Synthesis and photochromism of 1,2-bis(5-aryl-2-phenylethynylthien-3-yl)hexafluorocyclopentene derivatives

Article abstract of DOI:10.1016/j.jphotochem.2010.04.013

Photochromic dithienylethenes that possess elements of lipid complementarity, and undergo large, photoinduced changes in molecular geometry have been prepared. Further, a regioselective approach has been developed for the preparation of dithienylethenes containing phenylethynyl and various aryl substituents at C2 and C5 of the thiophene moieties, respectively. The prepared photochromic compounds were observed to undergo reversible photoisomerization. Their absorption properties, reaction quantum yields, and photoconversions were determined in n-hexane.

Full text of DOI:10.1016/j.jphotochem.2010.04.013

Journal of Photochemistry and Photobiology A: Chemistry 212 (2010) 176–182
Contents lists available at ScienceDirect
Journal of Photochemistry and Photobiology A:
Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
Synthesis and photochromism of
1,2-bis(5-aryl-2-phenylethynylthien-3-yl)hexafluorocyclopentene derivatives
Jianxin Cai, Amani Farhat, Pavel B. Tsitovitch, Vivek Bodani, R. Daniel Toogood, R. Scott Murphy∗
Department of Chemistry and Biochemistry, University of Regina, 3737 Wascana Parkway, Regina, SK S4S 0A2, Canada
a r t i c l e i n f o
a b s t r a c t
Article history:
Photochromic dithienylethenes that possess elements of lipid complementarity, and undergo large, pho-
toinduced changes in molecular geometry have been prepared. Further, a regioselective approach has
been developed for the preparation of dithienylethenes containing phenylethynyl and various aryl sub-
stituents at C2 and C5 of the thiophene moieties, respectively. The prepared photochromic compounds
were observed to undergo reversible photoisomerization. Their absorption properties, reaction quantum
yields, and photoconversions were determined in n-hexane.
Received 23 November 2009
Received in revised form 25 March 2010
Accepted 19 April 2010
Available online 24 April 2010
Keywords:
© 2010 Elsevier B.V. All rights reserved.
Dithienylethene
Photochromism
Photoisomerization
Quantum yield
1. Introduction
egy for dithienylethene derivatives of this general structure has
not been well developed. Further, the synthetic protocol for 1a is
Photochromism is defined as a light-induced reversible trans-
formation between two isomers having different absorption
spectra [1]. Most photochromic compounds, such as azobenzenes
and spirooxazines, undergo a thermally reversible isomerization
as well. On the contrary, the majority of dithienylethenes are
thermally irreversible and fatigue resistant [2]. As a result, these
photoresponsive molecules have promising applications in pho-
tonic devices [3,4].
The photomodulation of liquid crystal properties has recently
received considerable attention [5–10]. Notably, the photoisomer-
ization of a dithienylethene dopant has been used to reversibly
modulate the spontaneous polarization of ferroelectric smectic
liquid crystals [11]. This dithienylethene derivative contained
methyl substituents at the reactive carbons (i.e., C2) and 4-
heptyloxyphenyl substituents at C5 of the thiophene moieties
to improve its complementarity with the lamellar structure of
the liquid crystals. In this study, we have designed a similar
dithienylethene analog, 3a (Scheme 1), with aspects of lipid com-
plementarity by incorporating a 4-dodecylphenyl substituents at
C5. However, 3a contains bulky phenylethynyl substituents at C2
in place of the methyl substituents. As a result, the photoisomeriza-
tion of 3a will lead to a much larger change in molecular geometry
as has been shown for 1a [12]. Although the preparation of alkynyl
derivatives such as 1a have been reported [12–14], a synthetic strat-
limited to the derivatization of 2,3-dibromo-5-phenylthiophene,
which does not allow for the introduction of additional func-
tionality at C5 of the thiophene ring. Consequently, we present
a regioselective approach for the preparation of photochromic
dithienylethenes, like 3a, that possess elements of lipid comple-
mentarity, and undergo large, photoinduced changes in molecular
geometry. Regioselective modification of thiophene heterocycles at
three positions provides a versatile methodology for the incorpo-
ration of functional groups at C2 and C5 of the thiophene moieties
of dithienylethenes. In addition, we have examined the absorp-
tion properties and photochromic reactivity of these molecules
in solution. Compounds 1a–3a may be used in the development
of photoresponsive liposomes with reversible photocontrols over
membrane stability, and 3a may have interesting implications in
the development of photonic devices based on liquid crystals.
