J. Phys. Chem. 1987,91, 5847-5850 A
Figure 1. The three-atom system; introduction of yx, yv, and y.,
the total reactivity. It is noticed that the fit is reasonable, even at an energy value close to the resonance energy (approximately 0.977 eV). The results were calculated for yhO 63O, and the number of equations N needed to carry out this computation was 1000, where the number of basis functions was Nt = 9. As is noticed from eq 3, the final results may depend on the unspecified value of the angle yAo.Since the theory does not yield a value for this angle (in general, we have three angles, namely yAo, yd, and yA),but in the case of H H2one may assume them to be equal) the results should not depend on this angle. In Table I1 the probabilities as obtained for different values of y?oare compared, and indeed, a weak dependence was found. Still the
-
-
+
(13) 1922.
Walker, R. B.; Stechel, E. B.; Light, J. C. J. Chem. Phys. 1978,69,
5847
results change to some extent, and the question is which value should one take? Before answering this question, let us explain the physical meaning of the two extreme choices, i.e. yo = 0, ~ / 2 . (For reasons of convenience we dropped the index X.) Assuming yo= 0 leads to the decoupling of the various arrangement channel equations. Consequently, in order to obtain the correct answer, one has to include continuum states in the expansions of the Green functions. Since the main advantage of the BKLT approach is to avoid the continuum, this choice should be excluded. The other , each channel from itself extreme case, namely yo= ~ / 2 decouples which implies that the Green functions contain “most” of the information concerning the inelastic (nonreactive) process. In the present treatment of the H Hz system the Green functions contained only the elastic part and therefore the choice of yo = 7r/2 is probably not good enough. On the basis of what has been said so far, it is my belief that the better one solves the inelastic part of the problem, the weaker will be the dependence on yo. As for our particular case it is not obvious which is the best choice of yo. I chose an intermediate value for cos yowhich is close to 0.5, and the results in Table I were calculated for this value. I hope that in the future we will be able to say more about this problem.
+
Acknowledgment. I thank D. J. Kouri and D. G. Truhlar for sending me their papers prior to publication. I thank D. J. Kouri for valuable correspondence and Y . Shima for many valuable discussions. This work was supported by the United States-Israel Binational Science Foundation, Jerusalem. Registry No. H atomic, 12385-13-6; H2, 1333-74-0.
Thermal Decomposition of Methanol in the Sonolysis of Methanol-Water Mixtures. Spin-Trapping Evidence for Isotope Exchange Reactions C. Murali Krishna, Yves Lion,? Takashi Kondo,t and Peter Riesz* Radiation Oncology Branch, Clinical Oncology Program, Division of Cancer Treatment, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892 (Received: August 7, 1987)
The spin trap 3,5-dibromo-4-nitrosobenzenesulfonate was used to monitor the yield of free radicals produced during sonolysis of water-methanol mixtures. Methyl radicals and e H 2 0 H radicals were observed as well as the isotopically mixed radicals CH,D and CHD2 when CH30D:D20mixtures were studied. The results clearly show that thermal decompition of methanol to methyl radicals occurs in the gas phase. The methyl radical yield rises sharply at very low concentrations of methanol, reaches a maximum at 5 mol dm-’ in water and decreases to a smaller value in methanol. The yield of methyl radicals as a function of methanol concentration is discussed in terms of the different factors influencing the sonochemistry.
Introduction Numerous contributions to the study of the chemical effects of ultrasound in aqueous solutions have been made by Henglein, Hart, and their co-workers.’-1° The observed results have been explained in terms of reactions taking place in three different regions. One is the hot bubble created during cavitation (Le., creation, growth, and collapse of gas bubbles in solution by the acoustic field); in that region typical combustion reactions take place since the bubble implosion generates high local temperatures (several thousand degrees kelvin) and high pressures (hundreds of atmospheres). A typical gas reaction is the isotopic exchange between deuterium and water vapor occurring during the sonolysis of water in the presence of Permanent address: Institut de Physique, Universite de Liege, Sart Tilman 4000 Liege, Belgium. ‘Permanent address: Department of Experimental Radiology and Health Physics, Fukui Medical School, Matsuoka, Fukui 910-1 1 Japan.
The second region of interest is the interface between the hot bubbles and the bulk solvent. In this region, H and OH radicals can be scavenged by radical scavengers with an efficiency which is not predominantly, determined by the specific rate constants known from radiation chemistry but by other factors as well, (1) Henglein, A. Ultrasonics 1986, 25, 6. Rscher, Ch.-H.; Hart, E. J.; Henglein, A. J . Phys. Chem. 1986, 90,
(2) 222.
