Uniformly Carbon-Covered Alumina and Its Surface Characteristics

UV Raman spectra indicated that the carbon species formed were probably conjugated olefinic or polycyclic aromatic hydrocarbons, which can be consider...
0 downloads 0 Views 148KB Size
Download PDF
5040

Langmuir 2005, 21, 5040-5046

Uniformly Carbon-Covered Alumina and Its Surface Characteristics L. Lin, W. Lin, Y. X. Zhu,* B. Y. Zhao, and Y. C. Xie State Key Laboratory for Structural Chemistry of Unstable and Stable Species, College of Chemistry and Molecular Engineering, Peking University, 100871 Beijing, China

G. Q. Jia and C. Li State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 116023 Dalian, China Received November 25, 2004. In Final Form: March 21, 2005 Uniformly carbon-covered alumina (CCA) was prepared via the carbonization of sucrose highly dispersed on the alumina surface. The CCA samples were characterized by XRD, XPS, DTA-TG, UV Raman, nitrogen adsorption experiments at 77 K, and rhodamine B (RB) adsorption in aqueous media. UV Raman spectra indicated that the carbon species formed were probably conjugated olefinic or polycyclic aromatic hydrocarbons, which can be considered molecular subunits of a graphitic plane. The N2 adsorption isotherms, pore size distributions, and XPS results indicated that carbon was uniformly dispersed on the alumina surface in the as-prepared CCA. The carbon coverage and number of carbon layers in CCA could be controlled by the tuning of the sucrose content in the precursor and impregnation times. RB adsorption isotherms suggested that the monolayer adsorption capacity of RB on alumina increased drastically for the sample with uniformly dispersed carbon. The as-prepared CCA possessed the texture of alumina and the surface properties of carbon or both carbon and alumina depending on the carbon coverage.

Introduction Carbon and alumina probably are the two most widely used support materials for catalyst systems. Nevertheless, both of them have drawbacks.1,2 Most carbon materials either are microporous or have poor mechanical properties. Therefore, in some reactions involving large molecules on carbon-supported metal catalysts, the metal particles are deposited in the carbon micropores, making their effect on the catalytic activity trivial. In addition, most mesoporous carbon materials are not favorable supports because of their poor crushing strength, low bulk density, or low surface area. Alumina, under certain conditions, also has its own disadvantages, such as its acidity and strong interaction with metals. To overcome these shortcomings, a new approach was developed in which the Al2O3 surface was coated with a thin layer of carbon prior to metal impregnation.1 In this way, the favorable carbon surface properties are combined with the optimal textural and mechanical features of the Al2O3 support. This synergetic effect makes carbon-covered alumina (CCA) superior to pure alumina and pure carbon in some cases.1-8 * Author to whom correspondence should be addressed. Tel.: +86 10 62751703. Fax +86 10 62751725. E-mail: [email protected]. (1) Vissers, J. P. R.; Mercx, F. P. M.; Bouwens, S. M. A. M.; de Beer, V. H. J.; Prins, R. J. Catal. 1988, 114, 291. (2) Rao, K. S. R.; Rao, P. K.; Masthan, S. K.; Kaluschnaya, L.; Shur, V. B. Appl. Catal. 1990, 62, L19. (3) Masthan, S. K.; Prasad, P. S. S.; Rao, K. S. R.; Rao, P. K. J. Mol. Catal. 1991, 67, L1. (4) Boorman, P. M.; Chong, K.; Kydd, R. A.; Lewis, J. M. J. Catal. 1991, 128, 537. (5) Shekar, S. C.; Murthy, J. K.; Rao, P. K.; Rao, K. S. R. J. Mol. Catal. A 2003, 191, 45. (6) Shekar, S. C.; Murthy, J. K.; Rao, P. K.; Rao, K. S. R.; Kemnitz, E. Appl. Catal. A 2003, 244, 39. (7) Shekar, S. C.; Murthy, J. K.; Rao, P. K.; Rao, K. S. R. Catal. Commun. 2003, 4, 39.

