Zr

14552

J. Phys. Chem. C 2010, 114, 14552–14560

Effects of Surface Features on Adsorption of SO2 on Graphite Oxide/Zr(OH)4 Composites Mykola Seredych and Teresa J. Bandosz* Department of Chemistry, The City College of New York, 160 ConVent AVenue, New York, New York 10031 ReceiVed: June 4, 2010; ReVised Manuscript ReceiVed: July 26, 2010

Zirconium hydroxide/graphene composites were synthesized from zirconium chloride and graphite oxide with the content of the graphene component between 5 and 50%. They were used as adsorbents of sulfur dioxide at ambient conditions. The initial and exhausted materials were characterized using adsorption of nitrogen, infrared spectroscopy, potentiometric titration, scanning electron microscopy and thermal analysis. The results indicated enhanced adsorption of SO2 on the composites, which is linked to the formation of new basic sites and porosity as a result of interactions between zirconium hydroxide units and the oxygen groups attached to the graphene layers. Both physical adsorption and reactive adsorption of SO2 via formation of sulfites play a role in the retention process. The graphene component catalyzes oxidation of SO2 and leads to the formation of sulfates. Introduction Adsorption of sulfur dioxide on various adsorbents at ambient conditions has been extensively studied.1-21 Examples of adsorbents include activated carbons,1-9 activated carbon fibers,10-13 fly ash,14 sewage sludge derived materials,15 porous adsorbents of inorganic origin with basic properties,16-20 and exfoliated graphite.21 Removal of SO2 is important from both environmental (acid rain) and respiratory protection aspects. For the latter, filters being able to remove simultaneously species of different chemistries are in demand. Another important characteristic of respiratory filtration is the efficiency of air purification at ambient temperatures and in the presence of moisture. This can be a challenge since toxic industrial compounds (TICs) are often small molecule gases above their critical temperatures. Their heats of physical adsorption, which can be linked to the strength of the retention process, are usually low. For all of these reasons, reactive adsorption of TICs became a current research target.22,23 This process refers to the adsorption enhanced by chemical reactions on the surface of the adsorbents. These reactions may involve the reactions of the adsorbed molecules between each other or with oxygen and/or with the adsorbent surface and/or with species introduced to that surface to impose such reactions. SO2 is retained via physical adsorption in small pores and via interaction with surface basic groups on activated carbons, which are typical and the most versatile adsorbents used in respirators, .3,4,7,24 It was found that pores smaller than 7 Å in size are important for this process.3,4,7 The presence of water and the catalytic action of the carbon surface cause conversion of SO2 to sulfuric acid and thus the pollutant is strongly adsorbed on the surface. In our laboratory, the capacities of about 50 mg/ g, 70 mg/g, and 160 mg/g were reported at ambient conditions on coconut shell based carbon, wood-based carbon modified with nitrogen, and Centaur, respectively.3 Centaur is the catalytic carbon-containing nitrogen designed to remove acidic gases such as hydrogen sulfide via its oxidation to sulfur dioxide and then to sulfuric acid.25 On the adsorbents derived from sewage sludge, * To whom correspondence should be addressed. Phone: (212) 650-6017. Fax (212) 650-6107. E-mail: [email protected].

30 mg/g of SO2 was adsorbed.15 An SO2 removal capacity of 20 mg/g was also reported on virgin carbon by Carabineiro and co-workers.26 However, modifications of carbon by an introduction of small amounts of copper and vanadium (4%) increased the capacity of impregnated carbons to 140 mg/g.26 The presence of inorganic species was also indicated as important in the case of SO2 removal on sewage sludge derived adsorbents.15 In a search for a versatile adsorbent for respirator protection, Peterson and co-workers investigated the SO2 reactive adsorption on porous zirconium hydroxide.20 The capacity recorded was about 90 mg/cm3 of the bed volume and SO2 was strongly adsorbed via formation of sulfites.20 It was found that for this chemical reaction, only 10% of the available hydroxyl groups was used. In recent efforts to increase the versatility and efficiency of the adsorbents working at ambient conditions, new composites based on graphene components were synthesized. Their synthesis involves various inorganic species as iron,27,28 molybdenum/tungsten,29 silica,30 aluminum and zirconium polycations,31 and metal-organic framework (MOF).32,33 In the latter materials, formation of new porosity between the graphene layers and the MOF components was linked to the enhanced adsorption of ammonia at ambient conditions.33 On the basis of the above, the objective of this work is to synthesize new Zr(OH)4/graphite oxide composites, to characterize their surface, and to determine their sulfur dioxide sorption capacity. Differences in capacities result from differences in porosity, surface chemistry, and the conditions of the SO2 removal. All of these have an effect not only on the sorption capacity, but also on the chemical nature of the surface reaction products. Experimental Section Materials. Graphite oxide (GO) was synthesized by oxidation of graphite (Sigma-Aldrich) using the Hummers method.34 The details of synthesis of the particular GO used in this study are described elsewhere.31 The composite material GO/zirconium hydroxide was prepared by dispersing GO powder (5%, 20%, 50 wt % of the final mass of the material) in the 1.0 L sodium hydroxide (0.05 M). The resulting suspension was subsequently stirred over 3 h. The zirconium(IV) chloride solution (0.05 M) was then added (250

