Sei sulla pagina 1di 7
Chemical Engineering Journal 254 (2014) 252–258 Contents lists available at ScienceDirect Chemical Engineering

Contents lists available at ScienceDirect

Chemical Engineering Journal

journal homepage: www.else vier.com/locate/cej

Journal journal homepage: www.else vier.com/locate/cej Preparation of zinc oxide/silica nanocomposite particles via

Preparation of zinc oxide/silica nanocomposite particles via consecutive sol–gel and flame-assisted spray-drying methods

W. Widiyastuti a , , Iva Maula a , Tantular Nurtono a , Fadlilatul Taufany a , Siti Machmudah a , Sugeng Winardi a , Camellia Panatarani b

a , Sugeng Winardi a , Camellia Panatarani b a Department of Chemical Engineering, Institut Teknologi

a Department of Chemical Engineering, Institut Teknologi Sepuluh Nopember, Kampus ITS Sukolilo, Surabaya 60111, Indonesia b Department of Physics, Universitas Padjadjaran, Jl. Raya Bandung-Sumedang km. 21, Sumedang 45363, Indonesia

highlights

Water glass was used as a silica source to prepare ZnO/SiO 2 nanocomposite. The mixed ZnO–SiO 2 colloids were dried by flame-assisted spray-drying methods. The addition of SiO 2 to ZnO resulted in the increase of the particles photoluminescence intensity. Silica content of the nanocomposites is significantly affected the particles characteristics.

article info

Article history:

Received 7 February 2014 Received in revised form 19 May 2014 Accepted 23 May 2014 Available online 5 June 2014

Keywords:

Colloidal precursor Photoluminescence Flame spray method Particle Dryer Nanocomposites

abstract

ZnO/SiO 2 nanocomposites were prepared using consecutive sol–gel and flame-assisted spray-drying methods. Zinc oxide and silica sols were prepared individually from zinc acetate and water glass, respec- tively, via a sol–gel method. Subsequently, zinc oxide sol, silica sol, and a mixture of both were spray dried in a flame reactor to generate powder of pure zinc oxide, pure silica, and a ZnO/SiO 2 nanocomposite, respectively, in order to compare the characteristics. The generated particles were characterized via lumi- nescence spectrophotometer, Fourier Transform Infrared (FTIR), X-ray diffraction (XRD), and scanning electron microscopy (SEM) to obtain the photoluminescence spectra, chemical bonding, crystallinity, and morphology of the particles, respectively. The photoluminescence emission recorded at 250 nm exci- tation showed a peak at approximately 400 nm that was characterized by a violet band. The highest intensity was shown by a 75 mol% ZnO sample followed, in order, by 50 mol% ZnO, 25 mol% ZnO, and pure ZnO.

2014 Elsevier B.V. All rights reserved.

1. Introduction

New energy-conserving light sources have generated a tremen- dous amount of interest due to increasing energy demands and concerns of global warming. Solid-state light emitting diodes (LEDs) are now receiving increased attention as energy-saving sources, particularly phosphor-converted LEDs due to their high luminescence efficiency. Phosphors are usually limited to mono- chromatic emissions. Therefore, to generate white light, a blend of phosphors that emits in blue, green and red wavelengths is required. Zinc oxide is a wide-band-gap semiconductor that exhib- its a strong and stable broad-band emission. ZnO has been consid- ered as an alternative source to replace expensive GaN-based LEDs [1] , because they have semiconducting properties that compare

Corresponding author. Tel.: +62 31 5946240; fax: +62 31 5999282. E-mail address: widi@chem-eng.its.ac.id (W. Widiyastuti).

1385-8947/ 2014 Elsevier B.V. All rights reserved.

favorably to those of relatively expensive phosphor materials. Other applications of zinc oxide particles include the following:

high-density data storage systems, transparent electrodes, gas sensing, bio-detection, UV light emitters, window material for dis- plays, solar cells, and lasers [2] . Aerosol or wet processes can be utilized to synthesize ZnO nanoparticles. Particles on the nano scale are preferred due to the unique and improved properties that are a result of their size, composition and structure [3] . ZnO nanoparticles were synthesized using an integrated pulse combustion-spray pyrolysis process [4] . The fluctuations in the flow of hot gasses in pulse combustion sys- tems enhance energy transfer and promote an increase in crystal- line nanoparticles compared with the micron-sized particles that are produced by conventional spray pyrolysis. Other methods include gel-template combustion [5] and flame spray pyrolysis [6] . An aerosol-based spray-drying system for producing particles with a controlled morphology has been described in detail [7] .

