A R T I C L E S
Pestovsky and Bakac
different for the two metal(IV) aqua species. Craq(IV) is believed
k2H
(100 - %H)
2+
kH/kD )
to be an oxoaqua species, CraqO , owing to the great oxophi-
licity of chromium. The iron(IV), on the other hand, has been
k2D {100 - (%H × k /k )}
2H 2D
20
suggested to exist as a mixture of two hydrolytic forms, Feaq-
where k2H and k2D are the respective rate constants for reaction 2 (see
+
+
(
OH)22 and Feaq(OH)3 , related by a pKa of 2.0. The lack of
later) in H O and D O.
2
2
an oxo group may make the hydrogen atom transfer a preferable
path for iron provided such a path is thermodynamically allowed.
Here we report the results of a kinetic and mechanistic study
of the oxidation of a number of organic substrates with aqueous
Fe(IV). To simplify chemical equations, we will use the formula
Conductivity measurements were carried out at 26.5 °C with the
use of a TDI model VI stopped-flow apparatus equipped with a custom-
built TDI conductivity cell and ranging amplifier.25 The signal output
was amplified and offset to bring it into the dynamic range of the
instrument. An average of three independent runs was used to determine
the rate constants and the associated conductivity changes. Ultimately,
these data yielded the change in proton concentration during the run
IV 2+
Fe O to represent all forms of Fe(IV) in solution.
IV 2+
and the overall charge on the Fe O ion.
Experimental Section
Product Analysis. For UV-transparent substrates (methanol, ethanol,
Materials. Methanol, methanol-d
, -d , and -d , 2-butanol, benzyl alcohol, para-substituted benzyl
alcohols, cyclobutanol, cyclobutanone, propionaldehyde, benzaldehyde,
paraform, acetone, diethyl ether, THF, acetonitrile, acetonitrile-d , 1,10-
3
, ethanol, 2-propanol, 2-propanol-
2+
2-propanol, THF, formaldehyde, and cyclobutanol), the yields of Feaq
d
1
6
8
were obtained as a difference between the known initial concentration
2+
3+
of Feaq and the amount of Feaq determined by direct conventional
spectrophotometry at 240 nm immediately following the manual mixing
3
phenanthroline, sodium acetate, hydrogen peroxide, chromium(III)
perchlorate, iron(III) perchlorate, titanium(IV) oxysulfate, perchloric
acid, and D O were of highest purity from commercial sources (Aldrich,
2
3
of the reagents. For benzyl alcohol, p-CF -benzyl alcohol, acetone, and
acetonitrile, the reaction mixture was combined with 0.17 mM
phenanthroline (added as acetonitrile solution, giving a final acetonitrile
Fisher, Cambridge Isotope Laboratories) and were used as received.
Deionized water was obtained by passage of in-house distilled water
through a Millipore Milli-Q water purification system.
content of 3.9%) and 0.14 M sodium acetate. The absorbance reading
2+
was taken at 510 nm, where Fe(phen)
3
has an extinction coefficient
4
-1
-1
of 1.14 × 10 M cm . This method proved to be less precise than
2
+
2+
Feaq stock solutions were prepared by Zn/Hg reduction of 0.010
the direct spectrophotometry, because some Fe
(5-10 µM) was
aq
3
+
3+
M Feaq in 0.10 M HClO
4
. The amount of residual Feaq was
2 2
consumed in the reaction with H O during the analysis.
determined by UV-vis spectrophotometry, ꢀ240 (Feaq +) ) 4160 M-1
cm , and never exceeded 0.1% of initial Feaq . Ozone was generated
3
The yields of H O were determined by adding 20 µL of TiOSO
2
2
4
-
1
2+
2 4
reagent (1.42 M in 15% H SO ) to a 3-mL sample solution. The titanium
peroxo complex, ꢀ408 ) 723 M-1 cm-1, was formed immediately and
quantitatively. However, in experiments containing both Feaq2+ and
with an Ozonology L-100 apparatus. Aqueous solutions of ozone were
prepared by continuous bubbling of ozone through 0.10 M HClO
4
for
at least 30 min at room temperature. Such solutions, standardized
H O in the spent solution (see Results), the initially formed titanium
2
2
-
1
-1 24
26
spectrophotometrically, ꢀ260 (O
approximately 0.5 mM O . Titanium oxysulfate test (see below) showed
no detectable quantities (<2 µM) of H in ozone stock solutions.
3
) ) 3300 M cm
,
contained
peroxo complex decayed rapidly (minutes or less). The yields of H O
2
2
3
were obtained by extrapolation to zero time.
