Photolysis of cycloxydim, a cyclohexanedione oxime herbicide. Detection, characterization and reactivity of the iminyl radical

Article abstract of DOI:10.1039/c3pp50209k

Cyclohexanedione oxime herbicides have been reported to be readily photodegraded in the environment but the reaction mechanism has never been studied in detail. Here we investigated the photolysis of cycloxydim (CD) in acetonitrile and water, solvents in which CD is present as two distinct tautomeric forms: keto form in water and enol form with an intramolecular hydrogen bond between the enolic proton and the nitrogen atom of the oxime in acetonitrile O-H...N. CD (E isomer) undergoes photoisomerization in water but not in acetonitrile. This difference is attributed to the inhibiting effect of the intramolecular hydrogen bond existing in acetonitrile. In both solvents, irradiation of CD leads to the cleavage of the N-O bond as evidenced by the imine formation. The iminyl radical could be detected in acetonitrile by nanosecond laser-flash photolysis (λ_max = 280, 320 and 480 nm, τ ～ 100 μs). This radical is unreactive toward oxygen but readily abstracts an H atom from methanol (k = 1.8 × 10⁵ M^-1 s^-1). Quantum calculations confirm the assignment of the transient species to the iminyl radical by showing that (i) the most stable structure of the iminyl carries a large spin density on the ring carbon and on the nitrogen atoms, (ii) the enolic H atom is transferred to the nitrogen atom and (iii) the intramolecular hydrogen bond O...H-N is responsible for both the iminyl long wavelength absorption and its high hydrogen abstraction reactivity.

Full text of DOI:10.1039/c3pp50209k

An international journal
www.rsc.org/pps
Volume 12 | Number 12 | December 2013 | Pages 2059–2206
ISSN 1474-905X
PAPER
Claire Richard et al.
Photolysis of cycloxydim, a cyclohexanedione oxime herbicide. Detection,
characterization and reactivity of the iminyl radical
1474-905X(2013)12:12;1-T

Photochemical &
Photobiological Sciences
View Article Online
View Journal | View Issue
PAPER
Photolysis of cycloxydim, a cyclohexanedione oxime
herbicide. Detection, characterization and reactivity of
the iminyl radical
Cite this: Photochem. Photobiol. Sci., 2013,
12, 2067
Shirin Monadjemi,a Pascal de Sainte-Claire,a,b Isabelle Abrunhosa-Thomasb,c and
Claire Richard*a,d
Cyclohexanedione oxime herbicides have been reported to be readily photodegraded in the environment
but the reaction mechanism has never been studied in detail. Here we investigated the photolysis of
cycloxydim (CD) in acetonitrile and water, solvents in which CD is present as two distinct tautomeric
forms: keto form in water and enol form with an intramolecular hydrogen bond between the enolic
proton and the nitrogen atom of the oxime in acetonitrile O–H⋯N. CD (E isomer) undergoes photo-
isomerization in water but not in acetonitrile. This difference is attributed to the inhibiting effect of the
intramolecular hydrogen bond existing in acetonitrile. In both solvents, irradiation of CD leads to the
cleavage of the N–O bond as evidenced by the imine formation. The iminyl radical could be detected in
acetonitrile by nanosecond laser-flash photolysis (λmax = 280, 320 and 480 nm, τ ∼ 100 μs). This radical is
unreactive toward oxygen but readily abstracts an H atom from methanol (k = 1.8 × 105 M−1 s−1).
Quantum calculations confirm the assignment of the transient species to the iminyl radical by showing
that (i) the most stable structure of the iminyl carries a large spin density on the ring carbon and on the
nitrogen atoms, (ii) the enolic H atom is transferred to the nitrogen atom and (iii) the intramolecular
hydrogen bond O⋯H–N is responsible for both the iminyl long wavelength absorption and its high
hydrogen abstraction reactivity.