W. Widiyastuti et al. / Chemical Engineering Journal 254 (2014) 252–258

253

Besides aerosol processing, the synthesis of ZnO nanoparticles through wet-chemical processing is also an area of particular inter- est, because it provides an economical alternative using moderate temperatures and pressures. Particle size is controlled by the rate of nucleation and coagulation that is highly influenced by synthesis conditions such as temperature, precursor concentration, and sol- vent type. A synthesis of ZnO colloidal nanoparticles with sizes that range from 3 to 6 nm has been demonstrated via a sol–gel approach [8] . However, an increase in particle size after aging resulted in a significant color shift in the photoluminescence spec- tra [9] . Nanoparticles tended to coagulate and form larger particles, which led to a decrease in their functional characteristics. In an attempt to maintain the characteristics of nanoparticles, they were mixed with inert material to generate a nanocomposite. Further- more, trapping the particles in a solid matrix hindered the growth of ZnO nanoparticles, and the properties of the particles would not change. Nanocomposite structures are formed by incorporating functional nanoparticles into a transparent matrix material, which stabilizes the particle size and growth and maintains the enhanced nanoparticles [10]. Transparent silica nanoparticles are well suited for use in a matrix because they have almost no effect on the embedded nano- particles. Variations in synthesis methods cause ZnO nanoparticles to exhibit different optical properties when embedded into amor- phous silica nanoparticles by varying the particle size and compo- sition [11] . A matrix of silica particles can play an important role in determining the controllable luminescence position of ZnO nano- particles. Zinc oxide–silica nanocomposites have been prepared by sol–gel and spray-combustion methods using tetraethoxysilane (TEOS) and hexamethyldisiloxane (HMDSO), respectively [12,13]. However, TEOS and HMDSO are relatively expensive and toxic organic precursor materials. Zinc oxide nanoparticles embedded in a silica nanoparticle matrix were investigated by consecutive sol–gel and spray-drying methods by using commercial colloidal silica [14,15]. Water glass is a promising candidate for the synthe- sis of silica colloidal nanoparticles, and it is both relatively inexpensive and plentiful. The preparation of colloidal silica nano- particles from water glass has been accomplished under both acidic and basic conditions [16] . Silicic acid was generated before the formation of colloidal silica nanoparticles. In addition, silicic acid can be synthesized from biomass that is abundant in tropical regions. The objective of this work was to synthesize a ZnO/SiO 2 nano- composite using a flame-assisted spray drying method with mixed colloidal ZnO and silica precursors synthesized by sol–gel methods. Colloidal ZnO and silica were synthesized from zinc acetate and water glass, respectively. The use of water glass as a silica source was a novel attempt in a ZnO/SiO 2 nanocomposite. The silica con- tent in the colloidal precursor was studied for its effect on the char- acteristics of the generated particles.

2. Material and methods

Zinc acetate (Zn(CH 3 COO) 2 2H 2 O, p.a. 99.5%, Merck), lithium hydroxide (LiOH, 98%, Merck), ethanol (absolute, Merck), potas- sium hydroxide (KOH, Merck), and water glass (Na 2 O n SiO 2 m H 2 O, 8% Na 2 O, 27% SiO 2 , Merck) were used as received. Colloidal zinc oxide particles were prepared based on a method established by Spanhel and Anderson using 0.1 M zinc acetate and 0.14 M lith- ium hydroxide with ethanol as a solvent [8] . By using the correla- tion between average particle diameter ( d p ) and the absorption shoulder ( k 1/2 ), based on the UV–Vis absorbance measurement (spectrophotometer, Thermo Scientific Genesys), at the wave- length where the absorption was half of the excitonic peak, the size of the colloidal ZnO was 3.75 nm. The correlation was as fol- lows [17] :

