Laser flash photolysis study of the photoinduced oxidation of 4-(dimethylamino)benzonitrile (DMABN)

Article abstract of DOI:10.1039/c8pp00519b

Aromatic amines are aquatic contaminants for which phototransformation in surface waters can be induced by excited triplet states of dissolved organic matter (³DOM*). The first reaction step is assumed to consist of a one-electron oxidation process of the amine to produce its radical cation. In this paper, we present laser flash photolysis investigations aimed at characterizing the photoinduced, aqueous phase one-electron oxidation of 4-(dimethylamino)benzonitrile (DMABN) as a representative of this contaminant class. The production of the radical cation of DMABN (DMABN⁺) after direct photoexcitation of DMABN at 266 nm was confirmed in accord with previous experimental results. Moreover, DMABN⁺ was shown to be produced from the reactions of several excited triplet photosensitizers (carbonyl compounds) with DMABN. Second-order rate constants for the quenching of the excited triplet states by DMABN were determined to fall in the range of 3 × 10⁷-5 × 10⁹ M^?1 s^?1, and their variation was interpreted in terms of electron transfer theory using a Rehm-Weller relationship. The decay kinetics of DMABN⁺ in the presence of oxygen was dominated by a second-order component attributed to its reaction with the superoxide radical anion (O₂^?). The first-order rate constant for the transformation of DMABN⁺ leading to photodegradation of DMABN was estimated not to exceed ≈5 × 10³ s^?1

Full text of DOI:10.1039/c8pp00519b

Photochemical &
Photobiological Sciences
View Article Online
View Journal
PAPER
Laser flash photolysis study of the photoinduced
oxidation of 4-(dimethylamino)benzonitrile
Cite this: DOI: 10.1039/c8pp00519b
(DMABN)†
c
Frank Leresche,
‡
a,b Lucie Ludvíková,c Dominik Heger, *c Petr Klán,
a
Urs von Gunten a,b and Silvio Canonica
*
Aromatic amines are aquatic contaminants for which phototransformation in surface waters can be
induced by excited triplet states of dissolved organic matter (3DOM*). The first reaction step is assumed to
consist of a one-electron oxidation process of the amine to produce its radical cation. In this paper, we
present laser flash photolysis investigations aimed at characterizing the photoinduced, aqueous phase
one-electron oxidation of 4-(dimethylamino)benzonitrile (DMABN) as a representative of this contaminant
class. The production of the radical cation of DMABN (DMABN•+) after direct photoexcitation of DMABN
at 266 nm was confirmed in accord with previous experimental results. Moreover, DMABN•+ was shown
to be produced from the reactions of several excited triplet photosensitizers (carbonyl compounds) with
DMABN. Second-order rate constants for the quenching of the excited triplet states by DMABN were
determined to fall in the range of 3 × 107–5 × 109 M−1 s−1, and their variation was interpreted in terms of
electron transfer theory using a Rehm–Weller relationship. The decay kinetics of DMABN•+ in the pres-
ence of oxygen was dominated by a second-order component attributed to its reaction with the super-
oxide radical anion (O2•−). The first-order rate constant for the transformation of DMABN•+ leading to
Received 14th November 2018,
Accepted 13th December 2018
DOI: 10.1039/c8pp00519b
rsc.li/pps
photodegradation of DMABN was estimated not to exceed ≈5 × 103 s−1
.
of higher plants or from microbial metabolic activity and ubi-
quitously present in natural waters.10–13 Upon photoirradia-
1. Introduction
Aromatic amines constitute an important class of organic con- tion, the chromophoric components of DOM are promoted to
taminants present in wastewaters and natural waters.1–4 One their excited singlet states and subsequently excited triplet
of the major pathways contributing to their abatement in the states (3DOM*), which are likely to be the key reactive inter-
aquatic environment is phototransformation induced by mediates in the photosensitized transformation of aromatic
sunlight.5–9 Besides direct phototransformation following amines in surface waters.8,14,15 The initial reaction step in the
absorption of sunlight by the contaminant, aromatic amines photosensitized transformation is assumed to be an electron
are subject to transformation photosensitized by dissolved transfer from an aromatic amine to 3DOM*.16 This can be
organic matter (DOM). The latter is a complex mixture of inferred from the abundant studies on the photoreduction of
cross-linked organic compounds derived from the degradation excited triplet carbonyls by amines,17 arguments on the ener-
getics of the redox reactions between substituted anilines and
excited triplet states of aromatic ketones in aqueous solu-
tion,14 and recent results on the quenching of the excited
triplet state of methylene blue by substituted anilines.15
Formation of a radical cation of the aromatic amine, a primary
product of such an electron transfer reaction, is expected.
Studies performed during the last decade in our research
group have revealed another important effect of DOM on the
photosensitized transformation of anilines, which consists in
reducing the rates of transformation. Reduction of transform-
ation intermediates of the anilines by electron-donating (also
called antioxidant) moieties of the DOM leading to reforma-
tion of the parent compound was hypothesized to cause such
aEawag, Swiss Federal Institute of Aquatic Science and Technology,
Überlandstrasse 133, CH-8600 Dübendorf, Switzerland.
E-mail: silvio.canonica@eawag.ch; Tel: +41-58-765-5453
bSchool of Architecture, Civil and Environmental Engineering (ENAC),
Ecole Polytechnique Fédérale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland
cDepartment of Chemistry and RECETOX, Faculty of Science, Masaryk University,
Kamenice 5, 62500 Brno, Czech Republic. E-mail: hegerd@chemi.muni.cz;
Tel: +420 54949 3322
†Electronic supplementary information (ESI) available: Comprising further
experimental details and results as well as kinetic models and simulations
descriptions. See DOI: 10.1039/c8pp00519b
‡Present address: Department of Civil, Environmental and Architectural
Engineering, University of Colorado, Boulder, CO 80309, USA.
This journal is © The Royal Society of Chemistry and Owner Societies 2018
Photochem. Photobiol. Sci.

