Applied Surface Science 275 (2013) 369–373
Contents lists available at SciVerse ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Photocatalytic reactions of nanocomposite of ZnS nanoparticles and montmorillonite P. Praus ∗ , M. Reli, K. Koˇcí, L. Obalová VSˇ B – Technical University of Ostrava, 17, Listopadu 15, 708 33 Ostrava, Czech Republic
a r t i c l e
i n f o
Article history: Received 23 September 2012 Received in revised form 26 November 2012 Accepted 28 November 2012 Available online 20 December 2012 Keywords: ZnS nanoparticles Nanocomposite Montmorillonite Photocatalysis
a b s t r a c t ZnS nanoparticles stabilized by cetyltrimethylammonium bromide (CTAB) were deposited on montmorillonite (MMT) forming a ZnS–CTA–MMT nanocomposite. The nanocomposite was characterized by scanning electron microscopy (SEM), Fourier transformed infrared (FTIR) and UV diffuse reﬂectance spectra (DRS) spectrometry, X-ray powder diffraction (XRD) and speciﬁc surface area measurements. Thereafter, it was used for photocatalytic reactions under UV irradiation (Hg lamp) in three different reaction media with different pH: NaOH solution, HCl solution and water. Prior to the photocatalytic reactions the dispersions were saturated by carbon dioxide to buffer the systems. The main reaction products in gas phase determined by gas chromatography were hydrogen and methane. The reactions were monitored by measuring oxidation–reduction potentials. The highest yields of hydrogen were obtained in the dispersion acidiﬁed by HCl but the concentrations of methane were similar in all tested media. Hydrogen was supposed to be formed by the reaction of two hydrogen radicals. Methane was formed by the reduction of carbon dioxide and by the partial decomposition of CTAB. © 2012 Elsevier B.V. All rights reserved.
1. Introduction After the success of Fujishima and Honda in 1972 , when they photocatalytically split water into hydrogen and oxygen using a single-crystal TiO2 , many research papers devoted to heterogeneous photocatalysis have been published. There are several comprehensive overviews that summarize principles and applications of semiconductor photocatalysis (e.g. [2–6]). Generally, if semiconductors absorb a quantum of light (h) electrons are excited to the conduction band and holes remain in the valence band. Some electrons and holes interact with each other (recombination) and some of them migrate to the semiconductor surface where they react with adsorbed compounds. Their reaction mechanisms are described in details by, e.g. Ohtani . In our previous papers we investigated preparation of ZnS nanoparticles stabilized by a cationic surfactant CTAB and montmorillonite used as an inorganic support [7–10]. Montmorillonite and the ZnS nanoparticles formed a ZnS–CTA–MMT nanocomposite that exhibited photocatalytic activity for CO2 reduction  and decomposition of phenol . In this paper we aimed to compare the inﬂuence of different reaction mixtures on yields of the CO2 photocatalytic reduction. Three aqueous media with different pH were compared: 0.2 mol L−1
∗ Corresponding author. Tel.: +420 597321572. E-mail address: [email protected]
(P. Praus). 0169-4332/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apsusc.2012.11.155
sodium hydroxide with initial pH of 12.2; distilled water with initial pH of 6.7 and 0.001 mol L−1 hydrochloric acid with initial pH of 4.3. The photocatalytic experiments were performed with the nanocomposite of ZnS nanoparticles and montmorillonite [7,8]. 2. Experimental 2.1. Preparation and characterization of ZnS nanocomposite The ZnS nanoparticles were precipitated by zinc acetate and Na2 S in the presence of a stabilizing cationic surfactant cetyltrimethylamonnium bromide [7,8]. The nanocomposite of these ZnS nanoparticles and montmorillonite was prepared by adding montmorillonite into the ZnS nanodispersion and shaking for 24 h. The solid part was ﬁltered out and dried at 105 ◦ C and stored for photocatalytic experiments. The UV diffuse reﬂectance spectra of granulated ZnS–CTA–MMT (diameter of 0.25–0.50 mm) were recorded using a GBC CINTRA 303 spectrometer equipped with a spectralon-coated integrating sphere against a spectralon reference. The Kubelka–Munk function was calculated from the reﬂectance using the following formula: F(R) = (1 − R)2 /2R, where F(R) is the Kubelka–Munk function and R is the reﬂectance. Scanning electron microscopy was performed with an XL 30 Philips SEM instrument (Netherlands) equipped with a Robinson backscattered electron detector, which was used for examination
