Introduction

Titanium dioxide (TiO2) is an attractive material commercially being used for environmental purification because of its environmentally friendly photocatalytic activity (Ref 1). However, the main drawback of typical TiO2 is that photocatalytic activity is activated only under UV light irradiation. Many studies have focused on the preparation of higher functional TiO2 doping with carbon, results of which clearly show higher photocatalytic activity not only in UV, but also in visible light irradiation (Ref 2-9). Leary and Westwood (Ref 6) reviewed recent progress in the development of carbon-doped TiO2 (C-TiO2) and various methods of synthesizing C-TiO2 with solution precursors of TiO2, such as sol-gel, sputtering, chemical vapor deposition, and hydrothermal methods. However, few studies on synthesis of C-TiO2 or TiO2 using both the solution precursors of TiO2 and thermal plasma flow have been published (Ref 10-15).

Thermal plasma flow has been widely used for powder processing techniques such as direct current (DC) plasma and radio frequency inductively coupled plasma (RF-ICP) jets because they are characterized by short processing time, high thermal density, and high enthalpy (Ref 10-17). However, DC plasma jets have disadvantages such as smaller high-temperature volume and less particle residence time. Furthermore, RF plasma jets have other disadvantages such as relatively lower thermal density and enthalpy, and flow instability induced by the unsteady Lorentz force in the RF coils. To overcome these problems, a DC-RF hybrid plasma flow system combined with a DC plasma jet and RF-ICP plasma flow has received much attention (Ref 18-23). It has distinct advantages such as a larger high-temperature volume, higher energy flux, small input power and characteristic backflows caused by the Lorentz force of RF plasma flow and the high momentum of DC plasma jet. For preparation of powders using a solution or suspension precursor, liquid injection directly into plasma flow induces the reduction of plasma volume or the dissipation of plasma flow by thermal expansion of liquids (Ref 24, 25). Previously, we found that when water droplets were injected into the downstream plasma flow in a DC-RF hybrid plasma flow system, the droplets were successfully transported to the upstream plasma flame by the linked backflow created near the reactor walls in a DC-RF hybrid plasma flow system. This resulted in an increase in plasma enthalpy without such disappearance of plasma flow as those resulting from the mixing of dissociated hydrogen with argon plasma flow (Ref 23). Thus, it is feasible to prepare highly functional TiO2 nanopowder by droplet injection of solution precursor into the downstream plasma flow utilizing characteristic backflow transportation in the DC-RF hybrid plasma flow system under low electric power.

In the present study, C-TiO2 nanopowder was synthesized by microsized droplet injection of solution precursor utilizing the characteristic backflow transportation in a DC-RF hybrid plasma flow system, and the effect of C-TiO2 nanopowders on photodegradation of methylene blue solution, even under visible light irradiation, was investigated. Synthesized C-TiO2 nanopowder was characterized by optical emission spectroscopy, scanning electron microscopy (SEM), transmission electron microscopy (TEM), x-ray diffraction (XRD) patterns, spectra of x-ray photoelectron spectroscopy (XPS) and UV-Visible absorption.

Experimental Setup and Measuring Devices

DC-RF Hybrid Plasma Flow System

Figure 1 shows a schematic of a DC-RF hybrid plasma flow system with a two-phase atomizer. The experimental conditions are tabulated in Table 1. The DC-RF hybrid torch is cooled by a forced convective water flow. A gas-liquid atomizer (YS-03, Yaezaki, Japan) for liquid droplet injection was installed at z = 195 mm and r = 95 mm. The tip of the atomizer has an inner nozzle with a diameter of 0.3 mm for the supply of solution precursor and an outer nozzle with a diameter of 1.2 mm for the supply of atomizing gas. Plasma processes in a DC-RF hybrid plasma flow system can be divided into two parts: powder processing and plasma treatment. The purpose of powder processing is to synthesize C-TiO2 powders with solution precursor injection, while that of the plasma treatment is to investigate the effect of plasma irradiation on the properties of C-TiO2 deposited on the surface of the quartz tube at z = 200 mm without injections of atomizing gas and solution precursor.

