Photolysis of
Carbonatoamine Complexes of Cobalt(1I I )
+
+
with k , = k z k3 k4 and values given by Berlman34 for k, and kz, the results presented in Table IV are obtained. Comparison of results for 295°K in Tables III and IV shows that 122 and k4 increase and k 3 decreases with increase in the degree of benzene methylation. Such behavior of k z has been attributed to reduction of the symmetry and, thereby, removal of symmetry restrictions on the radiative t r a n ~ i t i o n .The ~ ~ increase in k4 may result from an increase in spin-orbit coupling with replacement of a hydrogen atom by a methyl groupz4 The decrease in k 3 is consistent with the finding by Sandrosl5 that k3 of toluene has a larger activation energy than k 3 of benzene. The information necessary for calculation of specific rates from values of x is not available for toluene and p-xylene at 195°K.
1111
The small yield of photoproduct observed in the present work (cf. Results) increases with increase in the concentration of trans-2-octene and, therefore, appears to be a product of photoaddition of the olefin to the aromatic. It has been suggested that the activated internal conversion represented by reaction 3 is associated with formation of benzene isomers (benzvalene and fulvene) and photoaddition of olefins to benzene.35 In accord with such a suggestion, the photoproduct yield increases with decrease in benzene concentration and changes in the same direction as k 3 / k , (cf. Tables I11 and IV) with change in temperature and in the degree of benzene methylation. (34) I. B. Berlman, "Handbook of Fluorescence Spectra of Aromatic Molecules," 2nd ed, Academic Press, New York, N. Y., 1971, p 69. (35) D. Bryce-Smith, Pure Appl. Chem., 16, 62 (1968).
Behavior of COS- Radicals Generated in the Flash Photolysis of Carbonatoamine Complexes of Cobalt(1 I I) in Aqueous Solution' Schoen-nan Chen, Virgil W . Cope,2 and Morton 2 . Hoffman* Department of Chemistry, Boston University, Boston, Massachusetts 02275 (Received October 27, 1972)
The flash photolysis of C O ( N H ~ ) ~ C O in ~aqueous + solution generates the C 0 3 - radical which is characterized by its absorption spectrum with A,, 600 nm and its second-order decay kinetics. The radical is also generated from Co(NH3)&03+ and, to a lesser extent, from Co(en)zC03+. By monitoring the decay of the C03- absorption as a function of the concentration of added scavenger solutes, values of k(CO3+ S) have been obtained for some organic and inorganic solutes. In particular, k(C03- indole-3-propionic acid) was determined as a function of ionic strength and pH; at pH 7, k is independent of ionic strength while a t pH 12 there is a positive dependence of log k us. &*. From the data it is concluded that a t pH 7 the radical must exist in its acidic form, C03H, with a pK, = 9.6 f 0.3. The dependence of k on pH at constant ionic strength for N-acetyltryptophan and P,P'-dithiodipropionic acid supports the model presented as does the lack of a pH dependence for indole.
+
Introduction The C03- radical can be generated radiolytically in aqueous solution through the interaction of OH radicals and ~ a r b o n a t e . ~However, -~ although the reaction of OH radicals with c0a2-to form COS- is only moderately fast ( k = 4.2 x 108 M-' sec-I), the corresponding reaction with C03H- is considerably slower (k = 1.5 X lo7 M-l sec-1).6 As a result, the use of the pulse radiolytic technique to observe and characterize the C03- radical has been limited to studies in rather strongly alkaline solution. The spectrum of C03-, with its absorption maximum a t 600 nm, has been recorded at pH values a t which the C032- concentration can be maintained near 0.1 M; the pK for the equilibrium C03H- F H+ C032- is 10.36. The radical spectrum has also been generated in the direct flash photolysis7 of c o s 2 - at p H 12.8 and in the flash photolysis of Hz02 solutionsa containing C 0 3 z - , In the latter case, the spectra obtained (from the photolytic decomposition of H202 to form OH radicals which then attack Co+) in the pH range 8-13 had identical
+
The observation of a lack of a spectral change for C 0 3 - as a function of pH was taken as evidence that the radical did not undergo acid-base equilibrium in that region; the acid-base forms of many radicals exhibit different absorption spectra.9 However, it must be pointed out that a change in the absorption spectrum is not a requirement for radical acid-base equilibria. For example, the A,.,
(1) This research was supported by the National Science Foundation through Grant No. GP11213 and the College Teachers Research Program. (2) Participant, NSF College Teachers Research Program, Summer 1971. Permanent address: Department of Chemistry, University of Michigan, Flint College, Flint, Mich. 48503. (3) G. E. Adams and J. W. Boag, Proc. Chem. SOC., London, 112 I1964). ( 4 ) G. E. Adams, J. W. Boag, and 6.D. Michael, Trans. Faraday SOC., 61, 1674 (1965). ( 5 ) G. E. Adams. J. W. Boag, and E. D. Michael. Proc. Roy. Soc., Ser. A. 287,321 (1965). (6) J. L. Weeks and J. Rabani, J. Phys. Chem., 70, 2100 (1966). (7) E. Hayonand J. J. McGarvey, J. Phys. Chem., 71, 1472 (1967). (8) D. Behar, G. Czapski, and I. Duchovny, J. Phys. Chem., 74, 2206 (1970). (9) P. Neta, M. Simic. and E. Hayon, J. Phys. Chem., 73, 4207 (1969).
