Relative rate constants for reactions of hydroxyl radicals from the

reaction of Fe(I1) or Ti(II1) with H2O2

W. A. ARMSTRONG
Defence Research Establishment Ottawa, Defence Research Board, Ottawa 4, Canada
Received March 6 , 1969

In an attempt to clarify the reactions of ferrous sulfate and titanous sulfate with hydrogen peroxide,
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 195.37.186.217 on 12/16/19

a novel method has been developed to determine the relative rate constants for reactions of the oxidizing
species generated in these systems. These species react with hydrogen peroxide to give perhydroxyl
radicals which combine with titanium(1V) ions to form the relatively stable TiO0.3+ radical. This
radical gives a strong electron spin resonance signal and the competition between hydrogen peroxide
and a scavenger for the oxidizing species can be followed by measuring the amplitude of this signal
in the presence of various concentrations of scavenger. The relative rate constants calculated in this
way for both the Fe(I1)-H202 and Ti(II1)-H202 systems at p H 1.0 agree with those reported for the
reactions of hydroxyl radicals in y-irradiated thymine solutions. This supports the view that hydroxyl
radicals are formed in these cases.
Under conditions of acidity favoring the hydrolysis of Fe(I1) to FeOH+ ions, hydroxyl radical
scavengers do not compete with hydrogen peroxide for the precursors of the TiO0.3+ radical. It is
suggested that the FeOH+ ions react with hydrogen- - -peroxide to -give a different oxidizing- species.
-
possibly the ferry1 ion.
Scavengers investigated were thymine, methanol, ethanol, formic acid, acetic acid, chloride ion, and
several amino acids.
Canadian Journal of Chemistry, 47, 3737 (1969)

+ scavenger
For personal use only.

Introduction [31 .OH -> X.

There has been much controversy concerning
the reaction of hydrogen peroxide with ferrous
or titanous ions in acidic aqueous solutions (1-7).
It is generally assumed that hydroxyl radicals are
formed in both reactions but there is some
evidence to suggest that the OH radicals are
complexed with titanium in the Ti(II1)-H202
system and that reaction with a scavenger takes
place in a [Ti(H,O),(OH)(s~avenger)]~+ complex
(4). Fenton's reagent has been characterized as
nucleophilic and Ti(II1)-H202 as electrophilic
(2)
The present investigation was designed to
measure the relative rate constants for the
reactions of a number of solutes with the oxidizing
species generated in each of these systems. These
constants could then be compared with the
reported absolute values obtained by radiolysis
experiments (8, 9). Reasonable agreement would
indicate that in each case free OH radicals are
being scavenged. The method is based on the
following reaction scheme
[I] Fe(I1) or Ti(II1) + H 2 0 2 ->
Fe(II1) or Ti(1V) + OH- + .OH
[21 .OH + H202 + H 2 0 + H02.
lIssued as D.R.E.O. Report No. 590.
[4] HOz. + Ti(IV) + Ti00.3+ + H+
The T i 0 0 . 3 + radical has a lifetime of several
seconds (4) and gives a single electron spin
resonance (e.s.r.) line with g = 2.013 (5). A
second paramagnetic species with g = 2.012 is
also observed but its identity is in doubt. Fischer
(5) suggests it is T i 0 0 T i 0 0 . 5 + , formed by the
reaction of TiO0.3+ with hydrogen peroxide and
titanium(1V) ions, but Takakura and Ranby (7)
favor the idea of a complex of OH and Ti(1V)
ions. However, Arthur and co-workers (10)
have observed this species (as well as the
T i 0 0 . 3 + ) upon the mixing of a Ce(1V) solution
with a solution of H 2 0 2 containing Ti(1V) ions
and it is unlikely that hydroxyl radicals are
produced in this reaction. Under the present
experimental conditions the concentration of this
second paramagnetic species is small and
decreases upon the addition of a scavenger in the
same manner as the TiO0.3+ (5). It has, therefore,
been ignored. As the lifetimes of most organic
radicals are much shorter than the lifetime of the
TiO0.3+ radical, the scavenger radicals formed
in reaction [3] are not observed.
The e.s.r. spectra observed with the Fe(I1)-
H202,Ti(IV) and Ti(II1)-H202,Ti(IV) systems
are shown in Fig. 1. For comparison, the spectrum
resulting from the illumination of a stream of
 CANADIAN JOURNAL O F CHEMISTRY. VOL. 47, 1969

should give a straight line with slope = k3/k2.

