Journal of Nanoparticle Research

, 11:1767

One-step flame method for the synthesis of coated composite nanoparticles


  • Sowon Sheen
    • National CRI Center for Nano Particle Control, Institute of Advanced Machinery and Design, School of Mechanical and Aerospace EngineeringSeoul National University
    • Samsung Electronics
  • Sangsun Yang
    • National CRI Center for Nano Particle Control, Institute of Advanced Machinery and Design, School of Mechanical and Aerospace EngineeringSeoul National University
    • Nano Functional Materials Group, Department of Powder MaterialsKorea Institute of Materials Science
  • Kimin Jun
    • National CRI Center for Nano Particle Control, Institute of Advanced Machinery and Design, School of Mechanical and Aerospace EngineeringSeoul National University
    • National CRI Center for Nano Particle Control, Institute of Advanced Machinery and Design, School of Mechanical and Aerospace EngineeringSeoul National University
Research Paper

DOI: 10.1007/s11051-009-9596-z

Cite this article as:
Sheen, S., Yang, S., Jun, K. et al. J Nanopart Res (2009) 11: 1767. doi:10.1007/s11051-009-9596-z


A simple in situ flame coating method has been developed by designing a new type of coflow diffusion flame burner having a sliding unit. The sliding unit was shown to be very effective in finding a right position where the precursor for coating layer should meet with core particles. SiO2-coated TiO2 nanoparticles were first prepared and whether most surfaces of particles were coated was examined by both direct observation of particles through a transmission electron microscope and Zeta potential measurements. Mean core sizes varied from 28 to 109 nm and mean coating thickness was about 2.4 nm for silica-coated titania particles. By simply changing chemical precursors, we demonstrated that SiO2-coated SnO2, SnO2-coated TiO2, SiO2–SnO2-coated TiO2 nanoparticles could be also synthesized.


Flame methodAerosol coatingCoated particlesComposite particles


Synthesis of composite nanoparticles attracts much attention since composite nanoparticles may not only exhibit unique properties that could not be expected from each constituent of the composites, but may also be necessitated for ensuring the stability of particles. In particular, coated composite nanoparticles have shown enhanced properties compared to non-coated particles. For example, more efficient photoluminescence and stability were reported from silica-coated CdSe quantum dots (Selvan et al. 2005) and silica-coated magnetic particles were shown to be more easily dispersed in organic and aqueous solutions (Santra et al. 2001). Layer by layer assembly of silica-coated Fe3O4 nanoparticle was proposed and coated magnetic particles showed reduction of cooperative magnetization switching between particles due to the barrier effect of silica layer (Aliev et al. 1999).

Liquid phase methods were mostly used to produce oxide-coated oxide nanoparticles. Water-in-oil microemulsion method was applied to produce silica-coated ceria nanoparticles for cosmetic applications (Tago et al. 2003) and silica-coated iron oxide nanoparticles for magnetic application (Santra et al. 2001). Precipitation was also used to produce silica-coated titania nanoparticles (Lin et al. 2002). Gas phase methods were also developed. Microwave plasma process was reported to prepare various coated composite oxide nanoparticles (Szabó and Vollath 1999). A continuous flow hot wall reactor was utilized to produce silica or alumina-coated titania particles (Fotou and Kodas 1997; Powell et al. 1997). Lee et al. (2002) also used a tubular furnace for the synthesis of silica-coated titania nanoparticles. Counterflow diffusion flame was used to form TiO2/SiO2 composite nanoparticles and depending on the flame condition, SiO2-coated TiO2 particles were obtained (Hung and Katz 1992). A premixed flat flame burner was also utilized to generate SiO2/TiO2 particles including coated particles (Ehrman et al. 1998). Recently, a coflow diffusion flame reactor was employed to prepare silica-coated titania nanoparticles (Teleki et al. 2005). Flame aerosol method has the capability to continuously produce high purity nanoparticles at mass quantity scale and its target is now moving toward the large scale production of size and crystalline phase controlled nanoparticles (Jiang et al. 2007), as well as more sophisticated composite nanoparticles (Stark and Pratsinis 2002; Akurati et al. 2006). Previous attempts for the synthesis of coated oxides in flames used the difference of reaction rates between mixed precursors. Therefore, this kind of approach has limitation for choosing precursors.

