NMR Kinetic Investigations of the Photochemical and Thermal Reactions of a Photochromic Chromene

Article abstract of DOI:10.1021/jo035279r

The photochromic behavior of 2,2-di(4-fluorophenyl) -6-methoxy-2H-1-chromene has been investigated by ¹⁹F NMR spectroscopy. Photocoloration under UV irradiation at low temperature led to the formation of three interconverting photoisomers including two merocyanines and an unprecedented allenyl-phenol isomer. Photobleaching with visible light, which was known to lead to reversion to the initial closed chromene, was shown to increase allenyl-phenol concentration. Thermal relaxation of the preirradiated system was also studied at various temperatures. In each case (UV and visible irradiations, thermal isomerization), the kinetics of each of the four species was monitored. Numerical analysis of concentration vs time profiles enabled us to unequivocally establish the global mechanism occurring in each of the experimental conditions and to interpret the specific reactivity of each photoisomer. It has been shown that, among the 12 possible isomerization processes, only some paths were active. For the first time, it has been possible to determine their corresponding thermal activation parameters and photochemical quantum yield ratios.

Full text of DOI:10.1021/jo035279r

NMR Kin etic In vestiga tion s of th e P h otoch em ica l a n d Th er m a l
Rea ction s of a P h otoch r om ic Ch r om en e
Ste´phanie Delbaere,*,† J ean-Claude Micheau,‡ and Gaston Vermeersch†
Laboratoire de Physique et LARMN, UMR CNRS 8009, Faculte´ de Pharmacie, Universite´ de Lille 2,
F-59006 Lille, France, and Laboratoire IMRCP, UMR CNRS 5623, Universite´ Paul Sabatier,
F-31062 Toulouse, France
sdelbaer@pharma.univ-lille2.fr
Received September 2, 2003
The photochromic behavior of 2,2-di(4-fluorophenyl)-6-methoxy-2H-1-chromene has been investi-
gated by 19F NMR spectroscopy. Photocoloration under UV irradiation at low temperature led to
the formation of three interconverting photoisomers including two merocyanines and an unprec-
edented allenyl-phenol isomer. Photobleaching with visible light, which was known to lead to
reversion to the initial closed chromene, was shown to increase allenyl-phenol concentration.
Thermal relaxation of the preirradiated system was also studied at various temperatures. In each
case (UV and visible irradiations, thermal isomerization), the kinetics of each of the four species
was monitored. Numerical analysis of concentration vs time profiles enabled us to unequivocally
establish the global mechanism occurring in each of the experimental conditions and to interpret
the specific reactivity of each photoisomer. It has been shown that, among the 12 possible
isomerization processes, only some paths were active. For the first time, it has been possible to
determine their corresponding thermal activation parameters and photochemical quantum yield
ratios.
In tr od u ction
eral photoisomers impedes the unequivocal extraction of
the photoconversion quantum yields and molar absorp-
tion coefficients.
Photochromic molecules have received great attention
in recent years due to their potential applications in
variable transmission glasses, in high-density optical
storage and switching, and in some attempts to manu-
facture light-driven molecular motors.1-5 However, the
mechanisms used by multi-isomeric photochromic com-
pounds remain unclear. Among photochromic molecules
involving multi-isomeric systems, benzopyrans (also known
as chromenes) are a promising class of compounds.
Becker and Michl first reported them in 1966.6 UV
irradiation of the colorless chromene proceeds through
C-O bond cleavage, producing a distribution of orange
to red isomeric open forms (merocyanines), which are
thought to be thermally and/or photochemically reverted
to the original benzopyran. In the presence of such
complex systems, multiwavelength absorbance vs time
matrixes are not a sufficient source of mechanistic
information7 because spectral overlapping between sev-
Recently, by use of 19F high-resolution NMR spectro-
scopy, we showed that irradiation of the 2,2-di(4-fluo-
rophenyl)-6-methoxy-2H-1-chromene (FC or fluorochro-
mene) generated, besides the two expected open mero-
cyanines (the transoid-cis (TC) and the transoid-trans
(TT) forms), an o-allenyl-p-methoxyphenol (AP).8 This
compound had never been detected by any other method
before, despite numerous investigations on benzo- and
naphthopyran compounds.9-18 We think that it is worth
investigating the kinetic behavior of this photochromic
benzopyran since the involvement of an allenyl-phenol
side product (AP) has recently been demonstrated to be
(7) Metelitsa, A. V.; Micheau, J . C.; Voloshin, N. A.; Voloshina, E.
