J . Chem. SOC., Faraday Trans. I , 1985, 81, 2569-2575 Reactions of Cobalt(II1) Compounds with some Free Radicals Derived from Uracil BY SUDHINDRA N. BHATTACHARYYA* AND PARIKSHIT C. MANDAL Nuclear Chemistry Division, Saha Institute of Nuclear Physics, Sector- 1, Block-' AF ', Bidhannagar, Calcutta 700 064, India Received 14th March, 1984 Reactions of Co"I complexes with the transient adducts of uracil (U), viz. U-, UH and UOH, have been studied by means of y-radiolysis of uracil in the presence of CoII'EDTA and Co'I'NTA (where EDTA is ethylenediamine tetra-acetate and NTA is nitrilotriacetate). The U-and UH radicals formed by the reactions of eiq and H with uracil reduce Co"' to Co" complexes. The UOH radicals, however, do not reduce the CoIII complexes although such reductions have been observed with Cu'I and FelI1 analogues as reported earlier. The absence of such electron transfer from UOH to CoII' could not be explained on the basis of the redox potential of the COI~~EDTA/CO'IEDTA couple, which was 0.38 V.However, this apparent anomaly has been explained on the basis of the individual structures of the respective complexes. When uracil in dilute aqueous solution is radiolysed, the water-derived radicals (eCq, H and OH) combine with the uracil molecule leading to the formation of the transient adductsl U-, UH and UOH. If metal ions, viz. FeIII and CuII, are present (either uncomplexed or complexed with aminopolycarboxylic acids) in the system, it has been reported earlier2, that such transients undergo electron-transfer process with the metal ions.CoIII constitutes another transition-metal species which has similar redox behaviour as that of FeIII and CuII. Incidentally, in connection with studies of the reactions of electron-affinic radiosensitizers with the radicals derived from pyrimidine bases it has been shown4 that the rate of electron transfer from the radicals to the electron-affinic compounds is dependent on their respective redox behaviour. It is known that the redox potentials of the relevant couples are FelI1/Ferl z 0.77 V,5 (CoIIIEDTA/Co*IEDTA z 0.38 V6 and CuI1/Cu1 z 0.17 V5 (EDTA and NTA repre- sent ethylenediamine tetra-acetate and nitrilotriacetate, respectively). Thus it is evident that the redox potential of the COIIIEDTA/CO~~EDTA couple lies between that of Fe1I1/FeI1 and Cur1/Cu1. It is therefore of interest to study whether the electron-transfer process also occurs with ColI1 ions.EXPERIMENTAL MATERIALS Uracil (Merck) was recrystallized three times from triply distilled water. [14Cz]uracil (14 mCi dm3 mmol-l) was obtained from BARC, Bombay. The Co'I' complexes, Co'IIEDTA and CoII'NTA, were prepared as reported earlier.'g All chemicals and solvents were of analytical- reagent grade. Deaeration was carried out by bubbling high-purity argon gas through the experimental solutions for ca. 30 min. Pure N,O was used throughout the investigation. IRRADIATION Irradiation was carried out with 6oCo y-rays and the dose rate was determined using a Fricke 2569 dosimeter.2570 REACTIONS OF CO"' COMPLEXES WITH URACIL ANALYSIS 2 x mol dmP3 uracil solution containing [14C,]uracil was radiolysed in the presence of the Co'II complexes, and the base degradation yield, G( - U), and the yields of the products were determined after separation of the different products by paper chromatography as described earlier.2* The reduction of the Co'" complexes was followed spectrophotometrically.RESULTS AND DISCUSSION The CoIII species, viz. CoIIIEDTA and CoIIINTA, have appreciable absorption in the range where uracil has maximum absorption, and hence the base decomposition of uracil in presence of the abovementioned metal complexes could not be measured by the direct measurement of loss of absorbance of uracil at the wavelength of its maximum