- Email: [email protected]
Sonochemical reactions at 640 kHz using an efficient reactor. Oxidation of potassium iodide
SONOCHEMISTRY
ELSEVIER
Ultrasonics Sonochemistry 4 (1997) 289 293
Sonochemical reactions at 640 kHz using an efficient reactor. Oxidation of potassium iodide James D. Seymour a, Henry C. Wallace b, Ram B. Gupta a,, a
Department ofChemicalEngineering, Auburn University, Auburn, AL 36849-5127, USA b Ultrasonic Energy Systems, P.O. Box 15215, Panama City, FL 32406, USA Received 8 August 1997; received in revised form 12 September 1997
Abstract
Ultrasound can be used to oxidize aqueous pollutants. However, due to economic reasons higher oxidation/destruction rates and higher energy efficiency are needed. Recent studies suggest that the higher ultrasound frequencies provide better oxidation rates than the conventional 20 kHz. Another area for improvement is reactor configuration. We have tested two new reactor configurations with proper focusing and reflection of ultrasound for maximum utilization. Reactor configuration plays an important role in the overall efficiency. In the new reactors, transducers and reaction mixture are separated by a polymer acoustic window which allows efficient transfer of ultrasound energy and not the heat from the transducer to the reaction mixture. One reactor at 640 kHz provides a 100% enhancement over the best reported rate for the oxidation of potassium iodide, on a perWatt basis. Experiments conducted at varying initial KI concentrations show interesting behavior. Increasing the KI concentration by over eight fold merely increases the iodine production rate by two fold. This suggests that in the oxidation region surrounding the bubble, the KI concentration is much different than in the bulk. It is proposed that the hydrophobic bubble region has lower and near saturation KI concentration. © 1997 Elsevier Science B.V. Keywords: Sonochemistry; Reactor; Oxidation; Ultrasound; High frequency
1. Introduction
Ultrasound is used in a wide variety of processes such as cleaning, sterilization, floatation, drying, degassing, de-foaming, soldering, plastic welding, drilling, filtration, homogenization, emulsification, dissolution, de-aggregation of powder, biological cell disruption, extraction, crystallization [1-4] and as a stimulus for chemical reactions [5]. By using ultrasound, complicated reactions are performed with inexpensive equipment, and often with fewer steps than conventional methods [6]. For example, syntheses of sodium phenylselenide and samarium diiodide were shown to be hundred times more energy efficient than the equivalent photochemical reactions [7]. In the gasification of aqueous-coal slurry, the use of ultrasound enhanced the hydrocarbon production rate by four fold [8]. In addition, ultrasound is used to oxidize toxic solvents, herbicides and pesticides such as quercetin, chlorinated hydrocarbons [9-12], parathion [13], atra-
* Corresponding author. Tel: (334) 844 2013; Fax: (334) 844 2063; e-mail: [email protected] 1350-4177/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PI1 S 1350-4177 (97) 00039-4
zinc [14], p-nitrophenol [ 15] and p-nitrophenyl acetate [16] in dilute aqueous wastes. The typical ultrasonic decomposition of toxic organics is 10 000 times faster than the natural aerobic oxidation; however, even faster decomposition is needed to carry out the oxidation at the commercial level [14] mainly because of poor energy efficiency. However, in a recent economic analysis of a dilute p-nitrophenol aqueouswaste treatment, the cost of sonochemical oxidation is found to be comparable to incineration [17]. To increase the oxidation efficiency, different reaction parameters such as temperature, static pressure, ultrasound frequency, solvent choice, and dissolved gas have been studied in depth [16,18]. The addition of NaC1 salt enhances oxidation rates by several fold due to hydrophobic interactions [9]. Higher frequencies provide a significant improvement over the common frequency of 20 kHz due to the lower bubble size [19,20]. These improvements are the right steps in making sonochemistry more efficient. An important variable which was not given proper attention is reactor configuration. The main objective of this work is to study reactor configurations while utilizing a high ultrasound fre-
,I.D. Seymour et al. / Ultrasonics Sonochemistry 4 (1997) 289 293
290
quency. Our goal is to provide a new reactor configuration that uses proper focusing and reflection of ultrasound for maximum utilization. The performances of reactor configurations are tested at 640 kHz by oxidation of potassium iodide. The other objective is to obtain a fundamental understanding of sonochemical oxidation by varying KI concentration.
