Journal of Colloid and Interface Science 436 (2014) 296–305
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Zinc (hydr)oxide/graphite oxide/AuNPs composites: Role of surface features in H2S reactive adsorption Dimitrios A. Giannakoudakis, Teresa J. Bandosz ⇑ Department of Chemistry and the Graduate School of CUNY, The City College of New York, 160 Convent Ave, New York, NY 10031, USA
a r t i c l e
i n f o
Article history: Received 22 May 2014 Accepted 22 August 2014 Available online 1 September 2014 Keywords: Reactive adsorption Hydrogen sulfide Desulfurization Zinc hydroxide Graphite oxide Gold nanoparticles
a b s t r a c t Zinc hydroxide/graphite oxide/AuNPs composites with various levels of complexity were synthesized using an in situ precipitation method. Then they were used as H2S adsorbents in visible light. The materials’ surfaces were characterized before and after H2S adsorption by various physical and chemical methods (XRD, FTIR, thermal analysis, potentiometric titration, adsorption of nitrogen and SEM/EDX). Significant differences in surface features and synergistic effects were found depending on the materials’ composition. Addition of graphite oxide and the deposition of gold nanoparticles resulted in a marked increase in the adsorption capacity in comparison with that on the zinc hydroxide and zinc hydroxide/ AuNP. Addition of AuNPs to zinc hydroxide led to a crystalline ZnO/AuNP composite while the zinc hydroxide/graphite oxide/AuNP composite was amorphous. The ZnOH/GO/AuNPs composite exhibited the greatest H2S adsorption capacity due to the increased number of OH terminal groups and the conductive properties of GO that facilitated the electron transfer and consequently the formation of superoxide ions promoting oxidation of hydrogen sulfide. AuNPs present in the composite increased the conductivity, helped with electron transfer to oxygen, and prevented the fast recombination of the electrons and holes. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Removal of toxic and odorous substances from the atmosphere is of public concern since some of those compounds cause either short or long-term environmental damage to human health and infrastructure. Hydrogen sulfide is a toxic gas present in the atmosphere as a result of anaerobic digestion, volcanic gas emission and various anthropogenic activities. Its oxidation in the atmosphere to SO2 results in an acid rain formation. Among the methods that have been used to remove H2S from air [1] or syngas [2,3], adsorption on surfaces of various adsorbents is preferred owing to low energy requirements and a relative small waste generation. Examples of the materials that are used as separation media include zeolites [4], modified alumina [5], metal oxides [6] and carbonaceous materials [7,8]. Since H2S is a small molecule with a low heat of physical adsorption, reactive adsorption, i.e. retention of molecules on the surface followed by chemical reactions such as acid–base interactions, ion exchange, redox and complexation, is expected to be the most effective method enhancing its retention on the surface, especially at
⇑ Corresponding author. Fax: +1 212 650 6107. E-mail address: [email protected] (T.J. Bandosz). http://dx.doi.org/10.1016/j.jcis.2014.08.046 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
ambient conditions [1,7]. Zinc oxide is known to be an efficient adsorbent of H2S at high temperatures [9] as well as at low temperatures [10]. Recently it has been shown that composites of zinc hydroxide with graphite oxide (GO), further improved the adsorption of hydrogen sulfide [11,12]. This improvement was attributed to the increase in the number and high dispersion of active terminal OH groups of GO that participated in oxidation reactions, as well as to the formation of superoxide ions via the photochemical path [13]. Gold nanoparticles, AuNPs, are of great interest owing to their stability in various matrices (ceramic, polymers, films porous solids), biocompatibility and catalytic properties. Since AuNP exhibit optoelectronic properties, similar to those of quantum dots [14] their involvement in the composite/reactive adsorbents can enhance the efficiency of surface reactions affecting the gas separation processes. Insertion of gold nanoparticles to GO/zinc oxide– hydroxide composites can further improve the H2S uptake by improving the material photocatalytic activity. The effect of AuNPs on photoactivity was shown by the degradation of dyes over graphene–gold nanocomposites [15], or reduction of 4-nitrophenol over TiO2/ZnO/Au [16] or over an Au–GO–Fe3O4 nanocomposite support [17] as well as by CO oxidation reaction using a bed reactor with Au/ZnO nanoparticles [18,19]. The AuNPs were also indicted as promoting the formation of electron and holes and preventing
D.A. Giannakoudakis, T.J. Bandosz / Journal of Colloid and Interface Science 436 (2014) 296–305
their recombination [10,11]. Among many synthesis methodologies, AuNPs can be obtained in a form of a stable colloidal dispersion from tetrachloroaureic (III) acid (HAuCl4), by a reduction of gold (III) with tri-sodium citrate dihydrate [20–22]. Up to now, to the best of our knowledge, a composite material consisting of zinc hydroxide–oxide phase, graphite oxide (GO) and gold nanoparticles (AuNP) has not been synthesized with the intend of its application in reactive adsorption and specifically for a gas phase desulfurization. Thus, the objective of this work is an investigation of H2S reactive adsorption capability of a new multifunctional composite material consisting of three active components: transition metal hydroxide, graphite oxide and AuNP. It is expected that the synergistic effect of the highly reactive surface of GO and the specific bonds (coordination chemistry) between zinc hydroxide [23], along with AuNP functionalization will enhance the reactive uptake of H2S. The performance of the new material is linked to the chemical and structural surface features and especially to the synergistic effects of the composite formation. Based on this, the reactive adsorption mechanism is proposed.