(C )
This approach is similar to that used in the pulse
radiolysis experiments where the reaction of O H
with I- or CNS- was used in the determination
of rate constants (1 1-13).

I Experimental
All solutions were prepared immediately before use
with triply distilled waler which had been saturated with
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 195.37.186.217 on 12/16/19

nitrogen. "Low manganese" ferrous sulfate (B and A

Laboratories) and "stabilizer-free" 30% hydrogen
peroxide (Fisher) were used in the reactant solutions.
Both 15 % titanous sulfate solution (British Drug Houses,
Ltd.) and material prepared in this laboratory by the
reduction of titanium basic sulfate in sulfuric acid
solution with zinc amalgam (14) were employed.
Hydrogen peroxide solutions were standardized by the
arsenious oxide method (15). The concentration of
titanous sulfate solutions was determined by titration of
an aliquot with standardized hydrogen peroxide to a
colorless end-point. It was assumed that 1 mole of
hydrogen peroxide oxidized 2 moles of titanous sulfate
(6). Titanium(1V) sulfate solutions of known concentra-
tion were prepared by similar titrations. The acidity of
the ferrous and titanous solutions was adjusted with
For personal use only.

sulfuric acid. The H20Z,Ti(IV) solutions for the second

stream were prepared by the addition of the desired
volume of a 0.4 M titanium(1V) sulfate solution in 2 M
sulfuric acid to the aqueous hydrogen peroxide solution.
FIG.1. The e.s.r. spectrum of the T i 0 D 3 + radical in he pH was then adjusted by the addition of either
O.l HzS04. (a) m M Fe(ll) - m M H2°2> sulfuric acid, sodium non no hydrogen phosphate, sodium
30 n i M Ti!IV), (b) 2.0 m M Ti(II1) - 80 m M H 2 0 2 ,
30 nlM Ti(IV), (c) ultraviolet light, nlM Hz02, tetraborate, or sodium hydroxide. At pH > 2.2 titanium
. m M- TifrV).
30. .
- -- . . ,.
- 1-
hydroxide precipitated.
The e.s.r. spectrometer and flow system have been
described elsewhere (16). T o permit the accurate adjust-
acidic H,O,,Ti(IV) solution with light of 2537 A ment of flow rates, the system was modified to include a
(a system which undoubtedly involves the flow meter and needle valve in each stream. Normally
formation of hydroxyl radicals) is included in the the flow of each stream was set at 115 ml/min and the
same figure. Apparently the same radicals are radicals were observed about 0.3 s after the mixing of the
reactants. Initially, a solution of ferrous or titanous
formed in each system. sulfate was mixed with a solution of the same acidity
If reactions [2] and [3] are the oi~lysignificant containing the desired amounts of hydrogen peroxide and
reactions involving O H radicals and all the HO, titaniuni(1V) sulfate. While these solutions' flowed
radicals react according to [4], then at a given through the mixing chamber and e.s.r. cell, the magnetic
field from 5 G below the value for resonance to 5 G above
time of observation and in the absence of a this value was scanned three times. A0 was the average
scavenger, [ T i 0 0 . 3 + ] = [OH], the total amount amplitude of this spectrum. After draining the flow
of hydroxyl radicals generated. In the presence system the process was repeated but with a known amount
of a scavenger of scavenger in the Ti(1V)-HZ02 solution. Values of A
were determined in this way for increasing concentrations
of scavenger until A z -kAO. Spectrometer noise pre-
vented the accurate determination of lower values of A .
T o check the stability of the spectrometer, A0 was
If A0 and A are the amplitudes of the T i 0 0 . , + redetermined at the end of each run but for a given set of
e.s.r. line in the absence and presence of a reaction conditions did not vary significantly over a
period of weeks.
scavenger
Results
A0 Ic, . [scavenger]
= 1 +-
--
The Systems at pH 1.0
A k2 [Hz021
In Fig. 2a are plotted values of AO/A vs.
Thus a plot o f ,AO/A vs. [scavenger]/[H,O,] [formic acid]/[H202] for the reaction of 1.2 mM
 ARMSTRONG: RELATIVE RATE CONSTANTS FOR REACTIONS OF HYDROXYL RADICALS 3739