Along with its use in catalysis, CCA also is a carbonmineral adsorbent that represents a new type of adsorption materials containing two components: mineral and carbonaceous species.9-15 The essential novelty of the carbon-mineral adsorbents lies in their dual functions as a polar inorganic adsorbent (e.g., Al2O3, SiO2, zeolites) and a nonpolar carbon adsorbent. This enables them to adsorb both organic and inorganic substances.15 The performance of such adsorbents in adsorption process depends on their physicochemical properties. Therefore, a detailed study on these dual-functional carbon-mineral adsorbents is necessary. Currently, knowledge is still lacking for the complex adsorbent structure, morphology and topography of the carbon deposits, and their surface properties in terms of their present and future applications.15 According to the above-mentioned literature, a synergistic effect exists between alumina and the carbon phases in CCA. This is expected to improve if the carbon phase is uniformly dispersed on the alumina surface. To the best of our knowledge, CCA is usually prepared by the pyrolysis of gaseous hydrocarbons, such as acetylene, cyclohexene, and ethane, in an inert atmosphere, otherwise known as chemical vapor deposition (CVD).1-8 Another conventional process involves carbon coating via the calcination of alumina powder premixed with carbon precursors, such as poly(vinyl chloride) (PVC), poly(vinyl alcohol) (PVA), and coal tar pitch, at temperatures between (8) Reddy, G. K.; Rao, K. S. R.; Rao, P. K. Catal. Lett. 1999, 59, 157. (9) Leboda, R. Mater. Chem. Phys. 1992, 31, 243. (10) Leboda, R. Mater. Chem. Phys. 1993, 34, 123. (11) Kyotani, T.; Sonobe, N.; Tomita, A. Nature 1988, 331, 331. (12) Sonobe, N.; Kyotani, T.; Tomita, A. Carbon 1990, 28, 483. (13) Sonobe, N.; Kyotani, T.; Tomita, A. Carbon 1991, 29, 61. (14) Bandosz, T. J.; Putyera, K.; Jagiello, J.; Schwarz, J. A. Carbon 1994, 32, 659. (15) Villie´ras, F.; Leboda, R.; Charmas, B.; Bardot, F.; Ge´rard, G.; Rudzinski, W. Carbon 1998, 36, 1501.

10.1021/la047097d CCC: $30.25 © 2005 American Chemical Society Published on Web 04/21/2005

Uniformly Carbon-Covered Alumina

Langmuir, Vol. 21, No. 11, 2005 5041

Table 1. Precursor Composition, Carbon Content, and Texture of CCA Samples sample

WS:WAa

carbon content (wt%)

SBET (m2 g-1)

Vb (mL g-1)

Al2O3 CCA01 CCA02 CCA03 CCA04 CCA05 CCA06 CCA08 CCA03-2 CCA03-3

0.1:1 0.2:1 0.3:1 0.4:1 0.5:1 0.6:1 0.8:1 -

0 2.0 3.9 5.7 7.4 8.8 10.0 12.7 9.6 12.5

128 133 136 139 140 141 152 164 141 143

0.48 0.46 0.45 0.43 0.40 0.39 0.38 0.35 0.40 0.35

a Weight ratio of sucrose to alumina in the precursors. b Sample pore volume.