2. Experimental 2.1. Materials All chemicals used in this research (HAuCl4, ZnCl2, NaOH, trisodium citrate dehydrate, and 3-mercapto-propionic acid) were reagent grade with a purity degree higher than 99%. The solutions were prepared in deionized water. Graphite oxide (GO) was prepared by oxidation of commercial graphite (Sigma–Aldrich) by the Hummers method [24]. Details of GO preparation are presented elsewhere [25]. For the synthesis of AuNPs, 20 mL aqueous solution of 1 mM tetrachloroaureic (III) acid (HAuCl4) was added in a 50 mL scratch-free flask on a stirring hot plate with reflux. The solution was heated to boil under vigorous stirring. Then 2 mL of a 1% solution of tri-sodium citrate dihydrate, Na3C6H5O72H2O and sodium 3-mercaptopropioneate (MPA-Na) was injected dropwise. The gold nanoparticles were gradually formed as the citrate reduced gold (III). The suspension was heated for 5 min, and the color of the colloid solution turned from yellow to red [20,22]. The 3-mercaptopropionate acid was neutralized with NaOH. Zinc hydroxide (ZnOH) was prepared using 1 L sodium hydroxide solution (0.05 M), that was added dropwise to a 0.5 L of zinc chloride (0.05 M) solution at a rate of 2.0 mL per minute, using a Titronic Universal (SCHOTT) [24]. The obtained material was filtered and washed with distilled water until no chloride was detected in the washing water. The ZnOH prepared was filtered and dried at 60 °C for 48 h. The composite of zinc hydroxide with GO (referred to as ZnGO) was prepared by dispersing GO (20 wt.% of the final mass of the material) in 0.5 L of zinc chloride (0.05 M) solution following the same procedure as in the preparation of ZnOH. The 20 wt.% was chosen in order a comparison to be made with previous findings where the composite of zinc hydroxide with GO without the addition of Au nanoparticles was examined [23]. Composite materials containing gold nanoparticles were synthesized by dispersing the AuNPs colloid solution (37.5 mL, 103 M) in the starting ZnCl2 solution in the presence or absence of GO. After 4 h stirring, sodium hydroxide solution, (0.05 M) was added dropwise at a rate of 2.0 mL per minute. After washing with distilled water, the composites were centrifuged for 30 min in 4 °C at 6000 rpm. An obtained gel phase was dried at 60 °C for 48 h. The zinc (hydr)oxide with AuNPs is referred to as AuZnOH and the composite with zinc (hydr)oxide, graphite oxide and AuNPs as AuZnGO.
297
3. Methods 3.1. H2S breakthrough capacity measurements To determine the adsorption capacities of the materials studied, dynamic breakthrough tests were performed at room temperature at ambient light. In a typical test, a flow of H2S diluted with moist (humidity 75%) air was passed through a fixed bed of an adsorbent (in a glass tube of 9 mm in diameter) with a total inlet flow rate of 500 mL/min for H2S with an initial concentration of 1000 ppm. Before the H2S adsorption experiments, the materials were prehumidified for 2 h by exposing them to air flow with 75% relative humidity. The adsorbent’s bed volume was about 1.4–1.6 cm3 and the particles had size between 0.4 and 1 mm. The concentrations of H2S and any SO2 produced in the adsorbent bed were measured in the outlet gas using an electrochemical sensor (Multi-Gas Monitor, RAE system). Once the outlet concentration reached the maximum detection limit of the sensor, the H2S feeding was immediately closed and the desorption of weakly adsorbed hydrogen sulfide was monitored. The experiments were repeated at least twice and the standard deviations were less than 15%. The adsorption capacity of each adsorbent was calculated in milligrams per gram of the material by integrating the area above the breakthrough curve. The exhausted materials are denoted with letter E.
3.2. FT-IR spectroscopy 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 64 times and corrected for the background noise.
3.3. XRD X-ray diffraction (XRD) measurements were conducted using powder diffraction procedures. Adsorbents were grounded to powder size and then analyzed by a CuKa radiation generated in a Phillips X’Pert X-ray diffractometer.
3.4. Thermal analysis Thermogravimetric (TG) curves were obtained using a TA Instruments Thermal Analyzer (New Castle, DE, USA). The heating rate was 10 °C min1 in a helium atmosphere at 100 mL min1 flow rate. The samples were heated up to 1000 °C. From the TG curves, differential TG (DTG) curves were derived.
3.5. Potentiometric titration Potentiometric titration measurements were performed with a DMS Titrino 716 automatic titrator (Metrohm). Subsamples of the initial and exhausted materials (0.050 g) were added to NaNO3 (0.01 M, 25 mL) and placed in a container maintained at 25 °C overnight for equilibrium. During the titration, the suspension was continuously saturated with N2 to eliminate the influence of atmospheric CO2. The suspension was stirred throughout the measurements. Volumetric standard NaOH (0.1 M) was used as the titrant starting from the initial pH of the material suspension up to pH 11. The experimental data was transformed into a proton-binding curve, Q, representing the total number of protonated sites [13,26]. These curves were deconvoluted using the SAIEUS procedure [13], and the pKa distributions for the species present on the surface were obtained.