Fig. 1
figure 1

Schematic of a DC-RF hybrid plasma flow system with a two-phase atomizer

Full size image
Table 1 Experimental conditions
Full size table

Preparation of Solution Precursor

Solution precursor, TTB (TiO4C16H36; 99.5%, Wako Pure Chemicals Industries, Ltd.) diluted in ethanol (C2H5OH; 99.5%, Wako Pure Chemicals Industries, Ltd.), was prepared to synthesize C-TiO2. For complete dispersion, the solution precursor was stirred for 10 min by a magnetic stirrer before injection. The injection condition of the solution precursor is tabulated in Table 1. Droplets of the solution precursor were injected into the downstream plasma flow at z = 195 mm by the atomizer.

An atomization model of viscous liquids demonstrated by Aliseda et al. (Ref 26) was used for estimation of the Sauter mean diameter (SMD) of droplets of a mixture liquid, which is defined by

$$ {\text{SMD}} = C_{1} \left( {1 + m_{\text{r}} } \right)\sqrt {\left( {\frac{{b_{\text{g}} }}{{d_{\text{in}} }}} \right)} \root{4} \of {{\left( {\frac{{\uprho_{\text{l}} }}{{\uprho_{\text{g}} Re_{d} }}} \right)}}\frac{1}{{\sqrt {We} }}\left\{ {1 + C_{2} \root{6} \of {{\left( {\frac{{d_{\text{in}} }}{{b_{\text{g}} }}} \right)}}\root{12} \of {{\frac{{\uprho_{\text{g}} Re_{d} }}{{\uprho_{\text{l}} }}}}\root{6} \of {We} \root{3} \of {Oh}^{2} } \right\}d_{\text{in}} , $$
(1)

where the coefficients C 1 and C 2 are dependent on nozzle geometry; m r is the mass flux ratio of precursor solution to atomizing gas; d in is the diameter of the inner nozzle of the atomizer; b g is difference between the outer and inner nozzle radii at the tip of the atomizer; ρl and ρg are the densities of the precursor solution and atomizing gas, respectively; and We, Oh, and Re d are the Weber, Ohnesorge, and Reynolds numbers, respectively. Using the estimated SMD in Table 1, the characteristic time of droplet vaporization is estimated to be approximately 17 ms from Ref 23 at a temperature of around 1000 K.

Analysis of C-TiO2 Characteristics

C-TiO2 nanopowders, CT-Up and CT-Down, were collected on the surfaces of upstream and downstream quartz tubes at z = 0-100 mm, and 200-510 mm in Fig. 1, respectively. The inner diameters of these two cylindrical quartz tubes were 45 and 68.5 mm and their lengths were 100 and 310 mm, respectively. The upstream quartz tube was installed at the DC-RF hybrid torch, and the downstream quartz tube was at z = 200 mm in the chamber. The mass fractions of C-TiO2 nanopowders were 1.59 and 87.25% on the surfaces of the two quartz tubes in the upstream and downstream regions, respectively, and 11.16% on the chamber wall on average to the total production rate of 0.502 g/h. To investigate the effect of plasma treatment on powder properties, C-TiO2 particles deposited on the surface of the quartz tube were consecutively irradiated by pure argon plasma flow for 5 min after powder processing. Single-phase anatase TiO2 powder (Wako Pure Chemicals Industries, Ltd.) was prepared for a clear comparison of its physicochemical characteristics with those of C-TiO2 nanopowders.

