Heteroatom Polymer-Derived 3D High-Surface-Area and Mesoporous

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Heteroatom Polymer-derived 3D High-surface-area and Mesoporous Graphene Sheet-like Carbon for Supercapacitors Haiyang Sheng, Min Wei, Alyssa D'Aloia, and Gang Wu ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b10099 • Publication Date (Web): 18 Oct 2016 Downloaded from http://pubs.acs.org on October 19, 2016

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.

ACS Applied Materials & Interfaces 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 38

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

ACS Applied Materials & Interfaces

Heteroatom Polymer-derived 3D High-surface-area and Mesoporous Graphene Sheet-like Carbon for Supercapacitors

Haiyang Sheng, Min Wei, Alyssa D’Aloia, and Gang Wu*

Department of Chemical and Biological Engineering, University at Buffalo, The State University of New York, Buffalo, New York 14260, United States Corresponding author: [email protected] (G. Wu)

1 Environment ACS Paragon Plus


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

ABSTRACT: Current supercapacitors suffer from low energy density due to the high degree of microporosity and insufficient hydrophilicity of their carbon electrodes. Development of a supercapacitor capable of simultaneously storing as much energy as a battery, along with providing sufficient power and long cycle stability would be valued for energy storage applications and innovations. Differing from often studied reduced graphene oxides, in this work we identified an inexpensive heteroatom polymer (polyaniline-PANI) as a carbon/nitrogen precursor, and applied a controlled thermal treatment at elevated temperature to convert PANI into 3D high-surface-area graphene-sheet-like carbon materials. During the carbonization process, various transition metals including Fe, Co, and Ni were added, which play critical roles in both catalyzing the graphitization and serving as pore forming agents. Factors including post-treatments, heating temperatures, and types of metal were found crucial for achieving enhanced capacitance performance on resulting carbon materials. Using FeCl3 as precursor along with optimal heating temperature 1000oC and mixed acid treatment (HCl+HNO3), the highest Brunauer-Emmett-Teller (BET) surface area of 1645 m2g-1 was achieved on the mesopore dominant graphene-sheet, similar to carbon materials. The unique morphologies featured with high-surface areas, dominant mesopores, proper nitrogen doping, and 3D graphene-like structures correspond to remarkably enhanced electrochemical specific capacitance up to 478 Fg-1 in 1.0 M KOH at a swiping rate of 5 mVs-1. Furthermore, in a real two-electrode system of a symmetric supercapacitor, a specific capacitance of 235 Fg-1 by using Nafion binder is obtained under a current density of 1 Ag-1 by galvanostatic chargedischarge tests in 6.0 M KOH. Long-term cycle stability up to 5000 cycles by using PVDF binder in electrode was systematically evaluated as function of type of metal and current densities.

KEYWORDS: Supercapacitors, high-surface-area carbon; graphene-sheet, heteroatom polymers, carbonization


Page 2 of 38

Page 3 of 38

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


1. INTRODUCTION An increase in popularity of portable, smart electronic devices, as well as electric vehicles has induced the need for reliable, stable and sustainable energy storage devices with desired energy and power density. Supercapacitors, an electrochemical double layer capacitor, are a relatively new energy storage technology compared to traditional batteries. Unlike electrochemical pseudocapacitors that usually use metal oxides or conducting polymers via faradaic electron transfer during the redox reaction, the electrochemical double-layer capacitors require high-surface areas carbon materials capable of providing much higher capacitance. In addition, upon charging with applied voltage, an electrical double layer is formed at the electrode surfaces with atomic scale charge separation, creating a very large charge storage capacity. Due to the rapid transfer of ions at very short distances (a few nanometers), supercapacitors provide high power density, long life cycles (>10k cycles) and operate in a safe manner.1-5 However, a major limitation to this technology is a much lower energy density (typically 5-10 Wh kg-1), relative to conventional Liion batteries around ~200 Wh kg-1.6 Hence, development of supercapacitors that are capable of storing as much energy as a battery and can provide sufficient power and long cycle stability would be a revolutionary advancement in all applications of energy storage technologies. Considering a variety of materials to be used for supercapacitor electrodes, activated carbons (ACs), graphene, carbon nanotubes (CNTs),7 certain metal oxides and conducting polymers have shown their potential applications.8-9 At present, carbon based materials have attracted substantial attention to prepare supercapacitor electrodes, due to their advantages including lower cost, excellent

electrical

conductivity,

easy

processing,

higher

specific

chemical/electrochemical stability, and wide operating temperature range.7,

10

surface

area,

Therefore, high

surface area carbon materials with appropriate porosity and surface chemistry are highly desirable



