Article pubs.acs.org/Langmuir
Carbon Xerogel Microspheres and Monoliths from Resorcinol− Formaldehyde Mixtures with Varying Dilution Ratios: Preparation, Surface Characteristics, and Electrochemical Double-Layer Capacitances Zulamita Zapata-Benabithe,†,§ Francisco Carrasco-Marín,† Juan de Vicente,‡ and Carlos Moreno-Castilla*,† †
Departamento de Química Inorgánica and ‡Departamento de Física Aplicada, Universidad de Granada, 18071 Granada, Spain S Supporting Information *
ABSTRACT: Carbon xerogels in the form of microspheres and monoliths were obtained from the sol−gel polymerization of resorcinol and formaldehyde in the presence of potassium carbonate as catalyst, using water as solvent and two different molar dilution ratios. The objectives of this study were as follows: to investigate the effect of the dilution ratio, polymerization reaction time, and temperature on the rheological properties of the sols used to prepare the carbon xerogel microspheres and monoliths; and to determine the influence of their preparation methods and shapes on their surface characteristics and electrochemical double-layer (EDL) capacitance. An increase in the molar dilution ratio produced a decrease in the apparent activation energy of the sol−gel transition. Carbon xerogel microspheres were steam-activated at different burnoff percentages. The morphology, surface area, porosity, and surface chemistry of samples were determined. The main difference between the carbon xerogel microspheres and monoliths was that the latter are largely mesoporous. Better electrochemical behavior was shown by carbon xerogels in monolith than in microsphere form, but higher gravimetric and volumetric capacitances were found in activated carbon xerogel microspheres than in carbon xerogel monoliths.
1. INTRODUCTION Carbon gels are porous carbon materials that have found applications in adsorption, catalysis, and electrochemical energy storage1−3 since their introduction by Pekala.4 They are considered nanostructured carbons, because their structure and porous texture can be designed and controlled at the nanometer scale.5 Carbon gels are obtained after carbonizing organic gels prepared by sol−gel polymerization of compounds such as resorcinol (R) and formaldehyde (F) using basic or acid catalysts. The synthesis and processing conditions influence the surface area and pore texture of the carbon gels and were recently reviewed.6 The RF polymerization reaction involves an addition and condensation mechanism4 that transforms a viscous liquid into a viscoelastic solid at the gelation point and then into a glassy material at the vitrification point. These transitions depend on the composition of the RF mixture and the reaction temperature and can be studied by examining the rheological properties of the mixtures,7 especially the complex modulus, G*, and complex viscosity, η*.8−11 The knowledege of these properties is essential to obtain carbon gels with different forms, for example, monoliths, films, microspheres, and microbeads, by casting them in molds or by emulsifying the RF mixture before its gelation, when it is in a liquid state. This © 2013 American Chemical Society
versatility in the carbon gel forms is one of the key properties of these materials. However, and despite the importance of the knowledge of the rheological properties of RF and similar mixtures for carbon gel preparation, very few published studies have determined the variation of the relevant rheological material functions, G* and η*,8−11 with polymerization reaction time and temperature. This is specially important in the case of the preparation of organic gel microspheres by the inverse emulsion of RF sols in the presence of the appropriate surface-active agent.9 Thus, the size of spherical RF carbon aerogels can be controlled by changing the apparent viscosity of the RF sol.11 In addition, the formation of spherical Cu-doped organic RF xerogels depends on the reaction time selected for the inverse emulsion.9,10 Synthesis of carbon gel microspheres is desirable because they have demonstrated attractive characteristics, including a smooth surface, high mechanical strength, and easy manageability,12 allowing their utilization as fillers for adsorption and catalysis applications in environmental remediation.13 Carbon microspheres are also used as electrochemical double-layer Received: February 27, 2013 Revised: April 23, 2013 Published: April 25, 2013 6166
dx.doi.org/10.1021/la4007422 | Langmuir 2013, 29, 6166−6173
Langmuir
Article
observed by SEM after cutting them, embedded in an epoxy resin, using an ultramicrotome (Reichert-Jung Ultracut S) with a diamond knife. Porous texture was characterized by N2 and CO2 adsorption isotherms at −196 and 0 °C, respectively, using an Autosorb 1 apparatus from Quantachrome after outgassing samples overnight at 110 °C under high vacuum (10−6 mbar). The Dubinin−Radushkevich (DR) equation was used to obtain micropore volumes, W0(N2) and W0(CO2), and mean micropore widths, L0(N2) and L0(CO2). Mesopore volume, Vmeso, was obtained from the difference between the amount of N2 adsorbed at relative pressures of 0.95 and W0(N2). Surface area, SBET, was estimated from the BET equation applied to the N2 adsorption isotherm. Particle density, ρ, was determined by mercury picnometry at atmospheric pressure. Surface chemistry was studied by X-ray photoelectron spectroscopy (XPS), using an Escalab 200R system (VG Scientific Co.) equipped with Mg Kα X-ray source (hγ = 1253.6 eV) and hemispherical electron analyzer. Survey and multiregion spectra were recorded at C1s and O1s photoelectron peaks. Each spectral region of photoelectron interest was scanned several times to obtain good signal-to-noise ratios. The C1s peak at 284.6 eV was used as internal standard in order to obtain the number of components, position of peaks, and peak areas. The depth of the XPS analysis is around 2−3 nm; therefore, the O content obtained by this method, OXPS, is the concentration on the outermost surface of the materials. 2.5. Electrochemical Measurements. Electrochemical measurements were carried out in a Biologic multichannel potentiostat at room temperature using 1 M H2SO4 as electrolyte in a typical threeelectrode cell, with Ag/AgCl as reference electrode and Pt wire as counter electrode. The working electrode was a rectangular (20 × 10 mm) graphite paper pasted with a homogeneous mixture of the carbon xerogel microspheres or finely ground carbon xerogel monoliths with acetylene black and binder (polytetrafluoroethylene, PTFE) at a mass ratio of 80:10:10. Cyclic voltammograms (CVs) were obtained at a scan rate of 0.5 mV/s in a potential window between 0 and 0.7 V. The gravimetric capacitance, CCV (F/g), was calculated from the CVs by eq 1
(EDL) capacitors because of the good electrochemical performance yielded by their high packing density and microstructure.14,15 The objectives of this work were as follows: to investigate the effect of the dilution ratio, D, and polymerization reaction time and temperature on the rheological properties of sols used to prepare carbon xerogel microspheres and monoliths; and to determine the influence of the preparation methods and shapes on their surface characteristics and EDL capacitance.