2. Experimental 2.1. Materials and analysis"
Potassium iodide (Fisher Scientific), iodine (Fisher Scientific) for calibration, and distilled water are used as received. Air sparging is done using an air bubbler. The concentration of iodine is analyzed using an UV spectrophotometer (Spectronic Genesis-2) at 355 nm. The calibration is obtained by using standard iodine solution (Fisher Scientific) of different concentrations; the average molar absorptivity for iodine is observed as 6000 M - 1 c m - 1. Supplemental baseline readings were taken at 600 nm to assure consistency of background absorbance. 2.2. Sonochemical reactor
We have used two reactors, A and B, at 640 kHz, designed and fabricated at Ultrasonic Energy Systems, Panama City, FL. Reactor A (Fig. 1 ) is a circular aluminum reactor attached with a 258 W transducer. The transducer is located outside of the reaction chamber, connected to the bulk of the reactor chamber by a brace. The transducer is separated from the reaction chamber by 1.5 in. Ultrasound enters the reaction chamber through a thin acoustic polymer window. This window allows full transmission of ultrasound energy incident on it. The polymer
window also contains the reacting fluid in the chamber. The chamber has exit and inlet ports which allow fluid delivery and recovery. The reaction chamber has a diameter of 5 in and can hold up to 750 mL liquid. The aluminum body of the reactor allows good heat transfer to the cooling water-bath. The circular shape ensures that the ultrasound is not reflected back on the transducer. Reactor B (Fig. 2) is a cylindrical glass reactor attached with a 396 W transducer. Again, the transducer is located on the outside of the reactor chamber. The separation between the transducer and chamber is 1.5 in. The ultrasound then enters a cylindrical chamber that is 3 in in diameter and 4 in long and can hold up to 600 mL liquid. There is a small fluid entry chamber at the top portion of the reaction chamber that is rectangular in shape. This chamber is l in in height and 3 in long with inlet and exit lines. Opposite to the transducer end of the reactor has a concave nose with an air pocket housed in an aluminum cap. This allows all ultrasound incident on the nose to be reflected back into the reacting solution creating a high intensity focus zone of ultrasonic energy. The reflector is a key feature in reactor B which allows for maximum ultrasound utilization. In both reactors, gas sparging is done externally, which ensures only gas-saturated liquid enters the reactor without any gas bubbles. To not have sparging gas bubbles inside the reactor is important because the bubbles can reflect and interfere with the sound field. Sparging is done in an external water-fall tank. One of the two sections of the water-fall tank has a submersible pump for liquid delivery to reactor, and the other side has a gas sparger. An exit line leaving the sonochemical reactor is then placed in the sparger side, so as to provide constant recirculation of the reacting solution. Residence time of the liquid in the tank is kept long enough so that sufficient gas saturation is achieved. 2.3. Acoustic power calibration
.
Line
Window
l"I Fig. 1. Reactor A.
The desire of the sonochemist is to know how much ultrasonic energy is being provided, and not to be confused by the 'waste heat' associated with the transducer's inefficiency. In many calorimetric calibrations, these two sources of heat are not suitably separated [21]. Consider, for instance, a transducer lowered into a water-bath of known volume, run for a known time at a known electrical power input, and the temperature rise is noted. Such a calibration cannot distinguish between the waste heat and the acoustic-phase heat. Indeed, if the transducer were replaced by an electrical heater, the results would be the same as for the transducer! Calibration of acoustic power for our reactors is fairly simple. With the ultrasonic power turned off, an immersion-heater is put into the vessel. The heater is turned
J.D. Seymour et al. / Ultrasonics Sonochemistry 4 (1997) 289 293
Exit Line t Aluminum Cap
|
Inlet Line ~ . 't,-
291
Polymer Window
I
AirP!c[et'"----i--nc-av~G(~i ..........] ~ "
~
Trans ducer
5"
Fig. 2. Reactor B.
on, and with a known heat input the vessel is allowed to stabilize to a measured temperature. The heater is then removed, and the ultrasonic power is increased to the point where the temperature stabilizes at the same temperature as the heater. The ultrasonic energy being absorbed in the vessel is the same as the known heat energy being previously supplied by the heater.
A
'•
300
f ....
2oot L
a ........
I ....
r .... 1 .....
4
~.
E-,------[]
2.4.Procedure All experiments were conducted in an identical manner. A large stock solution of KI in deionized water is prepared. The stock solution is diluted to the desired initial concentration. Each experiment is started by charging 2500 mL solution to a glass water-fall tank. The reactor is kept in a large ice-water bath for cooling. First gas sparging is started in the water-fall tank, then the ultrasound is turned on. 5 mL samples of the reaction solution are taken for analysis, at the beginning and at regular intervals during the reaction. Experiments using initial 3.5 wt.% KI aqueous solutions are performed using both reactors. The amount of iodine produced is measured at different ultrasound exposure times. After analysis and comparison of the two reactor designs, reactor B is used for further experiments due to its better performance. Two additional sets of experiments are performed with varying 1.2 and 0.4 wt.% initial KI concentration using reactor B.
3. Results
In the case of reactor A, 3.5wt.% KI solution, after 25 min of sonication, produced 266.4 gmol of iodine, yielding an average production rate of 2.47 lamol h - 1 W - 1. The production of iodine increased linearly with time passed in the reaction, while the rate of production remained fairly constant (Fig. 3). In the case of reactor B, 3.5 wt.% KI solution,
.
0
5
10
.
.
.
.
.
15
.
.
.
. . . .
20
. . . .
25
30
Time (rain.) Fig. 3. Oxidation of KI using sonochemical reactor A at 640 kHz and 258 W.
after 25min of sonication, produced 496 lamol of iodine, yielding an average production rate of 2.98 gmol h -1 W -1. The initial rate for reactor A is much higher than that in B. For reactor B, after 5 min, 157 gmol of iodine are produced (Fig. 4), yielding an initial production rate of 4.8 gmol h - l W - 1 (Fig. 5). Due to the better performance of reactor B, two additional experiments were carried out with it, by varying initial KI concentration. Similar trends are noticed for these two experiments. A lower reaction rate at lower K ! concentration are observed as expected from the law of mass action. For the 1.2wt.% KI experiment, the initial iodine production rate is 3.4 gmol h -~ W -x and for 0.4 wt.% KI experiment the initial rate is 2.4 gmol h-1 W - 1 (Figs. 4 and 5).
