Acoustic Bubble Sizes, Coalescence, and Sonochemical Activity in

Published on Web 07/01/2010 pubs.acs.org/ ... Acoustic bubble sizes, coalescence behavior, and sonochemical activity have been investigated in water i...
0 downloads 0 Views 820KB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

Acoustic Bubble Sizes, Coalescence, and Sonochemical Activity in Aqueous Electrolyte Solutions Saturated with Different Gases Adam Brotchie, Tom Statham, Meifang Zhou, Leena Dharmarathne, Franz Grieser,* and Muthupandian Ashokkumar* Particulate Fluids Processing Centre, School of Chemistry, University of Melbourne, VIC 3010 Australia Received April 28, 2010. Revised Manuscript Received June 14, 2010 Acoustic bubble sizes, coalescence behavior, and sonochemical activity have been investigated in water in the presence of various electrolyte additives (KCl, HCl, and NaNO3) and saturating gases-helium, air, and argon. A strong correlation was identified between the bubble radius and the dissolved gas concentration in the cavitation medium. The extent of bubble coalescence for each gas was also studied in different electrolyte solutions. A causal relationship between coalescence and bubble size was inferred. Importantly, the effects of the different electrolytes could be completely attributed to their “salting out” effect on the dissolved gas, providing valuable insight into the contentious issue of ion-specific coalescence inhibition. Extrapolation of the bubble size data to conditions where bubble coalescence is minimal, i.e., zero gas concentration and zero ultrasound exposure time, yielded a bubble radius of 1.5 ( 0.5 μm at an acoustic frequency of 515 kHz. In addition, the effects of electrolyte concentration and gas type on sonochemical activity were investigated. Sonochemical yields were increased by up to 1 order of magnitude at high electrolyte concentrations. This has been attributed to reduced gas and vapor content in the bubble core prior to collapse and a lower clustering density.

Introduction A direct consequence of ultrasound transmission through a liquid is the growth of adventitious bubble nuclei, which may collapse violently, generating localized “hot-spots” of extreme pressure and temperature. This phenomenon, termed acoustic cavitation, is responsible for the vast majority of applications involving ultrasound (e.g., in materials synthesis, environmental remediation, etc.) and also for the emission of broad wavelength light: sonoluminescence (SL).1-3 An understanding of how the various processes involved in acoustic cavitation are influenced by different parameters of the system is not only imperative for optimizing ultrasound reactors but is of far reaching fundamental importance to areas such as surface science, medicine and food technology. Many studies have examined the effect of certain ultrasound parameters, such as frequency, acoustic power and pulse modulation,4-11 and properties of the solution such as the nature of the solvent, presence of solutes and temperature.12-17 One of the most important properties of a *To whom correspondence should be addressed. E-mail: [email protected]; [email protected]. (1) Leighton, T. G. The Acoustic Bubble; Academic Press Inc.: London, 1994. (2) Suslick, K. S. In Wiley Encyclopedia of Electrical and Electronics Engineering; Webster, J., Ed.; John Wiley & Sons, Inc.: New York, 1999. (3) Suslick, K. S.; Price, G. J. Annu. Rev. Mater. Sci. 1999, 29, 295. (4) Okitsu, K.; Ashokkumar, M.; Grieser, F. J. Phys. Chem. B 2005, 109, 20673. (5) Sunartio, D.; Ashokkumar, M.; Grieser, F. J. Phys. Chem. B 2005, 109, 20044. (6) Kanthale, P.; Ashokkumar, M.; Grieser, F. J. Phys. Chem. C 2008, 111, 18461. (7) Brotchie, A.; Grieser, F.; Ashokkumar, M. Phys. Rev. Lett. 2009, 102, 084302. (8) Abe, S.; Choi, P. K. In 18th International Symposium on Nonlinear Acoustics; Nonlinear Acoustics Fundamentals and Applications; Stockholm, 2008; Vol. 1022, p 189. (9) Sostaric, J. Z.; Riesz, P. J. Phys. Chem. B 2002, 106, 12537. (10) Sponer, J. Czech. J. Phys. 1990, 40, 1123. (11) Tuziuti, T.; Yasui, K.; Lee, J.; Kozuka, T.; Towata, A.; Iida, Y. J. Phys. Chem. A 2008, 111, 12093. (12) Grieser, F.; Ashokkumar, M. Adv. Colloid Interface Sci. 2001, 89, 423. (13) Ashokkumar, M.; Guan, J.; Tronson, R.; Matula, T. J.; Nuske, J. W.; Grieser, F. Phys. Rev. E 2002, 65, 463101.