2. EXPERIMENTAL SECTION 2.1. Rheological Properties. Two organic gels (A and B) were prepared by the sol−gel polymerization reaction of resorcinol (R) with formaldehyde (F) in water (W) using potassium carbonate as catalyst (C). The R/F and R/C molar ratios were 0.5 and 800, respectively. The dilution molar ratio, D = W/R + F + C, was 2.53 and 4.43 and the solution pH was 6.1 and 5.9 for A and B, respectively. The rheological properties of the sol−gel mixtures were investigated using a torsional rheometer (Anton Paar MCR 501) under isothermal conditions. Small-amplitude oscillatory shear (SAOS) time sweeps were preferred because they are nonintrusive tests, unlike the more conventional steady shear measurements,10 because the deformation level is always kept at a low value in the linear viscoelastic region. Dynamic time sweeps were carried out in a parallel plate geometry (50 mm diameter and 1 mm gap thickness). The temperature ranged between 40 and 50 °C and was controlled with a P-PTD 200 Peltier system from Anton Paar. The strain amplitude was 0.1%, and the frequency 1 Hz. Cleaning the shearing tools after running a test is a challenging task, given that viscoelastic moduli increase by several orders of magnitude at the gelation point. Hence, carefully designed experimental tests involving steel disposable plates were used, allowing determination of both gelation and vitrification times. 2.2. Synthesis of the Carbon Xerogel Microspheres. RF polymerization reaction was carried out at 40 °C and at different gelation times, tg, selected from the rheological tests. Portions of 22 cm3 of the sols were removed and emulsified by pouring them into cyclohexane (1100 cm3) containing Span 80 (11 cm3) as surface-active agent, with a hydrophile−lipophile balance (HLB) of 4.3. Suspensions were stirred at a constant speed of 280 rpm for 1.5 h at 40 °C. Next, the suspensions were filtered and the gels were immersed in acetone for 5 days. Subsequently, the wet gels were dried at room temperature overnight, and 12 g of them were then dried in a microwave oven under N2 flow at 384 W for periods of 1 min until constant weight. Weight loss during the microwave treatment was around 5%. Times selected to prepare the microspheres were those when the storage (G′) and the loss (G″) moduli were equal or when the |η*| value of the A and B sols was equal. The organic xerogel (OX) microspheres are designated by addition of the letter A or B (according to the sol used) and of the letter G or η (according to the reaction time). Thus, organic xerogel OXAG indicates an organic xerogel prepared with recipe A and emulsified when G′ = G″. These samples were carbonized at 900 °C in N2 flow (300 cm3/min) for 5 h. Carbon xerogels are referred to in the text by replacing the letter O in the organic xerogels with the letter C. Finally, the carbon xerogels were steam-activated at 840 °C using a N2 flow (500 cm3/min) saturated with water vapor at 25 °C to obtain samples with different burnoff percentages, whose value is appended to the name of the carbon xerogels. 2.3. Synthesis of the Carbon Xerogel Monoliths. A and B sols were cast into glass molds (45 cm length × 0.5 cm i.d.) and sealed, and the mixtures were then cured for 5 days at different temperatures up to 80 °C. After the curing cycle, the hydrogel rods were cut into 5-mm pellets. Drying of the organic monolithic hydrogels and their subsequent carbonization were similar to the procedures for the microspheres. Monolithic carbon xerogels are designated MCXA or MCXB. 2.4. Surface Characterization. External sample morphology was examined by scanning electron microscopy (SEM) using Carl Zeiss SMT equipment. The internal morphology of the microspheres was
CCV =
∑ |I |Δt 2mΔV
(1)
where Σ|I|Δt is the area of the current (A) against time (s) curve, m is the mass of active material in the electrode (g), and ΔV is the potential window (V). Chronopotentiograms (CPs) were performed at current loading between 0.125 and 4 A/g in a potential interval of 0−0.7 V. The gravimetric capacitance, CCP (F/g), was obtained from these measurements by eq 2 CCP =
IdΔt mΔV
(2)
where Id is the discharge current (A), Δt is the discharge time (s), and ΔV is the potential interval (V). In addition, the equivalent series resistance (ESR) was calculated at the beginning of the discharge side of the CPs.
