Ultrasonics Sonochemistry xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Ultrasonics Sonochemistry journal homepage: www.elsevier.com/locate/ultson
Separation characteristics of alcohol from aqueous solution by ultrasonic atomization Keiji Yasuda a,⇑, Kyosuke Mochida a, Yoshiyuki Asakura b, Shinobu Koda c a
Department of Chemical Engineering, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8603, Japan Honda Electronics Co., Ltd., Oiwa-cho, Toyohashi 441-3193, Japan c Department of Molecular Design and Engineering, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8603, Japan b
a r t i c l e
i n f o
Article history: Received 30 November 2013 Received in revised form 14 February 2014 Accepted 14 February 2014 Available online xxxx Keywords: Atomization Droplet Vapor Ethanol Methanol Cluster
a b s t r a c t The generation rate of ultrasonically atomized droplets and the alcohol concentration in droplets were estimated by measuring the flow rate and the alcohol concentration of vapors from a bulk solution with a fountain. The effect of the alcohol concentration in the bulk solution on the generation rate of droplets and the alcohol concentration in droplets were investigated. The ultrasonic frequency was 2.4 MHz, and ethanol and methanol aqueous solutions were used as samples. The generation rate of droplets for ethanol was smaller than that for methanol at the same alcohol molar fraction in the bulk solution. For both solutions, at low alcohol concentration in the bulk solution, the alcohol concentration in droplets was lower than that in vapors and the atomized droplets were visible. On the other side, at high concentration, the concentration in droplets exceeded that in vapors and the atomized droplets became invisible. These results could be explained that the alcohol-rich clusters in the bulk solution were preferentially atomized by ultrasonic irradiation. The concentration in droplets for ethanol was higher than that for methanol at low alcohol concentration because the amount of alcohol-rich clusters was larger. When the alcohol molar fraction was greater than 0.6, the atomized droplets almost consisted of pure alcohol. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction When a liquid is irradiated with ultrasound, a fountain arises from the liquid surface and fine liquid droplets are generated from the fountain [1]. This phenomenon is called ultrasonic atomization. The ultrasonic atomization is utilized in various processes such as humidification, aroma diffusion and nanoparticle synthesis [2]. Recently, it was observed that ethanol was separated from the aqueous solution by ultrasonic atomization [3]. The effects of apparatus [4–7] and sample [8,9] conditions on the ethanol separation performance were investigated. It was reported that surfactants [10,11] and amino acids [12] were also enriched by the ultrasonic atomization. However, the separation mechanism is not fully elucidated because the ultrasonic atomization includes the vaporization from the bulk solution with the fountain in addition to droplets. In order to discuss the mechanism of separation by ultrasonic atomization, the atomized droplet amount and the vaporization amount should be estimated separately. The droplet diameter in ultrasonic atomization was investigated by Lang [13]. He changed ultrasonic frequency in the range ⇑ Corresponding author. Tel.: +81 527893623. E-mail address: [email protected] (K. Yasuda).
from 13 to 780 kHz and examined the droplet diameter of molten wax by taking photomicrographs. He proposed an equation describing the influence of ultrasonic frequency, surface tension and density of liquid on the droplet diameter. By using Lang’s equation, droplet diameters of water and ethanol at 2.4 MHz are calculated to be 2.3 and 1.7 lm, respectively. Effect of viscosity was studied by Ramisetty et al. [14]. Recently, Yano et al. [15] measured droplets of ethanol aqueous solution at 2.4 MHz by X-ray scattering and revealed the existence of nanometer-sized droplets at ethanol molar fractions of 0.2 and 1.0. Kobara et al. [16] used a scanning mobility particle sizer and a hand-held particle counter to measure droplet size distributions of ethanol aqueous solution at 2.4 MHz. They observed only nanometer-sized droplets with diameters of about 30 nm at the molar fraction of 0.50. Compared with the research on the droplet diameter, the research on the amount of atomized droplets are few. The reason is difficulty in collecting all droplets generated from an ultrasonic atomizer. The estimation of the atomization amount mainly had been conducted by means of mass change [3,4]. In this method, the sample mass is measured before and after the atomization and the amount of mass loss is regarded as the atomization amount. However, this atomization amount includes the vaporization amount from the bulk solution with a fountain since all
http://dx.doi.org/10.1016/j.ultsonch.2014.02.011 1350-4177/Ó 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
2
K. Yasuda et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
droplets must be ejected from an atomizer vessel by a carrier gas flow at high gas velocity. In our previous study [17], two apparatuses were used to estimate the generation rate of atomized droplets of water. The one was an apparatus with an ultrasonic transducer. Another was an apparatus with a liquid pump and a nozzle to measure the vaporization amount. The generation rate of droplets was able to be estimated by differences of mass change between these two apparatuses. We clarified that the mass generation rate of droplets decreased with increasing ultrasonic frequency at the same ultrasonic intensity. At the same apparent surface area of the fountain, the number of atomized droplets became larger as the ultrasonic frequency increased. The ultrasonic atomization was considered to arise from a combination of the capillary wave at the fountain surface and the cavitation inside the fountain. Ultrasonic atomization studies concerning alcohol aqueous solution have been mainly conducted against ethanol [3–9]. The information of alcohol concentration in droplets is important to elucidate the mechanism of separation by ultrasonic atomization. However, the amount of atomized droplets and the concentration in droplets have not been reported to our knowledge. In this study, ethanol and methanol aqueous solutions were used as atomizing liquids. The effect of the alcohol concentration in the bulk solution on the generation rate of droplets and the alcohol concentration in droplets were investigated at 2.4 MHz. 2. Experimental 2.1. Apparatus Fig. 1(a) shows the schematic diagram of the experimental apparatus with an ultrasonic transducer. The cylindrical vessel was made from transparent polyvinyl chloride resin. The height and inside diameter of vessel were 300 and 100 mm, respectively. A disc-shaped lead zirconate titanate (PZT) ultrasonic transducer (Honda Electronics Co. Ltd.) was installed at the central position of the bottom of vessel. The frequency and diameter of transducer were 2.4 MHz and 14 mm, respectively. Ultrasound was irradiated vertically upwards to the liquid surface. The transducer was driven by a power amplifier (AP400B, ENI) and a signal generator (1941,