Received 3rd July 2013,
Accepted 4th September 2013
DOI: 10.1039/c3pp50209k
www.rsc.org/pps
Cyclohexanedione oxime herbicides are reactive molecules
not only under irradiation but also in the dark in acidic
Introduction
Cyclohexanedione oxime herbicides are a class of compounds medium.1–9 Scheme
2 summarizes the main reactions
which show an excellent herbicidal activity against various observed in the dark and under light exposure: isomerization
weed species.1 Moreover, some maize varieties are tolerant to of the oxime, oxazole and imine formations. Isomerization of
this class of herbicides. This tolerance increases their potential the E-isomer into the Z-isomer has been reported in water, in
interest as herbicides and justifies studies aiming at a better the dark and under irradiation.2,5 The dark reaction, on the
understanding of their fate in the environment. The chemical one hand, is made possible by the acidic hydrolysis of the
structures of some of the main herbicides of this family are oxime which allows the rotation of the O-alkyl about the C–N
given in Scheme 1. These species bear two keto groups that bond.1 The photochemical isomerization of oximes, on the
can undergo enolization and an O-substituted oxime which other hand, is well-known and studied for a long time.10 The
may be represented by either one of the isomers E and Z. The formation of oxazole from cyclohexanedione oxime herbicides
E form is more active as a herbicide, and cyclohexanedione has also been reported.4 In the dark, this reaction occurs via
oximes are therefore marketed in this isomeric form.
N protonation followed by cleavage of the N–O bond, loss of
R3OH, Beckmann type rearrangement and final cyclization.
Oxazole can also be produced photochemically, the reaction
being observed in various media such as water, methanol
and soil surface. Lastly, the imine is the main photoproduct
in solid and in solution. Its formation is very minor in the
dark.
The oxime function is clearly at the core of the (photo)reac-
tivity of these compounds. Thus, this is important for ex-
plaining why photoisomerization of cyclohexanedione oxime
aClermont Université, Université Blaise Pascal, Institut de Chimie de Clermont-
Ferrand, Equipe Photochimie, BP 10448, F-63000 Clermont-Ferrand, France
bClermont Université, ENSCCF, Institut de Chimie de Clermont-Ferrand, BP 10448,
F-63000 Clermont-Ferrand, France
cClermont Université, Université Blaise Pascal, Institut de Chimie de Clermont-
Ferrand, Equipe SEESIB, BP 10448, F-63000 Clermont-Ferrand, France.
E-mail: Claire.richard@univ-bpclermont.fr; Fax: +33 04 73 40 77 00;
Tel: +33 04 73 40 71 42
dCNRS, UMR 6296, ICCF, F-63171 Aubiere, France
This journal is © The Royal Society of Chemistry and Owner Societies 2013
Photochem. Photobiol. Sci., 2013, 12, 2067–2075 | 2067

View Article Online
Paper
Photochemical & Photobiological Sciences
Scheme 1 Chemical structure of some cyclohexanedione oxime herbicides.
Experimental part
Chemicals
Cycloxydim (98.1%) was obtained from Fluka and used as
received. Acetonitrile was purchased from Prolabo (Hypersolv
Chromanorm, 99.9% purity) and methanol from Sigma
Aldrich (Chromasolv, 99.9% purity). The ultra-pure water used
was from a Millipore Milli-Q system.
Irradiation
Steady-state irradiation at 334 nm was conducted in a device
equipped with a high pressure Hg lamp (200 W) and a Bausch
and Lomb monochromator. For irradiation at 278 nm, we used
a high pressure xenon lamp (1600 W) equipped with a Bausch
and Lomb monochromator. Photon fluence rates were
measured using an Ocean Optics radiometer. These rates were
3.9 × 1014 photons cm−2 s−1 at 334 nm and 3.8 × 1013 photons
cm−2 s−1 at 278 nm. Laser flash photolysis was carried out
using a nanosecond laser-flash photolysis spectrometer from
Applied Photophysics (LKS.60) with laser excitation at 266 nm
from a Quanta-Ray GCR130-1 Nd:YAG laser. Experimental
details are described elsewhere.12 Solutions were deoxygenated
or suroxygenated directly in the cell under argon or oxygen
flow, respectively. The laser pulse energy (P) was measured
using a joulemeter from Ophir Optronics Ltd. The absorbance
at the pulse end (A) was measured for various P values to
check the linearity of the dependence of A on P. In this case,
the slope A/P is proportional to the ε × Φ product of the transi-
ent where ε is the molar absorption coefficient and Φ the
quantum yield of formation of the transient. Benzophenone
was used as a chemical actinometer and photolyzed at the
same absorbance at 266 nm as CD. The formation of the
benzophenone triplet excited state was monitored at 520 nm
(εT × ΦT = 6500 M−1 cm−1).13 The ε × Φ product of the CD tran-
sient was obtained using the relationship: ε × Φ = 6500 × (A/P)-
(A/P)BP−1. EPR experiments were conducted using a Bruker
EMX EPR spectrometer interfaced with a computer equipped
with WinEPR Acquisition System software. CD solutions were
Scheme 2 The main photochemical and dark reactions of cyclohexanedione
oximes reported in the literature.
herbicides occurs only in water or protic solvents. In this work,
this point is addressed by looking at the relationship between
solvent interactions and molecular configuration, with particu-
lar emphasis on the keto–enol tautomerism.