N.; Minkin, V. I. J . Phys. Chem. A 2001, 105, 8417-8422.
(8) Delbaere, S.; Micheau, J .-C.; Vermeersch, G. Org. Lett. 2002, 4,
3143-3145.
(9) Becker, R. S.; Dolan, E.; Balke, D. J . Chem. Phys. 1969, 50, 239-
245.
* To whom correspondence may be addressed. Fax: 333 2095 9009.
Tel: 333 2096 4023.
(10) Tyer, N. W.; Becker, R. S. J . Am. Chem. Soc. 1970, 92, 1289-
1294.
(11) Kolc, J .; Becker, R. S. Photochem. Photobiol. 1971, 13, 423-
429.
† Universite de Lille 2.
‡ Universite Paul Sabatier.
(1) Bertelson, R. C. In Photochromism; Brown, G. H., Ed.; Wiley:
New York, 1971; pp 45-ff and references therein.
(2) Guglielmetti, R. In Photochromism: Molecules and Systems,
Studies in Organic Chemistry 40; Du¨rr, H., Bouas-Laurent, H. Eds.;
Elsevier: Amsterdam, 1990; Chapters 8 and 23 and references therein.
(3) Van Gemert, B. In Organic Photochromic and Thermochromic
compounds; Crano, J . C., Guglielmetti, R. J ., Eds.; Plenum Press: New
York, 1999; Vol. 1, pp 111-140.
(12) Lenoble, C.; Becker, R. S. J . Photochem. 1986, 33, 187-197.
(13) Van Gemert, B.; Bergomi, M.; Knowles, D. Mol. Cryst. Liq.
Cryst. 1994, 246, 67-73.
(14) Aldoshin, S.; Chuev, I.; Filipenko, O.; Lokshin, V.; Pozzo, J . L.;
Pepe G.; Samat; A. Acta Crystallogr. C 1996, 52, 1537-1539.
(15) Ottavi, G.; Favaro, G.; Malatesta, V. J . Photochem. Photobiol.
A 1998, 115, 123-128.
(16) Ortica, F.; Levi, D.; Brun, P.; Guglielmetti, R.; Mazzucato, U.;
Favaro, G. J . Photochem. Photobiol. A 2001, 139, 133-141.
(17) J ockusch, S.; Turro, N. J .; Blackburn, F. R. J . Phys. Chem. A
2002, 106, 9236-9241.
(18) Zhao, W.; Carreira, E. M. J . Am. Chem. Soc. 2002, 124, 1582-
1583.
(4) Berkovic, G.; Krongauz, V.; Weiss, V. Chem. Rev. 2000, 100,
1741-1753.
(5) Koumura, N.; Zijlstra, R. W. J .; Van Delden, R. A.; Harada, N.;
Feringa, B. L. Nature 1999, 401, 152-155.
(6) Becker, R. S.; Michl, J . J . Am. Chem. Soc. 1966, 88, 5931-5933.
10.1021/jo035279r CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/10/2003
8968
J . Org. Chem. 2003, 68, 8968-8973

Investigations of a Photochromic Chromene
SCHEME 1. Mech a n istic Mod el Sh ow in g All th e
P ossible In ter con ver sion P a th s betw een th e F ou r
Isom er sa
F IGURE 1. 19F NMR spectra of FC (-115.16) at 228 K. The
times indicated are the cumulated periods of UV irradiation.
The photoisomers produced are TT (-112.00; -112.80), TC
(-112.06; -113.11), and AP (-114.94 ppm).
However, if the isomerizations are photochemically
induced, it is possible that inner filter effects induce
variations in the photokinetic factor, so giving rise to
more complex kinetic rate laws. We have shown that the
use of simultaneous nonlinear least-squares fitting of the
four-species concentration vs time profiles provides un-
ambiguous information about the location and the rela-
tive strengths of the various photochemical and thermal
isomerization pathways.
