GO composites as oxygen reduction catalysts

ARTICLE IN PRESS

JID: JECHEM

[m5G;January 20, 2016;12:1]

Journal of Energy Chemistry xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Journal of Energy Chemistry journal homepage: www.elsevier.com/locate/jechem

http://www.journals.elsevier.com/ journal-of-energy-chemistry/

S-doped carbon aerogels/GO composites as oxygen reduction catalysts Mykola Seredych a, Krisztina László b, Enrique Rodríguez-Castellón c, Teresa J. Bandosz a,∗

Q1

a

The City College of New York and The Graduate Center of CUNY, Department of Chemistry, 160 Convent Avenue, New York, NY 10031, USA Department of Physical Chemistry and Materials Science, Budapest University of Technology and Economics, Budapest 1521, Hungary c Departamento de Química Inorgánica, Universidad de Málaga, Málaga 29071, Spain b

a r t i c l e

i n f o

Article history: Received 12 October 2015 Revised 15 October 2015 Accepted 25 October 2015 Available online xxx Keywords: Oxygen reduction reaction Carbon aerogel/GO composites Porosity Surface chemistry Specific interactions

a b s t r a c t Composites of carbon aerogel and graphite oxide (GO) were synthesized using a self-assembly method based on dispersive forces. Its surface was modified by treatment in hydrogen sulfide at 650 and 800 °C. The samples obtained were characterized by adsorption of nitrogen, TA-MS, XPS, potentiometric titration, and HRTEM and tested as catalysts for oxygen reduction reactions (ORR) in an alkaline medium. The synergistic effect of the composite (electrical conductivity, porosity and surface chemistry) leads to a good ORR catalytic activity. The onset potential for the composite of carbon aerogel heated at 800 °C is shifted to a more positive value and the number of electron transfer was 2e− at the potential 0.68 V versus RHE and it increased to 4e− with an increase in the negative values of the potential. An excellent tolerance to methanol crossover was also recorded. © 2016 Science Press and Dalian Institute of Chemical Physics. All rights reserved.

1

1. Introduction

2

Platinum/carbon (Pt/C) based cathodes, which are used for the oxygen reduction reaction (ORR) [1,2], are expensive and show a low tolerance to fuel crossover. This caused an extensive search for effective non-noble metal catalysts [3–12] or even better, for non-metal based catalysts [13–19]. An example of the latter are heteroatom-doped carbonaceous materials including graphene oxide, graphene and nanoporous carbons [10,13–19] or graphiticcarbon nitride together with its composites with carbons [20,21]. Extensive research in this field indicated that heteroatoms such as nitrogen [5,19,20–22], sulfur [14–16,22], phosphorus [17,18], or boron [17,23,24] provide catalytic sites on which oxygen reduction is enhanced [15,16,23]. Among all heteroatom doped carbonaceous materials, the three-dimensional flowerlike nitrogen-doped carbon showed the highest number of electron transfer (n) and the kinetic-current density (Jk ) [25] reaching 3.96e− and 4.02 mA/cm2 at –0.50 V versus SCE (0.51 V versus RHE), respectively. On the other hand, for the nitrogen-modified graphene at the potential –0.50 V versus Ag/AgCl (0.48 V versus RHE) the number of electron transfer was close to 3.3e− and the kinetic current density—7.8 mA/cm2 [22]. Dual S and N doped mesoporous graphene showed n = 3.3e− and Jk = 24.5 mA/cm2 at –0.50 V versus Ag/AgCl (0.48 V versus RHE) and a high tolerance to methanol crossover [15].

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

∗

Corresponding author. Tel.: +1 212 650 6017; fax: +1 212 650 6107. E-mail address: [email protected] (T.J. Bandosz).

Even though doping with nitrogen was the first modification path to explore in order to increase the catalytic activity of carbon based materials [5,19,20–22], an introduction of sulfur along with nitrogen, as indicated above, resulted in even more promising properties based on a synergistic effect of these two heteroatoms [15]. Recent results have also shown that doping only with sulfur might result in efficient oxygen reduction catalysts [14–16,22]. On sulfur-doped graphene the number of electron transfer was found to be 3.82e− and the kinetic current density reached 9.34 mA/cm2 at –0.30 V versus Ag/AgCl (0.68 V versus RHE) [14]. The catalytic activity for oxygen reduction was linked to change in electronic properties of the carbon matrix [14–16], change in the spin and in charge density [14,15], and defects related to the polarity and larger size of sulfur [16,26]. Recently we have shown that hydrophobicity introduced by sulfur in thiophenic compounds is important for withdrawing oxygen from the electrolyte and its physical adsorption on the surface [26,27]. Moreover, location of these groups in small pores seemed to have a marked effect on the physical adsorption process. There was also an indication that other bulky sulfur surface groups in configurations with oxygen (sulfoxide, sulfones and sulfonic acids) located in larger pores (mesopores), owing to their sizes, attracted an electrolyte with dissolved oxygen to the pore system. Not without importance is the electrical conductivity of a catalyst, which promotes the electron transfer to oxygen. As a continuations of the previous research where carbon aerogels doped with sulfur were addressed as efficient ORR catalysts [26], the research presented in this paper focuses on evaluation of the ORR capability of the carbon aerogel/GO composites obtained

http://dx.doi.org/10.1016/j.jechem.2016.01.005 2095-4956/© 2016 Science Press and Dalian Institute of Chemical Physics. All rights reserved.

