
employing silica as hard template and PDMAC as structuring
agent was used to produce a hierarchical structure. This method 10 Z. A. C. Ramli and S. K. Kamarudin, Nanoscale Res. Lett.,
produced MCs with similar textural properties (SBET, V , V
2018, 13, 410.
and different mesopore distributions. In the presence of 11 E. Auer, A. Freund, J. Pietsch and T. Tacke, Appl. Catal., A,
PDMAC, Aerosil® produced a carbon with a single broad peak
1998, 173, 259–271.
pore distribution centered at 21 nm, while with Sipernat® the 12 H. Liu, C. Song, L. Zhang, J. Zhang, H. Wang and
m
T
)
PSD shows two peaks; a narrow one centered at 5 nm and
a broad one with a maximum ca. 30 nm.
D. P. Wilkinson, J. Power Sources, 2006, 155, 95–110.
13 L. Gong, Z. Yang, K. Li, W. Xing, C. Liu and J. Ge, J. Energy
Chem., 2018, 27, 1618–1628.
The electrochemical results and the fuel cell performance
measurements show the effect of the support textural properties 14 M. M. Bruno and F. A. Viva, in Direct Alcohol Fuel Cells, ed. H.
on the PtRu nanoparticles catalytic properties. The carbon
support synthesized presents an adequate surface area and
R. Corti and E. R. Gonzalez, Springer Netherlands,
Dordrecht, 1st edn, 2014, pp. 231–270.
microporosity that provides good particle size and distribution 15 S.-H. Liu, W.-Y. Yu, C.-H. Chen, A.-Y. Lo, B.-J. Hwang,
across the surface. The presence of the mesopores also allows
S.-H. Chien and S.-B. Liu, Chem. Mater., 2008, 20, 1622–1628.
the formation of the TPB and a sponge like structure which 16 C.-T. Hung, Z.-H. Liou, P. Veerakumar, P.-H. Wu, T.-C. Liu
facilitates the access of methanol onto the catalyst and the
departure of CO
and S.-B. Liu, Chin. J. Catal., 2016, 37, 43–53.
17 F. A. Viva, M. M. Bruno, M. Jobb ´a gy and H. R. Corti, J. Phys.
Chem. C, 2012, 116, 4097–4104.
2
.
18 H. Ishitobi, A. Ida, S. Anayama and N. Nakagawa, J. Chem.
Eng. Jpn., 2016, 49, 995–1001.
Conflicts of interest
1
9 A. M. Baena-Moncada, R. Coneo-Rodr ´ı guez, J. C. Calder ´o n,
J. Fl ´o rez-Monta n˜ o, C. A. Barbero, G. A. Planes,
J. L. Rodr ´ı guez and E. Pastor, Int. J. Hydrogen Energy, 2014,
There are no conicts of interest to declare.
39, 3964–3969.
2
2
0 J. B. Xu and T. S. Zhao, RSC Adv., 2013, 3, 16–24.
1 V. Rao, P. A. Simonov, E. R. Savinova, G. Plaksin,
S. Cherepanova, U. Stimming and G. Kryukova, J. Power
Sources, 2005, 145, 178–187.
Acknowledgements
The authors thank nancial support from Consejo Nacional de
Investigaciones Cient ´ı cas y Tecnol ´o gicas (CONICET, PIP
0
808CO). The author also thanks Dr Diego Lamas and Dr Ana L. 22 M. M. Bruno, M. A. Petruccelli, F. A. Viva and H. R. Corti, Int.
Larralde of Laboratorio de Cristalograf ´ı a Aplicada –UNSAM for
J. Hydrogen Energy, 2013, 38, 4116–4123.
the PXRD. EFQ thanks CONICET for his doctoral scholarship. 23 M. M. Bruno, F. A. Viva, M. A. Petruccelli and H. R. Corti, J.
GM is a graduate student of Escuela de Ciencia y Tecnolog ´ı a-
UNSAM and EFQ of Instituto Sabato-UNSAM. MMB, HRC and
FAV are staff members of CONICET.
Power Sources, 2015, 278, 458–463.
30638 | RSC Adv., 2020, 10, 30631–30639
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
2
2
2
4 W. Xu, Z. Wu and S. Tao, J. Mater. Chem. A, 2016, 4, 16272– 41 G. A. Planes, G. Garc ´ı a, E. Pastor, G. Garcia and E. Pastor,
6287. Electrochem. Commun., 2007, 9, 839–844.
5 P. V Samant, C. M. Rangel, M. H. Romero, J. B. Fernandes 42 E. A. Franceschini, M. M. Bruno, F. J. Williams, F. A. Viva and
and J. L. Figueiredo, 2005, 151, 79–84. H. R. Corti, ACS Appl. Mater. Interfaces, 2013, 5, 10437–10444.