298
D.A. Giannakoudakis, T.J. Bandosz / Journal of Colloid and Interface Science 436 (2014) 296–305
3.6. Sorption of nitrogen Nitrogen isotherms were measured using an ASAP 2020 (Micromeritics, Surface area and Porosity Analyzer Norcross, GA, USA) at 196 °C. The samples were outgassed at 120 °C to vacuum 104 Torr before the measurements. The surface area, SBET (Brunauer–Emmet–Teller method was used), the micropore volume, Vmic (calculated using the Dubinin–Radushkevich approach) the mesopore volume, Vmes, the total pore volume, Vt (calculated from the last point of the isotherms based on the volume of nitrogen adsorbed) were calculated from the isotherms. The volume of mesopores, Vmes, represents the difference between the total pore volume and the micropore volume. 3.7. SEM/EDX Scanning electron microscopy images were obtained using a Zeiss Supra 55 VP with an accelerating voltage of 5.00 kV. Scanning was performed in situ on a powder sample. SEM images with energy-dispersive X-ray (EDAX) analysis were obtained in the same instrument. 4. Results and discussion The measured H2S breakthrough curves on our adsorbents, with time normalized by a unit mass, are collected in Fig. 1. The curves for ZnOH, ZnGO and AuZnOH exhibit similar slopes, suggesting similar kinetics of a H2S removal process and/or of a surface reaction mechanism. The addition of AuNPs to the ZnGO composite resulted in a significant improvement in the performance. On the other hand, no positive effect was found in the case of hydroxide modification with AuNPs. Moreover, the slope of the curve for AuZnGO decreased suggesting alterations in the surface reaction mechanism. On all materials studied, the adsorption of hydrogen sulfide was very strong, which is demonstrated by steep desorption curves. No release of SO2 during the breakthrough experiment indicates either the lack of side reactions or adsorption of hydrogen sulfide oxidation products on the surface. The calculated capacities expressed both in milligrams per a unit mass and per a unit volume of an adsorbent bed are listed in Table 1. The ZnGO composite outperforms the zinc hydroxide and the measured capacity increased 30% upon the composite formation. These results are consistent with previous findings addressing the adsorption capacity of the ZnGO [12]. It is important to mention that the performance of our adsorbents is much better than that on virgin activated carbons (up to 20 mg/g) and in the range of that on catalytic carbons (about 100 mg/g) [3].
Fig. 1. H2S breakthrough curves.
Table 1 H2S breakthrough capacities. Sample
ZnOH AuZnOH ZnGO AuZnGO
H2S breakthrough capacity mg/g
mg/cm3
66.4 56.2 87.2 154.7
32.1 50.7 33.6 91.7
X-ray diffraction patterns of the ZnOH, AuZnOH, ZnGO and AuZnGO samples (Fig. 2) show differences in crystal morphologies. In the case of ZnOH and ZnGO, due to the solubility difference of zinc oxide and zinc hydroxide, the initial precipitation rate of Zn(OH)2 is faster than that of ZnO [27] and as a result of this Zn(OH)2 becomes the main phase. Low temperature and low basicity facilitate the nucleation and the growth of Zn(OH)2 [28]. The diffraction peaks at 20.2°, 20.9°, 27.2°, 27.8°, 32.9°, 39.5°, 40.8°, 42.1°, 52.4°, 57.9°, 59.5° and 60.4° can be indexed as those of orthorhombic e-Zn(OH)2 (JCPDS 38-0385) [29]. For the AuZnOH composite, the X-ray diffraction pattern indicates that the main crystallographic phase is ZnO. The diffraction peaks at 31.7°, 34.4°, 36.2° and 47.6° can be linked to a well-crystallized hexagonal wurtzite structure of ZnO (JCPDS 36-1451). Crystals of Zn(OH)2 are also detected. The addition of AuNPs to the AuZnGO composite leads to an entirely different crystallization progress and nucleation mechanism. In the XRD diffraction pattern of AuZnGO a high level of amorphicity is visible with a decreased intensity of the diffraction peak at 33° assigned to ZnO. This decrease could also be attributed to a small size of the ZnO crystals [16]. After H2S adsorption, no differences are found in XRD patterns for ZnOH-E and ZnGO-E suggesting that no new crystalline phases
Fig. 2. X-ray diffraction patterns for the initial and exhausted samples.