Here, we report a robust one-step flame method for the synthesis of versatile kinds of coated oxide nano-composite particles. We designed a new type of coflow diffusion flame burner having a sliding unit and demonstrated that this design could produce different kinds of oxide-coated oxide nanoparticles by simply changing chemical precursors. The sliding unit was shown to be very effective in finding a right position where the precursors for coating layer should meet with the core particles growing from the lower position near the burner surface. SiO2-coated TiO2 nanoparticles were first prepared and the particle sizes and the coating layer thickness were examined. The direct observation of a few hundred particles via a transmission electron microscope together with the comparison of Zeta potential measurements confirmed that most surfaces of particles were coated. Fourier transform infrared spectroscopy and X-ray Photoelectron Spectroscopy were also done to study the bonding between the coating layer and core particle. We also demonstrated that SiO2-coated SnO2, SnO2-coated TiO2, and SiO2–SnO2-coated TiO2 nanoparticles could be synthesized by simply changing chemical precursors.


Our flame burner having a sliding unit for synthesizing coated nanoparticles is illustrated in Fig. 1. A coating unit is mounted on the conventional diffusion flame burner composed of concentric stainless-steel tubes (Lee and Choi 2000; Cho and Choi 2000; Lee and Choi 2002; Choi et al. 2004). The precursor for a host particle is injected through the center nozzle of the burner. Shield gas (N2, 0.7 L/min) is injected through eight holes adjacent to the center nozzle to prevent deposition of formed particles on the nozzle. Fuel (H2, 1.8 L/min) and oxygen (O2, 4.0 L/min) are injected through the next two concentric annuli in order, forming the flame. The coating unit consisting of two concentric stainless-steel tubes and a honeycomb is designed to change its relative vertical position to the burner surface by sliding along the outer tube of the burner. Vapor-phase precursor for a coating layer is transported through an annular nozzle in the coating unit by a carrier gas (N2). Transported precursor is injected at the exit of the coating unit and then diffused toward the center of the flame where host particles are being produced. It is noted that the precursor for host particles was injected at a lower position than that for the coating layer by hr, which can be adjusted to find a right position ensuring particle coating by sliding the coating unit. Shield gas (N2, 70 L/min) through honeycomb in the coating unit prevents the injected vapor from mixing with surrounding air.
Fig. 1

Schematic illustration of the modified H2/O2 diffusion flame burner having a sliding type coating unit: a side view, hr and hp denote the distance between the burner surface and coating unit surface and the distance of particle sampling from burner surface, respectively, and b top view of the burner