absorption, Amax. However, the degradation yield and the yields of the different products formed in the radiolysis of uracil in the presence of cobalt ions could be determined after separating the products by paper chromatography as described earlier .2 ~ The radiolysis products were found to be similar to those obtained in the radiolysis of uracil in presence of other metal ions,2F3 e.g. CuII and FeIII. A typical plot of the formation of the various products arising from the degradation of the base as a function of absorbed dose is shown in fig. 1. The amounts of products formed were linear with dose, and hence the G values of each product were determined from the slope of the respective straight lines. However, the decomposition of the base was not linear at higher doses, and hence the value of G( - U) was determined from the initial slope of the line. The yields of the different products are shown in table 1, and table 2 shows the effect of varying the concentration of CoIIIEDTA on G(-U) and mol dmY3 uracil solution containing 5 x lop4 mol dm-3 CoIIIEDTA in the region 350-610 nm. As uracil has no absorption in this range, this absorbance is due to CoIIIEDTA, characterised by two peaks having A, at 380 and 535 nm, respectively.Fig. 2(b)-(4 represent the absorbances due to the same solution irradiated in an N,O-saturated medium to doses of 0.75 x 10l8, 1.5 x l0l8 and 3.0 x 10l8 eV ~ m - ~ , respectively. On irradiation the absorbances due to CoIIIEDTA at its absorption maximum decrease as the adsorbed dose increases. Hence, knowing the absorption coefficient, the decomposition yield of CoIIIEDTA, G(-CoIIIEDTA), was measured from the changes in absorbances at 535 nm.When the complex species was CoIIINTA, the decomposition of the complex was determined from the loss of absorbance at its absorption maximum, A,,, - 565 nm. Fig. 3 shows a typical plot for the decomposition of the CoIIINTA with absorbed dose. The decomposition of the complex was linear with dose and hence the G(-ColI1) values were determined from the slope of the straight line. The observed G values for the degradation of the ColI1 complexes are also included in table 1. Table 1 shows that when uracil is radiolysed in presence of CoI'I complexes in argon-saturated solutions, G( - U) z 1.1. Such low G( - U) values indicate that not all the primary radicals are involved in the radiolytic degradation of uracil.However, under the present experimental conditions, i.e. for the radiolysis of 2 x mol dm-3 uracil in the presence of 5 x mol dm-3 CoI1I complexes, the OH radicals react preferentially1* with uracil to give UOH. However, ca. 14.2 and ca. 57.7% of eiq will be scavenged9. lo by CoIIINTA and CoIIIEDTA, respectively. Nevertheless, this partial scavenging of eLq by the cobalt ions does not account for the large decrease in the G( - U) values. Hence it may be assumed that the U- radical, being a very good G( - CoI'IEDTA). Fig. 2(a) shows the absorption spectrum of 2 xS. N. BHATTACHARYYA AND P. C. MANDAL 257 1 6 h 6? - 4 9 ." x 1 0 0 1 2 3 4 Fig. 1. Radiolysis of 2 x 1 0-3 mol dm-3 uracil in the presence of CoIIIEDTA (5 x 1 0-4 mol dm-3) in N,O-saturated solution at neutral pH.@, Degradation of uracil; A, dimer; a, cis-uracil glycol; x , trans-uracil glycol; A, hydroxydihydrouracil. Table 1. Product yields in the radiolysis of 2 x lop3 mol dm-3 uracil in presence of various Co"I complexes (5 x mol dm-3) at neutral pH condition" G(products) A B C D G(dimer) G(cis-uracil glycol) G(trans-uracil glycol) G(hydroxydi hydrouracil) G(isobarbituric acid + dihydrouracil) G(dia1uric acid) G( - uracil) G( - CoI'I complex) 0.20 0.90 0.12 0.30 0.14 0.40 0.13 0.40 0.30 0.60 0.50 0.95 0.12 0.20 0.40 0.70 0.13 0.20 0.20 0.50 0.12 0.20 - - 1.10 2.50 1.20 2.90 2.70 0.70 3.30 1.10 a In the