500 and 1000 °C in an inert atmosphere.16,17 In general, it is difficult to uniformly cover the alumina surface with carbon, partially because of the diffusion limitation of the carbon-yielding hydrocarbons.1,18 We previously reported that sucrose could be monolayer-dispersed on the surface of alumina.19 If the sucrose can be pyrolyzed without aggregation, then uniformly carbon-covered alumina (CCA) would be obtained. Besides, sucrose as a carbon source has an advantage for its green chemistry and low cost. This method was proved feasible in our recent communication.20 In the present paper, the surface properties of as-prepared CCA are further characterized and the interaction between the alumina and the carbon species is discussed in detail. Experimental Section Typical Procedure for the Synthesis of CCA. The experimental procedure to prepare CCA is as follows. First, the sucrose/γ-Al2O3 precursors were prepared by impregnation of commercial γ-Al2O3 (SBA150, Engelhard) with aqueous solutions of sucrose. The sucrose loadings in the precursors are given in Table 1. After being dried at 90 °C, the precursors were calcined at 600 °C in N2 (flow rate ) 30 mL min-1) for 0.5 h. The final products discussed here were denoted as CCA01-CCA08. To obtain uniformly CCA with a higher carbon content, CCA03 was further impregnated with an aqueous solution of sucrose (weight ratio of sucrose to CCA03, 0.3:1), to give CCA03-2. Similarly, CCA03-3 was obtained by the re-impregnation of CCA03-2 in a sucrose solution (weight ratio of sucrose to CCA03-2, 0.3:1). The sucrose carbon as reference was prepared via the pyrolysis of sucrose in the same conditions. Characterization. XRD. Phase compositions of the samples were determined by a Rigaku D/MAX-200 X-ray powder diffractometer with Ni-filtered Cu KR radiation at 40 kV and 100 mA. The amount of residual crystalline sucrose was determined by a quantitative phase analysis with the internal standard, R-Al2O3.21 The peak areas of reflections (111) and (210) of sucrose and (113) of R-Al2O3 were measured. The peak intensity ratio, Isucrose/IR-Al2O3, should be proportional to the content ratio of crystalline sucrose to R-Al2O3, as given by

Isucrose/IR-Al2O3 ) k(Xsucrose/XR-Al2O3)

(1)

where I and X are the XRD intensity and weight percent, respectively; k is a proportionality constant determined from a (16) Inagaki, M.; Hirose, Y.; Matsunaga, T.; Tsumura, T.; Toyoda, M. Carbon 2003, 41, 2619. (17) Lecea, C. S.; Linares-Solano, A.; Dı´az-Aun˜o´n, J. A.; L’Argentie`re, P. C. Carbon 2000, 38, 157. (18) Baumgart, J.; Wang, Y.; Ernst, W. R.; Carruthers, J. D. J. Catal. 1990, 126, 477. (19) Zhu, Y. X.; Pan, X. M.; Xie, Y. C. Acta Phys.-Chimi. Sin. 1999, 15, 830. (20) Lin, L.; Lin, W.; Wang, P.; Zhu, Y. X.; Zhao, B. Y.; Xie, Y. C. Acta Phys.-Chim. Sin. 2004, 20, 1179. (21) Xie, Y. C.; Tang, Y. Q. Adv. Catal. 1990, 37, 1.

sample with a known phase composition. Thus, the weight percent of the crystalline sucrose in the samples can be derived from the intensity ratio of Isucrose/IR-Al2O3 when the R-Al2O3 content is known. DTA-TG. The carbon contents in the samples were measured by a DTA-TG instrument (Dupont model 1090) from room temperature to 600 °C at a heating rate of 10 °C min-1 in an air flow of 50 mL min-1; R-Al2O3 was used as a reference. UV Raman Spectra. UV Raman spectroscopy was carried out on a T64000 triple-stage spectrometer of Jobin Yvon S. A. equipped with three 2400 grooves mm-1 gratings and a specially coated UV-sensitive CCD detector. The spectra were recorded in air using a 180° backscattering geometry on a static sample with a laser power of 1-3 mW on the sample to avoid thermal damage. Each spectrum was collected for 300 s; two spectra were averaged to increase the signal-to-noise ratio. A He-Cd laser line of 325 nm was the excitation source. XPS. The surface composition of the samples was measured on a Kratos Axis Ultra System with monochromatic Al KR X-rays (1486.6 eV) at 15 V and 15 mA (emission current) in a chamber with a base pressure of approximately 10-8 Pa. N2 Adsorption Experiments at 77 K. The Brunauer-EmmettTeller (BET) surface areas (SBET), pore volumes, and pore size distributions were calculated from the adsorption isotherms of nitrogen at 77 K on a Micromeritics ASAP 2010 volumetric adsorption system. All samples were degassed at 200 °C before BET measurements. Rhodamine B (RB) Adsorption. Adsorption experiments of the organic dye RB were conducted in a glass vessel containing the sample (10 mg) and 100 mL of aqueous solutions of RB with concentrations of (0.8, 1, 1.5, 2, 4, 6, 8, and 10) × 10-5 mol L-1. The vessels were then sealed; the mixture was stirred and kept in the dark to ensure that adsorption-desorption equilibrium was established. The RB concentration was monitored by a UV-vis spectrometer at 554 nm.

