Mist Separation and Sonochemiluminescence under Pulsed Ultrasound

Article pubs.acs.org/JPCA

Mist Separation and Sonochemiluminescence under Pulsed Ultrasound Toru Tuziuti* National Institute of Advanced Industrial Science and Technology (AIST), 2266-98 Shimoshidami, Moriyama-ku, Nagoya 463-8560, Japan ABSTRACT: Differences in the amount of water−mist separation and the intensity of luminol chemiluminescence for pulsed and continuous-wave (CW) ultrasound at 135 kHz have been investigated. The amount of mist generated is estimated using the cooling rate of a copper plate sprayed with the mist. For pulsed operation with an appropriate duty cycle, the cooling rate and the cooling rate per input power to the transducer are higher by 4 and 12 times compared to CW operation, respectively. This is due to the amplitude of the pulsed ultrasound being higher than that for CW ultrasound. Relatively low power pulsed operation can successfully produce both a higher sonochemiluminescence (SCL) intensity and cooling rate than those for CW ultrasound. The sonochemical reaction for pulsed ultrasound occurs at the same input power threshold as that for mist separation, whereas for CW ultrasound, the former threshold is lower than the latter. A higher number of large bubbles is produced with CW ultrasound than that with pulsed ultrasound. To achieve a sound pressure amplitude sufficient for mist separation near the surface of a liquid, it is necessary to expel these bubbles by changing the sound field from resonant standing waves to progressive waves that give rise to capillary waves on the liquid surface.

1. INTRODUCTION Ultrasound waves incident on the interface between a gas and a liquid induce capillary waves that propagate along the liquid surface.1−3 When the sound pressure amplitude near the liquid surface is above a certain threshold, the top of the capillary waves breaks, yielding tiny liquid drops (mist) that disperse. An object placed close to this mist will be subjected to cooling due to evaporation. There have been a number of fundamental studies on atomization using ultrasound waves. Eisenmenger observed capillary waves on a liquid surface that led to atomization over a broad range of ultrasound frequencies from 10 kHz to 1.5 MHz.1 Lang measured the average size of mist droplets in the frequency range of 10−800 kHz4 and obtained the following relationship between the droplet size D, the surface tension T, the density of the liquid ρ, and the ultrasound frequency F D = 0.34(8πT /ρF 2)1/2

several thousand degrees Kelvin and pressures of several hundred atmospheres within the bubbles. Inside of the collapsing bubble, water vapor dissociates to create oxidants such as hydroxyl radicals, ozone, and hydrogen peroxide.7 Chemical reactions involving these oxidants are referred to as sonochemical reactions.8 The use of pulsed ultrasound can promote the creation of such oxidants because an appropriate choice of rest cycle suppresses excessive bubble growth and increases the number of active bubbles. Casadonte et al. showed that, with a suitable duty cycle, the amount of hydrogen peroxide generated by pulsed ultrasound was larger than that for continuous-wave (CW) ultrasound.9 Tuziuti et al. suggested that the mechanism involved in the enhancement of the sonochemical reaction efficiency involved residual sound during the rest cycle and a broadening of the spatial extent of pulsating bubbles.10 In the present study, the author investigated the power dependence of both the extent of mist separation based on the cooling rate and the sonochemiluminescence (SCL) intensity for pulsed ultrasound in order to clarify whether the effective power range for the physical and chemical effects is the same. A mechanism that explains the difference in the power dependence for pulsed and CW ultrasound is suggested. The author attempted to maximize the amount of water mist by varying the duty cycle of the pulsed ultrasound.

(1)

Sato et al. applied the ultrasonic atomization technique to a mixed solution of water and ethanol and successfully accomplished the condensation of ethanol.5 However, to the best of the author’s knowledge, there has been no report on the simultaneous production of both an ultrasonic physical effect (ultrasonic atomization in this paper) and a chemical effect due to ultrasonic cavitation bubbles. The propagation of intense ultrasound waves in a liquid causes cavitation bubbles to form, 6 and these bubbles repeatedly expand and contract at the period of the ultrasound waves. The bubbles collapse when the sound pressure exceeds 1 atm, momentarily giving rise to hot spots with temperatures of © 2012 American Chemical Society

Received: February 6, 2012 Revised: March 7, 2012 Published: March 26, 2012 3593

dx.doi.org/10.1021/jp301205q | J. Phys. Chem. A 2012, 116, 3593−3597

The Journal of Physical Chemistry A

Article

Figure 1. Experimental apparatus: (a) ultrasonic atomization of mist and temperature measurement of a copper plate to estimate the cooling rate due to mist evaporation, (b) SCL intensity measurement.

