Biobased Nano Porous Active Carbon Fibers for High-Performance

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN

Article

Biobased Nano Porous Active Carbon Fibers for High-Performance Supercapacitors Yuxiang Huang, Lele Peng, Yue Liu, Guangjie Zhao, Jonathan Chen, and Guihua Yu ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b02214 • Publication Date (Web): 25 May 2016 Downloaded from http://pubs.acs.org on May 27, 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 34

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

Biobased Nano Porous Active Carbon Fibers for High-Performance Supercapacitors Yuxiang Huang,a,b

Lele Peng,c

Yue Liu,b

Guangjie Zhao,a,* Jonathan Y. Chen ,b,*

and Guihua Yu c a

College of Materials Science and Technology, Beijing Forestry University, Tsinghua

East Road 35, Haidian 100083, Beijing, China b

School of Human Ecology, The University of Texas at Austin, Austin, TX 78712, USA

c

Materials Science and Engineering Program and Department of Mechanical Engineering,

The University of Texas at Austin, Austin, TX 78712, USA Abstract Activated carbon fibers (ACFs) with different pore structure have been prepared from wood sawdust using the KOH activation method. A study was conducted to examine the influence of the activation parameters (temperature, alkali/carbon ratio and time) on the morphology and structure of the as-prepared ACFs developed in the process of pore generation and evolution. Activation temperature was very essential for the formation of utramicropores (< 0.6 nm), which greatly contributed to the electric double layer capacitance. The significance of metallic potassium vapor evolved when the temperature was above 800 oC for the generation of 0.8- and 1.1-nm micropores cannot be ignored. Increasing the KOH/fiber ratio and prolonging the activation time enlarged to some extend the micropores to small mesopores within 2-5 nm. The sample with the optimal condition exhibited the highest specific capacitance (225 F g-1 at a current density of 0.5 A g-1). Its ability to retain capacitance corresponding to 10 A g-1 and 6 M KOH was 85.3 %, demonstrating a good rate capability. With 10000 charge-discharge cycles at 3 A g-1 the supercapacitor kept 94.2% capacity, showing outstanding electrochemical performance as promising electrode material.

* Correspondence to: Jonathan Y. Chen ([email protected]); Guangjie Zhao ([email protected]).

ACS Paragon Plus Environment


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

Keywords: Activated carbon fibers; pore evolution; electric double layer capacitor; capacitance, lignocellulose biomass; wood sawdust. 1. Introduction Supercapacitors, a sort of electrochemical capacitor also called ultracapacitors or electric double layer capacitors (EDLCs), have attracted more attention in applications of high power energy storage owing to their higher power density and long cycling durability.1 Supercapacitors are able to be charged or discharged instantly to a full capacity due to the lower energy density compared to batteries. This feature allows to complement some deficiencies of other power sources in the energy storage field, such as hybrid electric vehicles, load-leveling devices, and memory back-up systems.2-3 As the most common type of porous carbon materials, activated carbons (ACs) are commercially utilized as electrode materials for supercapacitors on account of their low cost and abundant porosity. Theoretically, high specific surface area (SSA) means that electrolyte have more contact to form a double layer capacitors, whereas there was no proportional relationship between SSA and specific capacitance of carbon materials.3 Because not all the pores are electrochemically accessible and the ability of the electrolyte ions to move is determined by the pore size, the pore size affects to a great extent the carbon-based supercapacitors’ performance. Initially, the larger pore size was thought to be advantageous for more accessible surface area. Shi reported that pores were usable for the electro-adsorption of hydrated ions when the pore size was above 0.5 nm.4 It was suggested that carbon-based materials with small mesopores (2-5 nm) would be beneficial to enhance the power capability as well as the energy density if their diameters were larger than those of two solvated ions.3 However, there is also considerable discussion over the capacitive contribution from micropores. Lozano-Castelló et al. attributed the high double layer capacitance to the micropores (< 2 nm) and found that, for the porous 2