time as shown in Fig. 2b. In this instance ethanol

was the scavenger and reaction times of 0.021
to 0.065 s were used. Similar linear plots for
methanol, acetic acid, cystine, and thymine are
given in Fig. 3. Repeated experiments indicated
that the measured slope of any plot is accurate
to about $- 15%.
The slope of each plot should equal k3/k2,
where k, and k2 are the rate constants for the
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 195.37.186.217 on 12/16/19

reactions of hydroxyl radicals with the scavenger

and hydrogen peroxide, respectively. However,
the literature values of k, range from 1.2 to 4.5
x lo7 M - I s-I (9). Although less than 2% of
the hydrogen peroxide is expended by reaction
[I], this does introduce some doubt as to the
08
[Formic Acid1 / [Hz021
1.2
exact concentration of peroxide which competes
for the OH radicals. In view of these factors it
FIG. 2. The effect of reaction conditions on the was decided to compare all of the results with
determination of relative rate constants at DH 1.0.
(a) 1.2 m M Fe(I1): 0, 55 m M H ~ o ~33, m M 'Ti(1~); those obtained with thymine as the scavenger.
a, 71 rnM HZO2, 25 m M Ti(1V); Q, 71 m M HzOz, The rate constant for the reaction of OH with
33 m M Ti(1V); 0 , 71 m M H20,, 49 m M Ti(1V); thymine has been determined accurately to be
+, 107 m M H 2 0 2 , 33 m M TiJIV). (b) 1.2 m M Fe(I1)
plus 60 m M H,O,, 31 m M Tl(1V); reactlon tlme: a, 4.3 $- 1 x 10' M-' s-I and has been used in
0.034 s ; a,0.065 s.
For personal use only.

0.021 s; 0, the determination of the relative reaction rates

of OH radicals with many scavengers (8).
ferrous sulfate with various concentrations of Assuming this value and employing the measured
hydrogen peroxide and titanium(1V) sulfate. All slope of the thymine plot (72.4), k, = 5.9 x lo7,
solutions described in this section were 0.1 M in a value somewhat higher than those quoted
sulfuric acid. Changing the concentration of above. Results for k3/k2 for a variety of scaven-
hydrogen peroxide from 55 to 107 mM and that gers along with rate constants based on the given
of titanium(1V) from 25 to 49 mM did not affect value for thymine are listed in Table I. For
the slope of this straight line. The results are comparison, the rate constants published by
independent of flow rate and hence reaction Scholes and Willson (8) and derived from studies
of the radiolysis of aqueous thymine solutions are
also included. Considering the magnitude of the
experimental errors, the agreement is good.
AO/Avs. [scavenger]/ [H20,] plots for 1.9 mM
Ti(II1) plus 80 m M H 2 0 2 , 38 mM Ti(IV), with
formic acid and ethanol as the scavengers are
given in Fig. 4. For formic acid, lc,/k, = 1.98,
in excellent agreement with the value obtained
with the Fe(I1)-H,O, system. Ethanol, however,
did not give a straight line plot. As the concentra-
tion of the ethanol was increased, the concentra-
tion of T i 0 0 . , + decreased to a greater extent
than predicted by reactions [2]-[4]. Reaction
of the ethanol radicals formed in [3] with the
T i 0 0 . 3 radicals could explain these results
+

FIG.3. Competition of scavengers with hydrogen

peroxide for OH radicals from Fenton's reagent: 1.2 m M
Fe(I1) plus 6 0 m M H2O2, 33 m M Ti(IV), scavengers
(pH = 1.0): 0, methanol; 9, acetic acid (multiply
abscissa by 100); a, I-cystine (n~ultiplyabscissa by 0.1);
0 , thymine (multiply abscissa by 0.25).
The importance of this reaction would depend
upon the concentration of T i 0 0 . , + and the
reactivity of X.. Under the conditions described
 CANADIAN JOURNAL O F CHEMISTRY. VOL. 47, 1969