O
2
As a double check for solutions containing both H O and Fe 2+
,
2
2 2 aq
2+
22
Craq
O
solutions were prepared as described previously and contained
the peroxide yields were also calculated from the initial rates of the
2
+
and 17 µM CraqOO2+. Stock solutions
2+
approximately 50 µM Craq
of formaldehyde were prepared by dissolving a known amount of
paraform in warm 1.0 M HClO
Kinetics All the kinetics studies were carried out at 25 ( 0.1 °C in
.10 M aqueous HClO , unless stated otherwise. For fast reactions, an
OLIS RSM-1000 stopped-flow apparatus was used. The reaction
O
Feaq /H O reaction and the mechanism in Scheme 1, where k is the
2
2
F
rate constant for the Fenton reaction.
4
.
According to the scheme, the hydrolysis of hydroxyalkyl hydro-
peroxides regenerates H O , which thus becomes a catalyst for the co-
2
2
2
+
0
4
oxidation of Fe
2
and alcohols by O . Pseudo-first-order kinetics with
aq
respect to [Feaq2 ] should be obeyed when sufficient amounts of O2
are present. A more complex kinetic behavior results when hydrogen
peroxide is consumed in side reactions or is not regenerated from a
particular peroxide. Any complications resulting from changing [H O ]
+
IV 2+
2+
2+
between Fe O and Feaq was studied at high [Feaq ] to ensure that
all the Fe O was generated in the first few milliseconds, which were
IV 2+
not used in the kinetic fitting. In the experiments on substrate oxidation
2
2
IV 2+
by Fe O , one stopped-flow syringe contained the ozone solution,
were avoided by use of initial rates. The method was checked with
2+
2+
and the other, a mixture of Feaq and substrate. After mixing, the typical
authentic reagents for the case Feaq (43 µM)/H O (30-50 µM)/
2
2
2
+
concentration of ozone was 0.25 mM, and that of Feaq , 0.10 mM.
i
cyclobutanol (0.102 M). As expected, the initial rates, V , obeyed the
IV 2+
2+
-1 -1
Under these conditions, Fe O was generated in the first few
rate law, V ) 2 × k × [H O ] × [Fe ], where k ) 58 M
s , the
i
F
2
2
aq
F
2+
27
milliseconds. The small, controlled excess of ozone was necessary to
aq 2 2
known rate constant for the Fe /H O reaction. The amount of
hydrogen peroxide produced in the reactions between Fe O and
2
+
IV 2+
scavenge any Feaq produced in 2-e steps (see Results) and, thus,
2
+
IV 2+
2+
prevent the Feaq / Fe O reaction and the associated absorbance
changes caused in large part by the formation and decay of dimeric
Fe(III). The data, obtained by rapid scanning (∆t ) 1 ms) in the 265-
substrates was calculated as [H O ] ) V /(2k × [Fe ]).
2
2
i
F
aq
1
H NMR spectra were recorded with a Varian VXR-400 spectrometer
at room temperature.
3
92 nm spectral range, were analyzed with the use of OLIS Global-
Results
Works v2.0.190 software. The rate constants were obtained as averages
of at least three experiments. Slower reactions were monitored with a
Shimadzu 3101 PC spectrophotometer. Kinetics of the reaction between
IV 2+
Fe O Formation and Decay. Literature data already exist
2+
IV 2+
for both the Feaq /O3 reaction and the decay of Fe O in
acidic aqueous solutions.
added substrates is complex (see later), precise kinetic data for
reactions 1 and 2 under our exact experimental conditions were
2+
2 2
Feaq and H O in the presence of alcohols were studied by monitoring
18-20
Because the reaction scheme with
3
+
the increase in Feaq concentration at 240 nm.
The analyses of kinetic data were performed with Kaleidagraph v3.51
for PC software. Kinetic simulations were carried out with Kinsim v4.0
software.
(
(
(
25) Knipe, A. C.; McLean, D.; Tranter, R. L. J. Phys. E: Sci. Instrum. 1974,
IV 2+
The kinetic isotope effect for the self-decay of Fe O was calculated
7, 586-590.
2
+
4 2 4
26) This rapid reaction between TiO(SO )/H SO and Feaq stands in contrast
53
from the expression
to the very slow reaction reported in the perchlorate medium.
27) Wang, W. D.; Bakac, A.; Espenson, J. H. Inorg. Chem. 1993, 32, 2005-
(24) Hart, E. J.; Sehested, K.; Holcman, J. Anal. Chem. 1983, 55, 46-49.
2009.
1
3758 J. AM. CHEM. SOC. VOL. 126, NO. 42, 2004
9