P. Praus et al. / Applied Surface Science 275 (2013) 369–373
of the nanocomposite morphology and zinc and sulphur distribution on the montmorillonite surface. The powdered samples were coated with gold and palladium in an ionization chamber before the examination. The elemental analysis of ZnS–CTA–MMT was performed by an EDAX detector. FTIR spectra were obtained by the KBr (Mid IR) and polyethylene (Far IR) tablet methods using a Nicolet NEXUS 470 Fourier transform spectrometer (ThermoNicolet, USA). For each spectrum, 64 scans were obtained with a resolution of 8 cm−1 . The recorded FTIR spectra were normalized for the same weight of MMT and processed by means of the program OMNIC 7.3. Speciﬁc surface area of MMT and the nanocomposite was measured with an instrument Sorptomatic 1990 (Thermo Electron Corporation, USA) using nitrogen as an adsorbing gas and calculated by the Advance Data Processing software according to the BET isotherm at a temperature of 77.3 K and p/p0 ratio of up to 0.3. 2.2. Photocatalytic reactivity experiments The photocatalytic reactions were carried out in a stirred batch annular reactor equipped with an on-line glass electrode for monitoring of pH, temperature and pressure. 0.1 g of the ZnS–CTA–MMT photocatalyst was suspended in 100 mL of the reaction medium and poured into the reactor. The volume of gas phase above liquid phase was 280 mL. Afterward the dispersion was saturated by pure carbon dioxide for 30 min until a constant pH was reached. The reaction was started by switching on the 8 W Hg UV lamp (254 nm). The temperature and pressure in the reactor were rapidly equilibrated at 30 ◦ C and 110 kPa, respectively. Gas products were analyzed by a gas chromatograph equipped with ﬂame ionization and thermal conductivity detectors. The relative standard deviation (RSD) of the CH4 analysis was better than 10%, while RSD of the H2 analysis was about 20%. The details of photocatalytic experiments and analytical methods were described in the previous work . It is important to minimize the inﬂuence of transport phenomena during kinetic measurements. The elimination of CO2 diffusion from the bulk of gas through the gas-liquid interface in the laboratory batch slurry reactor was accomplished by saturating the liquid phase with pure CO2 before the reaction was started [12,13]. The suitable volume of liquid phase in our annular photoreactor to fulﬁl the requirement of perfect mixing was found at 100 mL . 3. Results and discussion 3.1. Preparation of ZnS nanocomposite The ZnS nanoparticles were precipitated by mixing zinc acetate with sodium sulphide. The sulphide was added in excess (see Section 2) and thus sulphide ions were adsorbed on the nanoparticles surface according to the Paneth–Fajans rule causing their negative surface charge. The cationic surfactant CTAB was used to prevent their coagulation. The CTA cations were arranged around the nanoparticles forming positively charged ZnS–CTA micelles, which were created by ZnS cores and CTA bilayers formed on their surfaces. This was observed by measurements of their zeta-potential and sizes by the dynamic light scattering method and TEM and also conﬁrmed by molecular modelling . The mean size of the ZnS nanoparticles of about 5 nm was calculated from the band-gap energy of the ZnS–CTA dispersions and also evaluated from the TEM micrographs . The resulting nanodispersion was then mixed with a clay mineral montmorillonine to immobilize the ZnS nanoparticles on a solid and inert carrier. As observed earlier, the ZnS nanoparticles
Fig. 1. DRS spectra of ZnS nanoparticles and bulk particles on montmorillonite.
