CC BY
9свинца, цинка, натрия и калия при оценке качества питьевой воды и трансфузионных жидкостей. М., Вестник РАМН , №2, 2001.
THE CARBONIZATION TEMPERATURE EFFECT ON THE ELECTROCHEMICAL PROPERTIES OF ELECTROSPUN POLYACRYLONITRILE FIBER PYROPOLYMERS
Davydova E.S.
Rychagov A.Yu.
Kazanskii L.P.
Russian Academy of Sciences A.N. Frumkin Intfitute of Physical Chemitfry and Electrochemitfry, Moscow, Russia
Ponomarev I.I.
Ponomarev Iv.I.
Razorenov D.Yu.
Bukalov S.S.
Skupov K.M.
Russian Academy of Sciences A.N. Nesmeyanov Intfitute of Organoelement Compounds, Moscow, Russia
ABSTRACT
Pyropolymers were synthesized by the carbonization of electrospun polyacrylonitrile nanofibers at 600 - 2800 oC. The effect of the carbonization temperature on the composition and Sructure of the pyropolymers was inve^igated using the methods of elemental analysis, X-ray photoelectron spectroscopy and Raman spectroscopy. Electrochemical properties, such as capacitance and catalytic activity in the oxygen reduction reaction, were determined by means of cyclic voltammetry, electrochemical impedance spectroscopy, rotating disc electrode and rotating ring-disk electrode methods in H2SO4 and KOH solutions. A correlation between the capacitance and catalytic properties for polyacrylonitrile pyropolymers with the be& values at pyrolysis temperature of 850 -900 oC was shown. The capacitance of PAN-900 with the specific surface area of ~100 m2 g-1 can achieve 200 F g-1. It is assumed that such a high value of capacitance could be due to the bulk electroactivity of the pyropolymers, resulting from the potential-dependent charge accumulation ability of the bulk heteroconjugated Sructures. The presence of nitrogen was shown to be a necessary and sufficient condition for electrocatalysis of O2 reduction. However, the polyacrylonitrile pyropolymers provide rather sluggish kinetics of the oxygen reduction reaction proceeding predominantly via the two-electron route in acid solutions. The synthesis of highly active cataly^s, selective to four-electron process, the incorporation of transition metals into pyropolymers is required.
Keywords: capacitance; electrospun nanofiber; oxygen reduction reaction (ORR); polyacrylonitrile pyropolymer.
Introduction
Carbonaceous materials derived from polyacrylonitrile (PAN) heat treatment are called pyropolymers. It is well eflablished that heat treatment of PAN in the temperature range of 250 -2800 oC leads to the formation of various polymer flructures. They may vary from disordered ladder flructures, formed in the process of oxidative cyclization when treated at 200 - 350 oC in air [1,2], to graphite if treated at 2800 oC in inert atmosphere or vacuum [2-4].
In induflry, PAN is widely used for the synthesis of carbon fibers. It is shown [1] that the flage of oxidative cyclization at 200 - 350 oC in air is necessary for decreasing the duration of PAN heat treatment and, also, for improvement of carbon fiber mechanical properties. Oxidation flage precedes the carbonization one which is usually conducted at the temperature higher than 1500 oC. Thus, the temperature interval of 350 -1500 oC (referred to as intermediate temperature range) is not yet of interefl for the induflrial production.
The firfl fludies of PAN fiber carbonization and graphitization were mentioned by Shindo et al. in 1961 [5]. However, the firfl fludies of electrochemical properties of PAN, carbonized at intermediate temperature, appeared only at the end of 1980s [6]. Gupta et al. published the pioneer work fludying composite catalyfls based on PAN and iron and/or cobalt salts adsorbed on carbon support, pyrolyzed in a wide temperature range 300 -1000 oC. The authors showed the samples treated at 800 - 900
oC to be feasible catalyfls for O2 electroreduction (ORR) in acid and base solutions. According to [7], Gupta's fludy was one of the firfl works proving no necessity in macroheterocycles (such as phthalocyanines, porphins, etc.) usage for the synthesis of the non-platinum catalyfls for the ORR. Another paper invefligating the influence of heat treatment in the intermediate temperatures range on the electrochemical properties of PAN-based materials was published by Martins Alves et al. [8]. In this work, the authors fludied the flructural changes and catalytic properties evolution for the composite catalyfls based on PAN and cobalt acetate adsorbed on carbon support. The authors determined the heat treatment at 950 oC to provide the highefl catalytic activity of the composites.
Generally, the fludies of carbonized PAN electrochemical properties were directed to the temperature range of 800 - 900 oC. For example, Ohms et al.[9] invefligated the influence of ion metal nature on the electrocatalytic activity and, partially, corrosion flability in the ORR in acid and base solutions for PAN/carbon black/M (M=Zn, Mn, Ni, Cu, Fe, Co) composites obtained by pyrolysis at 850 oC. In more modern fludies, Wu et al.[10,11], Jeong et al. [12], Nakagawa et al.[13], etc. formed fibers from PAN solution with iron and/or cobalt salts and pyrolized them at 800 - 900 oC. The diflinction of these papers [10-13] as compared to the earlier mentioned [8,9] is that no carbon support was used for the catalyfl synthesis. Here, the
pyrolized PAN nanofiber mats (NFM) themselves play the role of the support and, to some extent, the role of the active mass.