4. Discussion
Decreasing reaction rates with time may be attributed to several factors: rise in temperature of the reactor, or due to inhibition by formed products. Due to the
J.D. Seymour et al. / Ultrasonics Sonochemistry 4 (1997)289 293
292 A
O 600
,,
m
, p....
F. . . .
~ ' ' ' - ' ~' '-'
i '°° /// / / / • 0
200 L / // // /
oo 3s
% wt.% KI t
loo
[]
wt.%
4
40
7 50
II/
i//,,
o~r,
m
12
o
,
.
i
0
. . 20
10
.
.
.
30
(rain.)
Time
Fig. 4. lodine production using sonochemical reactor B at 640 kHz and 396 W.
®
ee
t ' j x
4.5
=_o;4.o
.-
~v~
o
\
[]
X
3.5 wt.% KI
wt.,
o
0.4 wt.% KI
u~3.5:3 ° -
o
a.0
o,, ~ 2 . 5
.E
2.o
o
1.5
- -
,
0
,
,
I
10
,
,
,
,
I
,
20 Time
,
,
I
. . . .
30 (min.)
I
40
~
,
,
50
Fig. 5. Average iodine production rate using sonochemical reactor B at 640 kHz and 396 W.
complexity of the problem we have not explored it in detail at present. Our main goal is compare the overall reactor performance. The iodine production rate is constant in reactor A, perhaps, because of the aluminum body of the reactor which provides better heat dissipation than the glass wall. In a recent work by Entezari and Kruus [19] K I oxidation rates are reported at 20 kHz and at 900 kHz with air sparging, for initial K I concentration of 3.5wt.%. The initial iodine production rate is 0.07gmolh-lW-1 at 2 0 k H z and much higher 2.2 ~tmol h - 1 W - 1 at 900 kHz. Our initial rates of 2.5 g m o l h - t W -1 in reactor A and 4.8 gmol h - 1 W - ~ in reactor B at 640 kHz compares very favorably with the reported results of Entezari and Kruus; reactor B represents an enhancement of two fold over the best reported K I oxidation rate. Another positive aspect of reactor B is that the per-Watt enhancement is accomplished using fifteen-fold increase in power. Usually an increase in power results in a decrease in the per-Watt
oxidation efficiency. Higher powers are needed in the industrial scale sonochemical reactors. The reactors used in our experiments have an unique design feature in comparison to the conventional reactors. The transducer is not directly introduced into the reacting medium in our work. Except for ultrasonic baths which have a poor sonication efficiency, in all conventional sonochemical reactors the transducer is inserted directly into the reacting medium. For example, many low frequency experiments insert the tip of a titanium horn into the reacting solution for direct introduction of ultrasound into the system. All ultrasonic transducers operate at something less than 100% energy efficiency. Of the electrical energy delivered to a transducer, a fraction is converted directly into acoustical energy. This fraction defines the efficiency of the transducer, and the ultimate fate of this energy is to be absorbed as heat by liquid in the reactor. The other fraction of the electrical energy never produces acoustic energy. It is converted directly into heat in the process of providing the ultrasonic output. Thus all the electrical energy delivered to the transducer ultimately becomes heat energy. Some of it goes directly into heat and the rest goes through a phase as acoustic energy before being converted into heat. In reactors A and B, the transducer is mounted 1.5 in from the open end of the vessel, which is covered by a 0.003 in thick acoustic window made of polyethylene. This window allows the ultrasonic energy to enter the vessel, while preventing mixing of the tank water with the liquid in the reactor. In addition, the film provides a barrier to any metal particles breaking away form the transducer surface, hence the process fluid is not contaminated by these metal particles. As a practical matter, only energy in the acoustic phase will make the transit across the 1.5 in space of cold water. At the other end of reactor B is a concave nose that reflects ultrasound back to the center of the reactor, thus concentrating the ultrasound back into the reactor. The concave nose of the reactor has an air pad to reflect ultrasound. This reflection of ultrasound to a concentrated energy field in the reactor is the reason that higher production rates are noticed in reactor B as compared to reactor A. The circular shape of reactor A does not allow the ultrasonic energy to become as concentrated as the concave reflector does. Another interesting difference in the approach used is the external sparging of gas. In order that sparging bubbles do not interfere with sonochemical vessel operation, external sparging is recommended. The purpose is to introduce gas-saturated solution into the vessel, without bubbles, and without introducing sparging hardware into the sound field. This has been accomplished with an external vessel in which the sparging is accomplished by a sparger. A sump in this vessel is located at a low point, below which no bubbles can be seen. The gas-
J.D. Seymour et al. / Ultrasonics Sonochemistry 4 (1997) 289 293
saturated solution from this sump is pumped back into the sonochemical vessel. In practice, the solution from the external vessel can be seen to be free of bubbles, while the solution leaving the sonochemical vessel can be seen to contain many bubbles, evidence of the degassing of the solution within the sound field. Experiments conducted at varying initial KI concentration show an interesting behavior. In the cylindrical glass reactor increasing the KI concentration by over eight fold merely increases the iodine production rate by two fold. The increase is much less than expected from the first order kinetics. This suggests that in the oxidation region surrounding the bubble, K1 concentration is much different than in bulk. The bubble region is hydrophobic in nature [9] hence it will have a poor solubility of KI. Perhaps, by increasing the bulk KI concentration by eight fold, the interface concentration is increased only by two fold. Further work needs to be carried out in order to fully understand the KI oxidation mechanism and partitioning of electrolytes in the hydrophobic region.