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

for supercapacitors. The currently used carbon material for supercapacitors is activated carbon.1113

Generally, the specific capacitances of carbon materials exhibit the supercapacitor performance

with values in the range of 75-200 Fg-1 for aqueous electrolytes and 40-150 Fg-1 for organic electrolytes.14-16 The current issue for low capacity supercapacitors lies within the electrode and its inability to provide sufficient electrochemically accessible surface area, or inability to provide optimal morphology for mass/charge transfer. For example, as for one activated carbon (~500 m2/g), typically only about 10-20% of the “theoretical” capacitance can be achieved on the activated carbon electrode. This is due to the high degree of microporosity and insufficient hydrophilicity, which limits wetting deficiencies of electrolytes on the electrode surface. Attention has been shifted from traditional carbon electrodes to other advanced materials with increased surface areas and enhanced accessibility of ions. Nanotechnology has opened up new frontiers by offering unique enabling technologies and new materials for energy storage. In particular, graphitic carbon nanomaterials including carbon nanotubes (CNTs),17-22 carbon nanofibers,23-25 and reduced graphene oxides,6,

26-30

have played an important role in the

development of high-performance supercapacitors.31-35 Bing et al. 36 developed a two-step etching route to ultrathin carbon nanosheets for high performance supercapacitors showing the specific capacitance of 220 F g-1 at 0.5 A g-1. An et al. 37 discovered the single-walled carbon nanotubes, exhibiting the specific capacitance (180 Fg-1). However, these nanocarbon-based materials still suffer from low surface areas and insufficient charge transfer at the electrode/electrolyte interfaces. Therefore, the development of novel high-surface-area carbon with optimal morphology and nanostructures would be beneficial for enhanced supercapacitor performance. To overcome these issues, in this work we developed novel 3D high-surface area carbon materials dominant with graphene-sheet-like morphologies (not graphene itself), demonstrating


Page 4 of 38

Page 5 of 38

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


remarkable capacitance for supercapacitors. This synthetic method is facile and low cost, and is derived from an inexpensive heteroatom (e.g, polyaniline-PANI) via a high-temperature carbonization in the presence transition metal precursors (M: Fe, Co or Ni). In addition to the catalyzing function, the metal precursors served as self-sacrificing agents that were in-situ formed during carbonization, occupied their positions and then were eliminated by acid leaching, thus giving rise to the porous structures. With the highest specific surface area (1645 m2 g−1), an optimal sample (Fe-PANI-C) was derived from PANI and Fe exhibited a high capacitance of 478 F g-1 at a scan rate of 5 mV s-1 determined by using cyclic voltammetry in 1.0 M KOH. These newly prepared carbon materials all demonstrated smaller series resistance and charge transfer resistance than Black Pearl 2000 (BP 2000). Furthermore, in subsequent galvanostatic charge discharge analysis, real two-electrode supercapacitors were fabricated. Although the determined capacitance is reduced to certain degree when shifting from glassy carbon electrode to real supercapacitor electrode with an increased loading 3-4 mgcm-2 and using PVDF as a binder, M-PANI-Cs presented ideal cycle stability with the capacitance retention. 2. EXPERIMENTAL SECTION 2.1 Materials synthesis Within our typical approach to chemically polymerize aniline for in-situ preparation of polyaniline (PANI),26, 38-42 2.0 ml of aniline was dispersed in 100 ml 2.0 M hydrochloride acid (HCl), followed by the addition of 15 g metal precursors (FeCl3, CoCl2·6H2O or NiCl2·6H2O). Then 5.0 g ammonium peroxydisulfate (APS) was dissolved in another 100 ml 2.0 M HCl. The well dissolved APS solution was added into the pre-dispersed suspension of aniline and metal precursor to initialize the polymerization. Black Pearl 2000 (BP2000) was used as the carbon support. 0.4 g BP2000 was ultrasonically dispersed in 20 ml Milli-Q water with 2 ml isopropyl