The Journal of Physical Chemistry, Voi. 77, No. 9. 7973
1112
S.Chen, V. W. Cope, and M. Z. Hoffman
spectra of the COZH-COZ- and C204H-Cz04- lo pairs are virtually identical. The second-order rate constant (2k1) for the decay of C03-
sured 50 psec after the start of the flash as a function of monitoring wavelength. Rate constants were calculated from a least-squares computer fit of the data to the appropriate rate law. At p H ll),solutions were prepared diproduce useful values for 2k1 inasmuch as the C03HZOZreaction is kinetically competitive with reaction 1, rectly from the salt in order to minimize base-catalyzed The Weeks and Rabani pulse radiolytic study6 showed hydrolysis.21 Alternatively, in neutral or mildly alkaline that in the p H range 10.5-13.5, 2kl had a constant value solution, a fresh stock solution of 0.01 M C O ( N H ~ ) ~ C O ~ + that was affected only by ionic strength; a plot of log 2k1 was used to prepared the solutions for flashing. The solutions for flashing were prepared in less than 1 min and the us. p112/(l + O.4p1I2) was linear with a fitted slope of unity. An experiment at pH 8.4 in 1 m M HC03- is re- stock solution was discarded within 2 hr of preparation. For experiments involving Co(NH3)&03+, the flash soluported but the result falls significantly below the line. tion was prepared directly from the solid and was never at From these ionic strength studies, a value of 2kl = 1.25 x pH lo3 M-1 cm-l) ligand-to-metal charge transfer band can be generated cleanly from a strongly absorbing subgenerates Co2+ and C O ( N H ~ ) ~ ( O H with ~ ) ~ ~quantum + strate without resort to ionizing radiation and ultraviolet techniques involving weakly absorbing materials. In this yields in neutral solution of 0.06 and 0.1, respectively. Irradiation in the presence of 1 M 2-propanol in deaerated paper we shall examine the behavior of COS- generated by this technique with particular emphasis on the acidsolution causes +(Co2+) to double suggesting the presence of a radical which can oxidize the alcohol generating the base properties of the radical.12 strongly reducing (CH3)zCOH radical which can reattack Experimental Section the substrate. Co(NH3)&03+ (as the nitrate salt) was prepared acFigure 1 shows the absorption spectrum of the transient cording to literature procedure^^^ as was Co(NH3)&03+ species with ,A, 600 nm from the flash photolysis of 7.0 (as the chloride salt).14 Conversion of the bidentate comx 10-5 M Co(NH&C03+ solution at p H 6.4 (with the plex to the perchlorate salt was accomplished by dissolvoptical density read 50 psec after the start of the flash) ing the chloride salt in a minimum volume of water a t which is identical with that of the COS- radical generated 60", filtering quickly, and adding the filtrate to an equal by pulse r a d i o l y ~ i s ~ - ~from + B the reaction of OH radicals volume of 50% NaC104 solution (w/v). A bright purple with C032- or by the flash photolysis7 of COS2-. The solid was obtained after 1 hr at 0" which was filtered, intensity and position of the absorption maximum was not washed with alcohol-ether, and air dried. Recrystallizaaffected by the presence of 0 2 , by the pH of the solution tion did not change the absorption spectrum of the com(3-13), nor by the temperature (14-37"). The same tranplex which compared very well with the 1 i t e r a t ~ r e . l ~ sient absorption was also obtained from the flash photoly[Co(en)&03]C104 was prepared16 from [Co(en)&l]Cl which in turn had been prepared by the procedure of Bai(IO) N. Getoff, F. Schworer, V. M. Markovic, K. Sehested, and S. 0. lar.17 All solutions were prepared from triply distilled Nieisen, J. Phys. Chem., 75, 749 (1971). water which had been radiolyzed and photolyzed. Phos(11) V. W. Cope and M. 2. Hoffman, Chem. Commun., 227 (1972). (12) A preliminary account of this work has already been published: S. phate was used to buffer the solutions a t pH 4-9. -N. Chen and M. 2. Hoffman, Chem. Commun., 991 (1972). Flash photolysis was performed using a 22-cm optical (13) F. Basoioand R. K. Murmann, lnorg. Syn., 4, 171 (1953). cell and procedures which have already been reported.lsJQ (14) A. B. Lamb and E. B. Damon, J. Amer. Chem. SOC.,59, 383 (1937). For monitoring the 600-nm region, a Hamamatsu R136 (15) Y . Shimura and R. Tsuchida, Bull. Chem. SOC. Jap., 28, 572 PM tube was operated at about 600 V. The