5. Conclusion Two different reactor configurations are tested for KI oxidation efficiency at 640 kHz. Reactor configuration plays an important role in the efficient utilization of ultrasound. In the case of one reactor configuration at 640 kHz, a 100% enhancement is observed as seen by the potassium iodide oxidation over the best reported results by Entezari and Kruus at 900 kHz, on a perWatt basis. Increasing the KI concentration by over eight fold merely increases the iodine production rate by two fold. It is proposed that the hydrophobic bubble region has lower and near saturation KI concentration than the bulk.
Acknowledgements This research was partially supported by grants from Alabama NSF-EPSCoR Young Faculty Career Enhancement Program Award to RBG, the Alabama Legacy for Environmental Research Trust Fund (Alabama Department of Public Health), and Auburn University Grant-in-Aid.
References [1] T.J. Mason, J.P. Lorimer, Sonochemistry: Theory, Applications and Uses of Ultrasound in Chemistry, Ellis Horwood, New York, 1989.
293
[2] T.J. Mason, Chemistry with Ultrasound, Critical Reports in Applied Chemistry No. 28, Society for Chemical Industry, 1990. [3] T.J. Mason, Advances in Sonochemistry, Vol. 1, JAI Press, 1990. [4] S.V. Levy, C.M.R. Low, Ultrasound in Synthesis, SpringerVerlag, 1989. [5] K.S. Suslick, Synthetic applications of ultrasound, Modern Synth. Methods 4 (1986) 1 60. [6] C. Sullivan, Sonochemistry a sound investment, Chemistry & Industry 10 (1992) 365--365. [7] C.M.R. Low, Sonochemical generation of allylic phosphoranes an efficient route to conjugated dienes, Synlett. 2 (1991 ) 123 124. [8] D. Punwani, M.C. Mensinger, Slurry-phase gasification of carbonaceous materials using ultrasound in an aqueous media, US Patent 4,954,246, Sept. 4, 1990. [9] J.D. Seymour, R.B. Gupta, Oxidation of aqueous pollutants using ultrasound: salt induced enhancement, Ind. Eng. Chem. Res. 36 (1997) 3453 3457. [10] H.M. Cheung, A. Bhatnagar, G. Jansen, Sonochemical destruction of chlorinated hydrocarbons in dilute aqueous solution, J. Environ. Sci. Technol. 25 (1991) 1510 1512. [11] K. Inazu, Y. Nagata, Y. Maeda, Decomposition of chlorinated hydrocarbons in aqueous solutions by ultrasonic irradiation, Chemistry Letters, Jan. 1 (1993) 57 60. [12] A. Bhatnagar, H.M. Cheung, Sonochemical destruction of chlorinated C 1 and C 2 volatile organic compounds in dilute aqueous solution, J. Environ. Sci. Technol. 28 (1994) 1481 1486. [13] A. Kotronarou, G. Mills, M.R. Hoffmann, Decomposition of parathion in aqueous solution by ultrasound irradiation, J. Environ. Sci. Technol. 26 (1992) 1460-1462. [14] W.C. Koshkinen, K.E. Sellung, J.M. Baker, B.L. Barber, R.H. Dowdy, Ultrasonic decomposition of atrazine and alachlor in water, J. Environ. Sci. Technol. B29 (1994) 581-590. [ 15] M. Cost, G. Mills, P. Gilsson, J. Lakin, Sonochemical degradation ofp-nitrophenol in the presence of chemical components of natural waters, Chemosphere 27 (1993) 1737 1743. [16] I. Hua, R.H. Hochemer, M.R. Hoffmann, Sonolytic hydrolysis of p-nitrophenyl acetate: the role of supercritical water, J. Phys. Chem. 99 (1995) 2335-2342. Also: Disintegrate waste with ultrasound's tinny bubbles, Research & Development Magazine, January (1995) 55. [17] R. Hochemer, M.R. Hoffman, I. Hua, The sonochemical degradation ofp-nitrophenol in a parallel-plate near-field acoustical processor, Presented at the Second International Symposium on Advanced Oxidation Technologies, San Francisco, CA, Feb. 28 March 1 1996. [18] K.S. Suslick, Ultrasound: applications to materials chemistry, in: R.W. Cahn (Ed.), Encyclopedia of Materials Science and Engineering, Pergamon, Oxford, 3rd Suppl., 1992, p. 2093. [19] M.H. Entezari, P. Kruus, Effect of frequency on sonochemical reactions I: oxidation of iodine, Ultrasonics Sonochemistry 1 (1994) $75-$79. [20] M.H. Entezari, P. Kruus, Effect of frequency on sonochemical reactions lI: temperature and intensity effects, Ultrasonics Sonochemistry 3 (1996) 19 24. [21] T. Kimura, T. Sakamoto, J.-M. Leveque, H. Sohmiya, M. Fujita, S. Ikeda, T. Ando, Standardization of ultrasounic power for sonochemical reaction, Ultrasonics Sonochemistry 3 (1996) S157 S161.