10.1021/jp1051479  2010 American Chemical Society Published on Web 08/11/2010

Adsorption of SO2 on Graphite Oxide/Zr(OH)4 mL) to the dispersed graphite oxide with a rate 0.6 mL/min using a Titronic Universal (SCHOTT). Then, the obtained composite was extensively washed with distilled water until neutral pH and no traces of chloride ion were found in the suspension. Finally, the suspension was centrifuged and the gel formed was dried at 50 °C during 48 h. Zirconium hydroxide was prepared in the same way, without GO dispersed in NaOH solution. The graphite oxide obtained using Hummers methods is referred to as GO and zirconium hydroxide (hydrous zirconia) as ZrA. The composites obtained with 5%, 20%, and 50% of GO are referred to as ZrG-1, ZrG-2, and ZrG-3. The prepared materials were studied as adsorbents for sulfur dioxide in the dynamic tests described below under dry (D) and wet conditions (M). After the concentration of SO2 in the effluent gas reached 20 ppm, the samples were considered as exhausted for sulfur dioxide adsorption. Such samples are identified with letter “E” added to their names. Methods SO2 Breakthrough Capacity. Dynamic tests were carried out at room temperature to evaluate the capacity of the adsorbents for SO2 removal under two sets of conditions, wet and dry. For the former, adsorbent samples were packed into a glass column (internal diameter 9 mm). The bed volume used was 1.2 cm3 (with the mass of adsorbent between 1.27 to 1.45 g). Before the experiments in moist air, the samples were prehumidified with moist air (relative humidity 70% at 25 °C) for 2 h. The amount of water adsorbed was estimated from the increase in the sample weight. Dry or moist air (relative humidity 70% at 25 °C) containing 0.1% (1000 ppm) SO2 was then passed through the column of adsorbent at 500 mL/min. The breakthrough of SO2 was monitored using a MultiRae Plus monitoring system with an electrochemical sensor. The test was stopped at the breakthrough concentration of 20 ppm (sensor limit). The adsorption capacities of each adsorbent in terms of g of SO2 per g of carbon were calculated by integration of the area above the breakthrough curves, and from the SO2 concentration in the inlet gas, flow rate, breakthrough time, and mass of adsorbent. For each sample, the SO2 test was repeated at least twice. The determined capacities agreed to within 4%. To determine the capacity of the dry adsorbent, the experimental conditions were the same as those in the wet run except for the absence of water vapor. The samples run in dry are referred to with the letter D and those run in moist conditions with-M. The amount of weakly adsorbed SO2 was evaluated by purging the adsorbent column with air at 450 mL/min immediately after the breakthrough experiment. The SO2 concentration was monitored until its concentration dropped to 1 ppm. Nitrogen Adsorption. Nitrogen adsorption isotherms were measured using an ASAP 2010 analyzer (Micromeritics, Norcross, GA, USA) at -196 °C. Before the experiment, the samples were degassed at 120 °C to a constant pressure of 10-4 Torr. The isotherms were used to calculate the specific surface area, SBET; micropore volume, Vmic; total pore volume, Vt; and pore size distribution. The micropore volume was calculated using Dubinin-Radushkevich approach35 and the total pore volume seen by the nitrogen molecules, from the last point of the isotherms based on the volume of nitrogen adsorbed. The volume of mesopores, Vmes, represents the different between those two values. The relative microporosity was calculated as the ratio of the micropore volume to the total pore volume.

J. Phys. Chem. C, Vol. 114, No. 34, 2010 14553 pH of the Surface. A 0.1-g sample of dry adsorbent was added to 5 mL of deionized water and the suspension was stirred overnight to reach equilibrium. The sample was filtered and the pH of solution was measured using an Accumet Basic pH meter (Fisher Scientific, Springfield, NJ, USA). Thermal Analysis. Thermal analysis was carried out using TA Instruments Thermal Analyzer (New Castle, DE, USA). The heating rate was 10 °C/min in a nitrogen atmosphere at 100 mL/min flow rate. The samples were heated up to 1000 °C. SEM. Scanning electron microscopy (SEM) images were obtained using a Zeiss Supra 55 VP with an accelerating voltage of 5.00 kV. Scanning was performed on a sample powder previously dried (120 °C) and sputter coated with a gold to avoid charging. Elemental Analysis. The content of sulfur was evaluated using X-ray fluorescence (SPECTRO model 300T Benchtop Analyzer; ASOMA Instruments, Inc.) based on the calibration curve done for the samples with the internal standard of sulfur. The instrument has a titanium target X-ray tube and a highresolution detector. FTIR. Fourier transform infrared (FTIR) spectroscopy was carried out using a Nicolet Magna-IR 830 spectrometer using the attenuated total reflectance (ATR) method. The spectrum was generated and collected 16 times and corrected for the background noise. The experiments were done on the powdered samples, without KBr addition. Potentiometric Titration. Potentiometric titration measurements were performed with a DMS Titrino 716 automatic titrator (Metrohm). The instrument was set at the mode where the equilibrium pH is collected. Subsamples of the initial and exhausted materials (∼ 0.100 g) were added to NaNO3 (0.01 M, 50 mL) and placed in a container maintained at 25 °C overnight for equilibrium. During the titration, to eliminate the influence of atmospheric CO2, the suspension was continuously saturated with N2. The suspension was stirred throughout the measurements. Volumetric standard NaOH (0.1 M) was used as the titrant. The experiments were performed in the pH range of 3-10. Each sample was titrated with base after acidifying the sample suspension. No sharp changes in the slope were noticed which indicates the stability of the material in the experimental pH window. The experimental data was transformed into a proton binding isotherm, Q, representing the total amount of protonated sites, which is related to the pKa distribution by the following integral equation:

Q(pH) )

∫-∞∞ q(pH, pKa)f(pKa)dpKa

(1)

The solution of this equation is obtained using the numerical procedure,36,37 which applies regularization combined with nonnegativity constraints. On the basis of the spectrum of acidity constants and the history of the samples, the detailed surface chemistry was evaluated. Results and Discussion The SO2 breakthrough curves obtained on our adsorbents are collected in Figure 1. It should be mentioned here that GO alone does not retain any sulfur dioxide. In the case of ZrA, water in the system has a detrimental effect on the adsorption capacity and the breakthrough is reached about 25 min earlier (1/4 of the breakthrough time) compared to the experiment run in totally dry conditions. This suggests the competition between SO2 and water for the adsorption centers. Building the composites with

14554

J. Phys. Chem. C, Vol. 114, No. 34, 2010

Seredych and Bandosz

Figure 2. Comparison of the measured and hypothetical SO2 adsorption capacities in dry (A) and in moist (B) conditions.

Figure 1. SO2 breakthrough curves and desorption curves.

TABLE 1: SO2 Breakthrough Capacities, Amount of Water Preadsorbed and the Surface pH Values for the Initial and Exhausted Samples SO2 breakthrough capacity H2O adsorbed 3

sample

[mg/g]

[mg/cm ]

GO-ED GO-EM ZrA-ED ZrA-EM ZrG-1-ED ZrG-1-EM ZrG-2-ED ZrG-2-EM ZrG-3-ED ZrG-3-EM

0.0 0.0 116 83 146 99 112 106 59 50

0.0 0.0 147 106 178 120 136 128 71 57

[mg/g] 96 88 86 75 84

pH in

exh

2.62 2.62 6.58 6.58 7.38 7.38 6.85 6.85 6.04 6.04

4.51 5.02 4.45 5.61 4.76 5.06 4.92 5.20

5 or 20% of GO increases the breakthrough time, especially in moist conditions. Less sensitivity toward the detrimental effect of water is also observed, especially for ZrG-2. A sharp decrease in the SO2 concentration during air purging indicates a strong adsorption of sulfur dioxide in the case of all materials. Table 1 lists the calculated breakthrough capacities, amount of water preadsorbed, and surface pH values. The breakthrough capacities are reported per gram of an adsorbent and per a unit volume of the bed. The last values are important for practical applications. For ZrA and the composites with 5 and 20% of GO these capacities are about 20% to 100% greater than those reported by Peterson and co-workers for commercial hydrous zirconia.20 Even though in our work the breakthrough concentration is higher, the steep rise of the breakthrough curves makes these differences not so significant. However, a higher concentration of SO2 makes our test more accelerated, which could result rather in underestimated, not overestimated, capacities.38 For further evaluation of the adsorption process, the capacities per gram of adsorbents were analyzed. To visualize the

synergetic effect of the composite formation, the measured capacities were compared to those hypothetical ones calculated assuming the physical mixture of the components and their capacities. The results obtained are presented in Figure 2. The effect of the enhancement is especially seen for ZrG-1 and ZrG2. It is interesting that while for the former sample the adsorption in dry conditions seems to benefit from the composite formation, for the latter sample, the enhancing effect is more visible when the moisture is present in the system. For ZrG-2, there is a very small difference between the capacity measured in dry and moist conditions. The sample with the high content of GO exhibits only the capacity of the ZrA component with no visible synergetic effect. It is interesting that the amount of water adsorbed is not affected by the composition of the materials. This can be caused by the high affinity of GO present to retain water.39,40 Slightly less water adsorbed in the case of ZrG-2 can be related to the chemistry of composite formation, which is discussed below. Some interesting behavior is found in the trend of the surface pH values, which represent the average acidity of the synthesized materials. While Zr(OH)4 can be considered as chemically neutral, building a composite with a much more acidic component, GO, results in an unexpected increase in samples’ basicity. This is a direct indication of the new chemistry formed in this process. Since the basic character of the adsorbents was found to be important for SO2 adsorption, this factor has to be considered in derivation of the adsorption mechanism. After SO2 adsorption, the pH became much more acidic. We link this result to the chemical reaction and formation of acids and salts. Figure 3 shows the direct relationship between the decrease in the samples’ pH and the amount of SO2 adsorbed. This linear trend suggests a relatively simple chemistry of reactive adsorption. As discussed by Peterson and co-workers based on the results of the XPS study, zirconium sulfite, Zr(SO3)2 is expected as the major reaction product on hydrous zirconia.20 While in their work only about 10% of hydroxyl groups was involved in surface reactions, assuming that all SO2 reacted with Zr(OH)4, in the case of our materials, that “utilization” of surface chemistry in dry conditions would be 14%, 18%, 14%, and 7% for ZrA, ZrG-1, ZrG-2 and ZrG-3, respectively. In wet conditions, those number are 10%, 12%, 13%, and 6% for ZrA, ZrG1, ZrG-2, and ZrG-3, respectively. In the case of composite, that increase in the surface efficiency can be caused by synergetic effects of both, chemistry and porosity of the new

Adsorption of SO2 on Graphite Oxide/Zr(OH)4

J. Phys. Chem. C, Vol. 114, No. 34, 2010 14555 TABLE 2: Parameters of Porous Structure Calculated from Nitrogen Adsorption Isotherms

Figure 3. Dependence of the decrease in the pH after SO2 adsorption on the amount of SO2 adsorbed.