Results and Discussion Dispersion of Sucrose on the Alumina Surface. XRD results (Figure 1A) show that there is no crystalline sucrose in the precursors with low sucrose loadings, whereas the precursors with higher sucrose loadings (g0.40 g sucrose/g γ-Al2O3) exhibit crystalline sucrose peaks; these peak intensities increase with the sucrose loading. Thus, sucrose is highly dispersed on the γ-Al2O3 surface with low sucrose loadings, while crystalline sucrose appears only if its loading exceeds a threshold value. Figure 1B shows the residual crystalline content as a function of the sucrose loading. The threshold is deduced to be about 0.35 g sucrose/g alumina for the dispersion capacity of sucrose on alumina (Figure 1B). Texture Studies of CCA. Figure 2 presents the nitrogen adsorption/desorption isotherms and the corresponding pore size distributions of alumina, CCA03, and CCA06. The N2 adsorption/desorption isotherms show type IV hysteresis with a sharp capillary condensation effect at a partial pressure p/p0 of 0.7-0.9. With increasing carbon deposition, a new step appears at a lower partial pressure, 0.4-0.7, which is clearly observed for CCA06. This implies that a new type of pore appears in the highcarbon-content CCA samples. Figure 2B shows the pore size distribution curves of alumina, CCA03, and CCA06 calculated from the desorption branches of the isotherms according to the BJH (Barrett-Joyner-Halenda) method. CCA03 and alumina have similar pore size distributions with most probable pore diameter centers at 10.9 and 11.7 nm, respectively. Interestingly enough, the difference between these two diameters is about 0.8 nm, that is, about twice the thickness of a graphite monolayer. This change in diameter implies that carbon is uniformly deposited on the alumina surface.1,18 However, from sample CCA04 on, a new peak appears at a pore diameter of 3-4 nm and its intensity increases with carbon content.

5042

Langmuir, Vol. 21, No. 11, 2005

Lin et al.

Figure 1. (A) XRD patterns of (a) γ-Al2O3 and (b-f) sucrose/γ-Al2O3 with a sucrose loading of (b) 0.30, (c) 0.40, (d) 0.50, (e) 0.60, and (f) 0.80 g sucrose/g γ-Al2O3. (B) Residual crystalline sucrose content versus the total sucrose content.

Figure 2. Nitrogen adsorption/desorption isotherms (A) and pore size distributions (B) of Al2O3, CCA03, and CCA06.

layer thicknesses of these two samples are very similar. Because the carbon content in CCA03-2 is 9.6 wt%, close to the monolayer coverage of 9.1 wt%, the carbon coverage in CCA03-2 can be considered to be 1. Assuming that the pores of alumina are cylindrical, the theoretical distribution of uniformly CCA with one monolayer of carbon (OLCCA) can be calculated following the method adopted by Vissers et al.1 The pore volume, Vi(CCA), of a uniformly carbon-covered alumina (carbon thickness, d) was calculated from the experimentally derived pore volume, Vi(Al2O3), in association with the pore radius, ri(Al2O3) [equal to ri(CCA) + d], of the alumina support by

Vi(CCA) ) Vi(Al2O3) [1 - d/ri(Al2O3)]2 Figure 3. Comparison of experimentally determined pore size distributions of sample Al2O3, CCA03, CCA03-2, and CCA03-3 with theoretical ones of uniformly carbon-covered alumina with one and two layers of carbon, denoted as OLCCA and TLCCA, respectively.