2. EXPERIMENTAL DETAILS

with the carrier gas. The initial volume of water was 10 mL, and in some cases, this was topped up at a rate of 5.5 mL/min. The initial water temperature was 24 ± 1 °C. The cooling rate was estimated from the initial slope of the plate temperature versus time curve, measured just after sonication began. The size of the mist droplets of water, calculated using eq 1, was 11 μm. 2.2. Power Dependence of the SCL Intensity. The dependence of the SCL intensity on the input power was investigated using luminol (3-aminophthalhydrazide) with the experimental setup shown in Figure 1b. The sonication system was the same as that described in section 2.1. Luminol reacts with OH radicals generated in cavitation bubbles during intense ultrasound irradiation to yield aminophthalate anions and blue fluorescence.11 A solution consisting of 250 mM NaCO3 (Wako) and 2.5 mM luminol (Wako) was prepared using distilled water and was saturated with air. A volume of 10 mL of the luminol solution was placed in the pool on the transducer. Light emitted from the solution was detected using a photomultiplier tube (PMT; Hamamatsu R928), and the output voltage from the PMT was measured using a digital oscilloscope (Yokogawa DL1540C).

2.1. Cooling Rate Measurements. Figure 1a shows the experimental apparatus used to measure the temperature of a copper plate toward which mist due to ultrasonic atomization is directed. Evaporation of the mist from the plate absorbs heat and decreases the temperature of the plate. The plate temperature was measured using a thermocouple placed on its surface. A CW or pulsed sinusoidal signal of 135 kHz generated by a function generator (NF 1942) was amplified using a 55 dB power amplifier (ENI 1140LA). The signal amplitude was the same for both CW and pulsed operation. To produce ultrasound, a Langevin-type transducer (Kouwa) with a diameter of 45 mm was used together with a transducer plate. The transducer plate was surrounded by a circular flange to form a pool, into which distilled water was placed. This was then covered by a transparent plastic hood. The input power to the transducer was measured using a power meter (Towa TAW60A). The mist was transported through a transfer tube to the copper plate using a carrier gas (Ar, 0.05 MPa). Water droplets with large volume splashed from water surface occasionally appeared but were too heavy to be transported 3594



Article

2.3. Measurement of the Sound Pressure Amplitude. A hydrophone (RESON TC4038) was set near the surface in the liquid, and the sound pressure amplitude was measured for both CW and pulsed operation at different input power levels.

when the number of ON cycles is 300. The origin of the horizontal axis represents the CW condition. Whereas the input power decreased monotonically with increasing OFF cycles, the cooling rate initially increased, showed a maximum, and subsequently decreased. The highest cooling rate occurred for 300 ON cycles and 500 OFF cycles. Under these conditions, the cooling rate and the cooling rate per unit power were 4 and 12 times higher than those for CW ultrasound, respectively. The results are summarized in Table 1. Figure 4 shows the

3. RESULTS AND DISCUSSION 3.1. Dependence of Copper Plate Cooling Rate on Ultrasound Generation Conditions. Figure 2 shows the

Table 1. Cooling Rate and Cooling Rate Per Unit Power for CW and Pulsed Operation

power (W) (i) cooling rate (°C/s) (ii) cooling rate per power [°C/(s·W)]

(a) CW

(b) pulsing (300 cycles ON and 500 cycles OFF)

ratio (b)/ (a)

35 1.5 × 10−2 4.2 × 10−4

12 5.9 × 10−2 4.9 × 10−3

0.34 4.0 12

Figure 2. Change in plate temperature with time for CW operation and pulsed operation for a repetition of 300 ON cycles and 1000 OFF cycles.