Another interesting issue is the nature of the intermediate
species involved in the imine formation. The photocleavage of
the N–O bond is a well-known reaction and yields the iminyl
and alkoxy radicals.11 EPR was successfully used in the litera-
ture to characterize iminyl radicals. To our knowledge, this
radical was rarely observed by laser flash photolysis.
Our goal was to understand the photolysis mechanism of
these compounds, in particular to explain the absence of
photoisomerization in aprotic solvents and characterize the
intermediate species involved in the imine formation. Experi-
ments conducted on cycloxydim were rarely carried out in
acetonitrile and neutral water. In addition to the steady-state
irradiation that was used to measure the quantum yields of
photolysis and characterize the main photoproducts, we inves-
tigated the reaction by means of nanosecond laser flash pho-
tolysis. The mechanisms were supported by quantum
calculations.
irradiated using
Instruments).
a 1000 W Xe–Hg arc lamp (Schoeffel
2068 | Photochem. Photobiol. Sci., 2013, 12, 2067–2075
This journal is © The Royal Society of Chemistry and Owner Societies 2013

View Article Online
Photochemical & Photobiological Sciences
Paper
Analyses
UV-visible spectra were recorded using a Cary 3 (Varian)
spectrophotometer.
The HPLC analyses were performed using a Waters appar-
atus equipped with a pump model 515 and a detector model
996. The column was a Kinetex™ C18, 100 mm × 2.1 mm, with
a particle size of 2.6 µm, and a size pore of 100 Å. The eluent
was a water–acetonitrile mixture. The mixture was kept at
50 : 50 (v/v) for 3 min and then the percentage of acetonitrile
was increased linearly to reach 80% at 6 min. The flow was set
at 0.2 cm3 min−1. Mass spectrometry was performed using an
LC/QTOF apparatus equipped with an orthogonal geometry
Z-spray ion source (Waters/Micromass, Manchester, UK). Ana-
lyses were performed in positive mode. UV detection under MS
heading was done using a Waters Alliance 2695 photodiode
array detector system. The same conditions (column, gradient
and flow) as those described above were used.
Fig. 1 UV-spectrum of CD in acetonitrile (dotted line) and in neutral water
(full line).
The NMR spectroscopy analyses of CD were performed in
CD3CN and D2O and recorded using a Brüker 400 MHz
(101 MHz for 13C) spectrometer using TMS as an internal stan-
dard. The following data were obtained:
1H NMR (400 MHz, CH3CN) δ 15.04 (s, 1H), 4.11 (q, J =
7.0 Hz, 2H), 2.85 (dd, J = 7.7 Hz, 2H), 2.58 (td, J = 11.4, 3.0 Hz,
2H), 2.54–2.42 (m, 2H), 2.38–2.25 (m, 2H), 2.10–1.99 (m, 2H),
1.85 (m, 1H), 1.71–1.7 (m, 2H), 1.52 (sext, J = 7.8 Hz, 2H), 1.28
(t, J = 7.0 Hz, 3H), 1.21 (m, 1H), 0.93 (t, J = 7.5 Hz, 3H).
13C NMR (101 MHz, CD3CN) δ 195.6, 185.6, 163.7, 107.1,
69.6, 41.3, 36.3, 30.3, 28.7, 28.3, 27.8, 27.3, 19.6, 13.3, 13.20.
13C NMR (101 MHz, H2O + D2O + NaOH) δ 195.2, 195.1,
161.5, 111.0, 68.8, 41.2, 38.5, 37.9, 31.4, 30.6, 28.8, 27.9, 27.1,
18.3, 13.9, 13.8.
Scheme 3 Keto–enol tautomeric structures of CD.
tautomeric forms. The keto–enol tautomeric structures of CD
are shown in Scheme 3.
To delineate which form is predominant depending on the
solvent, 1H and 13C NMR spectra were recorded. Data obtained
by 13C-NMR confirm that CD is in equilibrium between keto
and enol forms and that this equilibrium depends on
the solvent used for the NMR analysis. As shown in Fig. 2, in
CD3CN, carbons C1 and C3 are not equivalent, the chemical
shift being 185.6 and 195.6 ppm. In D2O, carbons C1 and C3
are equivalent and symmetric and give a quasi-identical
chemical shift respectively at 195.1 and 195.2 ppm. We con-
clude that in D2O, CD is in the keto form, C1 and C3 belonging
to the carbonyls groups, while in CD3CN, CD is in the enol
form with C1 and C3 belonging to the enol and the carbonyl,
respectively. Moreover, the 1H-NMR spectroscopy analysis
performed in CH3CN reveals the presence of a signal at
14.14 ppm that is characteristic of the displacement of the
enolic hydrogen atom (Fig. 3). This confirms the presence of
an intramolecular hydrogen bond between the enol and the
nitrogen of the oxime.