In the papers [14-16] Ye et al. used PAN pyrolized at 900 oC as the supporting material for Pt catalyfls with ultra-low platinum loading. Probably, due to low catalytic activity, the catalyfls were not in demand for low temperature H2-O2 fuel cells. However, as it was shown in Ponomarev et al. [17,18], PAN pyropolymer-supported Pt catalyfls could be successfully used on the anodes of phosphoric acid (middle-temperature) fuel cells.
The interefl to PAN pyropolymers is not reflricted by catalysis issues only. The capacitive properties of different materials based on carbonized PAN at 850 - 900 oC were fludied by Davydova et. al. [19], Gouerec et al.[20], Wang et al.[21], Ania et al.[22], etc. Wang et al. fludied commercial activated PAN fibers (Challenge Carbon Technology Co., Taiwan) with specific surface area of ~1600 m2 g-1. The use of the fibers resulted in the achievement of the capacitance value ~300 F g-1 in the acid solution, which correspond to the specific capacitance of 17 - 24 ^F cm-2. Gouerec et al. [20] carried out activation of pyrolized PAN to produce microporous foamed films. This leads to the specific surface area increase from ~80 for the non-activated pyropolymers to ~600 m2 g-1 for the activated ones. As a result, Gouerec et al. achieved the capacitance value of ~ 170 F g-1 (or, the specific capacitance of 28 ^F cm-2) in base solution. The mofl significant specific surface area (>1600 m2 g-1 ) and, correspondingly, capacitance (~340 F g-1) were obtained in aqueous solutions by Ania et al. [22] due to the pyrolysis of PAN in the pores of etchable zeolite matrix. According to Ania, the major contribution to the value of capacitance is made by the electrochemical activity of pyropolymer heteroatoms.
Thus, in accordance with [22], the heteroatoms play the critical role in giving rise to the capacitance of PAN pyropolymers. However, to increase of mass-specific capacitance of PAN pyropolymers the activation flage via the development of microporous flructure is mandatory [20-23]. The activation results in the decrease of the specific capacitance to 20 - 30 ^F cm-2 as compared to the non-activated pyropolymers, the specific capacitance of which in some cases exceeds 100 ^F cm-2 [24-26].
The papers of Gavrilov et al. [24,25] also should be mentioned in which the polyaniline pyropolymers carbonized at 800 oC were fludied. The authors revealed the interrelation between the capacitive and electrocatalytic properties of polyaniline pyropolymers, however, only at single carbonization temperature.
Insummary, itshouldbe notedthatthe literature available covers the queflions predominantly dealing with the electrocatalytic properties of PAN pyropolymer-metal composites, not with pure PAN pyropolymers themselves. To the befl of our knowledge this
is the firfl fludy where electrocatalytic and capacitive properties of pure PAN pyropolymers are invefligated simultaneously in a wide range of pyrolysis temperature.
The main aim of our fludy is the invefligation of the capacitive and electrocatalytic properties of PAN pyropolymers, obtained by the carbonization of electrospun PAN nanofiber material (NFM) at the temperature interval of 600 - 1200 oC, in 0.5 M H2SO4 and 1 M KOH solutions.
Results and discussion
Some of results conflitute a part of E.S. Davydova's academic thesis.[27] Electrical conductivity of PAN-t NFM. PAN-t NFM electrical conductivity significantly increases when pyrolysis temperature increases. In the temperature range of 600 - 750 oC the mat electrical conductivity increases by three orders of magnitude: from ~2 mS cm-1 to ~2 S cm-1. After the heat treatment of PAN at 2800 oC graphite ordered layer flructure is formed and the conductivity value is over 300 S cm-1.
Cyclic voltammetry. The increase of the pyrolysis temperature (600 oC ^ ~850 oC) is accompanied by the current increase measured by cyclic voltammetry, while further temperature increase (900 oC ^ 1200 oC) leads to the decrease of the current (Fig. 1). For the pyrolysis temperature of 1200 oC (Fig. 1 d), the CV is reversible and the curve shape is close to CVs of traditional carbon materials [28]. CV of PAN-1200 contains reversible maximums which are usually related to the presence of O-containing surface functional groups [28].
For the PAN-600, PAN-750, PAN-850 and PAN-900 samples a flrong dependency of capacity from the potential is evident, which is not typical for traditional carbon materials in acidic electrolytes. Eliad et al. [29] have obtained a similar i - E dependency in sulfuric acid electrolyte on cellulose-based carbon fibers. The authors explain it by the flerical hindrance for large anions chemisorption (compared to H+) in the micropores of carbon materials at the potentials, higher then zero charge potential. Kawaguchi et al. [30] observed similar curve shape for the products of tetraazanaphthalene derivatives pyrolysis in acidic electrolyte. Gavrilov et al. [24] observed the same on polyaniline pyropolymers in basic electrolytes. The authors [24] assume that the CV curve shape is determined by pseudocapacitive properties of N-containing functional groups rather than by the flerical factor. Kawaguchi et al. [30] proposed a scheme of a possible faradaic (pseudo-capacitive) process of the pyridine nitrogen protonation. We also suggefl that the flerical factor does not have a significant influence on the charging character because, as it was shown multiple times in the literature [20,22,31] and was proven in the current paper, for non-activated carbonized polymers (such as PAN-t), the total specific surface is low due to low impact of the micropores.