View Article Online
Paper
Photochemical & Photobiological Sciences
an inhibition of transformation. Various pieces of evidence the hydrated electron (eaq−, very broad absorption centered at
have been presented to support this hypothesis.7,14,18–20
≈720 nm), and DMABN•+ (absorption maximum at ≈500 nm).
3
The effects of DOM on the overall phototransformation While DMABN* is formed through intersystem crossing from
rates of substituted anilines are now fairly well understood. the lowest excited singlet state of DMABN (1DMABN*, see
However, very little is known about the nature and fate of the eqn (1)) with substantial quantum yields (a value of 0.55 was
intermediates produced after the first oxidation step. The determined for an ethanol solution27), eaq− and DMABN•+ were
main goal of the present study is to characterize the kinetics of formed through photoionization of DMABN (eqn (2)). The
such intermediates derived from a model aniline in aqueous photoionization of DMABN was found to be an excitation wave-
solutions.
We
selected
4-(dimethylamino)benzonitrile length-dependent process mainly occurring at irradiation
(DMABN, see Chart 1; DMABN can be named alternatively N,N- wavelengths of <280 nm. This indicated that photoionization
dimethyl-4-cyanoaniline20) for various reasons. Firstly, DMABN did not involve 1DMABN* but some higher excited states of
was used in a preceding steady-state photolysis kinetic study as DMABN.27–29
a model compound to assess photosensitized transformation
1
3
DMABN þ hν ! DMABN* ! DMABN*
ð1Þ
ð2Þ
rates of aromatic amines and analogous compounds that
undergo DOM-induced inhibition of oxidation in surface
waters.20 Secondly, the results of that study showed that the
DOM-photosensitized transformation of DMABN very likely
proceeds by direct reaction between DMABN and 3DOM*,
whereby singlet oxygen plays a negligible role. Thirdly, the two
methyl substituents on the aniline group of DMABN make its
radical cation (DMABN•+, see Chart 1) more resistant to depro-
tonation than the radical cations of anilines bearing no or only
one substituent on the aniline group. Therefore, the relative
stability of DMABN•+ is expected to provide favorable con-
ditions for its direct observation in time-resolved experiments.
DMABN is the archetypal representative of aromatic com-
pounds bearing both an electron-donating and an electron-
accepting groups and exhibiting a dual fluorescence as well as
excited-state charge separation in solutions.21,22 The excited
singlet state characteristics and dynamics of such compounds,
in particular DMABN, have been a hot research topic in ultra-
fast spectroscopy for a few decades.21,23–26 Nevertheless, infor-
mation about the photophysics and photochemistry of
DMABN on a microsecond or longer timescale is limited. The
transient absorption spectrum observed on the microsecond
scale after laser flash photolysis (LFP) of an aqueous solution
of DMABN was interpreted as the superposition of various
species:27 these include the excited triplet state of DMABN
(3DMABN*, absorption maximum centered at 400 nm and a
secondary broad band with a maximum at ≈600 nm (ref. 27)),
ꢀ
DMABN þ hν ! DMABN•þ þ eaq
The formation of DMABN•+ was implicitly assumed, but not
directly observed, in a study on the reduction of excited triplet
methylene blue by several aromatic amines.30 DMABN•+ is
believed to be the main species formed in natural waters upon
oxidation of DMABN by 3DOM*.20
The present study focuses on DMABN•+ formation and the
description of its kinetics measured by LFP in aqueous solu-
tion. To form DMABN•+, direct photoexcitation and photosen-
sitized oxidation of DMABN were employed. Simulations of the
decay kinetics of DMABN•+ based on systems of reaction
equations were also performed to evaluate its main reaction
pathways. The results of these simulations were validated by
comparison with experimental data.
2. Experimental section
2.1. Chemicals and solutions
All chemicals were commercially available and used as
received. A complete list of chemicals is given in the ESI, Text
S1.† Water used for all experiments was obtained from an
Aqua Osmotic 02A purification system. All sample solutions,
made by diluting stock solutions of the reagents, were
buffered using 2 × 10−3 M phosphate (total concentration) at
pH 8.0, except when otherwise mentioned. Stock solutions
were made in water except for the photosensitizers 3-methox-
yacetophenone, 1-naphthaldehyde, 2-acetonaphthone and
1-acetonaphthone, for which acetonitrile (MeCN) was used as
a cosolvent due to the limited solubility of these compounds
in water. The concentration of MeCN in the sample solutions
did not exceed 10% (v/v).
2.2. Laser flash photolysis (LFP) apparatus
Nanosecond LFP experiments were conducted using a 4 × 1 ×
1 cm quartz cuvette containing the sample solution and laser
pulses of the second, third or fourth harmonic frequency from
a Nd:YAG laser (EKSPLA, model SL334). The LFP setup was
operated in a perpendicular arrangement of the pump and
probe beams. The laser pulses (pulse energies of 190–210 mJ
at λ = 532 nm, 150–180 mJ at λ = 355 nm, and 70–90 mJ at λ =
Chart 1 Structure of 4-(dimethylamino)benzonitrile (DMABN) and its
radical cation (DMABN•+).
Photochem. Photobiol. Sci.
This journal is © The Royal Society of Chemistry and Owner Societies 2018

View Article Online
Photochemical & Photobiological Sciences
Paper
266 nm; duration ≤170 ps) were dispersed on the 4 cm side of equivalent Eutech Instruments pH600 was used to measure
the cell using a cylindrical concave lens. Transient absorption the pH.
spectra were recorded using an ICCD camera (Andor iStar,
2.5 Kinetic models and simulations
model DH740i-18U-03) with an overpulsed xenon arc lamp as
a source of the probe light. Cut-off filters (480, 400, 380, 320 or
305 nm) were applied routinely to prevent the samples from
unnecessary irradiation from the xenon lamp. The filtered
light of the xenon lamp did not create transient spectra.
Kinetic traces were recorded using the software TekScope on a
Tektronix digital phosphor oscilloscope (model DPO7104C) at
a single wavelength in the range of 400–670 nm (see ESI,
Table S1† for the detection wavelength used in the measure-
ment of each particular transient) with a monochromator
using a Hamamatsu photomultiplier tube R928. Absorbance
values of the sample solutions were usually adjusted to 0.5–0.8
(for a 1 cm optical path length) at the excitation wavelength.
Samples were naturally aerated or degassed by applying three
freeze–pump–thaw cycles under reduced pressure (8 Pa) or
purged with a gentle stream of oxygen or N2O for 15 minutes
prior to measurements. For the determination of second-order
quenching rate constants, nominal equilibrium concentration
of dissolved gases for an ambient temperature of 21 °C and
The software Kintecus© was employed to simulate the buildup
and decay kinetics of DMABN•+ and various relevant species
occurring in DMABN-containing solutions after the short laser
flash excitation.31 Briefly, this software solves the system of
differential equations derived from the chemical reaction
equations for the system under study (i.e., the kinetic model),
using a set of known reaction rate constants and initial con-
centrations of the species involved. The output comprises the
time course of the concentrations of each considered species.
The systems of reaction equations and corresponding rate con-
stants used in the present study are described in the ESI, Texts
S2 and S3 and Tables S2 and S3.†
3. Results and discussion
3.1. Direct excitation of 4-(dimethylamino)benzonitrile
(DMABN)
atmospheric pressure of 99 kPa were assumed. These corres- Direct excitation of DMABN using laser pulses of 266 nm wave-
pond to 2.80 × 10–4 M and 1.33 × 10–3 M for dissolved oxygen length resulted in transient absorption spectra that are rep-
in aerated and O2-purged solution, respectively, and to 2.7 × resented in Fig. 1. A short-lived transient with a broad absorp-
10–2 M for dissolved N2O. Absorption spectra of the sample tion band at 600–700 nm is evident at 5–60 ns delay times
solutions were measured regularly between laser flashes to test after excitation in aerated and oxygen-purged solutions
for possible photodegradation of the solution components (Fig. 1A and B), while it is less apparent in an N2O-purged
−
using the later described diode-array spectrophotometer. solution (Fig. 1C). This transient can be assigned to eaq
,
Experiments were conducted in an air-conditioned room at the which is known to react with dissolved oxygen and N2O at
−
temperature of 21 1 °C.
diffusion-controlled rates.29 The quenching of eaq is more
efficient in N2O- than in O2-purged solution because of the
higher saturation concentration of N2O compared to O2
2.3 Kinetic analyses
The kinetic traces were generally fitted by single or multiple (vide infra). After the decay of eaq−, two broad bands are appar-
exponential decay functions using the software Flash Fit ent at all measured conditions. The first of these bands has an
v. 0.11. In the case of second-order kinetics, the corresponding absorption maximum at λ ≈ 400 nm and its absorbance is
differential equations were solved using the software Matlab strongly reduced 800 ns after the excitation, while the second
with the Levenberg–Marquard minimization algorithm or, band has an absorption maximum at λ ≈ 500 nm and is
alternatively, the software Kintecus© and applying the systems especially discernible at long delay times (up to 5 µs) after the
of equations and corresponding rate constants described in laser pulse. These species can be attributed to the excited
the ESI, Texts S2 and S3 and Tables S2 and S3.† The observed triplet state of DMABN (3DMABN*) and to the radical cation of
rate constants, determined from single-exponential fitting of DMABN (DMABN•+), respectively.27 The kinetics of the three
,
3DMABN* and DMABN•+, are
−
experimental data, are denoted in this paper with an key species, namely eaq
additional superscript “obs”, while the second-order rate con- described in more detail below.
stants obtained from fitting of experimental data or from
3.1.1. Excited triplet state of DMABN (3DMABN*). The
observed first-order rate constants are denoted with an observed decay kinetics of 3DMABN*, measured at 400 nm,
additional superscript “exp”. The uncertainties of the rate con- was first order except for degassed solution and exhibited
stants given in the tables are expressed as 95% confidence varying decay rate constants (see Table 1) depending on the
intervals and were calculated using the values of at least three concentration of dissolved gases. In degassed solution, the
independent measurements.
first-order deactivation rate constant of 3DMABN*, corres-
ponding to the rate constant of the spontaneous decay of the
triplet state, was determined to be (3.92 0.12) × 104 s−1 by
2.4. Analytical instrumentation
Absorption spectra in the ultraviolet (UV) and visible (Vis) fitting the measured data with a kinetic model that included a
range were measured on an Agilent Cary 100 UV-Vis or an second-order kinetic component to consider triplet–triplet
Agilent 8654 diode-array spectrophotometer. A BNC pHTestr annihilation (see ESI, Text S4†). The determined decay rate
10 pH meter equipped with a calibrated glass electrode or an constants in the presence of dissolved O2 or N2O were at least
This journal is © The Royal Society of Chemistry and Owner Societies 2018
Photochem. Photobiol. Sci.