Quantum calculations
The theoretical calculations were performed with the Gaussian
program14 at the DFT level with the hybrid B3LYP exchange
and correlation functional and the 6-31G(d,p) basis set.
Unrestricted Hamiltonians were used for the treatment of
radical species. Geometries were fully optimized in the ground
state and time dependent DFT (TD-DFT) calculations were per-
formed at the B3LYP/6-31G(d,p) level to obtain the absorption
spectra. The calculation of bond dissociation energies was also
performed at the B3LYP/6-31G(d,p) level.
Results and discussion
Determination of the CD structure in acetonitrile and water
The absorption spectra of CD in acetonitrile and neutral water
are given in Fig. 1. Significant differences are observed. In
acetonitrile, the main absorption band is centered at 250 nm
Steady-state photolysis
(ε = 11 200 M−1 cm−1) and a smaller one (ε = 6800 M−1 cm−1
)
Under light exposure, CD (E isomer) disappears rapidly in
appears at ∼280 nm. In water, the predominant band (ε = acetonitrile or in water. Photolysis quantum yields, ΦCD, were
15 600 M−1 cm−1) is centered at 282 nm. These spectral measured for a fixed concentration of CD (10−4 M) by monitor-
changes suggest that the solvent affects the keto–enol tauto- ing the CD loss by HPLC. Irradiations were performed at
merism and that acetonitrile and water stabilize two different 334 nm when acetonitrile was used as the solvent and at
This journal is © The Royal Society of Chemistry and Owner Societies 2013
Photochem. Photobiol. Sci., 2013, 12, 2067–2075 | 2069

View Article Online
Paper
Photochemical & Photobiological Sciences
Fig. 2 13C-NMR spectra of CD in CD3CN and D2O.
278 nm when the reaction was investigated in water. In both value of 282 corresponding to the loss of 44 u (O–CH2–CH2)
cases, CD was irradiated in the absorption band extending and is assigned to the imine. Photoproduct II has the same
above 295 nm, i.e. in the relevant one as far as environmental mass fingerprint as isomer E, but a retention time that is
photolysis is concerned. ΦCD are (3.9 0.4) × 10−2 for aerated much shorter (3.3 versus 10.2 min). It is assigned to isomer Z.
and (1.7 0.2) × 10−2 for argon-purged acetonitrile and (1.7
Photoproduct III gives an m/z value of 342 corresponding to
0.3) × 10−2 for aerated water. The latter value is higher than the addition of 16 u (O) and is assigned to an oxidation
those reported in the market authorization application of the product. Lastly, photoproduct IV shows an m/z value of 280
compound: 5.68 × 10−3 at pH 5, 1.87 × 10−4 at pH 7 and 2.02 × corresponding to the loss of 46 u (HO–CH2–CH3) and is con-
10−4 at pH 9 and obtained upon irradiation of CD with poly- sistent with the oxazole. These compounds are mentioned in
chromatic light.15 From our data, the quantum yield of CD the market authorization application.17 In acetonitrile, only
photolysis is significant in water and in acetonitrile (1.7 × 10−2 the imine(I) and the oxidation product (III) are found.
and 3.9 × 10−2 in air-saturated medium). Adding to this the
Laser flash photolysis
extending of the absorption band and the high extinction
coefficients of CD until 315 nm in water and 330 nm in aceto-
nitrile, one can expect a decay of CD under environmental
conditions as is reported for the other cyclohexanedione
oxime herbicides.1
The main photoproducts were assigned by HPLC-MS in
positive mode. Data are presented in Table 1. In water, four
main photoproducts are detected along with CD (m/z = 326 in
ES+, retention time = 10.2 min). Photoproduct I shows an m/z
The irradiation of aerated solutions of CD in acetonitrile gives
at the end of the pulse the transient absorption spectrum
shown in Fig. 4. This spectrum is very broad extending from
275 to 520 nm. The absorption maximum is located around
280 nm. Neither the formation nor the decay of this species is
affected by argon-purging of the solution. The transient
absorption decays slowly at 280, 350 and 500 nm with an
apparent first order rate constant of 9000 s−1 in air or in
2070 | Photochem. Photobiol. Sci., 2013, 12, 2067–2075
This journal is © The Royal Society of Chemistry and Owner Societies 2013

View Article Online
Photochemical & Photobiological Sciences
Paper
Fig. 3 1H-NMR spectra of CD in CD3CN.
