Journal of Photochemistry and Photobiology A: Chemistry 340 (2017) 8–14

Contents lists available at ScienceDirect

Journal of Photochemistry and Photobiology A:

Chemistry
journal homepage: www.elsevier.com/locate/jphotochem

Invited paper

Hydroxyl radical generation with a high power ultraviolet light

emitting diode (UV-LED) and application for determination of hydroxyl
radical reaction rate constants
Kazuhiko Takedaa,b,* , Katsunari Fujisawaa , Hitoshi Nojimaa , Ryota Katob , Ryuta Uekib ,
Hiroshi Sakugawaa,b
a
Graduate School of Biosphere Science, Hiroshima University, 1-7-1 Kagamiyama, Higashi-Hiroshima 739-8521, Japan
b
Faculty of Integrated Arts and Sciences, Hiroshima University, 1-7-1 Kagamiyama, Higashi-Hiroshima 739-8521, Japan

A R T I C L E I N F O A B S T R A C T

Article history:
Received 10 January 2017 We propose a simple, efficient and selective hydroxyl radical generation system based on the photolysis
Received in revised form 22 February 2017 of submicromolar concentrations of nitrite using a high-power ultraviolet light emitting diode (UV-LED).
Accepted 23 February 2017 Hydroxyl radical formation by the 6.75-W UV-LED was at least 10 times greater than that by a 300-W Xe
Available online 28 February 2017 lamp. In the UV-LED system, the hydroxyl radical formation rate from nitrite was about four orders of
magnitude larger than that from nitrate and two orders of magnitude larger than that from hydrogen
Keywords: peroxide. Such efficient and selective hydroxyl radical formation can be attributed to the overlap of the
Hydroxyl radical emission spectrum of the UV-LED and absorption of nitrite. The system was used to determine the
Reaction rate constant
reaction rate constants between hydroxyl radicals and chemicals based on the competition method with
Ultraviolet light emitting diode
terephthalate as the hydroxyl radical probe. The reaction rate constants between hydroxyl radicals and
Nitrite
Terephthalate some low-molecular-weight organic compounds and inorganic halide salts with various reaction rate
constants were determined. The values obtained ranged from 105 to 1010 M
1 s
1, and agreed well with
previously reported values. The potential of the developed method to determine the reaction rate
constants of hydroxyl radicals is discussed.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction the transformation and degradation of organic compounds.

Short-lived reactive oxygen and nitrogen species such as
hydroxyl radicals [1–3], superoxide anions [4], hydrogen peroxide
[5,6] and nitric oxide [7–9] are photochemically generated in water
in the natural environment. The hydroxyl radical, which is the most
oxidative free radical, is generated by photolysis of nitrate, nitrite
and hydrogen peroxide under natural sunlight [1–3] as well as by
the photochemical reaction of dissolved organic compounds in
natural water [2,3]. The photo-Fenton and Fenton-like processes
involving a transition metal such as iron or copper also generate
hydroxyl radicals [10]. Because of its strong oxidizing potential, the
hydroxyl radical reacts with many molecules with a high reaction
rate constant. Therefore, the hydroxyl radical plays major roles in
* Corresponding author at: Graduate School of Biosphere Science, Hiroshima
University, 1-7-1 Kagamiyama, Higashi-Hiroshima 739-8521, Japan.
E-mail address: takedaq@hiroshima-u.ac.jp (K. Takeda).
Particularly for persistent organic pollutants, which resist chemi-
cal, biological and photochemical degradation in natural environ-
ments, degradation by hydroxyl radicals can act as an important
possible sink in natural environments. Nakatani et al. studied the
photochemical degradation of bisphenol A in river water, finding
that hydroxyl radicals strongly influence the lifetime of bisphenol
A in the natural environment [11]. It is widely accepted that the
presence of hydroxyl radicals can strongly influence the fates of
natural and anthropogenic compounds discharged in water
environments [12–14]. The reaction rate constant between a
hydroxyl radical and chemical species in an aqueous environment
is a crucial parameter to characterize and understand the behavior
of pollutants. Because the hydroxyl radical is a toxic reactive
oxygen species in biological systems, it is a potent agent that can
cause many pathological effects and aging. The reactions of
hydroxyl radicals with biomolecules can lead to lethal damage in
biological systems. The scavenging of hydroxyl radicals in vivo and