Gas temperature was measured by using a W-Re thermocouple surrounded by a ceramic tube with a diameter of 3 mm. The measurement of gas temperature was limited to the downstream region because the uncertainty of the thermocouple was increased because of the thermal damage by the upstream plasma flame. Optical emission characterization was carried out by optical emission spectroscopy (Maya 2000PRO, Ocean Optics, USA) with wavelengths in the range of 160-390 nm to detect carbon and titanium species in plasma flow. Morphologic analysis was carried out by scanning electron microscopy (SU6600, Hitachi Co., Japan) and transmission electron microscopy (HF-2000, Hitachi Co., Japan). Particle size distribution was measured by a particle size analyzer (UPA EX250, Microtrac Inc.) after ultrasonic treatment of powder for 10 min. The crystal structure was analyzed by the XRD pattern (RINT-V, Rigaku Co., Japan). The carbon composition and properties of C-TiO2 were studied using XPS (KRATOS, Shimadzu Co., Japan). UV–Visible absorption characteristics were studied by an UV–Visible spectrometer (V-7200, JASCO Co., Japan). Photocatalytic performance of synthesized C-TiO2 nanopowders was assessed by the degradation rate of methylene blue (MB) solution under visible light irradiation. MB solution of 0.2 mg/L concentration was prepared with the addition of 1 mg C-TiO2 nanopowder. The MB solution was then irradiated by visible light of 0.58 W with a wavelength range of 420-700 nm for an hour. The degradation rate of the MB solution was measured by a turbidity photometer (AQUALYTIC®, Germany).

Experimental Results and Discussion

DC-RF Hybrid Plasma Flow Characteristics

Figure 2(a) and (b) show photographs of DC-RF plasma flow without and with droplet injection of solution precursor at z = 195 mm. DC-RF hybrid plasma flow, as a reaction zone, is considerably elongated by the DC plasma jet with high momentum, as shown in Fig. 2(a), which is divided into two plasma flows. Upstream and downstream plasma flows denote plasma flame with higher temperature at z = 0-100 mm at the DC-RF hybrid torch and free-stream plasma flow with lower temperature in the chamber, respectively. When injecting droplets of solution precursor into downstream plasma flow at z = 195 mm, the emission colors of plasma flow are widely observed to be cyan and orange from the upstream plasma flame below the RF coils to the downstream plasma flow, as shown in Fig. 2(b). It can be explained that the characteristic backflow transfers droplets of solution precursor along the chamber wall to the upstream plasma flame. That is, once injected droplets of solution precursor are transferred to the upstream plasma flame by the characteristic backflow, in-flight droplets actively react with the plasma flow before settling down.

Fig. 2
figure 2

Photographs of DC-RF plasma flow (a) without and (b) with droplet injection of solution precursor at z = 195 mm

Full size image

Figure 3 shows the optical emission spectra at z = 70 mm. Excited carbon peaks with higher emission intensity are clearly observed at wavelengths of 193 nm and 247 nm, and excited titanium peaks are observed at wavelengths in the range of 373-376 nm in the inset of Fig. 3. This indicates that droplets of solution precursor are successfully transported to the upstream plasma flame by such a characteristic backflow. TiC formation can occur near the upstream region because titanium and carbon combine to form TiC at more than 1000 °C (Ref 27) and TiO2 converts to TiC at approximately 1400 °C (Ref 28).

Fig. 3
figure 3

Optical emission spectra at z = 70 mm with the inset showing expanded spectra in the range of 372-378 nm

Full size image

Figure 4(a) and (b) show axial and radial distributions of gas temperature. TiC formation occurs in plasma flow near the upstream region where the gas temperature is expected to be around 1850 °C from Fig. 4(a). In Fig. 4(b), without atomizing gas injection, the downstream plasma flow is considerably elongated by the DC plasma jet, as shown in Fig. 2(a), which rapidly heats argon gas up to more than 1000 °C at −0.25 < r/R < 0.25. The temperature gradient is rather steep at surrounding downstream plasma flow because of low thermal conductivity of argon. When cold atomizing gas is injected without solution precursor injection, the distribution of gas temperature becomes asymmetric. The downstream plasma flow is entrained into the cold atomizing gas flow. Active heat transfer then occurs from the downstream hot plasma flow to the cold atomizing gas flow. As a result, the gas temperature gradually increases at 0.25 < r/R < 1 and rapidly increases at r/R ≤ 0.25 because of such active heat transfer. When injecting droplets of the solution precursor, the gas temperature slightly increases at r/R = 0. Although the thermal energy of the downstream plasma flow is consumed by evaporation of the droplets, heat is released by thermal decomposition of solution precursor. As found in Fig. 4(b), because the gas temperature is more than 600 °C even with atomizing gas injection, anatase TiO2 deposited on the surface of the downstream quartz tube can convert into rutile TiO2 during plasma treatment.