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

alcohol (IPA). The suspension was sonicated for 2 hours before it was added to the PANI suspension. Constant stirring of the suspension for 24 hours is crucial for allowing fully polymerization of PANI, which promotes uniform mixing of metal precursors and polyaniline, and both polyaniline chain and metal precursors covering the carbon support. Next, the suspension containing carbon, polymer, and transition metal(s) was moved to a beaker and heated to slowly evaporate the water. The resulting solid mixture was then heat-treated at a high temperature ranging from 850°C to 1000 °C in a tube furnace in inert nitrogen atmosphere for 1 h. The heattreated sample was leached in various acids including individual 0.5 M H2SO4, 30% HNO3 or mixed acid (30% HNO3 and 2.0 M HCl with a volumetric ratio of 1:3) at 80 °C for 5 h to remove unstable and metal aggregates, followed by a thorough wash with Milli-Q water. Finally, the washed samples were dried overnight in a vacuum oven at 80 °C. The as-prepared samples are labeled as this way: M-PANI-C-AAA-BBB (M: Fe, Co or Ni; AAA indicating heating temperatures: 850, 900, 950, 1000 or 1050oC; BBB labeling the acid for post treatment: H2SO4, HNO3 or Mix Acid). In the later part of this work, samples are only labeled as M-PANI-C (M: Fe, CO or Ni) because all the samples were prepared under exactly the same procedure (1000°C and mixed acids). To further explore the performance of our samples compared with commercial materials, BP 2000 carbon black was chosen as a reference. The scheme and flow chart of synthesis procedure are highlighted in Figure 1 and Figure S1.


Page 6 of 38

Page 7 of 38

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 1. Synthesis of M-PANI-Cs. (A) Polymerization of aniline by adding APS in the presence of metal precursors. (B) Mixing of high–surface-area carbon support with the suspension of PANI and metal precursors. (C) Heat treatment in N2 atmosphere. (D) Acid leaching, removal of metal aggregates.

2.2 Physical Characterization. The morphologies of samples were visualized through scanning electron microscopy (SEM) by Carl Zeiss AURIGA CrossBeam, operated at 3 kV. The crystalline phases present in each sample were identified using X-ray diffraction (XRD) by a Rigaku Ultima IV diffractometer with Cu Kα X-rays (λ = 1.5406 Å). Raman spectrum was measured and collected at an excitation wavelength



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

of 514.5 nm by a Renishaw inVia Raman system. BET surface area and N2 adsorption-desorption isotherms were measured on a Micromeritics TriStar II at 77 K. Samples were pre-treated at 150 °C for 3 hours under vacuum prior to nitrogen physisorption measurements. 2.3 Electrochemical measurements. Cyclic voltammetry (CV) was performed on a CHI 760E electrochemical workstation using a three-electrode test system in an N2 purging electrolyte. The three-electrode system includes an active material coated glass carbon electrode (GCE) or nickel foam electrode (NFE) as a working electrode, a Ag/AgCl (KCl saturated) reference electrode and a graphite rod counter electrode immersed in 1 M KOH over the potential range of 0-1 V at scan rates of 5 ~ 100 mV s−1. The mass loading of active materials on a GCE and a NFE was controlled at ~ 0.2 mg and ~ 3 mg, respectively. Electrochemical impedance spectroscopy (EIS) measurements were performed by Autolab (Metrohm) with the same three-electrode system with an active material coated NFE as the working electrode, with an amplitude of 5 mV and frequencies ranging from 100 kHz to 0.01 HZ. The obtained EIS data was fitted to an equivalent electrical circuit model using Nova 1.11.2 (Metrohm) software. Galvanostatic charge-discharge (GCD) curves of the symmetric capacitors in two-electrode mode were measured on a battery test station (Arbin) with MITS Pro software. The potential window was applied between 0 V and 1.0 V at different current densities (1 ~ 5 A g1

) for 5000 cycles. The working electrodes were prepared by mixing the active materials,

polyvinylidene fluoride (PVDF) and acetylene black at a ratio of 80 mw%:10 mw%:10 mw% in N-Methyl-2-pyrrolidone (NMP) to form a slurry and pressing the slurry onto a nickel foam substrate, then drying the electrodes under vacuum at 80°C for 12 hours. Then two electrodes with same mass loading of active material were pressed together and separated by a filter paper. Figure S2 demonstrates the general procedure of fabrication of a symmetric supercapacitor.