12690 DOI: 10.1021/la1017104

cavitation system is the nature and concentration of the dissolved gas. For instance, the high polytroptic index of noble gases results in much higher chemical yields and light emission intensities compared with diatomic gases generally attributed to higher temperatures of the former generated during compression.1 Nonadiabatic processes, such as heat transfer out of the bubble, serve to lower the collapse temperature. For noble gases, the bubble collapse temperature is generally considered to be a function of the gas thermal conductivity, i.e., increasing in the order: He< Ne< Ar99.5%) were obtained from Sigma Aldrich. Ammonium molybdate (>99%) and sodium hydroxide (>99%) were purchased from BDH. Potassium hydrogen phthalate (>99.8%) was procured from Ajax Finechem. Sodium nitrate (>97%) and potassium iodide (>99%) were obtained from Chem. Supply. Hydrochloric acid was supplied by Scharlau. High purity grade argon and helium were purchased from BOC gases. Bubble Size Determinations. Milli-Q water was left to stand overnight under atmospheric conditions to equilibrate and where helium and argon saturated solutions were required, solutions were sparged with the respective gas for at least 20 min. A blanket of gas was maintained above the liquid surface for the duration of the experiment in order to prevent redissolution of atmospheric gases. The ultrasound arrangement consisted of an Undatim Ultrasonics UL03/1 D-reactor with an in-house customized pulse modulator and a 515 kHz ceramic transducer (35 mm diameter plate) attached to a stainless steel plate. A cylindrical glass reaction vessel (diameter: 45 mm, height: 90 mm) was attached to the transducer and filled to a volume of 100 mL for all measurements. Unless otherwise specified, the pulse width and calorimetric power absorbed (measured under continuous wave sonication) were fixed at 5.8 ms and 10 W (0.1 W cm -3 ), respectively. SL measurements were taken by measuring the output of a Hamamatsu photomultiplier tube on a LeCroy digital oscilloscope. (22) Burdin, F.; Tsochatzidis, N. A.; Guiraud, P.; Wilhelm, A. M.; Delmas, H. Ultrason. Sonochem. 1999, 6, 43. (23) Mettin, R.; Luther, S.; Lauterborn, W. In Proc. 2nd Conf. on Applications of Power Ultrasound in Physical and Chemical Processing, Toulouse, France, 1999; p 125. (24) Tsochatzidis, N. A.; Guiraud, P.; Wilhelm, A. M.; Delmas, H. Chem. Eng. Sci. 2001, 56, 1831. (25) Luther, S.; Mettin, R.; Koch, P.; Lauterborn, W. Ultrason. Sonochem. 2001, 8, 159. (26) Labouret, S.; Frohly, J. Eur. Phys. J. Appl. Phys. 2002, 19, 39. (27) Chen, W. S.; Matula, T. J.; Crum, L. A. Ultrasound Med. Biol. 2002, 28, 793. (28) Lee, J.; Yasui, K.; Tuziuti, T.; Kozuka, T.; Towata, A.; Iida, Y. J. Phys. Chem. B 2008, 112, 15333.