6 M. Zhang, Z. Perry, J. Park and H.-C. Zhou, Polymer, 2014, 55, 43 D. Chu and S. Gilman, J. Electrochem. Soc., 1996, 143, 1685–
1
335–339.
1690.
2
2
7 X. Yu, F. Luo and Z. Yang, RSC Adv., 2016, 6, 98861–98866.
8 F. A. Viva, M. M. Bruno, E. A. Franceschini, Y. R. J. Thomas,
44 A. O. Neto, R. R. Dias, M. M. Tusi, M. Linardi and
E. V. Spinace, J. Power Sources, 2007, 166, 87–91.
G. Ramos Sanchez, O. Solorza-Feria and H. R. Corti, Int. J. 45 L. Ma, C. Liu, J. Liao, T. Lu, W. Xing and J. Zhang,
Hydrogen Energy, 2014, 39, 8821–8826. Electrochim. Acta, 2009, 54, 7274–7279.
9 E. Fuentes-Quezada, E. de la Llave, E. Halac, M. Jobb ´a gy, 46 E. Antolini, L. Giorgi, F. Cardellini and E. Passalacqua, J.
2
F. A. Viva, M. M. Bruno and H. R. Corti, Chem. Eng. J.,
019, 360, 631–644.
0 Y. R. J. Thomas, M. M. Bruno and H. R. Corti, Microporous
Solid State Electrochem., 2001, 5, 131–140.
47 E. Antolini and F. Cardellini, J. Alloys Compd., 2001, 315,
118–122.
2
3
3
3
Mesoporous Mater., 2012, 155, 47–55.
1 M. M. Bruno, N. G. Cotella, M. C. Miras and C. A. Barbero,
Colloids Surf., A, 2010, 362, 28–32.
2 AEROSIL® – Fumed Silica Technical Overview, https://
www.silica-specialist.com/product/silica/en/service-center/
downloads/pages/general.aspx.
48 J. W. Guo, T. S. Zhao, J. Prabhuram, R. Chen and
C. W. Wong, Electrochim. Acta, 2005, 51, 754–763.
49 J.-H. Choi, K.-W. Park, B.-K. Kwon and Y.-E. Sung, J.
Electrochem. Soc., 2003, 150, A973.
50 G. Wu and B. Q. Xu, J. Power Sources, 2007, 174, 148–158.
51 L. Li and Y. Xing, Energies, 2009, 2, 789–804.
3
3 G. Ramos-S ´a nchez, M. M. Bruno, Y. R. J. J. Thomas, 52 Z. Jusys, J. Kaiser and R. J. J. Behm, Electrochim. Acta, 2002,
H. R. Corti and O. Solorza-Feria, Int. J. Hydrogen Energy,
012, 37, 31–40.
4 R. Anderson, J. Catal., 1964, 3, 50–56.
47, 3693–3706.
2
53 H. Wang and H. Baltruschat, J. Phys. Chem. C, 2007, 111,
7038–7048.
3
3
5 S. J. Gregg, K. S. W. Sing and K. S. Sing, Adsorption, Surface 54 R. Rashidi, I. Dincer, G. F. Naterer and P. Berg, J. Power
Area, and Porosity, Academic Press, Cambridge, MA, 2nd
edn, 1991.
6 H. A. Gasteiger, N. M. Markovic, P. N. Ross and E. J. Cairns, J.
Phys. Chem., 1993, 97, 12020–12029.
7 Y. Takasu, T. Kawaguchi, W. Sugimoto and Y. Murakami,
Electrochim. Acta, 2003, 48, 3861–3868.
Sources, 2009, 187, 509–516.
55 H. Wang, L. R. Alden, F. J. DiSalvo and H. D. Abru n˜ a,
Langmuir, 2009, 25, 7725–7735.
56 A. Miura, M. E. Tague, J. M. Gregoire, X.-D. Wen, R. B. van
Dover, H. D. Abru n˜ a and F. J. DiSalvo, Chem. Mater., 2010,
22, 3451–3456.
3
3
3
3
8 Z. Jusys, J. Kaiser and R. J. Behm, Langmuir, 2003, 6759–6769. 57 F. A. Viva, G. A. Olah and G. K. S. Prakash, Int. J. Hydrogen
9 J. Jiang and A. R. Kucernak, J. Electroanal. Chem., 2002, 520,
4–70.
0 J. Jiang and A. Kucernak, J. Electroanal. Chem., 2003, 543,
87–199.
Energy, 2017, 42, 15054–15063.
58 M. Tian, S. Shi, Y. Shen and H. Yin, Electrochim. Acta, 2019,
293, 390–398.
6
4
1
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 30631–30639 | 30639