(3) Hart, E. J.; Henglein, A. J . Phys. Chem. 1986, 90, 5992. (4) Hart, E. J.; Henglein, A. J . Phys. Chem. 1985, 89, 4342. ( 5 ) Henglein, A. Z . Naturforsch. 1985, 406, 100. ( 6 ) Henglein, A.; Kormann, C. In?. J. Radiat. Biol. 1985, 48, 251. ( 7 ) Hart, E. J.; Fischer, Ch.-H.; Henglein, A. J . Phys. Chem. 1986, 90, 5989. (8) Fischer, Ch.-H.; Hart, E. J.; Henglein, A. J . Phys. Chem. 1986, 90, 3959. (9) Fischer, Ch.-H.; Hart, E. J.; Henglein, A. J . Phys. Chem. 1986, 90, 1954. (10) Hart, E. J.; Fischer, Ch.-H.; Henglein, A. J . Phys. Chem. 1987, 91, 4166.
This article not subject to U.S. copyright. Published 1987 by the American Chemical Society
5848 The Journal of Physical Chemistry, Vol. 91, No. 23, 1987
principally the hydrophobicity of the solute.6 Finally, in the third region, the bulk solvent, the radicals undergo the same scavenging reactions observed in radiolysis. H and OH radicals have been shown to be produced by ultrasound using spin-trapping experiments;" the stable spin adducts are formed by the reaction of the H and OH radicals diffusing into the bulk of the solution containing the spin trap. Studies of the sonochemistry of organic liquids have been reported by Suslick and his c o - ~ o r k e r s ~showing ~ - ~ ~ that organic liquids do support cavitation and the associated sonochemistry. Their results have been explained in terms of a hot spot mechanismZ5and show a vapor pressure dependence of the intensity of cavitational collapse: the lower the vapor pressure, the higher the rate of sonochemical reaction^.'^^'' We report the results of ESR and spin-trapping experiments during sonolysis of argon-saturated water-methanol mixtures. Methanol was chosen because it was expected that the high temperature reached in the collapsing gas bubbles .leads to the formation of methyl radicals whereas the H and OH radicals diffusing in the bulk of,the solution would lead to the formation of C H 2 0 H radicals. C H 3 and C H 2 0 H radicals can easily be discriminated by spin-trapping experiments using 3,s-dibromo4-nitrosobenzenesulfonate (DBNBS) as the spin trap.18 The sulfonate group ensures nonvolatility in the cavitation bubbles and several carbon-centered radical adducts show sufficiently detailed spectra to allow identification of the trapped radicals.'ssz6 Recently there has been a resurgence of interest in the use of ultrasound in organic s y n t h e s i ~ ; ' ~hence, . ' ~ it appeared to be useful to study sonochemical reactions of water-methanol mixtures.
Materials and Methods The spin trap 3,5-dibromo-4-nitrosobenzenesulfonic acid, sodium salt (DBNBS), was purchased from Sigma Chemical Co. C H 3 0 H , CH30D, CD30D, and D 2 0 were acquired from Aldrich Co. H202(30%) was obtained from Fischer Scientific Co. Sonolysis Experiments. A sample solution (4 mL) was placed in a closed system attached directly to the ESR quartz cell. This system was placed in the center of a sonication bath (Bransonic 1200; frequency 50 kHz; input power 80.5 W). Argon was bubbled for 15 min before and during sonication with a flow rate of 100 mL/min. The level of the liquid in the sample system was adjusted to the same level as that of the water in the sonicator and the temperature of the water in the sonication bath was maintained at 25 "C. The usual sonication time under these conditions was 15 min to achieve optimum spin adduct yield. Photolysis Experiments. H202(30%) a t a concentration of 50 pL/mL was added to the sample solutions containing the spin trap prior to photolysis. The photolysis experiments were carried out a t room temperature using a Schoeffel lOOOW Xe lamp coupled to a Schoeffel grating monochromator. The spin adducts were generated by UV irradiation at 275 f 10 nm of the argon-saturated solutions in an ESR quartz flat cell (60 X 10 X 0.25 mm) placed directly in the ESR cavity. The illumination time was usually between 1 and 2 min. ESR Measurements. The ESR spectra were recorded on a Varian E-9 X-band spectrometer operating at 100 lcHz modulation frequency and 9.5 GHz microwave frequency. Modulation amplitude was usually kept at 0.025 mT and the microwave power (11) Makino, K.; Mossoba, M. M.; Riesz, P. J . Phys. Chem. 1983, 87, 1369. (12) Mason, T. J. Ultrasonics 1986, 24, 245. (,13) Suslick, K. S. In Modern Synthetic Methods; Scheffold, R., Ed.; Springer-Verlag: New York, 1986; Vol. 4. (14) SusliEk, K. S.; Hammerton, D. A.; Cline, Jr, R. E. J. Am. Chem. Soc. 1986, 108, 5641. (15) Suslick, K. S.; Gawienowski, J. J.; Schubert, P. F.; Wang, H. H. J . Phys. Chem. 1983,87, 2299. (16) Suslick, K. S.; Schubert, P. F.; Goodale, J. W. Ultrason. Symp. Proc. 1981, 2, 612. (17) Suslick, K. S.; Gawienowski, J. J.; Schubert, P. F.; Wang, H. H. Ultrasonics 1984, 22, 33. (18) Kaur, H.; Leung, K. H. W.; Perkins, M. J. J . Chem. SOC.,Chem. Commun. 1982, 142.
