J . Phys. Chem. 1992, 96, 306-314
306
k, = 7.05 (f1.4) k2 = 6.8 (f1.4)
X X
cm3/molecule s cm3/molecule s
The value of k l can be compared with previous absolute measurements at 298 K and pressures below 10 Torr: 6.1 (*O.l);' 6.0 (f0.7);25.93 (10.45);3 5.5 (f0.6)4 X lo-" cm3/molecule s. Our rate constant is 20% larger than the average of these four measurements, well within the stated error limits, and is only 15% larger than the measurement1 with the smallest stated error limits that was carried out at the lowest pressure. Therefore, our value of kl agrees with all previous measurements to within experimental uncertainty, verifying that there is no pressure dependence to this reaction from to 700 Torr. Only one other absolute mea-
surement has been made of kz. At 298 K and 100 Torr of argon, Wine and Semmes'O obtain k2 = 8.04 (f0.57) X cm3/ molecule s, which again agrees with the value from our experiments to well within the error limits.
Acknowledgment. This work was carried out at Ford Motor Co. and was fully supported by the Chemical and Physical Sciences Laboratory of the Research Staff. E.S.also acknowledges partial support by the National Science Foundation. Registry NO. CI, 22537-15-1; C2H6, 74-84-0; C2HSC1, 75-00-3; Cl2, 7782-50-5; 02,7782-44-7;chloroethyl radical, 16520-13-1;2-chloroethyl radical, 16519-99-6. (10) Wine, P. H.; Semmes, D. H. J . Phys. Chem. 1983,87, 3572.
Perthiyl Radicals, Trisulfide Radical Ions, and Sulfate Formation. A Combined Photolysis and Radiolysis Study on Redox Processes with Organic Di- and Trisulfides Steven A. Everett: Christian Schoneich,t John H. Stewart,+and Klaus-Dieter Asmus*** Department of Applied Physical Sciences, University of Ulster, Shore Road, Newtownabbey Co. Antrim, BT 37 OQB, Northern Ireland, U.K., and Bereich S , Abteilung Strahlenchemie, Hahn- Meitner Institut Berlin, Posrfach 390128, 1000 Berlin 39, Germany (Received: September 3, 1991)
The formation of perthiyl and thiyl radicals has been studied in a combined photochemical and radiation chemical investigation of the di- and trisulfides of penicillamine and cysteine, together with glutathione disulfide, cystamine dihydrochloride, and dithiodiglycolicacid in aqueous solution. The highest RSS' yields are found from the trisulfides with little difference between PenSSSPen and CySSSCy in the case of photolysis, but with a (4-5):l ratio in favor in PenSS' over CySS' upon radiation chemically induced processes. Thiyl radicals are the other major species formed. The RSS'/RS' ratio changes significantly depending on the nature of the di- or trisulfide and the method of radical generation employed. Perthiyl radicals are found to be moderately good oxidants, weaker though than thiyl radicals, and readily react with molecular oxygen in an addition process. The following absolute rate constants have been measured: k(PenSS' + ascorbate) = (4.1 f 1.0) X lo6 M-' s-l and k(PenSS' + 4 ) = (5.1 1.0) X lo6 M-I PI.Probably the most interesting finding about perthiyl radicals is that inorganic Sod2is formed in their reaction with molecular oxygen. The mechanism of sulfate formation is suggested to proceed via the perthioperoxyl radical, RSSOO', structural rearrangement of this transient into the corresponding sulfonyl-type radical, RSSO;, addition of oxygen to the latter to yield a sulfonyl peroxyl, RSS0200', followed by a bimolecular radical-radical reaction of these peroxyl radicals (in analogy to the fate of peroxyl radicals in general) leading to RSS03'. Cleavage of SO3from RSSOj and hydrolysis would then result in SO,". Reduction of trisulfides yields (RSSSR)'- radical anions which decay into either RSS' + RS-(preferred pathway in case of PenSSSPen reduction; k(PenSSSPen + ea;) = 2.0 X 1O'O M-' s-l, k(PenSSSPen+ CO;-) = 2.3 X 10s M-I s-l), or RS' + RSS- (preferred pathway in case of CySSSCy reduction). Trisulfide radical anions could directly be identified in the case of the cysteine compound with (CySSSCy)'- absorbing at 425 nm and exhibiting a half-life of ca 4 ps. Radical cations (PenSSSPen)'+ are indicated in the oxidation of penicillamine trisulfide.