Table 1 Quantum yields of CD photolysis in various solvents and the main photoproducts
Quantum yield
Photoproducts in air saturated medium
Solvent (irradiation wavelength)
Water, neutral pH (278 nm)
Aerated medium
(1.7 0.3) × 10−2
Deoxygenated medium
(1.7 0.2) × 10−2
(m/z in ES+, retention time, function)
I (282, 3.0 min, imine)
II (326, 3.3 min, isomer Z)
III (342, 3.5 min, oxidation product)
IV (280, 4.6 min, oxazole)
Acetonitrile (334 nm)
(3.9 0.4) × 10−2
I (282, 3.0 min, imine)
III (342, 3.5 min, oxidation product)
argon-saturated medium. Similarly, varying the CD concen- Quantum calculations
tration within the range 5 × 10−5–2 × 10−4 M does not change
the decay rate. It can be concluded that a unique species is The formation of the imine indicates that the N–O bond is
formed which does not react with oxygen or CD itself. We tried cleaved. Such a bond was shown to be broken with the for-
to trap this species using methanol as an H-donor molecule. mation of two radicals.16 Therefore, the theoretical calcu-
The experiment was successful since the addition of methanol lations aimed first at identifying the most stable structure that
at levels between 10 and 66% (v/v) drastically increased the was obtained upon the homolytic N–O bond breaking. Three
transient decay rate. Plotting the apparent first order decay structures (see Scheme 4) of doublet spin multiplicity were
rate constant against methanol concentration gives a linear optimized in the ground state at the B3LYP/6-31G(d,p) level of
relationship and a bimolecular rate constant of reaction of theory.
1.8 × 105 M−1 s−1 (see the inset of Fig. 4). The intensity of
While the non-resonant C species is clearly disfavored, it is
absorbance at the pulse end increases linearly with the energy found here that A is significantly more stable than structure B
pulse demonstrating that the formation of this species is (EA − EB = −10 kcal mol−1) because the O⋯H intramolecular
monophotonic. By actinometry, ε × ϕ = 150 20 and 30 10 hydrogen bonding is strong in A (RO⋯H = 1.890 Å, RN⋯H
=
M−1 cm−1 were obtained at 350 and 500 nm, respectively. 1.028 Å). Note that B is twisted at the CCCN bond at this level
Assuming ϕ = 1.7 × 10−2, i.e. the value that was obtained by of theory (δ′ = −115.8°, see Scheme 4). Thus, B is expected to
steady-state irradiation in argon-saturated medium, one rearrange easily into A through CCOH internal rotation and
deduces that the molar extinction coefficient of the transient hydrogen shift. The spin density (SD) analysis on structure A
species is equal to 8800 1000 M−1 cm−1 at 350 nm and 1700 shows that the radical is mostly localized on the carbon atom
600 M−1 cm−1 at 500 nm. In water this transient species is that bridges the ring and the iminyl group: SDC(1) = 0.63, SDN
also observed but the signal is too weak to be followed.
= 0.41 (see Scheme 4 for labels). Note that the thianyl ring is
This journal is © The Royal Society of Chemistry and Owner Societies 2013
Photochem. Photobiol. Sci., 2013, 12, 2067–2075 | 2071

View Article Online
Paper
Photochemical & Photobiological Sciences
Fig. 5 Absorption spectrum of the iminyl radical (structure A in Scheme 4)
Fig. 4 Transient species measured at the pulse end upon irradiation of CD
(10−4 M) in acetonitrile. Absorbance at 266 nm = 0.5.
computed by the TD-DFT method at the B3LYP/6-31G(d,p) level.
twisted in the optimized structure (δCCCC = −55.7°) (see
Scheme 4 for labels).
expected to prevent such internal rotation and thus inhibits
photoisomerization. The excited E isomer yields back the
ground state isomer E. In contrast, in water, where the keto
form is favoured due to intermolecular hydrogen bonds
between the carbonyl and solvent OH, internal rotation is free
and isomer Z can be obtained. The inhibiting effect of intra-
molecular hydrogen bonds on photoisomerization has been
reported in the literature.17,18 Usually, photoisomerization is
not found in non-polar solvents but observed in alcohol or
acidic solvents because the latter are supposed to disrupt the
intramolecular hydrogen bond.