The temperature range of 850 - 900 oC is optimal and provides the maximum PAN-t capacity (Fig. S12).
Fig. 1. Cyclic voltammograms and O2 reduction polarization curves for PAN-750 (a), PAN-850 (b), PAN-1050 (c) and PAN-1200 (d). 0.5 M H2SO4, 60 °C, Ar, 0.1 V s-1, loading 200 ^g cm-2. [27]
Electrochemical impedance spectroscopy. By means of EIS method PAN-900 was invefligated in a wide range of potentials from -0.8 to 0.0 V with a flep of 0.1 V in 0.5 M H2SO4. The NFM EIS spectra (Fig. 2) are similar to the ones of the traditional carbon materials. Usually, for thin electrodes with highly developed surface the spectra are described by an equivalent scheme of Rs(CPEf(Rf))(CPEdl(RctZW)) [32]. The element (CPEf(Rf)) can be neglected due to high electrical conductivity of the
solution and (CPEdl(RctZW)) can be simplified to CdlZW. The possibility of this simplification is based on an assumption that the double layer capacitance is not shunted by the resiflance of charge transfer, and the double-layer rearrangement is quick. In the case of PAN-900 EIS data in a wide potential and frequency range of 10 kHz ^ 0.1 Hz corresponds to the equivalent circuit presented in Fig. 3. The advantage of the scheme is the avoidance of formal CPE elements.
Fig. 2. Experimental and model EIS spectra (Nyquifl plot) in a frequency range from 10 kHz to 0.1 Hz for PAN-900 electrode (2.1 mg) in 0.5 M H2SO4 at potentials: 1 - 0.1 V; 2 0.4 V. Insert: high frequency region for the spectra at different potentials is shown: 1 - 0.8 V; 2 - 0.7 V; 3 - 0.6 V; 4 - 0.3 V; 5 - 0.0 V.
Fig.3. Equivalent electrical circuit for PAN-900 electrode. [27]
As it was pointed out in Introduction, for PAN-based activated highly dispersed materials the specific mass capacity is in the range of 300 - 400 F g-1 [20-22]. In our case, for non-activated PAN-900 NFM with the low specific surface of ~100 m2 g-1 the capacitance values of ~200 F g-1 were obtained by means of cyclic voltammetry which corresponds to the specific capacity of ~200 |iF cm-2. The capacity value of ~200 |iF cm-2 cannot be entirely explained by double-layer charging since the normal specific capacity is within the range of 5 - 20 |iF cm-2 [33]. To elucidate the reasons of high capacity we examined dependencies of C1, R2 (Fig. 4) and C2 (Fig. 5) values on the electrode potential and hypothesized the physical significance of these elements. According to the Fig. 4, C1 dependency on potential has an extreme character passing through the non-symmetric maximum. The decrease of C1 when the potential shifts to > 0.6 V is correlated with a raise of the resiflance R2. A similar maximum was observed in [34] for activated carbon paper after long treatment in sulfuric acid. Usually, the presence of such kind of maxima is explained by the Faradaic processes of O-containing functional groups. In case of PAN-900 the oxygen content of 4.56 at.% cannot provide such a high value of charge which was observed under C1 maximum (~43 C g-1). The mofl possible explanation of C1 increase would be the process anion adsorption which flops as a result of the surface conductivity loss referred to as the surface "passivation". A comparison of the values of capacity derived from cyclic voltammetry data with the ones calculated from EIS data is shown on Fig. 5. Curve 1 is the capacity - potential dependency derived by means of
cyclic voltammetry at scan rate v = 0.005 V s-1 using the equation C=I v-1. Curve 2 was calculated for the frequency f = 0.1 Hz using the equation Z=(f C)-1, where Z is the imaginary part of the impedance. , The value of the capacity calculated according to the equation Z=(f C)-1 increases significantly with the decrease of the frequency f [19]. Curve 3 corresponds to the values of C2 element. As evident from the Fig. 5 the value of C2 gradually decreases from 220 to 75 F g-1 when passing from negative to positive potentials. The comparison of the curves shows that the capacitance obtained from the cyclic voltammograms (curve 1) is close to the values of the low frequency capacity C2 (curve 3). Thus, we consider the elements to fland for the following: R1 - the resiflance of the electrolyte and contacts; R2 - the charge transfer resiflance; C1 - the capacity of high frequency processes, namely, charging of the interphase boundary and anion adsorption; C2 - the capacity of low frequency processes, namely, charging of the pyropolymer solid phase. To explain the EIS and cyclic voltammetry data the following processes could be assumed to proceed: double layer charging at low potentials with capacitance value of C1~40 F g-1; anion adsorption at middle potentials and surface "passivation" at high potentials; counter-ion doping into the solid phase, compensating the bulk charge C2, arising from potential-dependent charge accumulation ability of the bulk heteroconjugated =C=N-containing flructures of PAN-t. The main factor which influences C2 decrease at positive potentials is, mofl probably, the loss of the solid phase electrical conductivity.