Figure 4. Nitrogen adsorption isotherms.

materials. It is interesting to note that even for pure hydrous zirconia, the efficiency of the stoichiometric reaction is higher than that for the commercial material.20 To propose feasible mechanisms of the composite formation and SO2 reactive adsorption the porosity of the materials studied has to be analyzed. The measured nitrogen adsorption isotherms are collected in Figure 4. Their shapes indicate the presence of micropores and mesopores. The volume of the latter increases with an increase in the content of GO, as evidenced by the shape of the hysteresis loop. The parameters of the porous structure calculated from the nitrogen adsorption isotherms for the initial and exhausted adsorbents are summarized in Table 2. The surface area and pore volume of ZrA consist of only about 60% of the values reported for the commercial material.20 Nevertheless, as indicated above, the capacity for SO2 removal is higher. All of the materials studied, with the exception of GO, can be considered as mesoporous. GO, as discussed previously, does not exhibit any porosity seen by nitrogen molecules.31,41 A synergetic effect of the composite formation is seen when the measured porosity is compared to that hypothetical one calculated assuming the physical mixture of the components (Figure 5). The most visible changes are those for ZrG-2 and ZrG-3. While larger pores can be formed between the GO flakes and the Zr(OH)4 units, the origin of an increased microporosity is

sample

SBET [m2/g]

Vt [cm3/g]

Vmeso [cm3/g]

Vmic [cm3/g]

Vmic/Vt

ZrA ZrA-ED ZrA-EM ZrG-1 ZrG-1-ED ZrG-1-EM ZrG-2 ZrG-2-ED ZrG-2-EM ZrG-3 ZrG-3-ED ZrG-3-EM

230 256 230 201 222 220 219 206 202 141 154 160

0.131 0.151 0.134 0.119 0.126 0.126 0.150 0.126 0.123 0.115 0.128 0.129

0.071 0.086 0.077 0.067 0.070 0.067 0.099 0.076 0.072 0.079 0.096 0.090

0.060 0.065 0.057 0.052 0.056 0.059 0.051 0.050 0.051 0.036 0.032 0.039

0.46 0.43 0.43 0.44 0.44 0.47 0.34 0.40 0.42 0.31 0.25 0.30

expected to be in the new chemistry formed between the components of the composites. After exposure to SO2, no significant changes in the porosity are noticed. This indicates that even though the chemical reaction could take place, the lattice of the materials’ structure is preserved, which is rather expected taking into account low stoichiometrical reactivity of surface hydroxyls. That reactivity can be limited to the terminal -OH in the predominantly mesoporous structure of the final materials, owing to their accessibility. If any SO2 was physically adsorbed in small pores, then it could be removed during outgassing at 120 °C. That outgassing could also remove chemically bound SO2 if the decomposition temperature of the products of surface reactions was close to that one used during outgassing. The X-ray diffraction patterns of graphite oxide, hydrous zirconia, and its composites are shown in Figure 6. For GO, the peak at 2Θ 11.7° corresponds the interlayer distance of 7.59 Å. Since hydrous zirconia represents the major component of the composites in graphite oxide, the XRD pattern of amorphous hydrous zirconia was expected. For the composites, the diffraction peak of GO is not seen and only a small shoulder is visible with an increase in the content of the graphite oxide. Thermal analysis of the materials before and after SO2 adsorption can throw some light on the products of surface reactions. The differential thermal gravimetric (DTG) curves are presented in Figure 7. For the ZrA sample, a gradual weight loss is assigned to the dehydration and dehydroxylation of a complex structure of zirconium hydroxide.42,43 In the case of composite, a new weight loss occurs at about 200 °C. It is seen as a shoulder for ZrG-1 and a well-defined peak for ZrG-3, and

Figure 5. Comparison of the measured and hypothetical micropore volumes (A) and total pore volume (B) in the composites.