http://dx.doi.org/10.1016/j.jphotochem.2017.02.020
1010-6030/© 2017 Elsevier B.V. All rights reserved.
 K. Takeda et al. / Journal of Photochemistry and Photobiology A: Chemistry 340 (2017) 8–14
in vitro is of considerable interest in life sciences and medicine.
Evaluation of the reaction rate constants of hydroxyl radicals and
chemical species in biological media is thus an important task.
Numerous attempts have been made to determine the reaction
rate constants between hydroxyl radicals and chemicals as the
antioxidant activities of biomolecules [15–18]. The transient
method using gamma ray pulse radiolysis has provided a huge
number of data sets to examine the reactions of hydroxyl radicals.
Some of them can be conveniently referred to on the website of the
Notre Dame Radiation Laboratory (NDRL) of the University of Notre
Dame, USA, in collaboration with the National Institute of
Standards and Technology (NIST), USA [19].
Steady-state methods based on competitive kinetics are
commonly used to determine reaction rate constants. In such
steady-state methods, hydroxyl radicals competitively react with a
probe compound (or reference substance) and target compound,
resulting in the formation of a stable product from the reaction
between the hydroxyl radicals and probe compound. Many probe
compounds to detect hydroxyl radicals have been reported,
including highly sensitive fluorescent probes, such as terephthal-
ate (TP) [15,17,20–22], coumarin [16], 3-hydroxycoumarin [23],
benzoate [24,25] and benzene [3,11]. These probes react with
hydroxyl radicals with high reaction rate constants (>109 M
1 s
1),
and generate stable fluorescent products. A hydroxyl radical
generation system is also an essential element for the competitive
method. The Fenton reaction, photolysis, and radiolysis have been
used to generate hydroxyl radicals. Louit et al. developed a
fluorometric assay to determine hydroxyl radical reaction rate
constants using a 137Cs gamma-ray source with coumarin [16].
They used a multi-well microplate for radiolysis to realize a simple,
rapid, high-throughput screening method of antioxidants, al-
though a gamma-ray source with N2O gas was essential. The
Fenton system using hydrogen peroxide and Fe(II) or Fe(II)–
ethylenediaminetetraacetic acid (EDTA) chelate has been frequent-
ly used to evaluate hydroxyl radical scavenging capacity [15,17,18].
However, the reaction mixtures contain submillimolar-order
ferrous compound and hydrogen peroxide, which can scavenge
the hydroxyl radicals and cause unexpected reactions. The simple
method to assess the relative reactivities of hydroxyl radicals
scavenging were also developed, and were applied to the biological
and medical fields [26,27].
The 2014 Nobel Prize for physics was awarded to three Japanese
scientists, Isamu Akasaki, Hiroshi Amano and Shuji Nakamura,
who invented an efficient blue light emitting diode (LED). Their
achievements have been used to develop short-wavelength LEDs
that emit in the ultraviolet (UV) region. Recently, high-power, low-
cost, long-lived UV-LED elements have become available on the
market. In this study, we propose a simple photolysis system to
generate hydroxyl radicals using a UV-LED. This system is used to
determine the reaction rate constant between hydroxyl radicals
and chemicals. The high-power 365-nm UV-LED used here
generates hydroxyl radicals by photolysis of submicromolar nitrite
aqueous solutions with an absorption maximum located at around
355 nm [28,29].
NO2
+ H2O + hv ! OH + NO + OH

We use TP as the hydroxyl radical probe, which is readily soluble
in water at neutral pH, stable, and not fluorescent and has low
toxicity. TP reacts with hydroxyl radicals with a reaction rate
constant of 3.3–4.4  109 M
1 s
1 to quantitatively form a strongly
fluorescent product, 2-hydroxyterephthalate (HTP) [20–22]. We
examine the features of hydroxyl radical generation by the UV-LED,
and the potential of the method to determine the reaction rate
constants of hydroxyl radicals.
9
2. Material and methods
2.1. Reagents
TP and HTP were obtained from Tokyo Chemical Industry Co.
Ltd. (Tokyo, Japan). Other chemicals and solvents were of reagent
or HPLC grade, and used without further purification. All solutions
were made using ultrapure water obtained from a Milli-Q Elix UV-
5 and Milli-Q Advantage (Millipore, Tokyo, Japan).
2.2. UV-LED light source
The UV-LED was purchased from Nitride Semiconductors Co.
Ltd (NS365L-6SMG, Naruto, Japan). According to the specification
sheet of the UV-LED, its typical peak wavelength is 360–370 nm,
full width at half-maximum is 9 nm, and typical optical output
power is 2750 mW at a DC forward current of 1.0 A. The UV-LED
was installed on an aluminum heat sink with a cooling fan driven
by a custom-made constant-current power supply at a DC forward
current of 0.9 A and forward voltage of 7.4–7.6 V.
2.3. Batch photochemical experiments
Photochemical experiments with a quartz photochemical
reaction cell were conducted by the batch method to characterize
hydroxyl radical formation using the UV-LED and compared with
that by a solar simulator equipped with a Xe lamp. The quartz
photochemical reaction cell, which was custom-made at our
university, was 80 mm in diameter and 12 mm thick (optical path).
The UV-LED was placed 10–15 mm above the top surface of the
reaction cell. All of the experiments were conducted in air-
saturated condition at room temperature.
A solar simulator (81160-1000, Oriel, USA) equipped with a
300-W ozone-free Xe lamp was also used as the light source. The
light intensity of the solar simulator was estimated using chemical
actinometry based on the photochemical degradation of 2-
nitrobenzaldehyde (2-NB). The degradation rate of 2-NB by the
solar simulator was estimated to be 0.0042 s
1. The degradation
rate of 2-NB by sunlight on October 7, 2002 at Higashi-Hiroshima
(at noon under clear sky conditions) was 0.0093 s
1, suggesting
that the light intensity of the solar simulator was about 45% of
natural sunlight at noon under a clear sky.
2.4. Determination of the rate constants of the reaction between
hydroxyl radicals and target compounds based on the competition
method
In a sample mixture including the target compound (X) and TP
as a probe compound (Supplementary content), the fraction of
hydroxyl radicals that reacted with TP in the competition with the
target compound was,
kTP;OH ½TP
f OH ¼ X ; ð1Þ
kTP;OH ½TP þ kX;OH ½X þ kScav
i;OH ½Scav
i
where kTP,OH and [TP] are the reaction rate constant and
concentration of TP, respectively, and kX,OH and [X] are the reaction
rate constant and concentration of the target compound,
respectively. kScav-i,OH and [Scav-i] are the reaction rate constant
and concentration of hydroxyl radical scavengers other than TP and
the target compound, respectively. The third term of the
P
denominator, kScav-i [Scav-i], is the subtotal of the scavenging
rate by all of the scavengers in the reaction mixture except TP and
the target compound, which include the hydroxyl radical source
(nitrite), components of the buffer solution (when used) and other
unknown scavengers in the reaction mixture. After generation of
10 K. Takeda et al. / Journal of Photochemistry and Photobiology A: Chemistry 340 (2017) 8–14

