Efficient Edge Plane Exposure on Graphitic Carbon Fiber for

Subscriber access provided by University of Florida | Smathers Libraries

Article

Efficient Edge Plane Exposure on Graphitic Carbon Fiber for Enhanced Flow-Battery Reactions Jun Maruyama, Shohei Maruyama, Tomoko Fukuhara, and Kei Hanafusa J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b07961 • Publication Date (Web): 12 Oct 2017 Downloaded from http://pubs.acs.org on October 21, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Efficient Edge Plane Exposure on Graphitic Carbon Fiber for Enhanced Flow-Battery Reactions Jun Maruyama,a,* Shohei Maruyama,a Tomoko Fukuhara,a and Kei Hanafusab a

Research Division of Environmental Technology, Osaka Research Institute of Industrial Science

and Technology, 1-6-50, Morinomiya, Joto-ku, Osaka 536-8553, Japan b

Power Systems R&D Center, Sumitomo Electric Industries, 1-1-3, Shimaya, Konohana-ku,

Osaka 554-0024, Japan

ABSTRACT: Surface treatments are often applied to carbon materials to impart specific functions to the surface. Surface oxidation is a typical treatment to form oxygen-containing surface functional groups on carbon fiber electrodes of redox flow batteries, which has attracted much attention as a large-scale electric energy storage system, in order to enhance the performance. At present, however, little attention has been paid to the effect of the edge plane exposure. In this study, fine etching of the graphitized carbon fiber surface was attained by coating the surface with a metal-containing carbonaceous thin film and thermal oxidation. The etching was caused by the catalysis of the metal species; the mechanism and the effect of the carbonaceous film were demonstrated by in-situ X-ray absorption fine structure measurements. The finely-etched surface possessed substantially enriched edge planes and an enhanced activity

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60

Page 2 of 29

for the positive and negative electrode reactions of the vanadium redox flow battery. The flow cell test with the carbon fiber electrodes after the tuned etching showed a significant decrease in the overpotential and increase in the efficiency as well as stable cycling performance.

INTRODUCTION Carbon electrodes have been used in a variety of applications, such as electric arc furnace steel production, zinc-carbon cells, Li-ion batteries, and electric double-layer capacitors. One of the applications that has recently attracted much attention is redox flow batteries (RFBs), in which carbon fiber in the form of felt or paper is conventionally used. The RFB is a promising electric energy storage system due to its suitability for large-scale energy storage and capability to withstand fluctuating power supplies.1–4 The RFB in the most advanced stage of research and development is the vanadium redox flow battery (VRFB) based on the following redox reactions.5–7 VO2+ + 2H+ + e– V2+

V3+ + e–

VO2+ + H2O

(positive electrode) (negative electrode)

The efficiency of the VRFB is largely influenced by the redox reaction rates8,9 and their enhancement has been the major issue in the VRFB technology. An efficient way to improve the reactivity is using electrodes with oxygen-containing functional groups on the carbon fiber surface, which have been attained by thermal oxidation,10–12 corona discharge,13 oxidation with


2

Page 3 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60


acids,14,15 and electrochemical oxidation.16,17 These treatments are often accompanied by surface roughening; however, the enhancement has been mostly explained by the involvement of the surface functional groups in the redox reactions, but only limited attention has been paid to the effect of the roughening. Carbon materials are generally considered as an assembly of graphite nanocrystallite, whose size and orientation are dependent on the degree of graphitization. The graphite surface consists of basal and edge planes and the roughening changes the amount of exposure of these planes. A few fundamental studies have reported the redox reactions using planar electrodes and favorable behavior on the edge plane,18–20 but homogeneous, fine, and effective treatment of practical electrodes composed of the carbon fiber focusing on the roughening and the efficient generation of the edge plane has not yet been performed. It has long been known that the surface etching of graphite is enabled by the catalytic effect of various metals loaded on the surface for carbon gasification in H2, O2, H2O, and CO2 gases. Most studies have been focused on the gasification reaction rate and the surface morphology; channels and pits were generated on the graphite surface.21–43 Quite recently, similar technique was used to form carbon fiber felt with porous surface.44 Although superior performance of the VRFB was demonstrated using the felt as the electrodes, the effect of the roughness and the edge planes are still not clear because the effects of the pores and the surface functional groups attached after the pore formation dominated. Repetitive treatments required for the higher performance than that obtained by simple thermal oxidation in air would be the point to be improved. In this study, we attempted fine etching by loading a series of typical transition metals