The photochemistry of CD in acetonitrile involves the
oxime. The N–O bond cleavage produces the iminyl and the
alkoxy radicals. Quantum calculations show that the stable
form of the iminyl radical has the electronic charge deloca-
lized on the O–C–C–N atoms with the highest density on the
nitrogen and carbon atoms. Moreover, the H-atom is borne on
the nitrogen. This implies that a fast H-atom transfer from the
enol to the nitrogen occurs after the N–O bond cleavage. Using
these data, it can be deduced that the long-lived transient
species observed in acetonitrile by laser flash photolysis is the
iminyl radical. In accordance, this radical does not react with
oxygen but is scavenged by methanol to yield the imine
(Scheme 5).
The absorption spectrum of structure A was calculated by
the TD-DFT method at the B3LYP/6-31G(d,p) level and is
shown in Fig. 5. The characteristic absorption at 455 nm, the
shoulder at 320 nm and the 240 nm absorption give strong
evidence that the transient species observed experimentally
(see Fig. 4) is the iminyl radical proposed in structure A. The
absorption band observed experimentally between 450 nm and
550 nm is built upon the n/π* (λtheo = 548 nm) and π/π*
transitions (λtheo = 444 nm). The strong O⋯H intramolecular
hydrogen bond stabilizes the π* orbital because the latter
bears local bonding character on the O⋯HN atoms. This tran-
sition is thus shifted towards lower energies. The molecular
orbitals involved in these transitions are shown in Fig. 6.
Mechanism
The photochemistry of CD is highly medium dependent. In
acetonitrile, CD gives mainly two photoproducts: the imine
and an oxidation product. In water there are two additional
pathways: photoisomerization and oxazole formation. To
rationalize the absence of photoisomerization in acetonitrile,
it is necessary to consider the effect of the medium on the
molecule structure. NMR and absorption data demonstrate
that in a non-protic solvent such as acetonitrile, the enol form
is stabilized by an intramolecular hydrogen bond with the
oxime nitrogen. After the π/π* oxime excitation, the CvN bond
is weakened. This allows isomerization about the CN bond.
However, here, the presence of the intramolecular bond is
Iminyl radicals have been intensively studied for synthesis
purposes to generate new rings by intramolecular cyclisation.
They can also abstract hydrogen atoms to yield imines: the rate
constants for H-abstraction by iminyl radicals are reported to
be more than an order of magnitude slower than that for
Scheme 4 Structures of the iminyl radical that were investigated in this work, δ = −55.7° and δ’ = −115.8°.
2072 | Photochem. Photobiol. Sci., 2013, 12, 2067–2075
This journal is © The Royal Society of Chemistry and Owner Societies 2013

View Article Online
Photochemical & Photobiological Sciences
Paper
Scheme 5 Formation and reactivity of the iminyl radical.
C-centered analogues.19 For instance, the abstraction of a
methanolic hydrogen (CH3O–H) by methyl radicals at 300 K is
∼102 M−1 s−1 20
Iminyl radicals should therefore be less reac-
.
tive. Thus, the rate constant of the alcohol H-atom abstraction
reaction between the radical obtained from CD and methanol
seems rather high, indicating that this specific radical is more
reactive than most of the iminyl species. Accordingly, attempts
to detect the iminyl derived from CD by EPR were
unsuccessful.
To explain such an unexpectedly high reactivity, dis-
sociation energy calculations of the N–H bond (DN–H in units
of kcal mol−1) were performed for three possible imine models
(Scheme 6). In (a), the model species is representative of the
iminyl radical investigated in this work: the key features of
electron delocalization and the intramolecular hydrogen bond
are accounted for in this structure. One gets DN–H = 98.4 kcal mol−1
Fig. 6 The B3LYP/6-31G(d,p) π, n and π* molecular orbitals involved in the
450–550 nm band of the absorption spectrum shown in Fig. 5.
Scheme 6 Theoretical N–H bond strengths of model molecules at the B3LYP/6-31G(d,p) level. D is purely electronic.