NF Corp.) to emit a continuous sinusoidal wave. An effective electric power applied to the transducer was calculated from a voltage at the transducer and a current measured using an oscilloscope (TDS3014B, Tektronix Inc.) and a current probe (TCP202, Tektronix Inc.). The effective power applied to the transducer was 20 W. In the case of water, the ultrasonic power determined by a calorimetric method was 13.4 W (8.78 W/cm2). The atomization threshold intensity was about 3 W/cm2. The carrier gas was dry nitrogen and flowed through the vessel to accompany the vapors and atomized droplets by a fan. The gas inlet was fitted at the position of 250 mm from the transducer surface. The initial height of sample was 30 mm. To eject all atomized droplets from the vessel, the flow rate of carrier gas was set at 3.3 104 m3/s (the superficial gas velocity based on the vertical cross-section of the vessel was 42 mm/s). The vessel was put on an electronic balance. The ultrasonic irradiation time was five minutes. After ultrasonic irradiation, the mass of the bulk liquid in the vessel was measured by the electric balance and the alcohol concentration was determined by the gas chromatograph equipped with a TCD detector (GC323, GL Science Inc.). The fountain shape and height were observed using a video camera. The sample temperature before ultrasonic irradiation was 293 K and the temperature rise in solution after ultrasonic irradiation was within 3 K. Fig. 1(b) shows the schematic diagram of the experimental apparatus with a liquid pump. In order to reproduce the fountain formed by ultrasonic irradiation, a nozzle and a liquid pump were used. The carrier gas was dry nitrogen and vapors were ejected from the vessel top by the fan. The flow rate of carrier gas was 3.3 104 m3/s. The experimental time was five minutes. After experiment, the mass and alcohol concentration of the bulk liquid in the vessel were measured. Special grade chemical ethanol and methanol were purchased from Wako Pure Chemical Industries, Ltd. These reagents were used without further purification. The distillated water was used. 2.2. Estimation of generation rate of droplets and alcohol concentration in droplets By using the apparatus with the ultrasonic transducer, the rate of mass change of bulk solution Ru and the alcohol concentration in the discharge solution Cu for the ultrasound are calculated from the material balances as follows:
Ru ¼ ðMu0 M ut Þ=t
ð1Þ
C u ðMu0 M ut Þ ¼ C 0 M u0 C ut Mut
ð2Þ
where Mu0 and C0 are the initial mass of bulk solution and initial alcohol concentration in the bulk solution, Mut and Cut are the mass of bulk solution and the alcohol concentration in the bulk solution after ultrasonic irradiation, and t is time. By using the apparatus with the liquid pump, the rate of mass change of bulk solution Rp and the alcohol concentration Cp in the discharge solution for the pump are calculated from the material balances as follows:
Fig. 1. Schematic diagram of experimental apparatus: (a) ultrasound and (b) pump.
RP ¼ ðMp0 M pt Þ=t
ð3Þ
C P ðMp0Mpt Þ ¼ C 0 Mp0 C pt M pt
ð4Þ
where Mp0 is the initial mass of bulk solution, Mpt, and Cpt are the mass of bulk solution and alcohol concentration in the bulk solution after the pump experiment. In the case of the apparatus with the ultrasonic transducer, the atomized droplets and the vapors, which are generated from the bulk liquid with a fountain, are ejected from the vessel, and the mass change is the amount of droplets and vapors. For the case of the apparatus with the liquid pump, the vapors, which are
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
3
K. Yasuda et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
generated from the bulk liquid with a fountain, are ejected from the vessel, and the mass change is the amount of vapors. The generation rate of atomized droplets Rd is obtained by next equation.
ð5Þ
And the alcohol concentration in droplets Cd is calculated from mass balance.
C d Rd ¼ C u Ru C p Rp
ð6Þ
3. Results and discussion 3.1. Amount of atomized droplets Fig. 2 shows the effect of the ethanol concentration in the bulk solution on the rates of mass change for apparatuses with the ultrasonic transducer and the liquid pump. In the case of the ultrasound, the mass change is attributed to droplets and vapors. For the case of the pump, the mass change is due to vapors. The rates of mass change for both cases increase with increasing ethanol concentration. This is because the vapor pressure becomes higher as the ethanol concentration becomes greater. Suzuki et al. [9] conducted ultrasonic atomization of ethanol aqueous solution at 2.4 MHz and collected droplets and vapors by a cascaded condenser. They reported that the collected volume increases with increasing ethanol concentration. In Fig. 2, we plotted the values of the rates of mass change for the ultrasound minus those for the pump at the same ethanol concentration (Eq. (5)). These values indicate the generation rate of atomized droplets [17]. The generation rate of droplets becomes higher as the ethanol concentration increases. Fig. 3 shows the effect of the alcohol concentration in the bulk solution on the generation rate of atomized droplets for methanol and ethanol solutions. With increasing methanol concentration, the generation rate of droplets at first increases and becomes almost constant value. At the same concentration, the generation rate of droplets for methanol solution is higher than those for ethanol solution. Burton [18] measured the sound absorption in alcohol aqueous solutions. The sound absorption is mainly caused by internal viscous friction and thermal conductivity. Compared with methanol solution, the sound absorption coefficient of ethanol
Methanol Ethanol
Generation rate of droplets [ g/min ]
Rd ¼ Ru RP
2.0
1.0
0.5
0.0 0
0.2 0.4 0.6 0.8 Alcohol molar fraction in bulk solution [ - ]
1.0
Fig. 3. Effect of alcohol concentration in bulk solution on generation rate of atomized droplets for methanol and ethanol solutions.
solution was large. From these results, it is considered that droplet generation by ultrasonic atomization is suppressed by an increment of sound absorption coefficient. In the solution with large sound absorption coefficient, the amplitude of capillary wave and the cavitation intensity become small because the ultrasonic intensity is attenuated in the solution.
