Regiospecific photoisomerization of fluorinated (E,E)-1,4-diphenyl-1,3- butadienes

Article abstract of DOI:10.1016/j.tetlet.2005.05.103

Photoisomerization of five fluorinated E,E-1-(R-phenyl)-4-phenyl-1,3- butadienes in solution (R = 1: p-monofluoro, 2: m,m′-difluoro, 3: m,m′,p-trifluoro, 4: o,o′,m,m′-tetrafluoro, 5: o,o′,m,m′,p-pentafluoro) was investigated via direct irradiation. Our results indicated that cis-trans photoisomerization of the fluorinated 1,4-diphenyl-1,3-butadienes in the excited singlet state took place exclusively at the CC bonds closer to the fluorine substituents.

Full text of DOI:10.1016/j.tetlet.2005.05.103

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 4953–4956
Regiospecific photoisomerization of fluorinated
(E,E)-1,4-diphenyl-1,3-butadienes
Jin Liu,* Kelly J. Boarman, Natalie L. Wendt and Lina M. Cardenas
Department of Chemistry, Murray State University, Murray, KY 42071, USA
Received 28 February 2005; revised 22 May 2005; accepted 24 May 2005
Available online 9 June 2005
Abstract—Photoisomerization of five fluorinated E,E-1-(R-phenyl)-4-phenyl-1,3-butadienes in solution (R = 1: p-monofluoro, 2:
m,m0-difluoro, 3: m,m0,p-trifluoro, 4: o,o0,m,m0-tetrafluoro, 5: o,o0,m,m0,p-pentafluoro) was investigated via direct irradiation. Our
results indicated that cis–trans photoisomerization of the fluorinated 1,4-diphenyl-1,3-butadienes in the excited singlet state took
place exclusively at the C@C bonds closer to the fluorine substituents.
Ó 2005 Elsevier Ltd. All rights reserved.
Cis–trans photoisomerization has been known as the
trigger for signal generation by natural visual pigments.1
During the past decades, many studies have been
devoted to understanding the complicated natures of
these natural pigments in the excited states, due to their
pivotal importance to visual systems.2 In an attempt
to better understand the regiospecificity of photo-
isomerization of retinal and other natural pigments,
photoisomerization of short polyenes in the excited
singlet state as model systems has been investigated
experimentally as well as theoretically.3–5 Although pho-
tochemical properties of the parent 1,4-diphenyl-1,3-
butadiene have been well elucidated,6 a few studies of
regioselectivity of photoisomerization of 1,4-diphenyl-
1,3-butadienes were reported.7 Recently, the unexpected
regioselectivity, favoring isomerization at one carbon
center of symmetric 1,4-diphenyl-1,3-butadienes via a
Hula–Twist process, was observed in organic glass at
liquid nitrogen temperature.8 However, it is still unclear
whether the regioselectivity of photoisomerization of
unsymmetric 1,4-diphenyl-1,3-butadienes can be dem-
onstrated in solution. Our results from an earlier study
showed the regioselective photoisomerization of cisoid
fluorinated 1,4-diphenyl-1,3-butadienes.9 Hence, it has
occurred to us that in order to reveal the critical factors
that determine the selectivity of photoisomerization of
linear 1,4-diphenyl-1,3-butadienes, systematical study
of fluorine effects on photoisomerization in the excited
singlet state is necessary. In this regard, we investigated
the electron-withdrawing effect on the regioselectivity of
photoisomerization of five fluorinated 1,4-diphenyl-1,3-
butadienes (DPB, 1–5).
1
R1 = p-monofluoro
;
2
R2 = m,m'-difluoro
;
1
3
3
R3 = m,m',p-trifluoro
R4 = o,o',m,m'-tetrafluoro
R5 = o,o',m,m',p-pentafluoro
;
;
2
4
R
4
5
.
To synthesize the (1E,3E) isomers, the substituted benz-
aldehydes were reacted with cinnamylidenetriphenyl-
phosphorane at 0 °C, respectively (Scheme 1), to yield
CHO
-78 oC
Ph3P
+
R
R
+
R
Keywords: Photoisomerization; Excited singlet state; DPB.
*
1Z,3E
1E,3E
Corresponding author. Tel.: +1 2707626626; fax: +1 2707626474;
e-mail: jin.liu@murraystate.edu
Scheme 1.
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.05.103