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60

Page 4 of 29

contained in a carbonaceous thin film on the graphitic carbon fiber surface, which was prepared through the sublimation, deposition, and pyrolysis of the corresponding metal phthalocyanine during a single heat treatment step in an Ar atmosphere,45–47 followed by the thermal oxidation in air. This approach was assumed to be advantageous and a new feasible method to form a highlydispersed fine metal oxide catalyst for homogeneous etching and the efficient edge plane exposure48. As well as conventional analyses, such as field emission scanning electron microscopy, Raman spectroscopy, and X-ray photoelectron spectroscopy, in-situ measurement of the X-ray absorption fine structure was performed to clarify the fine-etching mechanism. The electrochemical measurements including cyclic voltammetry and full cell tests demonstrated the significant enhancement of the performance and the effectiveness of the fine etching.

EXPERIMENTAL SECTION Materials. Graphitic carbon paper (TGP-H-090, Toray, abbreviated as TGP), manganese phthalocyanine (MnPc, Sigma-Aldrich), iron phthalocyanine (FePc, Tokyo Chemical Industry), cobalt phthalocyanine (CoPc, Sigma-Aldrich), nickel phthalocyanine (NiPc, Acros Organics), copper phthalocyanine (CuPc, Sigma-Aldrich), and Co(CH3CO2)2•4H2O (Nacalai Tesque), ethanol (99.5%, Nacalai Tesque) were used as received. TGP is often treated with polytetrafuluoroethylene (PTFE) to impart hydrophobicity in fuel cell applications, but TGP without PTFE was used in this study. High-purity water was obtained by circulating ionexchanged water through an Easypure water-purification system (Barnstead, D7403). Sulfuric acid (6 M, Kishida Chemical Co., Ltd.) was diluted with the high-purity water to prepare a 2 M H2SO4 solution. Oxovanadium sulfate hydrate, VOSO4·nH2O, was purchased from Sigma-


4

Page 5 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60


Aldrich (purity, > 99.99%) and Nacalai Tesque, which was dissolved in the 2 M H2SO4 to prepare VOSO4(1 M)–H2SO4(2 M). The number of waters of hydration, n, was provided by the manufacturer or determined in advance by thermogravimetry and a differential thermal analysis using an SSC/5200 thermal analyzer (Seiko Instruments). Fine etching of graphitic carbon fiber surface. Eight pieces of TGP (1 cm2) and 10 mg of MPc (M = Mn, Fe, Co, Ni, Cu) were placed in a crucible (15 cm3) with a cap and heat-treated at 700 ºC for 1 h after raising the temperature at 5 ºC min–1 in an Ar atmosphere to load the carbonaceous thin film containing M. The sample was labelled TGP-CMPc. The amount of Co loaded on the carbon fiber in TGP was 0.021 wt %, which was determined by inductively coupled plasma optical emission spectrophotometry (ICP-OES) using an iCAP6300 system (Thermo Fisher Scientific) after the combustion of the carbon matrix and the dissolution of the residue in boiling 0.5 M H2SO4 prepared from sulfuric acid (98%, Tama Chemicals, ultrapure analytical reagent) and high-purity water. The Co(CH3CO2)2-deposited TGP (TGP-Co) was also prepared for comparison by soaking TGP in the aqueous Co(CH3CO2)2 solution and drying it on a hot plate at 80 ºC. Its concentration was adjusted to load the same amount of Co on the TGP surface according to the manufacturer’s datasheet values of thickness (0.28 mm) and void fraction of TGP (0.78). The heat treatment in air was performed using TGP-CMPc and TGP-Co at T ºC (T = 500, 550, 600) for 1 h followed by a treatment with 6 M HCl (Nacalai Tesque) to remove any soluble metallic species from the carbon fiber surface. The samples obtained from TGP-CMPc and TGPCo were labeled TGP-CMPc-TAir and TGP-Co-TAir, respectively. For comparison, the heat treatment of TGP without M loading was also performed in air at 550 ºC for 1 h (TGP-550Air). The treatment conditions and the obtained samples are summarized in Table 1.