hydroxyl radicals by UV-LED irradiation of nitrite, the fluorescence

signal from HTP formed from TP and hydroxyl radicals is
F = ø YfROHfOH, (2)
where ROH is the amount of hydroxyl radicals generated, Yf is the
yield of HTP formed by the reaction of TP with hydroxyl radicals,
and ø is the coefficient of fluorescence detection. When F0 is the
fluorescence signal without target compound, i.e. [X] = 0, the
following relations can be deduced:
kTP;OH ½TP
F ¼ fY f ROH X ; ð3Þ
kTP;OH ½TP þ kX;OH ½X þ kScav
i;OH ½Scav
i

kTP;OH ½TP
F0 ¼ fY f ROH X ; ð4Þ
kTP;OH ½TP þ kScav
i;OH ½Scav
i

F0 kX;OH ½X 
¼1þ X : ð5Þ
F kTP;OH ½TP þ kScav
i;OH ½Scav
i

When it is assumed that the subtotal of the scavenging rate by the

scavengers except TP and the target compounds is negligible
P
compared with that of TP, i.e. kTP,OH [TP] >> kScav-i,OH [Scav-i],
Eq. (5) can be approximated as Fig. 1. Schematic diagrams of the (a) fluorescence flow measurement system with a
UV-LED and (b) U-shaped quartz glass tube. The U-shaped quartz glass tube was
F0 kX;OH ½X installed in the silver mirror-coated glass tube (dashed line in (a)), which reflected
¼1þ ð6Þ
F kTP;OH ½TP the UV-LED emission.

Eq. (6) indicates that a plot of F0/F against [X] at constant [TP]
Fluorescence signal was obtained from peak area as the mean of
should be a straight line with a y-intercept of 1 and a slope of kX,OH/
three independent injections of different mixtures at each
(kTP,OH [TP]). The reaction rate constant of the target compound
concentration of the target compound. All of the experiments
with hydroxyl radicals kX,OH can then be simply calculated from the
were conducted in air-saturated condition at room temperature.
slope of this plot, kTP,OH and [TP]. We chose a kTP,OH of 4.0  109
M
1 s
1 [21] to calculate the rate constants of target compounds.
2.6. Other analyses

2.5. UV-LED fluorescence flow measurement system

UV-vis absorption spectra were obtained using a Shimadzu UV-
2400 spectrophotometer (Kyoto, Japan) with a 10-mm optical-pass
The fluorescence signal of each sample mixture containing a
quartz cuvette. A pH meter and micro-pH electrode (D51 and
target compound and TP with nitrite was obtained using a semi-
9618S-10D, Horiba, Kyoto, Japan, respectively) were employed to
automatic fluorescence flow measurement system equipped with
measure pH.
the UV-LED, irradiation cell, sample injection valve, and fluores-
cence spectrometer. A schematic diagram of the system is shown in
3. Results
Fig. 1. A high-performance liquid chromatography (HPLC) plunger
pump (PU-980i, Jasco, Tokyo, Japan) was used to propel the Milli-Q
3.1. Hydroxyl radical formation by UV-LED irradiation
water employed as a carrier solution at a flow rate of 1.0 mL min
1.
The sample injection valve (9725i, Rheodyne, CA, USA) was
To characterize hydroxyl radical formation by the UV-LED,
equipped with a U-shaped quartz glass tube for UV-LED irradia-
batch photochemical experiments using the quartz photochemical
tion. The U-shaped quartz glass tube (o.d. = 3.0 mm, i.d. = 1.2 mm,
reaction cell were conducted with nitrite, nitrate, and hydrogen
total length of about 240 mm, and calculated total volume of about
peroxide as the photochemical hydroxyl radical source. The
270 mL) shown in Fig. 1(b) was custom-made at Hiroshima
formation of HTP from TP and hydroxyl radicals increased linearly
University. The U-shaped quartz glass was installed in a silver-
with irradiation time up to 20 min and with the concentration of
coated mirror glass reflector (o.d. = 25 mm, i.d. = 23 mm, and
nitrite, nitrate, and hydrogen peroxide. Table 1 lists the photo-
length = 110 mm). The fluorescence from HTP was detected using
chemical formation rates of HTP. The HTP formation rate is linearly
a HPLC fluorescence detector (FR-10AXL, Shimadzu, Kyoto, Japan)
proportional to the hydroxyl radical formation rate [15,17,20–22],
operated at 320 nm for excitation and 440 nm for emission.
and HTP yield was also systematically investigated as a function of
The standard procedure to obtain measurements was as
the pH and temperature of the aqueous solution [21]. The results in
follows. 1) Prepare sample mixtures containing a series of
Table 1 demonstrate that the hydroxyl radical formation rate from
concentrations of the target compound with 0.05 mM TP as a
nitrite by the UV-LED was about four orders of magnitude larger
hydroxyl radical probe and 0.1 mM nitrite as the photochemical
than that from nitrate and two orders of magnitude larger than
hydroxyl radical source just before measurements. The concentra-
that from hydrogen peroxide. In a previous study using a solar
tion range of the target compound depends on kX,OH, 2) Inject a
simulator and natural sunlight, the hydroxyl radical formation rate
sample mixture via injection valve in the “load” position, 3) Turn
of nitrite was estimated to be around 100 times larger than that of
on the UV-LED, 4) Turn off the UV-LED after 3 min of irradiation,
nitrate and 10 times larger than that of hydrogen peroxide [3].
and 5) Immediately switch injection valve to the “inject” position
Thus, the hydroxyl radical formation induced by the UV-LED was
and start to acquire fluorescence signal from the detector.
minor for nitrate and hydrogen peroxide, compared with that of
 K. Takeda et al. / Journal of Photochemistry and Photobiology A: Chemistry 340 (2017) 8–14 11

Table 1
Formation rates of 2-hydroxyterephthalate from nitrite, nitrate and hydrogen peroxide induced by the UV-LED and a solar simulator equipped with a 300-W Xe lamp.
Concentration ranges were 0–4 mM for nitrite, 0–400 mM for nitrate and 0–400 mM for hydrogen peroxide.