View Article Online
Paper
Photochemical & Photobiological Sciences
Fig. 1 Transient absorption spectra obtained upon 266 nm laser flash photolysis of (A) DMABN (83 µM) in aerated solution, (B) DMABN (83 µM) in
oxygen-purged solution, and (C) DMABN (133 µM) in N2O-purged solution. The spectra were recorded in the time range of 5 ns to 5 μs after the
laser pulse (see legend) using a 5 ns integration window, and data were smoothed by adjacent averaging over 20 data points (≈ 10 nm). All measure-
ments were performed in pH 8.0 phosphate-buffered aqueous solutions.
Table 1 Kinetic parameters determined for the deactivation of the main transient species formed during 266 nm laser flash photolysis of aqueous
DMABN at pH 8.0 in the presence of various dissolved gases
3DMABN*
eaq
DMABN•+
−
Transient species→
↓Dissolved gases
kd;obs
106 s−1
/
kq;exp
/
a
b
3DMABN*
3DMABN*;X
q;exp
keq;exp b/1010 M−1 s−1
k
d/109 M−1 s−1
k
/104 s−1
d;obs
109 M−1 s−1
ked;obs c/107 s−1
ꢀ;X
ꢀ
DMABN•þ;O2
DMABN•þ
•ꢀ
aq
aq
O2 (1.33 mM)
O2 (0.28 mM)
N2O (0.027 M)
None (degassed)
5.0 0.2
3.7
5.9
1.7 0.1
0.70 0.06
>20 f
1.3
2.5
9.4 0.2
8.1 0.5
n.a.
n.a.e
n.a.
1.70 0.02
0.647 0.002
0.039 0.001h
0.024
>0.74 f
n.a.
1.24 0.07g
0.48 0.12h
0.59 0.02i
n.a.
n.a.
a Observed first-order decay rate constants obtained by single-exponential fitting of the kinetic traces at λobs = 400 nm, unless where noted.
b Second-order quenching rate constants calculated by dividing the difference of the observed first-order decay rate constants in the presence and
absence of dissolved gas by the nominal gas concentration. c Observed first-order decay rate constants obtained by single-exponential fitting of
the kinetic traces at λobs = 600 nm. d Second-order decay rate constant obtained from mixed first- and second-order fitting from the kinetic traces
at λobs = 500 nm. e n.a.: not applicable. f Estimated as described in the text. g Observed first-order decay rate constant obtained by single-exponen-
tial fitting of the kinetic traces at λobs = 500 nm. h From ESI, Table S5, see text and ESI, Text S4 for the methods used for its determination.
i Second-order rate constant for triplet–triplet annihilation from ESI, Table S5, see text and ESI, Text S4 for the methods used for its
determination.
an order of magnitude higher than this value, and second- activation of 3DMABN* in the presence or absence of oxygen
3
order rate constants for the quenching of DMABN* by these can be rationalized in terms of the processes described by
two dissolved gases were obtained as given in Table 1. The de- eqn (3)–(9) and represented in Chart 2, whereby unimolecular
Photochem. Photobiol. Sci.
This journal is © The Royal Society of Chemistry and Owner Societies 2018

View Article Online
Photochemical & Photobiological Sciences
Paper
•ꢀ
eaqꢀ þ O2 ! O2
ð11Þ
−
In degassed solution, the decay of eaq was assumed to be
dominated by its reaction with DMABN, and the aforemen-
tioned fitting procedure (ESI, Text S4†) yielded a second-order
rate constant value of 1.05 × 1010 M−1 s−1 (ESI, Table S5†). This
value, which is close to the diffusion-controlled limit, is
similar to the ones observed for the reaction of eaq− with other
analogous compounds, such as benzonitrile (1.9 × 1010
M−1 s−1) and p-methylbenzonitrile (1.4 × 1010 M−1 s−1).29
Other reactions of eaq with buffer solution components (H+,
−
OH−, H2O, Na+, H2PO4 and HPO42−), with eaq itself, and
−
−
Chart 2 Relevant processes involving the excited triplet state of
DMABN (3DMABN*). Numbers correspond to the reaction equations
listed in the text.
with DMABN•+ were estimated to be negligible.
−
In N2O-purged solution, the signal corresponding to eaq
was not clearly detected even on a nanosecond time scale (see
Fig. 1C) due to fast scavenging by N2O, which is known to lead
to the formation of the hydroxyl radical and elementary nitro-
gen (eqn (12)).32,33 Considering that 5 ns after the laser pulse
<37% of the absorbance signal at 600 nm is present if com-
pared with an aerated solution (Fig. 1), the second-order rate
deactivation (including contributions by the solvent, eqn (3)–(5))
and triplet–triplet annihilation are dominant in degassed
solution, while quenching by oxygen (eqn (7)–(9)) is largely
prevalent in aerated or oxygen-purged solution. A distinction
of the processes within the groups of eqn (3)–(5) and (7)–(9) is
not possible in the frame of this study. However, we propose
that the processes of reactive deactivation (eqn (5)) and reactive
quenching by oxygen (eqn (9)) should have minor importance
in view of the small quantum yield (1.3 × 10–3) observed for
the direct phototransformation of DMABN under steady-state
irradiation in aerated solution.20
−
constant for the reaction of eaq with N2O (eqn (12)) can be
estimated as >7.4 × 109 M−1 s−1, which agrees well with the
selected value of 9.1 × 109 M−1 s−1 reported in the literature.29
•
eaqꢀ þ N2O þ H2O ! N2 þ OH þ OHꢀ
ð12Þ
The very reactive hydroxyl radical, formed according to eqn
(12), is expected to be completely scavenged by DMABN leading
to the formation of intermediate products. These products are
assumed not to be observable in the visible wavelength range
and not to significantly interact with the other relevant transient
species under the present experimental conditions.
3DMABN* ! DMABN þ hν ðphosphorescenceÞ
3DMABN* ! DMABN ðnon-radiative decayÞ
3DMABN* ! Products ðreactive deactivationÞ
ð3Þ
ð4Þ
ð5Þ
3.1.3. Radical cation of DMABN (DMABN•+). The decay of
DMABN•+ in oxygen-containing solution was best fitted using a
second-order kinetic component. In these solutions, the transi-
ent absorption traces measured at 500 nm were fitted using, in
addition to the mentioned second-order component, two expo-
nential functions with fixed decay constants to account for the
3
1
3DMABN* þ DMABN* ! DMABN*
ð6Þ
þ DMABN ðtriplet ꢀ triplet annihilationÞ
1
3DMABN* þ O2 ! DMABN þ O2 ðenergy transferÞ ð7Þ
observed first-order decay of DMABN* and eaq−, which have a
3
3DMABN* þ O2 ! DMABN þ O2 ðexcitation energy lossÞ ð8Þ
small but still significant absorption at the detection wavelength.
An example of such a fitting is given in the ESI, Fig. S2.† The
obtained second-order rate constants (Table 1) were assigned to
3DMABN* þ O2 ! Products ðreactive quenchingÞ
ð9Þ
−
−
the reaction of DMABN•+ (formed concomitantly with eaq
N2O, used to scavenge eaq (see the next sub-section) and
not known as an excited triplet state quencher, also contribu-
ted to the quenching of 3DMABN* (described by eqn (10)),
although at a much lower rate than O2.
according to eqn (2)) with O2•−, which is the product of the reac-
tion of eaq− with molecular oxygen (eqn (11)), see eqn (13).
•ꢀ
DMABN•þ þ O2 ! DMABN þ O2
ð13Þ
3DMABN* þ N2O ! DMABN þ N2Oðexcitation energy lossÞ
ð10Þ
3.1.2. Hydrated electron (eaq−). Upon LFP of DMABN in
This assignment is based on the fact that, once formed,
−
eaq is scavenged by O2 on the sub-microsecond time scale
and, consequently, its reaction with DMABN•+ (eqn (14)) does
not significantly contribute to the decay of the latter (see also
kinetic simulations in the ESI, Fig. S3†).
oxygenated aqueous solutions, the absorption corresponding
−
to eaq decayed by first-order kinetics with determined rate
constants as given in Table 1. The derived second-order rate
ꢀ
DMABN•þ þ eaq ! DMABN
ð14Þ
constants for the quenching of eaq− by O2 (Table 1) are in good
agreement with the literature value (1.9 × 1010 M−1 s−1) for the
A further postulated reaction is the transformation of
reaction of eaq with oxygen to yield the superoxide radical DMABN•+ to yield the N-demethylated product 4-(methylamino)
−
anion (O2•−, eqn (11)).29
benzonitrile (MABN).34 The proposed reaction sequence (eqn (15))
This journal is © The Royal Society of Chemistry and Owner Societies 2018
Photochem. Photobiol. Sci.