3. RESULTS AND DISCUSSION 3.1. Rheological Properties of the Sols. SAOS tests are ideal for examining the linear viscoelastic properties of sols during gelation and vitrification,7,10,16 allowing determination of all relevant material functions, including the complex modulus, G*, and complex viscosity, η*, which are given by eqs 3 and 4, respectively. G* = G′ + IG″
(3)
η* = η′ − iη″
(4)
The elastic or storage modulus, G′, is the amplitude of the portion of the stress wave that is in phase with the strain wave divided by the amplitude of the strain wave. It is a measure of the deformation energy stored by the sample during the 6167
dx.doi.org/10.1021/la4007422 | Langmuir 2013, 29, 6166−6173
Langmuir
Article
shearing process. The viscous or loss modulus, G″, represents the viscous behavior of the material and is a measure of the deformation energy lost by the sample during shearing. G″ is defined as the amplitude of the portion of the stress wave that is out of phase with the strain wave divided by the amplitude of the strain wave. The dynamic viscosity, η′, is given by the G″/ω ratio, and the storage viscosity, η″, out-of-phase component, is equal to the G′/ω ratio, where ω is the angular velocity. The loss factor or damping factor, tan δ, is calculated as the ratio of the lost to stored deformation, tan δ = G″/G′, where δ is the phase angle between the stress and strain. For a viscoelastic material, 0 < δ < 90°, and the gelation point is reached when δ = 45°; therefore, G′ = G″. The gelation time, tg, is obtained from the intersection of tan δ against time curves at various frequencies, for example, by using multiple waveform rheology to obtain an accurate determination.17,18 However, RF solutions exhibit an abrupt sol−gel transition that precludes the use of multiwave rheology tests. Therefore, the determination of the gelation point from the time course of G′ and G″ moduli, using dynamic time sweeps at a constant excitation frequency, is considered a sufficiently accurate method to evaluate gelation synthesis conditions.8 Figure 1a depicts, as an example, the experimental curves obtained with sample A for the change in storage (G′) and loss (G″) moduli as a function of reaction time at temperatures between 40 and 50 °C (curves at 42.5 and 47.5 °C are omitted for the sake of clarity). In the temperature range selected, both
moduli curves show the typical behavior for RF mixtures. Polymerization and cross-linking reactions between R and F take place during the sol−gel transition. The polymerization reaction mechanism has two main steps. The first is related to the formation of methylol derivatives by the addition reaction of F to R, and the second is the polycondensation reaction of methylol groups, which gives rise to a three-dimensional network.4,19 A sudden increase in viscoelastic moduli (molecular weight goes to infinity) is observed close to the gel transition. According to the tan δ criterion, the gelation point occurs when G′ = G″, which is in agreement with visual observation of the mixtures. After the gelation point, the curves of both moduli continue to increase until they reach a maximum value and stabilization, showing the complete formation of a polymeric structure. The gelation temperature is a significant factor for accelerating the crosslinking during the progress of chemical reactions. Table 1 compiles the values of G′, G″, |η*|, and tg obtained in the temperature range between 40 and 50 °C. Table 1. Rheological Properties of the Sol at Different Reaction Temperatures reaction temperature, °C
gelation time, tg, min
G′ = G″, Pa
|η*|, Pa·s
A
40.0 42.5 45.0 47.5 50.0
400 297 255 213 161
91 100 105 79 109
20 23 22 18 25
B
40.0 42.5 45.0 47.5 50.0
433 371 287 268 231
98 52 60 41 81
22 12 14 9 18
sol
The values of the viscoelastic moduli and the complex viscosity do not significantly change at the gel point, that is, at the reaction time where G′ = G″. This indicates that, as expected, all chemical reactions are the same regardless of the temperature. The deviations of the viscoelastic moduli and complex viscosity values for a given sample at different temperatures are within experimental error, acceptable in rheology. Thus, Figure 1a shows that there is a sudden increase of several order of magnitude in the moduli values in a narrow and well-defined time interval. Conversely, the tg measurement has a very low error. The increase in temperature significantly accelerates the gelation process due to the faster motion of molecules and clusters, diminishing the gelation time (tg). Apparent activation energy of the sol−gel transition, Ea, can be obtained by applying the Arrhenius eq 5 to the variation in tg with reaction temperature T (in Kelvin). 1/tg = A exp( −Ea /RT )
Figure 1. (a) Variation of the viscoelastic moduli G′ (thick line) and G″ (thin line) with time and temperature for sample A. (b) Arrhenius plot for the sol−gel transition, sample A (closed blue up-triangle) and sample B (closed red square). (c) Time dependence of the damping factor (tan δ), with time and temperature for sample A. (d) TTT diagram showing the cure temperature in function of cure time. Gelation time, circles; and vitrification time, triangles and squares. Vitrification criteria: peak in G″ (open and closed green up-triangle) and peak in tan δ (open and closed red square). Sample A, open symbols; sample B, closed symbols.
