3044
A.
(2) (a) J. N. Demas and A. W. Adamson, J. Amer. Chem. SOC., 93, 1800 (1971); (b) N. Sabbatini and V. Balzani, ibid., 94, 7587 '( 1972). (3) (a) P. Natarajan and J. F. Endicott, J. Amer. Chem. SOC.,94, 3635 (1972); 95, 2470 (1973); (b) P. Natarajan and J. F. Endicott, J. Phys. Chem., 77, 971 (1973); ( c ) J. N. Demas and A. W. Adamson, J. Amer. Chem. SOC.,95, 5159 (1973). (4) N. J. Turro, "Molecular Photochemistry," W. A. Benjamin, New York, N. Y., 1965, p 132. ( 5 ) J. Saittei, J. D'Agostino, E. D. Megarity, L. Metts, K. R. Neuberger, M. Wrighton, and 0. C. Zafiriou, Org. Phofochem., 3, 1 (1973).
Levy, D. Meyerstein, and M. Ottoienghi
(6) D. G. Whitten and M. T. McCali, J. Amer. Chem. SOC., 91, 5097 (1969). (7) G. Fischer, K . A. Muszkat, and E. Fischer, J . Chem. SOC. 6, 156 (1968). (8) L. Meites, "Polarographic Techniques," 2nd ed, Interscience, New York. N. Y . . 1965. 011671-711 (9) A. A. VlCek; Discus;. Faraday Soc., 26, 164 (1958). (10) F. G. Moses, R. S. H. Liu, and B. M. Monroe, Mol. Photochem., 1, 245 (1969), (11) C. G. Hatchard and C. A. Parker, Proc. Roy. SOC.,Ser. A, 235, 518 (1956),
Photodissociation of lodoaromatics in Solution A . Levy,* D. Meyerstein, Nuciear Research Centre-Negev,
Beer-Sheva 84190, israel
and M. Ottolenghi Department of Physical Chemistry, The Hebrew University, Jerusalem, lsrael (ReceivedJune 4, 7973) Pubiication costs assisted by the Nuciear Research Centre-Negev
Deiodination and isotopic-exchange processes are employed for determining the photodissociation yields of various iodoaromatic molecules in solution as a function of temperature and excitation wavelength. In the case of 1-iodonaphthalene the direct-excitation yields are compared with those obtained by photosensitization with benzophenone. The data indicate that dissociation takes place after thermal relaxation from either singlet or lowest triplet states. Photodissociation of these two excited states exhibits a different temperature dependence. Rate constants for the reaction of phenyl radicals with aromatic scavengers are determined and discussed along with the possibility of H-atom migration within the radical ring.
Introduction In previous publications1 we proposed a mechanism consisting of reactions 1-3 to account for the competition between the photoinduced exchange and deiodination in iodobenzene (PhI) solutions in the presence of radioactive iodine ( P I I ) .
Ph.
+
+
11311
I,
h
k
h~ 4 Ph*
k2
0,
21.
Ph.
+
---
PhI
+
1.
oxidation products
(1) (2)
(deiodination)
PhI131
+ 1. ( e x c h a n g e )
(3)
The final consequences of light absorption by the system are thus determined by the competition between dissolved oxygen and iodine on the phenyl radicals produced by photodissoc:ation of iodobenzene. Alternative mechanisms, such as exchange induced by photodissociation of 12, or uia excitation of the PhI.11311 charge-transfer complex, were ruled out. The observation that the photocleavage of iodobenzene (1) is wavelength dependent in the uv range, where part of the absorbed light leads directly to the triplet state of the molecule, raised questions relevant to the nature of the primary photodissociation step of iodoaromatics in solution. In the present work we have carried out photochemical experiments bearing principally on the following points. (a) The applicability of the proposed mechanism to other iodoaromatic molecules. (b) The details of the primary photodissociation step, such as the exact roles of the excited, thermalized, or nonthermalThe Journai of Physicai Chemistry, Vol. 77, No. 26, 1973
ized singlet and triplet states. (c) Properties of the aryl radical, related to its reactivity with added solutes and to the possibility of H atom migration along its ring. The results lead to a new insight into the photodissociation of iodoaromatics in solution for which only qualitative information was available.