SiO2-coated TiO2 particles, SiO2-coated SnO2 particles, and SnO2-coated TiO2 particles were synthesized from SiCl4, TiCl4, and SnCl4 precursors. These three liquid precursors were contained in bubblers. A bubbler containing SiCl4 liquid was maintained at 20 °C, whereas bubblers for TiCl4 and SnCl4 were kept at 60 °C. Carrier gas (N2) flow rates for the three precursors were 200 cc/min each. Carrier gases were assumed to be saturated with the precursor for given temperature and carrier gas flow rates (Choi et al. 2002, Akhtar et al. 1991). In order to control the titania core particle size, the feeding rate of TiCl4 was changed from 6.2 × 10−4 to 4.7 × 10−5 mol/min and the feeding rate of SiCl4 was fixed as 2.7 × 10−3 mol/min. In the case of SnO2- and SiO2-coated TiO2, vaporized SiCl4 and SnCl4 gases carried by N2 (150 cc/min, each) were mixed before being injected through the coating unit. Synthesized nanoparticles were characterized by a transmission electron microscope (TEM), Philips CM-20 operated at 200 kV, high resolution transmission electron microscope (HR-TEM), JEOL JEM-3000F operated at 300 kV, selected area electron diffraction (SAD) attached to CM-20, energy dispersive X-ray analysis (EDS) attached to CM-20, and JEM-3000F. Particles in the flame were sampled onto formvar-coated nickel grid by a local thermophoretic sampling device. The local thermophoretic sampling device consists of a sampling probe holding a TEM grid, a shield covering the TEM grid, two air cylinders, and three timers to control the insertion times of the probe and the shield independently (Choi et al. 1999; Cho and Choi 2000). The shield was designed to retract itself to expose the TEM grid only for the duration of sampling at the local position within a flame. Particles were also thermophoretically sampled onto a water-cooled quartz tube (6 mm in diameter) for zetapotential analysis to confirm the uniformity of particle coating. TEM samples were also made from the collected powder and particle morphology was double checked. Zeta-potential of particles was measured using a ZetaPlus potential analyzer (Brookhaven Instruments). The collected powder was dispersed in aqueous 0.01 M NaCl solutions and the colloidal solutions were treated in an ultrasonic bath for more than 30 min to break up powders into non-agglomerated particles. The pH values of the solutions were adjusted by adding NaOH (1.0 N solution in water, Aldrich) or HCl (1.0 N solution in water, Aldrich). The crystalline phase of coated particles was determined by X-ray powder diffraction (MXP18XHF-22, MAC/Science) using CuKα radiation (λcu = 1.54056 Å). Voltage and current outputs for measurements are 40.0 kV and 200.0 mA, respectively. Samples were measured at a step width of 0.02° and scanning speed was 0.5° or 2.0°/min. The chemical bonding of coated nanoparticles was examined by FTIR spectroscopy using a spectrometer (DA 8, Bomen). The powders were diluted with KBr (99+%, Aldrich) and the spectra were collected with a spectral resolution of 2 cm−1. X-ray Photoelectron Spectroscopy (XPS) studies were performed using an Electron Spectroscopy for Chemical Analysis (ESCA) instrument (ARIESARSC 10MCD 150, Vacuum Science Workshop). MgKα (1253.6 eV) X-ray radiation was used. Powders with or without the coating layer were pressed into compacted pellets (10 mm in diameter) for measurements. Charging effects were corrected by adjusting the main C 1 s peak to 284.5 eV binding energy. The vacuum in the main analytical chamber during measurement was about 5 × 10−9 Torr. The spectra of O 1 s, Si 2p, Ti 2p3/2, and C 1 s were recorded. The obtained spectra were deconvoluted into Gaussian curves. The width, intensity, and position of peaks were determined by applying least mean square algorithm using PeakFitTM. (PeakFitTM 4.0, 1997, SPSS Inc., USA).

Results and discussion

SiO2-coated TiO2 nanoparticles were first synthesized by injecting vaporized SiCl4 as a coating precursor and TiCl4 as a host precursor to show the validity of our method. SiCl4 was injected into the coating unit and TiCl4 was injected into the central nozzle as a host precursor. In order to find a proper position of the coating unit relative to the surface of a burner, particles were collected at hp = 160 mm using a thermophoretic sampling device by changing the sliding coating unit at different vertical positions (hr) and were examined through TEM. When the coating unit was placed at low positions (hr = 5 and 45 mm) (Fig. 2a, b), large aggregates and spherical particles were observed together and EDS analysis revealed that those were SiO2 aggregates and TiO2 spheres. At the lower positions of hr = 5 mm and 45 mm, diffused SiCl4 vapor is likely to form small aggregates by chemical reaction due to high flame temperatures at these low flame heights. By sliding the coating unit vertically to a further higher position at hr = 80 mm, it was possible to prevent the formation of independent SiO2 particles since flame temperatures decrease in the downstream region. As shown in Fig. 2c, silica-coated titania nanoparticles were prepared. A magnified TEM image confirms that the surface of spherical TiO2 nanoparticle is completely coated with a thin layer about 2.4 nm SiO2 in thickness (Fig. 2d). Suppression of silica aggregate formation for the case of hr = 80 mm was double checked in TEM from the powder collected onto the water-cooled quartz tube. An additional experiment with injecting only SiCl4 into the coating unit at hr = 80 mm without the core particle precursor of TiCl4 was also done to confirm the suppression of silica aggregates. TEM results showed the absence of particles, indicating that SiCl4 vapor precursor was just ventilated without particle formation for the case of hr = 80 mm. The suppression of aggregate formation indicates that the coating mechanism in the present study could be due to cluster scavenging and chemical vapor deposition on the surface of host particles rather than the formation of silica aggregates and subsequent collision of the aggregates with host particles (Fotou and Kodas 1997; Powell et al. 1997).
Fig. 2