Langmuir 2010, 26(15), 12690–12695

For each sample the first SL measurement was taken at a pulse separation of 20 ms. Subsequent measurements were made in increments of 25 ms until a certain separation was reached (after which the signal began to decay in intensity). At this point further measurements were taken at smaller time intervals until a point was reached where the signal fell abruptly to zero. The pulse separations at which this occurred were measured at least 5 times, and the median of the range used in the subsequent calculations. The values reported were recorded after the signal, at each particular pulse separation, had reached steady-state intensity (in cases, up to several minutes) to ensure that equilibrium conditions were achieved. The method used to correlate the ultrasound pulse separation with the cavitation bubble size has been described in detail elsewhere.14 Briefly, cavitation bubbles will experience growth, via rectified diffusion and coalescence during the ultrasound pulse, and will dissolve (or rise through buoyancy if sufficiently large) during the pulse off-time. Under conditions such as those used in the present study, it takes many pulses (ca. 50) to establish a steadystate level of SL emitting bubbles. At steady-state there exists equilibrium between the competing processes of bubble growth (during the pulse) and dissolution (between pulses). As the pulse separation increases, a greater fraction of the bubble population will dissolve, leading to a decrease in the bubble number population and consequently, a decrease in SL emission. If the time between pulses is sufficiently long, such that dissolution of the entire population occurs, no cavitation (or SL) is observed. The time of bubble dissolution (equal to the pulse separation) can be correlated with the bubble radius through the EpsteinPlesset equation ! ! DCs 1 RTFg R0 þ1 ð1Þ t ¼ 3 2Mγ Fg R0 2 where D is the solute diffusion coefficient (m2 s-1),29,30 Cs is the saturated dissolved gas concentration (g L-1), Fg is the gas density of the bubble (g L-1),31 R0 is the initial bubble radius (m), t is the dissolution time (s), R is the Universal gas constant (J kmol-1 K-1), T is the absolute temperature of the liquid (K), M is the molecular weight of dissolved gas (g mol-1), and γ is the surface tension (N m-1).31 The solubilities of helium, nitrogen, oxygen, and argon in pure water were taken from the literature.32 In order to calculate the extent of “salting out” of the gas in the various electrolyte solutions used, the following relationships were employed:   R0 ¼ KscR C ¼ hI ð2Þ log R h ¼ h - þ h þ þ hg

ð3Þ

where R0 and R are the Bunsen solubility coefficients in pure water and in the presence of electrolyte, Ksca is the Sechenov constant (m3 kmol-1), C is the electrolyteP concentration (M), I is the ionic strength of the solution [(1/2) cini2], h-, hþ, and hg are the empirical constants for the anion, cation, and gas.33,34 This approach permits an accurate estimation of values for a large range of anion-cation combinations and gases in both single and mixed electrolyte systems. Bubble Coalescence Measurements. Bubble coalescence was quantified following a method developed by Lee et al.35 A customized near-cylindrical 770 mL reaction cell fitted with a thin (29) Wise, D. L.; Houghton, G. Chem. Eng. Sci. 1966, 21, 999. (30) Himmelblau, D. M. Chem. Rev. 1964, 64, 527. (31) CRC Handbook of Chemistry and Physics, 56th ed.; CRC Press: Boca Raton, FL, 1975-1976. (32) http://www.engineeringtoolbox.com/. (33) Lang, W.; Zander, R. Ind. Eng. Chem. Fundamen. 1986, 25, 775. (34) Hermann, C.; Dewes, I.; Schumpe, A. Chem. Eng. Sci. 1995, 50, 1673. (35) Lee, J.; Kentish, S. E.; Ashokkumar, M. J. Phys. Chem. B 2005, 109, 5095.

DOI: 10.1021/la1017104

12691

Article

Brotchie et al. Table 1. Critical Sonoluminescence Decay Times (at which the intensity reaches zero) and the Corresponding Bubble Radii (μm) for Helium, Air, and Argon Saturated Aqueous Solutions in the Absence and Presence of Different Electrolytes SL decay times, ms (bubble radius, μm) Air

Ar

He

concentration (M)

NaCl

KCl

NaNO3

NaCl

NaCl

0 0.001 0.1 1 2 4

505 (4.3) 678 (4.7) 580 (4.4) 645 (4.0) 298 (2.7) 181 (1.8)

505 (4.3) 744 (4.9) 706 (4.7) 466 (3.7) 260 (2.6) 134 (1.6)

505 (4.3) 440 (4.0) 531 (4.3) 381 (3.4) 360 (3.0) 235 (2.1)

878 (5.8) 831 (5.7) 803 (5.5) 645 (4.7) 489 (3.8) 289 (2.5)