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60

Page 6 of 29

Table 1. Conditions of treatments for TGP Metal source Thermal oxidation temperature [ºC] TGP – – – TGP-CMPc CMPca TGP-550Air – 550 TGP-CMPc-550Air CMPca 550 TGP-Co-550Air Co(CH3CO2)2 550 TGP-CCoPc-500Air CCoPca 500 TGP-CCoPc-600Air CCoPca 600 a Carbonaceous thin film derived MPc (M = Mn, Fe, Co, Ni, Cu) Characterization of carbon fiber surface. A field-emission scanning electron microscope (FESEM, JSM-6700F, JEOL) was used to observe the surface structure. The Raman spectra were obtained in the backscattering mode by an NRS-3100 spectrometer (JASCO) using an Ar+-ion laser (532.05 nm, 0.3 mW) as the excitation source. The laser beam was focused on the surface of the carbon thin film, producing a spot (analysis area) of approximately 4 mm in diameter. Three different places within the sample were analyzed. The X-ray photoelectron spectroscopy (XPS) was performed using an AXIS ULTRA DLD system (Kratos Analytical) with Al Ka radiation (1486.6 eV). In-situ X-ray absorption fine structure measurement during air oxidation. The local structure around Co during the heat treatment in air was investigated by the in-situ measurement of the X-ray absorption fine structure (XAFS) at the Co-K edge. The X-ray absorption near-edge structure (XANES) was obtained using synchrotron radiation at the beam line of BL14B2 of SPring-8 at the Japan Synchrotron Radiation Research Institute. The measurements for TGPCCoPc and TGP-Co were performed in the fluorescence mode in air at room temperature, 100, 200, 300, 400, 450, 500, 550, and 600 ºC, consecutively. The spectra were acquired for 1 h at each temperature. The increasing temperature rate between the measurements was 5 ºC min–1. The XAFS spectra of the Co foil, CoPc, CoO (Koujundo Chemical Laboratory, 99.9%), Co3O4


6

Page 7 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60


(Alfa Aesar, 99.9985%) as the references were measured in the transmission mode. The gas evolution during the thermal oxidation was separately examined by the gas chromatograph – mass spectrometry (GCMS-QP2010Plus, Shimadzu) with the sample heated at 10 ºC min–1 in a thermogravimetric analyzer (TG, Thermo plus EVO, Rigaku) under the He/O2 gas flow of 0.3 dm3 min–1. Electrochemical properties. The cyclic voltammetry was performed using a three-electrode glass cell and an electrochemical analyzer, 100B/W (BAS). An Au wire as a lead was connected to the upper side of the 1-cm2 sample to form the working electrode, which was immersed in ethanol and then rinsed with high-purity water to fully wet the electrode and to minimize the influence of wetting.10,49 The counter electrode was carbon cloth (ElectroChem). The reference electrode was Ag/AgCl/NaCl(3 M) (0.212 V vs. standard hydrogen electrode). The electrolytes were H2SO4(2 M) and VOSO4(1 M)–H2SO4(2 M). The measurements were performed under an Ar atmosphere at 25 ºC. A flow cell test was performed using 3 layers of 3 cm2 of TGP-CCoPc550Air as the negative and positive electrodes, and Nafion 212 as the separator, incorporated into a flow cell similar to that used in a previous study.6 The number of the layers was also chosen according to the results in the study.6 TGP-CCoPc-550Air was immersed in ethanol and rinsed with high-purity water before the incorporation. The anolyte (40 cm3) and catholyte (20 cm3) were prepared by the electrolysis (charging) of 1 M VOSO4 + 2 M H2SO4 until the full conversion of VO2+ to VO2+ and V2+. After the electrolysis, half of the anolyte was removed and the pre-discharge was carried out at 25 mA cm–2, followed by measurement of the chargedischarge curve. The flow rate was 3 cm3 min–1. The current density was 50 mA cm–2. The flow cell using TGP was similarly tested for comparison.