This journal is © The Royal Society of Chemistry and Owner Societies 2013
Photochem. Photobiol. Sci., 2013, 12, 2067–2075 | 2073

View Article Online
Paper
Photochemical & Photobiological Sciences
in this case. In (b), where one unit of unsaturation is of compounds, the intramolecular hydrogen bond present in
removed, one finds that DN–H = 104.2 kcal mol−1. Clearly, the aprotic solvents governs a part of the photoreactivity. This
electronic delocalization weakens the N–H bond. In (c) the car- feature explains the lack of photoisomerization under specific
bonyl group was rotated to suppress the intramolecular hydro- conditions. The intramolecular hydrogen bond in the iminyl
gen bond. Two different cases (ca and cb) were investigated. radical also reinforces significantly the ability of this radical
The most stable structure is that obtained in (cb), where the to abstract hydrogen atoms in comparison to other iminyl
singly occupied orbital is directed toward the methyl substitu- radicals. Under environmental conditions, in surface water or
ent. The respective N–H bond strength is DN–H = 88.7 kcal on a soil surface, the iminyl radicals generated from cyclo-
mol−1. In this model, the corresponding radical iminyl no hexanedione oximes would have many possibilities to react
longer bears the intramolecular hydrogen bond. For the other with natural compounds, either by H abstraction or by
structure (ca) where the remaining H atom of the iminyl addition on double bonds. In this work, the iminyl radical was
radical is directed toward the methyl group, DN–H = 91.8 kcal successfully detected in acetonitrile and its reactivity toward
mol−1. These calculations demonstrate the drastic reinforcing H-donor molecules measured.
effect of the iminyl intramolecular hydrogen bond O⋯H–N on
the strength of the second N–H bond.
Using these data, it is possible to estimate the effect of the
intramolecular hydrogen bond on the iminyl H-abstraction
References
ability. Let us consider the reaction
1 P. Sandín-España, B. Sevilla-Morán, J. L. Alonso-Prados and
I. Santín-Montanya, Chemical Behaviour and Herbicidal
Activity of Cyclohexanedione Oxime Herbicides, in
Herbicides – Properties, Synthesis and Control of Weeds, ed.
M. N. Abd El-Ghany, Agricultural and Biological Sciences
series, 2012, pp. 75–102.
2 L. N. Falb, D. C. Bridges and A. E. Smith, Effects of pH and
adjuvants on clethodim photodegradation, J. Agric. Food
Chem., 1990, 38, 875–878.
RN• þ MeOH ! RNH þ MeO•
The activation energies may be expressed as Ea = Cte +
21
α ΔH, with ΔH = DMeO–H − DRN–H
.
If RN–H and R′N–H rep-
resent the reactant molecules (a) and (cb), respectively, the
ratio of the rate constants k/k′ is:
3 L. N. Falb, D. C. Bridges and A. E. Smith, Separation of
clethodim herbicide from acid and photodegradation
products by liquid chromatography, J. Assoc. Off. Anal.
Chem. Int., 1991, 74, 999–1002.
4 Metabolic pathways of agrochemicals, ed. T. R. Roberts,
Royal Society of Chemistry, Cambridge, UK, 1998.
5 B. Sevilla-Morán, P. Sandín-España, M. J. Vicente-Arana,
J. L. Alonso-Prados and J. M. García-Baudín, Study of
alloxydim photodegradation in the presence of natural
substances: Elucidation of transformation products,
J. Photochem. Photobiol., A, 2008, 198, 162–168.
k
k′
RNꢀHꢀDR′NꢀHÞ=RT
¼ eαðD
DRN–H − DR′N–H is 9.7 kcal mol−1. Taking α = 0.55,22 one gets k/
k′ = 8.1 × 103 at 298.15 K. Thus, the presence of the intramole-
cular hydrogen bond increases the abstraction rate constant by
four orders of magnitude. These calculations confirm the
experimental data.
The imine formation upon irradiation of CD in pure aceto-
nitrile requires the abstraction of a mobile H atom. Some
solvent impurities or CD itself might be the H-donor, but the
concentration used here is low, making this reaction improb-
able. Alternatively, the H atom could be abstracted from the
ethoxy radical that is produced along with the iminyl species
through the N–O bond cleavage.
6 P. Sandín-España, J. O. Magrans and J. M. García-Baudín,
Study of clethodim degradation and by-product formation
in chlorinated water by HPLC, Chromatographia, 2005, 62,
133–137.
Along with the imine, an oxidation product (III) corres-
ponding to CD + O is formed upon CD irradiation. Taking into
account that the quantum yield of CD phototransformation is
higher in the presence of oxygen than in its absence, it could
be produced by oxidation of the CD excited state by oxygen.
However, we did not observe the excited states by laser flash
photolysis, in particular the triplet excited state. This absence
may be due to several reasons: poor quantum yield of
formation, poor absorption coefficient or a short lifetime.
Another possibility for the formation of product III could be
the oxidation of CD by radicals obtained in the N–O cleavage.