Page 8 of 38

Page 9 of 38

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


3. RESULT AND DISCUSSION 3.1 Effect of heating temperatures The highly porous carbon materials are converted from PANI via a carbonization process. Therefore, the heating temperature is a critical factor in controlling the materials’ morphologies and surface chemistry. Using the Fe-PANI-C system, the effect of heating treatment temperatures on the materials morphologies and resulting electrochemical double layer capacitance was determined as function of temperature from 850 to 1050oC with incremental step of 50oC. With thiss effort is to optimize the temperature, we only used a simple acid treatment with 0.5 M H2SO4. Figure 2A shows the CV curves of Fe-PANI C-samples at a scan rate of 20 mV s-1 in 1.0 M KOH electrolyte. All five CV curves exhibit rectangular shapes (rare peaks showing), indicating ideal supercapacitor charge-discharge performance. Furthermore, the measured capacitance reaches the maximum value (158 F g-1) when the temperature of heat treatment is increased to 1000 °C and then starts to drop at a further increased temperature of 1050 °C. Figure 2B shows the correlation of the heating temperature and the resulting capacitance determined from their CV curves, clearly indicating that 1000°C is the optimal temperature for carbonizing PANI and generates maximum capacitance. Lower temperature such as 850 °C, 900 °C and 950 °C may be not sufficient to fully carbonize PANI that has very low surface areas (~5 m2g-1). This would result in less porous structures which would yield relatively low capacitance. Whereas the excessive high temperature (1050°C) may cause the deconstruction of the porous structure of the sample. As a result, 1000 °C was selected as the optimized temperature for heat treatment in the following sections.



Fe-PANI-C-850 Fe-PANI-C-950 Fe-PANI-C-1050

4

300

Fe-PANI-C-900 Fe-PANI-C-1000

-1

200

2

100

0

0

-2

-100

-4

-200 0.0

0.2

0.4

0.6

0.8

1.0

(B)

158

160 -1

6

Specific Capacitance (F g-1 ) Specific Capacitance (F g )

(A)

Current Density (A g )

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 38

140

115

120

111 100 88

83

80 850

Potential vs RHE (V)

900

950

1000

1050

Temperature of Heat Treatment (C)

Figure 2. (A) CV curves of Fe-PANI-C prepared from various heating temperature ranging from 850 to1050 oC at a scan rate of 20 mVs-1. (B) Calculated specific capacitance of Fe-PANI-Cs as a function of temperatures of heat treatment.

3.2 Effect of acidic leaching treatments Along with heat treatment, acid leaching was also an important factor to consider in this study. In principle, acid leaching treatment provides two key aspects in further enhancing the performance of carbon materials in supercapacitors. First, the in-situ formed metal aggregates can be removed by using the acidic treatment, opening porous structures and exposing increased surface areas. Second, proper acid treatment results in oxygen-containing functional group on the surface, thereby improving hydrophilicity of carbonized carbon materials. In this work, we applied three different acid solutions for acid leaching Fe-PANI-C samples: 0.5 M H2SO4, 30% HNO3 and mix acid (30% HNO3 and 2.0 M HCl with a volumetric ratio of 1:3). Figure 3 A shows the CV curves of Fe-PANI-C samples at a scan rate of 20 mVs-1 as a function of type of acids. Relative to traditional acid treated Fe-PANI-C-1000-H2SO4 sample, the Fe-PANI-C-1000-HNO3 exhibits slightly better performance. However, a significant improvement is found on Fe-PANI-C-1000-


Page 11 of 38

Mix Acid, with a specific capacitance of 444 Fg-1 at a scan rate of 20 mVs-1. Figure 3B further summarizes the comparison of capacitance of Fe-PANI-C samples treated by various acids. The mixed acid treatment is able to generate more than two times higher capacitance when compared to H2SO4 treated sample (158 Fg-1). Therefore, among studied acids, the mixed acid is most efficient to remove the metal species, create more pores inside the sample structure, and generated appropriate surface functional groups, when compared to individual 0.5 M H2SO4 and 30% HNO3. The combination of HCl and HNO3 makes the mixed acid with strong oxidizing property. Under the acid leaching process at 80 oC for 5 h, the mixed acid removes most of the leftover metals and simultaneously introduces large amount of oxygen-containing functional groups (e.g., hydroxyl, epoxy, and carboxyl) to the surface of carbon structures, which make the hydrophilicity of materials improved and more pore volume becomes accessible by electrolyte. Therefore, the mixed acid combination of HCl and HNO3 leads to an increase of electrochemically accessible surface areas, thereby increasing the capacitance.

800

15

-1

400 5

Fe-PANI-C-1000-HNO3

0

Fe-PANI-C-1000-H2SO4

200 0 -200

-5

-400 -10 -600 0.0

0.2

0.4

0.6

0.8

444 -1

Fe-PANI-C-1000-Mix Acid

10

(B) 500

600

Specific Capacitance (F g ) -1 Specific Capacitance (F g )

(A) Current Density (A g )

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


400 300 200

194 158

100 0

1.0


Fe-PANI-C -1000-H2SO4

Fe-PANI-C -1000-HNO3

Fe-PANI-C -1000-Mix Acid

Figure 3. (A) CV curves of Fe-PANI-C-1000 samples treated by using various acids including 0.5 M H2SO4, 30% HNO3 and mixed 30%HNO3+2.0 M HCl at a scan rate of 20 mVs-1. (B) Calculated specific capacitance for these Fe-PANI-C-1000 samples.