Figure 4. Images of mist; optimum condition (300 ON cycles and 500 OFF cycles) of pulse (right) and CW (left). These images were captured without a hood in Figure 1.

decrease in the copper plate temperature with time due to exposure to mist produced by ultrasonic atomization for both pulsed and CW operation. It can be seen that a larger and more rapid temperature drop occurs for pulsed ultrasound. Figure 3 shows the dependence of the input power to the transducer and the cooling rate on the number of OFF cycles

images of mist at the optimum condition (300 ON cycles and 500 OFF cycles) of pulse together with that at CW. It is clearly seen that the amount of mist at pulse is larger than that at CW. Thus, for appropriate pulse conditions, a higher cooling rate could be achieved than that for CW ultrasound. During pulsed operation, it is believed that the generation of large bubbles due to coalescence is suppressed, so that a high sound pressure amplitude is maintained near the surface of the liquid. This may be responsible for the larger amount of mist generated during pulsed operation. 3.2. Input Power Thresholds for Cooling and SCL Effects. Figure 5 shows the dependence of the cooling rate and the SCL intensity on the input power. The pulse conditions were a repetition of 300 ON cycles and 1000 OFF cycles. For CW ultrasound, SCL begins to occur at a low power threshold of less than 8 W. The SCL intensity initially increases with power, goes through a maximum, and subsequently decreases. In contrast, no appreciable cooling occurs until a much higher power threshold is reached, and the cooling rate then increases with power. However, at these power levels, the SCL intensity is already decreasing. In the case of pulsed ultrasound, cooling begins to occur at the same low power threshold at which SCL begins. Thus, at power levels of less than 8 W, pulsed operation gives rise to a higher mist production rate and SCL intensity than CW operation. Above the threshold, the cooling seems to increase continuously with power for both CW and pulsed operation, although the data are scattered. The cooling rate for relatively low CW or pulsed power levels seems to be lower for the case where the water was topped up at 5.5 mL/min. It is thought

Figure 3. Dependence of the transducer input power and cooling rate on the number of OFF cycles. 3595



Article

pressure amplitude on the input power for CW and pulsed operation at relatively low power. The pulse conditions were a repetition of 300 ON cycles and 1000 OFF cycles. It is noteworthy that for CW operation, the sound pressure amplitude saturates as the power increases, whereas it increases monotonically for pulsed operation. This result implies that the amount of bubbles created during CW operation is more than that for pulsed operation. Indeed, bubbles were observed at CW even at small power (0.8 W), as seen in Figure 7, where a

Figure 5. Input power dependence of the cooling rate and SCL intensity for CW operation and pulsed operation with 300 ON cycles and 1000 OFF cycles.

that the additional supply of liquid caused an increase in the load on the transducer, and the damping effect led to a reduction in the sound pressure amplitude at the liquid surface and therefore in the mist generation rate. It should be noted that the lower power threshold for SCL than that for mist generation is consistent with the results of Oyama et al.,12 who showed that the maximum yield in the sonochemical oxidation of a bilirubin aqueous solution occurred at an ultrasonic power lower than that for saturation of the physical emulsifying action in a water−fluidic paraffin mixture. 3.3. Dependence of Sound Pressure Amplitude on Input Power. Figure 6 shows the dependence of the sound

Figure 7. Captured image of bubbles (indicated with dotted circles) near the center of the pool at CW of 0.8 W. The laser was used for illumination.

laser (NEOARK LDP2-6835B) was used for illumination of bubbles. These bubbles lead to absorption and scattering of sound and cause a decrease in the sound pressure amplitude near the liquid surface. This suppresses the amount of mist generated and gives rise to a lower cooling rate than can be achieved for pulsed operation with an appropriate number of OFF cycles, as seen in Figure 3. The range of power that causes mist separation during CW operation is beyond the range shown in Figure 6. At sufficiently high CW power, the primary Bjerknes force13 may expel bubbles away from the sound beam axis, where a higher sound pressure amplitude is expected to exist. This is thought to be the reason for the decrease in SCL intensity at high CW power. For CW operation, the fact that the mist generation rate increases at higher ultrasonic amplitude whereas the SCL intensity decreases is supported by the conceptual sketch on a change of state for the sonicated liquid from SCL to mist generation by Hiramatsu and Watanabe.14,15 They showed that radiating and vanishing of SCL occur alternately and suggested that mist separation is caused during SCL vanishing. The periodic phenomenon on SCL comes from a periodic change in the dissolved gas concentration through the gas release due to ultrasonic degassing and dissolution of gas with capillary waves that provide wide surface area via the gas−liquid interface. It is reasonable that SCL vanishing occurs during surface vibration excited to a certain amplitude because a standing wave is out of resonance and then the sound field is ineffective for creating pulsating bubbles.16

Figure 6. Input power dependence of the sound pressure amplitude for CW operation and pulsed operation for repetition of 300 ON cycles and 1000 OFF cycles. 3596