14556

J. Phys. Chem. C, Vol. 114, No. 34, 2010

Figure 6. XRD patterns for the samples studied.

they represent the decomposition of the epoxy groups of the GO component (Figure 7).44,45 For the unexposed samples, a sharp peak at about 460 °C of decreasing intensity with an increase in the content of GO represents formation of metastable tetragonal zirconium oxide.43,46 After exposure of ZrA to SO2 in dry conditions, a gradual weight loss is revealed between 220 and 450 °C. It is likely related to the removal of water formed in the reaction of SO2 with OH groups and to the decomposition of Zr(SO3)2. The weight loss between 200-500 °C associated with adsorbed SO2 was also observed for commercial Zr(OH)4, studied by Peterson and co-workers.20 A phase transformation to zirconium oxide in ZrA-ED is represented by two sharp peaks at about 500 °C. This complexity is linked to the more complex chemistry of the exhausted adsorbent than that of the initial sample. Interestingly, when the experiments were run in moist air, the similar pattern to that for unexposed zirconium hydroxide was found. The only visible difference is at low-temperature peaks (∼100 °C), which we linked to the removal of physically adsorbed SO2 and water with dissolved SO2 (H2SO3). For the composites with 5 and 20% of GO, either run in dry or wet conditions, a gradual weight loss between 250 and 500 °C is found with visible sharp peaks of the phase transformation at 520 °C. Moreover, a well-marked weight loss is also noticed between 550 and 800 °C. It is linked to the decomposition of zirconium sulfate, Zr(SO4)2 with the emission of SO3.47 Formation of those species in the presence of the GO component can be explained by the oxidative capability of the carbonaceous material with incorporated oxygen functional groups.48,49 Moreover, since the GO used in the synthesis was obtained by the Hummers method, sulfonic groups present on the surface can also react with zirconium during the preparation step. This can be seen on the DTG curves about 600 °C for the ZrG-2 and ZrG-3. The sharp peaks at temperatures less than 150 °C are assigned to the removal of physically adsorbed water and SO2 dissolved in a water film. Nevertheless, knowing that the solubility of SO2 in water at 25 °C is 9.6 mg/100 mL50 and assuming that all pores are filled by water, only a small fraction of SO2 can exist in the latter form. That low temperature weight loss is especially well pronounced for ZrG-2, for which the adsorption at moist conditions was high and close to that one in the absence of water. The percentage of the weight loss between 210 and 700 °C, which is linked to the strong reactive adsorption, is presented in Table 3. Only the difference between the weight of the exhausted and the initial samples are listed in this table, along

Seredych and Bandosz with the comparison of the amount of SO2 adsorbed and sulfur detected in the materials using XRF. Assuming that the weight loss between 200-700 °C represents the removal of sulfur oxide reacting with terminal OH groups, the amounts of sulfur coming into the system and being detected are far from being balanced (it is assumed that sulfur from the sulfonic groups does not contribute significantly to the material balance). The differences are greater for the experiments run in dry conditions than for those in moist conditions. This supports our hypothesis that the significant amount of sulfur is physically adsorbed in small pores and removed at temperature less than 200 °C. This could also explain the unchanged porosity after the breakthrough experiments. That contribution of physically adsorbed sulfur dioxide seems to be especially high for ZrG-1 and ZrG-2. For these two samples, the synergetic effect of the enhancement in the SO2 removal capacity is high (Figure 2). Interestingly, the synergy in porosity was not important for ZrG-1. This suggests that the enhancement in the capacity of ZrG-1 is mainly caused by chemical factors, whereas for ZrG-2, the porosity effect is predominant. This is supported by the dramatic change in the shape of nitrogen adsorption isotherm between ZrG-1 and ZrG2. In the latter, the hysteresis loop is clearly visible. To investigate the aspects of changes in surface chemistry, the FTIR spectra are analyzed (Figure 8). Vibrations characteristic for graphite oxides appear at 1050, 1380, and 1630 cm-1. They correspond to the stretching of C-O bonds from carboxylic groups, to O-H bending from hydroxyl/phenol groups and to either O-H vibration in water and/or to the presence of oxygen surface compounds (cyclic ethers), respectively.51-53 The band at 1735 cm-1 is characteristic of CdO stretching vibration in carboxylic acids. Vibrations at 990 cm-1 correspond to epoxy/ peroxide groups.52 A band at 1228 cm-1 on the GO spectra might be related to SdO assymmetric stretching vibration in sulfonic groups and/or vibration of C-O in epoxides.52,54 It should be noted that the symmetric vibration of SdO from sulfonic groups appears at 1050 cm-1 as for the vibration of C-O. The absorption peaks at 1620 cm-1 and 1380 cm-1 for zirconium hydroxide are assigned to OH groups of water and O-H bending vibration from Zr-OH. The band at 854 cm-1 is attributed to Zr-O lattice vibration.43,55 For composites with low content of GO (5% and 20%), the vibrations of oxygen groups such as epoxide and carboxylic are not visible. When the content of GO reaches 50% these bands appear but they are of lower intensity than those for pure GO. Alternatively, the band at 854 cm-1 representing Zr-O lattice is seen for ZrG-1 and ZrG-2. After SO2 adsorption, the spectra for ZrA, ZrG-1 and ZrG-2 visibly change in the range between 850 and 1200 cm-1. Moreover, the bands at 1380 cm-1 and 1620 cm-1 are much less pronounced, suggesting replacement of some -OH groups as a result of their reactions with sulfur oxide. This is especially seen on the spectra of samples on which the highest amounts of SO2 were adsorbed. The absorption bands between 850 and 1200 cm-1 are linked to the vibrations from sulfur oxygen bonds. SO42- ions are represented by the bands at 1150 cm-1, and HSO4- by the bands at 850 cm-1 and 1030 cm-1.56,57 At 1150 cm-1, there is also a possibility of vibrations from SO32-. This is especially seen for the composite with the lowest content of GO. In the case of this material, it was hypothesized that the enhancement in the amount adsorbed was mainly owing to the new chemistry formed as a result of the synergy between the composite components. That chemistry was demonstrated by the increase in the pH values listed in Table 1. Sulfate bands are likely the result of SO2 oxidation by either oxygen dissolved