In this paper, on the basis of Scheme 1, we quantita-
tively analyzed all the possible photochemical and ther-
mal isomerization processes among the four identified
long-lived fluorine-containing species, which are able to
accumulate sufficiently.26
a
vij are the apparent rates. For the sake of clarity, FC ) 1, TC
) 2, AP ) 3, TT ) 4.
general as it has also been reported in nonfluorinated
compounds.19 The kinetics of this fluorochromene were
investigated at low temperature under continuous UV
and visible irradiation, while thermal relaxations in the
dark at various temperatures have also been monitored.
The purpose of these experiments was to obtain very
accurate concentration vs time profiles, from which we
carried out an extended kinetic analysis to identify the
isomeric forms and to determine which isomerization
pathways occurred in a given experimental situation. In
previous works, there are several examples of intercon-
version processes between photochromic isomers,20-23 but
a quantitative analysis of a four-species photochromic
system has, until now, never been published. For this
purpose, we considered that Scheme 1 was the most
general form for a four-isomer system in which every
componentsthe 2,2-di(4-fluorophenyl)-6-methoxy-2H-1-
chromene (FC), the two open merocyanines (TC and TT),
and the APsis assumed to react either thermally or
photochemically to form every other component.24
Such four-component interconverting systems have
been used to describe enzyme folding25 in biochemical
literature. They are modeled by a set of linear differential
equations
Resu lts a n d Discu ssion
UV a n d Visible Ir r a d ia tion a t Low Tem p er a tu r e.
The UV irradiation of a solution of 2,2-di(4-fluorophenyl)-
6-methoxy-2H-1-chromene (FC) in acetonitrile-d3 was
investigated at 228 K. At this temperature, there is no
noticeable thermal isomerization, so the evolution in the
photochromic system (the solution turns red) can only
be attributed to UV-induced photochemical processes.
Figure 1 shows the successive 19F NMR spectra recorded
under UV irradiation. The various photoisomers (pho-
tomerocyanine TT, TC, and AP) increase while the
starting FC decreases.8
In our experimental conditions, fluorine atom mass
balance was spontaneously respected and degradation
was negligible (i.e., less than 2%). The plots of the
photocoloration kinetics (Figure 2) show the simultaneous
evolution of the four isomeric species FC, TC, AP, and
TT. From this plot, homemade software27-29 was used to
perform kinetic analysis on the basis of Scheme 1. The
calculated evolution in concentrations was obtained by
numerical integration of the set of differential equations
4
4
d[Xi]
)
v
-
v
i ) 1-4 and m * i
(1)
∑
∑
mi
im
dt
m)1
m)1
giving rise to multiexponential relaxations.
(19) Delbaere, S.; Vermeersch, G. J . Photochem. Photobiol. A 2003,
159, 227-232.
(25) Bieri, O.; Wildegger, G.; Bachmann, A.; Wagner, C.; Kiefhaber,
T. Biochemistry 1999, 38, 12460-12470.
(20) Hobley, J .; Malatesta, V.; Millini, R.; Giroldini, W.; Wis, L.; Goto,
M.; Kishimoto, M.; Fukumura, H. Chem. Commun. 2000, 1339-1340.
(21) Chuev, I.; Aldoshin, S.; Samat, A.; Maurel, F.; Aubard, J .
THEOCHEM 2001, 548, 123-132.
(26) Aubin, L. B.; Wagner, T. M.; Thoburn, J . D.; Kesler, B. S.;
Hutchinson, K. A.; Schumaker, R. R.; Parakka, J . P. Org. Lett. 2001,
3, 3413-3416.
(22) Go¨rner, H.; Chibisov, A. K. J . Photochem. Photobiol. A 2002,
149, 83-89.
(27) Kaps, K.; Rentrop, P. Comput. Chem. Eng. 1984, 8, 393-396.
(28) Minoux, M. In Programmation Mathe´matique; Dunod, Ed.;
Bordas: Paris, 1983; Vol. 1, pp 95-168.