View Article Online
Paper
Photochemical & Photobiological Sciences
considers an H-atom shift in DMABN•+ to produce an intermedi- can be in the range of 1.8–2.4 V vs. standard hydrogen elec-
ate carbon-centered radical cation (abbreviated as DMABNr•+), trode (SHE) and are thus comparable to those of the strongest
which then deprotonates to give the radical DMABNr•.
radical oxidants.36,37 The oxidation reaction of DMABN by
ð15Þ
In degassed solution, application of the aforementioned 3Sens* is expected to produce DMABN•+ as well as the radical
fitting procedure that is described in the ESI, Text S4† yielded anion of the photosensitizer (Sens•−) within the lifetime of
a first-order rate constant of ≈0.5 × 104 s−1 (see Table 1), which 3Sens* (eqn (17)). Besides the unimolecular and bimolecular
3
we assign to the transformation of DMABN•+ (eqn (15)). In deactivation processes that were listed for DMABN* (eqn (3)–
N2O-purged solution, the decay of DMABN•+ fitted well first- (9)), oxidizing 3Sens* are expected to undergo the following
order kinetics with a rate constant that was ≈2.6 times larger additional reactions with DMABN: excitation energy loss (eqn
3
than the one determined in degassed solution (see Table 1). (18)) and triplet–triplet energy transfer from Sens* to DMABN
This first-order decay concurs with the expectation that (eqn (19)). Both reactions (17) and (18) have been observed
−
DMABN•+ cannot undergo second-order reactions with eaq
during the quenching of excited triplet methylene blue by
•−
(which is scavenged by N2O) or O2 (which cannot be formed DMABN in methanol solution with a ratio of 1 : 3 between elec-
under these solution conditions).
tron transfer and physical quenching.30 While we are not
3.1.4. pH effect. The decay kinetics of 3DMABN* and aware of any method that could be used to maximize the yield
DMABN•+ was studied in aerated and N2O-purged solutions at of DMABN•+ formation during quenching, the energy transfer
various pH values, with the aim of getting additional infor- reaction (eqn (19)), and consequently the formation of
mation about the processes involved. The results of the fit- 3DMABN*, can be suppressed by choosing photosensitizers
tings, performed as described in the previous sub-sections, are with a lower triplet energy than DMABN.
3
gathered in the ESI, Table S6.† For DMABN*, the deactivation
1
3
Sens þ hν ! Sens* ! Sens*
ð16Þ
ð17Þ
rate constant in aerated solution was not significantly affected
by pH in the range of 4.5–7.7 (average standard deviation:
3Sens* þ DMABN ! Sens•ꢀ þ DMABN•þ
ðoxidative quenchingÞ
(1.75 0.04) × 106 −1), while in N2O-purged solution it was
s
pH-dependent with an increase by a factor of ≈2 at pH 7.7
compared to the lower pH range. We tentatively attribute this
increase related to the quenching of 3DMABN* by N2O to an
3Sens* þ DMABN ! Sens þ DMABN ðexcitation energy lossÞ
2−
ð18Þ
effect of the used phosphate buffer, with HPO4 being the
dominant species at pH 7.7, while at the lower investigated pH
3
3Sens* þ DMABN ! Sens þ DMABN*
values H2PO4 is predominant. For DMABN•+ in aerated solu-
ð19Þ
−
ðtriplet ꢀ triplet energy transferÞ
tions, the obtained second-order rate constant values were
invariable in the pH range of 5.4–7.7 ((7.7 0.4) × 109 M−1
To avoid direct excitation of DMABN, the 355 nm and
532 nm excitation wavelengths of the laser were used to selec-
tively produce Sens*. The sensitizers were selected to signifi-
s
−1), but the value at pH 4.5 was significantly lower ((4.0 0.2)
× 109 M−1 s−1). This concurs with a major change in speciation
of O2•−, which is mainly protonated to its conjugated acid, the
3
cantly absorb light at one of these excitation wavelengths (see
the electronic absorption spectra in the ESI, Fig. S4†) and to
cover a range of one-electron reduction potentials that could
be used to estimate the oxidation potential of DMABN in
aqueous solution (presently unknown). Moreover, practical
aspects, such as the aqueous solubility of the photosensitizers
and the feasibility of their selective excitation were considered.
3.2.1. Transient absorption spectra. To disentangle the
spectra of 3Sens*, Sens•− and DMABN•+ present after LFP of
solutions containing a given photosensitizer and DMABN, and
hydroperoxyl radical (HO2 ), at pH 4.4 (note: the pKa of HO2• is
•
4.8).35 The reduction in rate constant is plausible, because
HO2 is a weaker electron donor than O2•−. For DMABN•+ in
•
N2O-purged solutions, the obtained first-order rate constant
values vary by only 15% in the pH range of 4.5–7.7, which let
us conclude that the transformation of DMABN•+ is not
affected by protonation reactions in this pH range.
3.2. Photosensitized formation of DMABN•+
Several photosensitizers were used as a means to oxidize to select suitable observation wavelengths for the kinetic
DMABN to DMABN•+. Upon excitation of a photosensitizer measurements, the following procedure was applied: (1) gene-
3
(Sens) its excited triplet state (3Sens*) is formed through inter- ration of the Sens* from a buffered solution of the photosen-
system crossing from the excited singlet states (eqn (16)). sitizer alone; (2) generation of 3Sens*, and subsequently
3Sens* can act in certain cases as a powerful oxidant. Sens•−, from a buffered solution containing the photosensiti-
Especially aromatic ketones and quinones possess standard zer and triethanolamine (TEA. Note: the radicals formed by
3
°*
red
one-electron reduction potentials, E ð Sens*=Sens•ꢀÞ, that the oxidation of TEA absorb outside of the considered spectral
Photochem. Photobiol. Sci.
This journal is © The Royal Society of Chemistry and Owner Societies 2018