4954
J. Liu et al. / Tetrahedron Letters 46 (2005) 4953–4956
crude products in 60–70% yield. Each crude mixture was
further separated by silica gel column chromatography
(0–5% ethyl acetate in hexane as an eluant) to afford
the (1E,3E) and (1Z,3E) isomers in a ratio of ꢀ3:2. To
unequivocally distinguish the (1Z,3E) isomers 1–5 from
the corresponding (1E,3Z) isomers that could possibly
be formed from photoisomerization in a minute yield,
we prepared the five (1E,3Z) isomers of 1–5. Synthesis
of the (1E,3Z) isomers was accomplished in two steps.
First, the substituted E-cinnamaldehydes were prepared
by C2-extension of the substituted benzaldehydes with
(triphenylphosphoranylidene)acetaldehyde, respectively.
Then, the fluorinated E-cinnamaldehydes were coupled
with benzyltriphenylphosphonium bromide using
sodium bis(trimethylsilyl)amide as a base at ꢁ78 °C, res-
pectively, to afford crude products containing the major
(1E,3Z) isomers plus the (1E,3E) isomers in an overall
yield of 85–90% (Scheme 2). Two E/Z isomers were fur-
ther separated by using the same chromatographic con-
ditions. All the synthetic DPBs were characterized by
spectroscopic methods (see Supplementary data). In
Figure 1, UV absorption band of the (1E,3E) isomer
(5) is shown to have a fine structure, which is character-
istic of UV absorption bands of all the (1E,3E) iso-
mers.15 Also, the cis bands have been noticed in the
UV–vis absorption spectra of all the (1E,3Z) and
(1Z,3E) isomers (1–5) (see Supplementary data).
200 W Hg medium-pressure lamp with a Pyrex filter
(k > 300 nm). At the beginning of 14 min, the progress
of the reaction was monitored by H NMR and HPLC
1
every 2 min, and then the photoreaction was checked
every 15 min until it reached a photostationary state.
During the irradiation course, it was found that the
(1E,3E)-isomer was isomerized to yield the (1Z,3E) iso-
mer (Fig. 2). On the other hand, the proton absorption
(dd) at d 7.64, which was assigned to H-2 of the (1E,3Z)
isomer, was undetected indicating that the (1E,3Z) iso-
mer was not formed from photoisomerization (Figs. 3
and 4). After irradiation, the remaining (1E,3E)-isomer
in the mixture was isolated by silica gel column chroma-
Direct irradiation of the (1E,3E)-isomer (5) in
chloroform-d (0.015–0.002 M) was carried out using a
CHO
-78 oC
Ph P
3
+
R
R
+
R
1E,3Z
1E,3E
Scheme 2.
1
Figure 2. 1H NMR spectra (from bottom to top) were recorded at
irradiation times of 15, 30, 60, 90, and 120 min, respectively. No
change was observed during the first 14 min.
0.8
0.6
0.4
0.2
0
210
260
310
360
410
Wavelength (nm)
Figure 1. UV Absorption maxima of the (1E,3E), (1Z,3E), and
(1E,3Z) isomers (5) in hexanes are 326, 307, and 312 nm, respectively.
The concentrations are varied.
Figure 3. 1H NMR spectrum of the (1Z,3E) isomer 5.