Experimental Section (a) Materials. All details concerning iodobenzene, iodine, iodine-131 and methylcyclohexane have been previously described.l 0 - and m-iodotoluenes (BDH) were purified by vacuum distillation in a dry nitrogen atmosphere. In the case of the para isomer a preparative gaschromatographic procedure was employed. The purity of all isomers was checked by glc analysis and by uv spectroscopy. Benzene, chlorobenzene, toluene, benzonitrile, and benzophenone (all Merck, analytical grade) as well as 1-iodonaphthalene (Fluka purum) were used without further purification. ( b ) Procedure. Deiodination and exchange measurements a t various excitation wavelengths were carried out as previously described.1 Iodotoluene isomers were separated gas chromatographically using a diethylene glycol adipate column (Analabs G P 35A 78 in. diameter, 6.5 m long). The retention times obtained at loo", with a 35 C C / min rate of gas flow, for the ortho, para, and meta isomers, were 37.08, 38.83, and 39.33 min, respectively. Under the above conditions, ortho-para and ortho-meta mixtures were readily separable with a 5% sensitivity. However, due to the close values of the corresponding re-
3045
Photodissociation of lodoaromatics in Solution TABLE I: Maximum Quantum Yield Values (@DE' and @EX')'' M Solutions of and Rate Constant Ratios Measured by 5 X lodotoluenes in Methylcyclohexane, Irradiated at 31 3 nm, M Iodine in the Presence of -5.3 X
Ortho
Meta Para
0.41 :k 0.02 0.41 :k0.02 0.39 :k0.02
0.41 f 0.02 0.41 f 0.02 0.38f 0.02
1.95f 0.2 2.30 & 0.2 1.70 f 0.2 1.93f 0.2 1.96f 0.2 1.96 & 0.2
a $DE" and q 5 are ~ the ~ deiodination ~ and exchange quantum yields measured correspondingly when k2[02] >> k3[12] and w h e n k3[12] >> k 2 [ 0 2 ] . According to the mechanism of reactions 1-3 they should be equal, representing the net photodissociation yield of iodobenzene. * Obtained from slopes of ~ / R D Evs. 1 / [ 0 2 ] plots. Obtained from slopes of 1 /REX vs. [02] plots.
I
Results (a) Photoinduced Exchange and Deiodination in Solutions of lodotoluenes Photochemical experiments were carried out in methylcyclohexane solutions of the three iodotoluene isomers, measuring the rates of deiodination and exchange (RDE and REX) as a function of the oxygen concentration. The mechanism of reactions 1-3 predicts linear relationships between ~ / R D and E l/[Oz], as well as between 1/REx and [ O Z ] .Both ~ relations are found to be accurately fulfilled in the above iodotoluenes system. The ratio k3/lzz, obtained from the slopes of such plots using 313-nm excitation, as well as the maximum yields, @DE' and @EXo, are presented in Table I. Similar experiments were also carried out a t 365 nm leading, in all cases, to the appreciably lower value of 0.23 f 0.02 for both DE' and @EX'. This wavelength effect on the yields is similar to that previously observed for iodobenzene.1 The feasibility of a photoinduced interchange between the isomers was investigated by exciting deaerated 0.1 M solutions a t 313 nm, varying the iodine concentration in the range between 2 x and 2 x M . In all cases the total amount of aryl radicals produced by irradiation (determined from the product of the absorbed light intensity, the irradiation time, and the photodissociation quantum yield) was above 20% of the iodotoluene concentration. Under such conditions no photoinduced isomerization could be detected. ( b ) Temperature Effects on the Photochemistry of Iodobenzene. Temperature effects on the deiodination of iodobenzene were studied by exciting air-saturated methylcyclohexane solutions a t 254, 313, and 365 nm. The formation of IZ was followed by measuring the absorbance at 520 nm as a function of the irradiation time. Figure 1 presents the values of $DE' obtained from the slopes of such initially linear curves, where kz[Oz] >> k3[Iz] and consequently @DE = @DE'. In the same figure we present values of @EXo measured in evacuated, or Nz saturated, PhI solutions where kz[Oz] ~ [ I z and ] @EX = &x'. In all cases we confirmed that the increase of the PhI absorption with temperature could be neglected. The excellent agreement, a t all temperatures, between the values of @EX' and @DE' indicates that genuine temperature effects on the net yield of photodissociation are actually observed,
,365
*4+J
254 nm 313 nm 365 nm
~J
1 1
nm
DEIOD.
I
.io-
tention times, we have not been able to separate mixtures of para and meta isomers. Temperature effects were measured using a metal cell mount, thermnstated (&0.2") by alcohol or water circulation.