TABLE I
Rate constants for OH radical reaction in solutions of pH I

Radiolysis of
Fe(I1)-H202 Ti(II1)-HZOz thymine?
+
k(OH scavenge;)* k(0H f scavenger)* k(OH f scavenger)
Scavenger k,/k, M - ' s - ' x 10- k3/k2 M-I S-I x lo-g M-' s-' x lo-g
Thymine 72.4
Methanol 15.0
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 195.37.186.217 on 12/16/19

Ethanol 28.9
Formic acid 1.95
Acetic acid 0.27
cr-Alanine 0.59
1-Cysteine 112
1-Cystine 130
DL-Penicillamine 79.2
Penicillamine disulfide 110
2-Mercaptoethylarnine 251
2,2'-Dithio,bis(ethylamine) 186
Glutathione (reduced) 186
Glutathione (oxidized) 143
'Based on k ( 0 H ithymine) = 4.3 x loq M-Is-'.
?Taken from Scholes and Willson (8).

above, the initial concentration of Ti00.3', as and cysteine did not give linear plots (Fig. 4),
determined by the magnitude of A', was more possibly because the CyS. radical (16) reacts
For personal use only.

than 100 times as great as that observed with
Fenton's reagent. When reactant concentrations
were reduced to 0.8 m M Ti(II1) plus 60 m M
H 2 0 2 , 3.3 m M Ti(IV), A0 was reduced by a
factor of 20 and the addition of ethanol resulted
in the straight line shown in Fig. 4. Thymine,
methanol, and acetic acid also gave linear plots.
The values of k,/k, are listed in Table I. Cystine
more rapidly with Ti00.3' than do the other
scavenger radicals. Attempts to decrease the
concentration of T i 0 0 . 3 + to the level attained
with the Fe(I1)-H202 system were unsuccessful,
as hydrolysis of the Ti(II1) prevented any
further decrease in its concentration.
The results of the addition of chloride ion to
both the Fe(I1)-H202 and the Ti(II1)-H202
reagents are shown in Fig. 5. The initial portion
of the plot is linear with a slope of 9.5 which
corresponds to

