A Microscopic View of Physical and Chemical Activation in the

9730

Langmuir 2006, 22, 9730-9739

A Microscopic View of Physical and Chemical Activation in the Synthesis of Porous Carbons J. I. Paredes,* F. Sua´rez-Garcı´a, A. Martı´nez-Alonso, and J. M. D. Tasco´n Instituto Nacional del Carbo´ n, CSIC, Apartado 73, 33080 OViedo, Spain ReceiVed May 11, 2006. In Final Form: July 26, 2006 The nanostructure and porosity of activated carbon fibers (ACFs) prepared by physical activation with CO2 and by chemical activation with H3PO4 of the highly ordered polymer poly(m-phenylene isophthalamide) have been investigated and compared by means of scanning tunneling microscopy (STM), scanning electron microscopy (SEM), and gas adsorption measurements. In general terms, both types of activation led to porous carbons with similar nanometer-scale structure, which consisted of relatively ordered and homogeneous arrays of platelets below 10-nm wide, the porous structure being mainly comprised by the network of narrow trenches present between neighboring platelets. This similarity was attributed to the influence of the crystalline structure of the polymeric precursor, which should favor a homogeneous, uniform transformation of the polymer into the final carbon material. Such influence was only lost in chemical activation with the use of very large amounts of activating agent. A comparison of samples before and after physical activation allowed a direct identification of the local areas where gasification (activation) took place. For chemical activation, the STM measurements suggested that porosity was developed at a lower temperature than the highly cross-linked nanographitic structure of the final ACF. This result was discussed in terms of the thermal transformation mechanism of the precursor polymer into a carbonaceous solid in the presence of H3PO4.

1. Introduction Because of their high surface area, large pore volume, chemical inertness, and good mechanical stability, porous carbons constitute a class of materials with important applications in fields such as gas adsorption and separation, ion exchange, or catalysis.1,2 Porous carbons are usually prepared from organic precursors of natural or synthetic origin, which are carbonized and then activated through a mild gasification under CO2 or H2O (physical activation) or by carbonization of the precursor previously mixed with an appropriate activating agent, such as KOH or H3PO4 (chemical activation).1 These activated carbons consist of complex, intricate assemblies of distorted graphitelike nanocrystallites, with the consequent disadvantage that uniformity in pore size and shape is typically difficult to control in such carbons.3 In this respect, activated carbon fibers (ACFs) have emerged in recent years as an attractive variant of activated carbon, since they possess a relatively uniform porosity (mainly microporosity) in comparison with that of more conventional (i.e., granular) activated carbons.4 For this reason, ACFs are being actively investigated not only for different practical purposes, which include water and air purification,5,6 catalyst supports,7,8 or gas storage,9,10 but also from a fundamental science perspective, for * Author to whom correspondence should be addressed. Telephone: (+34) 985 11 90 90; fax: (+34) 985 29 76 62; e-mail: [email protected]. (1) Handbook of Porous Solids; Schu¨th, F., Sing, K. S. W., Weitkamp, J., Eds.; Wiley-VCH: Weinheim, Germany, 2002. (2) Bandosz, T. J.; Briggs, M. J.; Gubbins, K. E.; Hattori, Y.; Iiyama, T.; Kaneko, K.; Pikunic, J.; Thomson, K. T. Chem. Phys. Carbon 2003, 28, 41-228. (3) Kyotani, T. Carbon 2000, 38, 269-286. (4) Inagaki, M. New Carbons. Control of Structure and Function; Elsevier: Amsterdam, 2000; Chapter 5. (5) Monneyron, P.; Faur-Brasquet, C.; Sakoda, A.; Suzuki, M.; Le Cloirec, P. Langmuir 2002, 18, 5163-5169. (6) Le Leuch, L. M.; Subrenat, A.; Le Cloirec, P. Langmuir 2003, 19, 1086910877. (7) Bulushev, D. A.; Yuranov, I.; Suvorova, E. I.; Buffat, P. A.; Kiwi-Minsker, L. J. Catal. 2004, 224, 8-17. (8) Mikkola, J.-P.; Aumo, J.; Murzin, D. Y.; Salmi, T. Catal. Today 2005, 105, 325-330. (9) Alcan˜iz-Monge, J.; de la Casa-Lillo, M. A.; Cazorla-Amoro´s, D.; LinaresSolano, A. Carbon 1997, 35, 291-297.