2. Experimental
2.1. Instrumentation
1H, 19F, and 13C NMR spectra were recorded at 300.18, 282.46,
and 75.48 MHz, respectively, on a Varian Mercury plus spec-
trometer (Palo Alto, CA, USA). Chemical shifts are referenced to
solvent signals. High-resolution mass spectral analyses were car-
ried out by the University of Saskatchewan on a VG 70SE mass
spectrometer (Manchester, UK), which was operated in electron
impact or electrospray ionization modes. X-ray data collection
were carried out by the University of Alberta on a Bruker PLAT-
∗
Corresponding author. Tel.: +1 306 585 4247; fax: +1 306 337 2409.
E-mail address: scott.murphy@uregina.ca (R.S. Murphy).
1010-6030/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochem.2010.04.013

J. Cai et al. / Journal of Photochemistry and Photobiology A: Chemistry 212 (2010) 176–182
177
Scheme 1. The reversible photoisomerization of dithienylethenes, where 1a–3a and 1b–3b are the open-ring and closed-ring isomers, respectively.
FORM diffractometer/SMART 1000 CCD (Madison, WI, USA) using
graphite-monochromated Mo K␣ radiation at −80 ◦C. Elemental
analyses were carried out by Guelph Chemical Laboratories on
a Fisons/Carlo Erba CHNS-O EA 1108 elemental analyzer (Lake-
wood, NJ, USA) or by the University of Calgary on a PerkinElmer
2400 Series II CHNS/O elemental analyzer (Waltham, MA, USA).
HPLC analyses were performed on a HP 1090 Series II liquid chro-
matograph (Santa Clara, CA, USA). A guard column containing
SecurityGuard silica cartridges (4 mm × 3 mm i.d.) was connected
to a Luna 5 ␮m Silica(2) analytical column (250 mm × 4.6 mm i.d.)
from Phenomenex (Torrance, CA, USA). The mobile phase was
95:5 n-hexane/ethyl acetate for 1 and 2, and n-hexane for 3 at
a flow rate of 1 mL min−1. The injection volume was 5 ␮L and
signals were monitored at 254 nm. Steady-state absorption spec-
tra were obtained at constant temperature (e.g., 21.0 0.1 ◦C)
on a Cary 300 Bio UV–vis spectrophotometer (Mississauga, ON,
Canada) equipped with a dual cell Peltier circulator accessory.
The absorption spectra were recorded at a scan rate, step size,
and integration time of 300 nm min−1, 0.5 nm, and 0.1 s, respec-
tively. All samples were irradiated with a 300 W xenon light source
from Luzchem (Ottawa, ON, Canada), measured in a 10 mm × 4 mm
quartz Suprasil absorbance cell from Hellma (Concord, ON, Canada),
and stirred. For ultraviolet irradiations, the xenon light source
was filtered with a bandpass filter (Hoya U-340, ꢀc = 340 42 nm)
and an aqueous solution of potassium chromate (ꢀc = 313 7 nm
[15]) circulated through a 50 mm × 22 mm cylindrical cell from
Hellma (A313 = 0.040 0.005). For visible irradiations, a bandpass
filter (Throlabs FB530-10, ꢀc = 530 5 nm) was used.
and 48% aqueous solution of hydrogen bromide (35 mL) was added
dropwise to mixture of thiophene (9.57 g, 0.114 mol), 48% aque-
ous solution of hydrogen bromide (50 mL), and diethyl ether
(35 mL) at 0 ◦C. The reaction mixture was allowed to warm to
room temperature and refluxed for 2 h. The biphasic mixture was
extracted with dichloromethane (3× 100 mL), dried over anhy-
drous magnesium sulfate, filtered, and concentrated under reduced
pressure. The crude product was passed through a silica gel column
(dichloromethane). Distillation of the liquid residue at 118–120 ◦C
(ca. 5 Torr) gave pure 4 as a colorless solid (20 g, 57%). 1H NMR
(CDCl3, ı): 6.90 (s, 1H, H4).