FIG. 4. Competition of scavengers with hydrogen

peroxide for OH radicals from Ti(II1)-Hz02: 1.9 m M
Ti(II1) plus 8 0 m M H z 0 2 , 38 m M Ti(IV), scavengers
(pH = 1.0): 0, formic acid (multiply abscissa by 10);
0 , ethanol. 0.8 m M Ti(II1) plus 60 m M HZOz,3.3 m M FIG. 5. Effect of chloride ions at pH 1.0: 9, 1.2
Ti(IV), scavenger: 8, ethanol; 8, cystine (multiply m M Fe(I1) plus 60 m M HZOZ,33 m M Ti(1V); 0, 0.8
abscissa by 0.1); 0, cysteine (multiply abscissa by 0.1). mM Ti(II1) plus 80 m M HzOZ,3.3 rnM Ti(1V).
 ARMSTRONG: RELATIVE RATE CONSTANTS I?OR REACTIONS O F HYDROXYL RADICALS
The reported values for this constant are 4.0 x
lo8 at pH 1 and 4.0 x lo9 at pH 0 (9). As the
chloride concentration was increased the con-
centration of T i 0 0 . 3 + decreased to a lesser
extent than predicted by reactions [2]-[4]. The
excess chloride ions may complex with the Fe(I1)
or Ti(II1) and be unavailable for reaction with
the hydroxyl radicals. An alternative explanation
is that the C1 atoms or C1,- radicals formed by
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 195.37.186.217 on 12/16/19
the reaction of hydroxyl radicals with chloride
ions (17, 18) react with hydrogen peroxide to
give perhydroxyl radicals which, by way of [4],
form T i 0 0 . 3 + . The yield of T i 0 0 . 3 + would
then become independent of the chloride ion
concentration. Attempts to observe the e.s.r.
spectrum of the C1,- radical were unsuccessful.
The other reactions which might interfere
with the proposed mechanism are
[6] Fe(I1) or Ti(II1) + OH +
Fe(II1) or Ti(1V) + OH-
[71 HO2. + scavenger -+ Y.
For personal use only.
[8] T i 0 0 . 3 + + scavenger -> Y.
Reaction [6] was minimized by keeping the
concentration of H 2 0 2 at least 50 times as great
as that of the ferrous or titanous ion. To investi-
gate reaction [7], hydroperoxyl radicals were
generated in the flow system by the reaction of
ceric ammonium sulfate with hydrogen peroxide.
Mixing 0.08 mM ceric ammonium sulfate in
100 m M sulfuric acid with a solution of 16.5 m M
Ti(1V) and 80 m M H 2 0 2 in 100 m M H2S04
resulted in the formation of an amount of
Ti00a3+ comparable with that formed in the
Ti(II1)-H202 reagent described above. Addition
to the Ti(1V)-H202 solution of a concentration
of scavenger 4 times greater than the maximum
amount used in the determination of relatiwrate
constants did not significantly decrease the
concentration of TiOO-3+.The first 8 scavengers
listed in Table I were investigated and the
maximum decrease observed was 3%.
A simple three-stream mixer constructed from
1 mm bore Pyrex tubing was used to investigate
reaction [S]. It was arranged so that the third
stream entered 0.15 s after the mixing of streams
1 and 2 and, therefore, after the completion of the
Ti(II1)-H202 reaction (-19). With 3 mM Ti(II1) -
100 m M H2S04 in stream 1, 16.5 m M Ti(1V) -
40 m M H 2 0 2- 100 m M H2S04in stream 2, and
lOOmM H,S04 added through stream 3, a
3741
concentration of Ti00a3+ radicals 10 times
greater than that observed with the Ti(II1)-H,02
system was generated. The addition of the same
scavengers to the third stream in the same
amounts as described in the preceding paragraph
did not appreciably affect the concentration of
radicals. The maximum decrease observed was
<4%. The low reactivity of TiO0.3+ radicals
with methanol has been demonstrated previously
(19, 20).
Under the present experimental conditions,
the most likely path by which the T i 0 0 . ~ +
radicals disappear is by the relatively slow
process of dimerization (19). Increasing the time
between the mixing of the reactants and the
observation of the radicals from 0.02 to 0.07 s
caused a decrease in the measured radical
concentration of less than 10%. Only when the
concentration of Ti0O.j' was high and the
reactivity of the scavenger radical X. great did the
addition of a scavenger have any apparent affect
on the dimerization reaction.
The EfSect of Increasing pH
The results obtained at pH > 1.0 are sum-
marized in Fig. 6. The slope of the plot of AO/A
vs. [ethanol]/[H202]at pH 1.2 is 30.0 for both
FIG. 6. Competition of scavengers with hydrogen
peroxide for oxidizing species at higher pH. (a) 0.6 m M
Fe(I1) plus 80 m M H z 0 2 , Ti(IV), scavenger. (i) 8.3 m M
Ti(IV), ethanol: 0, pH 1.2; Q, pH 1.6; @, pH of Fe(I1)
3.8, pH of H 2 0 2 ,Ti(IV), EtOH 2.2. (ii) 41 m M Ti(IV),
ethanol: 9, pH of Fe(I1) 3.8, pH of HzO1, Ti(IV),
EtOH 2.2. (iii) 8.3 m M Ti(IV), thymine: Q, pH of Fe(I1)
3.8, pH of H z 0 2 ,Ti(IV), thymine 2.2 (multiply abscissa
by 0.1). (b) 0.8 m M Ti(II1) plus 80 m M H z 0 2 , 8.3 m M
Ti(IV), scavenger. (i) Ethanol: 0, pH 1,2; D, pH 2.2.
(ii) Thymine: a, pH 2.2 (multiply ordlnate by 0.1).
(iii) 0.3 m M Fe(III), ethanol : pH 2.2.
 3742 CANADIAN JOURNAL O F CHEMISTRY. VOL. 47, 1969

the ferrous and titanous systems, in agreement the .CH,CHOH radical is formed by the reaction
with the value of 28.9 obtained at pH 1. With the of Fe(1I)-EDTA at pH 7 with H,O, in the
titanous system this slope decreases to 11.3 at presence of ethanol. We have confirmed this and
pH 2.2. Similarly, when thymine is the scavenger, also observed the same radical spectrum when
k3/k, decreases from 71.5 at pH 1.0 to 32.6 at the reaction was carried out at pH 1.2 in the
pH 2.2. If the thymine and ethanol results at absence of EDTA. However, when a 10 m M
pH 2.2 are compared and it is assumed that FeSO, solution at pH 3.8 was mixed with a
k(OH + thymine) = 4.3 x lo9 M-' s-l, then solution 150 mM in H,O, and 400 m M in
k(OH + ethanol) = 1.5 x lo9 M - I s-l. This is ethanol, no radicals were detected. Changing
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 195.37.186.217 on 12/16/19