3.2. Alcohol concentration in atomized droplets Fig. 4 shows the effect of the ethanol concentration in the bulk solution on the ethanol concentration in the discharge solution. When the ethanol molar fraction in the bulk solution is lower than 0.2, the ethanol concentration in the discharge solution for the ultrasound is lower than that for the pump. However, the ethanol concentration in the discharge solution for the ultrasound exceeds that for the pump at the ethanol molar fraction of 0.3 and above. Kirpalani and Suzuki [7] atomized ethanol aqueous solution at
1.0 Ultrasound Droplets Pump
1.5
1.0
0.5
Ethanol molar fraction in discharge solution [ - ]
2.0
Rate of mass change [ g/min ]
1.5
0.8
0.6
0.4
0.2
US Droplet Pump
0
0 0
0.2 0.4 0.6 0.8 1.0 Ethanol molar fraction in bulk solution [ - ]
Fig. 2. Effect of ethanol concentration in bulk solution on rates of mass change for apparatuses with ultrasonic transducer and liquid pump.
0
0.2 0.4 0.6 0.8 Ethanol molar fraction in bulk solution [ - ]
1.0
Fig. 4. Effect of ethanol concentration in bulk solution on ethanol concentration in discharge solution.
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
K. Yasuda et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
2.4 MHz and measured the ethanol concentration in collected mists which contained droplets and vapors. The ethanol concentration in collected mists was higher than that in vapors at the ethanol molar fraction above 0.35. The ethanol concentrations in droplets were estimated from Eqs. (1)–(6) and plotted in Fig. 4. The ethanol concentration in droplets is higher than those for the ultrasound and the pump at the ethanol molar fraction of 0.3 and above. When the ethanol molar fraction is higher than 0.5, the ethanol molar fraction in atomized droplets becomes almost one. Fig. 5 shows photographs of ultrasonic atomization for different ethanol concentrations in the bulk solution. In the cases of water and the ethanol solution at the molar fraction of 0.2, the droplets are observed. On the other hand, the droplets are invisible at molar fractions of 0.3 and 0.5. Yano et al. [15] revealed the existence of nanometer-sized droplets when ethanol solutions at molar fractions of 0.2 and 1.0 were atomized by ultrasound at 2.4 MHz. Kobara et al. [16] investigated droplet size distributions of ethanol solutions at 2.4 MHz. In the case at the molar fraction of 0.05, two peaks in the particle size distribution were observed at about 30 nm and 1 lm. On the other hand, for the case at the molar fraction of 0.5, they detected only nanometer-sized droplets with diameters of about 30 nm. In this figure, micrometer-size droplets are visible but nanometer-sized droplets are invisible. It is known that the ethanol aqueous solution is microscopically separated into phases consisted of ethanol-rich and water-rich clusters [19–22]. It is considered that micrometer-sized droplets mainly contain the bulk solution and the nanometer-sized droplets consist of ethanol-rich clusters. These droplets coexist when the ethanol molar fraction is lower than 0.5. As the ethanol molar fraction becomes higher in the range from 0 to 0.5, the amount of nanometer-sized droplets increases and the amount of micrometer-sized droplets decreases. At the molar fraction of 0.3, atomized
(a) 0
(c) 0.3
(b) 0.2
(d) 0.5
Fig. 5. Photographs of ultrasonic atomization for different ethanol concentration in bulk solution: (a) water, (b) ethanol solution at molar fraction of 0.2, (c) molar fraction 0.3, and (d) molar fraction 0.5.
droplets are difficult to be visible and the ethanol concentration in droplets exceeds that in vapors. When the ethanol molar fraction is greater than 0.5, only nanometer-sized droplets are generated and the ethanol molar fraction in droplets becomes almost one. Fig. 6 shows the effect of the methanol concentration in the bulk solution on the methanol concentration in the discharge solution. As the methanol concentration in the bulk solution becomes higher, the methanol concentrations in the discharge solution increase. When the methanol molar fraction in the bulk solution is lower than 0.5, the methanol concentration in droplets is lower than that for the pump. However, the methanol concentration in droplets exceeds that for the pump at the methanol molar fraction of 0.5 and above. Fig. 7 shows photographs of ultrasonic atomization for different methanol concentrations in the bulk solution. At methanol molar fractions of 0.2 and 0.3, the droplets are visible. On the other side, the droplets become invisible at the molar fraction of 0.5. These phenomena resemble those for ethanol solution. Fig. 8 shows the effect of the alcohol concentration in the bulk solution on the alcohol concentration in atomized droplets. At the alcohol molar fraction in the range from 0.05 to 0.5, the alcohol concentration in droplets for ethanol solution is higher than that for methanol solution. However, regardless of alcohol kinds, the alcohol molar fraction becomes almost one at high concentration. Yoshida and Yamaguchi [19] investigated alcohol clusters in aqueous solutions by low-frequency Raman spectroscopy. At the same molar fraction, contributions of pure alcohol clusters to Raman spectra in ethanol solution were large compared with methanol solution. They explained that hydrophobic interactions between ethanol molecules are stronger. Wakisaka et al. [20] examined microheterogeneities of methanol and ethanol aqueous solutions by analyzing mass spectra of clusters. They detected several different kinds of alcohol-rich and water-rich clusters for both solutions. With an increase in alcohol concentration, the amount of alcoholrich clusters became larger and the amount of water rich clusters became smaller. The amount of alcohol-rich clusters in ethanol solution was larger than that in methanol solution at the same molar fraction. Yoshida et al. [21] investigated the dynamics of water molecules in alcohol solutions by NMR. In the case of methanol, the tetrahedral-like water clusters remain at the molar fraction of 0.3. With increasing methanol molar fraction above 0.3, the water network was gradually ruptured and methanol-rich clusters grew. On the other hand, the case of ethanol, the water network started to
Methanol molar fraction in discharge solution [ - ]
4
1.0
0.8
0.6
0.4
0.2
Ultrasound Droplet Pump
0
0
0.2 0.4 0.6 0.8 1.0 Methanol molar fraction in bulk solution [ - ]
Fig. 6. Effect of methanol concentration in bulk solution on methanol concentration in discharge solution.
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
K. Yasuda et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
5
Fig. 7. Photographs of ultrasonic atomization for different methanol concentration in bulk solution: (a) methanol molar fraction of 0.2, (b) molar fraction 0.3, and (c) molar fraction 0.5.
aqueous solutions. To clarify the formation mechanism of nanometer-sized droplets in detail, the droplet size, the number of droplets and clusters in droplets generated by the ultrasonic atomization will be examined in the near future. Through such studies, we would like to fully elucidate the mechanism of separation by ultrasonic atomization.