ﬁlled MMT pores and after drying at 105 ◦ C they stayed enclosed inside them forming the nanocomposite ZnS–CTA–MMT . 3.2. Characterization of ZnS nanocomposite The UV absorption spectra of the ZnS-CTA dispersions were shown in our previous paper [7,8]. They exhibited the absorption edge at 320 nm (3.88 eV). Here, the DRS spectra of the ZnS–CTA–MMT nanocomposite and bulk ZnS precipitated with no stabilization and deposited on MMT (ZnS(bulk)–MMT) were used to estimate the ZnS band-gap energies (Fig. 1). As demonstrated in Fig. 1, the band-gap energies of the ZnS nanoparticles and bulk ZnS particles supported on MMT were 3.69 eV and 3.59 eV, respectively, which is lower than that of 3.70 eV corresponding to bulk ZnS. This effect of the MMT matrix is still unclear. The SEM micrograph (Fig. 2a) demonstrates morphology of ZnS–CTA–MMT and location of ZnS nanoparticles on MMT. The SEM device equipped with the EDAX detector was able to map the location of zinc and sulphur on the ZnS–CTA–MMT surface. From Fig. 2a it is obvious that the ZnS nanoparticles were equally distributed over the examined surface. The EDAX analysis found the presence of zinc, sulphur and carbon conﬁrming the presence of ZnS and CTA. Other elements were the components of montmorillonite. The prepared nanocomposite was also characterized by FTIR spectrometry. Fig. 3 shows the IR spectra of ZnS–CTA–MMT and ZnS(bulk)–MMT. The stretching vibrations of C H bonds at 2922 cm−1 (asym. CH2 ) and 2851 cm−1 (sym. CH2 ) conﬁrmed the presence of CTA in the nanocomposite. The large bands at higher wavenumbers correspond to the stretching O-H vibrations in MMT octahedra (3629 cm−1 ) and the stretching O-H vibrations of adsorbed water (3420 cm−1 ). It can be seen that the content of adsorbed water in the nanocomposite was lower than that in ZnS(bulk)–MMT because CTA made the MMT surface hydrophobic. Strong stretching vibrations of Si O bonds at 1045 cm−1 andbending vibrations at 520 cm−1 and 465 cm−1 corresponding to Al O Si and Si O Si bonds, respectively, were found as well. The absorption band at 1470 cm−1 corresponds to the C N stretching vibration of CTA. The spectra in the far IR region (not shown here) were also recorded and weak vibration bands at 281 cm−1 of ZnS were identiﬁed. X-Ray powder diffraction was used to recognize where the ZnS nanoparticles were located in the nanocomposite . The MMT interlayer distance increased from 1.23 nm to 1.83 nm, which indicates that CTA was intercalated between the MMT layers. Therefore,
P. Praus et al. / Applied Surface Science 275 (2013) 369–373
Fig. 2. (a) SEM micrograph of ZnS–CTA–MMT. EDAX maps of zinc and sulphur in ZnS–CTA–MMT; (b) EDAX spectrum of ZnS–CTA–MMT.
the ZnS nanoparticles with the mean size of 5 nm had to be located on the MMT external surface. These XRD results were conﬁrmed by a decrease of speciﬁc surface area of MMT and ZnS–CTA–MMT from 37.1 m2 g−1 to 3.7 m2 g−1 due to blinding the micro- and mesopores of MMT by CTA. The content of ZnS in the nanocomposite of about 7 wt.% was determined by chemical analysis of zinc and sulphur . The CTA content of 30 wt.% was determined by gravimetry. The stability of ZnS–CTA–MMT was veriﬁed by its shaking in water for 24 h. The dispersions were ﬁltered and UV–vis absorption spectra of the ﬁltrates were recorded. No typical absorption edge of ZnS nanoparticles at 320 nm was observed. This indicates that the ZnS nanoparticles were ﬁrmly deposited in the nanocomposite likely in its pores. 3.3. Photocatalytic production of hydrogen
Fig. 3. FTIR spectra of ZnS–CTA–MMT and ZnS(bulk)–MMT.