in agreement with the value measured at pH 1.0. reactant concentrations had no effect. When
These results may be explained by an increase in cc-alanine was substituted for ethanol, similar
k, with increasing pH. There is some other results were obtained. The alanine radical was
evidence suggesting this (9). observed in solutions of pH 1.2 but not at
The ferrous - hydrogen peroxide system be- pH 3.8. As k(Fe(I1) + H,O,) increases with
haves in a different manner. Addition of either increasing pH (23), it might be expected that the
thymine or ethanol in concentrations up to 100 concentration of scavenger radicals available for
times that used at pH 1 to an H,O,,Ti(lV) observation would be greater at the higher pH.
solution of pH 2.2 had no appreciable affect on
the amplitude of the T i 0 0 . 3 + signal observed Discussion
when this solution was mixed with a ferrous
The Systems at pH 1.0
sulfate solution of pH 3.8. The exact pH of the From the data obtained at pH 1.0 and pre-
mixture during the reaction is unknown and will
sented in Table I it is evident that the relative
For personal use only.

be referred to as pH > 2. Under these con- reaction rates of the oxidizing species from
ditions the scavengers did not compete for the
Fe(I1)-H,O,, Ti(II1)-H,O,, and the y-radiolysis
precursors of the Ti00a3+ radical.
of water are identical for a variety of substances.
The mechanism for the reaction of ferrous ions
This suggests that the same oxidizing species is
with hydrogen peroxide proposed by Haber and
produced in each system. A comparison of the
Weiss (21) and modified by Baxendale and co-
Fenton's reagent with the radiolysis of air-
workers (22) consists of reactions [I], [2], and
saturated aqueous solutions has been made
[6] as well as
recently by KraljiC (3) .Competition studies with
[9] H 0 2 - + Fe(I1) -> HOz- + Fe(II1) p-nitrosodimethylaniline indicated that the OH
radical is generated in both systems. The present
investigation confirms this conclusion and shows
Baxendale and co-workers (22) examined the that the Ti(II1)-H,O, reaction also produces
change in kinetics of the Fe(1I)-H,O, reaction hydroxyl radicals.
and the evolution of oxygen with changing The experimental results fit the mechanism
acidity and found that k,,lk, increases from outlined in eqs. [I]-[4]. It is almost identical
0.14 to 0.70 as the pH is increased from 1.35 to with that of Fischer (5), the main difference
2.35. To ensure that all the HO, reacted according being that reactions [2] and [4] replace
to [4], the concentration of Ti(IV) was raised
from 8.3 to 42 m M (as compared with 0.6 m M OH + Ti(IV)/H202 + TiO0.3+ + H 2 0 + H +
Fe(I1)) with no observable difference in the where Ti(IV)/H,O, is the titanium - hydrogen
results. Other evidence that the behavior of the peroxide complex and can be written as TiOZ2+
Fe(I1)-H,O,,Ti(IV) system at pH > 2 is not (24). Our results indicate that it is the hydrogen
due to reactions of the Fe(II1) species was peroxide rather than the complex which competes
obtained by the addition of Fe(1II) to the for hydroxyl radicals. Because the Ti(IV) is
H,O,,Ti(IV), ethanol solutions. This had no complexed, it might be more accurate to write
effect on the results when these solutions were eq. [4] as
mixed with Ti(II1) solutions at pH 2.2 (Fig. 6).
HOz + H + + TiOz2 + TiO0.3+ + H20z
The concentration of scavenger radicals formed
+

during the Fe(I1)-H,O,, scavenger reaction is Whether the T i 0 0 . 3 + species is a titanium