linearity of (1955). the detection system was checked and found to be satis(16) M. Linhard and G. Stirn, 2. Anorg. Chem., 268, 105 (1952). (17) J . C. Bailar, Inorg. Syn., 2, 222 (1946). factory. A glass filter (Corning 3-72) was always in place (18) G. Caspari, R. G. Hughes, J. F. Endicott, and M. 2. Hoffman, J. between the analyzing lamp (Osram 150w Xe lamp) and Amer. Chem. SOC., 92, 6801 (1970). (19) A. F. Vaudo, E. R. Kantrowitz, M. 2. Hoffman, E. Papconstantinou, the cell and an electric shutter restricted the exposure of and J. F. Endlcott, J. Amer. Chem. Soc., 94, 6655 (1972). the solution to the light. The solutions used were very di120) T. P. Dasauota - . and G. M. Harris, J. Amer. Chem. Soc., 91, 3207 lute ( - 5 X M ) to ensure homogeneous absorption of (1969). (21) D. J. Francis and R. B. Jordan, J. Amer. Chem. SOC.,89, 5591 the photolyzing flash and the distribution of transient (1967); 91, 6826 (1969). species in the cell. Transient absorption spectra were ob(22) T. P. Dasgupta and G. M. Harris, J. Amer. Chem. Soc., 90, 6360 (1968). tained from the optical density of the absorption mea-
+
+
,
The Journal of Physical Chemistry, Vol. 77, No. 9, 1973
I
1113
Photolysis of Carbonatoamine Complexes of Cobalt(I I I )
I O.O4
t
O.O1
'
I
I
,
time, msec
I
1
t P1
0'
I
5do
I
I
I
600 A,
m
I
nm
Figure 1. Spectrum of the COS- radical from the flash photolysis of 7.0 X M Co(NH3)4C03+at pH 6.4. Optical density read 50 ksec after the start of the flash. sis of Co(en)zC03+ (with a markedly reduced yield) and Co(NH3)5C03+. The absorbance of the solution a t 600 nm after the decay of the transient was the same as that before the flash. A second-order plot of the decay data (Figure 2) shows linearity through approximately 60% of the decay with the rate becoming faster than expected in the final stages of the reaction. Such deviations from complete linearity are also shown for the decay of C03- generated pulse radiolytically23 at pH 10.2. It is to be noted that in more alkaline solution, the decay of COa-, generated pulse radiolytically or flash photolytically, shows a more perfect second-order plot (Figure 2). This effect had been noted previously6 but no reason for this apparent catalytic behavior had been advanced. More recently,8 it was suggested that if the bimolecular recombination of C03- generates HzOz, then the rapid reaction of CO3- with HzO2 could produce this catalytic behavior as the Hz02 builds UP. Because the decay of the 600-nm absorption from Co(NH3)4C03+ also gave respectable first-order plots, an independent test of the order of the initial decay was performed by measuring the half-life of the decay as a function of the initial optical density of the transient. The intensity of the flash was varied by varying the voltage of the discharge and the capacitance of the energizing condensers. In the first place, the intensity of the initial transient absorbance was proportional to the intensity of the flash ( I a yZCvZ) indicating that the 600-nm absorption is generated in a monophotonic process. Secondly, Figure 3 shows that t 1 / 2 a l/(OD)o and that the decay of the transient generated from the complex is unquestionably second order. It should be mentioned that at no time was any formation of the 600-nm transient observed with the full absorption being present at the start of the observation period (50 wsec after the start of the flash). The decay of the transient was unaffected by the presence of 0 2 and small M ) concentrations of C032-, NO3-, and Clod-, and by changes in temperature. Assuming that the species that absorbs at 600 nm is the COS- radical and that its second-order decay corresponds to reaction 1, then 2kl can be obtained from the slope of the plots in Figure 2 taking7 e600 1830 M - l cm-l. The values of 2h1 from Co(NH3)4C03+ a t pH 7 are very weakly dependent on ionic strength; a t high pH values, 2kl becomes dependent on ionic strength with the rate constant
-
0
1
2
3 4 5 time, msec
6
7
Figure 2. Second-order decay kinetics of the COS- radical gerrerated by various techniques, monitoring wavelength 600 nni: 0 , 0.1 M KzC03, NZO-saturated solution, pH 10.2, pulse radiolyIM sis, 2-cm optical cell: 0 , same at pH 12.7; 0 ,5.0 X Co(NH3)4C03+. air-saturated solution, pH 7.0, flash photolysis, 22-cm optical cell; U, 0.1 M Na2C03, N2-saturated solution, p H 11.8, flash photolysis, 22-cm optical cell.