Please cite this article as: M. Seredych et al., S-doped carbon aerogels/GO composites as oxygen reduction catalysts, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2016.01.005

25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53

JID: JECHEM 2

54 55 56 57 58 59 60 61

62

ARTICLE IN PRESS M. Seredych et al. / Journal of Energy Chemistry xxx (2016) xxx–xxx

using a self-assembly methods [28]. This method, even though it employs the physical/dispersive forces, has been shown as resulting in composites of synergistic properties owing to chemical reactivity of both phases. The new materials obtained are extensively characterized and their activity is linked to surface features including both chemistry and porosity. The ORR performance of the composites was compared to that of a commercial Pt/C catalyst. 2. Experimental

63

2.1. Materials

64

The new synthesized composite consisted of 50:50 wt% graphite oxide:carbon aerogel or S-doped carbon aerogel. The former was obtained using Hummers method [29] and the latter was addressed in details elsewhere [26]. Carbon aerogel was obtained from resorcinol-formaldehyde polymer aerogel [30]. First, GO was well dispersed in water by sonication, and then finely ground carbon aerogel (CA) was added to the GO suspension. The mixture was sonicated for 1 h more and then stirred overnight. Afterward, the suspension was filtrated without washing and dried at 120 °C. The composite obtained is referred to as CA-GO. The S-doped composites, referred to as CA1-GO and CA2-GO, are built with sulfur-doped carbon aerogel, CA1 and CA2 addressed elsewhere, and GO [26]. Sulfur was doped by heating the initial aerogel, CA, at 650 °C (CA1) or 800 °C (CA2), respectively, in H2 S for 3 h (1000 ppm of H2 S balanced in nitrogen, flow rate 150 ml/min).

65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80

81

2.2. Methods

82

2.2.1. Electrochemical characterization The performance of our materials for electrochemical ORR was investigated in 0.1 M KOH using a three-electrode cell with Ag/AgCl/KCl (3 M) as a reference electrode. The measurements of chronoamperometry and long-term stability by applying 1300 potential cycles were carried out on VersaSTAT MC (AMETEK, Princeton Applied Research) with a scanning rate of 5 mV/s (cyclic voltammetry). The working electrode was prepared by mixing the active material with polyvinylidene fluoride (PVDF) and commercial carbon black (carbon black, acetylene, 50% compressed, Alfa Aesar) (8:1:1) in N-methyl-2-pyrrolidone (NMP) until a homogeneous slurry. The slurry was coated on a Ti foil (current collector) with the total surface area of 1 cm2 of an active material. Linear sweep voltammograms (LSVs) were obtained in 0.1 KOH using 757 VA Computrace (Metrohm) at various rotation rates (from 0 to 2000 rpm) with Ag/AgCl (3 M KCl) and Pt wire as a reference and a counter electrode, respectively. The measurements of cyclic voltammetry were carried out under O2 or N2 saturation in the electrolyte in the potential range of 0.19 to –0.8 V versus Ag/AgCl (1.17 V–0.18 V versus RHE) at a scan rate of 5 mV/s. The working electrode was prepared by dispersing 5 mg of the catalyst in 1 ml of deionized water and 0.5 ml of 1 wt% Nafion aqueous solution. About 5 μl of the prepared slurry were dropped (three times) on a polished glassy carbon electrode (Metrohm, Switzerland, diameter 2 mm) and dried at 50 °C in air. The potential was swept from 0.19 to –0.8 V versus Ag/AgCl (1.17 V–0.18 V versus RHE) at a scan rate of 5 mV/s. After each scan, the electrolyte was saturated with air (the source of O2 ) for 20 mins. All the experiments were carried out at room temperature. RHE conversion: The measured potentials versus the Ag/AgCl (3M KCl) reference electrode were converted to the reversible hydrogen electrode

83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113

[m5G;January 20, 2016;12:1]

(RHE) scale using the Nernst equation:

ERHE = EAg/AgCl + 0.059pH + E Ag/AgCl where ERHE is the converted potential versus RHE, EAg/AgCl is the experimental potential measured against the Ag/AgCl reference electrode, and Eo Ag/AgCl is the standard potential of Ag/AgCl (3 M KCl) at 25 °C (0.210 V). The electrochemical measurements were carried out in 0.1 M KOH (pH = 13) at room temperature; therefore, ERHE = EAg/AgCl + 0.977 V. 2.2.2. Evaluation of porosity Sorption of nitrogen at –196 °C was carried out using an ASAP 2020 (Micromeritics, Surface Area and Porosity Analyzer). Before the experiments, samples were out-gassed at 120 °C to constant vacuum (10−4 Torr). The BET surface area, total pore volumes, Vt , (from the last point of isotherm at relative pressure equal to 0.99), micropore volume, volume of pores less than 0.7 nm and 1 nm, V< 0.7 nm and V< 1 nm , mesopore volumes along with pore size distributions were calculated from the isotherms. The volume of mesopores, Vmeso , represents the difference between total pore and micropore volume. The volume of pores and pore size distributions were calculated using 2D-NLDFT (www.NLDFT.com) assuming a heterogeneous surface of pore walls [31]. 2.2.3. DC conductivity measurements The DC conductivity was measured using a 4-probe method on the pellets with the composition consisting of 90 wt% carbon materials and 10 wt% polytetrafluoroethylene as binder. The prepared composition was pressed by a Carver Press machine applying 2 tons pressure and disk-shaped well-packed pellets with diameter 8 mm were formed. The pellets were dried in an oven for 12 h. The thickness of the pellets was measured by a spring micrometer. The measurement of conductivity was carried out using a Keithley 2400 Multimeter.