also dependent on pH. Shiga (1) has shown that peroxyl radical or a hydroperoxyl radical com-
 ARMSTRONG: RELATIVE RATE CONSTANTS FOR REACTIONS O F HYDROXYL RADICALS
plexed by titanium(1V) ions is not material to the
overall reaction scheme. The identity of the
second paramagnetic species, present in very low
concentrations, is of greater importance. As it can
be formed under conditions where hydroxyl
radicals are not likely to be present (lo), the case
for a Ti(1V)-OH complex is not completely
convincing. However, the occurrence to a small
extent of
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 195.37.186.217 on 12/16/19
or the back reaction of OH, [6], would not
affect the values of the relative rate constants for
the reactions of OH radicals with the various
scavengers. It would interfere with the evaluation
of k,. That our value of k, is somewhat higher
than the figure reported previously (5.9 x lo7
vs. 4.5 x lo7 M - I s - ' ) may be related to
interference by either reaction [6] or [ll].
The suggestion, made by Turkevich and CO-
workers (4), that the reaction of the hydroxyl
radical with the scavenger takes place during the
For personal use only.
decomposition of a [Ti(H,O),(OH)(scaven-
ger)I4+ complex was based primarily on 2 obser-
vations. They found that, during the initial reac-
tion period of about 0.01 s, the scavenger radical
concentration decreased while the Ti00.3'
signal increased. Although this behavior shows
that the signal is not due to free OH radicals, it
has been pointed out by Florin et al. (6) that it
does riot require the postulation of this complex.
The second observation concerned an increase
in the intensity of the T i 0 0 . 3 + signal of more
than a factor of 2 when small amounts of
methanol were added to the hydrogen peroxide
solution before it was mixed with a titanium
trichloride solution. With higher methanol
concentrations, the signal decreased. We repeated
this experiment with titanous sulfate instead of
titanium trichloride. In all cases the signal
intensity decreased with increasing amounts of
methanol. Thus the phenomenon they observed
is apparently related to reactions of chloride ions.
It is evident that the presence of chloride ions
greatly complicates any study of the reaction
mechanism.
Although the identity of the scavenger radical
or radicals resulting from attack by the oxidizing
species of Fenton's reagent is often the same as
that observed with the Ti(II1)-H202 system (16),
there are some exceptions. As the evidence
presented here suggests that the oxidizing
3743
species is the same in both cases, the differences
in spectra may be due to differences in the
reactions of the scavenger radicals with Fe(II),
Fe(III), Ti(III), or Ti(1V) ions.
The Effect of Increasing pH
It is obvious that reactions [ l ] to [4] do not
explain the results observed with the ferrous -
hydrogen peroxide system at p H > 2. Studies
with the Ti(II1)-H,O, system at the same pH
show that changes in k3/k, with p H are not
sufficient to account for the decrease in the
scavenging ability of ethanol and thymine, and
also indicate that reaction [4] occurs at pH > 2.
As the pK for the dissociation of HO, is 4.4 (25),
only reactions of the unionized form are pertinent
to this investigation. The results can best be
explained on the assumption that hydroxyl
radicals are not formed by the reaction of
ferrous ions with hydrogen peroxide in solutions
of high pH. A possible mechanism is outlined
below
[I21 Fe(I1) f HzO 2 FeOH+ f H +
[13] FeOH+ + HzOz -> FeOZ+f HzO f OH-
[14] FeOZ+f H z 0 2 -> H 0 2 f OH- f Fe(II1)
[15] FeOZ+ f scavenger + product
All the iron ions are aquated.
It is known that K,, for reaction [12] is about
4 x lop4 and that the rate of reaction of ferrous
ion with hydrogen peroxide increases at pH > 3
(23). The competition studies described in the
present investigation were carried out by mixing
a ferrous sulfate solution of pH 3.8 with a
Ti(1V)-H,O, solution of the highest allowable
pH (2.2). Under these conditioils reaction [13]
could predominate over [I]. The experimental
results reported here show that k,,/lc,, << k,/k2.
The formation of the ferryl ion, F e 0 2 + , in the
reaction
Fe(I1) f H 2 0 z -> FeOZ+ f HzO
was originally proposed by Bray and Gorin (26).
Kolthoff and Medalia (27, 28) have pointed out
that it is possible to replace the hydroxyl radical
by the ferryl ion throughout the Haber-Weiss
reaction scheine without altering the kinetics
and that reaction with ethanol may take place
according to [15]. It is therefore suggested that,
under conditions favoring the presence of FeOH'
ions rather than Fe(I1) ions, reaction with
hydrogen peroxide produces an oxidizing species
 3744 CANADIAN JOURNAL OF CHEMISTRY. VOL. 47, 1969