Hydoxyl radical source Formation rates of 2-hydroxyterephthalate

Light source

UV-LED Solar simulator

NO2
8.6  10
5 M s
1 M(NO2
)
1 (1.00)a 5.0  10
6 M s
1 M(NO2
)
1
H2O2 1.1 10
6 M s
1 M(H2O2)
1 (0.013)a –
NO3
4.2  10
9 M s
1 M(NO3
)
1 (0.00005)a –
a
Ratios with formation rate of nitrite.

nitrite. The UV–vis absorption spectra of nitrite, nitrate, and
hydrogen peroxide along with the emission spectrum of the UV-
LED are shown in Fig. 2(a). The absorption spectrum of nitrite
overlapped well with the emission spectrum of the UV-LED. It is
well known that the spectrum of natural sunlight is a continuum
spectrum above 300 nm. Therefore, the specific formation of
hydroxyl radicals by the UV-LED can be attributed to the spectral
features (emission wavelength) of the UV-LED used in this study.
The photolysis of nitrite by the UV-LED can be used as an efficient
and selective hydroxyl radical source in aqueous media.
The absorption spectra of TP and HTP dissolved in Milli-Q water
are depicted in Fig. 2(b). Their absorption bands did not overlap
with the emission of the UV-LED. Page et al. determined the
hydroxyl radicals generated photochemically using a Hg lamp with
a 320-nm cutoff filter [22]. They found that HTP generated from TP
and hydroxyl radicals photochemically decomposed during
irradiation, and concluded that TP is an unsuitable probe for
studying photochemical systems under UV irradiation with
wavelengths shorter than 360 nm. However, the light emission
of the UV-LED used in this study should be spectroscopically inert
to HTP because of their spectral features, as shown in Fig. 2(b). We
carried out photochemical degradation experiments of 0.1 mM HTP
in Milli-Q water at neutral pH with UV-LED irradiation for 3 min.
The results indicated that there was no statistical difference in
fluorescence signal between irradiated and non–irradiated HTP.
Therefore, the photochemical degradation of HTP was negligible in
our system.
3.2. Methodology development with a UV-LED fluorescence flow
measurement system
In the fluorescence flow measurement system with the UV-LED,
the signal intensities of 2.0 mM TP solution after 3.0 min of
irradiation without a target compound were linearly proportional
to the nitrite concentration in 0.05–5 mM nitrite. The signal
intensities of 0.05 mM TP solution with 0.1 mM nitrite were linearly
proportional to the irradiation time up to 10 min. The relative
standard deviation of signal intensity for five measurements of
0.1 mM nitrite was 1.5%. This result demonstrates that 0.1 mM
nitrite with 3 min of irradiation provided strong enough signals to
detect the photochemically formed hydroxyl radicals. Therefore,
the concentration of nitrite used as the hydroxyl radical source was
0.1 mM with 3 min of irradiation.
The competition method can be realized if the concentrations of
both the probe compound (TP) and target compound can be
considered as constants during irradiation. The formation of HTP
generated from 0.05 mM TP, 0.1 mM nitrite, and irradiation for
3 min in the absence of target compound was 9.4 nM, which is
about 0.02% of the initial TP. This indicates that the concentration
of TP during the irradiation could be considered as constant at TP
concentrations exceeding 0.05 mM.
Eq. (6) was used to determine the reaction rate constant
between the target compound and hydroxyl radicals, when the
scavenging rate of the scavengers other than TP and the target
compound is sufficiently smaller than those of TP. In this study, the
scavenging by nitrite as a photochemical hydroxyl radical source
and components of phosphate buffer solution (when used) should
be considered. It is well known that nitrite is an effective scavenger
of hydroxyl radicals with reaction rate constants of 0.6–1.4  1010
M
1 s
1 [30,31] while it is also a photochemical source of hydroxyl
radicals. As shown in Table 2, the calculated scavenging rate of
0.1 mM nitrite was 0.6–1.4  103 s
1, which is negligible compared
with that by TP under the conditions employed. We sometimes
used phosphate buffer (pH 6–8) to stabilize the pH of reaction
mixtures. In Table 2, the rate constants and calculated scavenging
rates by components in the phosphate buffer solution are
summarized. Because the pKa values of phosphoric acid are

Table 2
Estimated scavenging rates of hydroxyl radicals by chemical components in the
reaction mixture.