Adsorption of SO2 on Graphite Oxide/Zr(OH)4

J. Phys. Chem. C, Vol. 114, No. 34, 2010 14557

Figure 7. DTG curves in nitrogen for the initial and exhausted samples.

in water or superoxide anions O2-* activated by GO layers. The presence of sulfates was suggested by DTG curves for the exhausted composites. That new chemistry of the composites is also seen on the pKa distributions for the initial samples (Figure 9). On the surface of GO carboxylic (pKa < 7) and -OH groups (pH > 7) are seen. However, the peaks at pKa about 4.0, 6.5, 7.8, 9.0 for zirconium hydroxide represent various Brønsted acidic groups of Zr-OH.58 Formation of the composite, even with 5% of GO, results in the appearance of groups with pKa about 5, 6.5 and 8.5, whose amounts cannot be linked proportionally to the content of GO. Also, a significant quantity of new groups at pKa about 10 is revealed. They might be responsible for the predominant basicity of the composites. The significance of the

TABLE 3: Analysis of Weight for the Exhausted Samples and the Expected and Measured Contents of Sulfur after the Breakthrough Experiments

sample

∆ [%] 210-400 °C

∆ [%] 400-700 °C

Σ∆ [%]

SSO2 [%]

SXRF [%]

SSO2/ SXRF

ZrA-ED ZrG-1-ED ZrG-2-ED ZrG-3-ED ZrA-EM ZrG-1-EM ZrG-2-EM ZrG-3-EM

1.0 1.46 0.93 0.93 0 0.49 0.73 0

3.06 2.54 1.43 2.12 0.94 1.79 1.86 0.55

4.06 4.00 2.36 3.05 0.94 2.28 2.59 0.55

5.77 7.32 5.58 2.93 4.16 4.96 5.30 2.34

4.4 4.6 4.1 2.5 4.2 3.6 3.5 2.4

1.31 1.59 1.36 1.17 0.99 1.37 1.51 0.97

14558

J. Phys. Chem. C, Vol. 114, No. 34, 2010

Seredych and Bandosz

Figure 8. FTIR spectra for the samples studied.

groups detected at the end of the experimental window should not be over interpreted owing to the assumptions used in the calculations.36,37 As indicated in the literature, zirconium hydroxide is envisioned as a two-dimensional square lattice, which is connected to each other by bridging hydroxyl groups.42 GO, however, consists of graphene layers with epoxy and -OH groups attached to the basal planes and carboxylic and sulfonic groups existing at the edges of these layer.44,51-53,59 Those -OH groups of graphene layers dispersed in water can work as bridging groups, resulting in chemical bonds between the two components of the mixture. Another possible mechanism of interactions is hydrogen bonding between the units of hydrous zirconia and the epoxy groups of GO. These two scenarios are visualized in Figure 10. Mesopores are likely formed as a result of the reactions of zirconium with carboxylic and sulfonic groups of graphene layers, which can lead to the house-of-cards structure. They can be also formed between the new units. That change in the zirconium environment is demonstrated by new peaks on the pKa distributions. A small amount of GO in the case of ZrG-1 can explain the lack of synergetic effect on the porosity (Figure 5). More GO causes that more OH groups are involved as bridging groups interacting with zirconium hydroxide components, and thus the synergetic effect on the porosity is greater. The lower volume of pores in ZrG-3 compared to other composites results in the smaller SO2 removal capacity. Also, the fact that more zirconium is involved in the reactions with the acidic groups of GO causes the edges of the zirconium oxide phase not to be active in the reactions of sulfate formation. Those terminal -OH were hypothesized by Peterson and co-workers as taking part in the surface reactions. This negatively compensates the synergetic effect caused by the porosity develop-

Figure 9. Distributions of acidity constants for the species present on the surface of the initial samples.

ment. Alternatively, the large enhancement in the SO2 removal capacity, especially in dry conditions, is caused by the increased basicity of the composite. Even though bridging zirconium units with some -OH groups of distorted graphene layers eliminates certain oxygen containing centers of both, GO and Zr(OH)4. The new centers remain polar/basic in the nature and thus SO2 can be adsorbed owing to its polarity. When water is present in the system, it competes with SO2 for those specific centers, fills up the small pores and SO2 can be adsorbed there only by dissolution. The similarity in the amounts of SO2 adsorbed on ZrG-2 in dry and wet conditions can be linked to more pore space between the graphene and zirconium hydroxides layers compared to that for ZrG-1. The surface in those pores, owing to the nature of graphene, is more hydrophobic than that of hydrous zirconia. Therefore, SO2 is attracted to those pores stronger than water owing to the differences in their polarity (dipole moment of SO2 is 1.61D60 compared to 1.87D for water50). The texture of the composites is presented on SEM micrographs (Figure 11). As seen, the Zr(OH)4 component deposits on the layers of GO. The thin composite layers arranged in agglomerates are especially seen for ZrG-3 owing to the high content of GO and reactions of zirconium hydroxide with surface OH groups. When the content of GO is very small, the GO layers in parallel positions are still seen but between them the large units of hydrous zirconia are present. This process of composite formation is visualized in Figure 10, where zirconium

Adsorption of SO2 on Graphite Oxide/Zr(OH)4

J. Phys. Chem. C, Vol. 114, No. 34, 2010 14559

Figure 10. Schematic view of composite formation via interactions of hydrous zirconia units with epoxy (A) and hydroxyl (B) groups of GO. N represents the large number of hydrous zirconia structural units.