510 (4.1) 470 (4.0) 418 (3.8) 513 (3.7) 303 (2.8) 283 (2.2)

Bubble Size. In order to determine the bubble size using the technique described earlier, a pulsed ultrasound field must be applied with the pulse width set at an arbitrary value. In previous studies the effect of this parameter was neglected. It can be seen in Figure 1, which contains size data as a function of pulse width in water and 1 M KCl solution, however, that the size is highly dependent on the pulse width. For a given pulse width, the size is larger in water and significantly lower in the electrolyte solution; both data sets converge to between 1.2 and 1.8 μm at zero pulse width. One may question whether these trends pertain to the assumption made in our calculations that bubbles which dissolve between pulses becoming inactive (i.e., not emitting SL) dissolve completely. It is likely that, rather than dissolving completely, bubbles only need to dissolve below a critical size from which, they cannot be grown to the Blake threshold during the period leading to the

next ultrasound pulse. A longer pulse width will effectively lower this critical size. We can dismiss the possibility that this influences the experimental data. Bubbles will dissolve completely from the Blake threshold in less than 10 ms. As the typical pulse separation required in this experiment is typically about 600 ms in air, it is improbable that a decrease in critical minimum size could have a measurable influence. Before further analysis is undertaken, it is expedient to define the parameter bubble size. In an acoustic field, the Blake threshold defines the boundary between smooth, stable radial oscillations and the explosive growth and collapse required to generate SL. We calculate that in this system, the Blake threshold is approximately 0.8 μm.1 A bubble can also be defined by its linear resonance size, which at 515 kHz is 5.8 μm. It is well established, however, that bubbles rarely reach their linear resonance size.38 It is clear from Figure 1 that the sizes measured in this system are mostly intermediate between these two theoretical sizes. In a bubble cluster, interior positioned bubbles are acoustically shielded from the sound field; these bubbles experience an attenuated acoustic pressure and consequently will have a larger Blake threshold than the calculated 0.8 μm. Thus, in a complex multibubble field, a range of threshold sizes can be expected reflecting the different bubble environments. The bubble size that is measured in our system is a composite of all bubbles sizes within a cluster. From the data however, it can be deduced that small bubbles, which are close to the theoretical Blake threshold do not significantly contribute to the measured bubble size; rather it is a much larger sized population that dominates. For an arbitrarily selected intermediate pulse width of 5.8 ms, the effects of a range of different electrolytes and gas type on the critical pulse separation time at which no SL emission was produced, and the corresponding cavitation bubble size are presented in Table 1. In our previous report,7 it was shown that bubbles which emit secondary chemical luminescence exhibit a broad size distribution corresponding to a slow decay of the emission intensity with increasing pulse separation. SL emitting bubbles, however, were shown to have an extremely narrow size distribution and a sharp cut off in the emission-pulse separation curve. It is for this reason that only single values for the critical decay time and mean bubble size are reported in the present study. There are two striking features of the data contained in Table 1. The first is that the gas type has a profound effect on the bubble size, with the size generally increasing in the order: He < Air < Ar. Second, the size decreases with the addition of all electrolytes used. A physicochemical effect of changing the gas type or adding a large amount of electrolyte is that total gas solubility in water changes. Upon examination of the size data as a function of gas

(36) Hochanadel, C. J. J. Phys. Chem. 1952, 56, 587. (37) Alegria, A. E.; Lion, Y.; Kondo, T.; Riesz, P. J. Phys. Chem. 1989, 93, 4908.

(38) Yasui, K.; Tuziuti, T.; Lee, J.; Kozuka, T.; Towata, A.; Iida, Y. J. Chem. Phys. 2008, 128, 184705.