(5)
The underlying assumption is that the cure kinetics involves only a single reaction with a single activation energy, although, in practice, Ea includes the addition and condensation reactions during gel formation. Figure 1b depicts the linear relationship between ln(1/tg) and 1/T. The Ea values obtained are 73.0 ± 4.7 and 55.0 ± 5.1 kJ/mol for samples A and B, respectively. Job et al.8 used a constant stress rheometer to study the gelation of RF mixtures in the presence of Na2CO3, with R/C 6168
dx.doi.org/10.1021/la4007422 | Langmuir 2013, 29, 6166−6173
Langmuir
Article
surface of the organic xerogel microspheres is very clean (Figure 2a), but after carbonization it exhibits adhered small fibrous-like particles (Figure 2b,c). These can derive from carbonaceous residues left on the microsphere surface by the SPAN 80 used for the emulsion polymerization.21 These carbonaceous residues are eliminated in the subsequent steamactivation (Figure 2d). Microsphere diameter distributions were obtained from the SEM micrographs used to calculate the characteristic microsphere diameters reported in Table 2. These diameters remain
molar ratios between 500 and 2000 and a constant dilution ratio of 5.7 in the temperature range between 50 and 90 °C. They found Ea to be independent of the R/C molar ratio and obtained a value of 80 kJ/mol, similar to that found for sample A in the present study. However, our results show that a higher dilution ratio (sample B) produces a decrease in the Ea value, which could be explained because a decrease in the catalyst concentration, due to the increase in dilution ratio, diminishes the growth rate of the colloidal particles.20 Thus, the smaller colloidal particles the lower energy to move them through the solution and cross-link each other. Vitrification is the second transition in R-F mixtures and takes place after long time periods, showing a shoulder in G′ and a local maximum in G″ (see Figure 1a). Vitrification is also evidenced by a peak in the plot of tan δ against reaction time (Figure 1c). Both criteria were used to determine the vitrification time. Figure 1d shows the time−temperature transformation (TTT) diagram plotting the gelation and vitrification against the corresponding gelation and vitrification times. This Figure demonstrates that both criteria for vitrification times yield highly similar results. The vitrification time increases with higher dilution ratio, similar to observations for the gelation time. 3.2. Surface Characteristics of the Carbon Xerogel Microspheres and Monoliths. OXAG and OXBG were prepared by inverse emulsion after 400 and 433 min of reaction time, when G′ = G″ at 40 °C (Table 1), whereas OXAη and OXBη were emulsified after 283 and 310 min of reaction time, when the |η*| value of both samples was equal to 0.05 Pa·s. Figure 2 depicts, as an example, typical SEM images of the organic and carbon xerogel microspheres. In all cases, the particles have a spherical shape, regardless of the dilution ratio or the tg value used for emulsion of the sol−gel mixture. The
Table 2. Characteristic Diameters (μm) and Standard Deviation, SD, of the Organic and Carbon Xerogel Microspheres sample
mean
median
mode
SD
OXAG CXAG CXAG-10 CXAG-30 OXBG CXBG CXBG-50 OXAη CXAη CXAη-20 CXAη-30 OXBη CXBη CXBη-20 CXBη-30
14 14 13 9 12 12 9 15 12 13 10 9 9 8 7
11 11 11 7 12 12 9 16 12 14 9 10 9 7 6
6 4 5 5 5 5 2 3 2 2 2 6 4 3 3
9 12 9 6 7 7 5 8 7 7 7 4 5 5 5
unchanged or lower with carbonization and systematically decrease with a greater degree of steam activation of the carbon xerogel. Samples from the organic xerogel OXBη show the lowest characteristic diameters, between 7 and 9 μm. The internal part of the microspheres (Figure 2e) is composed of fused primary particles, as also seen in the monoliths (Figure 2f). Figure 3 depicts, as an example, N2 adsorption−desorption isotherms on selected samples. All microsphere samples show type I adsorption isotherms22 with no hysteresis cycles. This is typical of microporous solids, although there is a slight increase in N2 uptake with the rise in relative pressure after micropore filling, indicating the presence of mesopores below 4 nm in size. Conversely, the carbon xerogel monoliths show type IV adsorption isotherms with type H hysteresis cycle,22 typical of mesoporous materials. In this case, the BJH method was applied to obtain the mean mesopore size, dBJH. Table 3 compiles the results obtained from these isotherms and from the CO2 adsorption isotherms at 0 °C. Results for microsphere samples (Table 3) indicate an increase in SBET when the organic xerogels are carbonized, producing a slight increase in particle density. SBET values of the carbon xerogel microspheres obtained at G′ = G″, samples CXAG and CXBG, increase with higher dilution ratio, whereas the values of the microspheres prepared at a similar |η*| value, samples CXAη and CXBη, are virtually unaffected by the dilution ratio. As expected, the SBET value is higher with a greater degree of activation. All samples are microporous, with a very low mesopore volume, and the activation produces a much greater increase in the microporosity than in the mesoporosity. All carbon xerogel
