Confirming the Formation of Hydroxyl Radicals in the Catalytic Decomposition of H2O2 on Metal Oxides Using Coumarin as a Probe

Article abstract of DOI:10.1002/cctc.201901316

Hydrogen peroxide can be catalytically decomposed to O₂ and H₂O on metal oxide surfaces in contact with aqueous solutions containing H₂O₂. The initial step in this process has been proposed to be the formation of surface-bound hydroxyl radicals which has recently been verified using tris as a radical scavenger. Here, we make use of the unique fluorescent product 7-hydroxycoumarin formed in the reaction between hydroxyl radicals and coumarin to probe the formation of surface-bound hydroxyl radicals. The experiments clearly show that 7-hydroxycoumarin is formed upon catalytic decomposition of H₂O₂ in aqueous suspensions containing ZrO₂-particles and coumarin, thereby confirming the formation of surface-bound hydroxyl radicals in this process. The results are quantitatively compared to results on the same system using tris as a probe for hydroxyl radicals. The effects of the two probes on the system under study are compared and it is concluded that coumarin has a significantly lower impact on the system.

Full text of DOI:10.1002/cctc.201901316

DOI: 10.1002/cctc.201901316
Communications
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
Confirming the Formation of Hydroxyl Radicals in the
Catalytic Decomposition of H O on Metal Oxides Using
2
2
Coumarin as a Probe
[a]
[a]
[a]
Annika Carolin Maier, Ena Herceglija Iglebaek, and Mats Jonsson*
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
Hydrogen peroxide can be catalytically decomposed to O and
It is important to note that the hydroxyl radicals formed
here are adsorbed to the catalyst surface (a prerequisite for this
reaction to occur spontaneously) and therefore probably display
a different reactivity from free hydroxyl radicals. The formation
2
H O on metal oxide surfaces in contact with aqueous solutions
2
containing H O . The initial step in this process has been
2
2
proposed to be the formation of surface-bound hydroxyl
radicals which has recently been verified using tris as a radical
scavenger. Here, we make use of the unique fluorescent
product 7-hydroxycoumarin formed in the reaction between
hydroxyl radicals and coumarin to probe the formation of
surface-bound hydroxyl radicals. The experiments clearly show
that 7-hydroxycoumarin is formed upon catalytic decomposi-
tion of H O in aqueous suspensions containing ZrO -particles
of hydroxyl radicals in catalytic decomposition of H
O on metal
2
2
oxides was later confirmed using tris(hydroxymethyl)-amino-
[5]
methane (tris) as a scavenger. The nature of surface-bound
hydroxyl radicals has been investigated fairly extensively using
[
6,7]
DFT.
formed in the first step of the catalytic decomposition of H
on oxide surfaces has also been found to play a major role in
the oxidative dissolution of used UO -based nuclear fuel under
waste repository conditions. In general, catalytic decomposi-
tion of H on metal oxide surfaces is an important process in
Quite recently, the surface bound hydroxyl radical
O
2
2
2
2
2
and coumarin, thereby confirming the formation of surface-
bound hydroxyl radicals in this process. The results are
quantitatively compared to results on the same system using
tris as a probe for hydroxyl radicals. The effects of the two
probes on the system under study are compared and it is
concluded that coumarin has a significantly lower impact on
the system.
2
[8]
O
2
2
nuclear technology (to understand radiation induced corrosion)
and it can also be used as an advanced oxidation technique
(AOT) to purify water from organic pollutants.
One of the products formed upon the reaction between
*
OH and tris is formaldehyde, which can be quantified fairly
easily. This method was first used to detect hydroxyl radicals
[
5]
produced upon catalytic decomposition of H O on ZrO2.
2
2
Highly reactive hydroxyl radicals can be produced in several
different ways. One of the most straightforward ways is
radiolysis of water where the absorption of ionizing radiation
Following the first study on ZrO , numerous experimental
2
studies on other oxides have been performed.
studies, methanol has been used as a scavenger for the
hydroxyl radical. Also, this reaction produces formaldehyde.
Both tris and methanol have also been used to quantify
[6,9,10]
In many
*
À
*
*
results in the formation of OH, e , H , H O , H , HO and
aq
2
2
2
2
+
[1]
H O . Heterogeneous photocatalysis using TiO2 or other
3
[
10]
semiconducting materials as well as the homogeneous Fenton
hydroxyl radicals produced by heterogeneous photocatalysis.