Page 2 of 34

Page 3 of 34

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 (SSA > 2000 m2 g-1), the existence of mesopores was not effective for the capacitance.5 The result of the experiment on the carbide-derived carbon with different average pore sizes from Chmiola and his co-workers suggested that pores would be available for charge storage when their diameters were smaller than those of solvated electrolyte ions.6 Raymundo-Pinero et al. also found that small micropores at 0.7 nm helped supercapacitors achieve the maximum capacitance in aqueous media.7 It can be interpreted that a part of ions desolvated to increase their penetration into the micropores.8 Thus, it is very necessary to design the pore size distribution (PSD) of porous carbons elaborately to achieve high capacitance and excellent rate capability. Generally, ACs have a relatively wide PSD in the 0.4-4 nm range,9 which may limit the capacitance and the performance in some supercapacitor applications. Activated carbon fiber (ACF) is another highly porous carbon material in the fibrous form, which has a typical diameter of ~ 10 µm and a very narrow PSD that is mainly microporous ( 0.1), demonstrating a non-microporous type of adsorption. However, the isotherms for the samples prepared at medium (A850-3-2) and high (A850-6-2) alkali/carbon ratio belonged to typical Type IV. This was confirmed by the results listed in Table 1. The SSA of A850-1-2, which Vmeso/Vtot ratio was as high as 82.7%, is close to 700 m2 g-1, mainly resulting from mesoporosity. SBET and Vtot of the ACF samples kept increasing gradually in response to a rising of the KOH/fiber ratio. The Smicro and Vmicro exhibited the same trend while the Smeso and Vmeso reduced slightly. As a consequence, the mesopore volume fraction fell to the lowest value (42.8%). This 10


Page 10 of 34

Page 11 of 34

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


provides an evidence that KOH activation is beneficial for introducing some micropores. The PSD result of A850-1-2 consisted mostly of mesopores with a pore size range of 3-10 nm and some micropores with a peak at 0.5 nm (Figure 2c). In addition to primary micropores at 0.5 nm, the PSD graphs of A850-3-2 and A850-6-2 present secondary and tertiary micropores located in the pore size ranges of 0.7-1.0 nm and 1.0-1.4 nm, respectively. The pore volume of these secondary and tertiary micropores also increased with an increase of the KOH/fiber ratio. The peaks for small mesopores (2.0-4.0 nm) could be observed clearly, but the pore volume of large mesopres (4-10 nm) decreased. According to the above analysis on activation temperature, there should be micropores distributed around the peaks of 0.8 and 1.1 nm when the temperature exceeded 800 oC. However, that was not the case until the KOH/fiber ratio reached 3. This is because of the limited availability of KOH attached on the surface when the KOH/fiber ratio was very low. KOH was exhausted shortly after the activation reaction began. Thus, the process in the coming period of time was substantially carbonization, where the already generated micropores were collapsed to mesopores.19 On the contrary, more micropores would generate by reaction between KOH and carbon as well as the resultant gas such as metallic potassium and H2 if there was sufficient KOH attached on the fibers. As seen from Figure 2a, the types of isotherms for ACFs prepared at different activation time varied from typical Type I (A850-6-1) to Type IV (A850-6-2), also demonstrating that the porous structure could be controlled by varying the activation time. Along with the increase of activation time, the SBET and Vtot increased and attained the highest values (2294 m2 g-1 and 1.412 cm3 g-1, respectively) at 3h owing to the activation depth. This ACF prepared from wood sawdust had better porosity than commercial activated carbon (YP80F) with SBET of ~2100 m2 g-1 and Vtot of ~0.94 cm3 g-1.The PSD of the ACFs produced with different activation time in Figure 2d shows that the pore distributions of 11



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

all the samples remain almost the same independent of activation time. However, the amount of pores in different regions changes markedly with the variation of activation time. The samples prepared with the shortest activation time (A850-6-1) have the largest amounts of micropores at 0.5 nm, which dramatically decreased as the activation time were prolonged to 2 or 3 h. But micropores in other region and mesopores increased gradually with activation time extending. This again confirms that pores at 0.5 nm were firstly generated at the initial stage of activation process. The deep activation process by prolonging the activation time resulted in the enlargement of a fraction of these micropores and the generation of more micropores with large size and mesopores by the activation reaction and derived sufficient potassium vapor. 3.4. Crystal structure Compared with X-ray diffraction (XRD), Raman spectroscopy technique is more sensitive to structural changes of disordered carbons. Thus, Raman spectra of all the KOH-activated samples were measured to characterize their microstructures. Commonly, two bands are presented in a first-order Raman spectrum of disordered amorphous carbon. One at 1600 cm-1 is graphite (G) band caused by the Raman active E2g mode of graphite. The other at ~ 1350 cm-1 is disorder carbon (D) band caused by A1g mode.20 Figure 3a shows typical Raman spectra for the KOH-activated ACFs where both G and D bands are depicted. The presence of the latter is very typical for activated carbon materials that usually have a highly disordered structure after activation. Both of the two bands exhibit asymmetric tailings and it is impracticable to use two Gaussian peaks to reproduce the observed spectrum for curve fitting. Thus, the two bands were further resolved into two pairs of peaks using the same deconvolution technique proposed by Shimodaira and Masui.21 The former pair of peaks (G1 & D1) represented the basal planes with small size. The latter was related to sp2 clusters like α-Cs with bond angle disorder 12