ELSEVIER
Ultrasonics Sonochemistry 4 (1997) 289 293
Sonochemical reactions at 640 kHz using an efficient reactor. Oxidation of potassium iodide James D. Seymour a, Henry C. Wallace b, Ram B. Gupta a,, a
Department ofChemicalEngineering, Auburn University, Auburn, AL 36849-5127, USA b Ultrasonic Energy Systems, P.O. Box 15215, Panama City, FL 32406, USA Received 8 August 1997; received in revised form 12 September 1997
Abstract
Ultrasound can be used to oxidize aqueous pollutants. However, due to economic reasons higher oxidation/destruction rates and higher energy efficiency are needed. Recent studies suggest that the higher ultrasound frequencies provide better oxidation rates than the conventional 20 kHz. Another area for improvement is reactor configuration. We have tested two new reactor configurations with proper focusing and reflection of ultrasound for maximum utilization. Reactor configuration plays an important role in the overall efficiency. In the new reactors, transducers and reaction mixture are separated by a polymer acoustic window which allows efficient transfer of ultrasound energy and not the heat from the transducer to the reaction mixture. One reactor at 640 kHz provides a 100% enhancement over the best reported rate for the oxidation of potassium iodide, on a perWatt basis. Experiments conducted at varying initial KI concentrations show interesting behavior. Increasing the KI concentration by over eight fold merely increases the iodine production rate by two fold. This suggests that in the oxidation region surrounding the bubble, the KI concentration is much different than in the bulk. It is proposed that the hydrophobic bubble region has lower and near saturation KI concentration. © 1997 Elsevier Science B.V. Keywords: Sonochemistry; Reactor; Oxidation; Ultrasound; High frequency
1. Introduction
Ultrasound is used in a wide variety of processes such as cleaning, sterilization, floatation, drying, degassing, de-foaming, soldering, plastic welding, drilling, filtration, homogenization, emulsification, dissolution, de-aggregation of powder, biological cell disruption, extraction, crystallization [1-4] and as a stimulus for chemical reactions [5]. By using ultrasound, complicated reactions are performed with inexpensive equipment, and often with fewer steps than conventional methods [6]. For example, syntheses of sodium phenylselenide and samarium diiodide were shown to be hundred times more energy efficient than the equivalent photochemical reactions [7]. In the gasification of aqueous-coal slurry, the use of ultrasound enhanced the hydrocarbon production rate by four fold [8]. In addition, ultrasound is used to oxidize toxic solvents, herbicides and pesticides such as quercetin, chlorinated hydrocarbons [9-12], parathion [13], atra-
* Corresponding author. Tel: (334) 844 2013; Fax: (334) 844 2063; e-mail: [email protected] 1350-4177/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PI1 S 1350-4177 (97) 00039-4
zinc [14], p-nitrophenol [ 15] and p-nitrophenyl acetate [16] in dilute aqueous wastes. The typical ultrasonic decomposition of toxic organics is 10 000 times faster than the natural aerobic oxidation; however, even faster decomposition is needed to carry out the oxidation at the commercial level [14] mainly because of poor energy efficiency. However, in a recent economic analysis of a dilute p-nitrophenol aqueouswaste treatment, the cost of sonochemical oxidation is found to be comparable to incineration [17]. To increase the oxidation efficiency, different reaction parameters such as temperature, static pressure, ultrasound frequency, solvent choice, and dissolved gas have been studied in depth [16,18]. The addition of NaC1 salt enhances oxidation rates by several fold due to hydrophobic interactions [9]. Higher frequencies provide a significant improvement over the common frequency of 20 kHz due to the lower bubble size [19,20]. These improvements are the right steps in making sonochemistry more efficient. An important variable which was not given proper attention is reactor configuration. The main objective of this work is to study reactor configurations while utilizing a high ultrasound fre-
,I.D. Seymour et al. / Ultrasonics Sonochemistry 4 (1997) 289 293
290
quency. Our goal is to provide a new reactor configuration that uses proper focusing and reflection of ultrasound for maximum utilization. The performances of reactor configurations are tested at 640 kHz by oxidation of potassium iodide. The other objective is to obtain a fundamental understanding of sonochemical oxidation by varying KI concentration.