D.A. Giannakoudakis, T.J. Bandosz / Journal of Colloid and Interface Science 436 (2014) 296–305
containing sulfur are formed as a result of reactive adsorption. Interestingly, for AuZnOH-E, all crystalline Zn(OH)2 on the surface was involved in the reactive adsorption of H2S, since the diffraction peaks at 40.8°, 42.1° and 52.4° disappeared after the exposure to hydrogen sulfide. For the exhausted AuZnGO-E sample a small peak appeared at 27–29° 2 Theta. This new peak might have its origin in the formed crystalline phase of ZnS [30]. Differences in the chemistry of zinc hydroxides, the composites, and their counterparts exposed to H2S are also seen on FTIR spectra (Fig. 3). On the spectrum for ZnOH the bands at 3455 and 3500 cm1 representing hydroxyl stretching vibrations of the molecules in the unit cell are clearly seen. The bands that originate from vibrations of AOH groups appear at 715, 830, 920, 1040, 1390, and 1500 cm1. The bands at 830 and 1040 cm1 correspond to the ZnAOH bending mode, whereas the band at 920 cm1 is assigned to the out-of-plane bending mode [31]. The strong band at 715 cm1 originates from vibrational modes of the OH groups or corresponds to vibrational modes of the H2O molecules [23,32]. The two peaks observed at 1390 cm1 (with small shoulders at 1360) and 1560 cm1 (with a splitting at 1500 cm1) are assigned to deformation vibrations of OAH groups bound to zinc hydroxide [11]. In the spectra for the ZnGO and AuZnGO composites the bands that originate from bending and deformational vibrations of AOH groups of ZnOH at 715, 920, 1040 cm1 are no longer observed. The bands associated to CAO, C@O bonds in carboxylic and epoxide groups of graphite oxide (GO) [15,26] are also not detected. This is an indication that these oxygen functionalities must have taken part in the formation of new chemical bonds specific for the composites [11]. After exposure to H2S, the intensity of the bands between 700 and 1600 cm1 as well as these at 1390 and 1500 cm1 representing hydroxyl groups significantly decreased for all samples, sug-
299
gesting the involvement of these groups in the reactive adsorption process. Similar results were found by Seredych and coworkers for the reactive adsorption of H2S and SO2 on ZnGO [12,25]. In spite of the large amounts of H2S adsorbed, especially on the composites with GO, the bands from sulfides are not expected to be seen, owing to their low intensity. Thermal analysis results were used to further evaluate the differences in surface chemistry of the initial composite samples, as well as the samples exposed to H2S. DTG curves measured in helium are presented in Fig. 4. For ZnOH, the main peaks are centered at about 150 and 200 °C. They are assigned to the removal of physically adsorbed water and the dehydration of Zn(OH)2, respectively. The shoulder at about 450 °C can be attributed to Zn(OH)2 dehydroxylation [33]. For the AuZnOH sample the peaks assigned to the removal of physically adsorbed water are less intense than those for ZnOH suggesting an increase in surface hydrophobicity. New peaks at 170, 330 and 410 °C are revealed, which we link to the removal of physically adsorbed water, dehydration and dehydroxylation of the small amount of zinc hydroxide present in this sample, respectively. For the ZnOH-E sample, after the exposure to H2S, the peak at a temperature less than 100 °C is attributed to the removal of water. Moreover, new peaks which appear at 150 and 210 °C can be assigned to weakly and strongly adsorbed/ reacted SO2, respectively [34]. The peak at 210 °C can be attributed to ZnSO3 [35]. This suggests that H2S is oxidized to SO2 on the surface. This SO2 has to be strongly adsorbed since its was not detected in the outlet gases. The peaks on the DTG curves located at about 600 °C are assigned to the decomposition of zinc sulfate. Its decomposition temperature/boiling point is expected at 600 °C [10,35]. For the AuZnOH-E sample an increase in the intensity of the peak at the temperature less than 100 °C is found and it is linked to the removal of water and/or physically adsorbed SO2
Fig. 3. FTIR spectra of the initial and exhausted samples.
300
D.A. Giannakoudakis, T.J. Bandosz / Journal of Colloid and Interface Science 436 (2014) 296–305
Fig. 4. DTG curves in helium for the initial and exhausted samples.
[12]. Additionally, the peaks at 150 and 210 °C, can be assigned to weakly and strongly adsorbed SO2, respectively, as in the case of Zn(OH)2-E, while the peak due to the decomposition of zinc sulfate (600 °C) is not found for this composite suggesting limited extent of H2S oxidation. A new peak appears at 350 °C that can be related to the removal of SO2 in the form of oxysulfides [36]. XRD results revealed that the addition of Au nanoparticles resulted in the formation of ZnO instead of Zn(OH)2, that leads to different adsorption products. For ZnGO and AuZnGO, the weight loss occurring at over 800 °C can be due to the further reduction of zinc oxide/ hydroxide by graphite oxide [11]. On the weight loss pattern for the samples after the exposure to H2S, the offset of a high temperature peak is visible at the end of the experimental window. It can be attributed to the zinc sulfide formed, that is expected to decompose at temperatures higher than 1000 °C [33]. In the case of the composite with graphite oxide (ZnGO), the weight loss patterns revealed the peaks centered at 150 and 200 °C, assigned to the removal of physically adsorbed water and the dehydration of Zn(OH)2, respectively. The latter peak presents a smaller intensity comparing to that of ZnOH. We link it to hydrophobicity introduced by the addition of the graphene-based phase. At temperature higher than 400 °C the weight loss related to the dehydroxylation of the zinc hydroxide phase is revealed [25]. Finally, the weight loss occurring at over 800 °C is assigned to the reduction of zinc oxide by the carbonaceous phase. After exposure to H2S, the low temperature peaks decrease significantly in intensity. A new intense peak at 280 °C appeared that might be linked to sulfur/oxysulfur species. It is suggested that water, OH, and O centers of Zn(OH)2 may take part in surface reactions towards sulfur/oxysulfur species [9]. The peak at 650 °C is assigned to the decomposition of zinc sulfates [37].