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 38

Figure 4. (A) N2 adsorption–desorption isotherms and BET surface area of Fe-PANI-C-900H2SO4, Fe-PANI-C-1000-H2SO4 and Fe-PANI-C-1000-Mix Acid. (B) DFT pore size distribution of Fe-PANI-C-900-H2SO4, Fe-PANI-C-1000-H2SO4 and Fe-PANI-C-1000-Mix Acid. (C) Pore volume distribution of Fe-PANI-C-900-H2SO4, Fe-PANI-C-1000-H2SO4 and Fe-PANI-C-1000Mix Acid.

To better understand how the optimal 1000°C and the mixed acid can generate the highest capacitance, three samples including Fe-PANI-900-H2SO4, Fe-PANI-1000-H2SO4, and Fe-PANI-


Page 13 of 38

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


1000-Mix acid were selected for subsequent N2 adsorption–desorption tests in terms of their porosity and BET surface area. Figure 4A shows the N2 adsorption–desorption isotherms and BET surface area of these selected samples. Compared with Fe-PANI-C-1000-H2SO4 (957 m2 g-1), consistent with the higher supercapacitor performance as observed in Figure 3A, Fe-PANI-C1000-Mix Acid exhibits a much larger BET surface area (1645 m2 g-1), which indicates more efficient removal of metal species to open pore structures by using mixed acid. On the other hand, Fe-PANI-C-900-H2SO4 shows smaller BET surface area (601 m2 g-1) than Fe-PANI-C-1000H2SO4, which agrees with the result shown in Figure 2, low capacitance measured with the lowtemperature treated samples (Fe-PANI-C-900-H2SO4). Figure 4B further compares the pore size distribution among these three samples. All samples exhibit two dominant pore distribution centering at 1.0 nm and 3.0 nm, belonging to the micropore and the mesopore region, respectively. Figure 4C summarize their pore volume distribution. Generally, all three studied samples are comprised of a dominant fraction of mesopore. Compared to 900oC treated sample, an increase in temperature to 1000oC leads to increased micropores and macropores, due to a high degree of carbonization at higher temperatures associated with micropores. In addition, increased temperature leads to generation of larger particles of metallic aggregates and more open macropore structures after acid leaching treatments. Meanwhile, the mixed acid treatment doesn’t change the various pore size distribution, except for slight increases of micro- and mesopores. Therefore, compared to acidic post treatment, heating temperature is the foremost factor to govern the overall porous structures of carbon samples. Table S1 compares more detailed information regarding the values of BET surface area and pore volume distribution of these studied samples.

3.3 Effect of transition metals during the carbonization



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 38

In addition to Fe, other two transition metals (Co, Ni) were also used during the synthesis due to their ability to catalyze graphitized carbon during the carbonization. As a result, Co-PANI-C and Ni-PANI-C samples were prepared using the optimal heating temperature of 1000 °C and acid leaching using mixed acid. Thus, in this section, three metal-derived carbon samples along with the BP 2000 reference were compared systematically in terms of their electrochemical properties by using CV, EIS, and galvanostatic charge-discharge (GCD) with a three-electrode system and a two-electrode system. CV curves (Figure 5A) obtained from active materials reveal a notable difference of capacitance among BP2000 and M-PANI-Cs. At a scan rate of 20 mVs-1, BP 2000 shows the lowest specific capacitance (77 F g-1). Once again, the Fe-PANI-C shows the highest specific capacitance 444 F g-1 followed by Ni-PANI-C (207 F g-1) and Co-PANI-C (175 F g-1). In addition, the measured double layer capacitance currents of Fe-PANI-C increases linearly with the scan rates (Figure 5B), indicating an ideal supercapacitor behavior and perfect accessibility of sample. It should be noted that the highest capacitance measured at the lowest scanning rate of 5 mVs-1 reaches 478 Fg-1 that is a value ever reported for carbon-based electrode for supercapacitors. Figure 5C gives an overall comparison of the specific capacitance for BP2000 and M-PANI-Cs on a GCE as a function of different scan rates. Fe-PANI-C has the highest specific capacitance at each scan rate but drops faster than the other samples at higher scan rates. Co-PANI-C and NiPANI-C maintain more stable performance at higher scan rates. Most likely this is due to a relatively larger fraction of macropore. In addition, specific capacitance for all of samples at different scan rates are listed in Table S2. The ion transfer behavior of BP2000 and M-PANI-Cs based nickel foam electrodes (NFEs) were investigated by using EIS. The representative Nyquist plots are shown in Figure 5D. The equivalent circuit model is shown in Figure S3. RS is the series resistance, RCT is the charge-