In correlation with the XRD results, these small pores can be attributed to the carbon formed by the pyrolysis of the excess crystalline sucrose. If the entire alumina surface is covered by a layer of graphitized carbon, the carbon content calculated is about 9.1 wt%. The estimated carbon coverage is about 62.6% for CCA03 with a carbon content of 5.7 wt%. Sucrose dispersion may preferentially occur on the exposed alumina surface rather than on carbon; then, uniformly CCA with higher carbon coverage can be synthesized by impregnation of CCA03 with aqueous sucrose solutions. According to the pore size distributions (shown in Figure 3), the most probable pore diameters of the samples, CCA03 and CCA03-2, are almost equal; thus, the carbon

(2)

Thus, the complete pore size distribution can be calculated for OLCCA with one layer of graphitized carbon (d ) 0.67 nm) uniformly deposited on the alumina surface. The theoretical curve of OLCCA is also plotted in Figure 3. Clearly, the experimentally determined pore size distribution of CCA03-2 is almost identical to that of OLCCA. Furthermore, on the basis of the same assumption, the squared ratio of the most probable pore diameter of Al2O3 to CCA03-2 should equal the volume ratio of these two samples, which is well supported by the experimental results. In Figure 3, the most probable pore diameter of CCA03-3 is only 10.2 nm, that is, 1.5 nm lower than that of alumina and about twice as large as the carbon thickness in CCA032. This implies that a second layer of carbon is found in CCA03-3. Regarding the carbon content, the coverage of the second carbon layer should be lower than unity. A comparison of the experimental and theoretical pore size distributions for two-carbon-layered CCA (TLCCA) is also shown in Figure 3. These two curves are almost the same

Uniformly Carbon-Covered Alumina

Figure 4. UV Raman spectra of sample CCA01, CCA06, CCA03, CCA03-2, CCA03-3, and sucrose carbon (labeled as SC).

Langmuir, Vol. 21, No. 11, 2005 5043

Figure 5. UV Raman spectrum of sample CCA03 showing triplet in the “D” region.

except that the experimental curve has a higher intensity than the theoretical one. Since the peak area represents the pore volume, it can be concluded that the coverage of the second carbon layer in CCA03-3 is lower than unity. UV Raman Spectroscopy of CCA. The processes of carbon matter formation from organic substances are very complex. A general picture of the carbonization process can be presented by the following scheme for the cracking of hydrocarbons:22

paraffins f olefins f high-molecular-weight olefins and naphthenes f unsaturated cyclic hydrocarbons f aromatic hydrocarbons f high-molecular-weight condensed aromatic hydrocarbons f asphalt-like substances f carboids f coke (3) The process of carbon deposit formation in complex carbon-mineral adsorbents may be initiated and terminated at any stage in eq 3 depending on the chemical nature of the carbonized substance, the porous structure, the chemical nature of adsorption and catalytic centers of the mineral matrix, etc. Raman spectroscopy is a nondestructive tool useful for the characterization of different types of carbon.23-27 However, the Raman spectra of our CCA samples collected at 514 nm are featureless because of background fluorescence.25 In contrast, UV Raman spectra (Figure 4) exhibit excellent signal-to-noise ratios for the CCA samples. The most important features are two broad peaks centered at about 1367 and 1607 cm-1, which are attributed to the “D” (disorder) and “G” peaks of sp2-bonded carbon, respectively.23-27 These two peaks are common for various forms of disordered, noncrystalline, and amorphous carbons in the Raman spectra collected with excitation sources from IR to the deep UV. Their widths, intensities, and dispersions vary for different carbons, providing a powerful means for the identification of the carbon species.23 The “G” peak is due to bond stretching of sp2 carbon atoms. It is positioned at 1582 cm-1 in crystalline graphite, but shifts to higher frequencies 1607 cm-1 in CCAs, which is possibly related to conjugated (22) Fitzer, E.; Mueller, K.; Schaefer, W. Chemistry and Physics of Carbon. Walker, P. L., Jr., Ed.; Marcel Dekker: New York, 1973; Vol. 7, p 237. (23) Ferrari, A. C.; Robertson, J. Phys. Rev. B 2001, 64, 075414. (24) Jackson, B. R.; Trout, C. C.; Badding, J. V. Chem. Mater. 2003, 15, 1820. (25) Li, C.; Stair, P. C. Catal. Today 1997, 33, 353. (26) Li, C.; Stair, P. C. Stud. Surf. Sci. Catal. 1997, 105, 599. (27) Li, C. J. Catal. 2003, 216, 203.