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60

Page 8 of 29

RESULTS AND DISCUSSION Finely-etched carbon fiber surface. The FESEM images of TGP, TGP-CCoPc, TGP550Air, and etched TGP are shown in Figure 1. Almost no appreciable change was observed by the formation of the CoPc-derived thin film on the carbon fiber (Figure 1(b)). The air oxidation of TGP-CCoPc at 550 ºC for 1h generated a roughened surface with many bumps and dents of around 100 nm over the whole surface of the carbon fiber (TGP-CCoPc-550Air, Figure 1(c)). The surface of TGP-550Air was nearly the same as that of TGP, but the bumps and dents were absent, indicating the necessity of the loading of the Co-containing carbonaceous thin film to generate the roughened surface. The importance of the coating of the carbonaceous thin film for

Figure 1. FESEM images of TGP, TGP-CCoPc, TGP-550Air, and etched TGP.


8

Page 9 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60


the fine surface structure was also suggested by the FESEM images of TGP-Co-550Air (Figure 1(e), Figure S1), which show a larger structure and the partial roughening due to the nonhomogeneous deposition of Co(CH3CO2)2 on the carbon fiber surface. The surface etching also proceeded on the binder contained in TGP (Figure S2(a)). The original surface structure was lost, and bumps surrounded by linear edges and holes different from those observed in the original surface were generated (Figure S2(b), (c)). Because the process was more complicated on the binder surface than on the fiber surface due to the presence of the original complex surface structure of the binder, the characterization was focused on the fiber surface to clearly show the effect of the surface etching. The surface structure was dependent on the thermal oxidation temperature. Limited roughening of the carbon fiber surface was observed after the thermal oxidation at 500 ºC, whereas the surface after the air oxidation at 600 ºC was severely roughened with large pores, losing the groove in the direction of the fiber axis originally present on TGP (Figures 1(f) and (g)). The temperature control in the thermal oxidation step is thus critical to avoid the insufficient or excessive roughening, especially the latter of which would lead to lower strength of the electrode. The surface structure was also dependent on the center metal of the phthalocyanine used for the carbonaceous thin film coating on the TGP surface (Figure S3). Exposure of edge plane by surface etching. The etched surface structure was further examined by Raman spectroscopy. Figure 2 shows the Raman spectra of TGP, TGP-CCoPc, TGP-CCoPc-550Air, and TGP-550Air. The peak ascribed to the ideal graphitic lattice (G peak) at 1580–1600 cm−1 dominated the spectrum of the TGP with a small peak ascribed to the graphene layer edges as a disordered graphitic lattice (D peak) at 1345–1355 cm−1. The spectrum