2.3.2. 2,4-Dibromothiophene (5)
Adapted from previously published procedures [18,19], n-
butyllithium (25.0 mL of a 2.50 M solution in hexanes, 62.5 mmol)
was added dropwise to a solution of 4 (20.0 g, 62.5 mmol) in diethyl
ether (200 mL) at −78 ◦C and under an atmosphere of argon as
shown (Scheme 2). After stirring for 10 min, the reaction mixture
was poured into cold water (250 mL) to give a biphasic mixture. This
mixture was extracted with diethyl ether (3× 60 mL), washed with
water (100 mL), dried over anhydrous magnesium sulfate, filtered,
and concentrated under reduced pressure. The crude product was
passed through a silica gel column (1:12 chloroform/hexanes). Dis-
tillation of the liquid residue at 78–80 ◦C (ca. 5 Torr) gave pure 5 as
a colorless liquid (7.9 g, 52%). 1H NMR (CDCl3, ı): 6.98 (d, J = 1.8 Hz,
1H, H3), 7.15 (d, J = 1.8 Hz, 1H, H5).
2.3.3. 4-Dodecylphenylboronic acid (6c)
The starting material, 1-bromo-4-dodecylbenzene, was syn-
thesized following reported procedures [20,21]. Adapted from a
previously published procedure [22], n-butyllithium (1.50 mL of
a 2.50 M solution in hexanes, 3.75 mmol) was added dropwise
to a solution of 1-bromo-4-dodecylbenzene (0.0400 g, 1.23 mmol)
in tetrahydrofuran (10 mL) at −78 ◦C and under an atmosphere
of argon. After stirring for 75 min, trimethoxyborane (0.418 mL,
3.75 mmol) was added dropwise to the mixture. The reaction
mixture was allowed to warm to room temperature, stirred for
16.5 h, and poured into 2 M HCl (50 mL) to give a biphasic mix-
ture. This mixture was washed with water (3× 30 mL), extracted
with dichloromethane (4× 50 mL), dried over anhydrous magne-
sium sulfate, filtered, and concentrated under reduced pressure to
give 6c as a pale yellow oil which was used for the next step without
further purification.
2.2. Materials
All reactants (99+%, Sigma–Aldrich, Oakville, ON, Canada),
deuterated solvents (99.9 at.% D, Sigma–Aldrich), trans-
dichlorobis(triphenylphosphine)palladium(II)
([PdCl2(PPh3)2])
(99.9+% Pd, Strem Chemicals, Newburyport, MA, USA), isopropyli-
denesuccinic acid diethyl ester (TCI America, Portland, OR, USA),
octafluorocyclopentene (99+%, SynQuest Laboratories, Alachua, FL,
USA), and tetrakis(triphenylphosphine)palladium(0) ([Pd(PPh3)4])
(99.9+% Pd Strem Chemicals) were purchased and used as received.
Carbon tetrachloride, copper iodide, diethyl ether, 1,4-dioxane,
tetrahydrofuran, and triethylamine were purified following
literature procedures [16]. Flash column chromatography was
performed on silica gel (200–400 mesh, 60 Å, Sigma–Aldrich).
2.3. Procedures for the synthesis of photochromic
2.3.4. 4-Bromo-2-phenylthiophene (7a)
compounds 1a–3a
Adapted from a previously published procedure [23], 5 (0.302 g,
1.25 mmol) and sodium carbonate (2.5 mL of a 2.0 M aqueous solu-
tion) were added dropwise to a stirred mixture of phenylboronic
acid (0.165 g, 1.35 mmol), Pd(PPh3)4 (0.136 g, 0.118 mmol), and
dioxane (15 mL) under an atmosphere of argon. After stirring at
2.3.1. 2,3,5-Tribromothiophene (4)
The bromination of thiophene was adapted from a previously
published procedure [17]. A mixture of bromine (61.5 g, 0.385 mol)

178
J. Cai et al. / Journal of Photochemistry and Photobiology A: Chemistry 212 (2010) 176–182
100 ◦C for 4.5 h, the reaction mixture was allowed to cool to room
temperature, diluted with diethyl ether (80 mL), poured into water
(80 mL) to give a biphasic mixture, and extracted with diethyl ether
(3× 80 mL). The combined organic extracts were washed with a sat-
urated aqueous solution of sodium chloride (3× 30 mL), dried over
anhydrous magnesium sulfate, filtered, and concentrated under
reduced pressure. Purification of the crude product by column chro-
matography (hexanes) gave pure 7a as a colorless solid (0.2 g, 72%).