Figure 2. SEM micrographs of samples: (a) OXAG, (b,c) CXAG, (d) CXAG-10, (e) CXAG-10, interior morphology of a microsphere, and (f) CAXG monolith. 6169
dx.doi.org/10.1021/la4007422 | Langmuir 2013, 29, 6166−6173
Langmuir
Article
and single C−OH bonds, respectively.25 In all samples, the relative surface concentrations of these functionalities are around 30−40% (double CO bonds) and 70−60% (single C−OH bonds), resulting in a predominance of the last functionalities, such as carboxyl, hydroquinone, and phenol groups. The SBET increases with higher dilution ratio in the carbon xerogel monoliths (Table 3), and the SBET values of MCXA and MCXB are very close to those of CXAη and CXBG, respectively. The main difference between the carbon xerogel microspheres and monoliths is that the latter are largely mesoporous. This may be related to the surface-active agent used to prepare the hydrogel microspheres by inverse emulsion, because the HLB value of the surface-active agent, in the range of 1.8−8.6, has been reported to influence the micro- and mesopore volume of the final carbon gel obtained, with a large reduction in the mesopore volume at higher HLB values.21 The Vmeso value of the carbon xerogel monoliths largely increases with a rise in the dilution ratio, and the same occurs with dBJH that increases from 4.9 to 29.3 nm. This produces a large decrease in ρ from 0.65 to 0.31 g/cm3. Both carbon xerogel monoliths show very similar surface oxygen content, which is much lower than that of the carbon xerogel microspheres. 3.3. Electrochemical Capacitances. The three-electrode system used in this work, although it is far from the real twoelectrode one, is useful for determining and comparing the specific capacitance of different materials.26 Figure 4 depicts typical CVs at 0.5 mV/s and CPs at 0.125 mA/g obtained with the carbon and activated carbon xerogel microspheres. Carbon xerogel microspheres show nonrectangular and asymmetric CVs (Figure 4a), indicating a deviation from the ideal EDL capacitor. They are narrower at higher potential due to diffusional limitations of the hydrated bisulfate ions into the micropores, producing an ion-sieving effect.27,28 The CP shape of these samples (Figure 4b) is not triangular and also reflects limitations on the diffusion of ions into the micropores. Conversely, CVs of the activated carbon xerogel microspheres (Figure 4c) have a quasi-rectangular shape and their CPs
Figure 3. N2 adsorption−desorption isotherms at −196 °C. Adsorption (open symbols) and desorption (closed symbols).
microspheres show W0(N2) ≤ W0(CO2), due to the presence of micropore constrictions23 or obstructions that are likely of the same nature as the carbonaceous residues adhered on the microsphere surfaces. However, because they are removed by the activation, W0(N2) > W0(CO2) in the activated carbon xerogel microspheres, indicating a heterogeneous micropore size distribution. Thus, L0(N2) ≈ L0(CO2) for the carbon xerogels but L0(N2) > L0(CO2) after their activation. Finally, particle density, as expected, decreases with activation. Surface oxygen content decreases with a greater degree of steam activation, likely due to oxygen functionalities fixed within the new micropores created, rendering these inaccessible to the XPS technique, as reported elsewhere. 24 The deconvoluted O1s XP spectra show two peaks at 531.8 and 533.3 eV, corresponding to functionalities with double CO
Table 3. Surface Area, Porosity, Particle Density, and Surface Oxygen Content of the Organic and Carbon Xerogel Microspheres and Monolithsa
a
sample
SBET, m2/g
OXAG CXAG CXAG-10 CXAG-30 OXBG CXBG CXBG-50 OXAη CXAη CXAη-20 CXAη-30 OXBη CXBη CXBη-20 CXBη-30 MCXA MCXB
9 554 1140 1292 23 746 1644 13 663 1242 1133 13 616 1154 1243 659 709
W0(N2), cm3/g
W0(CO2), cm3/g
L0(N2), nm
L0(CO2), nm
Vmeso, cm3/g
0.217 0.455 0.520
0.284 0.316 0.408
0.63 0.81 1.00
0.61 0.65 0.72
0.065 0.020 0.033
0.292 0.695
0.304 0.294
0.59 1.52
0.59 0.70
0.013 0.046
0.259 0.497 0.460
0.327 0.354 0.376
0.60 0.86 0.94
0.60 0.66 0.70
0.029 0.026 0.028
0.241 0.460 0.496 0.291 0.273
0.308 0.401 0.417 0.311 0.285
0.61 0.78 0.84 1.78 0.68
0.59 0.70 0.70 0.73 0.55
0.064 0.016 0.033 0.253 0.892
ρ, g/cm3
OXPS, wt %
0.88 0.91 0.85 0.80 0.76 0.80 0.61 0.74 0.77 0.70 0.66 0.88 0.90 0.67 0.67 0.65 0.31
n.d. 13.3 6.3 2.4 n.d. 12.1 6.5 n.d. 10.4 5.7 2.2 n.d. 9.9 4.0 3.2 5.4 6.5
n.d.: Not determined. 6170
dx.doi.org/10.1021/la4007422 | Langmuir 2013, 29, 6166−6173
Langmuir
Article