Page 15 of 38

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


transfer resistance. RS is derived from the high-frequency intersection of the Nyquist plot in the real axis, including ionic resistance of electrolyte, intrinsic resistance of substrate, and contact resistance at the interface of the active material/current collector. The diameter of the semicircle represents RCT at the electrode/electrolyte interface in the high-frequency region. In the low frequency region, the slope of the curve represents the Warburg resistance, ZW, which represents the diffusion of ions into the porous electrode in the intermediate frequency region.43 In high frequency region, M-PANI-Cs based NFEs presents perfect semicircles, indicating the ideal supercapacitor performance, whereas BP2000 shows an extreme small fraction of circle with large curvature, as shown in the inset. In low frequency region, M-PANI-C based NFEs present sharply increased and linear lines, which are characteristic of pure capacitive behavior, whereas BP2000 has a slowly increased curve (Figure 5D). RS and RCT of BP2000 and M-PANI-C based NFEs were obtained by fitting the Nyquist plots using the equivalent circuit model, which are also listed in Table S2. Details about fitted Nyquist plots of BP2000 and M-PANI-Cs based NFEs can be found in Figure S4. BP 2000 based NFE has both the highest RS (2.86 Ω) and RCT (7.22 Ω), which suggests slow charging and discharging rates are responsible for its lowest capacitance. Fe-PANIC, Co-PANI-C and Ni-PANI-C have similar RS values (1.77 Ω, 1.8 Ω and 1.83 Ω respectively), indicating almost the same series resistances. Fe-PANI-C based NFE has a RCT of 2.07 Ω, which is slightly larger than those of Co-PANI-C (1.60 Ω) and Ni-PANI-C (1.22 Ω) based NFEs, indicating a larger charge transfer resistance. This could be credited to the relatively higher graphitization degree and high content of nitrogen doping at edge sites (pyridinic N) obtained from Co- and Ni-PANI-C samples, corresponding to high charge transfer. Fe-PANI-C, Co-PANI-C, and Ni-PANI-C have very close RS: 1.77 Ω, 1.8 Ω and 1.83 Ω as shown in Table S2, indicating similar series resistance (ionic resistance of electrolyte, intrinsic resistance of substrate, and contact



resistance at the interface of the active material/current collector) of the samples. For the charge transfer resistance, RCT, Fe-PANI-C (2.07 Ω) is a little bit higher than Co-PANI-C (1.6 Ω) and NiPANI-C (1.22 Ω). This very interesting difference can be perfectly explained by the various nitrogen doping content and the relatively content of pyridinic nitrogen, which will be discussed in the XPS characterization. 600

(A) 10

(B) 40 -1

5

200

0

0 -200

-5

-400 BP2000 Co-PANI-C 0.0

0.4

0.6

0.8

BP2000 Fe-PANI-C Co-PANI-C Ni-PANI-C

400

0 -20 -40 5 mV 60 mV

-60

1.0


500

-600

300

0.0

200

0.2

(D) 35 25 20

40

60

80

100

0.8

1.0

5

15

4 3 Fe-PANI-C Co-PANI-C Ni-PANI-C

2 1 0 0

0 20

40 mV

BP2000 Fe-PANI-C Co-PANI-C Ni-PANI-C

5

0

20 mV 100 mV

0.4 0.6 Potential vs RHE (V)

10

100

10 mV 80 mV

30

-Z'' (Ohm)

-1

0.2

Fe-PANI-C Ni-PANI-C

20

-Z'' (Ohm)

-10

(C)

Specific Capacitance (F g ) -1 Current Density (A g )

-1

Current Density (A g )

400

Specific Capacitance (F g )

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 38

0

-1

Scan Rate (mV s )

5

10

15

1

20

2 3 Z' (Ohm)

25

4

30

5

35

Z' (Ohm)

Figure 5. (A) CV curves of BP2000 and M-PANI-Cs on a glassy carbon electrode (GCE) with a scan rate of 20 mV s-1. (B) CV curves of Fe-PANI-C on a GCE with different scan rates. (C) Specific capacitance of BP2000 and M-PANI-Cs on a GCE as a function of scan rate. (D) Nyquist plot of BP2000 and M-PANI-Cs coated nickel foam electrodes (NFEs). Inset: enlarged Nyquist plots for three M-PANI-C electrodes in high-frequency region.