The AuZnGO sample presents an entirely different weight loss pattern than that for the ZnGO composite. For it the peak representing dehydration at about 150 °C almost disappeared. This suggests that the surface of the composite is the most hydrophobic among all samples tested. At about 250 °C the decomposition of species in which both zinc and oxygen from epoxy groups of GO are involved takes place [25]. After exposure to H2S, on the DTG curve for the AuZnGO-E composite an intense peak at
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Zinc (hydr)oxide/graphite oxide/AuNPs composites: Role of surface features in H2S reactive adsorption Dimitrios A. Giannakoudakis, Teresa J. Bandosz ⇑ Department of Chemistry and the Graduate School of CUNY, The City College of New York, 160 Convent Ave, New York, NY 10031, USA
a r t i c l e
i n f o
Article history: Received 22 May 2014 Accepted 22 August 2014 Available online 1 September 2014 Keywords: Reactive adsorption Hydrogen sulfide Desulfurization Zinc hydroxide Graphite oxide Gold nanoparticles
a b s t r a c t Zinc hydroxide/graphite oxide/AuNPs composites with various levels of complexity were synthesized using an in situ precipitation method. Then they were used as H2S adsorbents in visible light. The materials’ surfaces were characterized before and after H2S adsorption by various physical and chemical methods (XRD, FTIR, thermal analysis, potentiometric titration, adsorption of nitrogen and SEM/EDX). Significant differences in surface features and synergistic effects were found depending on the materials’ composition. Addition of graphite oxide and the deposition of gold nanoparticles resulted in a marked increase in the adsorption capacity in comparison with that on the zinc hydroxide and zinc hydroxide/ AuNP. Addition of AuNPs to zinc hydroxide led to a crystalline ZnO/AuNP composite while the zinc hydroxide/graphite oxide/AuNP composite was amorphous. The ZnOH/GO/AuNPs composite exhibited the greatest H2S adsorption capacity due to the increased number of OH terminal groups and the conductive properties of GO that facilitated the electron transfer and consequently the formation of superoxide ions promoting oxidation of hydrogen sulfide. AuNPs present in the composite increased the conductivity, helped with electron transfer to oxygen, and prevented the fast recombination of the electrons and holes. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Removal of toxic and odorous substances from the atmosphere is of public concern since some of those compounds cause either short or long-term environmental damage to human health and infrastructure. Hydrogen sulfide is a toxic gas present in the atmosphere as a result of anaerobic digestion, volcanic gas emission and various anthropogenic activities. Its oxidation in the atmosphere to SO2 results in an acid rain formation. Among the methods that have been used to remove H2S from air [1] or syngas [2,3], adsorption on surfaces of various adsorbents is preferred owing to low energy requirements and a relative small waste generation. Examples of the materials that are used as separation media include zeolites [4], modified alumina [5], metal oxides [6] and carbonaceous materials [7,8]. Since H2S is a small molecule with a low heat of physical adsorption, reactive adsorption, i.e. retention of molecules on the surface followed by chemical reactions such as acid–base interactions, ion exchange, redox and complexation, is expected to be the most effective method enhancing its retention on the surface, especially at
⇑ Corresponding author. Fax: +1 212 650 6107. E-mail address: [email protected] (T.J. Bandosz). http://dx.doi.org/10.1016/j.jcis.2014.08.046 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
ambient conditions [1,7]. Zinc oxide is known to be an efficient adsorbent of H2S at high temperatures [9] as well as at low temperatures [10]. Recently it has been shown that composites of zinc hydroxide with graphite oxide (GO), further improved the adsorption of hydrogen sulfide [11,12]. This improvement was attributed to the increase in the number and high dispersion of active terminal OH groups of GO that participated in oxidation reactions, as well as to the formation of superoxide ions via the photochemical path [13]. Gold nanoparticles, AuNPs, are of great interest owing to their stability in various matrices (ceramic, polymers, films porous solids), biocompatibility and catalytic properties. Since AuNP exhibit optoelectronic properties, similar to those of quantum dots [14] their involvement in the composite/reactive adsorbents can enhance the efficiency of surface reactions affecting the gas separation processes. Insertion of gold nanoparticles to GO/zinc oxide– hydroxide composites can further improve the H2S uptake by improving the material photocatalytic activity. The effect of AuNPs on photoactivity was shown by the degradation of dyes over graphene–gold nanocomposites [15], or reduction of 4-nitrophenol over TiO2/ZnO/Au [16] or over an Au–GO–Fe3O4 nanocomposite support [17] as well as by CO oxidation reaction using a bed reactor with Au/ZnO nanoparticles [18,19]. The AuNPs were also indicted as promoting the formation of electron and holes and preventing
D.A. Giannakoudakis, T.J. Bandosz / Journal of Colloid and Interface Science 436 (2014) 296–305
their recombination [10,11]. Among many synthesis methodologies, AuNPs can be obtained in a form of a stable colloidal dispersion from tetrachloroaureic (III) acid (HAuCl4), by a reduction of gold (III) with tri-sodium citrate dihydrate [20–22]. Up to now, to the best of our knowledge, a composite material consisting of zinc hydroxide–oxide phase, graphite oxide (GO) and gold nanoparticles (AuNP) has not been synthesized with the intend of its application in reactive adsorption and specifically for a gas phase desulfurization. Thus, the objective of this work is an investigation of H2S reactive adsorption capability of a new multifunctional composite material consisting of three active components: transition metal hydroxide, graphite oxide and AuNP. It is expected that the synergistic effect of the highly reactive surface of GO and the specific bonds (coordination chemistry) between zinc hydroxide [23], along with AuNP functionalization will enhance the reactive uptake of H2S. The performance of the new material is linked to the chemical and structural surface features and especially to the synergistic effects of the composite formation. Based on this, the reactive adsorption mechanism is proposed.