View Article Online
Photochemical & Photobiological Sciences
Paper
Fig. 2 A−C: Transient absorption spectra following 355 nm laser pulse excitation of: (A) 2-acetonaphthone (2-AN, 5 × 10−4 M); (B) 2-AN (5 × 10−4
M) and triethanolamine (TEA, 1 × 10−2 M); (C) 2-AN (5 × 10−4 M) and DMABN (5 × 10−4 M). Spectra were recorded in the time range from 20 ns to
50 μs after the laser pulse using a 5 ns integration window. Spectral data were smoothed by adjacent averaging over 20 data points (≈ 10 nm). All
samples (pH 8.0) were made in aerated water containing 0.9% (v/v) MeCN. (D) A plot of the determined first-order decay rate constant of the triplet
state of 2-AN (obtained at the observation wavelength of 440 nm) vs. the DMABN concentration. Error bars represent 95% confidence intervals from
at least quadruplicate measurements.
range38); (3) production of the three species, i.e., 3Sens*, nation of decay rate constants of the various species. Fig. 2D
Sens•− and DMABN•+, in a buffered solution of the photosensi- displays an example of a linear regression plot of deactivation
tizer and DMABN. Fig. 2 shows an illustrative example of this rate constants of 32-AN* determined at various DMABN con-
procedure using 2-acetonaphthone (2-AN) as a model photo- centrations, which was used to determine the second-order
sensitizer (for a collection of transient absorption spectra for rate constant for the quenching of 32-AN* by DMABN.
the other photosensitizers, see the ESI, Texts S5−S9 and
3.2.2. Quenching of the excited triplet state of photosensi-
Fig. S5−S9†). The transient absorption spectrum of the excited tizers by DMABN. For photosensitizers having oxidizing
triplet state of 2-AN (32-AN*) has a maximum at λ ≈ 440 nm excited triplet states, quenching rate constants can be used as
(Fig. 2A), while the spectrum of the radical anion of 2-AN an indicator of the efficacy in producing the oxidized substrate.
(2AN•−) overlaps partly with that of 32-AN* but has its To produce DMABN•+ at sufficiently high concentration for an
3
maximum absorption at λ ≈ 400 nm (Fig. 2B). The transient accurate evaluation of its decay kinetics, high Sens* second-
absorption spectra shown in Fig. 2C contain a dominating order quenching rate constants are advantageous. Table 2 col-
superposition of 32-AN* and 2-AN•− at short delay times (up to lects the second-order quenching rate constants obtained in
°*
red
≈500 ns) after the laser pulse, but at longer delay times a band this study for six photosensitizers exhibiting E values in the
centered at ≈500 nm clearly emerged. This was assigned to range of 1.26–1.86 V vs. SHE. The corresponding linear
DMABN•+ based on its similarities to the spectra of DMABN•+ regression plots are given in Fig. 2D for 2-AN and in the ESI,
observed from direct photoexcitation of DMABN in this study Fig. S9† for the other photosensitizers. The quenching rate
(Fig. 1) and in previous studies.27,39 Note that the presence of constants vary over about two orders of magnitude and
oxygen is beneficial to the observation of DMABN•+, because it increase non-linearly with increasing E . Literature values of
°*
red
accelerates the decay of Sens•− (in this case 2-AN•−) according the triplet state energies of the photosensitizers are also col-
to eqn (20), leaving DMABN•+ as the only absorbing species in lected in Table 2 to evaluate the possibility of energy transfer
the considered spectral range at long delay times.
as a side-reaction (eqn (19)). Only 3-methoxyacetophenone
(ET = 3.14 eV)40 has a higher triplet state energy than DMABN
(2.81 eV),27 but no evidence for the formation of 3DMABN*
Sens•ꢀ þ O2 ! Sens þ O2
ð20Þ
•ꢀ
The transient absorption spectra shown in Fig. 2 were uti- could be found in the measured transient absorption spectra
lized to evaluate the optimal observation wavelength to be with this photosensitizer. For all other photosensitizers energy
used for the measurement of kinetic traces and the determi- transfer can be excluded a priori.
This journal is © The Royal Society of Chemistry and Owner Societies 2018
Photochem. Photobiol. Sci.

View Article Online
Paper
Photochemical & Photobiological Sciences
Table 2 Ground-state reduction potentials ðEr°edÞ, triplet-state reduction potentials ðE Þ, triplet state energies (ET) and determined second-order
°*
red
rate constants of triplet state quenching by DMABN ðkq;exp
3Sens*;DMABNÞ for the studied photosensitizers
Er°ed a/V vs. SHE
E
a/V vs. SHE
ET a/eV
kq;exp b/109 M−1 s−1
3Sens*;DMABN
°*
red
Photosensitizer
9,10-Anthraquinone-1,5-disulfonate
3-Methoxyacetophenone
Thionine
−0.5
1.86
2.36
3.14 c
1.70
2.45
2.59
2.52
5.02 0.16
4.0 0.2d
4.6 0.3
−1.43 c
−0.25
−1.11
−1.25
−1.26
1.71 c
1.45
1.34
1.34
1.26
1-Naphthaldehyde c
2-Acetonaphthone c
1-Acetonaphthone c
3.4 0.3e
0.21 0.03e
0.031 0.017e
a
°*
Standard one-electron reduction potentials of the photosensitizers in their electronic ground state ðEr°edÞ and excited triplet state ðE Þ as well
red
as triplet state energies (ET) obtained from ref. 36 unless otherwise noted. b Measured in aerated aqueous solution at pH 8.0. Errors represent
95% confidence intervals obtained from the linear regression lines. c From ref. 40. d Solutions containing 10% (v/v) MeCN. e Solutions containing
≈1% (v/v) MeCN.
In the following we analyze the rate constants for triplet 0.13 and 0.09 V, respectively) than a crude estimate of 1.19 V
state quenching by DMABN, kq, in the frame of electron trans- obtained by adding the difference between the oxidation
fer theory, more precisely by using the Rehm–Weller relation- potentials of 4-cyanoaniline (1.32 V)45 and aniline (1.00 V)45 to
ship (eqn (21)).41–43
the oxidation potential of N,N-dimethylaniline (0.87 V),46 all
kd
kq ¼
:
ð21Þ
8
<
9
=
20
4@
1
A
3
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ꢂ
ꢃ
ꢂ ꢃ
ꢂ
ꢃ
ꢂ
ꢃ
kd
kdZ
ΔrG°et
2
λ
ΔrG°et
2
ΔrG°et
RT
5
1 þ
exp
þ
2þ
=RT þ exp
:
;
4
The parameters are defined as follows: kd and k−d are the measured in aqueous solution. Note that a value of 1.11 V for
rate constants for the formation and separation of the precur- Er°edðDMABN•þ=DMABNÞ was determined in MeCN solution,47
sor complex, respectively, Kd = kd/k−d is the equilibrium con- which is even lower than the aforementioned crude estimate.
stant for the precursor complex formation, Z is the universal
3.2.3. Decay kinetics of DMABN•+. The DMABN•+ decay
collision frequency factor, R is the universal gas constant, T is kinetics was studied at various “initial” DMABN•+ concen-
the absolute temperature, λ is the reorganization energy, and
ΔrG°et is the standard molar free energy change of the electron
transfer reaction, i.e., the standard molar free energy differ-
ence between successor complex and precursor complex,
which was approximated using eqn (22).
ꢀ
ΔrG° ffi F Â Er°edðDMABN•þ=DMABNÞ
et
ð22Þ
ꢁ
ꢀE°* ð Sens* =Sens•ꢀÞ
3
red
To fit the determined second-order rate constants for
3
quenching of Sens* by DMABN, eqn (22) was inserted in eqn
3
(21), and E ð Sens*=Sens•ꢀÞ was used as the independent
°*
red
variable while Er°edðDMABN•þ=DMABNÞ was a fitting para-
meter. A similar procedure as in ref. 44 was applied, keeping
the ratio kd/(Kd × Z) fixed at 0.1. λ was used as an additional
fitting parameter, while kd was fixed at 5.0 × 109 M−1 s−1. Fits
to the Rehm–Weller model are represented in Fig. 3, whereby
fits with various fixed values of λ are also shown. The fit to the
Rehm–Weller model with two fitting parameters gave values of
Fig. 3 Determined second-order rate constants for the quenching of
excited triplet states of photosensitizers by DMABN (kq;exp
3Sens*;DMABN) (black
0
kJ mol−1 for
λ
and 1.32
0.04 V vs. SHE for
Er°edðDMABN•þ=DMABNÞ. Using a fixed value of λ = 20 kJ mol−1
(a value which was obtained for the quenching of the excited
triplet state of methylene blue by a series of substituted ani-
3
squares) plotted on a logarithmic scale against E ð Sens*=Sens•ꢀÞ, the
°*
red
one-electron reduction potential of the excited triplet states of the
photosensitizers. The lines represent fits to the Rehm–Weller model
with reorganization energy (λ) values as given in the legend. The fitting
procedure is explained in the text. Error bars represent 95% confidence
intervals obtained from plots of triplet state decay rate constants against
lines15) also yields
a
reasonably good fit with
Er°edðDMABN•þ=DMABNÞ ¼ 1:28 + 0:08 V vs: SHE (used as the
only fitting parameter). Both values are significantly higher (by DMABN concentration (see Fig. 2D and S10†).
Photochem. Photobiol. Sci.
This journal is © The Royal Society of Chemistry and Owner Societies 2018