Alcohol molar fraction in droplets [ - ]
1.0
0.8
0.6
4. Conclusions In this study, ethanol and methanol aqueous solutions were used as atomizing liquids. The influences of the alcohol concentration in the bulk solution on the generation rate of atomized droplets and the alcohol concentration in droplets were investigated. From the results obtained in the study, the following conclusions can be made:
0.4
0.2 Methanol Ethanol 0 0
0.2 0.4 0.6 0.8 Alcohol molar fraction in bulk solution [ - ]
1.0
Fig. 8. Effect of alcohol concentration in bulk solution on the alcohol concentration in atomized droplets.
rupture at the molar fraction of 0.2, which was lower than that for methanol solution. They explained that ethanol perturbs the water network more effectively than methanol because of greater hydrophobicity. Sato and Buchner [22] investigated cooperative and molecular dynamics of alcohol aqueous solutions by dielectric spectroscopy. For ethanol and methanol, alcohol molecules in solution formed a zigzag chain structure similar to pure alcohol at the molar fractions above 0.18 and 0.3, respectively. From these facts, it is thought that alcohol-rich clusters are preferentially atomized by vibration of capillary wave and/or explosion of cavitation. The interaction of water-rich clusters is mainly hydrogen bond. Since the hydrogen bond is much stronger than the hydrophobic interaction of alcohol-rich clusters, waterrich clusters are difficult to form nanometer-sized droplets. At low and middle alcohol concentrations, the alcohol concentration in atomized droplets becomes higher as the amount of alcohol-rich clusters in the bulk solution increases. Compared with methanol solution, the alcohol concentration in droplets for ethanol solution is high because of greater hydrophobicity. However, at high alcohol concentration, the amount of alcohol-rich clusters is enough to generate atomized droplets and the alcohol molar fraction in droplets becomes almost one, regardless of alcohol kinds. We were able to estimate the generation rate of atomized droplets and the concentration in droplets for ethanol and methanol
1. The generation rate of atomized droplets increases with increasing ethanol concentration. At the same concentration, the rate of atomized droplets for ethanol solution is smaller than that for methanol solution because of greater sound absorption. 2. At low alcohol concentration, the concentration in atomized droplets is lower than that in vapors and the atomized droplets are visible. On the other side, at high alcohol concentration, the alcohol concentration in atomized droplets is higher than that in vapors and the atomized droplets becomes invisible. These results could be explained that the alcohol-rich clusters are preferentially atomized by ultrasonic irradiation. 3. Compared with methanol solution, the concentration in droplets for ethanol solution is high at low concentration because the amount of alcohol-rich clusters is larger. When the alcohol molar fraction is higher than 0.6, the atomized droplets almost consist of pure alcohol.
Acknowledgement The authors are grateful for financial support by a Grant-in-Aid for Science Research (No. 20560713) from the Ministry of Education, Culture, Sports, Science and Technology of Japan. References [1] [2] [3] [4]
W.R. Wood, A.L. Loomis, Philos. Mag. 4 (1927) 417–436. S.E. Skrabalak, K.S. Suslick, J. Am. Chem. Soc. 127 (2005) 9990–9991. M. Sato, K. Matsuura, T. Fujii, J. Chem. Phys. 114 (2001) 2382–2386. Y. Kawase, T. Masuya, K. Yasuda, M. Nakamura, J. Chem. Eng. Jpn. 39 (2006) 334–339. [5] K. Matsuura, T. Fukazu, F. Abe, T. Sekimoto, T. Tomishige, AIChE J. 53 (2007) 737–740.
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
6
K. Yasuda et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
[6] S. Nii, M. Toki, S. Watanabe, K. Suzuki, K. Matsuura, T. Fukazu, J. Chem. Eng. Jpn. 43 (2010) 99–103. [7] D.M. Kirpalani, K. Suzuki, Ultrason. Sonochem. 18 (2011) 1012–1017. [8] K. Hamai, N. Takenaka, B. Nanzai, K. Okitsu, H. Bandow, Y. Maeda, Ultrason. Sonochem. 16 (2009) 150–154. [9] K. Suzuki, T.W. McCracken, D.M. Kirpalani, J. Chem. Eng. Jpn. 41 (2008) 855–861. [10] S. Nii, K. Matsuura, T. Fukazu, M. Toki, F. Kawaizumi, Chem. Eng. Res. Des. 84 (2006) 412–415. [11] B. Jimmy, S. Kentish, F. Grieser, M. Ashokkumar, Langumuir 24 (2008) 10133– 10137. [12] A. Suzuki, H. Maruyama, H. Seki, Y. Matsukawa, N. Inoue, Ind. Eng. Chem. Res. 45 (2006) 830–833. [13] R.J. Lang, J. Acoust. Soc. Am. 34 (1962) 6–8. [14] K.A. Ramisetty, A.B. Pandit, P.R. Gogate, Ultrason. Sonochem. 20 (2013) 254–264.
[15] Y.F. Yano, K. Matsuura, T. Fukazu, F. Abe, A. Wakisaka, H. Kobara, K. Kaneko, A. Kumagai, Y. Katsuya, M. Tanaka, J. Chem. Phys. 127 (2007) 0311011– 0311014. [16] H. Kobara, M. Tamiya, A. Wakisaka, T. Fukazu, K. Matsuura, AIChE J. 56 (2010) 810–814. [17] K. Yasuda, H. Honma, Z. Xu, Y. Asakura, S. Koda, Jpn. J. Appl. Phys. 50 (2011) 07HE2311–07HE2315. [18] C.J. Burton, J. Acoust. Soc. Am. 20 (1948) 186–199. [19] K. Yoshida, T. Yamaguchi, Z. Naturforsch, Zeitschrift für Naturforschung 56a (2001) 529–536. http://www.znaturforsch.com. [20] A. Wakisaka, K. Matsuura, M. Uranaga, T. Sekimoto, M. Takahashi, J. Mol. Liq. 160 (2011) 103–108. [21] K. Yoshida, A. Kitajo, T. Yamaguchi, J. Mol. Liq. 125 (2006) 158–163. [22] T. Sato, R. Buchner, J. Mol. Liq. 117 (2005) 23–31.