1240

k

1=2

¼ 3: 301 þ 294: 0

d

2

p

þ 1: 09

d

p

ð

1Þ

Colloidal silica particles were prepared by diluting water glass into distilled water to generate a 0.1 M colloidal solution at a tem- perature of 60 C. After the temperature was decreased to room temperature, the water glass solution was passed through a cation exchange resin to form a silicic acid solution. To the silicic acid solution, 0.1 M KOH was added drop-by-drop while the solution was stirred rigorously to adjust the pH to 8. Under these condi- tions, the solution was polymerized to form colloidal silica parti- cles [16] . The particle size distribution of the prepared colloidal silica, as measured by Malvern Particle Size Distribution, is depicted in Fig. 1 . The average geometric particle size ( d g ) and its geometric standard ( r g ) deviation were 11.25 and 1.29, respec- tively. The surface charges of colloidal silica and colloidal ZnO par- ticles were 13.10 and 5.00 mV, respectively, by measuring their zeta potentials (Zetasizer, Malvern). The colloidal zinc oxide and silica were mixed to obtain 25, 50, and 75 mol% ZnO. For comparison, pure colloidal zinc oxide and sil- ica were also used as precursors. The mixed colloidal solution was used as a precursor for the flame-assisted spray-drying system, as shown in Fig. 2 . In the flame reactor, propane gas was used as a fuel and air was used as an oxidizer. The flow rate of the fuel was 0.50 L/min and the flow rate of the oxidizer was 10-fold greater. In order to carry uncharged droplets from the nebulizer (ultrasonic nebulizer, NE-U17, Omron) to the flame reactor, air with a flow rate of 2 L/min was used as the carrier gas. In order to examine the water content of pure ZnO and pure sil- ica samples, thermogravimetry analysis (TGA, Mettler Toledo) was used to measure the change in weight of the generated particles by increasing the temperature at a heating rate of 10 C/min. Scanning electron microscopy (SEM, Zeiss) at 20 kV was used to examine the morphology of the generated particles. An X-ray diffractometer (XRD, Philips) using filter CuK a radiation ( k = 1.54 Å) operated at 40 kV and 30 mA was used to record the diffraction spectra with 2 h from 10 to 60 . The Fourier Transform Infrared (FTIR, Perkin Elmer) spectrum was analyzed for wave numbers ranging from 4000–500 cm 1 . FTIR was used to characterize the type of chemical bonding by evaluating the chemical bond’s vibrating energy that was altered by the absorbance of infrared radiation. A lumines- cence spectrometer (Perkin Elmer) equipped with a xenon source was used to examine the excitation and emission luminescence spectra.

to examine the excitation and emission luminescence spectra. Fig. 1. Particle size distribution of as-prepared colloidal

Fig. 1. Particle size distribution of as-prepared colloidal silica.

254

W. Widiyastuti et al. / Chemical Engineering Journal 254 (2014) 252–258

et al. / Chemical Engineering Journal 254 (2014) 252–258 Fig. 2. Flame-assisted spray-drying system. 3. Results

Fig. 2. Flame-assisted spray-drying system.

3. Results and discussion

SEM images of the samples are shown in Fig. 3 . When using pure ZnO, as shown in Fig. 3 a, smaller particles were observed which did not follow the one-droplet-to-one-particle mechanism that is common to particle formation using a spray-pyrolysis pro- cess [3] . This smaller-particle morphology was induced by the breakup of droplets contained in the precursor solution, because the use of ethanol as a solvent led to an increase in the evaporation rate compared to when water was used as the solvent. Ethanol has a lower boiling point and increases the energy for combustion, because it can also act as a fuel. The same result has also been reported when methanol has been used as a solvent—increasing the methanol content in a water–methanol solvent mixture led

to complete evaporation and resulted in nanoparticle formation [18] . Fig. 3 b shows that the content of ZnO was 75 mol%. The solvent contained 75 volume% of ethanol and the rest was water. The particles remained smaller compared with the particles gener- ated by an ethanol solvent with volume contents of 50% and 25%, as depicted by Fig. 3 c and d, respectively. The 50 and 25 mol% ZnO particles were generated one-droplet-to-one-particle, but the surfaces of the particles were not smooth compared to pure sil- ica using only water as a solvent, as shown in Fig. 3 e. This indicated that the selection of the solvent plays a role in forming distinctions in morphology size, shape, and surface. In order to examine the distribution of zinc oxide and silica nanoparticles in the composite, in the case of the 25 mol% ZnO, ele- mental mapping of Zn, Si, and O by EDX analysis was done, as shown in Fig. 4 , which shows that both the ZnO and silica nanopar- ticles were well-dispersed among the others. ZnO nanoparticles were smaller than the silica nanoparticles, and had a higher degree of kinetic energy. However, the surface charge of the silica nano- particles was larger than that of the ZnO nanoparticles even though both were negatively charged. The same charges led to the repul- sive interactions among colloidal nanoparticles. Size and surface charge contributed to the homogeneity of the generated nanocom- posite particles by maintaining the well-dispersed nature of the nanoparticles inside the droplet. Further explanation of the effects of colloidal surface charge, size and concentration on particle mor- phology has been discussed in detail for silica–polystyrene colloi- dal systems in a previous study [19] . The particle size distribution calculated from SEM images using the Image-J free software shown in Fig. 5 indicates that the mean particle size was increased with the silica content. The morphology of the particles was more spherical with increases in the silica con- tent. However, particles with a relatively broad size distribution were generated because droplets were uncharged or neutral during spraying. The morphology of particles generated from uncharged droplet had broad size distribution and agglomerated particles

[20].