2. Experimental 2.1. Materials and analysis"
Potassium iodide (Fisher Scientific), iodine (Fisher Scientific) for calibration, and distilled water are used as received. Air sparging is done using an air bubbler. The concentration of iodine is analyzed using an UV spectrophotometer (Spectronic Genesis-2) at 355 nm. The calibration is obtained by using standard iodine solution (Fisher Scientific) of different concentrations; the average molar absorptivity for iodine is observed as 6000 M - 1 c m - 1. Supplemental baseline readings were taken at 600 nm to assure consistency of background absorbance. 2.2. Sonochemical reactor
We have used two reactors, A and B, at 640 kHz, designed and fabricated at Ultrasonic Energy Systems, Panama City, FL. Reactor A (Fig. 1 ) is a circular aluminum reactor attached with a 258 W transducer. The transducer is located outside of the reaction chamber, connected to the bulk of the reactor chamber by a brace. The transducer is separated from the reaction chamber by 1.5 in. Ultrasound enters the reaction chamber through a thin acoustic polymer window. This window allows full transmission of ultrasound energy incident on it. The polymer
window also contains the reacting fluid in the chamber. The chamber has exit and inlet ports which allow fluid delivery and recovery. The reaction chamber has a diameter of 5 in and can hold up to 750 mL liquid. The aluminum body of the reactor allows good heat transfer to the cooling water-bath. The circular shape ensures that the ultrasound is not reflected back on the transducer. Reactor B (Fig. 2) is a cylindrical glass reactor attached with a 396 W transducer. Again, the transducer is located on the outside of the reactor chamber. The separation between the transducer and chamber is 1.5 in. The ultrasound then enters a cylindrical chamber that is 3 in in diameter and 4 in long and can hold up to 600 mL liquid. There is a small fluid entry chamber at the top portion of the reaction chamber that is rectangular in shape. This chamber is l in in height and 3 in long with inlet and exit lines. Opposite to the transducer end of the reactor has a concave nose with an air pocket housed in an aluminum cap. This allows all ultrasound incident on the nose to be reflected back into the reacting solution creating a high intensity focus zone of ultrasonic energy. The reflector is a key feature in reactor B which allows for maximum ultrasound utilization. In both reactors, gas sparging is done externally, which ensures only gas-saturated liquid enters the reactor without any gas bubbles. To not have sparging gas bubbles inside the reactor is important because the bubbles can reflect and interfere with the sound field. Sparging is done in an external water-fall tank. One of the two sections of the water-fall tank has a submersible pump for liquid delivery to reactor, and the other side has a gas sparger. An exit line leaving the sonochemical reactor is then placed in the sparger side, so as to provide constant recirculation of the reacting solution. Residence time of the liquid in the tank is kept long enough so that sufficient gas saturation is achieved. 2.3. Acoustic power calibration
.
Line
Window
l"I Fig. 1. Reactor A.
The desire of the sonochemist is to know how much ultrasonic energy is being provided, and not to be confused by the 'waste heat' associated with the transducer's inefficiency. In many calorimetric calibrations, these two sources of heat are not suitably separated [21]. Consider, for instance, a transducer lowered into a water-bath of known volume, run for a known time at a known electrical power input, and the temperature rise is noted. Such a calibration cannot distinguish between the waste heat and the acoustic-phase heat. Indeed, if the transducer were replaced by an electrical heater, the results would be the same as for the transducer! Calibration of acoustic power for our reactors is fairly simple. With the ultrasonic power turned off, an immersion-heater is put into the vessel. The heater is turned
J.D. Seymour et al. / Ultrasonics Sonochemistry 4 (1997) 289 293
Exit Line t Aluminum Cap
|
Inlet Line ~ . 't,-
291
Polymer Window
I
AirP!c[et'"----i--nc-av~G(~i ..........] ~ "
~
Trans ducer
5"
Fig. 2. Reactor B.
on, and with a known heat input the vessel is allowed to stabilize to a measured temperature. The heater is then removed, and the ultrasonic power is increased to the point where the temperature stabilizes at the same temperature as the heater. The ultrasonic energy being absorbed in the vessel is the same as the known heat energy being previously supplied by the heater.
A
'•
300
f ....
2oot L
a ........
I ....
r .... 1 .....
4
~.
E-,------[]
2.4.Procedure All experiments were conducted in an identical manner. A large stock solution of KI in deionized water is prepared. The stock solution is diluted to the desired initial concentration. Each experiment is started by charging 2500 mL solution to a glass water-fall tank. The reactor is kept in a large ice-water bath for cooling. First gas sparging is started in the water-fall tank, then the ultrasound is turned on. 5 mL samples of the reaction solution are taken for analysis, at the beginning and at regular intervals during the reaction. Experiments using initial 3.5 wt.% KI aqueous solutions are performed using both reactors. The amount of iodine produced is measured at different ultrasound exposure times. After analysis and comparison of the two reactor designs, reactor B is used for further experiments due to its better performance. Two additional sets of experiments are performed with varying 1.2 and 0.4 wt.% initial KI concentration using reactor B.
3. Results
In the case of reactor A, 3.5wt.% KI solution, after 25 min of sonication, produced 266.4 gmol of iodine, yielding an average production rate of 2.47 lamol h - 1 W - 1. The production of iodine increased linearly with time passed in the reaction, while the rate of production remained fairly constant (Fig. 3). In the case of reactor B, 3.5 wt.% KI solution,
.
0
5
10
.
.
.
.
.
15
.
.
.
. . . .
20
. . . .
25
30
Time (rain.) Fig. 3. Oxidation of KI using sonochemical reactor A at 640 kHz and 258 W.
after 25min of sonication, produced 496 lamol of iodine, yielding an average production rate of 2.98 gmol h -1 W -1. The initial rate for reactor A is much higher than that in B. For reactor B, after 5 min, 157 gmol of iodine are produced (Fig. 4), yielding an initial production rate of 4.8 gmol h - l W - 1 (Fig. 5). Due to the better performance of reactor B, two additional experiments were carried out with it, by varying initial KI concentration. Similar trends are noticed for these two experiments. A lower reaction rate at lower K ! concentration are observed as expected from the law of mass action. For the 1.2wt.% KI experiment, the initial iodine production rate is 3.4 gmol h -~ W -x and for 0.4 wt.% KI experiment the initial rate is 2.4 gmol h-1 W - 1 (Figs. 4 and 5).