2. Experimental 2.1. Materials All chemicals used in this research (HAuCl4, ZnCl2, NaOH, trisodium citrate dehydrate, and 3-mercapto-propionic acid) were reagent grade with a purity degree higher than 99%. The solutions were prepared in deionized water. Graphite oxide (GO) was prepared by oxidation of commercial graphite (Sigma–Aldrich) by the Hummers method [24]. Details of GO preparation are presented elsewhere [25]. For the synthesis of AuNPs, 20 mL aqueous solution of 1 mM tetrachloroaureic (III) acid (HAuCl4) was added in a 50 mL scratch-free flask on a stirring hot plate with reflux. The solution was heated to boil under vigorous stirring. Then 2 mL of a 1% solution of tri-sodium citrate dihydrate, Na3C6H5O72H2O and sodium 3-mercaptopropioneate (MPA-Na) was injected dropwise. The gold nanoparticles were gradually formed as the citrate reduced gold (III). The suspension was heated for 5 min, and the color of the colloid solution turned from yellow to red [20,22]. The 3-mercaptopropionate acid was neutralized with NaOH. Zinc hydroxide (ZnOH) was prepared using 1 L sodium hydroxide solution (0.05 M), that was added dropwise to a 0.5 L of zinc chloride (0.05 M) solution at a rate of 2.0 mL per minute, using a Titronic Universal (SCHOTT) [24]. The obtained material was filtered and washed with distilled water until no chloride was detected in the washing water. The ZnOH prepared was filtered and dried at 60 °C for 48 h. The composite of zinc hydroxide with GO (referred to as ZnGO) was prepared by dispersing GO (20 wt.% of the final mass of the material) in 0.5 L of zinc chloride (0.05 M) solution following the same procedure as in the preparation of ZnOH. The 20 wt.% was chosen in order a comparison to be made with previous findings where the composite of zinc hydroxide with GO without the addition of Au nanoparticles was examined [23]. Composite materials containing gold nanoparticles were synthesized by dispersing the AuNPs colloid solution (37.5 mL, 103 M) in the starting ZnCl2 solution in the presence or absence of GO. After 4 h stirring, sodium hydroxide solution, (0.05 M) was added dropwise at a rate of 2.0 mL per minute. After washing with distilled water, the composites were centrifuged for 30 min in 4 °C at 6000 rpm. An obtained gel phase was dried at 60 °C for 48 h. The zinc (hydr)oxide with AuNPs is referred to as AuZnOH and the composite with zinc (hydr)oxide, graphite oxide and AuNPs as AuZnGO.
297
3. Methods 3.1. H2S breakthrough capacity measurements To determine the adsorption capacities of the materials studied, dynamic breakthrough tests were performed at room temperature at ambient light. In a typical test, a flow of H2S diluted with moist (humidity 75%) air was passed through a fixed bed of an adsorbent (in a glass tube of 9 mm in diameter) with a total inlet flow rate of 500 mL/min for H2S with an initial concentration of 1000 ppm. Before the H2S adsorption experiments, the materials were prehumidified for 2 h by exposing them to air flow with 75% relative humidity. The adsorbent’s bed volume was about 1.4–1.6 cm3 and the particles had size between 0.4 and 1 mm. The concentrations of H2S and any SO2 produced in the adsorbent bed were measured in the outlet gas using an electrochemical sensor (Multi-Gas Monitor, RAE system). Once the outlet concentration reached the maximum detection limit of the sensor, the H2S feeding was immediately closed and the desorption of weakly adsorbed hydrogen sulfide was monitored. The experiments were repeated at least twice and the standard deviations were less than 15%. The adsorption capacity of each adsorbent was calculated in milligrams per gram of the material by integrating the area above the breakthrough curve. The exhausted materials are denoted with letter E.
3.2. FT-IR spectroscopy 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 64 times and corrected for the background noise.
3.3. XRD X-ray diffraction (XRD) measurements were conducted using powder diffraction procedures. Adsorbents were grounded to powder size and then analyzed by a CuKa radiation generated in a Phillips X’Pert X-ray diffractometer.
3.4. Thermal analysis Thermogravimetric (TG) curves were obtained using a TA Instruments Thermal Analyzer (New Castle, DE, USA). The heating rate was 10 °C min1 in a helium atmosphere at 100 mL min1 flow rate. The samples were heated up to 1000 °C. From the TG curves, differential TG (DTG) curves were derived.
3.5. Potentiometric titration Potentiometric titration measurements were performed with a DMS Titrino 716 automatic titrator (Metrohm). Subsamples of the initial and exhausted materials (0.050 g) were added to NaNO3 (0.01 M, 25 mL) and placed in a container maintained at 25 °C overnight for equilibrium. During the titration, the suspension was continuously saturated with N2 to eliminate the influence of atmospheric CO2. The suspension was stirred throughout the measurements. Volumetric standard NaOH (0.1 M) was used as the titrant starting from the initial pH of the material suspension up to pH 11. The experimental data was transformed into a proton-binding curve, Q, representing the total number of protonated sites [13,26]. These curves were deconvoluted using the SAIEUS procedure [13], and the pKa distributions for the species present on the surface were obtained.