instance, in studies of molecular confinement in nanometersized spaces.11,12 To take full advantage of their potential, a good knowledge of the physicochemical properties of ACFs is required. Particularly interesting is the investigation of the effect that the different activation methods have on the nanostructure and porosity of the resulting ACFs. Over the years, several techniques have been employed for the structural characterization of ACFs and more generally of porous carbons, including X-ray diffraction (XRD),2 small-angle X-ray scattering (SAXS),13-15 gas/vapor adsorption,16-18 scanning and transmission electron microscopies (SEM/ TEM),19-21 or scanning probe microscopies, mainly scanning tunneling microscopy (STM).13,22-25 STM is especially useful to image in real space the surface nanostructure and local porosity (10) de la Casa-Lillo, M. A.; Lamari-Darkrim, F.; Cazorla-Amoro´s, D.; LinaresSolano, A. J. Phys. Chem. B 2002, 106, 10930-10934. (11) Radhakrishnan, R.; Gubbins, K. E.; Sliwinska-Bartkowiak, M. Phys. ReV. Lett. 2002, 89, 076101. (12) Ohba, T.; Kanoh, H.; Kaneko, K. Chem. Eur. J. 2005, 11, 4890-4894. (13) Bo´ta, A.; La´szlo´, K.; Nagy, L. G.; Copitzky, T. Langmuir 1997, 13, 6502-6509. (14) Ehrburger-Dolle, F.; Morfin, I.; Geissler, E.; Bley, F.; Livet, F.; VixGuterl, C.; Saadallah, S.; Parmentier, J.; Reda, M.; Patarin, J.; Iliescu, M.; Werckmann, J. Langmuir 2003, 19, 4303-4308. (15) La´szlo´, K.; Marthi, K.; Rochas, C.; Ehrburger-Dolle, F.; Livet, F.; Geissler, E. Langmuir 2004, 20, 1321-1328. (16) Bradley, R. H.; Rand, B. J. Colloid Interface Sci. 1995, 169, 168-176. (17) Rodrı´guez-Reinoso, F.; Molina-Sabio, M. AdV. Colloid Interface Sci. 1998, 76-77, 271-294. (18) Nguyen, T. X.; Bhatia, S. K. Langmuir 2004, 20, 3532-3535. (19) Oshida, K.; Kogiso, K.; Matsubayashi, K.; Takeuchi, K.; Kobayashi, S.; Endo, M.; Dresselhaus, M. S.; Dresselhaus, G. J. Mater. Res. 1995, 10, 25072517. (20) Huang, Z.-H.; Kang, F.; Huang, W. L.; Yang, J.-B.; Liang, K.-M.; Cui, M.-L.; Cheng, Z. J. Colloid Interface Sci. 2002, 249, 453-457. (21) Lillo-Ro´denas, M. A.; Cazorla-Amoro´s, D.; Linares-Solano, A.; Be´guin, F.; Clinard, C.; Rouzaud, J. N. Carbon 2004, 42, 1305-1310. (22) Daley, M. A.; Tandon, D.; Economy, J.; Hippo, E. J. Carbon 1996, 34, 1191-1200. (23) Vignal, V.; Morawski, A. W.; Konno, H.; Inagaki, M. J. Mater. Res. 1999, 14, 1102-1112. (24) Brasquet, C.; Rousseau, B.; Estrade-Szwarckopf, H.; Le Cloirec, P. Carbon 2000, 38, 407-422. (25) Paredes, J. I.; Martı´nez-Alonso, A.; Tasco´n, J. M. D. Langmuir 2001, 17, 474-480.

10.1021/la061330l CCC: $33.50 © 2006 American Chemical Society Published on Web 10/06/2006