1H NMR (CDCl3, ı): 7.18 (d, J = 1.5 Hz, 1H, H3), 7.22 (d, J = 1.5 Hz,
1H, H5), 7.30–7.45 (m, 3H, Ar H), 7.54–7.61 (m, 2H, Ar H). 13C NMR
(CDCl3, ı): 110.8, 122.2, 125.9, 126.0, 128.5, 129.3, 133.4, 145.7.
HRMS-EI (m/z): [M]+ calcd for C10H7SBr, 237.9452; found 237.9449.
mp 81 ◦C). 1H NMR (CDCl3, ı): 7.10 (s, 1H, H4), 7.30–7.43 (m, 3H,
Ar H), 7.45–7.51 (m, 2H, Ar H). 13C NMR (CDCl3, ı): 110.1, 114.6,
125.5, 125.6, 128.6, 129.1, 132.7, 145.4. HRMS-EI (m/z): [M]+ calcd
for C10H6SBr2, 315.8557; found 315.8556.
2.3.8. 5-Biphenyl-2,3-dibromothiophene (8b)
Prepared by a method similar to that used for 8a, 8b was
obtained as a colorless solid (76%). 1H NMR (CDCl3, ı): 7.14 (s, 1H,
H4), 7.33–7.41 (m, 1H, Ar H), 7.42–7.50 (m, 2H, Ar H), 7.52–7.66
(m, 6H, Ar H). 13C NMR (CDCl3, ı): 110.3, 114.8, 125.7, 126.0, 127.1,
127.9, 127.9, 129.1, 131.8, 140.3, 141.6, 145.2. HRMS-ESI (m/z): [M]+
calcd for C16H10SBr2, 393.8849; found 393.8854. Anal. Calcd for
C
16H10SBr2: C, 48.76; H, 2.56. Found: C, 48.79; H, 2.84. All values
2.3.5. 2-Biphenyl-4-bromothiophene (7b)
are given as percentages.
Prepared by a method similar to that used for 7a, 7b was
obtained as a colorless solid (79%). 1H NMR (CDCl3, ı): 7.19 (d,
J = 1.5 Hz, 1H, H3), 7.25 (d, J = 1.5 Hz, 1H, H5), 7.33–7.41 (m, 1H, Ar H),
7.42–7.50 (m, 2H, Ar H), 7.58–7.66 (m, 6H, Ar H). 13C NMR (CDCl3, ı):
110.6, 121.9, 125.6, 126.1, 126.9, 127.6, 127.6, 128.9, 132.1, 140.2,
141.0, 145.0. HRMS-ESI (m/z): [M]+ calcd for C16H11SBr, 315.9744;
found 315.9750. Anal. Calcd for C16H11SBr: C, 60.96; H, 3.52. Found:
C, 60.25; H, 3.84. All values are given as percentages.
2.3.9. 2,3-Dibromo-5-(4-dodecylphenyl)thiophene (8c)
Prepared by a method similar to that used for 8a, 8c was
obtained as a colorless solid (71%). 1H NMR (CDCl3, ı): 0.88 (m,
3H, CH3), 1.25 (m br, 18H, CH2), 1.61 (m, 2H, CH2), 2.60 (m, 2H,
CH2), 7.05 (s, 1H, H3), 7.15–7.22 (m, 2H, Ar H), 7.36–7.42 (m, 2H, Ar
H). 13C NMR (CDCl3, ı): 14.1, 22.7, 29.2–29.7, 31.3, 31.9, 35.7, 109.4,
114.4, 125.0, 125.4, 129.1, 130.2, 143.8, 145.6. HRMS-EI (m/z): [M]+
calcd for C22H30SBr2, 484.0435; found 484.0426. Anal. Calcd for
C22H30SBr2: C, 54.33; H, 6.22. Found: C, 55.12; H, 6.21. All values
are given as percentages.