View Article Online
Photochemical & Photobiological Sciences
Paper
trations with the main aim of evaluating the relative impor-
tance of first- and second-order kinetic processes. A series of
experiments was conducted at constant DMABN concentration
(5 × 10−4 M) using 1-naphthaldehyde (1-NA) as the photosensi-
3
tizer. Thereby, the “initial” concentration of 1-NA* (measured
≈10 ns after the laser pulse) was tuned by varying the concen-
tration of 1-NA in the range of 5 × 10−5–3 × 10−4 M, and
additionally by varying the laser pulse intensity using metal-
grid filters (experiments conducted at [1-NA] = 5 × 10−5 M).
The transient absorbance change data were first analyzed
using conventional second-order kinetic plots as detailed in
the ESI, Text 10, Fig. S11 and Table S7.† Especially at the high
1-NA concentrations, and consequently at the high “initial”
31-NA* and DMABN•+ concentrations, this simple kinetic analysis
yielded consistent values for the second-order rate constant,
which was assigned to the reaction of DMABN•+ with O2•− (eqn
(13)). With decreasing initial concentration of the considered
transient species, deviation from the simple second-order
kinetic model became increasingly apparent, which was inter-
preted as an increasing impact of the transformation of
DMABN•+ (eqn (15)) on its decay. Kinetic simulations, per-
formed according to the details given in the ESI, Text S3 and
Table S3† and displayed in Fig. S12,† supported this hypoth-
esis. Moreover, these simulations revealed that the concen-
•−
trations of DMABN•+ and O2 at a given time point are gener-
ally different, meaning that the condition to apply the simple
second-order kinetic model are not fulfilled. Therefore, a com-
plete analysis in terms of differential equations representing
the kinetic system is required for an adequate evaluation of
the DMABN•+ decay data.
Data fitting was performed using Kintecus© and the kinetic
model presented in the ESI, Text S3 and Table S3.† For the
fittings, the initial concentration of the excited triplet state of
1-NA, [31-NA*]0, was estimated using the procedure described
in the ESI, Table S7† and employed as a fixed parameter, while
the molar absorption coefficient of DMABN•+, the second-
order rate constant for the reaction between DMABN•+ and
Fig. 4 Decay traces of DMABN•+ (red lines, observation wavelength
500 nm) formed by laser flash excitation of 1-naphthaldehyde (1-NA) at
O2•−, kDq;MexApBN•þ •ꢀ , and the first-order rate constant for the 355 nm. Experiments were performed with constant [DMABN]0 = 5 ×
;O2
10−4 M and various [1-NA]0 and metal grid filters (see text). Black curves
represent fits performed using Kintecus© as described in the text. Panels
transformation of DMABN•+ (see eqn (15)), kDd;MobAsBN•þ , were used
A−C contain decay traces measured at different initial concentrations of
31-NA*, [31-NA*]0, estimated as given in the ESI, Table S7.† The obtained
fit parameters are collected in the ESI, Table S8.†
as fitting parameters. Fig. 4 shows that the fittings were able to
accurately describe the experimental kinetic traces, however
the obtained best-fit parameters varied depending on the esti-
mated [31-NA*]0 (see ESI, Table S8†). With decreasing [31-NA*]0
q;exp
(from 63 to 1.9 μM), fitted k
values steadily
DMABN•þ;O2
•ꢀ
increased from 5.2 × 109 to 1.2 × 1010 M−1 s−1, while the fitted
values of the molar absorption coefficient of DMABN•+
increased from ≈2400 M−1 cm−1 to values in the rather narrow
from direct photoexcitation experiments in degassed solution
(see Table 1).
q;exp
A
preferred
value
of
k
¼ ð5:2 + 0:4Þ Â
DMABN•þ;O2
•ꢀ
109 Mꢀ1 sꢀ1 ðmean + sdÞ was calculated by averaging the
values obtained at the two highest [31-NA*]0. This was done
considering the fact that at these [31-NA*]0 values the reaction
range of ≈3500–3800 M−1 cm−1 for [31-NA*]0 ≤ 14 μM. For the
d;obs
DMABN
third fitting parameter, k
•þ , it was not possible to obtain
a reliable estimate (the lower limit of 100 s−1 that was set as a
constraint in the fitting procedure was attained for
most decays). The only successful estimation, obtained for
•−
between DMABN•+ and O2 largely dominates the decay of
DMABN•+, while at lower [31-NA*]0 values the increasing impor-
tance of the first-order decay process of DMABN•+ is expected
[31-NA*]0 = 1.92 μM, yielded a value of (1.2 0.5) × 103 s−1
,
q;exp
to bias the determination of k
•ꢀ . The preferred value
is significantly lower than the value obtained
DMABN•þ;O2
q;exp
d;obs
which is much lower than the estimate of k
obtained of k
DMABN•þ
DMABN•þ;O2
•ꢀ
This journal is © The Royal Society of Chemistry and Owner Societies 2018
Photochem. Photobiol. Sci.