TG-DTA analysis for the generated particles of pure silica and pure ZnO is shown in Fig. 6 . The analysis was performed under air atmospheric conditions within a temperature range of from 25 to 600 C. The water content of the generated particles was

600 C. The water content of the generated particles was Fig. 3. SEM image of the

Fig. 3. SEM image of the generated particles from a precursor of (a) pure colloidal zinc oxide, (b) 75 mol% zinc oxide, (c) 50 mol% zinc oxide, (d) 25 mol% zinc oxi de, and (e) pure colloidal silica.

W. Widiyastuti et al. / Chemical Engineering Journal 254 (2014) 252–258

255

al. / Chemical Engineering Journal 254 (2014) 252–258 255 Fig. 4. Elemental mapping of Si, Zn,

Fig. 4. Elemental mapping of Si, Zn, and O of ZnO/SiO 2 nanocomposite (25 mol% zinc oxide).

and O of ZnO/SiO 2 nanocomposite (25 mol% zinc oxide). Fig. 5. Particle size distribution of

Fig. 5. Particle size distribution of the generated particles from a precursor of (a) pure colloidal zinc oxide, (b) 75 mol% zinc oxide, (c) 50 mol% zinc oxide, (d) 25 mol% zinc oxide, and (e) pure colloidal silica.

examined based on decreases in sample weight at temperatures from 25 to 100 C. The weight of pure silica and pure ZnO samples at 100 C were 99.93% and 99.71% of the original amount, respec- tively. By assuming water evaporates completely at 100 C, the water content of the pure silica and pure ZnO samples were only 0.07% and 0.29%. At 600 C, the weight loss for pure silica and pure ZnO samples was only 2.43% and 5.45%, respectively. In addition, no exothermic or endothermic peaks in the heat flow curve indi- cated that no reaction or phase transformation had occurred. Nanocomposite particles were not analyzed by assuming the water content was in the range of pure silica and pure ZnO samples. Fig. 7 shows the XRD patterns of the generated particles for var- ious zinc oxide–silica compositions. For particles of 25, 50, 75, and 100 mol% ZnO, the ZnO structure could be characterized as hexag- onal wurtzite (JCPDS 36-1451). On the other hand, the diffraction

peak of silica at around 2 h = 22 can be assigned to the silica in an amorphous phase. The XRD pattern shows the characteristic peaks of silica as an amorphous material between 2 h = 20 and 40 . The intensity was decreased by the addition of silica concentra- tion, as shown in Table 1 . The crystalline size that was calculated based on the Scherer number used a half-width of the highest peak in the XRD pattern—the [10 1] peak at 2 h = 36.253 . The crystalline size was decreased by adding colloidal silica in the precursor, which was caused by a decrease in the final zinc precursor concen- tration after adding an amount of colloidal silica, as shown in Table 1 . Fig. 8 shows the FTIR pattern of the samples, which illustrates the type of chemical bonding by pure zinc oxide, pure silica, and the composites of zinc oxide–silica. For pure zinc oxide and 25–75% ZnO samples, infrared absorption reduction peaks were

256

W. Widiyastuti et al. / Chemical Engineering Journal 254 (2014) 252–258

et al. / Chemical Engineering Journal 254 (2014) 252–258 Fig. 6. Thermo-gravimetric and differential analysis

Fig. 6. Thermo-gravimetric and differential analysis (TG-DTA) for the generated particles of (a) pure silica, (b) pure zinc oxide.

observed near 3500 cm 1 , and were attributed to the hydroxyl group of water. This reduction peak was not observed for pure silica. At around 1100 cm 1 , a significant reduction peak was clearly detected for the pure silica sample and insignificant peaks were unmistakable for the nanocomposites samples, which indi- cated asymmetric stretching vibration modes of the Si–O–Si bridge

Table 1 Precursor composition of zinc oxide/silica particles.