Page 12 of 34

Page 13 of 34

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 3b).21 Table S1 in the Supporting Information presents details of deconvolution parameters in terms of peak intensity ratios (ID1/IG1, ID2/IG2, and IG2/IG1) and the position of the corresponding peaks as well as their full width at half maximum (FWHM). In order to obtain the clear behaviors of the Gaussian parameters, Raman shift (v) vs. FWHM (Γ) was plotted in two dimensions in Figure 3c. It can be observed that the center position of the G1 peak is very steady (1592±4 cm-1) as well as the FWHM (64±6 cm-1). Although the D1 peak also has a steady center position, it has a larger and scattered FWHM (128± 25 cm-1). As for the G2 and D2 peaks, both dimensions are widely scattered. These behaviors are consistent with the results found by Macedo and his co-workers.22 As shown in Figure 3d, the ID1/IG1 ratio increases with the increase of SSA in ACFs activated at 850 oC. According to the Tuinstra and Koenig formula,23 in-plane crystalline size (La) is calculated using the peak intensity ratio of ID/IG. Similarly, the ID1/IG1 is considered to vary inversely with graphitization degree or the in-plane size. Higher SSA meant that more edge planes existed at the fiber surface due to the destruction of microcrystallites by deep activation.24 However, temperature has a great influence on the determination of graphitization degree. The graphite-like microcrystalite structure of ACFs activated below 800 oC was initially formed and their degree of graphitization was very low. Therefore, although the ACFs samples activated below 800 oC had lower SSA than those activated at 850 oC except A850-1-2, their ID1/IG1 value was also higher due to the low graphitization degree. The ID2/IG2 is consistently in direct proportion to the sp2 cluster size.21 As seen in Table S1 in the Supporting Information, this ratio varied irregularly with SSA and activation conditions. It is probably that the differences in sp2 cluster size are not sufficient to account for difference in porosity. For activated carbon materials, the IG2/IG1 ratio is used empirically to indicate the 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

relatively disordered or ordered contents.21 This ratio has the same variation tendency as the ID1/IG1 ratio. It means that the proportion of disordered domains became higher with the process of deepened KOH activation in spite of raising the activation temperature, increasing the KOH/fiber ratio, or prolonging the activation time. 3.5. Surface chemistry For non-destructive characterization of material surface, the XPS measurement was frequently used to determine the element composition on the fiber surface and its binding characteristics. 25 The broad scan of XPS spectra and atomic concentrations on the surface of KOH-activated ACFs were presented in Figure 4a. The two main peaks including C 1s and O 1s appear in the binding energies of 1200-0 eV, suggesting that C and O were the major component elements for the KOH-activated ACF samples, which was consistent with the SEM-EDXA analysis. Apparently, the oxygen content decreases sharply from 14.96% to 6.58% with increasing KOH/fiber ratios, which revealed that oxygen suffered a great loss in abundance of available KOH although more carbon was consumed with increasing the amount of KOH according to the chemical reaction. Either raising the activation temperature or prolonging the activation time also brought about the decrease in the oxygen content, i.e., the removal of more surface oxides. In other words, KOH activation, like carbonization, was also a process of carbon enrichment. Curve fitting is optimized through the decomposition of C 1s into five peaks with binding energies according to our previous study (Figure 4b).26 The O 1s spectrum was fitted to the four components (Figure 4c): (I') carbonyl oxygen in COOR (C=O) (531.5 eV); (II') hydroxyl oxygen (-OH) or ether oxygen (C-O-C) (532.4 eV); (III') noncarbonyl O in COOR (533.8 eV); and (IV') chemisorbed oxygen (H2O, O2) (535.5 eV).27 Figure S3 in the Supporting Information shows the curve-fitted high resolution XPS scans of all the other ACF samples for C 1s and O 1s. 14


Page 14 of 34

Page 15 of 34

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


Table S2 in the Supporting Information lists the results of the fits of the C 1s and O 1s region. In agreement with the Raman analysis results, graphite content of the KOH-activated ACFs activated at 850 oC gradually decreases with increase of KOH/fiber ratios and activation time. It was speculated that the interaction between KOH and carbon resulted in the partial degradation of graphene layers, thereby disrupting the matrix structure. In addition, in the matrix, more unstable carbons were generated that led to mesopore formation. The A850-6-3 sample has the lowest of graphitic carbon fraction (type I=60.90%) as well as the largest mesopore volume (0.61 cm3/g). This low graphitization was available for storing metal ions owing to the disordered structure and huge interlayer distance.28 Dealing with oxygen fraction, hydroxyl oxygen (-OH) or ether oxygen (C-O-C) became the major oxygenated components along with the activation process. The results of the thermogravimetric analysis (TGA) for the precursor fiber of A850-6-3 and A650-3-2 are shown in Figure S4a in the Supporting Information. It can be observed that there was a weight loss of about 50 wt% in the temperature range of 0-1000 o

C, which was mainly attributed to dehydrogenation, oxidation, inter-molecular

cross-linking and inner molecular cyclizing reaction during the carbonization process.24 Upon comparison of TGA and derivative thermogravimetric (DTG) curves of precursor fiber, it was found that post-activation samples had higher thermal stability. With the temperature increasing, their TGA curves are almost two straight lines and DTG curves are flat from 100 to 1000 oC.