View Article Online
Paper
Photochemical & Photobiological Sciences
by the second-order fittings ((1.6 0.1) × 1010 M−1 s−1, see ESI, intended to assess the phototransformation rates of DMABN
Text S10 and Table S7†) and the values obtained from the and analogous aquatic contaminants in surface waters.
direct photoexcitation experiments in the presence of air and
oxygen ((8.1 0.5) and (9.4 0.2) × 109 M−1 s−1, respectively,
see Table 1). The difference to the latter values might be due
Conflicts of interest
to the different assumptions adopted when performing the
various types of fittings. Using this value of 5.2 × 109 M−1 s−1 There are no conflicts to declare.
q;exp
as fixed input parameter for k
•ꢀ , a two-parameters
DMABN•þ;O2
fitting of the [31-NA*]0 = 1.92 μM data was performed, which
yielded a value of (3.5 1.4) × 103 s−1 for the first-order trans-
formation of DMABN•+. This value is not far from (4.8 1.2) ×
Acknowledgements
103 s−1, the value obtained from direct photoexcitation of The authors would like to thank Luboš Jílek for the technical
DMABN in degassed solutions (see Table 1). To our knowledge, support with the laser system. This work was supported by the
this represents the first estimation of a deprotonation rate con- Swiss National Science Foundation (Project No. 200021-
stant for an N,N-dimethylaniline radical cation in aqueous 140815). The RECETOX research infrastructure was supported
solution. Previous studies on such radical cations performed by the projects of the Czech Ministry of Education (LO1214
using organic solvents and acetate as a proton acceptor48,49 and LM2011028).
support the assignment of this rate constant to a deprotona-
tion reaction as illustrated by eqn (15).
References
1 W. W. Stone, R. J. Gilliom and K. R. Ryberg, Pesticides in U.
S. streams and rivers: occurrence and trends during 1992-
4. Conclusions
This study focused on the formation and decay of the radical
cation of DMABN, DMABN•+, which was produced in aqueous
solution by laser flash excitation of either DMABN or a photo-
sensitizer inducing the oxidation of DMABN. DMABN•+, with a
maximum absorption at a wavelength of ≈500 nm, was the
longest-lived transient intermediate observed in the visible
spectral range, and was detected in a time delay range extend-
ing over several 100 μs. The photosensitized formation of
DMABN•+ was found to be efficient using sensitizers with esti-
mated triplet-state standard one-electron reduction potential
of ≳1.34 V vs. SHE, and for these sensitizers a rate constant for
the quenching of their excited triplet state by DMABN of >3 ×
109 M−1 s−1 was observed. On the nanosecond time scale, the
primary by-products of DMABN•+ formed upon laser flash
photolysis of DMABN and of a sensitizer (Sens) were the
hydrated electron, eaq−, and the radical anion of the sensitizer,
Sens•−, respectively. In aerated solution, these species reacted
with molecular oxygen with half-lifetimes of ≈100 ns and
≈0.7–2 μs, respectively, to yield the superoxide radical anion
(O2•−, not monitored in this study). In oxygen-containing solu-
2011, Environ. Sci. Technol., 2014, 48, 11025–11030.
2 C. Moschet, I. Wittmer, J. Simovic, M. Junghans,
A. Piazzoli, H. Singer, C. Stamm, C. Leu and J. Hollender,
How a complete pesticide screening changes the assess-
ment of surface water quality, Environ. Sci. Technol., 2014,
48, 5423–5432.
3 T. A. Ternes, Occurrence of drugs in German sewage treat-
ment plants and rivers, Water Res., 1998, 32, 3245–3260.
4 X. S. Miao, F. Bishay, M. Chen and C. D. Metcalfe,
Occurrence of antimicrobials in the final effluents of waste-
water treatment plants in Canada, Environ. Sci. Technol.,
2004, 38, 3533–3541.
5 K. Othmen and P. Boule, Photochemical behaviour of
dichloroanilines in water and formation of aminochloro-
phenoxazones, J. Photochem. Photobiol., A, 1999, 121, 161–
167.
6 D. Vialaton, J. F. Pilichowski, D. Baglio, A. Paya-Perez,
B. Larsen and C. Richard, Phototransformation of propico-
nazole in aqueous media, J. Agric. Food Chem., 2001, 49,
5377–5382.
tion under the LFP conditions of this study, the deactivation of
7 S. Canonica and H. U. Laubscher, Inhibitory effect of dis-
solved organic matter on triplet-induced oxidation of
aquatic contaminants, Photochem. Photobiol. Sci., 2008, 7,
547–551.
8 A. L. Boreen, W. A. Arnold and K. McNeill, Triplet-sensi-
tized photodegradation of sulfa drugs containing six-mem-
bered heterocyclic groups: identification of an SO2 extru-
sion photoproduct, Environ. Sci. Technol., 2005, 39, 3630–
3638.
9 J. J. Guerard, P. L. Miller, T. D. Trouts and Y. P. Chin, The
role of fulvic acid composition in the photosensitized
degradation of aquatic contaminants, Aquat. Sci., 2009, 71,
160–169.
•−
DMABN•+ appeared to be dominated by its reaction with O2
,
leading to the reformation of DMABN. For steady-state
irradiation conditions as typical for sunlit surface waters,
where much lower concentrations of O2•− were detected (10–12
–
10–9 M),50 the reaction with O2 should be negligible. From
direct photoexcitation experiments performed in degassed
solution as well as photosensitization experiments, the first-
order transformation of DMABN•+, which was primarily
assigned to its deprotonation, may be estimated to take place
with rate constants not exceeding ≈5 × 103 s−1. The reaction of
DMABN•+ with dissolved organic matter, an important natural
water component, will be the subject of a follow-up study
•−
Photochem. Photobiol. Sci.
This journal is © The Royal Society of Chemistry and Owner Societies 2018