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
Contents lists available at ScienceDirect
Ultrasonics Sonochemistry journal homepage: www.elsevier.com/locate/ultson
Separation characteristics of alcohol from aqueous solution by ultrasonic atomization Keiji Yasuda a,⇑, Kyosuke Mochida a, Yoshiyuki Asakura b, Shinobu Koda c a
Department of Chemical Engineering, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8603, Japan Honda Electronics Co., Ltd., Oiwa-cho, Toyohashi 441-3193, Japan c Department of Molecular Design and Engineering, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8603, Japan b
a r t i c l e
i n f o
Article history: Received 30 November 2013 Received in revised form 14 February 2014 Accepted 14 February 2014 Available online xxxx Keywords: Atomization Droplet Vapor Ethanol Methanol Cluster
a b s t r a c t The generation rate of ultrasonically atomized droplets and the alcohol concentration in droplets were estimated by measuring the flow rate and the alcohol concentration of vapors from a bulk solution with a fountain. The effect of the alcohol concentration in the bulk solution on the generation rate of droplets and the alcohol concentration in droplets were investigated. The ultrasonic frequency was 2.4 MHz, and ethanol and methanol aqueous solutions were used as samples. The generation rate of droplets for ethanol was smaller than that for methanol at the same alcohol molar fraction in the bulk solution. For both solutions, at low alcohol concentration in the bulk solution, the alcohol concentration in droplets was lower than that in vapors and the atomized droplets were visible. On the other side, at high concentration, the concentration in droplets exceeded that in vapors and the atomized droplets became invisible. These results could be explained that the alcohol-rich clusters in the bulk solution were preferentially atomized by ultrasonic irradiation. The concentration in droplets for ethanol was higher than that for methanol at low alcohol concentration because the amount of alcohol-rich clusters was larger. When the alcohol molar fraction was greater than 0.6, the atomized droplets almost consisted of pure alcohol. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction When a liquid is irradiated with ultrasound, a fountain arises from the liquid surface and fine liquid droplets are generated from the fountain [1]. This phenomenon is called ultrasonic atomization. The ultrasonic atomization is utilized in various processes such as humidification, aroma diffusion and nanoparticle synthesis [2]. Recently, it was observed that ethanol was separated from the aqueous solution by ultrasonic atomization [3]. The effects of apparatus [4–7] and sample [8,9] conditions on the ethanol separation performance were investigated. It was reported that surfactants [10,11] and amino acids [12] were also enriched by the ultrasonic atomization. However, the separation mechanism is not fully elucidated because the ultrasonic atomization includes the vaporization from the bulk solution with the fountain in addition to droplets. In order to discuss the mechanism of separation by ultrasonic atomization, the atomized droplet amount and the vaporization amount should be estimated separately. The droplet diameter in ultrasonic atomization was investigated by Lang [13]. He changed ultrasonic frequency in the range ⇑ Corresponding author. Tel.: +81 527893623. E-mail address: [email protected] (K. Yasuda).
from 13 to 780 kHz and examined the droplet diameter of molten wax by taking photomicrographs. He proposed an equation describing the influence of ultrasonic frequency, surface tension and density of liquid on the droplet diameter. By using Lang’s equation, droplet diameters of water and ethanol at 2.4 MHz are calculated to be 2.3 and 1.7 lm, respectively. Effect of viscosity was studied by Ramisetty et al. [14]. Recently, Yano et al. [15] measured droplets of ethanol aqueous solution at 2.4 MHz by X-ray scattering and revealed the existence of nanometer-sized droplets at ethanol molar fractions of 0.2 and 1.0. Kobara et al. [16] used a scanning mobility particle sizer and a hand-held particle counter to measure droplet size distributions of ethanol aqueous solution at 2.4 MHz. They observed only nanometer-sized droplets with diameters of about 30 nm at the molar fraction of 0.50. Compared with the research on the droplet diameter, the research on the amount of atomized droplets are few. The reason is difficulty in collecting all droplets generated from an ultrasonic atomizer. The estimation of the atomization amount mainly had been conducted by means of mass change [3,4]. In this method, the sample mass is measured before and after the atomization and the amount of mass loss is regarded as the atomization amount. However, this atomization amount includes the vaporization amount from the bulk solution with a fountain since all
http://dx.doi.org/10.1016/j.ultsonch.2014.02.011 1350-4177/Ó 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
2
K. Yasuda et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
droplets must be ejected from an atomizer vessel by a carrier gas flow at high gas velocity. In our previous study [17], two apparatuses were used to estimate the generation rate of atomized droplets of water. The one was an apparatus with an ultrasonic transducer. Another was an apparatus with a liquid pump and a nozzle to measure the vaporization amount. The generation rate of droplets was able to be estimated by differences of mass change between these two apparatuses. We clarified that the mass generation rate of droplets decreased with increasing ultrasonic frequency at the same ultrasonic intensity. At the same apparent surface area of the fountain, the number of atomized droplets became larger as the ultrasonic frequency increased. The ultrasonic atomization was considered to arise from a combination of the capillary wave at the fountain surface and the cavitation inside the fountain. Ultrasonic atomization studies concerning alcohol aqueous solution have been mainly conducted against ethanol [3–9]. The information of alcohol concentration in droplets is important to elucidate the mechanism of separation by ultrasonic atomization. However, the amount of atomized droplets and the concentration in droplets have not been reported to our knowledge. In this study, ethanol and methanol aqueous solutions were used as atomizing liquids. The effect of the alcohol concentration in the bulk solution on the generation rate of droplets and the alcohol concentration in droplets were investigated at 2.4 MHz. 2. Experimental 2.1. Apparatus Fig. 1(a) shows the schematic diagram of the experimental apparatus with an ultrasonic transducer. The cylindrical vessel was made from transparent polyvinyl chloride resin. The height and inside diameter of vessel were 300 and 100 mm, respectively. A disc-shaped lead zirconate titanate (PZT) ultrasonic transducer (Honda Electronics Co. Ltd.) was installed at the central position of the bottom of vessel. The frequency and diameter of transducer were 2.4 MHz and 14 mm, respectively. Ultrasound was irradiated vertically upwards to the liquid surface. The transducer was driven by a power amplifier (AP400B, ENI) and a signal generator (1941,