2.3.6. 4-Bromo-2-(4-dodecylphenyl)thiophene (7c)
Prepared by a method similar to that used for 7a, 7c was
obtained as a colorless solid (60%). 1H NMR (CDCl3, ı): 0.89 (m,
3H, CH3), 1.27 (m br, 18H, CH2), 1.62 (m br, 2H, CH2), 2.62 (m, 2H,
CH2), 7.13 (d, J = 1.5 Hz, 1H, H3), 7.15 (d, J = 1.5 Hz, 1H, H5), 7.17–7.23
(m, 2H, Ar H), 7.42–7.50 (m, 2H, Ar H). 13C NMR (CDCl3, ı): 14.3,
22.9, 29.5, 29.6, 29.7, 29.8, 29.9, 31.6, 32.2, 35.9, 110.6, 121.6, 125.4,
125.9, 129.3, 130.9, 143.6, 145.6. HRMS-EI (m/z): [M]+ calcd for
2.3.10. 3-Bromo-5-phenyl-2-phenylethynylthiophene (9a)
Adapted from a previously published procedure [23], ethynyl-
benzene (0.067 g, 0.66 mmol) in triethylamine (1 mL), PdCl2(PPh3)2
(2.3 mg, 3.3 ␮mol), and triphenylphosphine (1.1 mg, 4.2 ␮mol)
C22H31SBr, 406.1330; found 406.1332. Anal. Calcd for C22H31SBr:
C, 64.85; H, 7.67. Found: C, 65.41; H, 7.71. All values are given as
percentages.
were added dropwise to
a stirred mixture of 8a (0.207 g,
0.651 mmol) in triethylamine (1 mL) under an atmosphere of argon.
After stirring at 35 ◦C for 15 min, copper iodide (2.0 mg, 10 ␮mol)
was added to the reaction mixture and was stirred at 60 ◦C for
3.5 h. After allowing the reaction mixture to cool to room tem-
perature, it was diluted with ethyl acetate (25 mL) and poured
into water (25 mL) to give a biphasic mixture. This mixture was
extracted with ethyl acetate (3× 25 mL), dried over anhydrous mag-
nesium sulfate, filtered, and concentrated under reduced pressure.
Purification of the crude product by column chromatography (1:9
chloroform/hexanes) gave pure 9a as a colorless solid (0.14 g, 63%,
mp 110.0–110.6 ◦C). 1H NMR (CDCl3, ı): 7.21 (s, 1H, H4), 7.31–7.45
(m, 6H, Ar H), 7.52–7.62 (m, 4H, Ar H). 13C NMR (CDCl3, ı): 81.6, 98.0,
117.0, 120.2, 122.9, 125.9, 126.0, 128.7, 128.9, 129.0, 129.4, 131.8,
133.0, 145.4. HRMS-EI (m/z): [M]+ calcd for C18H11SBr, 337.9765;
found 337.9755.
2.3.7. 2,3-Dibromo-5-phenylthiophene (8a)
Adapted from
a previously published procedure [24], 7a
(1.34 g, 5.60 mmol) and 70% aqueous solution of perchloric acid
(100 ␮L, 1.16 mmol) were added to a stirred mixture of N-
bromosuccinimide (1.27 g, 7.14 mmol) and carbon tetrachloride
(8 mL) under an atmosphere of argon. After stirring at room tem-
perature for 36 h, the reaction mixture diluted with chloroform
(70 mL) and sodium carbonate (30 mL of a 2.0 M aqueous solu-
tion) to give a biphasic mixture. This mixture was extracted with
chloroform (3× 70 mL), washed with water (2× 50 mL), dried over
anhydrous magnesium sulfate, filtered, and concentrated under
reduced pressure. Purification of the crude product by column chro-
matography (pentane) gave pure 8a as a colorless solid (1.5 g, 83%,
Scheme 2. Synthesis of dithienylethenes 1a–3a. Reagents and conditions: (i) n-BuLi, Et2O, Ar, −78 ◦C, 10 min, H2O; (ii) [Pd0(PPh3)4], 6:1 dioxane/2 M Na2CO3, Ar, 100 ◦C,
4.5 h; (iii) NBS, 10 mol% HClO4 (70% in H2O), CCl4, rt, 36 h; (iv) ethynylbenzene, [PdII(PPh3)2]Cl2, PPh3, Et3N, Ar, 35 ◦C, 15 min; (v) CuI, Ar, 60 ◦C, 3.5 h; (vi) n-BuLi, THF, Ar,
−78 ◦C, 30 min; (vii) octafluorocyclopentene, Ar, −78 ◦C, 4 h. Assignment of C␣, C␣ꢀ, C␤, and C␤ꢀ is shown for 4, and used consistently for other thiophene derivatives.