Figure 6. XPS spectrum of CCA01 sample in the region of the O 1s binding energy.

olefinic species or polycyclic aromatic hydrocarbons (PAHs).28,29 The “D” peak of CCA03 can actually be fitted by three peaks (Figure 5). Different PAHs, viewed as molecular subunits of a graphitic plane,30 exhibit different UV Raman peaks in the spectral region, where the “D” peak is located.31 In Figure 4, the “G” peak of the sucrose carbon shifts to higher wavenumbers (1617 cm-1) with respect to that of the CCA samples. It’s believed that the “G” peak of the more-disordered sp2 carbon has a higher frequency. Therefore, carbon generated by the pyrolysis of sucrose highly dispersed on the alumina surface is more ordered than that from the residual crystalline sucrose. In other words, the alumina surface might favor the formation of ordered carbon species. XPS Investigation of CCA. X-ray photoelectron spectra of CCA were measured in the binding energy regions for Al 2p, O 1s, and C 1s electrons. The recorded Al 2p peaks are almost symmetric, whereas O 1s and C 1s core-level spectra can be deconvoluted into two and three individual peaks, respectively, as shown in Figures 6 and 7. The peak at 531.3 eV in Figure 6 may be attributed to Al-O bonds in the Al2O3 phase or CdO bonds in quinone groups. However, the intensity of this peak increases with Al2O3 content in the CCA samples (data not shown here). Therefore, this peak is probably due to the Al-O bond contribution and not the CdO bonds in quinone groups. The other peak at 532.7 eV can be assigned to hydroxyl (28) Angell, C. L. J. Phys. Chem. 1973, 77, 222. (29) Li, J.; Xiong, G.; Feng, Z.; Liu, Z.; Xin, Q.; Li, C. Micropor. Mesopor. Mater. 2000, 39, 275. (30) Negri, F.; Castiglioni, C.; Tommasini, M.; Zerbi, G. J. Phys. Chem. A 2002, 106, 3306. (31) Asher, S. A.; Johnson, C. R. Science 1984, 225, 311.

5044

Langmuir, Vol. 21, No. 11, 2005

Lin et al.

Figure 7. XPS spectrum of CCA01 sample in the region of the C ls binding energy.

Figure 9. The RB adsorption isotherms of samples alumina, sucrose carbon, CCA03, CCA03-2, and CCA03-3.

Figure 8. Change in the surface atomic ratio, (C/Al)S, versus total atomic ratio, (C/Al)T.

or carbonyl groups in quinines and anhydrides.32 For C 1s (Figure 7), the dominant peak at 284.8 eV can be assigned to carbon covering the alumina surface, whereas the components at 286.2 and 289.4 eV may be attributed to carbon bound to oxygen in surface species.32 Since the chemical shifts with respect to C-C are 1.5 eV for C-O, 3.0 eV for O-C-O or CdO, and 4.5 eV for OdC-O,32 these two high-energy C 1s components are probably related to the carboxylate species. The XPS data for the CCA samples show that the integral intensity of the C 1s and the overall Al 2p intensity change concurrently. Therefore, a comparison of the total and surface C/Al ratios could be important. Surface values, related to the probed surface layer, can be calculated using the following equation:

C/Al ) (IC/IAl)(SAl/SC)

(4)

where IC is the integral intensity of the C 1s, IAl is the intensity of the overall Al 2p peak, and SAl and SC are the atomic sensitivity factors of Al (0.193) and C (0.278) for the 2p and 1s electrons, respectively, determined from the empirical peak area values.32 The variation in the surface C/Al atomic ratio versus the total C/Al atomic ratio is shown in Figure 8. An obvious inflection occurs in the curve between CCA03 and CCA04, which is in good agreement with the results of XRD, the nitrogen adsorp(32) Plyuto, I. V.; Shpak, A. P.; Babich, I. V.; Plyuto, Y. V.; Sharanda, L. F.; Stoch, J.; Moulijn, J. A. Surf. Interface Anal. 1999, 27, 911.