presence of (A) CoII'EDTA in argon-saturated solution, (B) CoIIIEDTA in N,O saturated solution, (C) CoI"NTA in argon-saturated solution and (D) Co"INTA in N,O saturated solution. Table 2.Effect of Co'IIEDTA Concentration on G( - U) and G( - Co'IIEDTA) in the radiolysis of 2 x lop3 mol dmp3 uracil in the presence of Co"IEDTA in deaerated solution at pH ca. 5.5 ~ ~ ~ ~~~ [Co"'EDTA] /lop4 mol dm--3 G( - U) G( - Co'I'EDTA) 3 5 8 10 1.1 1.1 1.2 1.1 3 .O 2.7 2.7 2.82572 REACTIONS OF cO"* COMPLEXES WITH URACIL 0.001 1 I I I 3 50 L1 0 L70 530 590 A/nm Fig. 2. Change in absorption spectrum on irradiation of N,O-saturated uracil (2x lop3 mol dmp3) solution in the presence of 5 x lop4 mol dmp3 CoIIIEDTA at neutral pH. (a) Unirradiated; (b), (c) and (4, irradiated to doses of 0.75 x 10ls, 1.5 x 10l8 and 3.0 x 10l8 eV ~ m - ~ , respectively dose/ 1 0-17 ev c n i ~ ~ Fig. 3. Decomposition of CoII'NTA with absorbed dose in the radiolysis of 2 x lop3 mol dmp3 uracil in the presence of 5 x mol dmp3 CoI'INTA in N,O-saturated solution at neutral PH.S.N. BHATTACHARYYA AND P. C. MANDAL 2573 electron donor,ll transfers an electron to the ColI1 species, as a result of which the metal ion is reduced and the uracil molecule is regenerated : U-+Co'II -+ COII+U. (1) (2) Likewise, the UH radical may also reduce the metal ion: UH + Co'II --+ CoII + UH+. However, the uracil carbonium ion, UH+, does not give rise to product but reverts back to uracil:2p3 H,O UH+ -U+H30+. ( 3 ) It thus follows that no degradation of uracil can be accounted for by the primary reducing radicals. This has been justified from a study of the radiolytic disappearance of CoIII species.If the U- and UH radicals undergo reactions (1)-(3), then CoII should constitute one of the products of radiolysis. This has been proved by the following experiment. When hydrogen peroxide is added to the irradiated solution the absorbance lost due to irradiation is recovered. Since the recovery of absorbance occurs in the absence of any further addition of the ligand, it may be concluded that the product formed constitutes only the CoII species having no ligand degradation. Further evidence in favour of occurrence of reactions (1) and (2) was obtained by following the reduction yield of ColI1 species in the presence of varying concentrations of uracil and CoIII complexes. If the fate of U- and UH is only to react with ColI1 by reactions (1) and (2), then the reduction yield of ColI1 arising from these reactions (together with those of e;ts and H) should be independent of uracil and ColI1 concentrations, and it has been found that G( -CoIr1) is independent of both uracil concentration (fig.4) and CoIII concentration (table 2). From the above discussion it may then be stated that the degradation of the base is achieved primarily due to the subsequent reactions of UOH. The question arises as to whether the UOH radicals undergo reaction with CoI" species by electron transfer or undergo reactions between themselves, as was observed in the absence of a metal ion.12 If the former were the case, as seen2. in the presence of CuII and FeIII, then the G( - U) value would have corresponded to a value of ca. 2.7. However, the observed yield is much lower than this.This reduced yield might have arisen from the disproportionation of the UOH radicals, as has been proposed earlier1, in the absence of a metal ion. It may then be argued that UOH does not undergo electron transfer with Co'II. Were this process effective, the products arising from UOH should depend on the concentrations of CoIII species. However, it is evident from table 2 that the G( -U) value does not change as the ColI1 concentration varies, indicating the non- involvement of UOH in these reactions. That the UOH adduct does not undergo electron transfer