115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143

2.2.4. Thermal analysis-mass spectroscopy (TA-MS) Thermogravimetric (TG) curves were obtained using a TA Instrument thermal analyzer (SDT Q 600), which was connected to a gas analysis system (OMNI StarTM ) mass spectrometer. The samples were heated up to 1000 °C (10 °C/min) under a constant helium flow (100 ml/min). From the TG curves, differential TG (DTG) curves were derived. The composition of gases was measured by MS and m/z evolution profiles as a function of temperature were evaluated.

144

2.2.5. X-ray photoelectron spectroscopy (XPS) XPS analysis was collected using a Physical Electronics PHI 5700 spectrometer with non-monochromatic Mg Kα radiation (300 W, 15 kV, 1253.6 eV) for the analysis of the core level signals of C 1s, O 1s, S 2p and with a multichannel detector. Spectra of powdered samples were recorded with the constant pass energy values at 29.35 eV, using a 720 μm diameter analysis area. Under these conditions, the Au 4f7/2 line was recorded with 1.16 eV FWHM at a binding energy of 84.0 eV. The spectrometer energy scale was calibrated using Cu 2p3/2 , Ag 3d5/2 , and Au 4f7/2 photoelectron lines at 932.7, 368.3, and 84.0 eV, respectively. The PHI ACCESS ESCAV6.F software package was used for acquisition and data analysis. A Shirley-type background was subtracted from the signals. Recorded spectra were always fitted using Gauss–Lorentz curves, in order to determine the binding energy of the different element core levels more accurately. The error in BE was estimated to be ca. 0.1 eV.

153

2.2.6. Potentiometric titration Potentiometric titration measurements were performed with an 888 Titrando automatic titrator (Metrohm) set at the mode to collect the equilibrium pH. The samples (0.05 g) were dispersed in

Please cite this article as: M. Seredych et al., S-doped carbon aerogels/GO composites as oxygen reduction catalysts, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2016.01.005

114

o

145 146 147 148 149 150 151 152

154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173

ARTICLE IN PRESS

JID: JECHEM

[m5G;January 20, 2016;12:1]

M. Seredych et al. / Journal of Energy Chemistry xxx (2016) xxx–xxx

3

Fig. 1. Cyclic voltammograms on modified glassy carbon RDE in air-saturated 0.1 M KOH at scan rate of 5 mV/s for the graphite oxide (a), carbon aerogel/GO composite (b) and S-doped carbon aerogel/GO composites CA1-GO (c) and CA2-GO (d).

174 175 176 177 178 179 180

NaNO3 (25 ml, 0.01 M) in a container maintained at 25 °C, equilibrated overnight and continuously saturated with N2 . 0.1000 M NaOH was used as a titrant. The experimental pH window was 3– 10. The samples were acidified with HCl. The experimental data were transformed into a proton binding curves, Q, representing the total amount of protonated sites. The following integral equation relates it to the pKa distribution:

Q (pH ) = 181 182

183 184 185 186 187 188 189 190 191

192 193 194 195



∞ −∞

q(pH, pKa ) f ( pKa )dpKa

(1)

From them pKa distributions and the numbers of groups represented by certain pKa values were calculated [32,33].

2.2.7. Water affinity evaluation The changes in surface hydrophilicity were determined by measuring a water adsorption affinity. The initial carbon-based samples were dried at 120 °C to a constant mass and placed in a closed vessel with constant pressure of water vapor at ambient temperature. After 24 h the TA experiments were carried out using a TA instrument thermal analyzer (SDT Q 600). The weight lost in nitrogen between 30 °C and 120 °C was assumed as an equivalent to the quantity of water adsorbed on the surface.

2.2.8. High-resolution transmission electron microscopy (HRTEM) HRTEM was performed on a JEOL 2100 LaB6 instrument operating at 200 kV. Analyses were performed after the samples were dispersed in ethanol.