Figure 1. Effect of ultrasound pulse width on the bubble size in air saturated water and KCl solutions. capillary tube (internal diameter of 0.9 mm) was attached to a 55 mm diameter 355 kHz transducer plate (Allied Signal). An ELAC RF-generator/power amplifier was used to supply the electrical waveform. The solution height was adjusted with a syringe such that prior to sonication the fluid level was close to the base of the capillary. The solution was sonicated at 20 W (determined calorimetrically) for a nominal period of 30 s, sufficient sonication time for the fluid level in the capillary to rise an appreciable distance (ca. 10 cm in pure water), from which the total volume change was calculated. Hydrogen Peroxide Measurements. The yield of hydrogen peroxide formed during sonication was determined spectrophotometrically using a Varian (Cary) UV-visible spectrophotometer according to an analytical technique developed by Hochanadel.36 Sonication was performed using the same ultrasound equipment described above for the coalescence experiments, in a cylindrical glass cell (diameter: 65 mm, height: 120 mm) filled with 200 mL solution at 30 W calorimetric power. Freshly sonicated samples (2.5 mL) were withdrawn from the bulk solution and mixed with 1.25 mL of solution A (0.4 M KI, 0.1 M NaOH, 0.2 mM (NH4)6Mo7O24.4H2O) and solution B (0.1 M KHC8H4O4). After mixing, solutions were left to stand for 5 min prior to spectroscopic analysis. Fresh reagent solutions were prepared for each experiment and each experiment was repeated three times. This technique is based on the oxidation of iodide by hydrogen peroxide to I3- (λmax = 353 nm). A molar extinction coefficient of tri-iodide of 38 400 M-1 cm-1 was used.37

Results and Discussion

12692 DOI: 10.1021/la1017104

Langmuir 2010, 26(15), 12690–12695

Brotchie et al.

Article

Figure 2. Bubble radius as a function of dissolved gas concentration for helium, air and argon saturated aqueous solutions of electrolytes (NaCl, KCl and NaNO3). The bubble radii data used to construct this plot were taken from Table 1. The error bar shown is typical of all data points. The data points corresponding to pure water are indicated by dashed circles.

solubility, as shown in Figure 2, quite a strong correlation between bubble size and gas solubility appears. The experimental reproducibility of the results is estimated to be about 15%. [If the size data are normalized with respect to the size measured in pure water for each gas, and presented as a function of the normalized gas solubility (see the Supporting Information), the difference between data sets is eliminated. This suggests that there was a small systematic error associated with the collection of the data sets. This systematic error is most likely due to slight variations in solution temperature and ultrasound power variations from day to day.] The question that arises is how does gas concentration influence the bubble size? Bubbles can grow in the sound field through two mechanisms: coalescence and rectified gas diffusion. Changes in gas concentration and ultrasound exposure time can be expected to affect both processes. The extent of bubble growth through rectified diffusion will be slowed as the gas concentration is reduced as the quantity of gas diffusing across the interface is lower. The contribution of rectified diffusion to trends in bubble size, however, is likely to be small. This can be deduced from a consideration of the behavior of acoustic bubbles in the presence of surfactant. It is known for single bubble experiments that surfactant in the system accelerates rectified diffusion.15,39 However, in multibubble fields, rather than resulting in a larger bubble size, the presence of surfactant drastically reduces the bubble size14 owing to its inhibitory effect on coalescence.35 Alternatively, it can be proposed that the dissolved gas concentration, which can be controlled by the gas type and the addition of electrolytes, determines the extent to which bubbles coalesce and therefore the bubble size generated. The existence of this coalescence condition can also be used to account for the results of Figure 1. The longer the ultrasound exposure time (pulse width) the higher will be the number of bubble collisions and coalescence events, thereby leading to a larger bubble size. The smaller bubble sizes in the electrolyte solutions are consistent with the results of Figure 2. Direct investigation into bubble-bubble coalescence will be the focus of the following section. Bubble-bubble Coalescence. Under similar conditions to the bubble size experiments, the relative extent of coalescence was (39) Crum, L. A. J. Acoust. Soc. Am. 1979, 66, S45.