Component kX,OH/M
1 s
1 Scavenging rate/s
1

0.05 mM TP 4.0  109 [21] 2  105
2.0 mM TP 4.0  109 [21] 8  106
Fig. 2. (a) UV–vis absorption spectra of nitrite (10 mM, red solid line), nitrate 0.1 mM NO2
0.6–1.4  1010 [30,31] 0.6–1.4  103
(10 mM, green solid line), and hydrogen peroxide (10 mM, black dashed line), and 10 mM H3PO4 4.2  104 [32] 4.2  102
(b) UV–vis absorption spectra of HTP (0.1 mM, red solid line) and TP (0.1 mM, black 10 mM H2PO4
2  104 [33] 2  102
dashed line). The blue-solid lines in both panel (a) and (b) show the emission 10 mM HPO42
1.5  105 [33] 1.5  103
spectrum of the UV-LED from the specification sheet published by Nitride 10 mM PO43
<1 107 [34] 1 105
Semiconductors Co. Ltd.
12 K. Takeda et al. / Journal of Photochemistry and Photobiology A: Chemistry 340 (2017) 8–14
pKa1 = 2.1, pKa2 = 7.2 and pKa3 = 12.3, the scavenging rate of 10 mM
phosphate buffer at pH 6–8 was roughly predicted to be less than
<1.5  103 s
1. Thus, the hydroxyl radical scavenging by phosphate
buffer was also negligible. These results demonstrate that almost
all of the hydroxyl radicals generated reacted with TP or the target
compound rather than nitrite or components of the buffer solution,
suggesting that the reaction rate constant can be reliably
calculated by Eq. (6).
3.3. Determination of the reaction rate constant between the target
compound and hydroxyl radicals
The reaction rate constant between methanol and hydroxyl
radicals was determined. Fig. 3 plots F0/F against methanol
concentration. The plot was a straight line passing through a y-
intercept of 1. The reaction rate constants estimated from the slope
of the plot and concentration of TP were 0.85  109 M
1 s
1 for
0.05 mM TP, 0.84  109 M
1 s
1 for 0.5 mM TP and 0.88  109 M
1
s
1 for 1.0 mM TP, respectively. The reaction rate constants
obtained in 10 mM phosphate buffer (pH = 7.0) with 0.05 mM TP
was 0.83  109 M
1 s
1. These values were independent of the
concentration of TP in the range 0.05–1.0 mM and the presence of
buffer. The reported values summarized on the NDRL/NIST website
were 0.78–1.2  109 M
1 s
1 [19,35–38], indicating that the results
obtained using the proposed method agreed well with the values
reported previously.
We determined the reaction rate constants between hydroxyl
radicals and target compounds using the developed system. Some
low-molecular-weight organic compounds and inorganic halide
salts with various reaction rate constants were chosen. The results
are summarized in Table 3. The obtained values were compared
with the literature values quoted by the NDRL/NIST Solution
Kinetics Database Resources [19] and those by Louit et al. [16]. The
obtained reaction rate constants ranged from 105 to 1010 M
1 s
1,
and were in good agreement with the previously reported values,
except for chloride ions, bromide ions, cysteine, and nicotinamide.
The reaction rate constants for cysteine were more than 3 times
lower than the reported value. In aqueous solution, cysteine is
oxidized by air to cystine with a reaction rate constant of
2.1 109 M
1 s
1 [41]. This is one of the reasons for the lower
rate constants for cysteine than previously reported. For nicotin-
amide, it is unknown why a lower rate constant was obtained in
this study than elsewhere.
The reported reaction rate constants for chloride ions measured
by a pulse radiolysis method are 3.0–4.3  109 M
1 s
1 [19].
However, Arakaki et al. determined that the experimental net
reaction rate constant of chloride ions and hydroxyl radicals was
Fig. 3. Plots of F0/F against methanol concentration in the different concentration of
TP 0.05 mM (blue), 0.5 mM (green) and 1.0 mM (red). The concentration of nitrite
was 0.1 mM with 3 min of irradiation. Data plot and error bar in figure show the
means and standard deviation of three independent measurements of different
mixtures.
2.7  107 M
1 s
1 by using a photochemical steady-state method
[42], which was two orders of magnitude smaller than the previous
values. In our system, fluorescence signals were clearly detected,
and the plot against [Cl
] ([Cl
] = 0–0.9 M, [NO2
] = 0.1 mM and
[TP] = 0.05 mM) was a straight line passing through the y-intercept
at 1. The rate constants obtained for KCl and NaCl were two orders
of magnitude lower than those of Arakaki et al., and four orders of
magnitude lower than those measured by pulse radiolysis. Because
high concentrations of NaCl and KCl were used in our experiments,
the reaction rate constants obtained may include some unexpected
effects. However, the huge difference in the rate constants of
chloride ions may be explained by their fast back reaction as well as
forward reaction [42,43]. The results indicate that the apparent
reaction rate constant between chloride ions and hydroxyl radicals
could be quite low. At least, the results suggest that chloride ions
are inert in hydroxylation of TP by hydroxyl radicals, which
probably involves hydroxylation of the aromatic ring. The rate
constant determined for KBr was about one order of magnitude
lower than the values on the NDRL/NIST website. This discrepancy
is also associated with the fast back reaction between bromide ions
and hydroxyl radicals. Zehavi and Rabani estimated that the
apparent rate constant for the reaction OH + Br
! OH
+ Br was
1.0–1.9  109 M
1 s
1 when [Br
] < 10 mM [44], which is consistent
with our result. The results obtained here for chloride ions and
bromide ions provide important information to discuss hydroxyl
radical quenching in seawater.
4. Discussion
The reaction rate constants of hydroxyl radicals were deter-
mined successfully using the method developed in this study. The
important feature of the presented method is the hydroxyl radical
generation system using nitrite and a UV-LED. Because the
emission spectrum of the UV-LED employed overlapped well with
the absorption spectrum of nitrite, efficient and selective genera-
tion of hydroxyl radicals could be achieved with 0.1 mM nitrite. In
the Fenton system, which is commonly used to generate hydroxyl
radicals, Fe(II) ion or its complex compounds scavenge hydroxyl
radicals in the reaction mixture. For example, 0.03–1 mM Fe2+ with
an equivalent concentration of EDTA was used in previous studies
[15,17]. Because the reaction rate constants of [Fe(EDTA)]2
are
0.5–1.1 1010 M
1 s
1 [39,40], the scavenging rate by [Fe
(EDTA)]2
was estimated to be 0.15–11 106 s
1. In the Fenton
system without EDTA [18], the scavenging rate by 1 mM Fe(II) (kFe2
8
+,OH = 2.3–4.3  10 M