Figure 11. SEM micrographs for selected samples.

hydroxide units are linked to GO layers via reactions with surface oxygen groups. Conclusions The results presented in this work demonstrate the synthesis of hydrous zirconia/graphene composites. As a result of the chemical synergy, new pores and new surface chemistry are formed, which leads to the enhancement in sulfur dioxide

removal capacity under both dry and moist conditions. The addition of a graphene component makes the surface less competitive toward adsorption of water. It is hypothesized that the composites are formed via involvement of OH groups of the graphene layers bridging with Zr(OH)4 lattice and via reactions of zirconium with acidic groups present on the edges of the graphene layers. SO2 is retained on the surface via physical adsorption in small pores and via reactions with

14560

J. Phys. Chem. C, Vol. 114, No. 34, 2010

terminal -OH groups of hydrous zirconia. The latter reaction leads to the formation of sulfides. There is an indication that the small amounts of sulfates are also formed because of the oxidation of SO2 by oxygen activated by the specific chemistry of graphene layers. Acknowledgment. This study was supported by the ARO (Army Research Office) Grant No. W911NF-05-1-0537 and NSF Collaborative Grant No. 0754945/0754979. The authors are grateful to Dr. Jacek Jagiello for the SAIEUS program. References and Notes (1) Guo, J. Lua, A. C. J. Chem. Technol. Biotechnol. 2000, 75, 971. (2) Davini, P. Carbon 2001, 39, 2173. (3) Bagreev, A.; Bashkova, S.; Bandosz, T. J. Langmuir 2002, 18, 1257. (4) Raymundo-Pin˜ero, E.; Cazorla-Amoro´s, D.; Salinas-Martinez de Lecea, C.; Linares-Solano, A. Carbon 2000, 38, 335. (5) Moreno-Castilla, C.; Carrasco-Marin, F.; Utrera-Hidalgo, E.; RiveraUtrilla, J. Langmuir 1993, 9, 1378. (6) Lisovskii, A.; Semiat, R.; Aharoni, C. Carbon 1997, 35, 1639. (7) Molina-Sabio, M.; Mun˜ecas, A. M. A.; Rodrı´guez-Reinoso, F.; McEnaney, B. Carbon 1995, 33, 1777. (8) Lizzio, A. A.; DeBarr, J. A. Energy Fuels 1997, 11, 284. (9) Lo´pez, D.; Buitrago, R.; Sepu´lveda-Escribano, A.; Rodrı´guezReinoso, F.; Mondrago´n, F. J. Phys. Chem. C 2007, 111, 1417. (10) Mochida, I.; Miyamoto, S.; Kuroda, K.; Kawano, S.; Yatsunami, S.; Korai, Y. Energy Fuels 1999, 13, 369. (11) Daley, M. A.; Mangun, C. L.; DeBarrb, J. A.; Riha, S.; Lizzio, A. A.; Donnals, G. L.; Economy, J. Carbon 1997, 35, 411. (12) Mochida, I.; Korai, Y.; Shirahama, M.; Kawano, S.; Hada, T.; Seo, Y.; Yoshikawa, M.; Yasutake, A. Carbon 2000, 38, 227. (13) Lee, J. K.; Shim, H. J.; Lim, J. C.; Choi, G. J.; dae Kim, Y.; Min, B. G.; Park, D. Carbon 1997, 35, 837. (14) Ho, C. S.; Shih, S. M. Ind. Eng. Chem. Res. 1992, 31, 1130. (15) Bashkova, S.; Bagreev, A.; Locke, D. C.; Bandosz, T. J. EnViron. Sci. Technol. 2001, 35, 3263. (16) Srinivasan, A.; Grutzeck, M. W. EnViron. Sci. Technol. 1999, 33, 1464. (17) Kopac¸, T.; Kocaba, S. Chem. Eng. Process. 2002, 41, 223. (18) Goodman, A. L.; Li, P.; Usher, C. R.; Grassian, V. H. J. Phys. Chem. A 2001, 105, 6109. (19) Baltrusaitis, J.; Cwiertny, D. M.; Grassiano, V. H. Phys. Chem. Chem. Phys. 2007, 79, 5542. (20) Peterson, G. W.; Karwacki, C. J.; Feaver, W. B.; Rossin, J. A. Ind. Eng. Chem. Res. 2009, 48, 1694. (21) Llanos, J. L.; Fertitta, A. E.; Flores, E. S.; Bottani, E. J. J. Phys. Chem. B 2003, 107, 8448. (22) Seredych, M.; Portet, C.; Gogotsi, Y.; Bandosz, T. J. J. Colloid Interface Sci. 2009, 330, 60. (23) Bandosz, T. J.; Petit, C. J. Colloid Interface Sci. 2009, 338, 329. (24) Wang, Z. M.; Kaneko, K. J. Phys. Chem. B 1998, 102, 2863. (25) Matviya, T. M.; Hayden, R. A. Catalytic Carbon. U.S. Patent 5,356,849, 1994. (26) Carabineiro, S. A. C.; Ramos, A. M.; Vital, J.; Loureiro, J. M.; Orfao, J. J. M.; Fonseca, I. M. Catal. Today 2003, 78, 203.