Langmuir 2010, 26(15), 12690–12695

Figure 3. (a) Change in volume (ΔVT) in a capillary tube after 30 s sonication as a function of dissolved gas concentration for helium, air, and argon saturated solutions of KCl, HCl and NaCl, and NaNO3. The error bar shown is typical of all the data points shown. (b) Bubble radius (interpolated from the best fit linear regression of the data in Figure 2) shown as a function of the volume change. The data points corresponding to pure water are indicated by dashed circles.

quantified in helium, air and argon saturated aqueous solutions of different electrolytes: KCl, HCl and NaNO3. The absolute volume change (ΔVT) is presented in Figure 3a as a function of the dissolved gas concentration for each of the gases. Each set of data combines measurements made for all three electrolytes (0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 M). It is quite apparent that although the same trend is observed for each gas, the three data sets are quite distinct, more so than was observed for the bubble size data (Figure 1). [As discussed for the bubble size data, the differences between data sets may result from systematic error. Such errors can be eliminated by normalizing the volume data for each gas, in which case the data for all three gases map onto a single curve (see Figure S2 in the Supporting Information). From this representation of the data it is clear that there exists a linear dependence of the extent of coalescence upon the gas concentration.] Thus, a certain relative change in gas solubility brings about a fixed change in both the coalescence and the bubble size, irrespective of the gas used. This result provides very strong support for our argument that bubble coalescence is the main determinant of bubble size in an acoustic field. This conclusion is reinforced if the DOI: 10.1021/la1017104

12693

Article

bubble size is represented as a function of the volume change, shown in Figure 3b, in which case the data for all three gases map onto the same curve. The exact mechanism(s) through which gas concentration affects coalescence is not clear although it can be assumed that the rate of bubble nucleation will be lower in solutions containing less gas. An important outcome of this work relates to the contentious issue of ion-specific coalescence inhibition. In recent years, a number of studies have reported the ion-specific nature of electrolyte inhibition of bubble coalescence, albeit in static (nonacoustic) fields.40-46 Various explanations have been proposed, most implying a gas-liquid interfacial mechanism. Christenson and Yaminsky41 have reported a correlation between the inverse Marangoni factor, (dγ/dc)-2, and coalescence inhibition ability for several different electrolyte species. Some solutes which increase surface tension, however, are also observed to inhibit coalescence. Marcelja44 has argued that the tendency of certain ions to concentrate and others to avoid the bubble surface results in ion separation resulting in electric double-layer repulsions. Craig et al.40 proposed a mechanism based on the reduction in the range of the hydrophobic attraction between bubbles through their influence on the water structure. Henri et al.47 have provided evidence in mixed electrolyte systems that surface tension gradients and Gibbs elasticity do not correlate well with coalescence inhibition and instead, propose a mechanism based on ion separation in the interfacial region leading to electric double layer repulsion, and reduction in the magnitude of the hydrophobic attraction. Weissenborn and Pugh42 calculated the film rupture thickness to be much greater than the distance over which shortrange van der Waals and electrostatic forces are significant leaving the highly contentious long-range hydrophobic force as the only remaining factor in explaining their observed electrolyte coalescence inhibition. They have proposed that the hydrophobic force is affected by the reduction of dissolved gas, which decreases the microscopic bubble concentration therefore weakening a bridging attraction between macroscopic bubbles. In the present work, the concentrations required to induce a drop in the coalescence rate far exceed those reported in the literature. For example, Craig et al.48 recorded a transition concentration for NaCl of the order 10-2 M, with 100% coalescence inhibition at 10-1 M. In fact, we observe no change in coalescence below about 0.5 M and as can be seen in Figure 4, at 1 M there is only a 10-20% decrease in all salts except HCl. The fact that the HCl has a less marked effect on coalescence in our system can be completely attributed to its relatively lower “salting out” effect compared with the other salts (the Sechenov constants are 0.036 and 0.143 m3 kmol-1 for HCl and NaCl, respectively). In addition, salts which have been identified as noninhibitors (NaNO3 and NaClO4) and those that have been identified as inhibitors (NaCl and KCl) in quasi-static systems, can be seen to have an almost identical effect on coalescence in the acoustic system. Thus, differences in coalescence behavior of the different salts which have been attributed to specific ion effects are not (40) Craig, V. S. J.; Ninham, B. W.; Pashley, R. M. Nature 1993, 364, 317. (41) Christenson, H. K.; Yaminsky, V. V. J. Phys. Chem. 1995, 99, 10420. (42) Weissenborn, P. K.; Pugh, R. J. J. Colloid Interface Sci. 1996, 184, 550. (43) Deschenes, L. A.; Barret, J.; Muller, L. J.; Fourkas, J. T.; Mohanty, U. J. Phys. Chem. B 1998, 102, 5115. (44) Marcelja, S. J. Phys. Chem. B 2006, 110, 13062. (45) Henri, C. L.; Parkinson, L.; Ralston, J. R.; Craig, V. S. J. J. Phys. Chem. C 2008, 112, 15094. (46) Christenson, H. K.; Bowen, R. E.; Carlton, J. A.; Denne, J. R. M.; Lu, Y. J. Phys. Chem. C 2008, 112, 794. (47) Henri, C. L.; Dalton, C. N.; Scruton, L.; Craig, V. S. J. J. Phys. Chem. C 2007, 2007, 1015. (48) Craig, V. S. J.; Ninham, B. W.; Pashley, R. M. J. Phys. Chem. 1993, 97, 10192.