9


ID/IG TGP-CCoPc-600Air

0.822

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60

Page 10 of 29

0.463

TGP-CCoPc-500Air

0.395

TGP-550Air

0.558

TGP-CCoPc-550Air

TGP

0.255 G

0.333

Am D2

1800

1600

1400

D

TGP-CCoPc P

1200

1000

Raman shift/cm‒1

Figure 2. Raman spectra of TGP, TGP-CCoPc, TGP-550Air, and etched TGP. Deconvoluted components (D2, G, Am, P, D) and fitting result are shown for TGP-CCoPc as a typical example. Values of ID/IG for the corresponding samples are also shown. of TGP-CCoPc consisted of 3 other components as well as the G and D peaks; i.e., the peak attributed to the surface graphene layers as a disordered graphitic lattice (D2 peak) around 1620 cm−1; the peak attributed to amorphous carbon (peak Am) at 1490–1525 cm−1; and the peak attributed to the sp3-bonded carbon atoms (peak P) at 1150–1190 cm−1.50,51 The typical curve fitting result is shown for TGP-CCoPc. The presence of peaks Am and P indicates that the CoPc-derived carbonaceous thin film was amorphous. Peak Am decreased and peak P disappeared after the heat treatment in air at 550 ºC, showing the removal of the carbonaceous film from the surface. All the Raman spectra, except for TGP-CCoPc, were deconvoluted into 4 components and the ratios of the intensities of the D


10

Page 11 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60


peak to the G peak (ID/IG) were determined, which are generally recognized to be related to the concentration of the defects and the extent of the structural disorder.50 The high graphitization degree of TGP was reflected by the low ID/IG value. The significant increase in the D band intensity without an increase in peaks Am and P was observed for TGPCCoPc-550Air compared to TGP, which was due to the exposure of the edge plane on the carbon fiber surface.52 Because the limited increase was observed for TGP-Co-550Air, the result showed the effect of the coating by the Co-containing carbonaceous thin film. The ID/IG increase from TGP to TGP-CCoPc-550Air was more than twice, indicating the high efficiency of this technique. The ID/IG value was dependent on the thermal oxidation temperature and on the center metal of the phthalocyanine used for the carbonaceous thin film coating on the TGP surface (Figures 2 and S4), which was in agreement with the FESEM images. Hereafter, we focused our study on the fine etching with the Co species based on the previous favorable result using air,23 which is the most practical atmosphere, and the best redox activity among the metals examined in this study without the excessive etching, as described in detail in the following sections. Change in surface species during fine etching. Figure 3 shows the XPS spectra of N 1s, Co 2p, and O 1s in TGP, TGP-CCoPc, and TGP-CCoPc-550Air. The Co 2p spectrum in the sample before the acid treatment for TGP-CCoPc-550Air (TGP-CCoPc-550Air_woAW) is also shown. The detection of N and Co in TGP-CCoPc indicated the generation of the CoPc-derived carbonaceous thin film on the carbon fiber surface. The surface nitrogen almost diminished after the thermal oxidation due to the removal of the carbonaceous film from the surface, which was also observed in the Raman spectrum. The broad Co 2p satellite peak in TGP-CCoPc550Air_woAW indicated the generation of a mixture of Co oxides,53 which was not detected in TGP-CCoPc-550Air due to their removal by the acid treatment. The limited increase in the


11


(b) Co 2p

(a) N 1s

Intensity

Intensity

TGP-CCoPc

TGP-CCoPc-550Air 406 404 402 400 398 396 394

Binding energy/eV

TGP-CCoPc TGP-CCoPc-550Air-woAW TGP-CCoPc-550Air 810

800

790

780

770

Binding energy/eV

(c) O1s TGP-550Air

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60

Page 12 of 29

TGP-CCoPc-550Air TGP

538 536 534 532 530 528 526

Binding energy/eV

Figure 3. XPS spectra of (a) N 1s, (b) Co 2p, and (c) O 1s in TGP, TGP-CCoPc, TGP-550Air, and TGP-CCoPc-550Air. The Co 2p spectrum in the sample before the acid treatment for TGP-CCoPc-550Air (TGP-CCoPc-550Air_woAW) is also shown in (b). surface oxygen concentration only by the thermal oxidation was in agreement with the results in a previous study (Table 2).10 The surface oxygen increase on TGP-CCoPc-550Air (0.022 in the O/C ratio) could also regarded as minor comparing with those on the surface-oxidized carbon fiber reported in previous studies, in which the O/C ratio was in the range from 0.167 to 0.237.11 ,15,44,54 Fine-etching mechanism. The in-situ XAFS measurements were performed to clarify the fine-etching mechanism. The XANES spectra at the Co-K edge measured from room temperature to 600 ºC in air for TGP-CCoPc are shown in Figure 4(a). The experimentally