Article

Recently, Harada et al.17 investigated whether capillary waves on a liquid irradiated with 2.4 MHz CW ultrasound with an amplitude sufficient for mist separation emit sonoluminescence (SL),6,18,19 which is light produced in bubble cores by a mechanism different from that for SCL. They believed that ultrasonic focusing occurs to overcome the cavitation threshold near the gas−liquid interfaces where capillary waves are excited when mist droplets separate from the bulk solution. They measured the SL intensity at different positions between the surface of the liquid and a transducer and found that, however, it was highest at the position different from that where capillary waves were excited. Such capillary waves are unlikely to be a cause of SL.

(11) Pétrier, C.; Lamy, M.-F.; Francony, A.; Benahcene, A.; David, B.; Renaudin, V.; Gondrexon, N. J. Phys. Chem. A 1994, 98, 10514− 10520. (12) Oyama, H.; Ogata, S.; Nakajima, T.; Tomono, S.; Furukawa, G. Denki Gakkai Zasshi 1942, 62, 548. (13) Hatanaka, S.; Yasui, K.; Tuziuti, T.; Kozuka, T.; Mitome, H. Jpn. J. Appl. Phys. 2001, 40, 3856−3860. (14) Hiramatsu, S.; Watanabe, Y. Denshi Joho Tsushin Gakkai Ronbunshi A 1997, J80-A, 1675−1681. (15) Hiramatsu, S.; Watanabe, Y. Electr. Commun. Jpn. 1999, 82, 58− 65. (16) Tuziuti, T.; Yasui, K.; Kozuka, T.; Towata, A. J. Phys. Chem. A 2010, 114, 7321−7325. (17) Harada, H.; Iwata, N.; Shiratori, K. Jpn. J. Appl. Phys. 2009, 48, 07GH01. (18) McNamara, W. B.; Didenko, Y. T.; Suslick, K. S. Nature 1999, 401, 772−775. (19) Flannigan, D. J.; Suslick, K. S. Nature 2005, 434, 52−55.

4. CONCLUSIONS This study clarified that mist separation under pulsed ultrasound with appropriate duty cycle conditions gives rise to a cooling rate and a cooling rate per unit power that is 4 and 12 times higher than those for CW ultrasound, respectively. In addition, the power dependence of the SCL intensity and the cooling rate is different for CW and pulsed operation. At relatively low power, pulsed operation gave rise to a SCL intensity and a cooling rate that were higher than those at CW. Measurements of the sound pressure amplitude near the liquid surface suggested that the formation of large-sized bubbles during CW operation reduces the amplitude until the primary Bjerknes force expels these bubbles, allowing a sufficiently high amplitude for capillary wave excitation and mist separation. The fact that the creation of such large bubbles is suppressed during pulsed operation allows the inception and promotion of sonochemical reaction and mist separation for the same range of ultrasonic power.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported in part by grants from the Japan Society for the Promotion of Science and the Ministry of Economy, Trade and Industry, Japan.

■

REFERENCES

(1) Eisenmenger, W. Acustica 1959, 9, 327−340. (2) Yazawa, K. Denshi Joho Tsushin Gakkaishi 1989, 72, 441−446. (3) Yule, A. J.; Al-Suleimani, Y. Proc. R. Soc. London, Ser. A 2000, 456, 1069−1085. (4) Lang, R. J. J. Acoust. Soc. Am. 1962, 34, 6−8. (5) Sato, M.; Matsuura, K.; Fujii, T. J. Chem. Phys. 2001, 114, 2382− 2386. (6) Leighton, T. G. The Acoustic Bubble; Academic: London, 1996; pp 439−590. (7) Yasui, K.; Tuziuti, T.; Sivakumar, M.; Iida, Y. J. Chem. Phys. 2005, 122, 224706. (8) Mason, T. J. An introduction to the uses of power ultrasound in chemistry. Sonochemistry; Oxford University Press: New York, 1999. (9) Casadonte, D. J.; Flores, M.; Pétrier, C. Ultrason. Sonochem. 2005, 12, 147−152. (10) Tuziuti, T.; Yasui, K.; Lee, J.; Kozuka, T.; Towata, A.; Iida, Y. J. Phys. Chem. A 2008, 112, 4875−4878. 3597


Mist Separation and Sonochemiluminescence under Pulsed Ultrasound

Recommend Documents