After the characterization the ZnS nanocomposite was used for photocatalytic reduction of carbon dioxide dissolved in aqueous solutions with different pH. In all tested systems CTA was partially released from the nanocomposite, which was indicated by
P. Praus et al. / Applied Surface Science 275 (2013) 369–373
Table 1 pH values before and after saturation with carbon dioxide. Reaction medium
pH before saturation
pH after saturation
0.001 mol L−1 HCl H2 O 0.2 mol L−1 NaOH
4.30 6.70 12.20
4.65 4.80 6.75
pH during 24 h 4.65–4.69 4.80–5.09 6.75–6.61
moderate foaming of the dispersions. At the reaction temperatures of about 30 ◦ C CTA was supposed to stabilize the nanocomposite particles against their agglomeration. In our previous study, ZnS nanoparticles were found to be slightly (7 wt.%) oxidized to sulphate in the system of 0.2 mol L−1 NaOH . The three reaction dispersions with different initial pHs (Table 1) were used for the photocatalytic reactions. During the ﬁrst 30 min the dispersions were saturated by CO2 to obtain the buffered systems of dissolved carbon dioxide and hydrogencarbonate. After the saturation the photoreactions were started. The initial and ﬁnal pH values before and after the saturation and the pH changes during the photocatalytic reactions are shown in Table 1. The small pH changes between the start and end of the photoreactions indicate good buffering properties of these carbon dioxide systems. Reactions products in the gas phase were analyzed during 24 h. Prevailing products included hydrogen and methane. Their concentrations in dependence on time are demonstrated in Fig. 4a and b. No photocatalytic activity of montmorillonite itself was observed. Hydrogen (Fig. 4a) was the main and also the most desired product due to its further possible utilization. The highest yields of hydrogen were obtained in the dispersion acidiﬁed by HCl. During the ﬁrst 8 h hydrogen was not detected due to the high detection limit of the used gas chromatograph and then its concentrations gradually increased. Methane (Fig. 4b) was obtained in concentrations several times lower than those of hydrogen from the very beginning of the photoreactions. Taking into account RSD (10%) of the gas chromatography, the obtained yields were similar in all tested media. The mechanisms of methane formation are discussed bellow. Fig. 4. (a) Yields of hydrogen during photoreduction of carbon dioxide; (b) yields of methane during photoreduction of carbon dioxide.
3.4. Measurements of oxidation–reduction potentials Oxidation–reduction potentials were measured during the photocatalytic experiments in all tested media to understand mechanisms of hydrogen formation. The measured potentials are demonstrated in Fig. 5. After short stating periods they were stabilized at nearly constant values depending on pH of the reaction dispersions. The interesting aspect is that the highest concentrations of hydrogen were generated in the dispersion acidiﬁed by HCl. Actual oxidation–reduction potentials were about 230 mV and a standard formal oxidation–reduction potential of the H+ /H2 system at pH 7 is −413 mV. Therefore, under these conditions hydrogen could not be formed by reduction of hydrogen ions according to the reaction 2H+ + 2e− H2 and was likely produced by a reaction of two hydrogen radicals as follows: ZnS + h → ZnS + e− + h+
2H2 O + 4h+ O2 + 4H+
H+ + e− → H•
H• + H• → H2
Decomposition of water into hydrogen and hydroxyl radicals could not take place in these experiments because of low energy of UV irradiation at 254 nm .
Methane was formed by reduction of carbon dioxide by hydrogen CO2 (aq) + 4H2 → CH4 (g) + 2H2 O
Fig. 5. Oxidation–reduction potentials of photocatalytic reactions during 24 h.
P. Praus et al. / Applied Surface Science 275 (2013) 369–373
media. By measuring the oxidation–reduction potentials hydrogen was supposed to be formed by a reaction of two hydrogen radicals. Using blank tests, methane was found to be formed by reduction of carbon dioxide dissolved in reaction dispersions and by partial decomposition of CTA. The next research will be focused on preparation of ZnS nanoparticles and their nanocomposites with no stabilizing compounds (surfactants) to investigate their possibility of reducing carbon dioxide. Acknowledgements The ﬁnancial supports of the Grant Agency of the Czech Republic (P107/11/1918), EU project CZ.1.05/2.1.00/03.0069 and Regional Material Technology Research Center in Ostrava (CZ.1.05/2.1.00/01.0040) are gratefully acknowledged. References Fig. 6. Total and blank yields of methane after 24 h.