Table 2. Conditions of treatments for TGP C N O TGP 99.54 – 0.46 TGP-CCoPc 87.24 8.07 3.57 a 94.49 0.10 4.51 TGP-CCoPc-550Air_woAW TGP-CCoPc-550Air 97.44 0.35 2.20 TGP-550Air 99.12 – 0.88 a the sample before the acid treatment for TGP-CCoPc-550Air

Co – 1.12 0.90 – –


12

Page 13 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60


Figure 4. XANES spectra at the Co-K edge measured in situ during the thermal oxidation of (a) TGP-CCoPc and (b) TGP-Co. The simulated XANES curve obtained by the weighted addition of those for CoPc, Co foil, CoO, Co3O4, Co(CH3CO2)2, and that calculated by using FEFF8.2 for the five-atom model consisting of a Co atom surrounded by four nitrogen atoms in a square-planar coordination (Co–N4 model) are also shown. The XANES curves for CoPc and the Co–N4 model were combined in a 7:3 ratio and labeled as the CoPc + Co–N4 model. The relationships between the composition and the temperature are shown below the XANES spectra. obtained XANES spectrum at room temperature before the thermal oxidation well fit the curve obtained by the weighted addition of the spectrum for Co3O4, CoO, Co foil, CoPc, and that for the 5-atom model consisting of Co surrounded by 4 nitrogen atoms in a square-planar coordination (Co–N4 model). The fitting components of CoPc and the Co–N4 model suggested the presence of the intermediate structure between those similar to CoPc and the Co–N4 site with disordered surroundings of carbon matrix transformed from the macrocyclic ligand during the pyrolysis. The ratio of the CoPc and the Co–N4 model components was 7:3. The Co foil


13


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60

Page 14 of 29

component was attributed to Co metal aggregate nanoparticles formed by the decomposition of CoPc.46 These components for the CoPc + Co–N4 model and Co foil decreased from 300 ºC, and Co3O4 was generated. Above 500 ºC, the Co3O4 component decreased and that of CoO increased. Although these changes was limited, the tendency was demonstrated in the enlarged views of the XANES spectra (Figure S5). According to the FESEM images and the Raman spectrum, the fine etching proceeded above 500 ºC and the degree of etching became pronounced as the temperature was raised; thus, the XAFS results indicated that the catalyst particle for the etching consisted of CoO and Co3O4, and the mechanism of the fine etching was the following: 2 Co3O4 + C → 6CoO + CO2 6CoO + O2 → 2 Co3O4

Carbon monoxide could be another possible evolved gas; however, the CO2 generation was thermodynamically favorable below 700 ºC and the GCMS analysis indeed indicated the CO2 evolution (Figure S6). The XANES spectrum for TGP-Co at room temperature fit the curve obtained by the weighted addition of the spectrum for Co3O4, CoO, and Co(CH3CO2)2 (Figure 4(b)). The presence of the Co oxides might be caused by the drying process at 80 ºC to prepare TGP-Co and the generation of the bonds between the Co ion and the oxygen-containing surface functional groups on TGP. The Co3O4 component started to increase as low as 100 ºC and reached about 90% at 400 ºC. The Co3O4 content was clearly higher than that for for TGP-CCoPc. Because CoO was generated by the reaction between Co3O4 and the surface carbon, it is reasonable to