Seredych and Bandosz (27) Morishige, K.; Hamada, T. Langmuir 2005, 21, 6277. (28) Bashkova, S.; Bandosz, T. J. Ind. Eng. Chem. Res. 2009, 48, 10884. (29) Petit, C.; Bandosz, T. J. J. Phys. Chem. C 2009, 113, 3800. (30) Wang, Z. M.; Shishibori, K.; Hoshinoo, K.; Kanoh, H.; Hirotsu, T. Carbon 2006, 44, 2479. (31) Seredych, M.; Bandosz, T. J. J. Colloid Interface Sci. 2008, 324, 25. (32) Petit, C.; Bandosz, T. J. AdV. Mater. 2009, 21, 4753. (33) Petit, C.; Bandosz, T. J. J. Mater. Chem. 2009, 19, 6521. (34) Hummers, W. S.; Offeman, R. E. J. Am. Chem. Soc. 1958, 80, 1339. (35) Dubinin, M. M. Porous structure and adsorption properties of active carbons. In Chemistry and Physics of Carbon; Walker, P. L., Ed.; Marcel Dekker: New York, 1966; Vol. 2, pp 51-120. (36) Jagiello, J.; Bandosz, T. J.; Schwarz, J. A. Carbon 1994, 32, 1026. (37) Jagiello, J. Langmuir 1994, 10, 2778. (38) Elsayed, Y.; Seredych, M.; Dallas, A.; Bandosz, T. J. Chem. Eng. J. 2009, 155, 594. (39) Buchsteiner, A.; Lerf, A.; Pieper, J. J. Phys. Chem. B 2006, 110, 22328. (40) Seredych, M.; Bandosz, T. J. Langmuir 2010, 26, 5491. (41) Seredych, M.; Bandosz, T. J. J. Phys. Chem. C 2007, 111, 15596. (42) Chitrakar, R.; Tezuka, S.; Sonoda, A.; Sakane, K.; Ooi, K.; Hirotsu, T. J. Colloid Interface Sci. 2006, 297, 426. (43) Guo, G. Y.; Chen, Y. L.; Ying, W. J. Mater. Chem. Phys. 2004, 84, 308. (44) Lerf, A.; He, H.; Forster, M.; Klinowski, J. J. Phys. Chem. B 1998, 102, 4477. (45) Stankovich, S.; Dikin, D. A.; Piner, R. D.; Kohlhaas, K. A.; Kleinhammes, A.; Jia, Y.; Wu, Y.; Nguyen, S. T.; Ruoff, R. S. Carbon 2007, 45, 1558. (46) Stefanic, G.; Popovic, S.; Music, S. Thermochim. Acta 1995, 259, 225. (47) Juan, J. C.; Zhang, J.; Jiang, Y.; Cao, W.; Yarmo, M. A. J. Mol. Catal. A: Chem. 2007, 272, 91. (48) Stohr, B.; Boehm, H. P. Carbon 1991, 29, 707. (49) Strelko, V. V.; Kartel, N. T.; Dukhno, I. N.; Kuts, V. S.; Clarkson, R. B.; Odintsov, B. M. Surf. Sci. 2004, 548, 281. (50) Weast, R. C.; Melvin, J. A. Handbook of Chemistry and Physics, 62nd ed.; CRC Press: Boca Raton FL, 1981-1982. (51) Szabo, T.; Berkesi, O.; Dekany, I. Carbon 2005, 43, 3186. (52) Hontoria-Lucas, C.; Lopez-Peinado, A. J.; Lopez-Gonzalez, J.; de, D.; Rojas-Cervantes, M. L.; Martin-Aranda, R. M. Carbon 1995, 33, 1585. (53) Szabo, T.; Berkesi, O.; Forgo, P.; Josepovits, K.; Sanakis, Y.; Petridis, D.; Dekany, I. Chem. Mater. 2006, 18, 2740. (54) Jeong, H. K.; Lee, Y. P.; Lahaye, R. J. W. E.; Park, M. H.; An, K. H.; Kim, I. J.; Yang, C. W.; Park, C. Y.; Ruoff, R. S.; Lee, Y. H. J. Am. Chem. Soc. 2008, 130, 1362. (55) Ali, A. A. M.; Zaki, M. I. Coll. Surf. A 1998, 139, 81. (56) Sohn, J. R.; Seo, D. H. Catal. Today 2003, 87, 219. (57) Siriwardane, R. V.; Woodruff, S. Ind. Eng. Chem. Res. 1997, 36, 5277. (58) Bandosz, T. J.; Jagiello, J.; Schwarz, J. A. J. Phys. Chem. 1995, 99, 13522. (59) Petit, C.; Seredych, M.; Bandosz, T. J. J. Mater. Chem. 2009, 19, 9176. (60) Wang, Z. M.; Kaneko, K. J. Phys. Chem. 1995, 99, 16714.

JP1051479