[2,3]
reaction are other possibilities. It has been known for quite
some time that metal oxides can catalyse the decomposition of
Although very useful in assessing hydroxyl radical formation in
heterogeneous systems, it should be kept in mind that the
reaction with either tris or methanol yielding formaldehyde is
not necessarily unique for hydroxyl radicals. In principle, any
radical species capable of abstracting a hydrogen atom from
tris or methanol will produce formaldehyde. Hence, truly
unambiguous evidence for hydroxyl radical formation calls for
alternative radical scavengers where the reaction with hydroxyl
radicals yields a unique product that cannot be formed from
other hydrogen-abstracting radicals. A promising candidate is
coumarin that upon reaction with the hydroxyl radical yields a
number of hydroxylated products of which one is strongly
[4]
H O2 in aqueous solution producing O2 and H O. The
2
2
mechanism for this process was proposed to proceed via the
formation of hydroxyl radicals according to the following
[4]
mechanism [Eqs. (1) to (3)]:
.
H O ! 2 OH
(1)
(2)
(3)
2
2
.
.
OH þ H O ! H O þ HO
2
2
2
2
.
.
HO2 þ HO2 ! H O þ O
2
2
2
[11]
fluorescent. The primary step of the reaction is addition of
the hydroxyl radical to form a radical-adduct followed by
oxidation of the adduct (by O ) to form the hydroxylated
2
products. The fluorescent product is 7-hydroxycoumarin. Cou-
marin has been used to quantify hydroxyl radicals in radiation
chemistry as well as in heterogeneous photocatalysis.
[
a] A. C. Maier, E. H. Iglebaek, Prof. M. Jonsson
Department of Chemistry
KTH Royal Institute of Technology
Stockholm SE-100 44 (Sweden)
E-mail: matsj@kth.se
[11–17]
However, it has to the best of our knowledge not been used in
ChemCatChem 2019, 11, 1–5
1
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
experiments monitoring the catalytic decomposition of H O on
metal oxide surfaces (Scheme 1).
high. The results from the two sets of experiments are
presented in Figure 1.
2
2
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
In this work we have studied the possible formation of
hydroxyl radicals in the reaction between H O and ZrO using
In both cases when ZrO was present, H O was completely
2
2
2
consumed at the time of the sampling (24.5 h). The general
trends in the two sets of experiments look very similar. A
2
2
2
coumarin as a probe. The experimental results are compared to
the corresponding results using tris as a probe.
background signal (i.e., a signal in the absence of ZrO ) is
2
In heterogeneous photocatalysis, coumarin is sometimes
claimed to scavenge hydroxyl radicals in the bulk rather than
observed in both cases. The signal in the presence of ZrO is
2
significantly higher than the background after 24.5 h in both
cases. The fact that 7-hydroxycoumarin is formed at levels
significantly above the background in the system containing
H O and ZrO2 serves as an independent confirmation that
[18]
on the photocatalyst surface. The bulk hydroxyl radicals are
argued to be surface-bound hydroxyl radicals that have been
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
[18]
desorbed and subsequently diffused away from the surface.
Given the strong adsorption of hydroxyl radicals to oxide
2
2
hydroxyl radicals are indeed formed in the catalytic decom-
position of H O on ZrO .
[6,7]
surfaces demonstrated by both DFT calculations and experi-
ments assessing the reduction potential of the surface-bound
2
2
2
The relative yield of 7-hydroxycoumarin in the reaction
between hydroxyl radicals and coumarin is known to be 5.7%
[19]
hydroxyl radical compared to the free hydroxyl radical,
[12]
desorption of the radical would appear to be a very unlikely
process. Also, the concentration dependence on the rate of
coumarin hydroxylation in heterogeneous photocatalysis differs
significantly from that observed in homogeneous gamma
radiolysis (where hydroxyl radicals are produced homogene-
ously in solution) which points in the direction of surface-bound
from radiation chemical experiments. In the same way, the
formaldehyde yield for the reaction between the hydroxyl
[5]
radical and tris has been determined to 35%. By using the
relative yields we can calculate the total concentration of
scavenged hydroxyl radicals in the two systems.