3. Results and discussion

196

3.1. Electrochemical performance in alkaline medium

197

The cyclic voltammetry (CV) curves for GO, the initial composite and its hydrogen sulfide treated counterparts, are presented in Fig. 1. The first CV cycles of all materials show marked redox humps related to changes in the chemistry caused by the applied potential. These changes seem to be the most pronounced for CA1GO where the reduction humps are the broadest. Since in a subsequent CV cycle the humps do not appear and surface chemistry stabilizes, we link this behavior to the reduction of the species present on the surface, either adsorbed or introduced to the carbon aerogel/graphene oxide matrix. A comparison of CV curves run with and without oxygen in the system shows the oxygen reduction reaction at the potential between 0.86 V and 0.72 V versus RHE with the maximum 0.78 V. The ORR potential for the composites is shifted to more positive value than that for the GO sample (0.63 V versus RHE). The latter sample also shows some reversibility in the redox reaction involving oxygen groups (reversible humps on anodic and cathodic current). Moreover, the samples show the marked capacitive behavior with the highest specific capacitance for CA2-GO reaching 74 F/g. For CA-GO and CA1-GO the measured capacitance values are 53 and 48 F/g, respectively. These values for the carbon aerogels, CA1 and CA2, measured at the same conditions were reported as 16, 25 and 42 F/g, respectively [26]. The differences in the onset potential for ORR are seen in Fig. 2. Its shift to a more positive potential is clearly seen for

198

Please cite this article as: M. Seredych et al., S-doped carbon aerogels/GO composites as oxygen reduction catalysts, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2016.01.005

199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222

JID: JECHEM 4

ARTICLE IN PRESS

[m5G;January 20, 2016;12:1]

M. Seredych et al. / Journal of Energy Chemistry xxx (2016) xxx–xxx

Fig. 2. Linear sweep voltammograms on the modified glassy carbon RDE in air-saturated 0.1 M KOH at 2000 rpm and scan rate of 5 mV/s for the initial carbon aerogels (a) and graphite oxide and carbon aerogel/GO composites compared to 20 wt% Pt on Vulcan XC72 (b).

223 224 225 226 227 228 229

CA2-GO. The onset potentials estimated from LSV are 0.773 V, 0.819 V, 0.821 V and 0.829 V versus RHE for GO, CA-GO, CA1GO and CA2-GO, respectively. They are shifted to the more positive potential than those found for the carbon aerogel and its sulfur-doped counterparts (CA 0.780 V; CA1 0.766 V; CA2 0.809 V versus RHE). Even though the onset potential is less positive than that for Pt/Vulcan (0.947 V versus RHE), the ad-

dition of GO significantly improved the performance of the catalyst. LSV on modified glassy carbon RDE at different rotation speeds was run after the stabilization of the electrodes in the N2 purged electrolyte (Fig. 3). The number of electron transfer (n) is presented in Fig. 4(a). While practically no differences are found between CAGO and the composite built with the carbon aerogel modified with

Fig. 3. Linear sweep voltammograms on modified glassy carbon RDEs in air-saturated 0.1 M KOH at different rotation speeds and scan rate of 5 mV/s for the graphite oxide (a), carbon aerogel/GO composite (b) and S-doped carbon aerogel/GO composites CA1-GO (c) and CA2-GO (d).

Please cite this article as: M. Seredych et al., S-doped carbon aerogels/GO composites as oxygen reduction catalysts, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2016.01.005

230 231 232 233 234 235 236

JID: JECHEM

ARTICLE IN PRESS

[m5G;January 20, 2016;12:1]

M. Seredych et al. / Journal of Energy Chemistry xxx (2016) xxx–xxx

5

Fig. 4. (a) Number of electron transfer versus potential and (b) kinetic current density for the materials studied.

Fig. 5. (a) Chronoamperometry response for methanol tolerance and (b) stability testing by cycling and chronoamperometry at 0.73 V versus RHE.

237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263

hydrogen sulfide at 650 °C (CA1-GO), the composite containing the carbon aerogel treated at 850 °C (CA2-GO) shows a significant improvement with n reaching 4e− at the potential 0.18 V versus RHE. Interestingly, the kinetic current density for CA2-GO does not follow that trend and it is the highest at 0.68 V versus RHE reaching 6.0 mA/cm2 . Then CA2-GO, with a shift of the potential to a most positive value, outperforms the other two composites (Fig. 4b). The kinetic current density at 0.68 V versus RHE for CA-GO, CA1-GO and CA2-GO was 3.8 mA/cm2 , 3.1 mA/cm2 and 3.6 mA/cm2 , respectively [26]. The stability of the catalyst to a methanol crossover is shown in Fig. 5(a) where upon addition of three methanol spikes of 200 μl no change in the cathodic current density was found. This tolerance is better than that on a Pt/C catalyst (commercial 20 wt% Vulcan Pt/C), where some shift in a current from cathodic to a reversed anodic is observed [9,26]. The long-term stability of the catalyst by applying 1300 potential cycles shows initially a relatively low stability for CA2-GO compared to that for CA1-GO (drop to 84% after first 200 cycles) but then, during the cycling, the performance stabilizes. On the other hand, on the stability curves for CA1-GO two steps are found. After 400 and 1000 potential cycles, its kinetic current density decreases to 95 and 75% of its original value, respectively. The CV curves recorded after stabilization under oxygen saturation of the electrolyte and after 1300 potential cycles at 0.73 V versus RHE of ORR (Figure S1 of Supplementary Information) indicate that this decrease in a long-term stability is related to a decrease in pseudocapacitance. It is likely caused by

the decomposition of surface groups. The cathodic peak of ORR at 0.73 V versus RHE after applying 1300 potential cycles significantly decreases in its intensity and slightly shifts to a less positive potential indicating that the catalyst studied has low stability (Figure S1 of SI). S-doped carbon aerogel (CA1 and CA2) showed 32% and 30% decrease in a kinetic current density upon cycling at the same conditions [26].