14

Page 15 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60


assume that CoO was present only close to the carbon surface in the catalyst particle. The higher ratio of Co3O4 shown in the XAFS measurements thus indicated that the part of the larger part of the atoms inside the catalyst particle was apart from the carbon surface; i.e., the size of the catalyst particle was larger assuming a similar particle shape. It should also be noted that the CO2 evolution for TGP-Co was similar to that of TGP (Figure S6), which was in agreement of the FESEM image showing that the area of the etched surface was limited. The finer etching of TGP-CCoPc was explained by this result, which also implied that further finer etching would be possible using precursors leading to much finer catalyst particles. The in-situ XAFS measurements for the etching by the other metals and additional techniques, such as in-situ XRD measurements, to examine the catalyst particle size would be desired to further clarify the etching phenomena, but beyond the scope of this study. Electrochemical behavior at finely-etched carbon fiber surface without vanadium ions. The cyclic voltammograms in an acidic electrolyte without the vanadium ions for TGP, TGP550Air, TGP-CCoPc, TGP-CCoPc-550Air, and TGP etched at various temperatures are shown in Figures 5(a) and 5(b). The current in the voltammogram consisted of the electrochemical doublelayer charging current at the carbon–electrolyte interface and the Faradaic current due to the redox reactions of the surface functional groups and was dependent on the extent of the carbon surface exposure as the electrochemically active surface area and also the ratio of the basal and edge planes. The broad peaks around 0.5 V attributed to the redox reactions of the surface functional groups were only slightly observed, thus, the current was mainly due to the electrochemical double-layer charging.16 The current for TGP-CCoPc was significantly increased by the coating of the CoPc-derived carbonaceous thin film. According to the report by Yeager et al. that the specific capacitances of


15


the basal plane and the edge plane were 16 and 50 to 70 µF cm–2, respectively,55 the higher current was explained by the greater exposure of the edge plane in addition to the surface roughness due to its amorphous structure shown in the Raman spectrum, which was not detected 0.15

0.60 TGP TGP-550Air TGP-CCoPc TGP-CCoPc-550Air

Current/mA cm‒2

0.40

TGP-CCoPc-500Air TGP-CCoPc-550Air TGP-CCoPc-600Air

(b) 0.10

Current/mA cm‒2

(a)

0.20 0.00

0.05 0.00 -0.05

-0.20

-0.10

-0.40 -0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

-0.2

1.4

0.0

TGP TGP-CCoPc TGP-CCoPc-550Air TGP-550Air

Current/mA cm‒2

Current/mA cm‒2

0.6

0.8

1.0

1.2

1.4

(d)

10

0

-10


20

10

0

-10

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

0.4

Potential/V vs. Ag/AgCl/NaCl(3 M)

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

Potential/V vs. Ag/AgCl/NaCl(3 M) 150

150

(e)

100

Current/mA cm‒2

100

0.4

30

(c) 20

0.2


Potential/V vs. Ag/AgCl/NaCl(3 M) 30

Current/mA cm‒2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60

Page 16 of 29

50 0 -50 -100 TGP TGP-550Air TGP-CCoPc TGP-CCoPc-550Air

-150 -200 -250 -0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1


(f)

50 0 -50 -100 -150


-200 0.0

-250 -0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0.0


Figure 5. Cyclic voltammograms in Ar-saturated 2 mol dm–3 H2SO4 at 25 ºC for (a) TGP, TGP-550Air, TGP-CCoPc, and TGP-CCoPc-550Air, (b) TGP etched at various temperatures, and in Ar-saturated 1 mol dm–3 VOSO4 + 2 mol dm–3 H2SO4 at 25 ºC for (c, d) VO2+/VO2+, (e, f) V2+/3+ redox reactions at (c, e) TGP, TGP-550Air, TGP-CCoPc, and TGP-CCoPc-550Air, (d, f) TGP etched at various temperatures. Scan rate was (a, b, e, f) 50 mV s‒1 and (c, d) 1 mV s‒1.