298
D.A. Giannakoudakis, T.J. Bandosz / Journal of Colloid and Interface Science 436 (2014) 296–305
3.6. Sorption of nitrogen Nitrogen isotherms were measured using an ASAP 2020 (Micromeritics, Surface area and Porosity Analyzer Norcross, GA, USA) at 196 °C. The samples were outgassed at 120 °C to vacuum 104 Torr before the measurements. The surface area, SBET (Brunauer–Emmet–Teller method was used), the micropore volume, Vmic (calculated using the Dubinin–Radushkevich approach) the mesopore volume, Vmes, the total pore volume, Vt (calculated from the last point of the isotherms based on the volume of nitrogen adsorbed) were calculated from the isotherms. The volume of mesopores, Vmes, represents the difference between the total pore volume and the micropore volume. 3.7. SEM/EDX Scanning electron microscopy images were obtained using a Zeiss Supra 55 VP with an accelerating voltage of 5.00 kV. Scanning was performed in situ on a powder sample. SEM images with energy-dispersive X-ray (EDAX) analysis were obtained in the same instrument. 4. Results and discussion The measured H2S breakthrough curves on our adsorbents, with time normalized by a unit mass, are collected in Fig. 1. The curves for ZnOH, ZnGO and AuZnOH exhibit similar slopes, suggesting similar kinetics of a H2S removal process and/or of a surface reaction mechanism. The addition of AuNPs to the ZnGO composite resulted in a significant improvement in the performance. On the other hand, no positive effect was found in the case of hydroxide modification with AuNPs. Moreover, the slope of the curve for AuZnGO decreased suggesting alterations in the surface reaction mechanism. On all materials studied, the adsorption of hydrogen sulfide was very strong, which is demonstrated by steep desorption curves. No release of SO2 during the breakthrough experiment indicates either the lack of side reactions or adsorption of hydrogen sulfide oxidation products on the surface. The calculated capacities expressed both in milligrams per a unit mass and per a unit volume of an adsorbent bed are listed in Table 1. The ZnGO composite outperforms the zinc hydroxide and the measured capacity increased 30% upon the composite formation. These results are consistent with previous findings addressing the adsorption capacity of the ZnGO [12]. It is important to mention that the performance of our adsorbents is much better than that on virgin activated carbons (up to 20 mg/g) and in the range of that on catalytic carbons (about 100 mg/g) [3].
Fig. 1. H2S breakthrough curves.
Table 1 H2S breakthrough capacities. Sample
ZnOH AuZnOH ZnGO AuZnGO
H2S breakthrough capacity mg/g
mg/cm3
66.4 56.2 87.2 154.7
32.1 50.7 33.6 91.7
X-ray diffraction patterns of the ZnOH, AuZnOH, ZnGO and AuZnGO samples (Fig. 2) show differences in crystal morphologies. In the case of ZnOH and ZnGO, due to the solubility difference of zinc oxide and zinc hydroxide, the initial precipitation rate of Zn(OH)2 is faster than that of ZnO [27] and as a result of this Zn(OH)2 becomes the main phase. Low temperature and low basicity facilitate the nucleation and the growth of Zn(OH)2 [28]. The diffraction peaks at 20.2°, 20.9°, 27.2°, 27.8°, 32.9°, 39.5°, 40.8°, 42.1°, 52.4°, 57.9°, 59.5° and 60.4° can be indexed as those of orthorhombic e-Zn(OH)2 (JCPDS 38-0385) [29]. For the AuZnOH composite, the X-ray diffraction pattern indicates that the main crystallographic phase is ZnO. The diffraction peaks at 31.7°, 34.4°, 36.2° and 47.6° can be linked to a well-crystallized hexagonal wurtzite structure of ZnO (JCPDS 36-1451). Crystals of Zn(OH)2 are also detected. The addition of AuNPs to the AuZnGO composite leads to an entirely different crystallization progress and nucleation mechanism. In the XRD diffraction pattern of AuZnGO a high level of amorphicity is visible with a decreased intensity of the diffraction peak at 33° assigned to ZnO. This decrease could also be attributed to a small size of the ZnO crystals [16]. After H2S adsorption, no differences are found in XRD patterns for ZnOH-E and ZnGO-E suggesting that no new crystalline phases
Fig. 2. X-ray diffraction patterns for the initial and exhausted samples.