264

3.2. Characterization of carbon aerogel/GO composites

271

To explain the above discussed behavior and clearly indicate the advantage of the S-doped composites over the aerogel counterparts, the detailed surface characterization was carried out. Parameters of the porous structure (Table 1) show a decrease in a surface area and volume of micropores when the sulfur doped aerogels were used to form the composites. On the other hand, the volume of mesopores increased markedly. These values might be important for the transport of oxygen in electrolyte to ultramicropores where its reduction might take place. The enhanced adsorption potential in such pores might promote a strong oxygen adsorption on the catalytic sites. Interestingly, DC conductivity of the sulfurdoped composites decreased an order of magnitude compared to the conductivity of the carbon aerogel counterparts. An increase in the affinity to adsorb water from about 6% on CA to over/close 30% on the composites might be caused by the addition of oxygen rich graphite oxide phase. It is well known the surface hydrophilicity might help with transport of water-based electrolyte to the pore

272

Please cite this article as: M. Seredych et al., S-doped carbon aerogels/GO composites as oxygen reduction catalysts, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2016.01.005

265 266 267 268 269 270

273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288

ARTICLE IN PRESS

JID: JECHEM 6

[m5G;January 20, 2016;12:1]

M. Seredych et al. / Journal of Energy Chemistry xxx (2016) xxx–xxx Table 1. Parameters of the porous structure calculated from the nitrogen adsorption isotherms, the sulfur content, the amounts of water adsorbed, and the samples’ conductivity (σ ). Sample

SBET (m2 /g)

Vt (cm3 /g)

Vmeso (cm3 /g)

Vmic (cm3 /g)

V<

CA CA1 CA2 CA-GO CA1-GO CA2-GO

693 726 734 392 458 412

0.914 0.952 0.934 1.417 1.587 1.113

0.663 0.690 0.667 1.293 1.456 0.974

0.251 0.262 0.267 0.124 0.131 0.139

0.175 0.176 0.191 0.074 0.082 0.087

0. 7 nm

(cm3 /g)

V< 1 nm (cm3 /g)

Sulfur (XRF, wt%)

H2 O (wt%)

σ (S/m)

0.195 0.199 0.209 0.099 0.105 0.110

0.00 2.68 3.21 0.29 1.54 1.87

8.36 6.22 5.55 28.57 32.45 23.12

11.3 15.2 16.6 6.2 3.6 8.3

DC conductivity for GO is 1.1 × 10− 3 S/m; the amount of H2 O adsorbed by GO is 39.8 wt%; GO is non-porous (SBET = 6 m2 /g)

Fig. 6. Texture characterization of the materials studied: (a) GO, (b) CA-GO, (c) CA1-GO and (d) CA2-GO.

289 290 291 292 293 294 295

system [34]. The surface content of sulfur is less than that in H2 S treated carbon aerogels (2.68 wt% in CA1 and 3.21 wt% in CA2 [26]) owing to only 50 wt% of the S-doped aerogel components in the composites. On the HRTEM images for the initial and modified sulfur-doped carbon aerogel/GO composites only very corrugated and rather small units of GO are seen. In the majority of surface amorphous

microtexture is detected (Fig. 6). The visual analysis of the units resembling the graphite structure in the composites indicates their interlayer distance of 0.35 nm. The lack of obvious and separated units of GO, even with 50 wt% of this phase in the samples supports the composite formation. The surface chemical composition was determined by XPS and the atomic contents of elements are presented in Table 2. The

Please cite this article as: M. Seredych et al., S-doped carbon aerogels/GO composites as oxygen reduction catalysts, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2016.01.005

296 297 298 299 300 301 302

Q2

ARTICLE IN PRESS

JID: JECHEM

[m5G;January 20, 2016;12:1]

M. Seredych et al. / Journal of Energy Chemistry xxx (2016) xxx–xxx Table 2. Content of elements on the surface (in at%).

303 304 305 306 307 308 309 310 311

Sample

C

O

S

GO CA-GO CA1-GO CA2-GO

66.9 93.0 90.1 92.2

32.0 6.6 8.1 5.3

1.1 0.4 1.8 2.5

addition of 50 wt% GO with over 30 wt% of oxygen increased the surface content of oxygen only 1 wt% compared to CA [26] and this is a quite unexpected result suggesting a significant reduction of surface groups upon the composite formation [28,35]. Unexpected is also a higher content of sulfur on the surface of samples treated at 800 °C compared to that at 650 °C. Deconvolution of S 2p core energy level spectra (Fig. 7) and the binding energy of the detected species (Table 3) suggest that sulfur on the surface of GO, which is originally present as sul-

fates/sulfonic acids is reduced by the carbon aerogel phase increasing the diversity of its surface functionalities. Nevertheless, still its majority is in the oxidized form. On the other hand, on the surface of composites the reduced sulfur compounds predominate, and as expected [36], treatment at higher temperature lead to more sulfur with a binding energy of 164 eV. This is in agreement with a low oxygen content. This sulfur is expected to change the electronic properties of carbon creating active centers for oxygen reduction [15] and providing hydrophobicity in small pores for oxygen withdrawal from electrolyte/adsorption on the carbon surface [37]. Sulfur, when incorporated to aromatic ring results in a slightly positive charge on the neighboring carbon atoms owing to its higher electronegativity than that of carbon. These positive charged carbon atoms sites were indicated as sites for oxygen reduction [15]. That sulfur will also bring additional hydrophobicity enhancing molecular oxygen adsorption in these catalytically active pores [37]. The deconvolutions of O 1s and C 1s core energy level spectra also suggest that oxygen is in lower oxidation level

