Article pubs.acs.org/Langmuir
Activated Carbon Fibers with a High Heteroatom Content by Chemical Activation of PBO with Phosphoric Acid M. B. Vázquez-Santos,* F. Suárez-García, A. Martínez-Alonso, and J. M. D. Tascón Instituto Nacional del Carbón, INCAR-CSIC, Apartado 73, 33080 Oviedo, Spain S Supporting Information *
ABSTRACT: The preparation of activated carbon fibers (ACFs) by phosphoric acid activation of poly(p-phenylene benzobisoxazole) (PBO) fibers was studied, with particular attention to the effects of impregnation ratio and carbonization temperature on porous texture. Phosphoric acid has a strong effect on PBO degradation, lowering the temperature range at which the decomposition takes place and changing the number of mass loss steps. Chemical analysis results indicated that activation with phosphoric acid increases the concentration of oxygenated surface groups; the resulting materials also exhibiting high nitrogen content. ACFs are obtained with extremely high yields; they have well-developed porosity restricted to the micropore and narrow mesopore range and with a significant concentration of phosphorus incorporated homogeneously in the form of functional groups. An increase in the impregnation ratio leads to increases in both pore volume and pore size, maximum values of surface area (1250 m2/g) and total pore volume (0.67 cm3/g) being attained at the highest impregnation ratio (210 wt % H3PO4) and lowest activation temperature (650 °C) used; the corresponding yield was as large as 83 wt %. The obtained surface areas and pore volumes were higher than those achieved in previous works by physical activation with CO2 of PBO chars. phenylene isophtalamide) (PMIA)16 and poly(p-phenylene terephthalamide) (PPTA).17 The ACFs derived from these high-crystallinity polyaramids exhibited outstandingly homogeneous pore size distributions (PSDs),18 which made them attractive for demanding applications such as molecular sieves. Moreover, in the case of PPTA, phosphoric acid activation has led to ACFs not only with reasonably narrow PSDs, but also with an abundant (and simultaneous) occurrence of heteroatoms bound to the carbon matrix.17 The latter included not only O and P, but also N, which is present in the parent polymer and is partially retained in the derived carbons. This avoids the need for further treatments to introduce nitrogen using reagents such as ammonia,19,20 urea,21,22 nitric acid,23 or pyridine.24 More recently, in an attempt to widen the scope of available feedstock polymers for ACF production, we have investigated the preparation of porous carbons from poly(p-phenylene benzobisoxazole) (henceforth abbreviated as PBO). Physical activation with CO2 of PBO chars has led to ACFs with BET surface areas below 1000 m2/g, even at the highest burn-offs (BOs) studied.25 The obtained adsorbents were essentially supermicroporous with a certain contribution from narrow mesopores, and their PSDs became increasingly heterogeneous as the BO increased. Further work26 has shown that PBO
1. INTRODUCTION Phosphoric acid is a common chemical activation agent for the preparation of activated carbons (ACs). Besides contributing to develop a suitable pore structure, the use of H3PO4 leads to the insertion of different P- and O-containing functional groups in the resulting ACs. These functionalities have provided interesting properties for such applications as energy storage in electric double-layer capacitors,1 removal of metal ions from water,2 or removal of organic molecules from liquid fuels,3 to name just a few.4 The direct introduction of heteroatoms in carbons that takes place upon chemical activation with H3PO4, if exploited adequately, offers the potential advantage of avoiding the need for a further functionalization treatment, which most often requires exposure to oxidizing agents such as nitric acid,5−7 hydrogen peroxide,5,8 oxygen plasma,9,10 ammonium persulfate,11 or sulfuric acid12 (the surface chemical modifications of carbons connected in particular with water treatment have been recently reviewed in detail by Rivera-Utrilla et al.13). Note that oxidative treatments are also efficient to improve the performance of advanced adsorbents such as ordered mesoporous carbons.14 Apart from these postsynthesis oxidative treatments, “one-step” combined synthesis and functionalization approaches have been scarce to date and applied under rather uncommon conditions.15 In previous works from our laboratory, activated carbon fibers (ACFs), i.e., ACs with a fibrous morphology, were obtained by H3PO4 activation of polymers such as poly(m© 2012 American Chemical Society
Received: January 13, 2012 Revised: March 2, 2012 Published: March 8, 2012 5850
dx.doi.org/10.1021/la300189v | Langmuir 2012, 28, 5850−5860
Langmuir
Article
Samples are designated as ASXp_T, where AS means PBO in AS variety, Xp is the impregnation ratio (in wt %) and T is the activation temperature (in °C). 2.2. Methods. Thermogravimetric (TG) and derivative thermogravimetric (DTG) analysis were carried out using a CI Electronics thermobalance. A starting mass of 30 mg of impregnated samples was placed in a crucible made from quartz. The experiments were carried out at a heating rate of 10 °C/min using an argon (minimum purity: 99.9990 vol %) flow fixed at 50 STP cm3/min. Elemental analyses were made in a LECO CHNS-932 microanalyzer, with a LECO VTF-900 accessory for direct oxygen determination. Statistical errors were calculated from data derived from at least three replicated measurements. The porous texture was analyzed by physical adsorption of N2 at −196 °C and CO2 at 0 °C, measured in volumetric adsorption apparatuses, ASAP 2010 (Micromeritics) and NOVA 4200e (Quantachrome), respectively. From the N2 data, the surface area, SBET, was determined according to the BET equation, and the total pore volume, Vt, was calculated from nitrogen uptake at 0.975 relative pressure, assuming it to be in the liquid state (Gurvitsch rule). The mesopore volume, Vmp, was calculated as the difference between the total pore volume and the