//
I
I
I
20
40
60
c Figure 1. Temperature effects on the yield of iodobenzene photodissociation (4') in methylcyciohexane. (a) Yield determinations by deiodination were performed at 254 ( [ P h i ] = 0.001 M ) , 313 ( [ P h l ] = 0.03 M ) , and 365 n m ( [ P h l ] = 0.5 M ) ( b ) Yield determinations by isotope exchange were performed at 254 M and [ I 2 ] = 2 X M ) , 313 ( [ P h i ] = 3 X ([Phl] = IO-* M and [I2] = 2 X l O W 4 M ) , and 365 nm ([Phi] = 0.5 M and [I*] = 2-6 X M). 0
As clearly demonstrated in Figure 1 an identical temperature effect is observed a t 254 and 313 nm, differing from that observed a t 365 nm. (c) Photosensitized Exchange and Deiodination of 1lodonaphthalene. In order to clarify the role of the triplet state in the photodissociation of iodoaromatics, we have carried out photochemical experiments attempting to obtain selective triplet population uia triplet-triplet energy transfer using benzophenone as sensitizer. Irradiation of deaerated benzophenone solutions in the presence of IZ resulted in an efficient iodine consumption, ultimately leading to total bleaching of the solution. This process, suppressed by molecular oxygen, was attributed to the triplet reaction
+
k4
iodo-organic products (4) Energy transfer from the triplet state of benzophenone to an iodoaromatic (ArI) acceptor can thus occur only if reaction 5 3(Ph CO Ph)*
3(Ph CO Ph)*
+
I,
k
ArI
Ph CO Ph
+
3(ArI)*
(5)
competes efficiently with reaction 4. In view of the relative inefficiency of reaction 5, in the case of iodobenzene, this requirement could not be fulfilled even at low [Iz] and high [PhI] values. However, the addition of l-iodonaphthalene was found to inhibit bleaching process 4 leading to its complete suppression when [ArI]/[Iz] 2 lo2. Thus, under such conditions, any organic radioiodine originating in reaction 4 could be neglected. An additional factor in choosing the exact concentrations in the photosensitization experiments is associated with the observation that relatively high benzophenone concentrations lead t o deiodination of ArI even in deaerated solutions. The effect which, as shown below, is observed for several aromatic compounds is attributed to the Ar. radical scavenging reaction The Journai of Physical Chemistry, Voi. 7 7 , No. 26, 1973
3046
A. Levy, D. Meyerstein, and M. Ottolenghi
Ar.
+
k
Ph CO Ph -2Laddition products
followed by
I.
+
1.
-
(6)
I, (deiodination)
/
where 1. and Ar. are formed from the dissociation of 3 ( ~ r 1 ) 2* 3(ArI)*
--L
-
Ar.
+
1.
i
I
; )
.4 I-
(7)
Reaction 6 competes with Ar.
+
113'1
k
ArI131
+
1. (exchange)
i
.3 -
./
(8)
To suppress reaction 6 we have carried out all experiments a t [Ph CO Ph]/[I2] 5 20, where no deiodination is observed. The exchange rates induced a t various temperatures by benzophenone photosensitization via reactions 5 , 7 , and 8 are presented in Figure 2. The temperature effect is compared with those observed in the case of direct excitation of 1-iodonaphthalene a t 365 and 313 nm. ( d ) Iodobenzene Deiodination and Exchange in Aromatic Solvents. Solutions of iodobenzene and iodine in deaerated aromatic solvents undergo an iodine exchange process which is accompanied by deiodination as in the case of aerated methylcyclohexane solutions. The dependence of the exchange yields on the iodine concentration in the cases of benzene, toluene, and chlorobenzene are shown in Figure 3a. The irradiation time a t 313 nm was always short so as to minimize the change in [I21 during the exposure. We have also carried out experiments in methylcyclohexane as solvent to which varying amounts of aromatic compounds were added. The relevant data obtained with 365-nm excitation for iodobenzene and cyanobenzene are presented in Figure 3b. In all cases the exciting light was exclusively absorbed by iodobenzene, thus avoiding possible complications due to the photolysis of other aromatics.
.1
/
-
I
I
I
I
I
0
20
40
60
"C
Figure 2. Temperature effects on the exchange quantum yield of I-iodonaphthalene (Ari) in deaerated methylcyclohexane. (a, m) Exchange induced by direct excitation of 1 -iodonaphthalene (at 313 n m ) leading to Sl:[Ari] = 10-3 M ; [ I 2 1 = M. (b, A) Exchange induced by direct excitation (at 365 nm) of l-iodonaphthalene to the triplet state: [Arl] = l o - ' M ; [ I 2 ] = 2 X M . (The data have been obtained after subtracting the contribution of excitation to Sl).(c, 0 ) Exchange induced by benzophenone photosensitization at 365 nm: [ A r l ] = IO-' M ; M ; [PhCOPh] = 2.5 X M. The data have [I21 = 2 X been corrected for a small (-20%) contribution of direct excitation of 1-iodonaphthalene at this wavelength. The correction was carried out by measuring the exchange when l-iodonaphthalene was directly excited at 365 nm, taking into account the light fraction absorbed by I-iodonaphthalene in the presence of benzophenone.