Fig. 7. C 1s, O 1s and S 2p core energy levels for the materials studied.

Please cite this article as: M. Seredych et al., S-doped carbon aerogels/GO composites as oxygen reduction catalysts, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2016.01.005

7

312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329

ARTICLE IN PRESS

JID: JECHEM 8

[m5G;January 20, 2016;12:1]

M. Seredych et al. / Journal of Energy Chemistry xxx (2016) xxx–xxx Table 3. The results of deconvolution of C 1s, O 1s and S 2p core energy levels for the materials studied. Energy (eV)

Bond assignment

C 1s 284.6–284.9 286.3–286.9 287.6 288.9–289.2 290.8

C–(C, S) (graphitic carbon) C–O, C–H (phenolic, alcoholic, etheric) C=O (carbonyl or quinone) O–C=O (carboxyl or ester) Carbonate, occluded CO, π -electrons in aromatic ring

49.09 49.20

O 1s 530.8–531.5 532.6–533.1 534.3–534.7

O=C/O=S (in carboxyl/carbonyl or sulfoxides/sulfones) O–C/O–S (in phenol/epoxy or thioethers/sulfonic) –O– (in carboxyl, water or chemisorbed oxygen species)

15.52 82.19 2.30

S 2p3/2 163.9–164.2 165.7–166.2 168.2–168.4 169.8–170.5

Ph–SH (in thioles), thiophenes C–S–C/R–S2 –OR (in sulfides or thioethers) R–SO2 –R, R–SO3 H, SO4 2− (in sulfone, sulfonic acids, sulfate) RO2 –S–S–R, R–SO3 H (in sulfonic acids)

Fig. 8. The amount of total acidic groups for the initial GO and composites, after washing with H2 O and filtrate.

330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 Q3 352 353 354 355 356 357

in CA2-GO than that in CA1-GO, which should lead to higher hydrophobicity of the former samples. Since the CV experiments on a modified glassy carbon RDE (Fig. 1) showed a marked instability of the surface of our composites in the first cycle run, and the contents of oxygen detected by XPS are rather small, we feel that this behavior requires further investigation. XPS is a surface analysis carried out after a high vacuum outgassing and it is possible that some small molecule compounds retained on the surface might be removed during the sample preparation. This hypothesis was formulated after the analysis of the potentiometric titration results which showed a marked, and unexpectedly high, surface acidity after the treatment with H2 S, even that at 800 °C (Fig. 8). Such high number of surface acidic groups would result in the high oxygen content and it was not the case. Therefore, we extensively washed the composite samples with distilled water and then performed the titration of the filtrates and the surface. The results showed that washing removed strong acids from the materials. The pH for filtrate was 3.68, 3.54 and 3.20 for CA-GO, CA1-GO and CA2-GO, respectively. These acids might explain the cathodic reduction on the initial CV runs. It is likely that get reduced to H2 S and removed from the system (purged with nitrogen). Of course, even after removal of these acids some hydrophilic groups containing oxygen are present on the surface and they promote the transport of electrolyte with dissolved oxygen to small pores. Fig. 8 shows a marked decrease in the number of acidic groups after washing and their relocation to the solution. The results suggest that some acids were in the free form, deposited on the sur-