the internal surface of micropore walls is inaccessible, due to ion diffusion, at high scan rates.30 The decrease in CCV with higher scan rate is greater with the carbon xerogel microspheres (43−64%) than with their activated equivalents. CXAη-30 shows the lowest decrease in CCV at the scan rate used, 11%, likely due to its more appropriate microporosity. Activated carbon xerogel microspheres also show lower ESR and higher CCP values in comparison to the corresponding carbon xerogel microspheres, attributable to diffusional limitations of the electrolyte into the micropores of the latter samples. The CCP values of the activated carbon xerogel microspheres are very high, between 190 and 250 F/g, and they can be related to their micropore texture. Thus, Figure S1 in the Supporting Information shows that CCP tends to decrease when the total micropore volume W0(N2) and its mean width L0(N2) increase, but it tends to increase when the narrower micropore volume W0(CO2) and its mean width L0(CO2) increase within the micropore volume and width range studied. These results indicate the importance of the narrower micropores (below about 0.7 nm width) in the EDL formation on these samples. Activated carbon xerogel microspheres show a very large volumetric capacitance, VCCP, (between 115 and 166 F/cm3) due to their elevated particle density. Samples CXAG-30 and CXAη-30 have a similar VCCP, 166 F/cm3, indicating that this value is not influenced by the differences in the preparation of their organic xerogel microspheres. Figure 5a shows that CCP decreases when current loading increases from 0.125 through 4 A/g. The capacitance retention at the highest current loading (Table 4) is 60−70%, except for CXBG-50 that is smaller due to its wider microporosity. Figure
Figure 4. (a) CVs at 0.5 mV/s and (b) CPs at 0.125 A/g in 1 M H2SO4 for samples: CXAG (solid blue circle), CXAη (solid green square), CXBG (solid red tilted square), and CXBη (solid black uptriangle). (c) CVs at 0.5 mV/s and (d) CPs at 0.125 A/g in 1 M H2SO4 for samples: CXAG-30 (open blue circle), CXAη-30 (open green square), and CXBη-30 (open black up-triangle).
(Figure 4d) are triangular, indicating that these samples behave as ideal EDL capacitors. This is because the activation increases the mean micropore width and volume. CVs of these activated samples show small and wide humps at around 0.3 V, related to their surface functionalities.29 Table 4 compiles the gravimetric capacitances, CCV and CCP, obtained from these curves by applying eqs 1 and 2. CCV decreases when the scan rate increases between 0.5 and 10 mV/s because ions can reach the internal and external surface of porous electrodes at low scan rates, whereas part of Table 4. Gravimetric Capacitance, CCV, at Different Scan Ratesa CCV (F/g) sample CXAG CXAG10 CXAG30 CXBG CXBG50 CXAη CXAη-30 CXBη CXBη-30 MCXA MCXB
0.5 mV/s
3.5 mV/s
10 mV/s
CCP, F/g
VCCP, F/cm3
RCP, %
ESR, mΩ
107 173
78 167
61 135
81 191
74 162
n.d. 64
21.1 3.5
201
180
140
207
166
70
3.2
19 154
11 125
8 104
38 188
30 115
n.d. 45
33.3 3.4
67 221 25 208 119 161
36 215 13 186 n.d. n.d.
28 196 9 160 n.d. 125
60 251 48 207 102 169
46 166 43 139 66 56
n.d. 56 n.d. 60 74b 85b
30.7 4.0 37.1 4.0 3.6 3.5
Figure 5. (a) Variation in gravimetric capacitance with current load for samples: CXAG-10 (blue half-circle), CXAG-30 (open blue circle), CXAη-30 (open green square), CXBη-30 (open black up-triangle), and CXBG-50 (open red tilted square). (b) Variation of the gravimetric capacitance (CCP) with the number of charge−discharge cycles at 0.5A/g in the potential window between 0 and 0.7 V. Samples CXAη-30 (open green square) and CXBη-30 (open black up-triangle).
a
Gravimetric and volumetric capacitances from chronopotentiometry, CCP and VCCP, respectively, at 0.125 A/g. Capacitance retention at 4 A/g, RCP, and equivalent series resistance, ESR. n.d.: Not determined. b Capacitance retention values at 1 A/g. 6171
dx.doi.org/10.1021/la4007422 | Langmuir 2013, 29, 6166−6173
Langmuir
Article
5b depicts the CCP variation with the number of charge− discharge cycles at 0.5 A/g and in a potential window between 0 and 0.7 V for selected samples. Results obtained show no fading after 2000 cycles. Figure 6 depicts CVs at 0.5 mV/s and CPs at 0.125 A/g obtained with the carbon xerogel monoliths. They have
the main difference between the carbon xerogel microspheres and monoliths was that the latter were largely mesoporous. This was related to the surface-active agent used to prepare the hydrogel microspheres by inverse emulsion. Activated carbon xerogel microspheres evidenced lower ESR values and higher gravimetric capacitances in comparison to the corresponding carbon xerogel microspheres; the activated microspheres also had very large volumetric capacitances due to their high particle density. The capacitance of the activated carbon xerogel microspheres increased as the narrower micropore volume and its mean width increased. Better electrochemical behavior was shown by the carbon xerogels in monolith form than by those in microsphere form. However, gravimetric and volumetric capacitances were higher in the activated carbon xerogel microspheres than in the carbon xerogel monoliths.