Discussion ( a ) Primary Dissociation Process. The results presented above show that the basic deiodination and exchange mechanism of reactions 1-3 previously proposed for iodobenzene1 is also valid for the iodotoluene isomers as well as for 1-iodonaphthalene, The question arises as to the details of photodissociation process l. In all cases, excitation a t 254 and 313 nm leads directly to the lowest excited singlet state of the molecule (&). Figure 1 shows that within the 0-65" range the quantum yield for photodissociation at 254 nm is identical with that a t 313 nm. This lack of a wavelength effect indicates that photodissociation in the SIstate occurs after complete thermal relaxation. Figure 1 shows, however, that the quantum yields measured with 365-nm excitation are usually lower, exhibiting a different temperature behavior, than those a t 254 and 313 nm. This apparent discrepancy can be readily rationalized by the fact that excitation a t 365 nm is within the tail of the SO -*SIband where the superimposed direct transition to the T1) contributes substantially lowest triplet state (So ( -50%) to the extinction coefficient.l.3 Thus, the difference observed between excitation at 254 or 313 nm and that at 365 nm can be attributed to differences in the photodissociation efficiencies of SIand TI. Direct evidence for photodissociation of the triplet state of iodoaromatics is derived by comparing the photosensitized dissociation yields in the 1-iodonaphthalene system with those obtained by direct excitation to T1 a t 365 nm. Figure 2 shows that absolute quantum yields of the two
-
The Journalof Physical Chemistry, Voi. 77,
No. 26, 1973
J
I I
Figure 3. Kinetic plots showing the competition between 12 and aromatic molecules on the scavenging of phenyl radicals: (a) excitation at 313 nm, [Phl] = 0.03 M in ail three aromatic solvents; ( b ) ( 0 ) excitation at 365 nm in methylcyclohexane. [I*] = 2 X M; ( A )excitation at 365 nm in methylcycloM, [Phl] = 0.1 M . hexane, [I2] = 5.2 X processes, as well as their dependence on temperature, are exactly identical. The data not only prove that dissociation may occur via triplet energy transfer from benzophenone but also that, in respect to dissociation, the triplet TI excitation is indistinguishapopulated by direct So ble from that obtained by photosensitization. The same data also suggest that, as in the case of SI,dissociation occurs from T1 after thermalization. The conclusion that +
3047
Photodissociation of lodoaromatics in Solution TABLE II: Relative Rate Constants for the Reaction between
Phenyl Radicals and Some Aromatic Compounds kPhx/kPhH
k d a
PhX
PhH PhCN Ph I PhCi PhCH3
I
React ion
C oord ina te
Figure 4. Schematic energy level diagram showing the thermally activated dissociation of Si and T i of iodoaromatics.