Fig. 4. The dependency of C1 (1), R2 (2) and W (3) values, calculated according to equivalent circuit (Fig. 3) from the potential. [27]
Fig. 5. Volt-farad dependencies for PAN-900 with in 0.5 M solution: 1 - from cyclic voltammetry data at scan rate 0.005 V s-1; 2 - from impedance data at 0.1 Hz; 3 - C2 capacitance. [27]
The ORR regularities
RDE and RRDE data on PAN-t NFM and PAN-900/KjBl. Materials such as PAN-t and composites on the basis of PAN are of interefl as nitrogen-doped metalless electrocatalyfls for the ORR. The fludy of the ORR regularities on such catalyfls could provide fundamental information on whether N-doping can catalyze direct water formation or not.
0.5 M H2SO4. According to the polarization curves given on Fig. 1 metalless PAN-t pyropolymers provide rather poor catalytic activity of ORR. The activity does not exceed 0.4 F g-1 at 0.6 V (Fig. S13). On PAN-1050 and PAN-1200 the current rises exponentially with potential shift (Fig. 1 c, d). Tafel slope for PAN-1050 is -0.145 V per decade at the potential region of 0.6 - 0.5 V RHE and -0.250 V per decade at 0.45 - 0.3 V RHE. The slope is -0.280 V per decade at 0.7 - -0.4 V RHE is observed for PAN-1200.
For PAN-600, PAN-750, PAN-850 and PAN-900 bends and ill-defined limiting currents are observed on the polarization curves (Fig. 1 a, b), arising not from O2 diffusion limitations but rather from the deficiency of active cites. Tafel slopes are -0.145-^-0.155 V per decade at 0.7 - 0.4 V RHE; the same slope is observed on the carbon black Vulcan XC72 at 0.5 - 0.3 V RHE.
The values of onset potentials for PAN-750, PAN-1050 and PAN-1200 are close to the thermodynamic equilibrium potential of the reaction (0.695 V NHE). The onset potential for the mofl active PAN-850 and PAN-900 lies in the range of 0.75 - -0.78 V RHE, probably due to the proceeding of 4-electron direct water formation reaction according to the equation (1.229 V NHE), but Sill with the domination of the hydrogen peroxide formation reaction. The maximum of PAN-t catalytic activity observed at 900 °C correlates with the capacitance maximum position.
Polarization curves of O2 reduction (ID) for PAN-900 and PAN-900/KjBl in 0.5 M H2SO4 at 150 rpm are shown on Fig 6. The values of H2O2 oxidation currents (IR) on the ring electrode are related to the collection efficiency N=0.25 and normalized to the disc electrode surface (0.2 cm2). The value of limiting diffusion current calculated according to the equation (3) for n=4 equals 1.3 mA cm-2 at 150 rpm which is shown on Fig. 6 with dashed line. If H2O2 is not formed in the ORR, the value of IR/N would be close to zero and the value ID would be equal to for n=4. From the other point, if H2O2 is formed only, then the values of IR/N and ID would be equal. As it is seen from Fig. 6, for PAN-900 and PAN-900/KjBl, IR/N is lower than ID, and ID is lower than , i.e. the ORR undergoes according to parallel-sequential mechanism [35].
I N I mA cm
/„/mA cm'
Fig. 6. O2 reduction polarization curves on the disc electrode (ID) and H2O2 oxidation currents on Pt-ring (IR) for PAN-900/Fe/ KjBl (1, 1'), PAN-900/KjBl (2, 2'), Ketjenblack EC-300 (3, 3') and PAN-900 (4, 4'). RRDE, 1 M KOH, 02, 0.001 V s-1, 25 °C, 150 rpm, loading 200 ^g cm-2. for n=4 (1.3 mA cm-2) is shown with dashed lines.
Thus, on Fig. 6, by means of RRDE method the H2O2 yield on PAN-900 and PAN-900/KjBl was shown to be lower than 100%. According to Fig. 7 and equation (4) the ratio k1/k2 for PAN-900/KjBl was shown to be 0.2 - 0.4 for in the potential interval of 0.1 - 0.45 V RHE proving the partial proceeding of the ORR via the direct 4-electron route. The composite PAN-900/KjBl has the same Tafel slope as PAN-900 (-0.155 V per
decade). Comparing the curves 2(2') and 3(3') (Fig. 6) it is seen that PAN-900 is more active than PAN-900/KjBl (1.2 vs 0.8 A g-1 at 0.45 V RHE). However, the catalytic activity normalized to the mass unit of PAN pyrolysis product (PAN-900 contains 100 wt.% and PAN-900/KjBl - 22 wt.% of PAN pyrolysis products) was higher for PAN-900/KjBl (3.6 vs 1.2 A g-1 at the same potential).
IdXN/Ir 2
1.6
1.2
Д
0.1
Д *
0.15
0.2
ю-1'5 / rad-05 s-0'5
A 0.4 Ж 0.3 + 0.2 -0.1
0.25
Fig. 7. IDxN/IR - œ-05 dependencies for PAN-900/KjBl. RRDE, 0.5 М H2SO4, 25 °C. Collection efficiency N=0.25.
1.