*
Introduction The disulfide bond provided by cystine plays a major role in the structure and function of many proteins.' Numerous studies have shown that organic disulfides are easily oxidized through free radicals like 'OH,2-4 reactive oxygen species like 035 and 102,63' and many other oxidizing radical^.*^^ Such reactions thus constitute potent ways of protein inactivation. Moreover, disulfides have been shown to suffer bond breakage by one-electron reductionlOJ1or photochemical p r o c e s s e ~ . While ~ ~ ~ ~splitting ~ of the sulfursulfur bond, with generation of RS' radicals, appears to be the major route of radical-induced disulfide destruction, carbon-sulfur bond rupture may also occur, particularly when tertiary carbon substituents are attached to the disulfide bridge. The latter would lead to the perthiyl radical, RSS', which constitutes the sulfur analogues of the well-known and chemically rather well-characterized peroxyl radicals, ROO'. I4 Whereas research has focused particularly on the reactivity of oxygen-centered radical species in biological environment, e.g., 'University of Ulster. Hahn-Meitner Institut Berlin.
*
with respect to the pathology of many d i s e a ~ e s , ' ~surprisingly J~ minor attention was paid for a long period to the biochemical role (1) Huxtable, R. J. Biochemistry of Sulphur; Plenum Press: New York, 1986. (2) Miickel, H.; BonifaEiE, M.; Asmus, K.-D. J . Phys. Chem. 1974,78,282. (3) BonifaEiE, M.; Schlfer, K.; Mockel, H.; Asmus, K.-D. J. Phys. Chem.
1975, 79, 1496. (4) Gilbert, B. C.; Laue, H. A. H.; Norman, R. 0. C.; Sealy, R. C. J . Chem. SOC.,Perkin Trans. 2 1975, 892. (5) Murray, R. W.; Smetana, R. D.; Block, E. Tetrahedron Lett. 1971,4, 299. (6) Murray, R. W.; Jindal, S . L. J. Org. Chem. 1972, 37, 3516. (7) Stevens, 8.; Perez, S. R.; Small, R. D. Photochem. Photobiol. 1974, 19, 315.
(8) BonifaEiE, M.; Asmus, K.-D. Inr. J . Radial. Biol. 1984, 46, 35. (9) BonifaEiC, M.; Asmus, K.-D. J. Phys. Chem. 1976, 80, 2426. (10) Hoffman, M. 2.; Hayon, E. J . Am. Chem. SOC.1972, 94, 7950. (1 1 ) Purdie, J. W.; Gillis, H. A.; Klassen, N. V. Can. J . Chem. 1973, 51, 3 132. (12) Morine, G. H.; Kuntz, R. R. Photochem. Photobiol. 1981, 33, 1. (13) Grant, D. W.; Stewart, J. H. Phorochem. Photobiol. 1984, 40, 285. (14) Sonntag, C. v. The Chemical Basis of Radiation Biology; Taylor 8c Francis: London, 1987. (15) Slater, T. F. Biochem. J . 1984, 222, 1.