(29) Deniel, M. H.; Lavabre, D.; Micheau, J . C. In Organic Photo-
chromic and Thermodynamic Compounds; Crano, J . C., Guglielmetti,
R., Eds.; Plenum Press: New York, 1999; Vol. 2, pp 167-209.
(23) Rini, M.; Holm, A. K.; Nibbering, J .; Fidder, H. J . Am. Chem.
Soc. 2003, 125, 3028-3034.
(24) Matsen, F. A.; Franklin, J . L. J . Am. Chem. Soc. 1950, 72,
3337-3341.
J . Org. Chem, Vol. 68, No. 23, 2003 8969

Delbaere et al.
F IGURE 3. 19F NMR spectra of the photochromic mixture at
228 K during visible irradiation. The initial spectrum (t ) 0)
corresponds to the end of UV irradiation. The times indicated
are the cumulated periods of visible light irradiation.
SCHEME 2. P h otoisom er iza tion P r ocesses
Occu r r in g u n d er UV Ir r a d ia tion a t 228 Ka
F IGUR E 2. Photocoloration kinetics under UV irradiation
(signs are experimental concentrations, solid lines are best fits
from our model).
(eq 1). The photochemical rates30 vij are given by
vij ) ΦijꢀilI0F[Xi] ) hij[Xi]
(2)
By minimization of the residual quadratic error ø2
between the calculated and the NMR measured experi-
mental concentrations, only those processes remained
which were kinetically significant, indicating the location
of the photoactive pathways and delivering the corre-
sponding values of hij. We found that the fitting procedure
was very selective among the various possible processes.
The simulated kinetic curves were significantly distant
from their corresponding experimental data if some
important parameter was omitted or if the presence of a
forbidden pathway was artificially imposed (see Support-
ing Information for the details). Results are grouped in
Scheme 2 and Table 1.
Close examination of the results in Table 1 shows that
the major part of the photochemical reactivity of FC
concerns its photoconversion to TC. UV irradiation causes
C-O bond cleavage, leading to a sterically crowded cis-
cisoid isomer through a zwitterionic intermediate32 (non-
monitorable under our conditions) whose further rear-
rangement gives rise to TC. This path requires only a
C3-C4 one-bond rotation. From FC, minor paths to AP
a
The sizes of the arrows give a rough indication of the relative
quantum yield values.
TABLE 1. Ap p a r en t P h otoch em ica l Ra tes a n d Rela tive
Wa velen gth -Aver a ged Qu a n tu m Yield s for P r ocesses
Occu r r in g u n d er UV Ir r a d ia tion 31
relative quantum yields
apparent
photochemical
process
rate constants
from FC
from TC
UV
UV
UV
UV
UV
UV
UV
FC f TC
FC f AP
FC f TT
TC f FC
TC f AP
TC f TT
TT f TC
h
) 1.50 × 10-3
) 2.16 × 10-4
) 1.30 × 10-4
) 6.22 × 10-4
) 1.83 × 10-4
) 5.50 × 10-4
) 3.48 × 10-4
0.81
0.12
0.07
12
13
14
21
23
24
42
h
h
h
h
h
h
0.46
0.13
0.41
and TT are also observed. On the other hand, absorption
of UV by TC induces the reverse reaction, i.e., ring closure
to FC. Also of importance is bond rotation around C4-
C4a, leading to the second merocyanine transoid isomer
TT. There is also a minor path to AP. As ΦAP/ΦTT from
FC ) 1.66 * ΦAP/ΦTT from TC ) 0.33, it is likely that
there is no common intermediate from FC and TC to TT
and AP. As for TT, it was found that its only photochemi-
cal isomerization was the C4-C4a bond rotation to TC.
More interesting is the complete lack of photoreactivity
from the newly discovered AP that is likely to be an
indication of some forbidden photochemical pathways.
Under visible light irradiation, the red solution is
bleached and, as illustrated in Figure 3, TC and TT
gradually disappear while AP increases.
(30) Φij is the quantum yield of the photochemical transformation
of compound “Xi” into compound “Xj”, ꢀi is the molar absorption
coefficient of compound Xi at the irradiation wavelength, l is the optical
path, I0 is the incident photon flux, and F is the photokinetic factor.