ActiVation in the Synthesis of Porous Carbons

of carbons but has not been employed so far to study different physically and chemically activated carbons obtained from the same precursor, so that the effect of the specific type of activation can be directly visualized and compared at the nanoscale. Indeed, to the best of our knowledge, this kind of information has never been reported in the literature. Therefore, we present here an STM study, complemented by gas adsorption and SEM, of the nanostructure and porosity of different ACFs prepared by physical and chemical activation of high crystallinity polyaramid fibers. These ACFs are especially interesting because of their very homogeneous pore texture, which leads to distinctive adsorption properties. The porosity characteristics and potential uses of these polyaramid-derived ACFs have been very recently reviewed.26 Promising applications include highly efficient air separation (carbon molecular sieves)27 and electrochemical energy storage (electrodes for supercapacitors).28 2. Experimental The ACFs studied in this work were prepared using the polyaramid fiber referred to as Nomex [poly(m-phenylene isophthalamide)] as a precursor. ACFs were produced both by physical activation with CO2 of fresh Nomex or Nomex preimpregnated with a small amount of H3PO4 and by chemical activation with H3PO4. Details of the production process of the different ACFs can be found elsewhere,29-31 so only a brief description of the preparations will be given here. Impregnation of Nomex with a small amount of H3PO4 for the physical activation, or with a larger amount for the chemical activation, was accomplished in a rotary evaporator, using an aqueous solution of this agent at an appropriate concentration to attain the desired impregnation ratio [defined as the weight gain after impregnation relative to the initial mass of Nomex, i.e., (g H3PO4/g Nomex) × 100]. Impregnated samples are denoted as NPx, where x is the impregnation ratio (x ) 9, 75, or 140 for the samples used in this work). Nonimpregnated Nomex is simply denoted as N. Pyrolysis/physical activation of fresh Nomex (N) or Nomex preimpregnated with a small amount of H3PO4 (NP9) was carried out in a vertical quartz reactor. Pyrolysis was achieved at a heating rate of 10 °C min-1 under an Ar flow up to a temperature of 850 °C. Samples denoted as N-0 and NP9-0 (chars) were obtained at this stage. Then, the temperature was lowered to 800 °C and physical activation with CO2 was performed for a certain time interval to attain a given burnoff degree. Physically activated samples are denoted as N-bo or NP9-bo, where bo is the burnoff degree (%). Chemical activation was accomplished in a U-shaped quartz reactor under Ar flow at a heating rate of 10 °C min-1 up to the desired temperature. Once this temperature was attained, the samples were rapidly cooled to room temperature under the same Ar atmosphere. The chemically activated samples thus prepared are designated as NP75-t or NP140-t, where t is the activation temperature (°C). Likewise, to investigate the effect of chemical activation time on the nanostructure of the ACFs, additional samples were prepared by keeping NP75-t specimens at the activation temperature t for certain time intervals, ranging from half an hour to several hours. These samples are denoted as NP75-t-time interval (h). All the samples pyrolyzed/activated in the presence of H3PO4 (i.e., NP9-bo, NP75-t, and NP140-t) were washed with Milli Q water (26) Villar-Rodil, S.; Sua´rez-Garcı´a, F.; Paredes, J. I.; Martı´nez-Alonso, A.; Tasco´n, J. M. D. Chem. Mater. 2005, 17, 5893-5908. (27) Villar-Rodil, S.; Denoyel, R.; Rouquerol, J.; Martı´nez-Alonso, A.; Tasco´n, J. M. D. Chem. Mater. 2002, 14, 4328-4333. (28) Leitner, K.; Lerf, A.; Winter, M.; Besenhard, J. O.; Villar-Rodil, S.; Sua´rezGarcı´a, F.; Martı´nez-Alonso, A.; Tasco´n, J. M. D. J. Power Sources 2006, 153, 419-423. (29) Sua´rez-Garcı´a, F.; Paredes, J. I.; Martı´nez-Alonso, A.; Tasco´n, J. M. D. J. Mater. Chem. 2002, 12, 3213-3219. (30) Sua´rez-Garcı´a, F.; Martı´nez-Alonso, A.; Tasco´n, J. M. D. Carbon 2004, 42, 1419-1426. (31) Sua´rez-Garcı´a, F.; Martı´nez-Alonso, A.; Tasco´n, J. M. D. Microporous Mesoporous Mater. 2004, 75, 73-80.