The slow weight loss maybe caused by the pyrolysis of

residual phenolic resin that was not completely carbonized in synthesizing the ACF. The A850-6-3 sample presented the highest thermal stability and the highest residual mass (92%), when compared with A650-3-2 (81%). This can be interpreted as lowest oxygen content (6.58%) in A850-6-3 according to the XPS results because the nature and number 15



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 carbon-oxygen surface groups were closely related with the weight loss during the activation process. The evolved CO2 came from the acidic groups, while CO was evolved from the no-acidic groups such as phenol and quinone or carbonyl groups.29 The good surface stability of A850-6-3 will contribute to the improvement of electrochemical performance.30 The evolved CO (m/z 28) and CO2 (m/z 44) mass spectrometer (MS) signal plot measured by TG-MS technique was shown in arbitrary unit during the thermal degradation of A850-6-3 and A650-3-2, respectively (Figure S4b in the supporting information). Apparently, A650-3-2 had more CO and CO2 evolution than A850-6-3, confirming that the former had larger amount of oxygen surface groups. Both of them show most of the CO evolution at the high temperature around 750 oC due to decomposition of phenol or quinone groups. The amount of CO2 evolved from both of them was significantly lower than that corresponding to CO evolution, demonstrating that the concentration of anhydride, lactonic and carboxyl surface groups became lower. This was consistent with the XPS results. Both of the samples displayed at least two overlapped peaks between 200 and 1000 oC, corresponding to a decomposition of the different acidic surface groups. 3.6. Electrochemical performance The as-prepared ACFs were assembled to supercapacitors to investigate their true performance, rather than to analyze the voltages or the faradic reactions at a single surface. Figures 5a-c show the typical CV behaviors of all the ACF-based supercapacitors in a 6 M KOH electrolyte. All the electrodes present the well-defined rectangular CV curves in the range of 0-1 V, owing to the typical capacitive behavior of the ACF electrodes. The CV curve loop areas is getting larger as the activation temperature increased, indicating larger specific capacitance (Figure 5a). Raising the KOH/fiber ratio (Figure 5b) and 16


Page 16 of 34

Page 17 of 34

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


prolonging the activation time (Figure 5c) also improved the specific capacitance of the ACFs. Galvanostatic discharge, a widely accepted method to determine the performance of packaged supercapacitors,31 is carried out to determine the capacitance of the ACF-based electrodes. Figure 5d shows the GC profiles of the all the supercapacitors at 0.5 A·g-1 current density. The linear voltage-time dependence also demonstrated the typical capacitive behavior of the electrode. Moreover, no obvious IR drop was found in the curves, indicating good conductivity of these ACF electrode materials. Calculated from the discharge curve, the specific capacitance for A850-6-3, A850-6-2, A850-3-2, A850-6-1, A750-3-2, A650-3-2, and A850-1-2 at 0.5 A g-1 are 225, 192, 177, 160, 146, 113, and 102 F g-1, respectively. The capacitive performance achieved in this study was higher than that of the supercapcitors made of activated carbons eucalyptus wood sawdust, cellulose and potato starch in 6 M KOH electrolyte,32 with reported specific capacitance of 143 F g-1 and high SSA of 2967 m2 g-1. The dependence of SSA, specific capacitance, and normalized capacitance by SSA at 0.5 A g-1 on the activation temperature (Figure 6a), KOH/fiber ratio (Figure 6b) and activation time (Figure 6c) is plotted to illustrate the influence of activation conditions as well as porosity on the capacitive behavior of ACF-based electrodes. Corresponding to the CV curves, the thorough activation enhanced the capacitance. The variation tendency of specific capacitance was very similar to that of SSA. This confirms that carbon with high SSA has a strong ability to store charges. However, it was reported that there was no clear relationship between SSA and specific capacitance,4 which was also presented by the variation of normalized capacitance in Figure 6. The sample A750-3-2 with low SSA of 1013 m2/g has the largest capacitance per unit area. According to the PSD (Figure 2b), the volume of ultramicropores (

Biobased Nano Porous Active Carbon Fibers for High-Performance

Recommend Documents