For comparison we also calculate the fraction of trapped
hydroxyl radicals per probe molecule (concentration of trapped
hydroxyl radicals divided by total probe concentration) and the
hydroxyl
photocatalysis.
radicals
in
the
case
of
heterogeneous
[
12,20]
Although a high coumarin concentration (resulting in a
larger amount of probe being adsorbed on the surface of the
catalyst) would provide a higher radical scavenging capacity,
the competing absorption of the excitation light by coumarin in
the fluorescence measurement of 7-hydroxycoumarin drastically
[20]
reduces the sensitivity of the method. There are several ways
to handle this problem. By far, the easiest way is to perform the
calibration of the fluorescence measurements using solutions
with the same coumarin concentrations as used in the experi-
ments. If different coumarin concentrations are used, the
samples can be diluted to the same coumarin concentration
before the fluorescence measurement. Alternatively, calibra-
tions can be performed at different coumarin concentrations.
The material of the reaction vessel as well as vigorous
stirring or gas purging and frequent sampling turned out to
affect the reproducibility of the experiments. In the end, the
optimum conditions of the coumarin experiment were the ones
described in the experimental section. Five parallel experiments
were performed and the reproducibility turned out to be very
Figure 1. Concentration of formaldehyde (a) and 7-hydroxycoumarin (b) as a
function of exposure time in experiments using tris (a) and coumarin (b) to
scavenge hydroxyl radicals.
Scheme 1. Principle for probing hydroxyl radical formation by using
coumarin.
ChemCatChem 2019, 11, 1–5
www.chemcatchem.org
2
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
fraction of scavenged hydroxyl radicals per decayed H O . These
results are summarized in Table 1.
In a system containing a radical scavenger there are two
competing reactions that consume the hydroxyl radical formed
in reaction (1). The first reaction is hydrogen abstraction from
H O , reaction (2), and the second is the reaction between the
2
2
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
The scavenging of surface-bound hydroxyl radicals is a
surface reaction and its efficiency will depend on the amount of
adsorbed radical scavenger. This situation is analogous to
probing the efficiency of a heterogeneous photocatalyst. For
heterogeneous photocatalysis, Langmuir-Hinshelwood kinetics
can often be used to describe the probe concentration
2
2
hydroxyl radical and the scavenger (probe). At high scavenger
concentrations the yield of O (reaction (3)) will be significantly
2
[24]
reduced. From Table 1 it can be seen that tris scavenges 0.09
hydroxyl radicals per decomposed H O . Theoretically, two
2
2
[21]
dependence. Maximum scavenging capacity is reached when
the photocatalyst surface is saturated with probe molecules. To
reach this point, a sufficiently high concentration of the probe
hydroxyl radicals could be formed per decomposed H O2
2
provided that the scavenger (tris) concentration is high enough
to outcompete hydrogen abstraction from H O , i.e., H O is
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
2
2
2
2
must be used. In a recent work on TiO -photocatalysis, the
only consumed in reaction (1). In a system without an added
scavenger, the theoretical number of hydroxyl radicals formed
per decomposed H O is one (since H O is consumed in the
2
coumarin concentration dependence was explored and it was
concluded that the solubility of coumarin is too low to get even
2
2
2
2
[20]
close to surface saturation. However, the affinity of coumarin
reaction with surface-bound hydroxyl radicals, reaction (2)).
Hence, quantitative scavenging of hydroxyl radicals can only be
achieved when the overall reaction system is severely altered
by the presence of the probe. At the concentration used in this
work, tris will have a small but significant effect on the reaction
under study. Coumarin on the other hand only scavenges 3×
towards the TiO -surface was assessed to be similar to that of
2
[20]
tris. Nevertheless, a fraction of the coumarin is adsorbed to
the surface and capable of scavenging hydroxyl radicals. In a
previous paper, the tris concentration dependence on the
formaldehyde production rate in a system where H O was
2
2
[22]
À 4
catalytically decomposed on ZrO was studied. Interestingly,
10 hydroxyl radicals per decomposed H O . This is considered
2
2
2
the concentration dependence paralleled the trend seen for
to be negligible. In this respect, a probe like coumarin can be
seen as an alternative with minimal influence on the chemistry
of the system since the sensitivity is quite reasonable also at
low probe concentrations. However, under such conditions the
probe only provides qualitative information.
TiO -photocatalysis (the rate of formaldehyde formation be-
2
[23]
comes independent of tris-concentration above 100 mM)
indicating that the affinity of tris towards ZrO is similar to the
2
affinity towards TiO2. The difference in the amount of
scavenged hydroxyl radicals between the tris and coumarin
experiments in the present work is around a factor of 300 in
favour of tris. Given the fact that the tris concentration is three
orders of magnitude higher than the coumarin concentration
and that the affinity of tris and coumarin towards oxide surfaces
can be assumed to be similar, the observed difference in
hydroxyl radical scavenging capacity between the two systems
appears quite reasonable.