However, this would not explain the quantum yield increase in
the presence of oxygen.
7 B. Sevilla-Morán, M. M. Mateo-Miranda, J. L. Alonso-
Prados, J. M. García-Baudín and P. Sandín-España, Inter-
nat. Sunlight transformation of sethoxydim-lithium in
natural waters and effect of humic acids, J. Environ. Anal.
Chem., 2010, 90, 487–496.
8 B. Sevilla-Morán, J. L. Alonso-Prados, J. M. García-Baudín
and P. Sandín-España, Indirect photodegradation of cletho-
dim in aqueous media. byproduct identification by quadru-
pole time-of-flight mass spectrometry, J. Agric. Food Chem.,
2010, 58, 3068–3076.
9 P. Sandín-España, B. Sevilla-Morán, L. Calvo, M. Mateo-
Miranda and J. L. Alonso-Prados, Photochemical behavior
of alloxydim herbicide in environmental waters. Structural
elucidation and toxicity of degradation products, Micro-
chem. J., 2013, 106, 212–219.
To conclude, this study rationalizes the different results
reported on cyclohexanedione oxime herbicides. For this class
2074 | Photochem. Photobiol. Sci., 2013, 12, 2067–2075
This journal is © The Royal Society of Chemistry and Owner Societies 2013

View Article Online
Photochemical & Photobiological Sciences
Paper
10 A. Padwa and F. Albrecht, Photoisomerization about the
carbon-nitrogen double bond of an oxime ether, J. Am.
Chem. Soc., 1972, 94, 1000–1002.
11 A. R. Forrester, M. Gill, C. J. Meyer, J. S. Sadd and
R. H. Thomson, Iminyls. Part 1. Generation of iminyls,
J. Chem. Soc., Perkin Trans, 1979, 3, 606–611.
Gaussian09, Revision C.01, Gaussian, Inc., Wallingford,
CT, 2010.
15 Austrian Agency for Health and Food Safety. CLH report.
Proposal for harmonised classification and labelling, based
on regulation (EC) No 1272/2008. Cycloxydim. Annex VI.
part 2.
12 F. Bonnichon and C. Richard, Phototransformation of 16 A. J. McCarroll and J. C. Walton, Photolytic and radical
3-hydroxybenzonitrile in water, J. Photochem. Photobiol., A,
1998, 119, 25–32.
induced decompositions of O-alkyl aldoxime ethers,
J. Chem. Soc., Perkin Trans. 2, 2000, 1868–1875.
13 I. Carmichael and G. L. Hug, Triplet-triplet absorption 17 F. D. Lewis and A. Beth, Influence of the intramolecular
spectra of organic molecules in condensed phases, J. Phys.
Chem. Ref. Data, 1986, 15, 1–250.
14 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
hydrogen bonding on the structure and E-Z photoisomeri-
zation of urocanic acid derivatives, Res. Chem. Intermed.,
1995, 21, 749–763.
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, 18 D. A. Lightner and Y. T. Park, Synthesis, photooxidation
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino,
and Z-E photoisomerization of benzalpyrrolinones,
J. Heterocycl. Chem., 1977, 14, 415–422.
G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, 19 M.-H. Le Tadic-Biadatti, A.-C. Callier-Dublanchet,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida,
T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven Jr.,
J. A. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark,
J. H. Horner, B. Quiclet-Sire, S. Z. Zard and M. Newcomb,
Absolute Rate Constants for Iminyl Radical Reactions,
J. Org. Chem., 1997, 62, 559–563.
J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, 20 P. Neta, J. Grodbowski and A. B. Ross, Rate constants for
T. Keith, R. Kobayashi, J. Normand, K. Raghavachari,
A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi,
reactions of aliphatic carbon-centered radicals in aqueous
solution, J. Phys. Chem. Ref. Data, 1996, 25, 709–1051.
M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, 21 M. G. Evans and M. Polanyi, Inertia and driving force of
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, chemical reactions, Trans. Faraday Soc., 1938, 34, 11–24.
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, 22 S. Korcek, J. H. B. Chenier, J. A. Howard and K. U. Ingold,
R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas,
J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox,
Absolute Rate Constants for Hydrocarbon Autoxidation.
XXI. Activation Energies for Propagation and the Corre-
lation of Propagation Rate Constants with Carbon-Hydro-
gen Bond Strengths, Can. J. Chem., 1972, 50, 2285–2297.
This journal is © The Royal Society of Chemistry and Owner Societies 2013
Photochem. Photobiol. Sci., 2013, 12, 2067–2075 | 2075