the optically populated TI state is photochemically active is in variance with the suggestion that direct excitation to T1 leads to negligible photodissociation.2 The early deiodination experiments were carried out in deaerated solutions where reaction 2 does not occur and reaction 1 is followed by the recombination of Ar. and I.. In these photosensitization experiments benzophenone not only acted as a triplet sensitizer but also replaced O2 in reaction 2. This explains why in the presence of benzophenone the photosensitized deiodination yields in deaerated solutions2 are essentially identical with the present exchange quantum yields, both representing the yield of triplet photodissociation. When attempting to understand the temperature effects on the thermal dissociation of either SIor T1 as shown in Figures 1 and 2, the complex nature of photodissociation in solution should be considered, Temperature may affect the primary yield of radicals initially formed in a photochemical cage as well as the net yield of pairs which escaped secondary geminate recombination.4 A tentative expression for the photodissociation yield is (9) where kd and k , are competitive rate constants as shown in Figure 4. If kd = k d a e - A E i Rwith T AE, kd (and kr) being different for SI and TI, the factor k d / ( k d 4- hr) will exhibit a different temperature dependence for the two corresponding states. 0 is the probability of escaping secondary recombination uia a random walk process, and depends on the initial radical separation determined by the solvent viscosity5 and will thus increase with the temperature. Our data (Figure 1 and 2) indicate that 4' approaches a plateau at high temperatures suggesting t h a t kd8/(kd + k,) approaches unity. Since our limiting value is (for iodobenzene) 4' =- 0.51 < 1, an additional transmission factor K determining the fraction escaping deactivation to SOat point A (see Figure 4) should be introduced. This interpretation is consistent with data obtained using isopentane as solvent, in experiments similar to those described for methylcyclohexane in Figure 1 . I C Higher 4' values are obtained a t low temperatures, fitting the characteristic S-shaped curve which levels off a t the same plateau around 0.51. In view of the lower viscosity of isopentane these results may be rationalized by higher 0 ( r ) and kd values, as well as by solvent-independent K . According to this general picture TI and SI exhibit similar K value for iodobenzene but different ones in the case of l-iodonaphthalene. A quantitative analysis of eq 9 requires a knowledge of the temperature effect on 0. When assuming that 0 is temperature independent and plotting log [ ( + O m a x / 4') - 11 against 1/T we obtained apparent AE values in
kphx
a
2.0 x 104 5.1 x 103 0.9 x 104
1.3x 104 4.0x 104
Present measurements
1 .o 3.9 2.2 1.5 0.5
Ref 7and8
1 .o 3.7 1.8
1.4 1.7
Calculated from data in Figure 3.
the range between 6 and 12 kcal/mol. These values are unreasonable since, in view of the time scale in which the photodissociation takes place, they predict unacceptably high frequency factors (kd'). Lower values for AE and k d are obtained when assuming a temperature effect on 8. However in the absence of suitable models describing quantitatively the function 0( 5") no reliable values could be obtained. ( b ) Reactiuity of the Phenyl Radicals. The failure to observe a photoinduced isomerization between iodotoluene isomers indicates that, in agreement with previous suggestions,6 no H atom migration around the aromatic ring takes place within the lifetime of the radicals. Assuming M-1 sec-l for the scavenging rate conthe value of stant of Armby 12 (which is an upper limit), one obtains (with [I2] = 2 X ' l O - 6 M ) a value of T 1 / 2 = 0.693/k[Iz] = 0.035 msec for the half-life of the aromatic radical. In view of the sensitivity of our analytical methods this implies a lower limit of 0.1 msec for the half-life of the hydrogen migration process. The relative rate constants for the reaction of the phenyl radical with aromatic scavengers, as calculated from the data of Figure 3, are presented in Table 11. With the exception of toluene the table shows a fair agreement with previous data obtained from thermal decomposition experiments a t elevated temperature^.^^^ According to the present room-te-mperature data, it appears that the reaction of P h . with aromatics is nucleophilic in nature as compared to electrophilic reaction of aromatics with iodine atoms.g This may be consistent with a charge-transfer intermediate [ArX -Ph+] analogous to the species [ArXf .I-] which has been observed spectroscopically a t room temperature.1°
Acknowledgment. The authors wish to thank Professor A. Treinin for valuable discussions. They, are indebted to Mr. S. Shrem for his help in carrying out photochemical experiments and to Mr. Ch. Klein for the gas-chromatographical determinations. References and Notes (1) (a) A. Levy, D. Meyerstein, and M. Ottolenghi, J. Phys. Chem., 75, 3350 (1971); (b) J . Appl. Radiat. Isotopes, in press; (c) A. Levy, Ph.D. Thesis, Hebrew University, Jerusalem, 1973. (2) F. Wilkinson, il. Phys. Chem., 66, 2569 (1962). (3) E. Olaertsand J. C. Jungers, Discuss. Faraday Soc., 2, 222 (1947). (4) R. M. Noyes, Progr. React. Kinet. 1, 129 (1961). (5) D. Booth and R. M. Noyes, J . Amer. Chem. SOC.,82,1868 (1960). (6) R. K. Sharma and N. Kharasch, Angew. Chem., Int. Ed. Engl., 7, 36 (1968). (7) C. Walling, "Free Radicals in Solution," Wiley, London, England, 1957, p 484. (8) G. H. Williams, "Homolytic Aromatic Substitution," Pergamon Press, London, England, 1960. (9) M . Nakashima, C. Y . Mok, and R. M. Noyes, J. Amer. Chem. SOC., 91, 7635 (1969). (10) S. J. Rand and R. L. Strong, J . Amer. Chem. SOC.,82, 5 (1960). The Journal of Physical Chemistry, Vol. 77, No. 26, 1973