Fig. 4
figure 4

(a) Axial and (b) radial distributions of gas temperature

Full size image

Synthesized C-TiO2 Characteristics

Figure 5(a) and (b) shows images of SEM and TEM for the powder morphology of C-TiO2 in CT-Down. Agglomerated C-TiO2 particles in Fig. 5(a) are extensively observed in all the samples. In-flight C-TiO2 nanoparticles are soon melted by plasma flow and agglomerated because carbon with thermal conductivity higher than those of argon and TiO2 increases the heat exchange rate between plasma flow and C-TiO2 particles. Nanocrystalline anatase, rutile, TiC, and graphite are formed by short residence time in plasma flow and observed in Fig. 5(b).

Fig. 5
figure 5

Images of (a) SEM with the inset showing the agglomerated C-TiO2 and (b) TEM for the powder morphology of C-TiO2 nanopowder in CT-Down

Full size image

Figure 6 shows the particle size distributions of C-TiO2 nanopowders. The size distribution of CT-Down shows good agreement with the observation of the SEM image in Fig. 5(a). The median particle size of CT-Down is larger than that of CT-Up, although the particles are collected in the lower temperature region. It can be explained that nanoparticles of CT-Down agglomerate to form larger clusters in plasma flow because of longer residence time.

Fig. 6
figure 6

Particle size distributions of C-TiO2 nanopowders where d 50 denotes the median particle size

Full size image

Figure 7 shows XRD patterns of synthesized C-TiO2 nanopowders with plasma treatment for 5 min and without such treatment. TiO2 peaks of anatase (JCPDS 01-070-7348) and rutile (JCPDS 01-084-1283), and TiC peaks (JCPDS 065-8417) are observed in all the samples. Since anatase-rutile mixed-phase TiO2 is well known for its photocatalytic performance which is higher than that of single-phase anatase TiO2 (Ref 3, 6), C-TiO2 nanopowder with anatase-rutile mixed-phase TiO2 is expected to exhibit better photocatalytic performance. However, TiC is not a photocatalyst. As discussed, TiC is produced by an active reaction with in-flight particles and high-temperature plasma flow. The anatase-rutile contents and crystalline sizes of synthesized C-TiO2 nanopowders are summarized in Table 2. Calcination was carried out for the estimation of TiC contents. Because the oxidation of TiC occurs at 350 °C in air atmosphere, TiC in C-TiO2 nanopowder completely converts into anatase at 400 °C for 2 h in air atmosphere (Ref 4, 9). Assuming that the amount of carbon impurity is negligible and that rutile phase transformation does not occur, TiC contents are roughly estimated from the increment of anatase content and the constant rutile content before and after the calcination. The estimated TiC contents are 24.8, 29.2, and 35.1% in CT-Up, CT-Down, and CT-P-Down, respectively. CT-Up collected in the upstream region has rather larger anatase content and smaller TiC content than that of CT-Down collected in the downstream region. This is due to two factors: first, the effect of upstream plasma flame on particles deposited on the surface of the quartz tube in the upstream region is less, because the upstream plasma flame is entrained to the center below the RF coils by a DC plasma jet. Second, particle heating is suppressed by the injection of cold swirling sheath gas and by the water-cooled quartz tube in the DC-RF hybrid torch. The nanocrystalline sizes of anatase and rutile in CT-Down are approximately consistent with the observation of the TEM image in Fig. 5(b). After plasma treatment for 5 min, anatase content decreases with the increase in TiC content because the very high surface energy of nanocrystalline C-TiO2 affords easier phase transformation even in a downstream plasma flow of pure argon less than 700 °C (Ref 29).