■
ASSOCIATED CONTENT
S Supporting Information *
Figure S1. Relationship between the chronopotentiometric capacitance and the micropore texture: Total and narrower micropore volumes and their mean sizes. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Tel.: +34-958 243 323. Fax: +34958 248 526. Present Address
§ ́ Z.Z.-B.: Facultad de Ingenieriá Quimica, Universidad Pontif́ Colombia. icia Bolivariana, 050031 Medellin,
Figure 6. (a) CVs at 0.5 mV/s and (b) CPs at 0.125 A/g in 1 M H2SO4 for samples: MCXA (solid green square) and MCXB (solid blue circle).
Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.
rectangular and triangular shapes, respectively; hence, these samples behave as ideal EDL capacitors. Results obtained from these curves are compiled in Table 4. CCP values of the carbon xerogel monoliths are higher than those of the carbon xerogel microspheres because they do not have diffusional limitations in the micropores and have a more developed mesoporosity. However, CCP values were lower in the carbon xerogel monoliths than in the activated carbon xerogel microspheres due to the higher surface area and total micropore volume W0(N2) of the latter. VCCP values of the monoliths are also much lower than those of the activated carbon xerogel microspheres, attributable to a lower CCP (MCXA) or ρ (MCXB) value. The higher volumetric capacitance of the activated carbon xerogel microspheres could make them more appropriate than the carbon xerogel monoliths for small-volume device applications.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This research was financed by the Junta de Andalucia.́ Z.Z.-B. acknowledges a predoctoral fellowship from COLCIENCIAS, Colombia. J.d.V. acknowledges the financial support received from MICINN MAT 2010-15101, P10-RNM-6630, and P11FQM-7074 projects (Spain).
■
REFERENCES
(1) Fairén-Jiménez, D.; Carrasco-Marín, F.; Moreno-Castilla, C. Adsorption of Benzene, Toluene and Xylenes on Monolithic Carbon Aerogels from Dry Air Flows. Langmuir 2007, 23, 10095−10101. (2) Moreno-Castilla, C. In Carbon Materials for Catalysis: Carbon gels in catalysis; Serp, P., Figueiredo, J. L., Eds.; John Wiley & Sons: New York, 2009; pp 373−400. (3) Zhang, L. L.; Zhao, X. S. Carbon-Based Materials as Supercapacitor Electrodes. Chem. Soc. Rev. 2009, 38, 2520−2531. (4) Pekala, R. W. Organic Aerogels from the Polycondensation of Resorcinol with Formaldehyde. J. Mater. Sci. 1989, 24, 3221−3227. (5) Inagaki, M.; Kaneko, K.; Nishizawa, T. Nanocarbons. Recent Research in Japan. Carbon 2004, 42, 1401−1417. (6) El Khatat, A. M.; Al Muhtaseb, S. A. Advances in Tailoring Resorcinol-Formaldehyde Organic and Carbon Gels. Adv. Mater. 2011, 23, 2887−2903. (7) de Vicente, J. In Structural and Functional Properties of Colloidal Systems: Rheological Models for Structured Fluids; Hidalgo-Alvarez, R., Ed.; CRC Press: Boca Raton, FL, 2010; pp 235−260.