16

Page 17 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60


by the FESEM image.45 The surface nitrogen also increased the electrochemical double-layer charging current.56 The current above 0.9 V was attributed to the electrochemical carbon surface oxidation. The thermal oxidation of TGP-CCoPc to form TGP-CCoPc-550Air decreased the current, but it was higher than that for TGP. This higher current was due to the exposure of the edge plane on the surface of the carbon fiber, as indicated by the Raman spectrum. The exposure could be roughly estimated by the electrochemical double-layer charging current in the cyclic voltammograms. The capacitance of TGP and TGP-CCoPc-550Air were calculated from the charges to be 379 and 907 µF cm–2, respectively. If all the effective surface area was assumed to be attributed to the basal plane and to the fiber, the specific surface area of the fiber in TGP and the diameter was estimated according to the apparent density of TGP provided by the manufacturer’s datasheet values (0.44 g cm–3) and the specific capacitance of the basal plane (16 µF cm–2) to be 0.19 m2 g–1 and 10 µm, respectively. The estimated diameter of the fiber was not so deviated from that shown in the FESEM image (Figure S1) and thus suggested approximate correctness of the assumption. The change in the capacitance between TGP and TGP-CCoPc550Air was then attributable to the edge plane, and its specific surface area was estimated from the specific capacitance of the edge plane (50–70 µF cm–2) to be 0.061–0.086 m2 g–1., corresponding to 32–45% increase. Only a slight increase was observed for TGP-550Air compared to TGP, which was in agreement with the FESEM and Raman results. The behavior of TGP-CCoPc-500Air and TGPCCoPc-600Air also agreed with these results, although the current change was lower than expected from the ID/IG values. One possible reason would be the conflicting effects of the


17


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60

Page 18 of 29

etching: the exposure of the edge planes to increase the capacitance; the disconnection of the sp2 bonds in the graphene layers to decrease the electron conduction through the graphene layer and partly lose the electrochemically effective surface area by the limited interlayer electron conduction,57 but clarifying the reason would require further studies. Redox reactions of vanadium ions at finely-etched carbon fiber surface. Figures 5(c–f) show the cyclic voltammograms in the potential ranges corresponding to the positive and negative electrode reactions in an acidic electrolyte containing the vanadium ions for TGP, TGP550Air, TGP-CCoPc, TGP-CCoPc-550Air, and TGP etched at various temperatures. Regarding the VO2+/VO2+ redox reaction (Figures 5(c), 5(d)), TGP-CCoPc-550Air showed a slightly negative shift of the VO2+ oxidation peak potential and the higher VO2+ reduction peak current, indicating the activity enhancement compared to TGP, although the peak current for the VO2+ oxidation was slightly lower than that for TGP. The enhancement was also supported by electrochemical impedance spectroscopy (EIS, Supporting information, Figure S7). The limited activity enhancement in spite of the development of the edge planes on the surface might be due to the adsorption of VO2, which would increase with an increase in the oxygen-containing surface functional groups and inhibited the redox reaction.52,54 The similar XANES spectra at the V-K edge for TGP-CCoPc-550Air was observed after the immersion in an acidic solution containing VO2+ and VO2+ as that observed for VO2 on the carbonaceous electrode formed with reduced graphite oxide in our previous study,51 which indicated the presence of the vanadium species (Figure S8). The lower activity for TGP-CCoPc and TGP-550Air than TGP was also attributed to this inhibition. The activity was dependent on the thermal oxidation temperature and it reached a maximum at 550 ºC. The development of the edge planes on the surface of the carbon fiber enhanced the activity, whereas the excessive treatment led to the probable inhibition


18

Page 19 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 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 54 55 56 57 58 59 60


by VO2. The cyclic voltammograms in the potential range corresponding to the V2+/3+ redox reaction (Figures 5(e), 5(f)) demonstrated the significant current increase for TGP-CCoPc-550Air as well as the slight increase for TGP-550Air and TGP-CCoPc. The scan rate of 50 mV s–1 was chosen in order to observe the oxidation current of V2+, which was generated by the negative scan, before its loss caused by the reaction between V2+ and VO2+ to generate V3+. The shape of the voltammograms was distorted and clear V2+/3+ redox peak was absent; nevertheless, the information about the order of the activity of the electrode (TGP < TGP-CCoPc ≈ TGP-550Air

Efficient Edge Plane Exposure on Graphitic Carbon Fiber for

Recommend Documents