GO

1.71

100

CA-GO

CA1-GO

CA2-GO

73.01 12.52 7.45 3.88 3.15

63.72 17.98 10.35 4.67 3.28

74.47 12.44 6.42 3.67 3.00

23.35 69.04 7.62

18.28 71.23 10.50

33.67 58.39 7.94

16.08 3.98 70.15 9.80

59.18 13.03 21.91 5.88

66.10 9.73 19.83 4.34

face. That decrease in % is much greater for the composites than for GO and its extent is similar to all three samples. Knowing the ability of carbon surface to activate oxygen [38], it is possible that some sulfur present on the surface got strongly oxidized when samples were exposed to air and adsorbed water resulted in the deposition oxysulfur acids in the pore system. This behavior is a specificity of these composites and we did not encounter it previously in our study of surface modification of nanoporous using hydrogen sulfide. Apparently the carbon aerogel, its low density and thin pore walls in combination with GO lead to this effect. Very peculiar is also a low content of oxygen on the surface for the composite not modified with H2 S and having only residual sulfur from GO. Potentiometric titration of its surface also shows a very high acidity and high number of surface groups. From its surface a marked number of free of acids was removed. In this case, some acids detected on the surface and removed during XPS pretreatment should be based also on oxygen and carbon. In fact the first CV run for CA-GO differs from those for the sulfur treated composites in the huge reduction peak at more negative potential which might represent the reduction reactions of a specific acid. Formed hydrocarbons or hydrogen sulfide might be removed by nitrogen purging. DTG, DTA and m/z 18 thermal profiles (Fig. 9a, b and c, respectively) show a marked weigh loss below 100 °C. The decomposition of epoxy groups of the GO component is at about 190 °C and the shift to higher temperature than that in GO indicates the change in their chemical environment that leads to higher stability. m/z 48 represents the removal of SO from the decomposition of sulfur containing surface groups. Obviously, in the presence of CO2 from the decomposition of epoxy groups, the reduced sulfur compounds can get oxidized [39]. Nevertheless, the pattern shows difference in the position and the width of the main peak on the m/z 48 thermal profile. While for GO it is broad and at relatively low temperature (sulfonic acids), for the composites the maximum of the peaks is at about 350 °C. At this temperature, compounds in more reduced forms decompose. The maximum of the peaks and the weight loss in the temperature range between 200 and 400 °C are similar for all composites regardless their composition. This suggests that it is the GO phase, which oxidized the carbon aerogel and thus it got reduced and led to the formation of small molecule acids on the surface. Based on the results obtained, sulfur species in CA1 and CA2 were the most susceptible to that oxidation process. Thus on the surface of the catalytically active composites sulfur is mainly in the form of reduced thiophenic compounds (as shown by XPS), which can exist in small pores, and which contribute actively to the ORR on these materials. Small pores usually enhance the adsorption potential. Oxygen, which needs to be reduced, is dissolved in water. Hydrophilicity in larger pores promotes the transport of

Please cite this article as: M. Seredych et al., S-doped carbon aerogels/GO composites as oxygen reduction catalysts, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2016.01.005

358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405

JID: JECHEM

ARTICLE IN PRESS

[m5G;January 20, 2016;12:1]

M. Seredych et al. / Journal of Energy Chemistry xxx (2016) xxx–xxx

9

Fig. 9. TA-MS results: (a) differential thermal gravimetry (DTG) curves, (b) differential thermal analysis (DTA) curves and (c–f) m/z thermal profiles for the materials studied.

412

that water to small pores and there, owing to hydrophobic surface, oxygen is withdrawn from water efficiently (owing to enhanced adsorption potential) and then it is strongly adsorbed on catalytic sites on the surface. Then ORR takes place. Apparently the high amount of sulfur doped to the composite matrix combine with its higher contribution in thiophenic compounds results in a better catalytic performance.

413

4. Conclusions

414

The results presented in this paper further support solid phase reactivity between the two phases of GO-carbon aerogel composites. Strong oxidation of carbon aerogel by GO phase results in the presence of “free” acids on the surface. Upon a catalytic application of these materials, the acids have to be removed. Sulfur, originating either from GO or from carbon aerogel, is stabilized as

406 407 408 409 410 411

415 416 417 418 419

predominantly reduced surface species upon the composite formation. Thiophenic compounds located in small pores create centers for oxygen reduction. Moreover, their hydrophobicity increases the efficiency of oxygen adsorption on the surface sites. That oxygen is supplied to the small pores by the developed system of hydrophilic transport pores creating a friendly environment for interactions with the electrolyte.

421 422 423 424 425 426

Acknowledgments

427

The authors are grateful to Mr. Balázs Nagy for the aerogel synthesis. This work was supported by the Spanish Ministry of Economy and Competitiveness (Project CTQ2012-37925-C03-03) and FEDER funds, and by the Hungarian National Fund OTKA K109558.

428

Please cite this article as: M. Seredych et al., S-doped carbon aerogels/GO composites as oxygen reduction catalysts, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2016.01.005

420

429 430 431

JID: JECHEM 10

ARTICLE IN PRESS

[m5G;January 20, 2016;12:1]

M. Seredych et al. / Journal of Energy Chemistry xxx (2016) xxx–xxx

432

Supplementary materials

433 434

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jechem.2016.01.005.