J. Cai et al. / Journal of Photochemistry and Photobiology A: Chemistry 212 (2010) 176–182
179
H). 13C NMR (CDCl3, ı): 81.2, 100.3, 122.5, 123.0, 124.0, 126.4,
127.1, 127.6, 127.9, 128.6, 129.1, 131.6, 131.9, 132.8, 140.5, 141.3,
145.5. 19F NMR (CD3COCD3, ı): −110.5 (t, J = 5.2 Hz, 4F), −132.7 (m,
J = 5.2 Hz, 2F). HRMS-ESI (m/z): [M]+ calcd for C53H30S2F6, 844.1693;
found 844.1729. Anal. Calcd for C53H30S2F6: C, 75.34; H, 3.58.
Found: C, 75.50; H, 3.28. All values are given as percentages. UV–vis
(n-hexane) ꢀmax, nm (log ε): 326 (4.90), 591 (4.43).
2.3.11. 5-Biphenyl-3-bromo-2-phenylethynylthiophene (9b)
Prepared by a method similar to that used for 9a, 9b was
obtained as a colorless solid (79%). 1H NMR (CDCl3, ı): 7.25 (s, 1H,
H4), 7.33–7.41 (m, 4H, Ar H), 7.43–7.50 (m, 2H, Ar H), 7.55–7.66 (m,
8H, Ar H). 13C NMR (CDCl3, ı): 81.6, 98.0, 117.0, 120.0, 122.8, 126.0,
126.3, 127.2, 127.9, 128.0, 128.7, 129.0, 129.1, 131.8, 131.9, 140.4,
141.8, 145.0. HRMS-ESI (m/z): [M]+ calcd for C24H15SBr, 416.0057;
found, 416.0069. Anal. Calcd for C24H15SBr: C, 69.40; H, 3.64. Found:
C, 68.73; H, 3.75.
2.3.15. 1,2-Bis(5-(4-dodecylphenyl)-2-phenylethynylthien-3-yl)
hexafluorocyclopentene (3a)
2.3.12. 3-Bromo-5-(4-dodecylphenyl)-2-
Prepared by a method similar to that used for 1a, 3a was
obtained as a yellow solid (51%). 1H NMR (CD3COCD3, ı): 0.88 (m,
6H, CH3), 1.29 (m br, 36H, CH2), 1.64 (m br, 4H, CH2), 2.63 (m, 4H,
CH2), 7.13–7.24 (m, 8H, Ar H), 7.25–7.35 (m, 10H, Ar H), 7.44 (s,
2H, H4). 13C NMR (CD3COCD3, ı): 13.7, 22.7, 29.2, 29.6, 29.7–29.8,
31.5, 32.0, 35.5, 80.6, 100.3, 122.0, 122.4, 123.4, 125.9, 128.8, 129.2,
129.3, 130.0, 131.5, 132.3, 144.1, 146.7. 19F NMR (CD3COCD3, ı):
−110.4 (t, J = 5.2 Hz, 4F), −132.7 (m, J = 5.2 Hz, 2F). HRMS-ESI (m/z):
[M]+ calcd for C65H70S2F6, 1028.4823; found 1028.4808. Anal. Calcd
for C65H70S2F6: C, 75.84; H, 6.85. Found: C, 76.08; H, 7.05. All values
are given as percentages. UV–vis (n-hexane) ꢀmax, nm (log ε): 312
(4.76), 580 (4.59).
phenylethynylthiophene (9c)