Langmuir, Vol. 22, No. 23, 2006 9731 in a Soxhlet extractor to remove the excess of H3PO4 and its decomposition products. Washing was done repeatedly until the conductivity of the washing liquid was below 3 µS cm-1. Finally, the resulting carbon materials were vacuum-dried overnight at 110 °C. A general characterization of the samples was carried out by N2 (-196 °C) and CO2 (0 °C) physisorption, as well as by SEM. N2 adsorption isotherms were obtained with an automatic volumetric adsorption apparatus (ASAP 2010, from Micromeritics), whereas CO2 adsorption isotherms were measured in a semiautomatic volumetric adsorption apparatus (NOVA 1200, from Quantachrome). SEM images were obtained both with a Zeiss DSM 942 and a JEOL JSM-840 instrument. STM studies were performed under ambient conditions with a Nanoscope IIIa Multimode apparatus (Veeco Instruments). The ACF samples were mounted onto STM sample holders by means of doublesided carbon adhesive tape. Images were recorded in the constant current mode (variable height) using mechanically prepared Pt/Ir (80/20) tips. Typical tunneling parameters employed were 0.5 nA and 500 mV for the tunneling current and bias voltage, respectively, except for the atomic scale images, which were recorded at 5-8 nA (tunneling current) and at 50-100 mV (bias voltage). Unless otherwise stated, the images presented here were obtained under these conditions. However, it was observed that such tunneling parameters could be varied to a large extent for most samples without significant changes in the features observed in the images. The only exception to this rule was some chemically activated samples prepared at relatively low temperatures. This point will be addressed in detail below. To check for the reproducibility of the images and to verify that they were representative of the studied samples, so that meaningful comparisons between samples could be made, the following imaging protocol was implemented. For each ACF sample, a given fiber was first selected under an optical microscope coupled to the Multimode apparatus and was imaged by STM at several different locations with a new, previously unused tip. Then, to ensure that the images were not affected by tip artifacts, the same fiber was imaged with several additional new tips. After this point was checked, these tips were used to image many other different fibers of the same sample. In general terms, for a given sample, good reproducibility was obtained with different tips on the same fiber and for the same tip on different fibers, indicating that the nanostructure of the ACF sample was rather homogeneous.

3. Results and Discussion 3.1. Porous Texture by N2 and CO2 Adsorption. Figure 1 shows the N2 adsorption isotherms (a) and corresponding pore size distributions (PSDs) derived by the nonlocal density functional theory (DFT) method (b) of the different ACF samples studied in this work. The main textural parameters of all the samples, as deduced from the N2 and CO2 adsorption data, are shown in Table 1. For the samples chemically activated with an impregnation ratio of 75, that is, for samples NP75-t, the N2 adsorption isotherms were of type Ia, indicating a very narrow porosity.32 By contrast, the ACFs prepared by chemical activation with an impregnation ratio of 140 (NP140-700) and the physically activated materials (N-89 and NP9-72) exhibited type Ib isotherms and adsorbed considerably larger amounts of N2 compared to the NP75-t carbons, which reflects the presence of a significant number of wide micropores and small mesopores in the former materials.32 These general observations are confirmed by inspection of the corresponding PSDs (Figure 1b). The samples prepared by physical activation at high burnoffs as well as sample NP140-700 displayed a range of pore sizes going from ultramicropores (∼0.5 nm) to small mesopores (∼3-5 nm). On the (32) Rouquerol, F.; Rouquerol, J.; Sing, K. S. W. Adsorption by Powders and Porous Solids. Principles, Methods and Applications; Academic Press: San Diego, CA, 1999.

9732 Langmuir, Vol. 22, No. 23, 2006

Paredes et al.

Figure 1. Nitrogen adsorption isotherms (a) and corresponding pore size distributions obtained by the nonlocal density functional method (b) of the studied activated carbon fibers. In a, open symbols denote adsorption and full symbols denote desorption. Table 1. Textural Parameters of the Different ACF Samples Deduced from N2 (-196 °C) and CO2 (0 °C) Adsorptiona sample N-0 N-89 NP9-0 NP9-72 NP75-600 NP75-700 NP75-950 NP140-700

SBET Vt (0.975, N2) Vµp (RS) Vµp (DR, N2) Vµp (DR, CO2) 2804

1.44

1.35

1.22

2303 951 1021 1130 1628

1.09 0.39 0.41 0.46 0.74

1.00 0.37 0.39 0.44 0.68

1.02 0.38 0.40 0.42 0.67

0.18 0.36 0.20 0.38 0.35 0.36 0.36 0.41

a

SBET, specific surface area according to the BET method. Vt (0.975, N2), total pore volume calculated from the N2 uptake at a relative pressure of 0.975. Vµp (RS), micropore volume calculated by the RS method. Vµp (DR, N2), micropore volume deduced from the N2 isotherm using the Dubinin-Radushkevich (DR) equation. Vµp (DR, CO2), micropore volume from the CO2 isotherm using the DR equation.

other hand, the chemically activated fibers NP75-t revealed a porosity strictly confined to the micropore range (