Its variations during the UV irradiation period have been neglected.
This assumption is fully justified as F is strictly constant when the
irradiation wavelength is close to an isosbestic point or if the absor-
bance of the photochemical reacting solution is sufficiently low. As the
irradiation light was not strictly monochromatic, Φij and ꢀi values must
be considered as wavelength averaged. Then hij terms correspond to
apparent first-order rate constants.
(31) Relative errors for the stated parameters vij are estimated to
be less than (5% for the most sensitive (or (10% for the corresponding
ratios).
Careful observation of the concentration vs time kinetic
curves (Figure 4) shows that only the two photomero-
(32) Kolc, J .; Becker, R. S. J . Phys. Chem. 1967, 12, 4045-4048.
8970 J . Org. Chem., Vol. 68, No. 23, 2003

Investigations of a Photochromic Chromene
F IGURE 4. Photobleaching kinetics under visible light ir-
radiation (signs are experimental concentrations, solid lines
are best fits from our model).
SCHEME 3. P h otoisom er iza tion P r ocesses
Occu r r in g u n d er Visible Ligh t Ir r a d ia tion a t 228
Ka
F IGURE 5. Time evolution of concentrations of FC, TC, AP,
and TT at two different temperatures. Dots are experimental
data; continuous lines are fitted curves from the Scheme 4
model.
TABLE 2. Ap p a r en t P h otoch em ica l Ra te Con sta n ts a n d
Wa velen gth -Aver a ged Rela tive Qu a n tu m Yield s fr om TC
u n d er Visible Ir r a d ia tion
apparent
photochemical
rate constants
relative
quantum
yields from TC
a
The sizes of the arrows give a visual indication of the relative
values of the corresponding quantum yields.
process
Vis
Vis
Vis
TC f AP
TC f FC
TT f TC
h
h
h
) 1.20 × 10-2
) 1.79 × 10-3
) 7.22 × 10-3
0.87
0.13
23
21
42
cyanines TC and TT were photoreactive but that they
did not follow mono- or multiexponential decays. They
appeared as zero order. This particular kinetic feature
has been accounted for by a strong variation in photo-
kinetic factor throughout visible irradiation.33
Only three photoisomerization paths were detected:
TC f FC, TC f AP, and TT f TC, confirming, as
expected, that FC and AP are not reactive under visible
irradiation due to the lack of visible absorption. (Scheme
3 and Table 2).
The predominant reaction appears to be the photoeno-
lization of TC (TC f AP). This reaction is likely to occur
via a [1,5]-hydrogen shift. In the TC isomer, this rear-
rangement is possible because there is a helical trienic
open chain (H3C3C4C4aC8aO), in which continuous overlap
can be maintained. Such geometrical requirements are
not possible within the TT structure. The absence of
photochemical pathway from TT f AP can then be
justified. The marked differences between TC reactivities
under UV and visible light (ΦFC/ ΦAP
ΦAP ) 0.15 can be accounted for by the presence of
)UV ) 3.40 and (ΦFC/
)
vis
various vibrational levels within the excited state. The
visible photon only makes it possible to reach a less
energetic state from which ring closure to FC is less
favorable. Also less favorable is the photoisomerization
TC f TT, which was not detected under visible irradia-
tion.
Th er m a l Rela xa tion s. When kept in the dark at
higher temperatures, there was slow but complete disap-
pearance of both merocyanines TT and TC and AP as well
as recovery of the initial FC. The recorded kinetics of the
thermal relaxation processes after UV + visible irradia-
tion (Figure 5) show the influence of temperature on the
relative amplitude and curvatures of the concentration
vs time profiles.
(33) In the photokinetic factor F ) (1-10-Abs′)/Abs′, Abs′ is the total
absorbance of the sample at the irradiation wavelength. If the
photoproducts do not absorb at the irradiation wavelength, the
photochemical rate reduces to vij ) Φij′I0(1-10-Abs′) with Abs′ ) (ꢀ′TC
-
[TC] + ꢀ′TT[TT])l. In the first part of the evolution, as TC and TT remain
sufficiently high, the term 10-Abs′ f 0 while f 1 in the final part, hence
the zero-order-like kinetics.