NF Corp.) to emit a continuous sinusoidal wave. An effective electric power applied to the transducer was calculated from a voltage at the transducer and a current measured using an oscilloscope (TDS3014B, Tektronix Inc.) and a current probe (TCP202, Tektronix Inc.). The effective power applied to the transducer was 20 W. In the case of water, the ultrasonic power determined by a calorimetric method was 13.4 W (8.78 W/cm2). The atomization threshold intensity was about 3 W/cm2. The carrier gas was dry nitrogen and flowed through the vessel to accompany the vapors and atomized droplets by a fan. The gas inlet was fitted at the position of 250 mm from the transducer surface. The initial height of sample was 30 mm. To eject all atomized droplets from the vessel, the flow rate of carrier gas was set at 3.3 104 m3/s (the superficial gas velocity based on the vertical cross-section of the vessel was 42 mm/s). The vessel was put on an electronic balance. The ultrasonic irradiation time was five minutes. After ultrasonic irradiation, the mass of the bulk liquid in the vessel was measured by the electric balance and the alcohol concentration was determined by the gas chromatograph equipped with a TCD detector (GC323, GL Science Inc.). The fountain shape and height were observed using a video camera. The sample temperature before ultrasonic irradiation was 293 K and the temperature rise in solution after ultrasonic irradiation was within 3 K. Fig. 1(b) shows the schematic diagram of the experimental apparatus with a liquid pump. In order to reproduce the fountain formed by ultrasonic irradiation, a nozzle and a liquid pump were used. The carrier gas was dry nitrogen and vapors were ejected from the vessel top by the fan. The flow rate of carrier gas was 3.3 104 m3/s. The experimental time was five minutes. After experiment, the mass and alcohol concentration of the bulk liquid in the vessel were measured. Special grade chemical ethanol and methanol were purchased from Wako Pure Chemical Industries, Ltd. These reagents were used without further purification. The distillated water was used. 2.2. Estimation of generation rate of droplets and alcohol concentration in droplets By using the apparatus with the ultrasonic transducer, the rate of mass change of bulk solution Ru and the alcohol concentration in the discharge solution Cu for the ultrasound are calculated from the material balances as follows:
Ru ¼ ðMu0 M ut Þ=t
ð1Þ
C u ðMu0 M ut Þ ¼ C 0 M u0 C ut Mut
ð2Þ
where Mu0 and C0 are the initial mass of bulk solution and initial alcohol concentration in the bulk solution, Mut and Cut are the mass of bulk solution and the alcohol concentration in the bulk solution after ultrasonic irradiation, and t is time. By using the apparatus with the liquid pump, the rate of mass change of bulk solution Rp and the alcohol concentration Cp in the discharge solution for the pump are calculated from the material balances as follows:
Fig. 1. Schematic diagram of experimental apparatus: (a) ultrasound and (b) pump.
RP ¼ ðMp0 M pt Þ=t
ð3Þ
C P ðMp0Mpt Þ ¼ C 0 Mp0 C pt M pt
ð4Þ
where Mp0 is the initial mass of bulk solution, Mpt, and Cpt are the mass of bulk solution and alcohol concentration in the bulk solution after the pump experiment. In the case of the apparatus with the ultrasonic transducer, the atomized droplets and the vapors, which are generated from the bulk liquid with a fountain, are ejected from the vessel, and the mass change is the amount of droplets and vapors. For the case of the apparatus with the liquid pump, the vapors, which are
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
3
K. Yasuda et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
generated from the bulk liquid with a fountain, are ejected from the vessel, and the mass change is the amount of vapors. The generation rate of atomized droplets Rd is obtained by next equation.
ð5Þ
And the alcohol concentration in droplets Cd is calculated from mass balance.
C d Rd ¼ C u Ru C p Rp
ð6Þ
3. Results and discussion 3.1. Amount of atomized droplets Fig. 2 shows the effect of the ethanol concentration in the bulk solution on the rates of mass change for apparatuses with the ultrasonic transducer and the liquid pump. In the case of the ultrasound, the mass change is attributed to droplets and vapors. For the case of the pump, the mass change is due to vapors. The rates of mass change for both cases increase with increasing ethanol concentration. This is because the vapor pressure becomes higher as the ethanol concentration becomes greater. Suzuki et al. [9] conducted ultrasonic atomization of ethanol aqueous solution at 2.4 MHz and collected droplets and vapors by a cascaded condenser. They reported that the collected volume increases with increasing ethanol concentration. In Fig. 2, we plotted the values of the rates of mass change for the ultrasound minus those for the pump at the same ethanol concentration (Eq. (5)). These values indicate the generation rate of atomized droplets [17]. The generation rate of droplets becomes higher as the ethanol concentration increases. Fig. 3 shows the effect of the alcohol concentration in the bulk solution on the generation rate of atomized droplets for methanol and ethanol solutions. With increasing methanol concentration, the generation rate of droplets at first increases and becomes almost constant value. At the same concentration, the generation rate of droplets for methanol solution is higher than those for ethanol solution. Burton [18] measured the sound absorption in alcohol aqueous solutions. The sound absorption is mainly caused by internal viscous friction and thermal conductivity. Compared with methanol solution, the sound absorption coefficient of ethanol
Methanol Ethanol
Generation rate of droplets [ g/min ]
Rd ¼ Ru RP
2.0
1.0
0.5
0.0 0
0.2 0.4 0.6 0.8 Alcohol molar fraction in bulk solution [ - ]
1.0
Fig. 3. Effect of alcohol concentration in bulk solution on generation rate of atomized droplets for methanol and ethanol solutions.
solution was large. From these results, it is considered that droplet generation by ultrasonic atomization is suppressed by an increment of sound absorption coefficient. In the solution with large sound absorption coefficient, the amplitude of capillary wave and the cavitation intensity become small because the ultrasonic intensity is attenuated in the solution.