4. Discussion
Decreasing reaction rates with time may be attributed to several factors: rise in temperature of the reactor, or due to inhibition by formed products. Due to the
J.D. Seymour et al. / Ultrasonics Sonochemistry 4 (1997)289 293
292 A
O 600
,,
m
, p....
F. . . .
~ ' ' ' - ' ~' '-'
i '°° /// / / / • 0
200 L / // // /
oo 3s
% wt.% KI t
loo
[]
wt.%
4
40
7 50
II/
i//,,
o~r,
m
12
o
,
.
i
0
. . 20
10
.
.
.
30
(rain.)
Time
Fig. 4. lodine production using sonochemical reactor B at 640 kHz and 396 W.
®
ee
t ' j x
4.5
=_o;4.o
.-
~v~
o
\
[]
X
3.5 wt.% KI
wt.,
o
0.4 wt.% KI
u~3.5:3 ° -
o
a.0
o,, ~ 2 . 5
.E
2.o
o
1.5
- -
,
0
,
,
I
10
,
,
,
,
I
,
20 Time
,
,
I
. . . .
30 (min.)
I
40
~
,
,
50
Fig. 5. Average iodine production rate using sonochemical reactor B at 640 kHz and 396 W.
complexity of the problem we have not explored it in detail at present. Our main goal is compare the overall reactor performance. The iodine production rate is constant in reactor A, perhaps, because of the aluminum body of the reactor which provides better heat dissipation than the glass wall. In a recent work by Entezari and Kruus [19] K I oxidation rates are reported at 20 kHz and at 900 kHz with air sparging, for initial K I concentration of 3.5wt.%. The initial iodine production rate is 0.07gmolh-lW-1 at 2 0 k H z and much higher 2.2 ~tmol h - 1 W - 1 at 900 kHz. Our initial rates of 2.5 g m o l h - t W -1 in reactor A and 4.8 gmol h - 1 W - ~ in reactor B at 640 kHz compares very favorably with the reported results of Entezari and Kruus; reactor B represents an enhancement of two fold over the best reported K I oxidation rate. Another positive aspect of reactor B is that the per-Watt enhancement is accomplished using fifteen-fold increase in power. Usually an increase in power results in a decrease in the per-Watt
oxidation efficiency. Higher powers are needed in the industrial scale sonochemical reactors. The reactors used in our experiments have an unique design feature in comparison to the conventional reactors. The transducer is not directly introduced into the reacting medium in our work. Except for ultrasonic baths which have a poor sonication efficiency, in all conventional sonochemical reactors the transducer is inserted directly into the reacting medium. For example, many low frequency experiments insert the tip of a titanium horn into the reacting solution for direct introduction of ultrasound into the system. All ultrasonic transducers operate at something less than 100% energy efficiency. Of the electrical energy delivered to a transducer, a fraction is converted directly into acoustical energy. This fraction defines the efficiency of the transducer, and the ultimate fate of this energy is to be absorbed as heat by liquid in the reactor. The other fraction of the electrical energy never produces acoustic energy. It is converted directly into heat in the process of providing the ultrasonic output. Thus all the electrical energy delivered to the transducer ultimately becomes heat energy. Some of it goes directly into heat and the rest goes through a phase as acoustic energy before being converted into heat. In reactors A and B, the transducer is mounted 1.5 in from the open end of the vessel, which is covered by a 0.003 in thick acoustic window made of polyethylene. This window allows the ultrasonic energy to enter the vessel, while preventing mixing of the tank water with the liquid in the reactor. In addition, the film provides a barrier to any metal particles breaking away form the transducer surface, hence the process fluid is not contaminated by these metal particles. As a practical matter, only energy in the acoustic phase will make the transit across the 1.5 in space of cold water. At the other end of reactor B is a concave nose that reflects ultrasound back to the center of the reactor, thus concentrating the ultrasound back into the reactor. The concave nose of the reactor has an air pad to reflect ultrasound. This reflection of ultrasound to a concentrated energy field in the reactor is the reason that higher production rates are noticed in reactor B as compared to reactor A. The circular shape of reactor A does not allow the ultrasonic energy to become as concentrated as the concave reflector does. Another interesting difference in the approach used is the external sparging of gas. In order that sparging bubbles do not interfere with sonochemical vessel operation, external sparging is recommended. The purpose is to introduce gas-saturated solution into the vessel, without bubbles, and without introducing sparging hardware into the sound field. This has been accomplished with an external vessel in which the sparging is accomplished by a sparger. A sump in this vessel is located at a low point, below which no bubbles can be seen. The gas-
J.D. Seymour et al. / Ultrasonics Sonochemistry 4 (1997) 289 293
saturated solution from this sump is pumped back into the sonochemical vessel. In practice, the solution from the external vessel can be seen to be free of bubbles, while the solution leaving the sonochemical vessel can be seen to contain many bubbles, evidence of the degassing of the solution within the sound field. Experiments conducted at varying initial KI concentration show an interesting behavior. In the cylindrical glass reactor increasing the KI concentration by over eight fold merely increases the iodine production rate by two fold. The increase is much less than expected from the first order kinetics. This suggests that in the oxidation region surrounding the bubble, K1 concentration is much different than in bulk. The bubble region is hydrophobic in nature [9] hence it will have a poor solubility of KI. Perhaps, by increasing the bulk KI concentration by eight fold, the interface concentration is increased only by two fold. Further work needs to be carried out in order to fully understand the KI oxidation mechanism and partitioning of electrolytes in the hydrophobic region.