12694 DOI: 10.1021/la1017104

Brotchie et al.

Figure 4. Normalized volume change in argon saturated aqueous solutions of different electrolytes.

reproduced for acoustic bubbles and to reiterate, the coalescence inhibition observed is solely due to a reduction is gas concentration. It should also be noted that recent direct force AFM measurements support this observation, revealing no difference in the coalescence between two air-filled gas bubbles in the presence of salts which are considered to be “inhibiting” and “noninhibiting” for quasi-static systems.49 The possibility cannot be excluded however, that the differences in coalescence behavior with added salts observed between the acoustic and nonacoustic systems may relate to the different time domains over which measurements are recorded. Acoustic bubbles are formed with a fresh interface and exist for less than 1 ms,50 whereas bubbles investigated in nonacoustic systems survive for several seconds. If selective ion adsorption is indeed responsible for effects observed in the quasistatic systems, it is possible that due to the limited time frame of the acoustic bubbles this may preclude this mechanism in the latter system. Sonochemical Activity. The influence of two electrolytes, NaNO3 and NaClO4,51 on sonochemical activity, quantified through the yield of hydrogen peroxide, was investigated in argon and helium saturated solutions. The absolute product yields are presented in Figure 5a as a function of gas concentration and the normalized yields for both gases shown in Figure 5b as a function of the normalized gas concentration. The substantially higher absolute yields in argon may be expected owing to its lower heat conductivity compared with helium, which is widely believed to result in a hotter collapse temperature.18,19 It should be noted, however, that recent studies into the effect of gas type on the bubble temperature challenge this supposition. Yasui et al.20 and Okitsu et al.21 have reported that the temperature is not sensitive to the thermal conductivity of the gas; rather it is determined by the water vapor trapped in the bubble. The dependence of the yield upon the gas concentration requires further explanation. Wall et al.,52 investigated multibubble SL from air saturated aqueous solutions of (49) Vakarelski, I. U.; Manica, R.; Tang, X.; O’Shea, S. J.; Stevens, G. W.; Grieser, F.; Dagastine, R. R.; Chan, D. Y. C. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 11177. (50) Sunartio, D.; Ashokkumar, M.; Grieser, F. J. Am. Chem. Soc. 2007, 129, 6031. (51) The use of salts comprising a chloride anion (i.e., HCl, NaCl and KCl) was found not be possible for analysis of the sonochemical hydrogen peroxide formation due to what is assumed to be side reactions mediated by chloride radicals. (52) Wall, M.; Ashokkumar, M.; Tronson, R.; Grieser, F. Ultrason. Sonochem. 1999, 6, 7.

Langmuir 2010, 26(15), 12690–12695

Brotchie et al.