As can be seen in the table, the fraction of probe molecules
consumed during the duration of the experiment is quite low in Experimental Section
both cases. Only around 1% of the coumarin is consumed while
the consumption of tris corresponds to less than 0.4%. In other
words, the concentration of probe molecules can be considered
constant throughout the experiment in both cases.
In this work we have confirmed the formation of hydroxyl
radicals in the catalytic decomposition of H O on ZrO using
2
2
2
coumarin as a selective probe. Apart from yielding a unique
product upon reaction with hydroxyl radicals, coumarin is also
demonstrated to be a sensitive probe capable of detecting
hydroxyl radicals at probe concentrations low enough not to
alter the overall surface chemistry.
Two sets of experiments were performed using either coumarin or
tris as radical scavengers. Purified water (18.2 MΩ.cm, Merck MilliQ)
was used throughout and chemicals were of reagent grade or
higher unless otherwise stated.
In general, the presence of a radical scavenger may
influence the overall mechanism as well as the kinetics for
catalytic decomposition of H O . The kinetics of H O decom-
For experiments were coumarin was used as radical scavenger,
1
.25 g ZrO2 powder were suspended in aqueous coumarin
solutions. H O was added to the suspensions to reach a total
2
2
2
2
2
2
2
volume of 50 ml, a coumarin concentration of 0.05 mM and a H O2
position is affected unless reaction (1) is the rate limiting step.
concentration of 2 mM. In experiments were tris was used instead
of coumarin, the pH of the tris solution was adjusted to 7.5 using
HCl before adding ZrO
suspensions to reach a total volume of 50 ml, a tris concentration
of 50 mM and H O concentration of 2 mM. All powder
. Afterwards, H O was added to the powder
2
2 2
Table 1. Summary of results from the hydroxyl radical scavenging experi-
ments.
a
2
2
suspensions (containing either tris or coumarin) where kept in
polypropylene test tubes which were protected from light and
constantly mixed using a test tube rocking mixer. Both sets of
experiments were repeated 5 times to assure reproducibility.
Background measurements with either coumarin or tris were also
Tris/CH O
2
Coumarin/7-hy-
droxycoumarin
Concentration of product due to
reaction on ZrO
64�5 μM
35�5 nM
2
Concentration of scavenged hydroxyl 180�10 μM 620�90 nM
performed in sets of five but in the absence of ZrO . 2 ml aliquots
2
radicals
were taken from the powder suspensions and the blank samples
just before, 5 min as well as 24.5 h after the addition of H O . The
À 3
À 2
Fraction of scavenged hydroxyl radi-
cals per probe molecule
Fraction of scavenged hydroxyl radi-
3.6×10
0.09
1.2×10
2
2
À 4
aliquots were filtered through 0.2 μm cellulose acetate syringe
filters to separate the powder from the solution. The concentrations
3×10
2 2
cals per decayed H O
ChemCatChem 2019, 11, 1–5
www.chemcatchem.org
3
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
*
of H O as well as the products from the reaction between HO and
coumarin or tris were determined from the aliquots.
[2] A. Fujishima, K. Honda, Nature 1972, 238, 37–38.
3] C. Walling, Acc. Chem. Res. 1998, 31, 155–157.
[4] A. Hiroki, J. A. LaVerne, J. Phys. Chem. B 2005, 109, 3364–3370.
[5] C. M. Lousada, M. Jonsson, J. Phys. Chem. C 2010, 114, 11202–11208.
[6] C. M. Lousada, A. J. Johansson, T. Brinck, M. Jonsson, J. Phys. Chem. C
2
2
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
[
H O concentrations were measured spectrophotometrically at
2
2
[
25]
3
50 nm using the Ghormley triiodide method.
2
012, 116, 9533–9543.
*
Formaldehyde formed upon the reaction between tris and HO can
indirectly be measured by spectrophotometry at 368 nm using the
[7] C. M. Lousada, A. J. Johansson, T. Brinck, M. Jonsson, Phys. Chem. Chem.
Phys. 2013, 15, 5539–5552.
[
26,27]
[
8] A. Barreiro Fidalgo, Y. Kumagai, M. Jonsson, J. Coord. Chem. 2018, 71,
799–1807.