3.2. Alcohol concentration in atomized droplets Fig. 4 shows the effect of the ethanol concentration in the bulk solution on the ethanol concentration in the discharge solution. When the ethanol molar fraction in the bulk solution is lower than 0.2, the ethanol concentration in the discharge solution for the ultrasound is lower than that for the pump. However, the ethanol concentration in the discharge solution for the ultrasound exceeds that for the pump at the ethanol molar fraction of 0.3 and above. Kirpalani and Suzuki [7] atomized ethanol aqueous solution at
1.0 Ultrasound Droplets Pump
1.5
1.0
0.5
Ethanol molar fraction in discharge solution [ - ]
2.0
Rate of mass change [ g/min ]
1.5
0.8
0.6
0.4
0.2
US Droplet Pump
0
0 0
0.2 0.4 0.6 0.8 1.0 Ethanol molar fraction in bulk solution [ - ]
Fig. 2. Effect of ethanol concentration in bulk solution on rates of mass change for apparatuses with ultrasonic transducer and liquid pump.
0
0.2 0.4 0.6 0.8 Ethanol molar fraction in bulk solution [ - ]
1.0
Fig. 4. Effect of ethanol concentration in bulk solution on ethanol concentration in discharge solution.
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
K. Yasuda et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
2.4 MHz and measured the ethanol concentration in collected mists which contained droplets and vapors. The ethanol concentration in collected mists was higher than that in vapors at the ethanol molar fraction above 0.35. The ethanol concentrations in droplets were estimated from Eqs. (1)–(6) and plotted in Fig. 4. The ethanol concentration in droplets is higher than those for the ultrasound and the pump at the ethanol molar fraction of 0.3 and above. When the ethanol molar fraction is higher than 0.5, the ethanol molar fraction in atomized droplets becomes almost one. Fig. 5 shows photographs of ultrasonic atomization for different ethanol concentrations in the bulk solution. In the cases of water and the ethanol solution at the molar fraction of 0.2, the droplets are observed. On the other hand, the droplets are invisible at molar fractions of 0.3 and 0.5. Yano et al. [15] revealed the existence of nanometer-sized droplets when ethanol solutions at molar fractions of 0.2 and 1.0 were atomized by ultrasound at 2.4 MHz. Kobara et al. [16] investigated droplet size distributions of ethanol solutions at 2.4 MHz. In the case at the molar fraction of 0.05, two peaks in the particle size distribution were observed at about 30 nm and 1 lm. On the other hand, for the case at the molar fraction of 0.5, they detected only nanometer-sized droplets with diameters of about 30 nm. In this figure, micrometer-size droplets are visible but nanometer-sized droplets are invisible. It is known that the ethanol aqueous solution is microscopically separated into phases consisted of ethanol-rich and water-rich clusters [19–22]. It is considered that micrometer-sized droplets mainly contain the bulk solution and the nanometer-sized droplets consist of ethanol-rich clusters. These droplets coexist when the ethanol molar fraction is lower than 0.5. As the ethanol molar fraction becomes higher in the range from 0 to 0.5, the amount of nanometer-sized droplets increases and the amount of micrometer-sized droplets decreases. At the molar fraction of 0.3, atomized
(a) 0
(c) 0.3
(b) 0.2
(d) 0.5
Fig. 5. Photographs of ultrasonic atomization for different ethanol concentration in bulk solution: (a) water, (b) ethanol solution at molar fraction of 0.2, (c) molar fraction 0.3, and (d) molar fraction 0.5.
droplets are difficult to be visible and the ethanol concentration in droplets exceeds that in vapors. When the ethanol molar fraction is greater than 0.5, only nanometer-sized droplets are generated and the ethanol molar fraction in droplets becomes almost one. Fig. 6 shows the effect of the methanol concentration in the bulk solution on the methanol concentration in the discharge solution. As the methanol concentration in the bulk solution becomes higher, the methanol concentrations in the discharge solution increase. When the methanol molar fraction in the bulk solution is lower than 0.5, the methanol concentration in droplets is lower than that for the pump. However, the methanol concentration in droplets exceeds that for the pump at the methanol molar fraction of 0.5 and above. Fig. 7 shows photographs of ultrasonic atomization for different methanol concentrations in the bulk solution. At methanol molar fractions of 0.2 and 0.3, the droplets are visible. On the other side, the droplets become invisible at the molar fraction of 0.5. These phenomena resemble those for ethanol solution. Fig. 8 shows the effect of the alcohol concentration in the bulk solution on the alcohol concentration in atomized droplets. At the alcohol molar fraction in the range from 0.05 to 0.5, the alcohol concentration in droplets for ethanol solution is higher than that for methanol solution. However, regardless of alcohol kinds, the alcohol molar fraction becomes almost one at high concentration. Yoshida and Yamaguchi [19] investigated alcohol clusters in aqueous solutions by low-frequency Raman spectroscopy. At the same molar fraction, contributions of pure alcohol clusters to Raman spectra in ethanol solution were large compared with methanol solution. They explained that hydrophobic interactions between ethanol molecules are stronger. Wakisaka et al. [20] examined microheterogeneities of methanol and ethanol aqueous solutions by analyzing mass spectra of clusters. They detected several different kinds of alcohol-rich and water-rich clusters for both solutions. With an increase in alcohol concentration, the amount of alcoholrich clusters became larger and the amount of water rich clusters became smaller. The amount of alcohol-rich clusters in ethanol solution was larger than that in methanol solution at the same molar fraction. Yoshida et al. [21] investigated the dynamics of water molecules in alcohol solutions by NMR. In the case of methanol, the tetrahedral-like water clusters remain at the molar fraction of 0.3. With increasing methanol molar fraction above 0.3, the water network was gradually ruptured and methanol-rich clusters grew. On the other hand, the case of ethanol, the water network started to
Methanol molar fraction in discharge solution [ - ]
4
1.0
0.8
0.6
0.4
0.2
Ultrasound Droplet Pump
0
0
0.2 0.4 0.6 0.8 1.0 Methanol molar fraction in bulk solution [ - ]
Fig. 6. Effect of methanol concentration in bulk solution on methanol concentration in discharge solution.
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
K. Yasuda et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
5
Fig. 7. Photographs of ultrasonic atomization for different methanol concentration in bulk solution: (a) methanol molar fraction of 0.2, (b) molar fraction 0.3, and (c) molar fraction 0.5.
aqueous solutions. To clarify the formation mechanism of nanometer-sized droplets in detail, the droplet size, the number of droplets and clusters in droplets generated by the ultrasonic atomization will be examined in the near future. Through such studies, we would like to fully elucidate the mechanism of separation by ultrasonic atomization.