TEM images of collected nanoparticles with different sliding unit positions: ahr = 5 mm, hp = 160 mm. bhr = 45 mm, hp = 160 mm. chr = 80 mm, hp = 160 mm. d Magnified TEM image of a SiO2-coated TiO2 nanoparticle, hr = 80 mm, hp = 160 mm

It is important to examine the uniformity of particle coating to ensure that almost all particles are coated. In addition to the confirmation of silica coating for hundreds of particles via TEM image examination, the zeta-potential analyses of pure TiO2, pure SiO2, and the present SiO2-coated TiO2 particles were also carried out to confirm whether most particles were coated (see Fig. 3). Pure TiO2 particles were prepared at the same condition as in the generation of coated particles without SiCl4 gas through the coating unit. Pure SiO2 particles were also prepared using the same flame burner. As shown in Fig. 3, the SiO2-coated TiO2 particles show almost the same electrokinetic behavior as the pure SiO2. The isoelectric point (IEP) was shifted to a smaller value and the estimated IEP point of coated TiO2 well matched with that of pure SiO2. Together with TEM examination, this result confirms that most surfaces of TiO2 particles are well coated with SiO2. Using the zeta-potential measurements, the process of coating could be also monitored. Table 1 shows the measured zeta-potentials of powders collected at different flame heights from the burner surface. Collected powders were dispersed in aqueous 0.01 M NaCl solutions of a fixed pH value (pH 6.6 ± 0.2). The measured zeta-potential value decreased with the increase of the collection height and approached the value of pure SiO2 near hp = 160 mm, where the surface of TiO2 nanoparticles was considered to be almost completely coated by SiO2.
Fig. 3

Zeta-potential values of SiO2, TiO2, and SiO2-coated TiO2 nanoparticles; hr = 80 mm, hp = 160 mm

Table 1

Zeta-potential values of TiO2, SiO2, and SiO2-coated TiO2 nanoparticles collected at various heights from the burner surface


Zeta potential (mV)

Pure TiO2

−33.3 ± 0.5

Pure SiO2

−71.6 ± 2.5

86 mm

−45.3 ± 1.2

100 mm

−54.7 ± 0.8

120 mm

−76.3 ± 1.3

140 mm

−72.2 ± 1.2

160 mm

−72.7 ± 1.3

Distributions of coating thickness and core sizes were measured for three different TiCl4 feeding rates (see Fig. 4). Size distributions were determined by measuring more than 300 particles from the TEM images per case. By decreasing the feeding rate of TiCl4 from 6.2 × 10−4 (case C) to 9.3 × 10−5 mol/min (case B), the mean core size of TiO2 particles decreased from 109 nm to 66 nm. In the case of 4.7 × 10−5 mol/min (case A) of TiCl4 feeding rate, the geometric mean core size of TiO2 particles was about 28 nm. In the inset of Fig. 4, the distribution of SiO2 coating thickness is shown for the case of TiCl4 feeding rate of 6.2 × 10−4 mol/min. The geometric mean coating thickness is 2.44 nm and the geometric standard deviation is 1.44. Typical TEM images of silica-coated titania particles with different core sizes are shown in Fig. 5a and b for the cases of A and B (Fig. 2c, d for the case C). A high resolution TEM image in Fig. 5c for case C shows lattice fringe of the core crystalline TiO2 and amorphous SiO2 thin coating layer. HR-TEM analyses confirmed that crystalline TiO2 particle was well coated with amorphous SiO2. XRD analysis showed that the crystalline phase of TiO2 core was anatase (Sheen 2003).
Fig. 4