The hydrogen radicals should also consequently break the CH3 N bonds of CTA adsorbed on the ZnS nanoparticle surface producing methane as follows: H• + CH3 • → CH4
At the used pH values carbon dioxide also exists in the form of hydrogencarbonate but this anion should be repelled from the negatively charged ZnS nanoparticles as well as the montmorillonite particles. Therefore, only hydrated carbon dioxide was supposed to react with hydrogen according to Eq. (5). The mentioned ways of methane formation were veriﬁed by blank tests performed only after 24 h. Blank dispersions were prepared by removal of dissolved carbon dioxide with helium and then the photocatalytic experiments were performed as described above. Obtained methane yields are shown in Fig. 6. Methane obtained from the blank dispersions was most likely formed by radical reactions according to Eq. (6) while methane in the other tested dispersions was formed also by reduction of CO2 with hydrogen. Solubility of hydrogen in water (1.6 mg L−1 ) is lower than that of methane (22 mg L−1 ) therefore no concentrations of hydrogen measured during the ﬁrst 8 h (Fig. 4a) can be explained by its consumptions for the reduction of CO2 in this period. 4. Conclusion ZnS nanoparticles were deposited on montmorillonite forming the ZnS–CTA–MMT nanocomposite that was used for photocatalytic reactions in aqueous dispersions with different pH. The main reaction products in the gas phase were hydrogen and methane. The highest yields of hydrogen were obtained in the dispersion acidiﬁed by HCl but the concentrations of methane were similar in all tested
 A. Fujishima, K. Honda, Electrochemical photolysis of water at a semiconductor electrode, Nature 238 (1972) 37–38.  A. Henglein, Catalysis of photochemical reactions by colloidal semiconductors, Pure and Applied Chemistry 56 (1984) 1215–1224.  A. Mills, S. Le Hunte, An overview of semiconductor photocatalysis, Journal of Photochemistry and Photobiology A 108 (1997) 1–35.  S.D. Bhatkhande, G.V. Pangarkar, A.A.C.M. Beenackers, Photocatalytic degradation for environmental applications – a review, Journal of Chemical Technology and Biotechnology 77 (2001) 102–116.  D. Bahnemann, Photocatalytic water treatment:solar energy applications, Solar Energy 77 (2004) 445–459.  B. Ohtani, Photocatalysis A to Z – what we know and what we do not know in a scientiﬁc sense, Journal of Photochemistry and Photobiology C 11 (2010) 157–178.  O. Kozák, P. Praus, K. Koˇcí, M. Klementová, Preparation and characterization of ZnS nanoparticles deposited on montmorillonite, Journal of Colloid and Interface Science 352 (2010) 244–251. ´ P. Pustková, O. Kozák, Precipitation of ZnS nanoparticles  P. Praus, R. Dvorsky, and their deposition on montmorillonite, Advanced Science, Engineering and Medicine 3 (2011) 113–118. ´ P. Horínková, M. Pospíˇsil, P. Kováˇr, Precipitation, sta P. Praus, R. Dvorsky, bilization and molecular modelling of ZnS nanoparticles in the presence of cetyltrimethylammonium, Journal of Colloid and Interface Science 377 (2012) 58–63.  P. Praus, J. Matys, O. Kozák, Critical evaluation of montmorillonite catalytic activity by means of photodecomposition of phenol, Journal of the Brazilian Chemical Society 23 (2012) 1900–1906, PII:S0103-505320120050000063. ˇ ´ J. Jirkovsky, ´ O. Solcová, Effect  K. Koˇcí, L. Obalová, L. Matˇejová, D. Plachá, Z. Lacny, of TiO2 particle size on the photocatalytic reduction of CO2 , Applied Catalysis B 89 (2009) 494–502.  I.H. Tseng, W-C. Chang, J.C.S. Wu, Photoreduction of CO2 using sol-gel derived titania and titania-supported copper catalysts, Applied Catalysis B 37 (2002) 37–48.  H. Inoue, T. Torimoto, T. Sakata, H. Mori, H. Yoneyama, Effect of size quantization of zinc sulphide microcrystallites on photocatalytic reduction of carbon dioxide, Chemistry Letters 9 (1990) 1483–1486. ´ Effect of temperature, pressure and  K. Koˇcí, L. Obalová, D. Plachá, Z. Lacny, volume of reacting phase on photocatalytic CO2 reduction on suspended nanocrystalline TiO2 , Collection of Czechoslovak Chemical Communications 73 (2008) 1192–1204.  T. Alapi, A. Dombi, Comparative study of the UV and UV/VUV-induced photolysis of phenol in aqueous solution, Journal of Photochemistry and Photobiology A 188 (2007) 409–418.