with the CorT1 species is further evident from a radiolysis study in N,O-saturated solution. When uracil (2 x mol dmP3 CoIII species in an N,O-saturated solution, the G( - U) value is double that observed in a deaerated solution.The observed G( - U) value of ca. 2.5-2.9 may be explained as being due to disproportionation of the UOH radicals, since the OH yield under these conditions should increase because of the reaction mol dm-3) is radiolysed in the presence of 5 x eiq+N,O -+ OH+OH-+N,. (4)2574 REACTIONS OF CO'" COMPLEXES WITH URACIL 1ci4 1 o - ~ lo-* [ uracil]/mol dm-3 Fig. 4. Effect of uracil concentration on the reduction yield of CoIII in the radiolysis of uracil in the presence of 5 x lop4 mol dm-3 Co"' complexes in Ar-saturated solutions at neutral pH. 0, CoIIIEDTA; a, CoIIINTA. From the observed yield of G( - CoIII) it is evident that eZq does not totally react with N,O. A consideration of the rate-constant data for eZq with N20,13 CoIIINTAlO and Co1I1EDTA9 indicates that G( -ColI1) arising from e;ts corresponds to ca.0.4 and ca. 0.7 for CoIIINTA and CoIIEDTA, respectively. Hence the expected G( - CoIII) values would correspond to 0.4+GH = 1.0 and 0.7+GH = 1.3 in the presence of CoIIINTA and CoIIIEDTA, respectively. The observed results for G( - CoIII) are close to those expected. The absence of electron transfer from UOH to ColI1 in its complex with EDTA, unlike that in case of CuI1 and FelI1, does not concur with its redox potential. This indicates that the redox potential is not the only factor responsible for determining the nature of the electron-transfer process. Differences in the rates of electron transfer in the case of metal complexes, despite the fact that they have similar redox potentials, have also been reported ea~1ier.l~ The mechanism of electron transfer has been thought of as occurring by tunnelling through the potential barrier. Evidently the tunnelling process will depend on the height and width of the potential barrier.The width will also depend on the distance of closest approach of the electron substrate, which will evidently depend on the structure of the metal complex concerned. Thus CoIIIEDTA is known to have a definite octahedral c o n f i g u r a t i ~ n , ~ ~ ~ whereas the corresponding FelI1 and CuII analogues are known to have pentagonal-bipyramida115 and distorted-octahedral15 structures, respectively. These structural considerations indicate that the electron-transfer process should be more difficult in the former case than in the latter two complexes. G. A. Infante, E. J. Fendler and J. H. Fendler, Radiat. Res. Rev., 1973, 4, 301. S. N. Bhattacharyya and P. C . Mandal, Znt. J . Radiat. Biol., 1983, 43, 141. S. N. Bhattacharyya and P. C . Mandal, J. Chem. Soc., Faraday Trans. I , 1983, 79, 2613. C . L. Greenstock, J . Chem. Educ., 1981,58, 156 Handbook of Chemistry, ed. N. A. Lange (McGraw Hill, New York, 1961), p. 1212. Encyclopedia of Electrochemistry of the Elements, ed. A, J. Bard (Marcel Dekker, New York, 1975), vol. 111, p. 53. ' F. P. Dwyer, E. C . Gyarfas and D. P. Mellor, J. Phys. Chem., 1955, 59, 296.S. N. BHATTACHARYYA AND P. C. MANDAL 2575 M. Mori, M. Shibata, E. Kyuno and V. Okubo, Bull. Chem. SOC. Jpn, 1958, 31, 940. M. Anbar, and P. Neta, Int. J. Appl. Radiat. hot., 1967, 18, 493. lo S. N. Bhattacharyya and E. V. Srisankar, Radiat. Res., 1977, 71, 325. l1 J. Cadet and R. Teoule, J. Radiat. Res., 1977, 18, 93. l2 S. N. Bhattacharyya and P. C. Mandal, J. Chem. SOC., Faraday Trans. 1, 1984,80, 1205. l3 J. P. Keent, Radiat. Res., 1964, 22, 1. l4 E. J. Hart and M. Anbar, The Hydrated Electron (Wiley-Interscience, New York, 1970), p. 189. l5 F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry (Wiley-Eastern, New Delhi, 1969), (a) p. 148; (b) p. 849; (c) p. 894. (PAPER 4/41 5)