which is not the hydroxyl radical and may be
the ferry1 ion. If the Fe(I1) are cOnl~lexed
with EDTA, the formation of FeOH' is pre-
vented and hydroxyl radicals are generated at
higher pH.
In addition to providing information about
the reaction mechanisms of ferrous and titanous
ions with hydrogen peroxide, the technique
described offers a
method for determining
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 195.37.186.217 on 12/16/19
9. M. ANBARand P. NETA. Int. J. Appl. Radiat.
Isotop. 18, 493 (1967).
10. M. S. BAINS,J. C. ARTHUR,JR., and 0. HINOJOSA.
J. Phys. Chem. 72, 2250 (1968).
1;: &KE,Tf~~~~; JT$~ri,","~,"~a~~R$
MICHAEL.Pulse radiolysis. Acadenxc Press, Inc.,
New York and London. 1965. P. 131.
13. J. H. BAXENDALE, P. L. T. BEVAN,and D. A. STOTT.
Trans. Faraday Sot. 64, 2389 (1968).
14. H. WILKINSON and A. G . TYLER. J. SOC. Dyers
Colour. 43, 114 (1927).
15. A. I. VOGEL. A text-book of quantitative inorganic

the relative rate constants for analysis. 3rd ed. Longmans, Green, and Co., London.
hydroxyl radicals with a wide variety of organic 1961. p. 379.
16. W. A. ARMSTRONG and W. G. HUMPHREYS. Can. J.
compounds in solutions of pH 1. Chem. 45, 2589 (1967).
17. T. J. SWORSKI.Rad~at.Res. 2, 26 (1955).
18. M. ANBARand J. K. THOMAS.J. Phys. Chem. 68,
Acknowledgment 3829 (1964).
19. J. STAUFFand H. J. HUSTER. Z. Phys. Chem. 55,
The author thanks Mr. G. Humphreys 39 (1967).
for valuable technical assistance. 20. C. R. E. JEFCOATE and R. 0. C. NORMAN.J. Chem.
Soc. B, 48 (1968).
21. F. HABERand J. WEISS. Proc. Roy. Soc. London,
1. T. SHIGA. J. Phys. Chem. 69, 3805 (1965). Ser. A, 147, 332 (1934).
2. T. SHIGA,A. BOUKHORS, and P. Douzou. J. Phys. 22. W. G. BARB,J. H. BAXENDALE, P. GEORGE,and
Chem. 71, 4264 (1967). K. R. HARGRAVE. Trans. Faraday Soc. 47, 462
3. I. K R A U I ~ .The chemistry of ionization and (1951).
excitation. Taylor and Francis Ltd., London. 1967. 23. C. F. WELLSand M. A. SALAM. J. Chem. Soc. A,
p. 303. 24 (1968).
For personal use only.

4. Y. S. CHIANG,J. CRADDOCK, D. MICKEWICH, and 24. J. A. CONNORand E. A. V. EBSWORTH.Advan.

J. TURKEVICH.J. Phys. Chem. 70, 3509 (1966). Inorg. Chern. Radiochem. 6, 280 (1964).
5. H. FISCHER. Ber. Bunsenges. Phys. Chem. 71, 25. G. CZAPSKI and L. M. DORFMAN.J. Phys. Chem.
685 (1967). 68, 1169 (1964).
6. R. E. FLORIN, and L. A. WALL. J. Phys.
F. SICILIO, 26. W. C. BRAYand M. H. GORIN. J. Amer. Chem. Soc.
Chem. 72, 3154 (1968). 54, 2124 (1932).
7. K. TAKAKURA and B. RANBY. J. Phys. Chem. 72, 27. A. I. MEDALIA and I. M. KOLTHOFF.J. Polym. Sci.
164 (I 968). 4, 377 (1949).
8. G. SCHOLES and R. L. WILLSON. Trans. Faraday 28. I. M. KOLTHOFF and A. I. MEDALIA.J. Amer.
SOC.63, 2983 (1967). Chem. Soc. 71, 3777 (1949).