Figure 3. Dependence of t 1 / 2 for the decay of the COS- radical on the initial radical concentration (optical density): [Co(NH3)4C03+]= 5.0 X M, pH 7.0, air-saturated SOILtion, monitoring wavelength 600 nm, ionic strength = 0.1 M. Inasmuch as t 1 / 2 = '/2k[CO3- 0 for bimolecular second-order decay, taking 6600 1.83 X l o 3 M - 1 cm-l and an optical cell path length of 22 cm gives 2k = 2.8 X 1 O7 M - sec- l . increasing as ionic strength is increased. As Table I shows, whereas 2/21 for COa- generated from pulse radiolysis is virtually independent of pH (10.2-12.7)with the variation in rate attributable to the change of ionic strength, the radical from the complexes shows a marked increase in rate a t pH >10 with 2/21 rising to a maximum a t pH l:!. Even in highly alkaline solution the observed decay is second order and the spectrum and its intensity are the same as in neutral solution. (23) The pulse radiolysis facilities at the U S Army Natlck Laboratories were used for this experiment and the authors thank Dr E Hayon for the use of the instrument The Journal of Physical Chemistry, VoI. 77,
No. 9. 1973
1114
S.Chen, V. W. Cope, and M . Z . Hoffman
TABLE I:
Decay of C 0 3 -
[scavenger], M
Q
0
System
0.1 M K&O$
0.1 M Na2C03d 5 x 10-5 M CO( N H 3 ) 4 C 0 3 +
5 x 10-5 M CO( N H3) & 0 3 +
PH
10.2 12.7 11.8
1.9 x 107 2.9 x 107 6.9 x 107
7.1 7.6 8.7 9.5 10.0 10.6 10.8 11.0 11.6 12.0 12.7 13.0
3.8 x 4.1 x 4.0 x 6.1 x 7.6 x 2.0 x 3.1 X 1.1 x 8.0 X 1.1 x 1.2 x 1.3 x
107 107 107 107 107 108 lo8 109 lo8 109 109 109
9.5 9.8 10.0 10.2 10.5 10.9 11.0 11.2 11.5 11.9 12.2 12.6 13.0
1.1 x 1.7 X 1.3 X 1.8 X 1.9 x 4.1 X 5.3 x 7.4 x 8.0 X 1.4 x 1.3 x 1.3 x 1.3x
108 lo8 lo8 lo8 108 loB 108 108 lo8 109 109 109 109
a Determined at 600 nm from flash photolysis in air-saturated solution unless otherwise noted; ionic strength controlled at 0.1 M with NaC104. Determined from the slope of a plot of 1/OD vs. time taking ~ 6 0 01830 M - - 1 cm-- 1 and a cell path lengh of 22 cm (2 cm path length in the pulse radiolysis experiments). Pulse radiolysis experiment in NzO-saturated solution: ionic strength not controlled. Ionic strength not controlled: solution deaerated with NP.
The value of 2kl for Co(NH3)4C03- is constant below pH 7 to about pH 3 where the decay rate of the radical again increases sharply. A similar sharp increase in the decay rate was shown for C 0 3 - from Co(NH3)&03+ a t pH