0022-3654/92/2096-306%03.00/0 0 1992 American Chemical Society
Formation of Perthiyl and Thiyl Radicals of sulfur-centered radical species. This has changed, however, with the in situ discovery of RS' radicals in biological material"J8 and the establishment of a quite versatile and interesting general chemistry of RS*.19*20It is noteworthy, for example, that thiyl free radicals, RS', appear to be stronger oxidizing species than simple aliphatic peroxyl radicals, R00'.21-23 They exhibit, in fact, only a slightly lower reactivity toward many biologically available electron donors than their oxyanalogues, R0'.24 Biochemically, RS'-generating systems have even been found to initiate lipid p e r o x i d a t i ~ n . This ~ ~ ~ would ~ ~ indicate that sulfurcentered radicals play a more important role in free-radicalmediated tissue damage than believed previously. Very little is still known until today about the chemistry and biochemistry of perthiyl radicals, except that they apparently cannot abstract hydrogen from formate but may oxidize reduced flavins.27 It seems that RSS' are generally less reactive than RS'. Physically, perthiyl radicals are, however, reasonably well characterized through ESR s p e c t r o ~ c o p y ~ *and - ~ ~to some extent through a moderately strong optical absorption peaking around 374 nm,1227.33 The present study provides, in particular, information concerning the formation of cysteine- and penicillamine-derived perthiyl radicals (CySS' and PenSS'), generated from different di- and trisulfides by means of radiation chemical and photochemical methods. The important aspect of this study is the influence of molecular oxygen, and the formation of inorganic sulfate from perthiyl radicals. In connection with the redox chemistry of trisulfides it will be shown that trisulfide radical anions and cations must be considered as intermediates besides perthiyl and thiyl radicals. These trisulfide-derived species are not only interesting from the purely chemical point of view. It should be mentioned that various organic trisulfides are found, for example, in extracts from garlic and onion which are known for their capacity in healing or preventing of many pathological syndromes.34
Experimental Section L-Cystine (CySSCy), cystamine dihydrochloride (CyaSSCya.2HCl), and dithiodiglycolic acid (DDGA) were supplied (16) Basaga, H. Biochem. Cell. Biol. 1990, 68, 989. (17) Mason, R. P.; Ramakrishna Rao, D. N. In Sulphur-Centered Reactive Intermediates in Chemistry and Biology; Chatgilialoglu, C., Asmus, K.-D., Eds.; NATO-AS1 Series A, Life Sciences, Vol. 197; Plenum Press: New York, 1990; p 401. (18) Mason, R. P.; Maples, K. R. In Sulphur-Centered Reactive Intermediates in Chemistry and Biology; Chatgilialoglu, C., Asmus, K.-D., Eds.; NATO-AS1 Series A, Life Sciences, Vol. 197; Plenum Press: New York, 1990; p 429. (1 9) Asmus, K.-D. In Radioprotectors and Anticarcinogens; Nygaard, 0. F., Simic, M. G., Eds.; Academic Press: New York, 1984; p 23. (20) Asmus, K.-D. Methods Enzymol. 1990, 186, 168. (21) Forni, L. G.; Mijnig, J. Mora-Arellano, V. 0.; Willson, R. L. J. Chem. Soc., Perkin Trans. 2 1983, 961. (22) Forni, L. G.; Willson, R. L. Biochem. J . 1986, 240, 905. (23) Forni, L. G.; Willson, R. L. Biochem. J. 1986, 240, 897. (24) Erben-Russ, M.; Michel, C.; Bors, W.; Saran, M. J. Phys. Chem. 1987, 91, 2362. (25) Tieu, M.; Bucher, J. R.; Aust, S . D. Biochem. Biophys. Res. Commun. 1982, 107, 279. (26) Searle, A. J. F.; Willson, R. L. Biochem. J . 1983, 212, 549. (27) Elliot, A. J.; McEachern, R. J.; Armstrong, D. A. J. Phys. Chem. 1981, 85, 68. (28) Hadley, J. H. Jr.; Gordy, W. Proc. Natl. Acad. Sci. U.S.A. 1975, 72, 3486. (29) ESR and Endor Analysis: "Radiation Effects"; Box, H. C., Ed.; Academic Press: New York, 1977; p 154. (30) Banazzola, L.; Fackir, L.; LeRay, N.; Roncin, J. Radial. Res. 1984, 97, 462. (31) Franzi, R.; Geoffroy, M.; Bernadinelli, G. Mol. Phys. 1984, 52, 947. (32) Becker, D.; Swarts, S.;Campagne, M.; Sevilla, M. D. Int. J. Radial. Biol. 1988, 53, 767. (33) Burkley, T.J.; Hawari, J. A.; Losing, F. P.; Lusztyk, J.; Sutcliffe, R.; Griller, D. J . Org. Chem. 1985, 50, 4966. (34) Block, E. In Sulphur-Centered Reactive Intermediates in Chemistry and Biology; Chatgilialoglu, C., Asmus, K.-D., Eds.; NATO-AS1 Series A, Life Sciences, Vol. 197; Plenum Press, New York, 1990; p 283.
The Journal of Physical Chemistry, Vol. 96, No. 1, 1992 301 TABLE I: Yields of Initially Formed Perthiyl and Thiyl Free R.di~rts00 Flash Photolysis of NZ-SPtumted Aqueous S o l ~ t i ~p nH~ , 5.57, Containing 5 X lWSM of the Respective Di- or Trisulfide, and Quantum Yields of Sulfate Obtained on Steady-State Photolysis of Oxygen-Saturated, Aqueous Solutions Containing 3 X M of Mor Trisulfide at Various pH Qs0,Z-
compd
RSS'; pM RS'; pM pH 1 2 p H 5-6
PenSSSPen
13.6
cyssscy PenSSPen
11.3 8.5
cysscy CyaSSCya DDGA GSSG
4.3 10
0.003
50 X 1017photons. At higher doses the S042-vs dose plots show similar characteristics as in the absence of ascorbate and eventually level off at the same plateau values (curves 2 and 4 in Figures 2, A and B, respectively). The efficiency of various concentrations of ascorbate in the lowdose lag-phase region (ca. (0-50) X lOI7 photons) is displayed in the insert in Figure 2. Similar observations were made for the other disulfides and trisulfides (not shown here). Our experimental results indicate a correlation between the initial yield of perthiyl radicals and sulfate formation. This is particularly evident for PenSSSPen, CySSSCy, PenSSPen, CySSCy, and GSSG, i.e., polysulfides which contain both an amino and a carboxyl function, while such a relationship is not as pronounced for DDGA and CyaSSCya which lack either of these two functional groups, respectively.
+
TABLE Ik Quantum Yields of Sulfate Obtained on 254-nm Photolysis of Oxygen-Saturated Aqueous Solutions, pH 5.5, Containing Respective Disulfide Concentrations mncn of @Sop RSSR, 10-3 M PenSSPen CyaSSCya 0.5 0.006 0.004 1.o 0.008 0.006 3.0 0.011 0.008 5.0 0.013 0.009
As will be discussed later in the mechanism section, the production of sulfate requires the existence of perthiyl radicals and the presence of oxygen, and occurs via formation of oxidizing intermediates. An unambiguous proof for the need of perthiyl radicals will be given in the y-radiolysis section where sulfate formation could be observed in a system (formate/PenSSSPen) which exclusively yields PenSS' (see below). It can therefore be suspected that the particularly low initial yields of perthiyl radicals from CyaSSCya, DDGA, and GSSG (as measured in the flash experiments) refer, indeed, only to the early processes, and additional perthiyl species are generated from these compounds a t a later stage via as yet unidentified mechanisms. Since photolytic oxidation of sulfite to sulfate is a well-established procas, we had to check whether sulfate was a direct result from interaction of the perthiyl free radical with oxygen or whether it was produced via sulfite as an intermediate subsequently oxidized to sulfate. Investigation of this problem is possible by means of modem ion chromatography which allows fast and quantitative separation of both these anions.54 Photolysis of an air-saturated solution, p H 5.5, containing 5 X M sulfite leads, in fact, to a complete conversion of S 0 3 2 - to S042-. This process takes a period of about 5 min though, and after photolysis for only 1 min considerable amounts of sulfite (ca. 50%) are still left. Upon 1-min photolysis of di- and trisulfides, however, never any trace of sulfite could be detected. We can therefore conclude that sulfate is formed directly and not via a free sulfite intermediate. This conclusion is corroborated by the sulfate yields obtained on yirradiation of di- and trisulfides in the absence of photolyzing light (see section on y-radiolysis). Concentration Dependence. Photolytic production of sulfate is dependent on the employed concentration of the di- and trisulfides. This is exemplified by PenSSPen and CyaSSCya in Table 11. As can be seen, this effect is more pronounced at low than a t high solute concentrations. All comparative steady-state photolysis experiments on sulfate formation were, therefore, carried out in the high concentration range (typically 3 X M) of the respective di- or trisulfide. Influence of pH. Sulfate formation is, to some extent, also dependent on pH as displayed for three different pHs in the last three columns in Table I. While the difference in sulfate yield is not very large between the pH 5-6 and pH 10 (investigation of trisulfides at pH > 10 could not be performed because of increasing instability against hydrolysis above pH 7).42 The drastic reduction of the sulfate formation in the alkaline is complemented by the formation of at least two new, but as yet unidentified, products as becomes evident from the ion chromatogram (not shown here). (Sulfite, P.enS03H,and PenSOzH could be eliminated as possible candidates by comparison with authentic samples or analogy with the Cy-substituted sulfinic and sulfonic acids.) Since the initial photochemical formation of perthiyl radicals from PenSSPen is not dependent on pH the lower sulfate yield at high pH is likely to be due to a pH-dependent change or influence on the mechanism of sulfate formation. Such an influence could be, for example, the deprotonation of the amino group (pK 8-9). As sulfate appears to be formed via oxidizing intermediates (as shown in a later section) these species might, in fact, also oxidize the deprotonated amino group, with this mechanism competing against sulfate formation.
(53) Alfassi, 2.B.; Huie, R. E.; Neta, P.; Shoute, L. C. T. J . Phys. Chem. 1990, 94,8800.
(54) Deister, U.; Warneck, P. J . Phys. Chem. 1990, 94, 2191.
310 The Journal of Physical Chemistry, Vol. 96,No.
"1 E
Everett et al.
I, 1992 B
A
20
I
C 8 4
1-1
"0
200
100
0
100
200
300
10 l7 photons Figure 2. Plots of sulfate yields vs photon flux obtained on the 254-nm photolysis of 3 X M PenSSSPen, pH 5.5 under different conditions: (A) oxygen-saturated solution in the absence (curve 1) and presence of 2 X lo4 M ascorbate (curve 2); (B) air-saturated solution in the absence (curve 3) and presence of 2 X lo4 M ascorbate (curve 4). Insert C: Sulfate yield obtained in the low-dose region at various concentrations of ascorbate on photolysis of oxygen-saturated aqueous solutions, pH 5.5, containing 3 X IO-' M PenSSSPen, and (a) no, (b) 5 X and (d) 2 X (c) 1 X M ascorbate; x axis, 10'' photons; y axis, [S042-]/10-5M.
I
10 -
fu
8-
E
6-
c
twice the yield (curve b in Figure 3). In this system PenSS' radicals are generated via reactions 2 and 5-7.
'OH/H'
:c' I i COS'-
I
\
+ HC02-
20
W
40
+ PenSSSPen
+ COS'Pens- + PenSS' + C02
-+
HzO/H2
+ PenSS' + Penst PenSS-
380
340
420
460
500
540
hlnm Figure 3. Optical absorption spectra obtained on pulse radiolysis of (a) N,-saturated aqueous solution, pH 5.7, containing 1.0 M tert-butyl alcohol and 2 X lo4 M PenSSSPen; (b) N2-saturated aqueous solution, pH 5.6, containing 0.1 M NaHC0, and 2 X lo4 M PenSSSPen; and (c) N,-saturated aqueous solution, pH 5.6, containing 0.1 M NaHC0, and 2 X lo4 M PenSSPen. Insert d: Trace of optical absorption vs time obtained under experimental conditions as in (b); x axis, time in lod s; Gt [M-' cm-'1. y axis, optical absorption in
Pulse Radiolysis. One-Electron Reduction of PenSSSPen. Pulse irradiation or an N,-saturated aqueous solution, pH 5.7, M PenSSSPen containing 1.0 M tert-butyl alcohol and 2 X leads to the formation of a transient absorption with A, = 374 nm, fully developed immediately after the pulse, and shown in curve a of Figure 3. By comparison with the flash photolysis results12 [and an analogous transient observed in the one-electron reduction of bis(2-hydro~yethyl)trisulfide]~~ it is assigned to the PenSS' radical which is generated according to reactions 2-5 (k, = 2.0 X 1Olo M-I s-l). H,O
'OH/H'
-
ea,,-. H', 'OH,
+ CH,C(CHj),OH
(2)
+
+ 'CHZC(CH3)zOH PenSH + PenSS' Pens- + PenSS'
H2O/H2
-
+ PenSSSPen eaq- + PenSSSPen
H'
...
(7)
In analogy to the reduction of disulfides, reactions 5 and 7 can be expected to proceed via the trisulfide radical anion which, in principle, could decay via two different pathways (8a) and (8b).