1
1
s [19]) was estimated to be 0.23–
6
1
0.43  10 s . In these systems, hydroxyl radical scavenging by
the hydroxyl radical source may not be negligible. To avoid
hydroxyl radical quenching by components in the reaction mixture
other than the probe and target compounds, a probe compound
with a high reaction rate should be used (see Eq. (5)). However,
under such conditions, it is difficult to determine the rate constant
of compounds with a low rate constant and/or low solubility. As
shown in Table 2, hydroxyl radical scavenging by nitrite was
negligible in our system. When the lowest limit of estimation for a
reaction rate constant is defined as F0/F = 1.10, in which 10% of
hydroxyl radicals are quenched by the target compound with
0.05 mM TP, it is estimated that kX,OH = 2  104 M
1 s
1 at
[X] = 1.0 M, kX,OH = 2  105 M
1 s
1 at [X] = 0.1 M and kX,OH = 2  106
M
1 s
1 at [X] = 0.01 M. Furthermore, the system with nitrite and a
UV-LED can avoid unexpected reactions of ferrous ions in the
reaction mixture, such as reduction and complex formation.
Radiolysis is the simplest method to generate hydroxyl radicals
without additional reagents, and it avoids unexpected reactions
and quenching of hydroxyl radicals. However, a gamma-ray
radiation source is not a common instrument, and N2O gas is
essential to eliminate the reactions of hydrated electrons
 K. Takeda et al. / Journal of Photochemistry and Photobiology A: Chemistry 340 (2017) 8–14 13

Table 3
Reaction rate constants of hydroxyl radicals determined in this study with literature values for comparison. All of the results were obtained under air-saturated condition and
at room temperature.

Substance Reaction rate constant/109 M
1 s

This study valuea Literature value

In Milli-Q water In phosphate buffer (pH)b NDRL/NIST [19] Louit et al. [16]

Potassium thiocyanate 10.2  1.4 (pH = 5.8) (n = 3) 10.3  0.5 (pH = 6.8) (n = 4) 9.6–13 10
Thiourea 7.5 (pH = 5.6) (n = 1) 3.9–9.8 10.1
Cysteine 7.5, 6.4 (pH = 5.3) (n = 2) 6.9, 4.9 (pH = 6.9) (n = 2) 19–47 23.3
Thymine 4.9  0.6 (pH = 5.3) (n = 3) 5.3  0.4 (pH = 7.0) (n = 4) 4.6–8.7 7.2
Sodium formate 3.0 (pH = 5.9) (n = 1) 2.1–5.1
Acetophenone 3.7 (pH = 6.2) (n=1) 3.3 (pH = 6.8) (n = 1) 5.4–6.5
Glucose 2.0, 1.8 (pH = 6.0) (n = 2) 1.6, 1.8 (pH = 6.9) (n = 2) 1.0–2.3
Ethanol 2.1  0.4 (pH = 6.1) (n=4) 1.4  0.1 (pH = 6.5) (n = 4) 1.6–2.8 2.5
2-Propanol 2.0  0.2 (pH = 5.5) (n = 3) 1.8  0.3 (pH = 6.9) (n = 5) 1.6–2.3 2.2
Glycerine 1.5  0.4 (pH = 5.9) (n = 3) 1.5–2.1
Ethylene glycol 1.5, 1.4 (pH = 6.0) (n = 2) 1.4–2.4
Nicotinamide 0.32, 0.30 (pH = 6.2) (n = 2) 0.35, 0.43 (pH = 6.9) (n = 2) 1.4–1.5
Methanol 0.84  0.01 (pH = 5.8) (n = 3) 0.83  0.02 (pH = 7.0) (n = 3) 0.78–1.2 1.2
Acetone 0.12  0.01 (pH = 5.8) (n = 4) 0.099  0.004 (pH = 7.0) (n = 3) 0.083–0.14
Sodium acetate 0.073, 0.070 (pH = 6.6) (n = 2) 0.078, 0.075 (pH = 7.1) (n = 2) 0.074–0.1
Alanine 0.070  0.009 (pH = 5.8) (n = 3) 0.072  0.010 (pH = 7.0) (n = 3) 0.065–0.43
Acetonitrile 0.0095  0.0004 (pH = 6.5) (n = 3) 0.022

NaI 11, 10 (pH = 5.6) (n = 2) 10–12 (I
)

KBr 1.8, 1.6 (pH = 5.7) (n = 2) 11 (Br
)
KCl 0.0001 (pH = 6.4) (n = 1) 3.0–4.3 (Cl
)
NaCl 0.0002 (pH = 6.4) (n = 1) 3.0–4.3 (Cl
)
a
Average  standard deviation of replicate (n > 3).
b
1–10 mM phosphate buffer.