5. Conclusion Two different reactor configurations are tested for KI oxidation efficiency at 640 kHz. Reactor configuration plays an important role in the efficient utilization of ultrasound. In the case of one reactor configuration at 640 kHz, a 100% enhancement is observed as seen by the potassium iodide oxidation over the best reported results by Entezari and Kruus at 900 kHz, on a perWatt basis. Increasing the KI concentration by over eight fold merely increases the iodine production rate by two fold. It is proposed that the hydrophobic bubble region has lower and near saturation KI concentration than the bulk.
Acknowledgements This research was partially supported by grants from Alabama NSF-EPSCoR Young Faculty Career Enhancement Program Award to RBG, the Alabama Legacy for Environmental Research Trust Fund (Alabama Department of Public Health), and Auburn University Grant-in-Aid.
References [1] T.J. Mason, J.P. Lorimer, Sonochemistry: Theory, Applications and Uses of Ultrasound in Chemistry, Ellis Horwood, New York, 1989.
293
[2] T.J. Mason, Chemistry with Ultrasound, Critical Reports in Applied Chemistry No. 28, Society for Chemical Industry, 1990. [3] T.J. Mason, Advances in Sonochemistry, Vol. 1, JAI Press, 1990. [4] S.V. Levy, C.M.R. Low, Ultrasound in Synthesis, SpringerVerlag, 1989. [5] K.S. Suslick, Synthetic applications of ultrasound, Modern Synth. Methods 4 (1986) 1 60. [6] C. Sullivan, Sonochemistry a sound investment, Chemistry & Industry 10 (1992) 365--365. [7] C.M.R. Low, Sonochemical generation of allylic phosphoranes an efficient route to conjugated dienes, Synlett. 2 (1991 ) 123 124. [8] D. Punwani, M.C. Mensinger, Slurry-phase gasification of carbonaceous materials using ultrasound in an aqueous media, US Patent 4,954,246, Sept. 4, 1990. [9] J.D. Seymour, R.B. Gupta, Oxidation of aqueous pollutants using ultrasound: salt induced enhancement, Ind. Eng. Chem. Res. 36 (1997) 3453 3457. [10] H.M. Cheung, A. Bhatnagar, G. Jansen, Sonochemical destruction of chlorinated hydrocarbons in dilute aqueous solution, J. Environ. Sci. Technol. 25 (1991) 1510 1512. [11] K. Inazu, Y. Nagata, Y. Maeda, Decomposition of chlorinated hydrocarbons in aqueous solutions by ultrasonic irradiation, Chemistry Letters, Jan. 1 (1993) 57 60. [12] A. Bhatnagar, H.M. Cheung, Sonochemical destruction of chlorinated C 1 and C 2 volatile organic compounds in dilute aqueous solution, J. Environ. Sci. Technol. 28 (1994) 1481 1486. [13] A. Kotronarou, G. Mills, M.R. Hoffmann, Decomposition of parathion in aqueous solution by ultrasound irradiation, J. Environ. Sci. Technol. 26 (1992) 1460-1462. [14] W.C. Koshkinen, K.E. Sellung, J.M. Baker, B.L. Barber, R.H. Dowdy, Ultrasonic decomposition of atrazine and alachlor in water, J. Environ. Sci. Technol. B29 (1994) 581-590. [ 15] M. Cost, G. Mills, P. Gilsson, J. Lakin, Sonochemical degradation ofp-nitrophenol in the presence of chemical components of natural waters, Chemosphere 27 (1993) 1737 1743. [16] I. Hua, R.H. Hochemer, M.R. Hoffmann, Sonolytic hydrolysis of p-nitrophenyl acetate: the role of supercritical water, J. Phys. Chem. 99 (1995) 2335-2342. Also: Disintegrate waste with ultrasound's tinny bubbles, Research & Development Magazine, January (1995) 55. [17] R. Hochemer, M.R. Hoffman, I. Hua, The sonochemical degradation ofp-nitrophenol in a parallel-plate near-field acoustical processor, Presented at the Second International Symposium on Advanced Oxidation Technologies, San Francisco, CA, Feb. 28 March 1 1996. [18] K.S. Suslick, Ultrasound: applications to materials chemistry, in: R.W. Cahn (Ed.), Encyclopedia of Materials Science and Engineering, Pergamon, Oxford, 3rd Suppl., 1992, p. 2093. [19] M.H. Entezari, P. Kruus, Effect of frequency on sonochemical reactions I: oxidation of iodine, Ultrasonics Sonochemistry 1 (1994) $75-$79. [20] M.H. Entezari, P. Kruus, Effect of frequency on sonochemical reactions lI: temperature and intensity effects, Ultrasonics Sonochemistry 3 (1996) 19 24. [21] T. Kimura, T. Sakamoto, J.-M. Leveque, H. Sohmiya, M. Fujita, S. Ikeda, T. Ando, Standardization of ultrasounic power for sonochemical reaction, Ultrasonics Sonochemistry 3 (1996) S157 S161.