Page 17 of 38

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


During the carbonization process, PANI directly converted into carbon materials. Meanwhile, nitrogen atoms originally in PANI were found able to dope into carbon structures during the thermal treatment.44-45 XPS is a powerful technique to determine nitrogen doping including its doping content and position according to corresponding binding energy. As shown in Figure 6, the peak at relatively low binding energy around 398.6 eV is due to the pyridinic N that nitrogen atoms are doped at the edge sites of carbon plane. At higher binding energy, another dominant peak around 401.2 eV usually can be detected, which is associated with nitrogen atoms doped at the center of carbon planes labelled as graphitic N. It should be noted that transition metals play important role in affecting nitrogen doping during the PANI carbonization. Among studied metals (e.g., Fe, Co, and Ni), Ni induces the highest nitrogen doping content (8.9 at%) with the relatively highest pyridinic nitrogen. Oppositely, Fe is able to yield nitrogen doping with lowest content (4.5 at%) as well as the smallest pyridinic fraction. This suggests that Ni is the most effective to induce nitrogen doping at the edge site of carbon during the PANI carbonization. Oppositely, Fe is favorable for doping nitrogen at the center of carbon planes. In general, pyridinic N often provides active sites for ion adsorption thereby improving ion diffusion and transfer at the interface. This is in agreement with the lowest charge transfer resistance measured with the Ni-PANI-C samples as shown in Figure 5D. On the other hand, graphitic N tends to modify the electron distribution on carbon plane, resulting in polarization and improving hydrophilicity.45 This is partially responsible for the significant increase in electrochemical surface areas of Fe-PANI-C, associated with the highest capacitance determined using CV.



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 6. Nitrogen doping in these M-PANI-C samples determined by using XPS: (A) Fe-PANIC, (B) Co-PANI-C, and (C) Ni-PANI-C. Two major types of doping were identified as pyridinic N at edge sites and graphitic N at center of carbon planes. Therefore, we believe the most favorable nitrogen doping play an important role in enhancing the charge transfer for the non-faradaic process associated with supercapacitors. However, the 18 Environment ACS Paragon Plus

Page 18 of 38

Page 19 of 38

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


overwhelmingly high surface areas obtained from Fe-based carbon materials, as discussed in next part, lead to the largest capacitances. In our ongoing effort, we are trying to combined Fe and Ni to prepare bimetallic carbon materials with an aim to improve the surface areas and nitrogen doping simultaneously. Among studied M-PANI-C, Fe-based carbon materials exhibit the highest BET surface areas (1645 m2g-1) and the largest capacitance up to 478 Fg-1. These remarkable morphologies were further characterized using electron microscopy. Figure 7 represent the overall morphologies for the Fe-PANI-C before and after the carbonization process. The in-situ polymerized PANI exhibits uniform and porous fiber-like network morphologies with a fiber diameter of 100 nm (Figure 7A and 7B). After 1000oC treatment in N2 atmosphere, these well-defined fiber networks change to porous graphene-sheet like morphologies (Figure 7C-7F). The transparent bubble-like structure suggests thin-layered carbon planes. As a reference, SEM images for well-known reduced graphene oxide, are shown in Figure S5. It would be an interesting method to generate thin-layered carbon sheet structures via high-temperature decomposition of polymer, instead of traditional physical and chemical methods for exfoliation of graphite. In addition, the highly porous 3D graphene sheet like structures were further justified by using TEM images (Figure 8). The thinlayered carbon sheet tangled and twisted together in a porous structure is responsible for the highest surfaces area with improved mass transfer determined for the Fe-PANI-C carbon materials. Furthermore, Figure S6 shows the morphologies of BP 2000 and other M-PANI-C samples. BP 2000 presents the amorphous and porous structures with a uniform distribution of carbon particles. Both Co-PANI-C and Ni-PANI-C exhibit less porous structures. A small fraction of graphitized structure was detected, covered by some metal aggregate particles. In particular, Ni-PANI-C



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

contained some short carbon nanotubes that were encapsulated and surrounded by a large amount of amorphous carbon.

Figure 7. SEM images for (A-B) Fe-PANI-C before heating treatment, (C-F) Fe-PANI-C after heating treatment at 1000oC and acid leaching in mixed acid.


Page 20 of 38

Page 21 of 38

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 8. Two tEM images (A) and (B) showing the highly porous and 3D morphology for the Fe-PANI-C samples.