1.4
1
1 M KOH. O2 reduction polarization curves on carbon black Ketjenblack EC-300 (curve 3), PAN-900 (curve 4) and PAN-900/KjBl (curve 2), and corresponding H2O2 oxidation currents in 1 M KOH are shown on Fig. 8. It is well known [36] that the ORR on carbon materials in highly basic electrolytes proceeds via the sequential mechanism with H2O2 producing. As it is seen from RRDE data (curves 2 - 4, Fig. 8), this mechanism is also true for the ORR on highly-N-doped carbon materials, namely
I W'/mA cm"2
R
2.0
PAN-900 and PAN-900/KjBl composite. This is proved by the following formal features: the appearance of two half-waves with the heights equal to the half value of for n=4, and the equality of the values for IR/N and ID at the region of the firfl half-wave. From Fig. 8 it is evident that the catalytic activity of PAN-t catalyfls is rather poor in the ORR in basic solution. Low activity of PAN-based metal-free carbon fibers was mentioned earlier by Jeong et al. [12].
£ vs RHEfV
Fig. 8. O2 polarization curves on the disc electrode (ID) and H2O2 oxidation currents on Pt-ring (IR) for PAN-900/Fe/KjBl (1, 1'); PAN-900 (2, 2'); PAN-900/KjBl (3, 3'). RRDE, 1 M KOH, 02, 0.001 V s-1, 25 °C, 660 rpm, loading 200 ^g cm-2. for n=4 (2.4 mA cm-2) is shown with dashed line.
RRDE data on the composite PAN-900/Fe/KjBl 0.5 M H2SO4. Comparing the curves 1 and 3 (Fig. 6) shows that the addition of 2 wt% Fe into the composite PAN-900/Fe/ KjBl leads to significant shift of the polarization curves toward the positive potentials. The ORR flarts at ~0.85 V RHE on PAN-900/Fe/KjBl catalyfl. PAN-900/Fe/KjBl catalyfl activity is 0.5 - 8 A g-1 at 0.8 V (or, 1.4 - 1.8 A g-1 at 0.76 V). The RRDE measurements made for the catalyfl loading of 200 ^g
cm-2 show draflic increase in the ORR selectivity to H2O: H2O2 yield decreases to 40 %. But it is known from literature [37] and from our fludies [38] that for Pt-free catalyfls, that the higher the catalyfls loading (or, the catalyfl layer thickness) the lower H2O2 yield. At 200 ^g cm-2 the ratio k1/k2 is ~2 - 3 at 0.65 - 0.75 V RHE with gradual decrease to ~1 at 0.2 V RHE as evidenced from Fig. 9 and equation (4).
I D?N /I R
■ 0.7 X0.6 O 0.5 -0.4 «0.3 □ 0.2
0.15 5 / rad- "
Fig. 9. IDxN/IR - ra-0-5 dependencies for PAN-900/Fe/KjBl. RRDE, 0.5 M H2SO4, 25 °C. Collection efficiency N=0.25. [27]
1 M KOH. Compared to metalless PAN-900/KjBl the composite PAN-900/Fe/KjBl catalyzes the ORR via parallel-sequential mechanism with k1/k2 ratio of ~4 - 5 at 0.7 - 0.8 V RHE which decreases to ~1 at 0.2 V RHE. As in the case of acid media, in alkaline solution the contribution of electrochemical H2O2 reduction and its chemical decomposition is also negligible. PAN-900/Fe/KjBl composite has high catalytic activity (up to 3 A g-1 at 0.98 V RHE). The average value of Tafel slopes are ~-0.031 V per decade at 0.96 - 0.94 V RHE and ~-0.066 V per decade at 0.94 - 0.88 V RHE.
Thus, PAN-t pyropolymers primarily catalyze H2O2 formation in 0.5 M H2SO4 with the maximum reaction rate at 900 °C pyrolysis temperature. PAN-900, as well as metalless PAN-900/KjBl composite, provides the ORR via parallel-sequential mechanism with the domination of H2O2 formation (k1/k2=0.2 - 0.4). Fe-containing PAN-900/Fe/KjBl composite was shown to be a feasible cathode catalyfls in acid media, providing the ORR with k1/k2=2 - 3 in acid medium and 4 - 5 in alkaline medium at low values of loading. In contrafl to PAN-900/Fe/ KjBl composite, PAN-t and metalless composites on its basis are poorly active in 1 M KOH.
Conclusions
It was shown that the temperature of PAN nanofibers pyrolysis has the key role on the electrochemical properties of resulting N-doped carbon materials. The highefl values of capacity and catalytic activity in respect to ORR in alkaline and acid media for PAN nanofiber mats are observed with the pyrolysis temperature of ~900 °C. Metalless N-doped can facilitate mainly 2-electron route of H2O2 formation with k1/k2= 0.2 - 0.4 in acid medium and k1/k2= 0 in alkaline medium. Sharp increase in catalytic activity and selectivity both in acid (with k1/k2= 2 -3) and alkaline (with k1/k2= 4 - 5) is observed for Fe-containing PAN-based composites. High values of specific capacity of PAN nanofiber mats amounting to ~200 ^F cm-2 for PAN-900 were revealed in acid solutions. We supposed that this fact could be referred to volumetric charge accumulating ability of N-rich PAN pyropolymers with the potential-dependent capacity value of C2.