Letters
I
Ultrasound CH,OH H20 5 95
Ullrapound CH,OD D,O 5 95
Figure 1. ESR spectra of spin-trapped radicals obtained by sonolysis of argon-saturated water-methanol mixtures ( 5 9 5 , v/v) in the presence of mol dm-3 DBNBS: (a) CH,OH:H,O; (b) C H 3 0 D : D 2 0 .
a t 20 mW. The observed ESR spectra were analyzed with the help of an isotropic ESR simulation program.
Results Photolysis of CH30H:H20(5:95, v/v) in the presence of H202 and DBNBS ( mol dm") resulted in a spin adduct exhibiting an ESR spectrum with a primary nitrogen splitting (aN = 1.37 mT) and a secondary triplet in the intensity ratio 1:2:1 (a2H= 0.92 mT). This is characteristic of a CHzOH adduct. A similar spin adduct has been observed previously.18 Sonolysis of argonsaturated C H 3 0 H : H 2 0 (5:95, v/v) at room temperature in the mol dm-3) yielded a spin adduct expresence of DBNBS ( hibiting a different ESR spectrum (Figure la) compared to that obtained by photolysis in the presence of HzOZ. The ESR spectrum consists of six groups of lines. This spectrum has been analyzed as due to a primary nitrogen triplet (aN = 1.45 mT). Each line of this triplet is further split into a 1:3:3:1 quartet by three equivalent protons (a3" = 1.35 mT). An additional splitting of 0.075 mT due to two equivalent meta hydrogens of the spin trap has been observed. Based on these ESR parameters, the spin adduct has been characterized as the CH3 spin addduct of DBNBS. This spin adduct has also been observed in the O H radical reactions of aqueous dimethyl sulfoxide by Kaur et a l l s Sonolysis of argon saturated CD30D:Dz0 (5:95, v/v) in the presence of DBNBS mol dm-3) resulted in a spin adduct exhibiting three groups of lines (Figure 2a) due to a primary nitrogen splitting of 1.45 mT. The M I = +1 component of the primary nitrogen triplet is shown in Figure 2b. This group consists of 21 lines which have been analyzed as a multiplet of lines in the intensity ratio 1:3:6:7:6:3:1 due to three equivalent deuterium nuclei ( I = 1). Each line of the multiplet is further split into a 1:2:1 triplet by two equivalent meta hydrogens of the spin trap (azH= 0.075 mT) resulting in the observed pattern. Based on the observed splitting constants, this spin adduct has been identified as a CD3 adduct of DBNBS. These experiments indicate the formation of CH3 radicals in the sonolysis of an argon-saturated C H 3 0 H : H 2 0 (5:95, v/v) mixture at room temperature. This is in contrast to the O H radical reactions with C H 3 0 H where only
The Journal of Physical Chemistry, Vol. 91, No. 23, I987
Letters @I
5
Ultrasound
5849
5r
CD,OD DeO 5 95
M,-+I
10
0
15
I
0
20
40
60
20
25
I
I
80
(MI
100 (%\
METHANOL
Figure 3. Effect of methanol concentration on the yields of CH, and CH20H radicals obtained by sonolysis of argon-saturated water-
methanol mixtures.
Figure 2. ESR spectrum of spin-trapped radicals obtained by sonolysis of CD30D:D20 mixture (5:95, v/v) in the presence of mol dm-' DBNBS: (a) complete spectrum; (b) expansion of M,= 1 group of (a).
the CH20Hradical is observed. The possibility of the formation of CH, from CH,OH indirectly through H and OH radicals produced by sonolysis of water can be n e g l e ~ t e d . ' ~ ~ ~ ~ Figure l b shows the ESR spectra recorded after sonication of argon-saturated CH30D:D20 (295, v/v) at room temperature in the presence of mol dm-, DBNBS. The ESR spectrum indicates the presence of five groups of lines in addition to the lines due to the CH, radical. These lines have been identified as due to the isotopically mixed methyl radical adducts of DBNBS, i.e., CH2D and CHD2. Stick diagrams for the ESR spectra of DBNBS adducts of e H 2 D and CHD2 have been constructed with the primary nitrogen splitting of 1.45 mT and the secondary exact hydrogen and deuterium splittings of 1.35 and 0.205 mT. match has been observed between the observed lines of CH2D, CHD2, and those expected from the stick diagram. The ratio of the splitting constants of the hydrogen and deuterium of the CH2D and e H D 2 spin adducts is in exact agreement with the ratio of the magnetic moments of the hydrogen and deuterium nuclei and this observation gives additional support to the identification of the isotopically mixed methyl radicals. The yields of the radicals CH3, CH,D, and CHD, were found to be in the ratio 80:18:
