1337
N otizen
OH -Radical-Induced O xidation o f
Benzene in the Presence o f O xygen:
R ’+ 0 2^ R 0 2‘ Equilibria in
Aqueous Solution. A Pulse Radiolysis Study
Xian-M ing Pan and Clemens von Sonntag*
rapidly attained that it can be measured by pulse
radiolysis without too much interference of subse­
quent reactions.
In the present study OH radicals have been ge­
nerated radiolytically (reactions (1) and (2), for de­
tails see ref. [19]).
M ax-P lanck-Institut fü r Strahlenchem ie,
S tiftstraße 3 4 -3 6 , D-4330 M ülheim an d er R u h r,
G erm any
Z. N aturforsch. 45b, 1337-1 3 4 0 (1 9 9 0 );
received A pril 5, 1990
Benzene, H ydroxyl R adical,
H ydroxycyclohexadienyl R adical,
Peroxyl R adical, Pulse R adiolysis
T he reaction o f hydroxycyclohexadienyl ra d i­
cal w ith oxygen was studied by pulse radiolysis.
A t room tem perature the establishm ent o f the
equilibria between the hydroxycyclohexadienyl
radical ( 1), oxygen an d the hydroxycyclohexadienperoxyl radicals 5-hydroxy-1,3-cyclohexadiene- 6-peroxyl radical (3) an d 6-hydroxy-1,4-cyclohexadiene-4-peroxyl radical (4), w as observed.
From the kinetic analysis o f the decay o f radical 1
at 310 nm overall forw ard an d reverse rate c o n ­
stants o f k = 3.1 x 108 d m 3 m o l -1 s _1 an d k =
1.2* 104 s “1 were obtained. F ro m the analysis o f
the transient spectra an d the p ro d u cts o f pulse-ir­
radiated solutions it is concluded th a t in the e q u i­
librium radical 4 is m uch favoured over radical 3.
H 0 ionizing_^.QH
radiation
e;q + N20
h -5 h
o
n 2j H+ (1)
OH + N2 + OH-
(2)
The radiolytic yields of OH and H formed are
known (G ('O H ) = 5.8* 10-7 mol J~] and G(H') =
0.58 x 10“7 mol J~'). The OH radicals react with
benzene to give the hydroxycyclohexadienyl radical
(1) (reaction (3)). In the presence of oxygen, 1,
which exists in two mesomeric forms, is converted
into the peroxyl radicals 3 and 4 (reactions (5) and
(7)). The H atom plays only a minor role. It reacts
both with benzene (reaction (4), k 4 = 9.1 x 108 dm 3
m ol-1 s-1) and oxygen (reaction (10), k ]0 =
2.1 x IO10 dm 3 m ol-1 s“1) [16]. At a typical benzene
concentration o f 2x 10-3 mol dm -3 the contribu­
tion of radical 2 and hence radicals 5 and 6 is even
less (ca. 20% ) than suggested by the primary H
atom yield. In addition, cyclohexadienyl peroxyl
radicals reform benzene in 60% yield by H 0 2'-elim ination [20]. Thus, only radicals 3 and 4 (and
some H 0 27 0 2*-) have to be considered as inter­
mediates.
Introduction
There is a long-standing interest in the degrada­
tion of benzene by OH radicals in the presence of
oxygen [1-15]. In aqueous solution the OH addi­
tion to benzene is fast and irreversible at room tem ­
perature (reaction (3), k 3 = 7.8 x 109 dm 3 m o l'1 s“1)
[16]. This also usually holds for the reaction o f ox­
ygen with carbon-centered radicals [17], and it has
been believed also to be the case for the hydroxy­
cyclohexadienyl radical formed in reaction (3)
(reactions (5) and (7)) [9].
There is experimental evidence that the openchain analogs of the hydroxycyclohexadienyl radi­
cal such as the pentadienyl radicals formed as
intermediates in the autoxidation of polyunsatur­
ated fatty acids undergo reversible oxygen addi­
tion [18] (for a review see ref. [19]). Here we will
show that this also holds for the hydroxycyclo­
hexadienyl radical and that the equilibrium is so
* R eprint requests to Prof. D r. C. v. Sonntag.
Verlag der Zeitschrift für N aturforschung, D-7400 Tübingen
0932-0776/90/0900-1337/$ 01.00/0
5
H*
+
02
-------►
H 0 2'
6
(10)
Unauthenticated
Download Date | 6/17/17 8:26 AM
N otizen
1338
In this paper it will be shown that our present
view o f the 0 H /b en z en e/0 2 system has to be re­
vised: In oxygenated solutions the hydroxycyclohexadienyl radicals 1 are in equilibrium with their
corresponding peroxyl radicals 3 and 4 (reactions
(5)~(8)).
25
20
15
10
Experimental
Benzene (Merck) was used as supplied. Solu­
tions were made up in triply distilled or Millipore
Milli-Q filtered water presaturated with N 20 / 0 2
(4:1 v:v) (Messer Griesheim) prior to adding the
required am ount of benzene through a serum cap.
Different N 20 / 0 2 mixtures were prepared with the
help o f a Brooks gas mixer. Pulse radiolysis was
carried out using a Van de G raff 2.8 MeV electron
generator delivering 0 .4 - 1 /.is electron pulses of
3 -3 0 Gy. The pulse radiolysis setup as well as the
data processing procedures have been described
elsewhere [21, 22].
Results and Discussion
The hydroxycyclohexadienyl radical 1 is charac­
terized by a strong absorption at 310 nm. Our
value agrees well with that reported in the litera­
ture (e(310) = 3600 dm 3 m ol-1 cm “1) [23]. The un­
substituted cyclohexadienyl radical 2 also absorbs
at this wavelength (£(310 nm) = 4400 dm 3 mol
cm -1) [24], In the presence of oxygen 2 rapidly
decays yielding 5 and 6 (reaction (9); k 9 =
1 .2 x l0 9 dm 3 m ol-1 s“1) [20]. Because of very fast
subsequent unimolecular reactions of the peroxyl
radicals 5 and 6 the reverse of reaction (9) cannot
be m onitored [20],
When the hydroxycyclohexadienyl radical 1
reacts with oxygen its absorption at 310 nm rapid­
ly decays, but after this rapid decay a considerable
absorption at this wavelength persists (cf. the “pla­
teau” at around 3 0 -4 0 //s in Fig. 1). This residual
absorption at 310 nm which subsequently decays
only slowly has been attributed to be due to the ab ­
sorption o f the resulting peroxyl radicals 3 and 4
[9], It will be shown below that this is only partly
correct. In fact, a considerable part of the rem ain­
ing absorption at 310 nm is due to hydroxycyclo­
hexadienyl radicals 1 in equilibrium with its corre­
sponding peroxyl radicals, the reactions (5) and (7)
being reversible (reactions (6) and (8)).
In cyclohexadienyl radicals the highest spin
density is at their central position (spin density
0.51) and considerably less at their terminal posi­
tions (spin density 0.35) [25]. The dimer distribu­
tion from the combination of cyclohexadienyl rad­
5
0
260
280
300
320
340
------- X / n m ------- ►
Fig. 1. UV spectra o f p ulse-irradiated N :0 -s a tu ra te d
2 //s after the pulse (...; right scale) and N 2Ö / 0 2-satu rated aq u eo u s solu tio n o f benzene (left scale): ( • ) 10 /is,
( x ) 20 ps, ( A ) 30 jus, (□ ) 40 ps and (O ) 170 /us after the
pulse.
icals 2 shows a 2:1 preference of the reaction from
their central vs. their terminal position(s) [24],
There is strong evidence that oxygen is not quite as
selective and adds to 2 with about equal probabili­
ty at the central position and the two terminal po­
sitions [20]. Hence it can be expected that also in
the case of 1 reactions (5) and (7) will be about
equally fast. However, the equilibrium concentra­
tions o f 3 and 4 are not only determined by the for­
ward reactions (5) and (7) but also by the reverse
reactions (6) and (8). There is strong evidence that
in the equilibrium 4 is much favoured over 3.
Peroxyl radical 3 has a 1,3-cyclohexadienyl
structure while 4 has a 1,4-cyclohexadienyl struc­
ture. 1,3-Cyclohexadienes have high extinction
coefficients at 260 nm (e ~ 104 dm 3 m ol-1 cm -1),
but 1,4-cyclohexadienes do not absorb at this
wavelength. Assuming that peroxyl radical 4 has
an extinction coefficient of around 1000 dm 3 m ol-1
cm “1 at 260 nm (as many peroxyl radicals have) it
is calculated from the data shown in Fig. 1 that
peroxyl radical 3 can only contribute a few per­
cent. Supporting evidence that in the equilibrium
mixture of 1, 3 and 4 peroxyl radical 3 can only be
a m inor com ponent comes from a product study of
pulse-irradiated solutions [20].
It can be shown that for the simple case, i.e. that
there is only one peroxyl radical (R' + 0 2 R 0 2')
the kinetics o f the decay of R' into R 0 2' can be de­
scribed by k obsd = £fonvard[ 0 2] + £reverse. In such a
case a plot o f k obsd vs. [0 2] will yield a straight line
whose slope equals A'forward and an intercept which
shows the value of /rreverse. The present system is
Unauthenticated
Download Date | 6/17/17 8:26 AM
N otizen
1339
more complex, but since peroxyl radical 4 predom­
inates, this form ulation also holds for the present
system.
In Fig. 2 the observed first-order rate constant
of the decay o f the 310 nm absorption is plotted as
a function of the oxygen concentration. A straight
line with a slope of k = 3.1 x 108 dm 3 m ol-1 s“1 is
obtained. The pronounced intercept indicates that
the reverse reactions (6) and (8) occur with an
overall rate constant of k = 1.2><104 s_1. From
these values an equilibrium constant of K =
2.6 x 104 dm 3m ol“ 1is calculated.
cm “1. When the right e(4) is chosen a plot of A' vs.
[0 2] will yield a straight line, when e(4) is chosen
too high the curve will bend upward, and when
chosen too low, downward. As can be seen from
Fig. 3 a good fit is obtained when e(4) = 500 dm 3
m ol“1 cm “ 1 is assumed. With this value K =
3.2 x 104 dm 3 m ol“1 is calculated. This value agrees
well with the one calculated from the data shown
in Fig. 2.
I
-------
[ 0 2l / 10"
mol dm -3
Fig. 3. Plots o f A ' vs. [ 0 2]. ( x ) e(4) = 650, ( A ) e(4) = 500
and (O ) e(4) = 400 d m 3 m o l -1 c m ' 1(see text).
[ 0 21 / 10~5 mol d m '3
Fig. 2. O bserved first-order rate co n stan t o f the decay o f
the 310 nm a b so rp tio n o f pulse-irradiated N 20 / 0 2sa tu ra te d aqueous benzene solution as a function o f
oxygen co n cen tratio n .
This equilibrium constant can be obtained by an
additional approach. The equilibrium constant as
defined by K = [4]eq/[l]eq[0 2j can be reformulated
as K = ([l]0- [ l ] eq)/[l]eq[0 2] where the subscript
“eq” and “o ” indicate “equilibrium'’ and “initial” ,
respectively. The initial absorbance is A0 =
e (l)x [l]0 (e(l) = 3600 dm 3 mol“1 cm “1, see above).
The absorbance in the equilibrium is given by
A = £(1) x [l]eq + «(4) x [4] = £(l)[l]eq + £(4) x
([IJo - [I]«,)- From these basic equations the fol­
lowing equation is derived
1 + K x [O,] =
2
f (4)} * ft*
e ( l ) x Aeq- e ( 4 ) x A 0
= A'
(a)
In this equation there are two unknowns, K and
e(4). From Fig. 1 it may be estimated that e(4) may
range somewhere between 400 and 800 dm3 m ol“ 1
Additional remarks. The hydroxycyclohexadienyl radical presents us with the rather fortunate op­
portunity to study the reversibility of the oxygen
addition reaction, because the equilibrium is estab­
lished at times when the subsequent reactions, bimolecular decay, HOV-elimination and intram ole­
cular cyclization of the peroxyl radical and fixa­
tion of this endoperoxyl radical by oxygen do not
yet play a decisive role (details o f these reactions
will be published elsewhere). In the unsubstituted
cyclohexadienyl radical these unimolecular proc­
esses are so fast (e.g. the H 0 2‘-elimination is
> 8 x l 0 5 s_1) that it was not possible to observe
such an equilibrium. In the case o f the hydroxycyclohexadienyl radicals derived from benzoic acid
the equilibrium appears to ly much more on the
side o f the biallylic radical than in the present sys­
tem and with hydroxycyclohexadienyl radicals de­
rived from some other substituted benzenes (e.g.
nitrobenzene and benzonitrile) a reaction with
oxygen is barely observable on the pulse radiolysis
time scale [26].
Unauthenticated
Download Date | 6/17/17 8:26 AM
N otizen
1340
[1] M. D aniels, G . Seholes, and J. W eiss, J. C hem . Soc.
1956, 832.
[2] E. V. B arelko, L. I. K artash ev a, P. D. N ovikov, and
M. A. P ro sk u rn in , Proc. 1st. A ll-U nion C onf. R adiat. Chem . M oscow (1957); N ew Y ork: C o n su lt­
an ts B ureau (1959), p. 81.
[3] I. L oeff and G . Stein, J. C hem . Soc. 1963, 2623.
[4] L. I. K artash ev a an d A. K. Pikaev, Russ. J. Phys.
C hem . 41, 1534,(1967).
[5] T. K. K. S rinivasan, I. B alak rish n an , and M . P.
R eddy, J. Phys. C hem . 73, 2071 (1969).
[6] I. B alakrish n an an d M. P. R eddy, J. Phys. C hem .
74,850(19 7 0 ).
[7] I. B alakrishnan a n d M . P. R eddy, J. Phys. C hem .
76,1273(1972).
[8] L. I. K artash ev a an d A. K. P ikaev, H igh Energy
Chem . 6 , 76(1972).
[9] G . R am an an , J. In d ian C hem . Soc. 53, 957 (1976).
[10] L. C astle, J. R. L indsay Sm ith, a n d G . V. B uxton, J.
M ol. C atal. 7, 235 (1980).
[11] N. N a rita and T. T ezuka, J. Am. C hem . Soc. 104,
7316(1982).
[12] G . C ohen and S. E. O fodile, in G . C ohen and R. A.
G reenw ald (eds): Oxy R adicals an d T heir Scavenge
System s, p. 74, N ew Y ork: Elsevier, Vol. I, M o lecu ­
lar A spects (1983).
[13] L. L atrian o , A. Z accaria, B. D. G oldstein, and
G . W itz, J. Free R adicals Biol. M ed. 1, 363 (1985).
[14] S. D as, O. J. M ieden, X .-M . P an, M . R epas, M . N.
S chuchm ann, H .-P. S chuchm ann, C. von S on n tag ,
an d H. Z egota, in M . G . Simic, K. A. T ay lo r, J. F.
W ard , and C. von S onntag (eds): O xygen R adicals
in Biology and M edicine, Vol. 49, p. 55, N ew Y ork:
Plenum , Basic Life Sciences.
[15] M. A l-Sheikhly and M. G . Simic, in C. R ice-E vans
an d B. Halliwell (eds): Free R adicals: M eth o d o lo g y
an d C oncepts, p. 481, Richelieu Press, L o n d o n
(1988).
[16] G . V. Buxton, C. L. G reenstock, W. P. H elm an , and
A. B. Ross, J. Phys. C hem . Ref. D a ta 17, 513 (1988).
[17] G .E . A dam s and R. L. W illson, T rans. F a ra d a y
Soc. 65,2981 (1969).
[18] N. A. P orter, L. S. L ehm an, B. A. W eber, an d K. J.
Sm ith, J. Am. Chem . Soc. 103, 6447 (1981).
[19] C. von S onntag, T he Chem ical Basis o f R ad iatio n
Biology, T aylor an d Francis, L ondon (1987).
[20] X .-M . Pan and C. von S onntag, to be published.
[21] S. Steenken, J. Buschek, and R. A. M cC lellan d , J.
Am . Chem . Soc. 108, 2808 (1986).
[22] A. J. S. C. Vieira an d S. Steenken, J. A m . C hem .
Soc. 109, 7441 (1987).
[23] L. M . D orfm an, I. A. T aub, and R. E. B ühler, J.
C hem . Phys. 36, 3051 (1962).
[24] X .-M . Pan, E. B astian, an d C. von S on n tag , Z. N atu rfo rsch . 43b, 1201 (1988).
[25] R. W. Fessenden an d R. H. Schuler, J. C hem . Phys.
39,2147(1963).
[26] X .-M . Pan and C. von S onntag, unpublished results.
N ach d ru ck - auch auszugsw eise - n u r m it schriftlicher G enehm igung des Verlages g estattet
Satz und D ruck: A llgäuer Z eitungsverlag G m b H , K em pten
Unauthenticated
Download Date | 6/17/17 8:26 AM