Article

Plesset equation1 under the conditions of the experiment and found that although reducing the equilibrium bubble size in argon from 7 to 4 μm in argon would lead to a more severe collapse (indicated by the radius ratio: Rmax/Rmin), the change in size for helium from about 2 to 1 μm would not have such an effect. Thus, it is unlikely that the influence of the initial bubble size on radial dynamics can account for the observed trends. The sharp reduction in coalescence in the electrolyte solutions is indicative of less dense bubble clusters. Dense clustering is known to acoustically shield bubbles positioned in the interior of the cluster as well as scattering and attenuating the sound wave.53,54 Tuziuti et al.55 have shown that the dissolved gas concentration strongly affects the spatial distribution of cavitation. They reported that at high gas concentration, the sonochemical activity was marginalized to the sides and liquid surface of the reaction cell. Further, bubbles are strongly influenced by interaction with neighboring bubbles, with the expansion during rarefaction being suppressed.54 We therefore propose that in addition to a reduced amount of gas and vapor entering the bubble prior to collapse, elevating the collapse temperature, the reduction in bubble density at low gas concentrations, leads to a more symmetric and efficient collapse process, which serves to further increase the core temperature. The discrepancy between the relative increases observed in argon and helium (the sonochemical yield is increased 10-fold in helium and only 3-fold in argon at 3 M of the two salts) remains unclear.

Summary and Conclusions

Figure 5. (a) Concentration of hydrogen peroxide formed after 30 min sonication as a function of dissolved gas concentration for helium and argon saturated aqueous NaNO3 and NaClO4 solutions. (b) Hydrogen peroxide concentration normalized with respect to the pure water value for both gases as a function of the normalized dissolved gas concentration.

a range of electrolytes. They reported up to 3-fold enhancements in SL intensity at high salt concentration and identified that the only solution parameter that correlated with the SL trends was the gas solubility. It was suggested that the severity of bubble collapse could be increased at high electrolyte concentration due to a lower amount of gas present in the collapsing bubble. It is thought that neither the equilibrium bubble size nor the Blake threshold is significantly affected in the presence of salt. Thus, if it is assumed that there is nonequilibrium gas diffusion during the explosive growth that precedes the transient collapse that produces SL (and sonochemistry), the effect of salt can be expected to reduce the total amount of gas (and vapor) present in the bubble prior to collapse. A lower initial internal pressure within the collapsing bubble will serve to increase the collapse temperature and therefore increase SL emission and sonochemical yields.1 It is possible that such a nonequilibrium gas diffusion effect serves to intensify collapse conditions however, in light of the results presented in the previous sections, we can extend upon this interpretation. It was demonstrated in the present study that the bubble size is reduced by a decrease in gas concentration. It is reasonable to assume that a smaller size distribution of bubbles will influence the severity of collapse. We solved the RayleighLangmuir 2010, 26(15), 12690–12695

The relationship between dissolved gas concentration, interbubble coalescence and the size of acoustically active bubbles in an ultrasound field has been investigated for the first time. A strong correlation was found between the dissolved gas concentration of argon, air and helium in different electrolyte solutions, and both the extent of bubble coalescence and the bubble size. It was deduced that the bubble number and density are functions of the gas concentration and that the latter parameter controls the rate of bubble coalescence, which in turn determines the bubble size. Importantly, no evidence of electrolyte induced coalescence inhibition was observed over the typical ranges reported in the literature in quasi-static systems and the reduction in the extent of coalescence observed at elevated electrolyte concentrations was found to depend only on the “salting out” qualities of the electrolyte. In light of the data from the quasi-static systems showing ion-specific inhibition and from recent AFM force curve measurements for individual bubbles, which show no such ionspecific effects, the issue of electrolyte coalescence inhibition remains an enigma. The present work contributes an important insight into this poorly understood area. Acknowledgment. The financial support from the Australian Research Council is gratefully acknowledged. A.B. also acknowledges the receipt of an Australian Postgraduate Award and a David Lachlan Hay Postgraduate Writing-up Award from The University of Melbourne. Supporting Information Available: Normalized bubble size and coalescence data and bubble dissolution data. This material is available free of charge via the Internet at http:// pubs.acs.org. (53) Hatanaka, S.; Yasui, K.; Kozuka, T.; Tuziuti, T.; Mitome, H. Ultrasonics 2002, 40, 655. (54) Yasui, K.; Iida, Y.; Tuziuti, T.; Kozuka, T.; Towata, A. Phys. Rev. E 2008, 77, 016609. (55) Tuziuti, T.; Yasui, K.; Sivakumar, M.; Iida, Y. Ultrasonics 2006, 44, 357.

DOI: 10.1021/la1017104

12695