435

References

436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463

[1] B.C.H Steele, A. Heinzel, Nature 414 (2001) 345–352. [2] V.R Stamenkovic, B. Fowler, B.S. Mun, G. Wang, P.N. Ross, C.A. Lucas, N.M. Markovic, Science 315 (2007) 493–497. [3] L. Wang, A. Ambrosi, M. Pumera, Angew. Chem 125 (2013) 14063–14066 Angew. Chem. Int. Ed. 52 (2013) 13818-13821. [4] L. An, W. Huang, N. Zhang, X. Chen, D. Xia, J. Mater, Chem. A 2 (2014) 62–65. [5] M. Lefevre, E. Proietti, F. Jaouen, J.-P. Dodelet, Science 324 (2009) 71–74. [6] J.B. Xu, P. Gao, T.S. Zhao, Energy Environ. Sci 5 (2012) 5333–5339. [7] S. Guo, S. Zhang, S. Sun, Angew. Chem 125 (2013) 8686–8705 Angew. Chem. Int. Ed. 52 (2013) 8526–8544. [8] M. Jahan, Z. Liu, K.P. Loh, Adv. Funct. Mater 23 (2013) 5363–5372. [9] C.O. Ania, M. Seredych, E. Rodriguez-Castellon, T.J. Bandosz, Appl. Catal. B: Environ 163 (2015) 424–435. [10] M. Seredych, E. Rodriguez-Castellon, T.J. Bandosz, J. Mater. Chem. A 2 (2014) 20164–20176. [11] M. Lefevre, J.P. Dodelet, P. Bertrand, J. Phys. Chem. B 106 (2002) 8705–8713. [12] J. Du, F. Cheng, S. Wang, T. Zhang, J. Chen, Sci. Rep 4:4386 (2014) 1–7. [13] Y. Li, W. Zhou, H. Wang, L. Xie, Y. Liang, F. Wei, J.-C. Idrobo, S.J. Pennycook, H. Dai, Nat. Nanotechnol 7 (2012) 394–400. [14] Z. Yang, Z. Yao, G. Li, G. Fang, H. Nie, Z. Liu, X. Zhou, X. Chen, S. Huang, ACS Nano 6 (2012) 205–211. [15] J. Liang, Y. Jiao, M. Jaroniec, S.Z. Qiao, Angew. Chem 124 (2012) 11664–11668 Angew. Chem. Int. Ed. 51 (2012) 1–5. [16] M. Seredych, T.J. Bandosz, Carbon 66 (2014) 227–233. [17] C.H. Choi, S.H. Park, S. Woo, ACS Nano 6 (2012) 7084–7091. [18] D. Deak, E.J. Biddinger, K. Luthman, U.S. Ozkan, Carbon 48 (2010) 3637–3639. [19] M. Zhong, E.K. Kim, J.P. McGann, S.-E. Chun, J.F. Whitacre, M. Jaroniec, K. Matyjaszewski, T. Kowalewski, J. Am. Chem. Soc 134 (2012) 14846–14857.

[20] J. Liang, Y. Zheng, J. Chen, J. Liu, D. Hulicova-Jurcakova, M. Jaroniec, S.Z. Qiao, Angew. Chem 124 (2012) 3958–3962 Angew. Chem. Int. Ed. 51 (2012) 3892– 3896. [21] Y. Zheng, Y. Jiao, J. Chen, J. Liu, J. Liang, A. Du, W. Zhang, Z. Zhu, S.C. Smith, M. Jaroniec, G.Q.M. Lu, S.Z. Qiao, J. Am. Chem. Soc 133 (2011) 20116–20119. [22] S. Yang, L. Zhi, K. Tang, X. Feng, J. Maier, K. Mullen, Adv. Funct. Mater 22 (2012) 3634–3640. [23] S. Wang, L. Zhang, Z. Xia, A. Roy, D.W. Chang, J.-B. Baek, L. Dai, Angew. Chem. Int. Ed 51 (2012) 4209–4212. [24] C.H. Choi, M.W. Chung, H.C. Kwon, S.H. Park, S.I. Woo, J. Mater. Chem. A 1 (2013) 3694–3699. [25] H. Li, W. Kang, L. Wang, Q. Yue, S. Xu, H. Wang, J. Liu, Carbon 54 (2013) 249– 254. [26] M. Seredych, K. László, T.J. Bandosz, ChemCatChem 7 (2015) 2924–2931. [27] M. Seredych, M. Biggs, T.J. Bandosz, Micropor. Mesopor. Mater 221 (2016) 137– 149. [28] M. Seredych, T.J. Bandosz, Carbon 95 (2015) 580–588. [29] W.S Hummers, R.E. Offeman, J. Am. Chem. Soc 80 (1958) 1339. [30] O. Czakkel, K. Marthi, E. Geissler, K. Laszlo, Micropor. Mesopor. Mater 86 (2005) 124–133. [31] J. Jagiello, J.P. Olivier, Adsorption 19 (2013) 777–783. [32] J. Jagiello, Langmuir 10 (1994) 2778–2785. [33] J. Jagiello, T.J. Bandosz, J.A. Schwarz, Carbon 32 (1994) 1026–1028. [34] J.K. Brennan, K.T. Thomson, K.E. Gubbins, Langmuir 18 (2002) 5438–5447. [35] M. Seredych, E. Rodriguez-Castellon, T.J. Bandosz, ChemSusChem 8 (2015) 1955–1965. [36] G.N. George, M.L. Gorbaty, S.R. Kelemen, M. Sansone, Energy Fuels 5 (1991) 93–97. [37] D.A. Doshi, E.B. Watkins, J.N. Israelachvili, J. Majewski, PNAS 102 (2005) 9458– 9462. [38] V.V. Strelko, N.T. Kartel, I.N. Dukhno, V.S. Kuts, R.B. Clarkson, B.M. Odintsov, Surf. Sci 548 (2004) 281–290. [39] T.J. Bandosz, M. Seredych, E. Rodríguez-Castellón, Y. Cheng, L.L. Daemen, A.J. Ramírez-Cuesta, Carbon 96 (2016) 856–863.

Please cite this article as: M. Seredych et al., S-doped carbon aerogels/GO composites as oxygen reduction catalysts, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2016.01.005

464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497