Alcohol molar fraction in droplets [ - ]
1.0
0.8
0.6
4. Conclusions In this study, ethanol and methanol aqueous solutions were used as atomizing liquids. The influences of the alcohol concentration in the bulk solution on the generation rate of atomized droplets and the alcohol concentration in droplets were investigated. From the results obtained in the study, the following conclusions can be made:
0.4
0.2 Methanol Ethanol 0 0
0.2 0.4 0.6 0.8 Alcohol molar fraction in bulk solution [ - ]
1.0
Fig. 8. Effect of alcohol concentration in bulk solution on the alcohol concentration in atomized droplets.
rupture at the molar fraction of 0.2, which was lower than that for methanol solution. They explained that ethanol perturbs the water network more effectively than methanol because of greater hydrophobicity. Sato and Buchner [22] investigated cooperative and molecular dynamics of alcohol aqueous solutions by dielectric spectroscopy. For ethanol and methanol, alcohol molecules in solution formed a zigzag chain structure similar to pure alcohol at the molar fractions above 0.18 and 0.3, respectively. From these facts, it is thought that alcohol-rich clusters are preferentially atomized by vibration of capillary wave and/or explosion of cavitation. The interaction of water-rich clusters is mainly hydrogen bond. Since the hydrogen bond is much stronger than the hydrophobic interaction of alcohol-rich clusters, waterrich clusters are difficult to form nanometer-sized droplets. At low and middle alcohol concentrations, the alcohol concentration in atomized droplets becomes higher as the amount of alcohol-rich clusters in the bulk solution increases. Compared with methanol solution, the alcohol concentration in droplets for ethanol solution is high because of greater hydrophobicity. However, at high alcohol concentration, the amount of alcohol-rich clusters is enough to generate atomized droplets and the alcohol molar fraction in droplets becomes almost one, regardless of alcohol kinds. We were able to estimate the generation rate of atomized droplets and the concentration in droplets for ethanol and methanol
1. The generation rate of atomized droplets increases with increasing ethanol concentration. At the same concentration, the rate of atomized droplets for ethanol solution is smaller than that for methanol solution because of greater sound absorption. 2. At low alcohol concentration, the concentration in atomized droplets is lower than that in vapors and the atomized droplets are visible. On the other side, at high alcohol concentration, the alcohol concentration in atomized droplets is higher than that in vapors and the atomized droplets becomes invisible. These results could be explained that the alcohol-rich clusters are preferentially atomized by ultrasonic irradiation. 3. Compared with methanol solution, the concentration in droplets for ethanol solution is high at low concentration because the amount of alcohol-rich clusters is larger. When the alcohol molar fraction is higher than 0.6, the atomized droplets almost consist of pure alcohol.
Acknowledgement The authors are grateful for financial support by a Grant-in-Aid for Science Research (No. 20560713) from the Ministry of Education, Culture, Sports, Science and Technology of Japan. References [1] [2] [3] [4]
W.R. Wood, A.L. Loomis, Philos. Mag. 4 (1927) 417–436. S.E. Skrabalak, K.S. Suslick, J. Am. Chem. Soc. 127 (2005) 9990–9991. M. Sato, K. Matsuura, T. Fujii, J. Chem. Phys. 114 (2001) 2382–2386. Y. Kawase, T. Masuya, K. Yasuda, M. Nakamura, J. Chem. Eng. Jpn. 39 (2006) 334–339. [5] K. Matsuura, T. Fukazu, F. Abe, T. Sekimoto, T. Tomishige, AIChE J. 53 (2007) 737–740.
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
6
K. Yasuda et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
[6] S. Nii, M. Toki, S. Watanabe, K. Suzuki, K. Matsuura, T. Fukazu, J. Chem. Eng. Jpn. 43 (2010) 99–103. [7] D.M. Kirpalani, K. Suzuki, Ultrason. Sonochem. 18 (2011) 1012–1017. [8] K. Hamai, N. Takenaka, B. Nanzai, K. Okitsu, H. Bandow, Y. Maeda, Ultrason. Sonochem. 16 (2009) 150–154. [9] K. Suzuki, T.W. McCracken, D.M. Kirpalani, J. Chem. Eng. Jpn. 41 (2008) 855–861. [10] S. Nii, K. Matsuura, T. Fukazu, M. Toki, F. Kawaizumi, Chem. Eng. Res. Des. 84 (2006) 412–415. [11] B. Jimmy, S. Kentish, F. Grieser, M. Ashokkumar, Langumuir 24 (2008) 10133– 10137. [12] A. Suzuki, H. Maruyama, H. Seki, Y. Matsukawa, N. Inoue, Ind. Eng. Chem. Res. 45 (2006) 830–833. [13] R.J. Lang, J. Acoust. Soc. Am. 34 (1962) 6–8. [14] K.A. Ramisetty, A.B. Pandit, P.R. Gogate, Ultrason. Sonochem. 20 (2013) 254–264.
[15] Y.F. Yano, K. Matsuura, T. Fukazu, F. Abe, A. Wakisaka, H. Kobara, K. Kaneko, A. Kumagai, Y. Katsuya, M. Tanaka, J. Chem. Phys. 127 (2007) 0311011– 0311014. [16] H. Kobara, M. Tamiya, A. Wakisaka, T. Fukazu, K. Matsuura, AIChE J. 56 (2010) 810–814. [17] K. Yasuda, H. Honma, Z. Xu, Y. Asakura, S. Koda, Jpn. J. Appl. Phys. 50 (2011) 07HE2311–07HE2315. [18] C.J. Burton, J. Acoust. Soc. Am. 20 (1948) 186–199. [19] K. Yoshida, T. Yamaguchi, Z. Naturforsch, Zeitschrift für Naturforschung 56a (2001) 529–536. http://www.znaturforsch.com. [20] A. Wakisaka, K. Matsuura, M. Uranaga, T. Sekimoto, M. Takahashi, J. Mol. Liq. 160 (2011) 103–108. [21] K. Yoshida, A. Kitajo, T. Yamaguchi, J. Mol. Liq. 125 (2006) 158–163. [22] T. Sato, R. Buchner, J. Mol. Liq. 117 (2005) 23–31.
Please cite this article in press as: K. Yasuda et al., Separation characteristics of alcohol from aqueous solution by ultrasonic atomization, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/j.ultsonch.2014.02.011