%ZnO Volume of sol ZnO

Volume of

Final ZnO

Final silica

d

c

sol silica

concentration

concentration

[nm]

 

[mL]

[mL]

[M]

[M]

100

50

0

0.100

0

24.20

75

50

16.7

0.075

0.025

21.82

50

50

50

0.050

0.050

18.19

25

50

150

0.025

0.075

16.79

0

0

50

0

0.100

of the siloxane link [21] . Reduction peaks at around 1600 cm 1 were only observed for samples that contained zinc oxide. These reduction peaks were attributed to the bending vibration from absorbed water [22] . The combination of the bend vibration of Si–O–Si and the stretching modes of Zn–O are indicated by the reduction peaks at around 400 cm 1 that were only apparent in ZnO/SiO 2 nanocomposite samples: ZnO 25%, ZnO 50%, and ZnO

75%.

When the samples were excited at a wavelength of 250 nm at room temperature, PL emission peaks of 25%, 50%, 75%, and 100% ZnO were located at 401.5, 394.5, 396, and 399.5 nm, respectively, as shown in Fig. 9 . The peak intensities did not significantly change the wavelength position with silica concentration in the compos- ites. The PL emission peaks of nanocomposite samples were not dependent on the particles size of the bulk powder. This result indicated that the samples showing a noticeable near ultra violet – violet region were matched to near band-edge ZnO emissions as a result of the recombination of free excitons [23] . An emission spectra that consists of near-UV, violet, blue, and green reportedly will reveal bands of 3.4, 3.07, 2.8, and 2.4 eV, respectively, and wavelengths of 365, 404, 443, and 517 nm, respectively [24] . As the figure shows, the half band gap covered by wavelengths rang- ing from 345–480 nm corresponds to the near ultra violet – green bands. In this case, the silica concentration only affected the PL intensity peak of the composites, but was incapable of shifting the band. The maximum intensities of PL emissions of 25%, 50%, and 75%, were 2.8-, 3.6-, and 3.9-fold higher, respectively, than pure ZnO. The emission band covering the visible region was broad, which can be attributed to the deep-level defects of ZnO at different levels of energy. Photoluminescence excitation detected at an emission wave- length of 500 nm is shown in Fig. 10 . The highest intensities were observed at 247, 248.5, and 250 nm wavelengths for samples of 25%, 50%, and 75% volumes of ZnO, respectively. On the other hand,

and 75% volumes of ZnO, respectively. On the other hand, Fig. 7. X-ray diffraction pattern of

Fig. 7. X-ray diffraction pattern of particles at various mole percentages (a) at the same scale of intensity and (b) at a higher scale of intensity only for sil ica XRD pattern.

W. Widiyastuti et al. / Chemical Engineering Journal 254 (2014) 252–258

257

al. / Chemical Engineering Journal 254 (2014) 252–258 257 Fig. 8. Fourier Transform Infrared pattern of

Fig. 8. Fourier Transform Infrared pattern of particles at various mole percentages.

Infrared pattern of particles at various mole percentages. Fig. 9. Photoluminescence emission spectra of particles at

Fig. 9. Photoluminescence emission spectra of particles at an excitation of 250 nm.

emission spectra of particles at an excitation of 250 nm. Fig. 10. Photoluminescence excitation spectra of

Fig. 10. Photoluminescence excitation spectra of particles at an emission of 500 nm.

for the same emission wavelength, pure ZnO at a wavelength of approximately 250 nm showed the lowest intensity. That result indicated that composing ZnO with silica resulted in a higher per- formance for photoluminescence by comparison with pure ZnO.

4. Conclusion

In summary, ZnO/SiO 2 nanocomposites were synthesized by using consecutive sol–gel and flame-assisted spray-drying meth- ods. A flame reactor was used to spray the mixed colloidal zinc oxide and silica precursor from zinc acetate and water glass, respectively. Water glass used as a silica source was a novel attempt in a ZnO/SiO 2 nanocomposite. The use of water and etha- nol as solvents in colloidal silica and zinc oxide, respectively, increased the silica content in the precursor, and led to an increase in the particle size. The pure silica sample had an amorphous struc- ture. The ZnO/SiO 2 nanocomposite was characterized as a hexago- nal wurtzite of ZnO, which was the same structure as pure zinc oxide. The crystalline size was increased with an increase in the ZnO content. The highest photoluminescence intensity was observed for the 75 mol% ZnO sample followed by the 50, 25 mol%, and pure ZnO particles, in order, which emitted at a wavelength approximately of 400 nm when they were excited at a wavelength of 250 nm, which corresponded to a violet band.

Acknowledgements

This research was supported by the Indonesia Toray Science Foundation (ITSF) under the scheme of Science and Technology Research Grant 2012 (STRG 2012). We also extend our gratitude to Ms. Paulina Ruliawati for assistance with the experiment.

References

258

W. Widiyastuti et al. / Chemical Engineering Journal 254 (2014) 252–258