Experimental
The materials invefligated in the paper were synthesized by the pyrolysis of electrospun PAN fibers or impregnated PAN/ carbon black composite materials in the wide temperature interval (600 - 2800 oC). The details of the synthesis methodologies are given in Supporting Information. The methodology and results of elemental analysis, Raman spectroscopy, X-ray photoelectron spectroscopy and electron microscopy are provided ibidem.
Measurements in three-compartment electrochemical cell. The experiments were performed in glass thermoflated three-
electrode electrochemical cell. Pt grid was used as the counter electrode. Hg/Hg2SO4/0.5 M H2SO4 with a potential of +0.695 V RHE was used as the reference electrode in acidic solutions; Hg/HgO/1 M KOH with a potential of +0.925 V RHE - in base solutions. A disc electrode made of pyrographite with working area 0.2 cm2 was used as the working electrode. A suspension was prepared by ultrasound dispersion of 2 mg of sample in 1 ml of ethanol with Nafion solution (5 wt%) an aliquot of which was applied on the disc to obtain the catalyfls loading of 200 ^g cm-2. This quantity was chosen in accordance with our previous paper [38] in order to form equally accessible layer.
Measurements in two-compartment cell. The fludies were made in a filter press cell [39] with carbon electrical leads and polypropylene separator (200 ^m). Non-porous foil made of thermosplitted graphite was used for electrical leads. The electrodes and separators were cut in the form of a disc with 0.2 cm in diameter for the electrodes and 0.22 cm for the separators. Activated charcoal CH-900 (Japan) with potential +0.5 V RHE was used as the counter electrode. The measurements were conducted in 0.5 M H2SO4 at room temperature.
Cyclic voltammetry (CV). Cyclic voltammograms were recorded at 0.005 - 0.1 V s-1 potential scan rate in a potential range of 0 - 1.2 V RHE using IPC-Pro-3A potentioflat. CVs were measured in two- and three-compartment cells. The mass charge density, Q was calculated according to equation 1.
1.2
! JIdE
2 = 2 V (1)
Electrochemical impedance spectroscopy (EIS). The measurements were performed in two-electrode cell. For the EIS spectra measurements the Solartron 1286 potentioflat with Solartron 1255 frequency analyzer were used. The EIS spectra were recorded from 100 kHz to 0.1 Hz with 10 mV amplitude.
Rotating disc electrode (RDE) method. The polarization curves for O2 reduction were recorded at potentiodynamic regime, using RDE at 150 - 660 rpm speeds with potential scan rate 1 mV s-1. The measurements were performed in three-compartment cell in 0.5 M H2SO4 and 1 M KOH at 60 oC.
Rotating ring-disc electrode (RRDE) method. O2 reduction reaction selectivity was determined by RRDE method using Pine Inflrument bipotentioflat. The disc electrode is made of glassy carbon, the ring electrode is platinized Pt. The diameter of glassy carbon disc is 0.5 cm. Internal and external diameters of Pt-ring are 0.65 and 0.75 cm correspondingly. The area Sgeom. of the disc electrode is ~0.2 cm2; Sgeom. of the ring was ~0.11 cm2. The roughness of Pt surface Sreal/Sgeom is 80 - 90. The measurements were performed in three-compartment cell in
6
4
2
0
0
0.05
0.1
0.2
0.25
v)
0.5 M H2SO4 and 1 M KOH at 25 oC. The experimental value of the collection efficiency N was determined according to the methodology described in [40] using K3[Fe(CN)6] and was equal to 0.25. Calculations of H2O2 yield were conducted according to equation 2. The ratio k1/k2 where k1 is the rate conflant of 4-electron O2 reduction to H2O (OH-) and k2 is the rate conflant of 2-electron O2 reduction to H2O2 ( ) was determined graphically as the intercept on the y-axis of the linear graph - versus -w-1/2- The values of limiting diffusion current for 4-electron transfer were calculated using Levich equation (equation 3). The following parameters [41] were taken for 0.5 M H2SO4 at 25 oC: oxygen solubility = 1.1 mol m-3, molecular oxygen diffusion coefficient =1.4*10-9 m2 s-1; electrolyte solution viscosity v=1.0*10-6 m2 s-1. The calculated values of limiting diffusion current in 0.5 M H2SO4 at 150 is 1.3 mA cm-2. For 1 M KOH at 25 oC: =0.89 mol m"3; =1.6*10-9 m2 s-1; v= 0.952*10-6 m2 s-1 [42]. In 1 M KOH, at 660 rpm equals 2.4 mA cm-2. All the; potential values are related to reversible hydrogen electrode (RHE).
H,O, = 2Ir/ N x100% 2 2 Id + IR/ N
(2)
Ilim=0-62 X n x F X DO/3 XV-1'6 X ^ (3)
■D—-— = 1 + 2
Ir k2
k3 + (k3 + k4)| 1 + 2
0.62 x DH
(4)
Acknowledgements
This fludy was financially supported by Russian Academy of Science and Russian Foundation for Basic Research (RFBR), projects 13-03-00317, 11-03-12115, 13-03-00993, 14-29-04011, 14-03-31964. The authors would like to thank O.M. Zhigalina and V.G. Zhigalina.