View Article Online
Photochemical & Photobiological Sciences
Paper
10 J. A. Leenheer and J. P. Croué, Characterizing aquatic dissolved 25 J. Catalán, On the dual emission of p-dimethyl-
organic matter, Environ. Sci. Technol., 2003, 37, 18A–26A.
aminobenzonitrile and its photophysical implications,
11 C. Richard and S. Canonica, in Handbook of Environmental
Phys. Chem. Chem. Phys., 2013, 15, 8811–8820.
Chemistry, ed. O. Hutzinger, Springer-Verlag Berlin, 26 S. I. Druzhinin, V. A. Galievsky, A. Demeter,
Heidelberger Platz 3, D-14197 Berlin, Germany, 2005, vol.
2, pp. 299–323.
12 J. Hoigné, B. C. Faust, W. R. Haag, F. E. Scully and
R. G. Zepp, Aquatic humic substances as sources and sinks
of photochemically produced transient reactants, ACS
Symp. Ser., 1989, 219, 363–381.
S. A. Kovalenko, T. Senyushkina, S. R. Dubbaka, P. Knoche,
P. Mayer, C. Grosse, D. Stalke and K. A. Zachariasse, Two-
state intramolecular charge transfer (ICT) with 3,5-
dimethyl-4(dimethylamino)benzonitrile (MMD) and its
meta-isomer mMMD. Ground state amino twist not essen-
tial for ICT, J. Phys. Chem. A, 2015, 119, 11820–11836.
13 D. Vione, M. Minella, V. Maurino and C. Minero, Indirect 27 G. Köhler, G. Grabner and K. Rotkiewicz, Nonradiative de-
photochemistry in sunlit surface waters: photoinduced pro-
duction of reactive transient species, Chem. – Eur. J., 2014,
20, 10590–10606.
activation and triplet states in donor aryl acceptor com-
pounds (dialkylaminobenzonitriles), Chem. Phys., 1993,
173, 275–290.
14 J. Wenk and S. Canonica, Phenolic antioxidants inhibit the 28 G. Köhler, N. Getoff, K. Rotkiewicz and Z. R. Grabowski,
triplet-induced transformation of anilines and sulfonamide
antibiotics in aqueous solution, Environ. Sci. Technol.,
2012, 46, 5455–5462.
Electron photoejection from donor−aryl-acceptor mole-
cules in aqueous solution, J. Photochem., 1985, 28, 537–546.
29 G. V. Buxton, C. L. Greenstock, W. P. Helman and
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,
1988, 17, 513–886.
15 P. R. Erickson, N. Walpen, J. J. Guerard, S. N. Eustis,
J. S. Arey and K. McNeill, Controlling factors in the rates of
oxidation of anilines and phenols by triplet methylene blue
in aqueous solution, J. Phys. Chem. A, 2015, 119, 3233–3243.
16 R. G. Zepp, P. F. Schlotzhauer and R. M. Sink, 30 R. H. Kayser and R. H. Young, The photoreduction of
Photosensitized transformations involving electronic
energy transfer in natural waters - Role of humic sub-
stances, Environ. Sci. Technol., 1985, 19, 74–81.
methylene blue by amines, A flash photolysis study of the
reaction between triplet methylene blue and amines,
Photochem. Photobiol., 1976, 24, 395–401.
17 S. G. Cohen, A. Parola and G. H. Parsons, Photoreduction 31 J. C. Ianni, Kintecus Windows version 5.20, 2014. http://www.
by amines, Chem. Rev., 1973, 73, 141–161. kintecus.com (accessed August 2016).
18 J. Wenk, U. von Gunten and S. Canonica, Effect of dis- 32 F. S. Dainton and D. B. Peterson, Forms of H and OH pro-
solved organic matter on the transformation of contami-
nants induced by excited triplet states and the hydroxyl
radical, Environ. Sci. Technol., 2011, 45, 1334–1340.
duced in radiolysis of aqueous systems, Proc. R. Soc.
London, Ser. A, 1962, 267, 443–463.
33 D. Zehavi and J. Rabani, Pulse radiolytic investigation of
Oaq radical ions, J. Phys. Chem., 1971, 75, 1738–1744.
19 J. Wenk, M. Aeschbacher, M. Sander, U. von Gunten and
S. Canonica, Photosensitizing and inhibitory effects of ozo- 34 E. Baciocchi, M. Bietti, M. F. Gerini and O. Lanzalunga,
nated dissolved organic matter on triplet-induced contami-
nant transformation, Environ. Sci. Technol., 2015, 49, 8541–
8549.
Electron-transfer mechanism in the N-demethylation of
N,N-dimethylanilines by the phthalimide-N-oxyl radical,
J. Org. Chem., 2005, 70, 5144–5149.
20 F. Leresche, U. von Gunten and S. Canonica, Probing the 35 B. H. J. Bielski, D. E. Cabelli, R. L. Arudi and A. B. Ross,
photosensitizing and inhibitory effects of dissolved organic
matter by using N,N-dimethyl-4-cyanoaniline (DMABN),
Environ. Sci. Technol., 2016, 50, 10997–11007.
Reactivity of HO2/O− radicals in aqueous solution, J. Phys.
Chem. Ref. Data, 1985, 14, 1041–1100.
36 I. Loeff, J. Rabani, A. Treinin and H. Linschitz, Charge-
transfer and reactivity of nπ* and ππ* organic triplets,
including anthraquinonesulfonates, in interactions with in-
organic anions: A comparative study based on classical
Marcus theory, J. Am. Chem. Soc., 1993, 115, 8933–8942.
2
21 Z. R. Grabowski, K. Rotkiewicz and W. Rettig, Structural
changes accompanying intramolecular electron transfer:
Focus on twisted intramolecular charge-transfer states and
structures, Chem. Rev., 2003, 103, 3899–4031.
22 A. Demeter and K. A. Zachariasse, Triplet state dipole 37 K. McNeill and S. Canonica, Triplet state dissolved organic
moments of aminobenzonitriles, J. Phys. Chem. A, 2008,
112, 1359–1362.
23 W. Rettig, External and internal parameters affecting the
matter in aquatic photochemistry: reaction mechanisms,
substrate scope, and photophysical properties, Environ.
Sci.: Processes Impacts, 2016, 18, 1381–1399.
dual fluorescence of p-cyano-dialkylanilines, J. Lumin., 38 K. Kalyanasundaram, J. Kiwi and M. Gratzel, Hydrogen
1981, 26, 21–46.
24 S. I. Druzhinin, N. P. Ernsting, S. A. Kovalenko,
L. W. Lustres, T. A. Senyushkina and K. A. Zachariasse,
evolution from water by visible-light, a homogeneous 3
component test system for redox catalysis, Helv. Chim. Acta,
1978, 61, 2720–2730.
Dynamics of ultrafast intramolecular charge transfer with 39 R. J. Visser, P. C. M. Weisenborn, J. Konijnenberg,
4-(dimethylamino)benzonitrile in acetonitrile, J. Phys.
Chem. A, 2006, 110, 2955–2969.
B. H. Huizer and C. A. G. O. Varma, Solute solvent exci-
plexes in the photoinduced formation of radical ions of 4-
This journal is © The Royal Society of Chemistry and Owner Societies 2018
Photochem. Photobiol. Sci.

View Article Online
Paper
Photochemical & Photobiological Sciences
N,N-dimethylaminobenzonitrile and related compounds, 46 G. Merényi and J. Lind, in N-centered radicals, ed.
J. Photochem., 1986, 32, 217–233.
Z. B. Alfassi, John Wiley & Sons Ltd, Chichester, 1998, ch.
40 H. Shizuka and H. Obuchi, Anion-induced triplet quench-
20, pp. 599–613.
ing of aromatic ketones by nanosecond laser photolysis, 47 R. J. Visser and C. A. G. O. Varma, Source of anomalous
J. Phys. Chem., 1982, 86, 1297–1302.
fluorescence
from
solutions
of
4-N,N-dimethyl-
41 D. Rehm and A. Weller, Kinetics and mechanics of electron
transfer during fluorescence quenching in acetonitrile, Ber.
Bunsenges. Phys. Chem., 1969, 73, 834–839.
42 D. Rehm and A. Weller, Kinetics of fluorescence quenching
by electron and H-atom transfer, Isr. J. Chem., 1970, 8, 259–
271.
aminobenzonitrile in polar solvents, J. Chem. Soc., Faraday
Trans. 2, 1980, 76, 453–471.
48 X. M. Zhang, S. R. Yeh, S. Hong, M. Freccero, A. Albini,
D. E. Falvey and P. S. Mariano, Dynamics of alpha-CH depro-
tonation and alpha-desilylation reactions of tertiary amine
cation radicals, J. Am. Chem. Soc., 1994, 116, 4211–4220.
43 R. A. Marcus, On the theory of oxidation-reduction reac- 49 G. W. Dombrowski, J. P. Dinnocenzo, P. A. Zielinski,
tions involving electron transfer .1., J. Chem. Phys., 1956,
24, 966–978.
44 S. Canonica, B. Hellrung and J. Wirz, Oxidation of phenols
S. Farid, Z. M. Wosinska and I. R. Gould, Efficient unimole-
cular deprotonation of aniline radical cations, J. Org.
Chem., 2005, 70, 3791–3800.
by triplet aromatic ketones in aqueous solution, J. Phys. 50 J. M. Burns, W. J. Cooper, J. L. Ferry, D. W. King,
Chem. A, 2000, 104, 1226–1232.
B. P. DiMento, K. McNeill, C. J. Miller, W. L. Miller,
B. M. Peake, S. A. Rusak, A. L. Rose and T. D. Waite,
Methods for reactive oxygen species (ROS) detection in
aqueous environments, Aquat. Sci., 2012, 74, 683–734.
45 M. Jonsson, J. Lind, T. E. Eriksen and G. Merenyi, Redox
and acidity properties of 4-substituted aniline radical
cations in water, J. Am. Chem. Soc., 1994, 116, 1423–1427.
Photochem. Photobiol. Sci.
This journal is © The Royal Society of Chemistry and Owner Societies 2018