[9] C. M. Lousada, M. Yang, K. Nilsson, M. Jonsson, J. Mol. Catal. A 2013,
79, 178–184.
modified Hantzsch method,
where a dihydropyridine derivative
1
forms when 1.5 ml formaldehyde containing sample is left to react
for 15 min with ammonium acetate (2.5 ml, 4 M) and acetoacetani-
lide (1 ml, 0.2 M) at 40 C.
3
°
[
10] Å. Björkbacka, M. Yang, C. Gasparrini, C. Leygraf, M. Jonsson, Dalton
Trans. 2015, 44, 16045–16051.
[11] K. Gopakumar, U. R. Kini, S. C. Ashawa, N. S. Bhandari, G. U. Krishnan, D.
Krishnan, Radiat. Eff. 1977, 32, 199–203.
12] G. Louit, S. Foley, J. Cabillic, H. Coffigny, F. Taran, A. Valleix, J. P. Renault,
S. Pin, Rad. Phys. Chem. 2005, 72, 119–124.
[13] K. Ishibashi, A. Fujishima, T. Watanabe, K. Hashimoto, Electrochem.
Commun. 2000, 2, 207–210.
14] H. Czili, A. Horvàth, Appl. Catal. B 2008, 81, 295–302.
15] U. Cernigoj, U. L. Stangar, P. Trebse, M. Sarakha, J. Photochem. Photobiol.
A 2009, 201, 142–150.
16] J. Zhang, Y. Nosaka, J. Phys. Chem. C 2013, 117, 1383–1391.
17] J. Zhang, Y. Nosaka, Appl. Catal. B 2015, 166–167, 32–36.
18] J. Zhang, Y. Nosaka, J. Phys. Chem. C 2014, 118, 10824–10832.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
*
Coumarin reacts with HO to form fluorescent 7- hydroxycoumarin
(among other hydroxylated species), which was measured using a
fluorescence spectrophotometer. A calibration for aqueous solu-
tions containing 7- hydroxycoumarin (various concentrations) and
0.05 mM coumarin was used for quantification. The excitation
wavelength was set to 326 nm and the emission was measured at
[
[
[
456 nm using a Cary Eclipse Fluorescence Spectrophotometer.
[
[
[
Acknowledgements
[19] D. Lawless, N. Serpone, D. Meisel, J. Phys. Chem. 1991, 95, 5166–5170.
20] V. Leandri, J. M. Gardner, M. Jonsson, J. Phys. Chem. C 2019, 123, 6667–
[
The Swedish Nuclear Fuel and Waste Management Company
6674.
[
21] M. A. Fox, M. T. Dulay, Chem. Rev. 1993, 93, 341–357.
[22] M. Yang, M. Jonsson, J. Mol. Catal. A 2015, 400, 49–55.
23] V. Diesen, M. Jonsson, J. Adv. Oxid. Technol. 2012, 15, 392–398.
24] C. M. Lousada, J. A. LaVerne, M. Jonsson, Phys. Chem. Chem. Phys. 2013,
(SKB) is gratefully acknowledged for financial support.
[
[
1
5, 12674–12679.
[25] J. A. Ghormley, A. C. Stewart, J. Am. Chem. Soc. 1956, 78, 2934–2939.
26] Q. Li, P. Sritharathikhun, S. Motomizu, Anal. Sci. 2007, 23, 413–417.
Conflict of Interest
[
The authors declare no conflict of interest.
[27] T. Nash, Biochem. J. 1953, 55, 416.
Keywords: Hydrogen peroxide · Catalytic decomposition ·
Metal oxide · Surface-bound hydroxyl radicals · Coumarin
Manuscript received: July 23, 2019
Accepted manuscript online: September 9, 2019
Version of record online: ■■■, ■■■■
rd
[
1] J. W. T. Spinks, R. Woods, An introduction to radiation chemistry, 3 ed.
Wiley, New York, NY 1990.
ChemCatChem 2019, 11, 1–5
www.chemcatchem.org
4
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
COMMUNICATIONS
New probes: By using fluorescence
spectroscopy to quantify the
A. C. Maier, E. H. Iglebaek, Prof. M.
Jonsson*
formation of 7-hydroxycoumarin we
confirmed the formation of hydroxyl
radicals upon the catalytic decompo-
sition of H O on ZrO in aqueous
1 – 5
Confirming the Formation of
Hydroxyl Radicals in the Catalytic
2
2
2
powder suspension.
Decomposition of H
Oxides Using Coumarin as a Probe
2 2
O on Metal
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5