References
1. Konkin A. A. Carbon fibers and other heat-resiflant fibrous materials. -Khimiya, Moscow, 1974. [in Russian]
2. Morita K., Murata Y., Ishitani A., Murayama K., Ono T., Nakajima A. Characterization of commercially available PAN (polyacrylonitrile)-based carbon fibers. // Pure Appl. Chem. 58, 1986, 455-468.
3. Morgan P. Carbon fibers and their composites. -Taylor & Francis, Boca Raton, London, New York, Singapore, 2005.
4. Henrici-Olive G., Olive S. The chemiflry of carbon fiber formation from polyacrylonitrile. // Adv. Polym. Sci., 51,1983, 1-60.
5. Shindo A. Studies on Graphite Fiber. // J. Ceram. Assoc. Japan, 69, 1961, C195.
6. Gupta S., Tiyk D., Bae I., Aldred W., Yeager E., Heat-treated polyacrylonitrile-based catalyfls for oxygen electroreduction. // J. Appl. Electrochem., 19, 1989, 19-27.
7. Jaouen F., Herrannz J., Lefevre M., Dodelet J.-P., Kramm U. I., Herrmann I., Bogdanoff P., Maruyama J., Naqaoka T., Garsuch A., Dahn J. R., Olson T., Pylypenko S., Atanassov P., Uflinov E. A. Cross-laboratory experimental fludy of non-noble-metal electrocatalyfls for the oxygen reduction reaction. // ACS Appl. Mater. Interfaces, 1, 2009,1 623-1639.
8. Martins A. M. C., Dodelet J.-P., Guay D., Ladouceur M., Tourillon G. Origin of the electrocatalytic properties for oxygen reduction of some heat-treated polyacrylonitrile and phthalocyanine cobalt conpounds adsorbed on carbon black as
probed by electrochemiflry and X-ray absorption spectroscopy. // J. Phys. Chem., 96, 1992, 10898-10905.
9. Ohms D., Herzog S., Franke R., Neumann V, Wiesener K., Gamburcev S., Kaisheva A., Iliev I. Influence of metal ions on the electrocatalytic oxygen reduction on carbon materials prepared from pyrolized polyacrylonitrile. // J. Power Sources, 38, 1992, 327-334.
10. Wu J., Higgins D., Chen Z. Electrospun iron/ polyacrylonitrile derived nanofibrous catalyfls for oxygen reduction reaction. // ECS Trans., 50, 2013, 1807-1814.
11. Wu J., Park H. W., Yu A., Higgins D., Chen Z. Facile synthesis and evaluation of nanofibrous iron-carbon based non-precious oxygen reduction reaction catalyfls for Li-O2 battery applications. // J. Phys. Chem. C, 116, 2012, 9427-9432.
12. Jeong B., Uhm S., Lee L. Iron-cobalt modified electrospun carbon nanofibers as oxygen reduction catalyfls in alkaline fuel cells. // ECS Trans., 33, 2010, 1757-1767.
13. Nakagawa N., Abdelkareem M. A., Takino D., Ishikawa T., Tsujiguchi T. PAN based carbon nanofibers as an active ORR catalyfl for DMFC. // ECS Trans., 41, 2011, 2219-2229.
14. Ye S., Vijh A. K., Dao L. H. Carbonized aerogel platinum composites as fuel cell electrocatalyfls: some electrochemical and surface effects. // J. New Mat. Electr. Sys., 1, 1998, 17-24.
15. Ye S., Vijh A. K., Dao L. H. New fuel cell electrocatalyfl based on highly porous carbonized polyacrylonitrile foam. The nature of platinum-support interaction. // J. Electrochem. Soc., 144, 1997, 90-95.
16. Ye S., Vijh A. K., Dao L. H. A new fuel cell electrocatalyfl based on highly porous carbonized polyacrylonitrile foam with very low platinum loading. // J. Electrochem. Soc., 143, 1996, L7-L9.
17. Ponomarev I. I. Ponomarev Iv. I., Filatov I. Yu., Filatov Yu. N., Razorenov D. Yu., Volkova Yu. A., Zhigalina O. M., Zhigalina V. G., Grebenev, V. V, Kiselev N. A. Design of electrodes based on a carbon nanofiber nonwowen material for the membrane electrode assembly of a polybenzimidazole-membrane fuel cell. // Dokl. Phys. Chem., 448, 2013, 23-27.
18. Kondratenko M. S., Ponomarev I. I., Gallyamov M. O., Razorenov D. Y., Volkova Y. A., Kharitonova E. P., Khokhlov A. R. Novel composite Zr/PBI-O-PhT membranes for HT-PEFC applications. // Beilflein J. Nanotechnol., 4, 2013, 481-492.
19. Davydova E. S., Rychagov A. Yu., Ponomarev I. I., Ponomarev I. I. Electrocatalytic and capacitive properties of pyrolized polyacrylonitrile nanofibers synthesized by electrospinning. // Russ. J. Electrochem., 49, 2013, 1010-1011.
20. Gouerec P., Miousse D., Tran-Van F., Dao L. H. Characterization of pyrolized polyacrylonitrile aerogel thin films used in double-layer supercapacitors. // J. New Mat. Electr. Sys., 2, 1999, 221-226.
21. Wang K.-P., Teng H. The performance of electric double layer capacitors using particulated porous carbons derived from PAN fibers and phenol-formaldehyde resin. // Carbon, 44, 2006, 3218-3225.