J. Liu et al. / Tetrahedron Letters 46 (2005) 4953–4956
4955
Table 1. The percentages (%) of the E/Z isomers at photostationary
state
Compound
1E,3E
1Z,3E
1E,3Z
1
2
3
4
5
72
48
39
16
9
28
52
61
84
91
0
0
0
0
0
DPB.7 Moreover, the exclusive formation of the
(1Z,3E) isomers 1–5 from photoisomerization of the
fluorinated (1E,3E) isomers was found in other solvents
such as hexane and acetonitrile.
Figure 4. 1H NMR spectrum of the (1E,3Z) isomer 5.
The above photo-studies show that photoisomerization
of the fluorinated (1E,3E) isomers 1–5 is a regiospecific
reaction, and the E/Z photoisomerization conversion
rate is strongly dependent on the number of fluorine
substituents on the phenyl rings. As indicated by our
computational calculation (DFT-B3LYP/6-31G**), the
ground-state energy difference between two E,Z isomers
of the fluorinated DPBs 1–5 is less than 1 kcal/mol.
Because of the small difference, the observed regioselec-
tivity must be influenced by the photochemical proper-
ties of 1–5 in the excited singlet state, which are clearly
sensitive to the electron-withdrawing fluorine substitu-
ents. In 1970, the geometrical relaxation in the excited
state of a 1,3-diene caused by a highly polar twisted
allylmethylene species (zwitterion) was postulated by
Dauben to elucidate the stereoselective cyclization of
ethylidenecyclooctene.10 Since then, the postulated zwit-
terionic excited states as intermediates have been used to
account for the regioselectivity of photoisomerization of
some unsymmetrical short polyenes.11–14 Meanwhile,
other explanations have been proposed.15–17 Due to
the polar nature of fluorine substituents, we examined
all the possible zwitterionic intermediates. Although
the excited zwitterion is short-lived, it could play a cru-
cial role in determining the regioselectivity of photoiso-
merization.18,19 Of the four possible zwitterionic
intermediates (Fig. 6), it is clear that structures 6a and
6c could lead to selective isomerization at the 1,2-bond.
However, due to the nature of the substituents in these
intermediates, the participation of zwitterionic structure
6c is most unlikely. Thus, the most preferred zwitter-
ionic intermediate is 6a, especially while the number of
fluorine substituents increases. The preferentially
tography. The photo-conversion yield (91%) of the
(1E,3E) isomer (5) to the (1Z,3E) isomer at the photo-
stationary state was determined. The result from the di-
rect irradiation of 5 indicated that a dominant cis–trans
photoisomerization took place at the carbon–carbon
double bond closer to the fluorine substituents.
In Figure 5, the similar UV absorption bands of the
three isomers (1) are shown. Upon direct irradiation
with a Pyrex filter (k > 300 nm), the (1E,3E)-isomer (1)
in chloroform-d (0.015–0.002 M) was also found to
photoisomerize at the C@C double bond near the fluo-
rine atom. After purification, the (1Z,3E)-isomer (1) as
the only photoproduct was further confirmed by using
spectroscopic methods (see Supplementary data). Fur-
thermore, the cis–trans photoisomerization of the
(1E,3E) isomers 2–4 was carried out in a manner ana-
logous to 1 and 5. The data suggested the preferred iso-
merization at the C@C double bonds closer to fluorine
substituents. Also, the photoconversion yield increased
as the number of fluorine substituents on the phenyl
rings increased (Table 1). At photostationary state, the
ratios of (1Z,3E)/(1E,3E) of 1, 2, 3, 4, and 5 were
1:2.6, 1.1:1, 1.6:1, 5.2:1, and 10.1:1, respectively. The re-
sults are in accordance with the similar regioselectivity
found in photoisomerization of a cyano-substituted
1
0.8
0.6
0.4
0.2
0
+
-
-
+
R
R
6b
6a
-
+
210
260
310
360
410
Wavelength (nm)
+
-
R
R
Figure 5. UV absorption maxima of the (1E,3E), (1Z,3E), and (1E,3Z)
isomers (1) in hexanes are 326, 310, and 307 nm, respectively. The
concentrations are varied.
6c
6d
Figure 6. Zwitterionic intermediates.