60 80 PI
(PenSSSPen)'-
300
(6)
(3) (4)
(5)
Pulse irradiation of an N2-saturated aqueous, pH 5.6, solution of 2 X M PenSSSPen, and 0.1 M NaHCO, instead of tert-butyl alcohol leads to formation of a similar band but at almost
(8a)
+ Pens'
(8b) The absence of any shoulder a t 330 nm clearly indicates that reaction 8b is only of minor if any importance. Seemingly the overall thermodynamics (radical and anion stabilization energies) are in favor of pathway 8a. The optical measurements did not provide any direct evidence for (PenSSSPen)'- radical anions which, in analogy to disulfide radical anions, are expected to absorb in the blue range of the visible spectrum. The lifetime of (PenSSSPen)'- must therefore be < 1 ps under the experimental conditions. This may, however, be expected since the corresponding disulfide radical anion which exists in the (PenSSPen)'- s Pens'
+ Pens-
(9)
equilibrium is, in fact, also characterized through very short lifetimes ( T ~ < 1 ps) in the absence of excess thiolate.20 The instability o/trisulfides in the presence of thiols, particularly in basic solution, prevented any attempt to stabilize the trisulfide radical anion through the back reaction of equilibrium 8a (at least in aqueous environment). The PenSS" radicals are formed with the radiation chemical yield of Gt = 6280 and 11 000 M-' cm-' in the tert-butyl alcohol and formate containing systems, respectively. In the first system the trisulfide is directly reduced by the full yield of hydrated electrons (G = 2.8) and ca. 2/3 of the H' atoms Qi.e.,with G = 0.4). (The latter value is based on the assumption that H'reacts with trisulfides as fast as with disulfides and that this reaction has to compete with the H' + tert-butyl alcohol reaction).5s On this basis, Le., a total G = 3.2 for reduction, an extinction coefficient of 1960 M-' cm-l is calculated. This value is slightly higher than the photochemically derived t = 1700 M-l cm-1.12Additional perthiyl generation through a displacement reaction of the tertbutyl alcohol radicals with the trisulfide in our radiation chemical system would aCcOunt for this discrepancy. In the formate system PenSSSPen is reduced by eaq- and C0,'- with a yield corre-
,
( 5 5 ) Buxton, G . V.;Greenstock, C. L.; Helman, W. P.;Ross, A. B. J . Phys. Chem. Re5 Data. 1988, 17, 513.
The Journal of Physical Chemistry, Vol. 96, No. I, 1992 311
Formation of Perthiyl and Thiyl Radicals
and (Gt)374 = 2200 M-' cm-l, the yields are calculated to be G(CyS') = 5.0 and G(CySS') = 1.3 (taking tcys.,330= 320 M-I (&,47 and assuming that Q ~ .= ,1750 ~ M-l ~ cm-I). ~ Combining both G values we derive that CySSSCy is reduced to the respective products in the formate containing system with CG = 6.3, i.e., with similar yield as PenSSSPen. In contrast to the latter, however, the electron adduct decays to an extent of about only 20% via reaction 1l a into CySS' radicals while the majority of 80% yields CyS' via pathway 11b. Both reactions 1 l a and 11b
10 -
. 6'E V
8-
W
eaq-/C02'-
+ CySSSCy
(CySSSCy)'-
(10)
+ 0.2 CySS' s 0.8 CyS' + 0.8 CySS-
(lla)
+
(CySSSCy)'- s 0.2 CyS-
300
350
400
450
500
550
hlnm Figure 4. Optical absorption spectra obtained on pulse radiolysis of an N,-saturated, aqueous solution, pH 5.6, containing 0.1 NaHCO, and 2 X lo-" M CySSSCy (a) 5 ps and (b) 17.5 ps after application of a ca. 1.3-Gy electron pulse.
+
+
sponding to G(e,,H' 'OH) = 6.2. With this figure an extinction coefficient of 1770 M-' cm-' is calculated which is in much better agreement with the photochemically derived value. Considering all these aspects the currently best value for the extinction coefficient appears to be tpenSS.,374 = 1750 f 200 M-' cm-l. The buildup kinetics of the 374 nm transient in the formate containing system are displayed in Figure 3d. They clearly consist of two separable processes, the fast one being due to reaction 5 whereas the slower one is caused by reaction 7. The second-order rate constant k7 = 2.3 X lo8 M-' s-l (derived from plots of k = In 2/t1,2 for the slower process vs [PenSSSPen]) is in good agreement with a corresponding rate constant (5 X IO8 M-' s-I 1 reported for the C02'- induced reduction of bis( 2-hydroxyethyl) tri~ulfide.~~ The PenSS' radicals decay by pure second-order kinetics (presumably yielding the tetrasulfide) with a rate constant of 2k = (1.9 f 0.1) X IO9 M-I S-I , in very good agreement with a value of 1.4 X lo9 M-I s-I (ref 42) for the bimolecular decay of (hydroxyethy1)perthiyl radicals. For comparison, and in support of the assignment of the 374-nm band in the reduction of the trisulfide, we also investigated the reduction of penicillamine disulfide in N,-saturated aqueous soM PenSSPen and 0.1 M lution, pH 5.6, containing 2 X NaHC02. The transient absorption obtained in such system (curve c of Figure 3) shows only one band with a maximum a t 330 nm. This is unambiguously attributable to the Pens' species. From Gt = 3000 M-l cm-l and t330,pcnS. = 1200 M-l cm-' the Pens' yield is evaluated to G = 2.5 which is close to the full yield of hydrated electrons. This, in turn, indicates that C02'- does not react with PenSSPen to any significant extent. One-ElectronReduction of CySSSCy. The mechanisms outlined for PenSSSPen apply, in principle, also to the reduction of CySSSCy (and presumably any other trisulfide). Pulse radiolysis of N2-saturated aqueous solutions, pH 5.6, containing 3 X M CySSSCy alone (for reduction by ea,-) or the same concentration of CySSSCy plus 0.1 M N a H C 0 2 (for reduction by e,; and C02'-) reveals further information in support of the trisulfide radical anion, (CySSSCy)'-, as an intermediate. Thus a strongly absorbing transient is observed which exhibits a maximum at A, = 425 nm (curve a in Figure 4) and decays with a half-life of ca. 3.9 pus. The assignment seems reasonable in view of quite similar characteristics of the corresponding disulfide radical anions which also absorb in the same region and exhibit short lifetimes in the absence of stabilizing t h i ~ l a t e . ' ~ *The * ~ decay of (CySSSCy)'leads to the transient spectrum which, e.g., for the formate-containing system, is shown in curve b of Figure 4. It is characterized through two maxima peaking at 330 and 374 nm suggesting that it is a composite of CyS' and CySS' spectra. From the radiation (Ge)330= 1600 M-' cm-l chemical yields at the respective A,,
(1 1b) are reasonably formulated as equilibria. As in the penicillamine system it is not possible though to verify the respective back reactions for experimental reasons. The apparent extinction coefficient of t = 1600 M-' cm-I for (CySSSCy)'- (as calculated from the observed Gt i= 104 M-I an-') is much lower than that of corresponding disulfide radical anions 8900 M-I ~ m - 9 .However, ~~ due to the slow formation and (t short lifetime of the trisulfide radical anion its measurable yield will be lower than the maximum possible G = 6.2 used for the calculation. Furthermore, it cannot be excluded that C02'- does not only react via electron transfer (eq 10) but possibly also via displacement reactions:
RSSSR
+ C02'-
+
+ RSS' RSSC02- + RS'
RSC02-
(12a)
(12b) This additional mechanism seems feasible in view of the low S-S bond energies in trisulfide^.^^ A satisfactory explanation for the drastically different behavior of both trisulfides with respect to thiyl vs perthiyl radical formation cannot be given yet. It is interesting, however, to note that the perthiyl radical yields from CySSSCy are lower than those from PenSSSPen not only in radiolysis but also in photolysis. A full understanding would presumably require exact knowledge of all relevant thermodynamic data on the stability of radical and ionic products, as well as information about the electronic and conformational structures of the trisulfides and their radical anions in solution. The longer lifetime of (CySSSCy)'- relative to (PenSSSPen)'- may, however, be explained by the electronic structure of these species. As in the case of disulfide radical anions it can be expected that the incoming electron is accommodated in an antibonding, Le., bond weakening, u* orbital of the sulfur bridge.20 Furthermore, it is known that electron-donating substituents (increase of electron density in the antibonding orbital) and steric hindrance (less sulfursulfur porbital overlap) lead to an additional weakening of any such u/u* bond.57 The two a-positioned methyl substituents in the penicillamine moiety (as compared to hydrogen substituents in the cysteine moiety) thus provide the necessary rationale for the lower stability of the penicillamine-derived species. Comparing the di- and trisulfides, in general, it appears that trisulfides are somewhat easier to be reduced than disulfides. This becomes particularly evident for the C02'--induced processes which readily occur with PenSSSPen and CySSSCy but not with the respective disulfides. (Except for cyclic disulfides, like lipoic acid, the rate constants for the reduction of RSSR by Cot'- are