X-ray diffraction (XRD) results show the crystalline structure of BP 2000 and M-PANI-C (Figure 9A). Similar to the peaks of BP 2000 at 25° and 44° associated with carbon,46 there are no Fe aggregates identified from the Fe-PANI-C catalysts, suggesting the mixed acid treatment is able to remove all of the metallic species, leaving porous carbon. Unlike the Fe-PANI-C, in addition to the broad peak around 25o assignable to carbon (002), the distinct XRD peaks associated with metal aggregates in Ni-PANI-C and Co-PANI-C (shown by green and blue stars) could be assigned to nickel carbide (Ni0.98C0.02, JCPDS No. 01-074-5561) and Co9S8 (JCPDS No. 00-056-0002), respectively. The sulfur sources of originates from the APS that was used as an effective oxidant to polymerize aniline during the material synthesis. The carbide could be formed during the heat treatment of Ni-PANI-C. The presence of nickel carbide (Ni0.98C0.02) and Co9S8 indicated that acid leaching cannot entirely remove all the metal species, especially for those enclosed into graphitized carbon shells. This is in agreement with the observation of SEM images, showing graphitic carbon in Co and Ni-derived samples. Also, relative to the major Fe aggregates 21 Environment ACS Paragon Plus


(FeS) in Fe-PANI-C before acid leaching, Ni0.98C0.02 and Co9S8 themselves have stronger acid resistance. The carbon structures in BP2000 and M-PANI-C samples were further analyzed using Raman spectroscopy as shown in Figure 9B. In principle, the relative intensity of D and G bands along with their ratios (ID/IG) can be referred to a degree of disorder of carbon material. All carbon samples demonstrate obvious D and G bands, indicating they have both the sp3 bonds and sp2 bonds, which are typically in diamond and planar sheets carbon structure, respectively.47-48 Calculated ID/IG ratios range from 0.87 to 1.04, indicative of dominant amorphous carbon structures. Among various M-PANI-C samples, Co-PANI-C has the highest degree of graphitelike material, as evidence by its smallest value of ID/IG (0.87).

(2 0 0)



 Ni0.98C0.02

(3 1 1)



(4 4 0)

(4 2 2)



(5 1 1)

(3 3 1)

(4 0 0)





D

G

Ni-PANI-C

 

(2 2 2)

(1 1 1)

 (2 2 0)

 (2 0 0)





(B) 

 Co9S8

Co-PANI-C

Fe-PANI-C

Intensity (a.u.)

(1 1 1)



(2 2 0)

(A)

Intensity (a.u.)

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 22 of 38

20

30

40 50 2 (degree)

60

70

80

Ni-PANI-C

ID/IG= 0.87

Co-PANI-C

ID/IG= 1.04

Fe-PANI-C

ID/IG= 0.9

BP2000 10

ID/IG= 0.99

1000

1500

2000

BP2000

2500

3000

-1

Raman Shift (cm )

Figure 9. (A) XRD patterns and (B) Raman Spectra of BP 2000 and M-PANI-Cs. N2 adsorption–desorption was employed to study the BET surface area, pore size distribution (PSD), and pore volume distribution. As a high-surface-area carbon black material, BP2000 presents a high BET surface area of 1426 m2 g-1. Fe-PANI-C exhibits increased BET surface area of 1645 m2 g-1 indicating that the iron compound acts as an effective sacrificial pore-forming agent during carbonization of PANI and subsequent acid leaching treatment. Co-PANI-C and Ni-PANI-


Page 23 of 38

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


C show similar BET surface area of 542 m2 g-1 and 496 m2 g-1, respectively, which are much smaller than those of Fe-PANI-C and BP2000 (Figure 10A). Blocked pores due to the presence of nickel carbide (Ni0.98C0.02) and Co9S8 in corresponding Ni-PANI-C and Co-PANI-C may be the cause for their low surface areas. Also, BP2000 contains the largest volume of marcopores (> 50 nm), evidenced by its isotherm plot with a sharp increase of P/P0 in the region close to P/P0 = 1.0. This feature was not observed for M-PANI-C samples. DFT pore size distribution of BP2000 and M-PANI-Cs were shown in Figure 10B. BP2000 showed a PSD centered at 0.85 and 1.2 nm belongs to the micropore (0~2 nm) region along with the mesopore region (~3 nm and larger than 10 nm). However, all pores detected in M-PANI-C were located in a relatively small diameter region (

Heteroatom Polymer-Derived 3D High-Surface-Area and Mesoporous

Recommend Documents