4956
J. Liu et al. / Tetrahedron Letters 46 (2005) 4953–4956
formed zwitterionic intermediates, which appear to be
controlled by fluorine substituents, ultimately deter-
mined the specific position of photoisomerization. We
believe that the current system involving the different
number of polar fluorine substituents further strength-
ens the fact that the polar character of the excited state
has a significant role in directing the position of photo-
isomerization. In addition, the same regioselectivity of
photoisomerization of 1–5 found in different solvents
states crucial substituent effect as compared to solvent
effect.
References and notes
1. Wald, G. Science 1968, 162, 230–239.
2. Mathies, R. A.; Lugtenburg, J. In Handbook of Biological
Physics; Stavenga, D. G., De Grip, W. J., Pugh, E. N.,
Eds.; Elsevier Science: New York, 2000; Vol. 3.
3. Waldeck, D. H. Chem. Rev. 1991, 91, 415–436.
4. Allen, M. T.; Whitten, D. G. Chem. Rev. 1989, 89, 1691–
1702.
5. Papper, V.; Pines, D.; Likhtenshtein, G.; Pines, E. J.
Photochem. Photobiol. A 1997, 111, 87–96.
6. Yee, W. A.; Hug, S. J.; Kliger, D. S. J. Am. Chem. Soc.
1988, 110, 2164–2169.
7. Singh, A. K.; Mahalaxmi, G. R. Photochem. Photobiol.
2000, 71, 387–396.
8. Yang, L. Y.; Liu, R. S. H.; Boarman, K. J.; Wendt, N. L.;
Liu, J. J. Am. Chem. Soc. 2005, 127, 2404–2405.
9. Liu, J.; Suits, E. L.; Boarman, K. J. Tetrahedron Lett.
2003, 44, 8103–8107.
In summary, we have demonstrated that the selective
formation of zwitterionic intermediates, which are sensi-
tive to the substituent effects, controlled the regiospecific
photoisomerization of the fluorinated DPBs in the ex-
cited singlet state. As mentioned at the beginning of this
letter, photoisomerization of all-trans retinal protonated
Schiff base (RPSB) in the protein environment (rhodop-
sin) is regiospecific. In contrast, direct irradiation of all-
trans RPSB in solution yields several photoproducts
including the main 11-cis isomer. Our results from this
study provide useful information for understanding the
regioselectivity of photoisomerization of chromophores
in solution, which can be potentially applied to other
media.20,21
10. Dauben, W. G.; Ritscher, J. S. J. Am. Chem. Soc. 1970, 92,
2925–2926.
ˇ ´
´
11. Bonacic-Koutecky, V.; Bruckmann, P.; Hiberty, P.; Kou-
tecky´, J.; Leforestier, C.; Salem, L. Angew. Chem., Int. Ed.
Engl. 1975, 14, 575–576.
12. Muthuramu, K.; Liu, R. S. H. J. Am. Chem. Soc. 1987,
109, 6510–6511.
13. Squillacote, M.; Semple, T. J. Am. Chem. Soc. 1987, 109,
892–894.
14. Schilling, C. L.; Hilinski, E. F. J. Am. Chem. Soc. 1988,
110, 2296–2298.
15. Squillacote, M.; Wang, J.; Chen, J. J. Am. Chem. Soc.
2004, 126, 1940–1941.
Acknowledgments
16. Olivucci, M.; Bernardi, F.; Ottani, S.; Robb, M. A. J. Am.
Chem. Soc. 1994, 116, 2034–2048.
17. Olivucci, M.; Ragazos, I. N.; Bernardi, F.; Robb, M. A.
J. Am. Chem. Soc. 1993, 115, 3710–3721.
18. Baretz, B. H.; Singh, A. K.; Liu, R. S. H. Nouv. J. Chim.
1981, 5, 297–303.
19. Satiel, J.; Ko, D.-H.; Fleming, S. A. J. Am. Chem. Soc.
1994, 116, 4099–4100.
20. Uda, M.; Mizutan, T.; Hayakawa, J.; Momotake, A.;
Ikegam, M.; Nagahata, R.; Arai, T. Photochem. Photobio.
2002, 76, 596–605.
J.L. thanks Kentucky NSF EPSCoR and Howard
Hughes Medical Institution for financial support.
Supplementary data
Spectroscopic data (1H NMR, 13C NMR, and UV) of
new compounds included in the Supplementary data
associated with this article can be found, in the online
version, at doi:10.1016/j.tetlet.2005.05.103.
21. Liu, J.; Boarman, K. J. Chem. Commun. 2005, 340–341.