J . Org. Chem, Vol. 68, No. 23, 2003 8971

Delbaere et al.
TABLE 3. Sta n d a r d En tr op y a n d En th a lp y of
SCHEME 4. Th er m a l P r ocesses Occu r r in g d u r in g
Rela xa tion a
Activa tion , ∆Sq a n d ∆Hq, Ca lcu la ted fr om Eyr in g’s
Equ a tion
processes
∆Sq (J mol-1 K-1
)
∆Hq (kJ mol-1
)
TC f TT
TC f FC
TT f TC
AP f TC
27.5 ( 2.1
20.7 ( 0.8
-61.3 ( 8.7
-106.7 ( 6.2
85.6 ( 4.2
82.2 ( 2.2
69.4 ( 5.0
58.3 ( 1.3
itself and ∆Sq from the rearrangement of the solvent.
solv
∆Sq ) ∆Sq + ∆Sq
(3)
a
The sizes of the arrows are a visual indication of the relative
intr
solv
values of the corresponding rate constants.
Moreover, it can be argued that polar acetonitrile is more
oriented near polar open forms than near the less polar
closed forms. Higher positive ∆Sq
are therefore ex-
solv
pected during the open to closed reactions rather than
during open to open rearrangements. This is a possible
explanation for the positive value found for the open to
closed TC f FC process. Among the remaining processes,
the lowest negative activation entropy is for AP f TC
isomerization. This reaction has already been shown in
a previous paper35 to occur via a [1,5]-sigmatropic hy-
drogen shift. According to the Woodward-Hoffmann
rules for sigmatropic rearrangements,36,37 the [1,5] mi-
gration of hydrogen involves a six-membered-ring (i.e.,
highly ordered) transition state in accordance with the
highly negative ∆Sqintr. Because an odd number (3) of
pairs of electrons (two π bonds and a pair of π electrons)
are involved, the ground-state HOMO is symmetrical.
Then, the selection rules require that the [1,5]-hydrogen
shift suprafacial rearrangement occurs under thermal
conditions and is forbidden under photochemical condi-
tions in complete agreement with our conclusions (AP f
TC is thermal but not photochemical).
F IGURE 6. Eyring plots for the four thermal relaxation
processes.
Analysis of the variations of the TC f TT (k24) and TT
f TC (k42) rate constants enabled us to estimate the
relative thermodynamic stability of the two photomero-
cyanine isomers. Although the equilibrium between TT
and TC was never reached in our experimental condi-
tions, the expected equilibrium constant can be calculated
from the kinetic rate constants of the reversible process.
We obtained a value [TT]/[TC]eq ) k24/k42 ) Keq, which
lies between ca. 15 and 50, depending on the tempera-
ture. This result confirms that the TT conformation is
more stable than the TC and that the free enthalpy ∆G°
is around 6.5 kJ mol-1 at 260 K.
These kinetics were also analyzed according to general
Scheme 1 where the rates vij were considered as thermal
terms, being equal to kij[Xi] where kij are the correspond-
ing first-order rate constants. Numerical fitting of the
whole set of kinetics showed that only four paths (among
the possible 12) were needed to account for all the various
temperature kinetic results. AP undergoes a slow ir-
reversible process to TC. TC is the only structure able to
be converted directly to FC by a single bond rotation and
acts as a key intermediate in the pathway from AP to
FC. On the contrary, the TT isomer, needing a double
bond rotation for reversion, does not proceed directly to
FC. The slightly transient TT concentration increase
indicates a slow equilibration process with TC (Scheme
4).
Con clu sion s
For the first time, photocoloration under UV, photo-
bleaching by visible light, and thermal bleaching of a
fluorinated photochromic benzopyran have been inves-
tigated by 19F NMR spectroscopy. UV excitation of ring-
closed FC yields TC and TT merocyanine isomers, as well
as AP. On the basis of a quantitative measurement of
concentrations and subsequent numerical kinetic analy-
sis using a without a priori model, the most plausible
From the Eyring plots ln(kij/T) vs 1/T (Figure 6), the
standard entropy and enthalpy of activation, ∆Sq and
∆Hq, respectively, were obtained (Table 3).