Dubinin−Radushkevich (DR) micropore volume obtained from N2 adsorption, Vμp (DR). The revision of the αS method for high-resolution isotherms proposed by Kaneko and coworkers32,33 was used to determine the micropore volume, Vμp (αS), the ultramicropore volume, Vuμp (αS), and the external surface area, Sext (αS); Spheron 6 carbon black was used as reference. Pore size distributions were obtained by applying the nonlocal density functional theory (NLDFT) using the DFT Plus software implemented by Micromeritics.34,35 The CO2 isotherms were analyzed by means of the DR equation, to obtain the corresponding DR micropore volume. The different volumes and surfaces were calculated by fitting of the results to straight lines within the relative pressure ranges characteristic of each of the methods described previously. The corresponding errors were calculated using the least-squares method. A field emission scanning electron microscope (Quanta FEG-650FEI) was used to characterize the morphology of the fibers. Samples were attached to an aluminum tap using conducting double-sided adhesive tape. The AS0_650 sample was sputter-coated with a layer of Ir of approximately 10 nm thickness to reduce charging effect; the rest of the samples were examined without any coating. An AMETEKEDAS microanalyzer with Apollo X detector was used for elemental microanalysis. For this elemental microanalysis, statistical errors were calculated from at least three replicated measurements. X-ray diffractograms were taken in a Bruker D8 Advance diffractometer using Cu Kα (λ = 0.15406 nm) radiation with a step size of 0.02° and a step time of 1 s. The goniometer has a reproducibility of ±0.0001°, and the smallest measurable angular step size is 0.0001°. The spacing between layers, d002, and the crystallite size along the c-axis, Lc, were obtained from the (002) peak, and the crystallite size along the a-axis, La, from the (10) band. The d002 parameter was calculated from Bragg equation. Lc and La were calculated from the Scherrer equation assuming values of 0.9 and 1.84, respectively, for the shape factor, K. The errors associated with d002, Lc, and La were calculated taking into account the intrinsic error to the measurement and using the error propagation method. The calculated errors are small, especially in the case of the interlayer spacing, since this is only affected by the measurement of the angle corresponding to the maximum intensity of the (002) peak.
impregnation with small amounts (5−15 wt %) of H3PO4 prior to physical activation with CO2 restricted porosity development to the range of micropores and narrow mesopores (above 10 nm, there was practically no contribution to the pore volume). Nevertheless, even in the latter case, the obtained ACFs exhibited more heterogeneous PSDs than equivalent materials prepared from PMIA, a polymer with a much lower degree of crystallinity than PBO. Let us remember here that PBO exhibits the highest thermal stability, tensile strength, and modulus among all known polymeric commercial fibers,27 all these properties being a consequence of the high degree of crystalline order of this polymer. The main aim of the present work is to explore chemical activation of PBO with H3PO4 as a possible way to better control the porous texture of PBO-derived ACFs while introducing abundant O-, N-, and P-containing functionalities at their surface (as with PPTA, the presence of nitrogen in the parent polymer is expected to result in a certain retention of this element). To this end, PBO pyrolysis was carried out in the presence of different amounts of H3PO4, and the effects of impregnation ratio and activation temperature on the porous texture and chemical composition of the resulting materials were investigated. Let us mention as well that the growing application of PBO in advanced composites is expected to generate an increasing amount of waste that will be difficult to process due to the high thermal and chemical stability of the polymer. The simplest way to dispose of this PBO-derived waste is combustion, but it produces small amounts of toxic gases such as hydrogen cyanide, sulfur oxides, and nitrogen oxides28,29 besides carbon oxides and water. Otherwise, the storage of PBO wastes in dumps does not seem to be a good alternative, since PBO is sensitive to light (the polymer is always protected by outer covers). The presence of acids30 and/or humidity can accelerate this degradation.31 For these reasons, use as feedstock for carbon adsorbents can provide an attractive and profitable recycling route for PBO-derived waste, thus adding an environmental benefit to ACF production from PBO.
2. EXPERIMENTAL SECTION 2.1. Preparation of ACFs. The starting material was PBO in the as-spun (AS) variety, provided by Toyobo Japan. Impregnation with the chemical activating agent was carried out in a rotary evaporator using 10 g of PBO and 200 cm3 of solutions of H3PO4 at different concentrations in order to vary the impregnation ratio. The temperature was increased progressively until complete evaporation of water; this process lasted for 6 h. Then, the samples were dried overnight at 110 °C in a vacuum furnace. The impregnation ratio (Xp) was calculated as the quotient between the mass increase following impregnation and the initial mass of PBO. All the activations were done in a ceramic crucible, which was introduced in an horizontal furnace, using 3 g of phosphoric acidimpregnated PBO, a flow of Ar (minimum purity: 99.9990 vol %) of 400 STP cm3/min, and a heating rate of 10 °C/min until the maximum activation temperature was reached. Once this temperature was attained, the samples were cooled down to room temperature. The effect of the activation temperature was studied by preparing materials at temperatures between 650 and 1000 °C. The effect of the impregnation ratio was studied by preparing materials with Xp values of 70, 145, and 210 wt %. To remove residues from the H3PO4 additive, the samples were fully washed in a Soxhlet extractor with hot distilled water until the conductivity in the washing liquids was