22. Ania C. O., Khomenko V., Raymundo-Pinero E., Parra J. B., Beguin F., The large electrochemical capacitance of microporous doped carbon obtained by using a zeolite template. // Adv. Funct. Mater., 17, 2007, 1828-1836.
23. Kim C. Electrochemical characterization of electrospun activated carbon nanofibers as an electrode in supercapacitors. // J. Power Sources, 142, 2005, 382-388.
24. Gavrilov N., Pafli I. A., Mitric M., Travas-Sejdic J., Citic-Marjanovic G., Mentus S. N. Electrocatalyfl of oxygen
reduction reaction on polyaniline-derived nitrogen-doped carbon nanoparticle surfaces in alkaline media. // J. Power Sources, 220, 2012, 306-316.
25. Gavrilov N., Pafli I. A., Vujkovic M., Travas-Sejdic J., Citic-Marjanovic G., Mentus S. N. High-performance charge florage by N-containing nanoflructured carbon derived from polyaniline. // Carbon, 50, 2012, 3915-2927.
26. Wang T. Electrospun carbon nanofibers for electrochemical capacitor electrodes. Ph. D. Thesis, Georgia Inflitute of Technology, Atlanta, USA, 2007.
27. Davydova E. S. O2 reduction electrocatalysis on Co and Fe porphyrin and polyacrylonitrile pyropolymers: active site flructure. Academic Thesis, Frumkin Inflitute of Physical Chemiflry and Electrochemifliy of Russian Academy of Science, Moscow, Russia, 2014. [in Russian]
28. Andreas H. A., Conway B. E. Examination of the double-layer capacitance of an high specific-area C-cloth electrode as titrated from acidic to alkaline pHs. // Electrochim. Acta, 51, 2006, 6510-6520.
29. Eliad L., Salitra G., Pollak E., Soffer A., Aurbach D. Enhanced anion electroadsorption into carbon molecular sieve electrodes in acidic media. // Langmuir 21, 2005,10615-10623.
30. Kawaguchi M., Itoh A., Yagi S., Oda H. Preparation and characterization of carbonaceous materials containing nitrogen as electrochemical capacitor. // J. Power Sources, 172, 2007, 481-486.
31. Gouerec P., Talbi H., Miousse D., Tran-Van F., Dao L. H., Lee K. H. Preparation and modification of polyacrylonitrile microcellular foam films for use as electrode supercapacitors. // J. Electrochem. Soc., 148, 2001, A94-A101.
32. Mendoza-Sanchez B., Rasche B., Nicolosi V., Grant, P. S. Scalable ultra-thin and high power density graphene electrochemical capacitor electrodes manufactured by aqueous exfoliation and spray deposition. // Carbon, 52, 2013, 337-346.
33. Pandolfo A.G., Hollenkamp A.F. Carbon properties and their role in supercapacitors. // J. Power Sources, 2006, 157, 11-27.
34. Rychagov A. Yu., Volfkovich Yu. M., Interaction of activated carbon electrodes with sulphuric acid solutions. // Russ. J. Electrochem., 2007, 43, 1273-1278.
35. Tarasevich M. R., Sadkowski A., Yeager E. Oxygen electrochemiflry. In: Comprehensive treatise of electrochemiflry. Conway B. E., Bockris J. O'M., Yeager E., Khan S. U. M., White R. E. Eds.-Plenum Press, New York, 1983, Vol. 7.
36. Soto G., Samano E. C., Machorro R., Caflillion F. F., Farias M. H., Cota-Araiza L. XPS, AES and EELS fludy of the bonding character in CNx films. // Superf. Vacio, 15, 2002, 34-39.
37. Biddinger E. J., von Deak D., Singh D., Marsh H., Tan B., Knapke D. S., Ozkan U. S. Examination of catalyfl loading effects on the selectivity of CNx and Pt/VC ORR catalyfls using RRDE. // J. Electrochem. Soc., 158, 2011, B402-B409.
38. Davydova E. S. Effect of thickness of nonplatinum cobalt-tetramethoxyphenylporphyrin based catalyfl layer on the selectivity of O2 electrochemical reduction. // Russ. J. Electrochem., 49, 2013, 733-740.
39. Rychagov A. Yu, Volfkovich Yu. M. Interaction of activated carbon electrodes with sulphuric acid solutions. // Russ. J. Electrochem., 2009, 45, 304-310.
40. Schmidt T. J., Gafleiger H. A. Rotating thin-film method for supported catalyfls. In: Handbook of fuel cells. Vielflich W., Lamm A., Gafleiger H. A., Eds. -John Wiley & Sons, Wefl Sussex, 2003, Vol. 2.
41. Borja-Arco E., Jimenez Sandoval O., Escalante-Garcia J. Sanoval-Gonzalez A. Microwave assifled synthesis of ruthenium electrocatalyflsfor oxygen reduction reaction in the presence and absence of aqueous methanol. Int. J. Hydrog. Energy, 36, 2011, 103-110.
42. Davis R. E., Horvath G. L., Tobias C. W. The solubility and diffusion coefficient of oxygen in potassium hydroxide solutions. // Electrochim. Acta, 12, 1967, 287-297.