4. CONCLUSIONS Increases in the molar dilution ratio diminished the apparent activation energy of the sol−gel transition in the temperature range between 40 and 50 °C. The surface of the organic xerogel microspheres was very clean except for the postcarbonization adherence of small fibrous-like particles that disappeared after steam-activation. Higher dilution ratios increased the surface area of carbon xerogel microspheres obtained at G′ = G″. However, the dilution ratio had virtually no effect on the surface area of microspheres prepared at a similar |η*| value. The microporosity markedly increased with higher degree of steam-activation. The surface area of the carbon xerogel monoliths augmented at higher molar dilution ratios. However, 6172
dx.doi.org/10.1021/la4007422 | Langmuir 2013, 29, 6166−6173
Langmuir
Article
(8) Job, N.; Panariello, F.; Crine, M.; Pirard, J. P.; Léonard, A. Rheological Determination of the Sol-Gel Transition During the Aqueous Synthesis of Resorcinol-Formaldehyde Resins. Colloids Surf,. A 2007, 293, 224−228. (9) Zapata-Benabithe, Z.; de Vicente, J.; Carrasco-Marín, F.; MorenoCastilla, C. Importance of the Rheological Properties of ResorcinolFormaldehyde Sols in the Preparation of Cu-Doped Organic and Carbon Xerogel Microspheres. Carbon 2013, 53, 402−404. (10) Zapata-Benabithe, Z.; de Vicente, J.; Carrasco-Marín, F.; Moreno-Castilla, C. Synthesis, Surface Charactersitics, and Electrochemical Capacitance of Cu-Doped Carbon Xerogel Microspheres. Carbon 2013, 55, 260−268. (11) Horikawa, T.; Hayashi, J.; Muroyama, K. Size Control and Characterization of Spherical Carbon Aerogel Particles from Resorcinol-Formaldehyde Resin. Carbon 2004, 42, 169−175. (12) Yamamoto, T.; Endo, A.; Ohmori, T.; Nakaiwa, M.; Mukai, S. R.; Tamon, H. Effect of Drying Method on Gas Adsorption Characteristics of Carbon Gel Microspheres. Drying Technol. 2005, 23, 2119−2129. (13) Menéndez, J. A.; Juárez-Pérez, E. J.; Ruisánchez, E.; Calvo, E. G.; Arenillas, A. A Microwave-Based Method for Synthesis of Carbon Xerogel Spheres. Carbon 2012, 50, 3555−60. (14) Chaichanawong, J.; Yamamoto, T.; Kim, S.-I.; Ohmori, T. Preparation and Characterization of Nickel-Modified Carbon Cryogel Beads with Uniform Particle Size. J. Non-Cryst. Solids 2009, 355, 1605−12. (15) Wang, X.; Wang, X.; Liu, L.; Bai, L.; An, H.; Zheng, L.; Yi, L. Preparation and Characterization of Carbon Aerogel Microspheres by an Inverse Emulsion Polymerization. J. Non-Cryst. Solids 2011, 357, 793−797. (16) Ponton, A.; Barboux-Doeuff, S.; Sanchez, C. Physico-Chemical Control of Sol−Gel Transition of Titaniumalkoxide-Based Materials Studied by Rheology. J. Non-Cryst. Solids 2005, 351, 45−53. (17) Winter, H. H.; Chambon, F. Analysis of Linear Viscoelasticity of a Crosslinking Polymer at the Gel Point. J. Rheol. 1986, 30, 367−382. (18) Raghavan, S.; Chen, L. A.; McDowell, C.; Khan, S. A.; Hwang, R.; White, S. Rheological Study of Crosslinking and Gelation in Chlorobutyl Elastomer Systems. Polymer 1996, 37, 5869−5875. (19) Pekala, R. W.; Alviso, C. T.; Kong, F. M.; Hulsey, S. S. Aerogels Derived from Multifunctional Organic Monomers. J. Non-Cryst. Solids 1992, 145, 90−98. (20) Yamamoto, T.; Mukai, S. R.; Endo, A.; Nakaiwa, M.; Tamon, H. Interpretation of Structure Formation During the Sol-Gel Transition of a Resorcinol-Formaldehyde Solution by Population Balance. J. Colloid Interfce Sci 2003, 264, 532−537. (21) Horikawa, T.; Ono, Y.; Hayashi, J.; Muroyama, K. Influence of Surface-Active Agents on Pore Characteristics of the Generated Spherical Resorcinol-Formaldehyde Based Carbon Aerogels. Carbon 2004, 42, 2683−2689. (22) Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.; Pierotti, R. A.; Rouquerol, J.; Siemieniewska, T. Pure Appl. Chem. 1985, 57, 603−619. (23) Cazorla-Amorós, D.; Alcañiz-Monge, J.; De la Casa-Lillo, M. A.; Linares-Solano, A. CO2 as an Adsorptive to Characterize Carbon Molecular Sieves and Activated Carbons. Langmuir 1998, 14, 4589− 4596. (24) Zapata-Benabithe, Z.; Carraco-Marín, F.; Moreno-Castilla, C. Electrochemical Performance of Cu- and Ag-Doped Carbon Aerogels. Mater. Chem. Phys. 2013, 138, 870−876. (25) Desimoni, E.; Casella, G. I.; Salvi, A. M. XPS/XAES Study of Carbon Fibres During Thermal Annealing Under UHV Conditions. Carbon 1992, 30, 521−526. (26) Raymundo-Piñero, E.; Béguin, F. In Activated Carbon Surfaces in Environmental Remediation: Application of Nanotextured Carbons for Supercapacitors and Hydrogen Storage; Bandosz, T. J., Ed.; Elsevier: Oxford,UK, 2006; pp 293−343. (27) Salitra, G.; Soffer, A.; Eliad, L.; Cohen, Y.; Aurbach, D. Carbon Electrodes for Double-Layer Capacitors - I. Relations Between Ion and Pore Dimensions. J. Electrochem. Soc. 2000, 147, 2486−2493.
(28) Barbieri, O.; Hahn, M.; Herzog, A.; Kötz, R. Capacitance Limits of High Surface Area Activated Carbons for Double Layer Capacitors. Carbon 2005, 43, 1303−1310. (29) Kinoshita, K. Carbon: Electrochemical and Physicochemical Properties; John Wiley: New York, 1988. (30) Kotz, R.; Carlen, M. Principles and Applications of Electrochemical Capacitors. Electrochim. Acta 2000, 45, 2483−2498.
6173
dx.doi.org/10.1021/la4007422 | Langmuir 2013, 29, 6166−6173