generated by radiolysis of water. Our system is easy to assemble
anywhere at low cost.
In our system, the photolysis of nitrite was utilized. The
photolysis of nitrite involves the formation of nitric oxide, complex
chemistry of nitrite ions and the photonitration or photonitrosa-
tion of organic compounds, as well as the photochemical
generation of hydroxyl radical [7–9,45,46]. In these reactions,
nitric oxide is primary component during nitrite photolysis.
Reaction rate constant between hydroxyl radical and nitric oxide
is 1 1010 M
1 s
1 [45], which is relatively high. However, the
lifetime of nitric oxide in environmental water was 0.05–1.3 s,
indicating that the reactivity of nitric oxide were relatively low
comparing with the reactions between hydroxyl radical and
chemical species in natural water. The steady state concentration
of nitric oxide was reported to be around 10
11 M [9]. The
scavenging rate of hydroxyl radical by nitric oxide can be estimated
to be only 10
1 s
1, which is quite low comparing with the results
shown in Table 2. The quantitative effects of nitrogen species
involved from nitrite photolysis in our system would be negligible.
In this study, experiments were conducted under constant
intensity of UV-LED. The light intensity of the UV-LED would be
stable, since the UV-LED was driven in the constant-current mode
on the aluminum heat sink with a cooling fan. As we mention in
Section 3.2, all of the experiments in this study were carried out in
“linear range” of irradiation time and nitrite concentration. It was
supposed that the hydroxyl radical formation is also linear, at least
up to the light intensity employed in this experiments. suggesting
that the light intensity dose not affected the result of the reaction
rate constants.
This hydroxyl radical generation system using nitrite and a UV-
LED will be useful for disinfection, degradation, and hydroxylation
systems in aqueous media. Because it is anticipated that UV-LEDs
will become more efficient, powerful and cheaper in the future,
systems with a UV-LED coupled with nitrite are promising to
generate hydroxyl radicals in aqueous solution.
In conclusion, we have developed a simple method to
determine the hydroxyl radical reaction rate constants using the
high-power UV-LED. In our method, reaction rate constants ranged
from 105 to 1010 M
1 s
1 were obtained without gamma ray and the
Fenton reagent. Results obtained agreed well with the previous
reported. Because rate constants determined in here are apparent
reaction rate constants based on steady–state method, the
difference of chloride ion between our method and gamma pulse
radiolysis should be attributed the fast back reaction.
Acknowledgments
We thank the technical staffs of the Glass Blowing Factory in the
Innovation Plaza of Hiroshima University. This work was supported
by JSPS KAKENHI Grant Number 16K00516.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.
jphotochem.2017.02.020.
References
[1] K. Mopper, X. Zhou, Hydroxyl radical photoproduction in the sea and its
potential impact on marine processes, Science 250 (1990) 661.
[2] P.P. Vaughan, N.V. Blough, Photochemical formation of hydroxyl radical by
constituents of natural waters, Environ. Sci. Technol. 32 (1998) 2947.
[3] K. Takeda, H. Takedoi, S. Yamaji, K. Ohta, H. Sakugawa, Determination of
hydroxyl radical photoproduction rates in natural waters, Anal. Sci. 20 (2004)
153.
[4] A.J. Anifowose, K. Takeda, H. Sakugawa, A novel fluorometric method for the
determination of production rate and steady-state concentration of
photochemically generated superoxide radical in seawater using 3',6'-
(diphenylphosphinyl)fluorescein (PF-1), Anal. Chem. 87 (2015) 11998.
[5] K. Fujiwara, T. Ushiroda, K. Takeda, Y. Kumamoto, H. Tsubota, Diurnal and
seasonal distribution of hydrogen peroxide in seawater of Seto Inland Sea,
Geochem. J. 27 (1993) 103.
14 K. Takeda et al. / Journal of Photochemistry and Photobiology A: Chemistry 340 (2017) 8–14
[6] W.L. Miller, D.R. Kester, Hydrogen-peroxide measurement in seawater by
(para-hydroxyphenyl)acetic acid dimerization, Anal. Chem. 60 (1988) 2711.
[7] O.C. Zafiriou, M. McFarland, Nitric oxide from nitrite photolysis in the central
equatorial pacific, J. Geophys. Res. 86 (1981) 3173.
[8] F. Olasehinde, K. Takeda, H. Sakugawa, Photochemical production and
consumption mechanisms of nitric oxide in seawater, Environ. Sci. Technol.
44 (2010) 8403.
[9] A.J. Anifowose, K. Takeda, H. Sakugawa, Photoformation rate, steady-state
concentration and lifetime of nitric oxide radical (NO) in a eutrophic river in
Higashi-Hiroshima, Japan, Chemosphere 119 (2015) 302.
[10] N. Nakatani, M. Ueda, H. Shindo, K. Takeda, H. Sakugawa, Contribution of the
photo-fenton reaction to hydroxyl radical formation rates in river and rain
water samples, Anal. Sci. 23 (2007) 1137.
[11] N. Nakatani, N. Hashimoto, H. Sakugawa, An experimental study of
photodegradation of bisphenol A in water: reaction with hydroxyl radical,
Geochem. Soc. Spec. Publ. 9 (2004) 233.
[12] P. Calza, D. Vione, A. Novelli, E. Pelizzetti, C. Minero, The role of nitrite and
nitrate ions as photosensitizers in the phototransformation of phenolic
compounds in seawater, Sci. Total Environ. 439 (2012) 67.
[13] T. Hayashi, Y. Okamoto, K. Ueda, N. Kojima, Formation of estrogenic products
from benzophenone after exposure to sunlight, Toxicol. Lett. 167 (2006) 1.
[14] M.S. Diaz-Cruz, M. Llorca, Marta D. Barcelo, Organic UV filters and their
photodegradates, metabolites and disinfection by-products in the aquatic
environment, Trends Anal. Chem. 27 (2008) 873.
[15] X.-F. Yang, X.-Q. Guo, Fe(II)–EDTA chelate-induced aromatic hydroxylation of
terephthalate as a new method for the evaluation of hydroxyl radical-
scavenging ability, Analyst 126 (2001) 928.