Fig. 7
figure 7

XRD patterns of synthesized C-TiO2 nanopowders with plasma treatment for 5 min (CT-P-Down) and without such treatment (CT-Up and CT-Down)

Full size image
Table 2 Properties of synthesized C-TiO2 nanopowders
Full size table

Figure 8(a) and (b) shows XPS spectra in Ti 2p and O 1 s regions for C-TiO2 nanopowder. In the Ti 2p XPS region, two peaks of Ti 2p3/2 and Ti 2p1/2 ascribed to the formation of TiO2 are observed at 458.5 and 464.2 eV, respectively (Ref 12, 13, 30, 31). Particularly, a peak at 457.1 eV shifted by the carbon substitution into oxygen vacancy is found only in CT-Down because the relatively small size of the carbon atom can be introduced without too much surface strain and lattice distortion (Ref 2, 6, 32). This C-TiO2 with substitutional carbon is expected to increase UV–Visible light absorption and photocatalytic performance by narrowing the band gap (Ref 12, 13, 30). In addition, the peak at 454.1 eV assigned to Ti-C bond is found only in CT-Up (Ref 33, 34). In the O 1 s XPS region, three peaks at 529.8, 531.8, and 533.1 eV are ascribed to Ti-O, O-H, and H2O bonds, respectively (Ref 6, 10, 30). A peak at 528.3 eV ascribed to carbon bonding with oxygen is found only in CT-Down (Ref 30). Consequently, the XPS results suggest the presence of a Ti-O-C carbonaceous bond only in CT-Down.

Fig. 8
figure 8

XPS spectra in (a) Ti 2p and (b) O 1 s regions for C-TiO2 nanopowder

Full size image

C-TiO2 Photocatalytic Characteristics

Figure 9 shows the effect of C-TiO2 nanopowder on UV–Visible absorption characteristics. C-TiO2 nanopowder in CT-Down is observed to absorb more UV–Visible light than that in CT-Up, which has a cut-off wavelength at around 560 nm. It can be explained that C-TiO2 with a Ti-O-C bond reduces band gap energy (Ref 2-4, 9). On the other hand, C-TiO2 nanopowder in CT-Up has a lower cut-off wavelength of approximately 375 nm. Based on these results, the synergistic effect of C-TiO2 with Ti-O-C bond and graphite improves UV–Visible light absorption.

Fig. 9
figure 9

Effect of C-TiO2 nanopowder on UV–Visible absorption characteristics

Full size image

Figure 10 shows the effect of C-TiO2 nanopowder on the degradation rate of MB solution for an hour under visible light irradiation. Synthesized C-TiO2 nanopowder exhibits a better degradation rate of MB solution than that of single-phase anatase TiO2, in spite of being TiC. In particular, C-TiO2 nanopowder in CT-Down perfectly completes degradation of MB solution in 50 min because of an increase in visible light absorption and reduction of band gap energy induced by carbon substitution into TiO2. On the other hand, C-TiO2 nanopowder in CT-Up exhibits a degradation rate of 60% of MB solution. Single-phase anatase TiO2 demonstrates an insufficient degradation rate of 37% of the solution. Therefore, highly functional C-TiO2 nanopowder synthesized by droplet injection of solution precursor in a DC-RF hybrid plasma flow system is shown to have more effective photocatalytic activity even under visible light irradiation, in spite of being TiC.

Fig. 10
figure 10

Effect of C-TiO2 nanopowder on degradation rate of MB solution for an hour under visible light irradiation

Full size image

Conclusions

C-TiO2 nanopowder was successfully prepared by droplet injection of solution precursor into downstream plasma flow utilizing the backflow transportation in a DC-RF hybrid plasma flow system. The obtained results are summarized as follows:

  • When injecting droplets of solution precursor into the downstream plasma flow, the droplets are transported by the characteristic backflow, and C-TiO2 nanoparticles are formed by the active reaction with carbon and plasma flow.

  • Synthesized C-TiO2 nanopowder contains anatase-rutile mixed-phase TiO2 and TiC, the contents of which depend on the location of the powder collection, the temperature, and the duration of plasma treatment.

  • Highly functional C-TiO2 nanopowders exhibit a degradation rate of methylene blue higher than that of single-phase anatase TiO2 even under visible light irradiation, in spite of being TiC. Particularly, C-TiO2 nanopowder collected in the downstream region is shown to have more effective photocatalytic activity even under visible light irradiation.