D.A. Giannakoudakis, T.J. Bandosz / Journal of Colloid and Interface Science 436 (2014) 296–305
containing sulfur are formed as a result of reactive adsorption. Interestingly, for AuZnOH-E, all crystalline Zn(OH)2 on the surface was involved in the reactive adsorption of H2S, since the diffraction peaks at 40.8°, 42.1° and 52.4° disappeared after the exposure to hydrogen sulfide. For the exhausted AuZnGO-E sample a small peak appeared at 27–29° 2 Theta. This new peak might have its origin in the formed crystalline phase of ZnS [30]. Differences in the chemistry of zinc hydroxides, the composites, and their counterparts exposed to H2S are also seen on FTIR spectra (Fig. 3). On the spectrum for ZnOH the bands at 3455 and 3500 cm1 representing hydroxyl stretching vibrations of the molecules in the unit cell are clearly seen. The bands that originate from vibrations of AOH groups appear at 715, 830, 920, 1040, 1390, and 1500 cm1. The bands at 830 and 1040 cm1 correspond to the ZnAOH bending mode, whereas the band at 920 cm1 is assigned to the out-of-plane bending mode [31]. The strong band at 715 cm1 originates from vibrational modes of the OH groups or corresponds to vibrational modes of the H2O molecules [23,32]. The two peaks observed at 1390 cm1 (with small shoulders at 1360) and 1560 cm1 (with a splitting at 1500 cm1) are assigned to deformation vibrations of OAH groups bound to zinc hydroxide [11]. In the spectra for the ZnGO and AuZnGO composites the bands that originate from bending and deformational vibrations of AOH groups of ZnOH at 715, 920, 1040 cm1 are no longer observed. The bands associated to CAO, C@O bonds in carboxylic and epoxide groups of graphite oxide (GO) [15,26] are also not detected. This is an indication that these oxygen functionalities must have taken part in the formation of new chemical bonds specific for the composites [11]. After exposure to H2S, the intensity of the bands between 700 and 1600 cm1 as well as these at 1390 and 1500 cm1 representing hydroxyl groups significantly decreased for all samples, sug-
299
gesting the involvement of these groups in the reactive adsorption process. Similar results were found by Seredych and coworkers for the reactive adsorption of H2S and SO2 on ZnGO [12,25]. In spite of the large amounts of H2S adsorbed, especially on the composites with GO, the bands from sulfides are not expected to be seen, owing to their low intensity. Thermal analysis results were used to further evaluate the differences in surface chemistry of the initial composite samples, as well as the samples exposed to H2S. DTG curves measured in helium are presented in Fig. 4. For ZnOH, the main peaks are centered at about 150 and 200 °C. They are assigned to the removal of physically adsorbed water and the dehydration of Zn(OH)2, respectively. The shoulder at about 450 °C can be attributed to Zn(OH)2 dehydroxylation [33]. For the AuZnOH sample the peaks assigned to the removal of physically adsorbed water are less intense than those for ZnOH suggesting an increase in surface hydrophobicity. New peaks at 170, 330 and 410 °C are revealed, which we link to the removal of physically adsorbed water, dehydration and dehydroxylation of the small amount of zinc hydroxide present in this sample, respectively. For the ZnOH-E sample, after the exposure to H2S, the peak at a temperature less than 100 °C is attributed to the removal of water. Moreover, new peaks which appear at 150 and 210 °C can be assigned to weakly and strongly adsorbed/ reacted SO2, respectively [34]. The peak at 210 °C can be attributed to ZnSO3 [35]. This suggests that H2S is oxidized to SO2 on the surface. This SO2 has to be strongly adsorbed since its was not detected in the outlet gases. The peaks on the DTG curves located at about 600 °C are assigned to the decomposition of zinc sulfate. Its decomposition temperature/boiling point is expected at 600 °C [10,35]. For the AuZnOH-E sample an increase in the intensity of the peak at the temperature less than 100 °C is found and it is linked to the removal of water and/or physically adsorbed SO2
Fig. 3. FTIR spectra of the initial and exhausted samples.
300
D.A. Giannakoudakis, T.J. Bandosz / Journal of Colloid and Interface Science 436 (2014) 296–305
Fig. 4. DTG curves in helium for the initial and exhausted samples.
[12]. Additionally, the peaks at 150 and 210 °C, can be assigned to weakly and strongly adsorbed SO2, respectively, as in the case of Zn(OH)2-E, while the peak due to the decomposition of zinc sulfate (600 °C) is not found for this composite suggesting limited extent of H2S oxidation. A new peak appears at 350 °C that can be related to the removal of SO2 in the form of oxysulfides [36]. XRD results revealed that the addition of Au nanoparticles resulted in the formation of ZnO instead of Zn(OH)2, that leads to different adsorption products. For ZnGO and AuZnGO, the weight loss occurring at over 800 °C can be due to the further reduction of zinc oxide/ hydroxide by graphite oxide [11]. On the weight loss pattern for the samples after the exposure to H2S, the offset of a high temperature peak is visible at the end of the experimental window. It can be attributed to the zinc sulfide formed, that is expected to decompose at temperatures higher than 1000 °C [33]. In the case of the composite with graphite oxide (ZnGO), the weight loss patterns revealed the peaks centered at 150 and 200 °C, assigned to the removal of physically adsorbed water and the dehydration of Zn(OH)2, respectively. The latter peak presents a smaller intensity comparing to that of ZnOH. We link it to hydrophobicity introduced by the addition of the graphene-based phase. At temperature higher than 400 °C the weight loss related to the dehydroxylation of the zinc hydroxide phase is revealed [25]. Finally, the weight loss occurring at over 800 °C is assigned to the reduction of zinc oxide by the carbonaceous phase. After exposure to H2S, the low temperature peaks decrease significantly in intensity. A new intense peak at 280 °C appeared that might be linked to sulfur/oxysulfur species. It is suggested that water, OH, and O centers of Zn(OH)2 may take part in surface reactions towards sulfur/oxysulfur species [9]. The peak at 650 °C is assigned to the decomposition of zinc sulfates [37].
The AuZnGO sample presents an entirely different weight loss pattern than that for the ZnGO composite. For it the peak representing dehydration at about 150 °C almost disappeared. This suggests that the surface of the composite is the most hydrophobic among all samples tested. At about 250 °C the decomposition of species in which both zinc and oxygen from epoxy groups of GO are involved takes place [25]. After exposure to H2S, on the DTG curve for the AuZnGO-E composite an intense peak at