The enthalpy of activation values agree with those
generally obtained for photochromic compounds (from
about 40 to 120 kJ mol-1).1,2,34 The highest ∆Hq values
correspond to processes coming from TC, while the lowest
corresponds to processes going to TC. ∆Sq values are more
difficult to interpret. This parameter can be considered
(35) Delbaere, S.; Micheau, J .-C.; Teral, Y.; Bochu, C.; Campredon,
M.; Vermeersch, G. Photochem. Photobiol. 2001, 74, 694-699.
(36) Woodward, R. B.; Hoffmann, R. J . Am. Chem. Soc. 1965, 87,
395-397.
(37) Woodward, R. B.; Hoffmann, R. J . Am. Chem. Soc. 1965, 87,
2511-2513.
as the sum of two contributions: ∆Sq from the reaction
intr
(34) Hobley, J .; Malatesta, V. Phys. Chem. Chem. Phys. 2000, 2, 57-
59.
8972 J . Org. Chem., Vol. 68, No. 23, 2003

Investigations of a Photochromic Chromene
equate filters (Schott 011FG09, 259 < λ < 388 nm with λmax
)
mechanism was deduced and the values of the relative
quantum yields were extracted. It was found that under
UV, the main photochemical process is the TC f FC
photoequilibrium.
However, both compounds also undergo some minor
rearrangements to TT and AP. AP has been shown not
to be photochemically reactive. Under visible light, only
the two merocyanines TT and TC are photochemically
reactive. TC is converted to both AP and ring-closed
chromene FC. Merocyanine TT is only photoisomerized
to TC.
330 nm for UV irradiation, and Oriel 3-74, λ > 400 nm for
visible irradiation). Temperature was regulated with a home-
built apparatus described elsewhere.34 After irradiation, the
sample was transferred into the thermoregulated probe of a
Bruker Avance-DPX NMR spectrometer (1H, 300 MHz; 19F,
1
282 MHz). H and 19F NMR spectra at 228 K were recorded.
For thermal relaxation studies, only 19F NMR spectra were
recorded at regular intervals at higher selected temperatures
(235, 242.8, 244.8, 248.9, 253.2, 254.3, 259.4, 260.4, 261.7,
264.2 K).
Thermal bleaching relaxation was extensively studied
at various temperatures. The main path is isomerization
of the AP to the ring-closed chromene FC. This path
occurs in two steps, TC playing the role of intermediate.
But, TC has also been found to be involved in a pseu-
doequilibrium with TT. The relative stability of the two
photomerocyanines TC and TT has been determined, TT
conformation being more stable than TC. We think that
these results possess a general character and apply
whatever the detailed structure of the chromene under
investigation. Our conclusions will undoubtedly open the
way to further theoretical modeling of the chromene
photochromic reaction.
Ack n ow led gm en t. The 300-MHz NMR facilities
were funded by the Re´gion Nord-Pas de Calais (France),
the Ministe`re de l’Education Nationale, de l’Enseigne-
ment Supe´rieur et de la Recherche, and the Fonds
Europe´ens de De´veloppement Re´gional. We thank Dr.
M. Campredon (University of Me´diterrane´e) for the gift
of fluoro-[2H]chromene. Part of this collaborative work
was performed within the framework of the “Groupe de
Recherche: Photochromes Organiques, Mole´cules, Me´can-
ismes, Mode`les”, GDR CNRS No. 2466.
Su p p or tin g In for m a tion Ava ila ble: UV spectrum of FC,
1H and 19F NMR spectra, and details of the data analysis and
fitting procedure in PDF format. This material is available free
of charge via the Internet at http://pubs.acs.org.
Exp er im en ta l Section
Samples ([FC]0 ) 10-2 mol L-1 in CD3CN) were irradiated
directly in the NMR tube (5 mm) using a 1000 W Xe-Hg high-
pressure filtered short-arc lamp (Oriel), equipped with ad-
J O035279R
J . Org. Chem, Vol. 68, No. 23, 2003 8973