tion isotherms, and the corresponding pore size distributions. This confirms that carbon is uniformly dispersed on the alumina surface for the samples with low carbon loadings, while aggregated carbon particles appeared in samples with higher carbon loadings.20 RB Adsorption of CCA. RB is a cationic organic dye commonly used for tracer studies in water research. Its adsorption in aqueous media was investigated for the further characterization of the surface properties of CCA. RB adsorption was performed on pure alumina, sucrose carbon, CCA03, CCA03-2, and CCA03-3. The corresponding RB adsorption isotherms, C/G versus C, are shown in Figure 9. These isotherms are well fitted with a correlation coefficient, r, greater than 0.99 using the Langmuir adsorption model as follows

C/G ) C/Gmax + 1/(KaGmax)

(5)

where C is the equilibrium concentration of the substrate in the solution in moles; Ka is the adsorption equilibrium constant in L mol-1; G is the experimental equilibrium adsorption amount in mol g-1; and Gmax is monolayer adsorption capacity in mol g-1. The Gmax can be estimated from the reciprocal of the line slope of C/G versus C; the values of adsorption equilibrium constant, Ka, can be obtained from the intercepts of the lines. When Gmax is obtained, the relative surface area, S, can be calculated according to the general surface area equation:33,34

S ) GmaxNAm × 10-20

(6)

where, S is the relative surface area; N is Avogadro’s (33) Potgieter, J. J. Chem. Ed. 1991, 68, 349. (34) Amjad, H. E.; Alan, P. N.; Hafid, K. A.; Suki, P.; Neil, C. J. Anal. Appl. Pyrol. 2004, 71, 151.

Uniformly Carbon-Covered Alumina

Langmuir, Vol. 21, No. 11, 2005 5045

Table 2. Saturated Adsorption Capacity (Gmax), Adsorption Equilibrium Constant (Ka), and BET and RB Relative Surface Areas of Different Samples in RB Adsorption Experiments samples

Gmax (10-5 mol g-1)

Ka (L mol-1)

SBET (m2 g-1)

RB relative surface area S (m2 g-1)

Gsmax (10-7 mol m-2)

Al2O3 SC CCA03 CCA03-2 CCA03-3

0.753 0.563 6.40 9.99 10.1

6.74 4.28 11.7 13.9 14.3

128 20 139 141 143

7.25 5.42 61.3 96.2 97.4

0.588 2.81 4.61 7.08 7.07

number; and Am is the cross-sectional area of the adsorbate taken as 160 Å2 for RB.35,36 All the results are listed in Table 2. Evidently, CCA samples have much higher adsorption capacities than the corresponding pure alumina and sucrose carbon. The saturated adsorption capacities (Gmax) of pure alumina and sucrose carbon are 0.753 × 10-5 and 0.563 × 10-5 mol g-1, respectively, whereas the Gmax for CCA03 is as high as 6.40 × 10-5 mol g-1. The CCA03-2 sample with a higher carbon coverage exhibits an even higher Gmax of 9.99 × 10-5 mol g-1. These results demonstrate that the deposition of carbon on the alumina surface leads to significant changes in the surface properties of alumina. However, it should be noted that the Gmax of CCA03-3 with a second carbon layer is similar to that of CCA03-2, indicating that the second layer of carbon has little effect on the increase in RB monolayer adsorption capacity of CCA. To better understand these special properties of CCA, the BET-surface-area-normalized monolayer adsorption capacities for RB on all the samples, labeled as Gsmax, were calculated and listed in Table 2. The monolayer adsorption capacity, expressed per unit area, Gsmax, represents the intrinsic adsorption ability or the surface density of active adsorption sites of different materials for RB. It is self-evident that alumina shows the lowest Gsmax; the Gsmax of sucrose carbon is nearly five times that of alumina. When the alumina surface is partially covered by carbon, as in the case of CCA03, its adsorption ability greatly increased. Nevertheless, both CCA03-2 and CCA03-3 have a carbon coverage of 1, and their Gsmax of 7.08 × 10-7 and 7.07 × 10-7 mol m-2, respectively, are almost the same despite the existence of the second carbon layer in CCA03-3; this suggests a very weak interaction between carbon species and the alumina surface and that the adsorption ability of the carbon layer on the alumina surface in CCA samples is independent of its thickness with a Gsmax of about 7.08 × 10-7 mol m-2. From the above-mentioned results, it is reasonable to attribute the increased adsorption ability of CCA with a higher Gsmax to the “dispersion” of carbon on the alumina surface. Therefore, the CCA with a carbon coverage that is lower than unity should have a combination of surface properties characteristic of alumina and the dispersed carbon. CCA03 with a carbon content of 5.7 wt% has a carbon coverage of about 62.6% based on the assumption that the carbon is in the form of a graphene layer. Therefore, the Gsmax of CCA03 should be equal to the sum of the contribution of alumina and carbon covering the surface of alumina:

0.626Gsmax(CCA03-2) + (1 - 0.626)Gsmax(Al2O3) ) 4.65 × 10-7 mol m-2 (7) which is very close to the experimental Gsmax (4.61 × 10-7 (35) Sorensen, B. L.; Wakeman, R. J. Water Res. 1996, 30, 115. (36) Amina, A. A.; Badie, S. G.; Soheir, A. K. J. Chem. Technol. Biotechnol. 2003, 78, 611.

mol m-2) of CCA03. This calculation convincingly supports our conclusion that the as-prepared CCA possesses the surface properties of carbon or both carbon and alumina depending on the carbon coverage. It must be pointed out that the specific adsorption properties of carbon in CCA and pure sucrose carbon should be similar because of the poor interaction between carbon species and alumina. However, the Gsmax of carbon in CCA is much higher than that of pure sucrose carbon, as shown in Table 2. This may be due to the different pore sizes of CCA and sucrose carbon. Krupa and Canon37 have stated that the correlation between pore volume and adsorption capacity of dyes may indicate pore filling as the dominant adsorption mechanism of dyes. Although we have not found any information on the diameter of pores accessible to RB, it is indeed reported36 that another dye, methylene blue (MB), can be used to measure the area accessible in pores with a diameter greater than 15 Å. From the cross-sectional area of RB, about 160 Å2 (this is larger than that of MB, about 120 Å2), it can be estimated that the diameter of the pores accessible to RB should be larger than 17 Å. From the nitrogen adsorption results, the most probable pore diameter of sucrose carbon is smaller than 14 Å. Therefore, the highly microporous nature of the sucrose carbon should suppress the adsorption of RB to a certain extent. As shown in Table 2, the RB relative surface area is only one-fourth of the total surface area. In contrast to sucrose carbon, carbon in CCA is highly dispersed on the γ-Al2O3 surface, whose most probable pore diameter is as large as 117 Å. Accordingly, RB should be able to reach most of the carbon in CCA. Thus, CCA has the texture of alumina, and carbon highly dispersed in the pores with a large pore diameter enables the effective use of the active adsorption sites of carbon. Conclusions 1. Sucrose can be highly dispersed on the alumina surface when its loading is kept under a threshold value. Pyrolysis of such a highly dispersed sucrose yields uniformly carbon-covered alumina (CCA). 2. The coverage and number of the carbon layers in CCA can be controlled by the tuning of the sucrose content in the precursor and the impregnation times. 3. Carbon species dispersed on the alumina surface in CCA are probably conjugated olefinic or polycyclic aromatic hydrocarbons that can be viewed as molecular subunits of a graphitic plane. 4. The carbon/alumina composite possesses the texture of alumina and the surface properties of carbon or both carbon and alumina depending on the carbon coverage. Alumina is a widely used material, and its pore size distribution can be adjusted. Therefore, the synthesis of uniformly carbon-covered alumina with different textures and carbon coverage is of great significance to meet a (37) Krupa, N. E.; Canpn, F. S. J. Am. Water Works Assoc. 1996, 88, 94.

5046

Langmuir, Vol. 21, No. 11, 2005

variety of applications. In addition, this simple approach is also suitable for the preparation of other carbonmineral composites such as carbon-covered titania and silica. This type of composite material has potential for applications as catalysts or catalyst supports and adsorbents.

Lin et al.

Acknowledgment. The authors are grateful to the National Science Foundation of China (20173002), and the Major State Basic Research Development Program (Grant No. G2000077503) for their financial support. LA047097D