Size distributions for core particles and coating thicknesses for three different feeding rates. (A: TiCl4 feeding rate of 4.7 × 10−5 mol/min, B: 9.3 × 10−5 mol/min, and C: 6.2 × 10−4 mol/min.) The inset plot shows shell thickness distribution for Condition C
Fig. 5

Typical TEM images of SiO2-coated TiO2 nanoparticles: a case A. b case B. c HR-TEM image of TiO2 core lattice fringe and SiO2 amorphous coating layer (case C)

FTIR study was done to confirm the chemical bonding of SiO2 coating (see Fig. 6). The large absorption peak around 700 cm−1 is caused by the Ti–O–Ti bonding of anatase-phase TiO2 in both Fig. 6a and b. In Fig. 6b, the case of SiO2-coated TiO2 particle shows two absorption bands in the range 1,000–1,300 cm−1, which is caused by the Si–O–Si asymmetric vibration of amorphous SiO2 under the effect of Ti for coated particles (Lin et al. 2002). XPS study was also done which revealed the existence of a binding energy peak of SiO2 and the shift of the binding energy of Ti 2p3/2 to a higher value for silica-coated TiO2 particles (see Fig. 7), indicating Ti–O–Si bond formation (Lin et al. 2002).
Fig. 6

FTIR spectra of TiO2 (a) and SiO2-coated TiO2 nanoparticles (b). Nanoparticles were collected under the condition of hr = 80 mm and hp = 160 mm
Fig. 7

XPS Ti 2p spectra of TiO2 (a) and SiO2-coated TiO2 (b) nanoparticles. Nanoparticles were collected under the condition of hr = 80 mm and hp = 160 mm. (…) original spectrum, (—) sum of deconvoluted Gaussian peaks

The present method can synthesize various kinds of coated nanoparticles by changing chemical precursors. By changing the host precursor from TiCl4 to SnCl4, SiO2-coated SnO2 nanoparticles were produced (Fig. 8a). High crystallinity of core SnO2 could be inferred from the selected area electron diffraction (SAD) pattern (Fig. 8a, inset). SnO2-coated TiO2 particles were also synthesized by replacing the coating precursor SiCl4 with SnCl4 (see Fig. 8b). Multi-component coating onto the surface of TiO2 nanoparticle was also possible by simply feeding mixed gases (SiCl4 and SnCl4) through the coating unit. Amorphous coating layer composed of SnO2–SiO2 is presented in Fig. 8c and EDS (Fig. 8d) analysis shows that both Si and Sn species are present in the coating layer. This method is worth extending to oxide-noble metal nanoparticle system (e.g. Pt/TiO2 or Pt/SiO2) or polymer-oxide nanoparticles with supplying corresponding precursors. Since it is well known that flame methods in general can be easily scaled up compared to other routes (Pratsinis 1998), various kinds of coated composite nanoparticles are expected to be produced at a mass production scale by using our one-step flame method.
Fig. 8

Various types of coated composite nanoparticles: a SiO2-coated SnO2 nanoparticle. b SnO2-coated TiO2 nanoparticle. c Magnified image of SnO2 and SiO2-coated TiO2 nanoparticle. d EDS analysis data for c)


In summary, we report a novel in situ flame coating method by modifying the widely used H2/O2 diffusion flame burner with the addition a sliding type coating unit that allows us to find a right position for producing coated composite nanoparticles. This coating method could produce various kinds of coated nanoparticles in one-step, particularly, oxide-coated oxide nanoparticles, at atmospheric condition. SiO2-coated TiO2 nanoparticles were first synthesized to show the validity of the present method. TEM examination and Zeta-potential measurements confirmed that most surfaces of the particles were well coated. Size distributions of particles and coating thicknesses were also measured. By simply changing the chemical precursors, we also demonstrated that other coated particles such as SiO2-coated SnO2, SnO2-coated TiO2, and SiO2- and SnO2-coated TiO2 could be synthesized.


This work was supported by the acceleration research program and BK21 program of the Ministry of Education, Science and Technology, Korea.

Copyright information

© Springer Science+Business Media B.V. 2009