[16] G. Louit, M. Hanedanian, F. Taran, H. Coffigny, J.P. Renault, S. Pin, Determination
of hydroxyl rate constants by a high-throughput fluorimetric assay: towards a
unified reactivity scale for antioxidants, Analyst 134 (2009) 250.
[17] B. Bekdeser, M. Ozyurek, K. Guclu, R. Apak, Novel spectroscopic sensor for the
hydroxyl radical scavenging activity measurement of biological samples,
Talanta 99 (2012) 689.
[18] F. Yu, D. Xu, R. Lei, N. Li, K. Li, Free-Radical Scavenging Capacity Using the
Fenton reaction with rhodamine B as the spectrophotometric indicator, J.
Agric. Food Chem. 56 (2008) 730.
[19] NDRL/NIST Solution Kinetics Database on the Web. <http://kinetics.nist.gov/
solution//>.
[20] M. Saran, K.H. Summer, Assaying for hydroxyl radicals: hydroxylated
terephthalate is a superior fluorescence marker than hydroxylated
benzoate, Free Radic. Res. 31 (1999) 429.
[21] T. Charbouillot, M. Brigante, G. Mailhot, P.R. Maddigapu, C. Minero, D. Vione,
Performance and selectivity of the terephthalic acid probe for OH as a function
of temperature, pH and composition of atmospherically relevant aqueous
media, J. Photochem. Photobiol. A 222 (2011) 70.
[22] S.E. Page, W.A. Arnold, K. McNeill, Terephthalate as a probe for
photochemically generated hydroxyl radical, J. Environ. Monit. 12 (2010) 1658.
[23] G. Baldacchino, T. Maeyam, S. Yamashita, M. Taguchi, A. Kimura, Y. Katsumura,
T. Murakami, Determination of the time-dependent OH-yield by using a
fluorescent probe. Application to heavy ion irradiation, Chem. Phys. Lett. 468
(2009) 275.
[24] D.K. Das, G.A. Cordis, P.S. Rao, X.K. Liu, S. Marty, High-performance liquid
chromatographic detection of hydroxylated benzoic acids as an indirect
measure of hydroxyl radical in heart: its possible link with the myocardial
reperfusion injury, J. Chromatogr. 536 (1991) 273.
[25] L. Sun, H. Chen, H.A. Abdulla, K. Mopper, Estimating hydroxyl radical
photochemical formation rates in natural waters during long-term
laboratory irradiation experiments, Environ. Sci. Process. Impacts 16 (2014)
757.
[26] S. Uzunboy, S.D. Cekic, E. Eksin, A. Erdem, R. Apak, CUPRAC colorimetric and
electroanalytical methods determining antioxidant activity based on
prevention of oxidative DNA damage, Anal. Biochem. 518 (2017) 69–77.
[27] N. Loganayaki, P. Siddhuraju, S. Manian, Antioxidant activity and free radical
scavenging capacity of phenolic extracts from Helicteres isora L. and Ceiba
pentandra L, J. Food Sci. Technol. 50 (2013) 687–695.
[28] O.C. Zafiriou, M.B. True, Nitrite photolysis as a source of free radicals in
productive surface waters, Geophys. Res. Lett. 6 (1979) 81.
[29] A. Treinin, E. Hayon, Absorption spectra and reaction kinetics of NO2 N2O3,
and N2O4 in aqueous solution, J. Am. Chem. Soc. 92 (1970) 5821.
[30] T. Logager, K. Sehested, Formation and decay of peroxynitrous acid: a pulse
radiolysis study, J. Phys. Chem. 97 (1993) 6664.
[31] G.V. Buxton, Pulse radiolysis of aqueous solutions. Some rates of reaction of OH
and O- and pH dependence of the yield of O3-, Trans. Faraday Soc. 65 (1969)
2150.
[32] P.-Y. Jiang, Y. Katsumura, M. Domae, K. Ishikawa, R. Nagaishi, K. Ishigure, Y.
Yoshida, Pulse radiolysis study of concentrated phosphoric acid solutions, J.
Chem. Soc. Faraday Trans. 88 (1992) 3319.
[33] P. Maruthamuthu, P. Neta, Phosphate radicals. Spectra, acid-base equilibria,
and reactions with inorganic compounds, J. Phys. Chem. 82 (1978) 710.
[34] E.D. Black, E. Hayon, Pulse radiolysis of phosphate anions H2PO4- HPO42-,
PO43-, and P2O74- in aqueous solutions, J. Phys. Chem. 74 (1978) 3199.
[35] N. Motohashi, Y. Saito, Competitive measurement of rate constants for
hydroxyl radical reactions using radiolytic hydroxylation of benzoate, Chem.
Pharm. Bull. 41 (1993) 1842.
[36] G.V. Buxton, C.L. Greenstock, W.P. Helman, A.B. Ross, Critical review of rate
constants for reactions of hydrated electrons, hydrogen atoms and hydroxyl
radicals (OH/O-) in aqueous solution, J. Phys. Chem. Ref. Data 17 (1988) 513.
[37] B.S. Wolfenden, R.L. Willson, Radical-cations as reference chromogens in
kinetic studies of one-electron transfer reactions: pulse radiolysis studies of 2
2'-azinobis-(3-ethylbenzthiazoline-6-sulphonate), J. Chem. Soc. Perkin Trans.
2 (7) (1982) 805.
[38] G.E. Adams, J.W. Boag, B.D. Michael, Trans. Faraday Soc. 61 (1965) 1417.
[39] J. Lati, D. Meyerstein, Oxidation of first-row bivalent transition-metal
complexes containing ethylenediaminetetra-acetate and nitrilotriacetate
ligands by free radicals: a pulse-radiolysis study, J. Chem. Soc. Dalton Trans.
1978 (1978) 1105.
[40] M.J. Zhao, L. Jung, C. Tanielian, R. Mechin, Kinetics of the competitive
degradation of deoxyribose and other biomolecules by hydroxyl radicals
produced by the Fenton reaction, Free Radic. Res. 20 (1994) 345.
[41] T. Masuda, S. Nakano, M. Kondo, Rate constants for the reactions of OH radicals
with the enzyme proteins as determined by the p-nitrosodimethylaniline
method, J. Radiat. Res. 14 (1973) 339.
[42] T. Arakaki, T. Miyake, M. Shibata, H. Sakugawa, Photochemical formation and
scavenging of hydroxyl radical in rain and dew waters, Nihon Kagaku Kaishi
1999 (1999) 335 (in Japanese).
[43] G.G. Jayson, B.J. Parsons, A.J. Swallow, Some simple, highly reactive, inorganic
chlorine derivatives in aqueous solution, J. Chem. Soc. Faraday Trans. 1 (9)
(1973) 1597.
[44] D. Zehavi, J. Rabani, The oxidation of aqueous bromide ions by hydroxyl
radicals: a pulse radiolytic investigation, J. Phys. Chem. 76 (1972) 1079.
[45] J. Mack, J.R. Bolton, Photochemistry of nitrite and nitrate in aqueous solution: a
review, J. Photochem. Photobiol. A 128 (1999) 1.
[46] E. De Laurentiis, M. Minella, S. Berto, V. Maurino, C. Minero, D. Vione, The fate
of nitrogen upon nitrite irradiation: formation of dissolved vs. gas-phase
species, J. Photochem. Photobiol. A 307–308 (2015) 30.