Prepared by a method similar to that used for 9a, 9c was
obtained as a colorless solid (67%, mp 63.1–63.7 ◦C). 1H NMR (CDCl3,
ı): 0.88 (m, 3H, CH3), 1.26 (m br, 18H, CH2), 1.61 (m, 2H, CH2), 2.61
(m, 2H, CH2), 7.15 (s, 1H, H4), 7.16–7.22 (m, 2H, Ar H), 7.32–7.38
(m, 3H, Ar H), 7.42–7.48 (m, 2H, Ar H), 7.51–7.59 (m, 2H, Ar H). 13
C
NMR (CDCl3, ı): 14.4, 22.9, 29.5, 29.6, 29.7, 29.8, 29.9, 31.6, 32.2,
35.9, 81.7, 97.7, 116.8, 119.4, 122.9, 125.5, 125.8, 128.6, 128.9, 129.4,
130.4, 131.8, 144.1, 145.7. HRMS-EI (m/z): [M]+ calcd for C30H35SBr,
506.1643; found 506.1623. Anal. Calcd for C30H35SBr: C, 70.99; H,
6.95. Found: C, 71.41; H, 6.65. All values are given as percentages.
2.3.13. 1,2-Bis(5-phenyl-2-phenylethynylthien-3-yl)
hexafluorocyclopentene (1a)
2.4. Procedures for the synthesis of photochromic actinometer 11
Adapted from previously published procedures [3,25], n-
butyllithium (0.17 mL of a 2.5 M solution in hexanes, 0.42 mmol)
was added dropwise to a solution of 9a (0.139 g, 0.410 mmol) in
tetrahydrofuran (10 mL) at −78 ◦C and under an atmosphere
of argon. After stirring for 30 min, octafluorocyclopentene
(0.0275 mL, 0.205 mmol) was added dropwise to the reaction
mixture at −78 ◦C. The mixture was stirred for another 4 h at
−78 ◦C and then allowed to warm to room temperature. After
2 h, hydrochloric acid (50 mL of a 1.2 M aqueous solution) was
added. Following vigorous stirring, tetrahydrofuran was removed
under reduced pressure. After neutralization of the acid with
saturated aqueous solution of sodium bicarbonate, the mixture
was extracted with chloroform (3 × 20 mL), dried over anhydrous
magnesium sulfate, filtered, and concentrated under reduced
pressure. Purification of the crude product by column chromatog-
raphy (1:3 toluene/hexanes) and recrystallization (n-hexane) gave
pure 1a as yellow crystals (0.077 g, 54%). 1H NMR (CD3COCD3, ı):
7.18–7.25 (m, 4H, Ar H), 7.27–7.41 (m, 16H, Ar H), 7.50 (s, 2H, H4).
13C NMR (CD3COCD3, ı): 80.5, 100.5, 121.9, 123.0, 123.9, 128.8,
129.0, 129.3, 131.5, 132.3, 132.5, 146.4. 19F NMR (CD3COCD3,
ı): −110.5 (t, J = 5.2 Hz, 4F), −132.7 (m, J = 5.2 Hz, 2F). HRMS-ESI
(m/z): [M]+ calcd for C41H22S2F6, 692.1067; found, 692.1066. Anal.
Calcd for C41H22S2F6: C, 71.09; H, 3.20. Found: C, 71.32; H, 2.97.
All values are given as percentages. UV–vis (n-hexane) ꢀmax, nm
(log ε): 308 (4.63), 574 (4.07).
2.4.1. (E)-2-(1-(2,5-Dimethylfur-3-yl)ethylidene)-3-
isopropylidenesuccinic acid (10)
Adapted from a previously published procedure [26], a mix-
ture of diethyl isopropylidenesuccinate (3.69 g, 17.2 mmol) and
3-acetyl-2,5-dimethylfuran (2.38 g, 17.2 mmol) was added to a
stirred suspension of sodium hydride (1.38 g of a 60% dispersion
in oil, 34.2 mmol) in toluene (35 mL) at 60 ◦C and under an atmo-
sphere of argon as shown (Scheme 3). A drop of ethanol was added
to initiate the reaction. After 1.5 h, the brownish reaction mix-
ture was allowed to cool to room temperature and poured onto
crushed ice (60 g) to give a biphasic mixture. The mixture was
extracted with water (2× 50 mL) and the combined aqueous layers
were acidified to a pH of 2 with hydrochloric acid (6.0 M aque-
ous solution) to liberate an oil. The organic phase was washed
with water, extracted with ethyl acetate (3× 80 mL), dried over
anhydrous magnesium sulfate, filtered, and concentrated under
reduced pressure. The (E)- and (Z)-half-esters were obtained as a
red gum (5.0 g, 96%). A 6% ethanolic solution of potassium hydrox-
ide (50 mL) was added to the half-esters, refluxed for 15 h, and
concentrated under reduced pressure. Water (60 mL) was added to
the mixture and washed with diethyl ether (2× 40 mL). The aque-
ous layer was acidified to a pH of 2 with hydrochloric acid (6.0 M
aqueous solution), extracted with ethyl acetate (3× 60 mL), dried
over anhydrous magnesium sulfate, filtered, and concentrated
under reduced pressure. The brownish oil was recrystallized (ethyl
acetate/hexane) and an off-white solid, the (Z)-diacid, was removed
by filtration (1.3 g). The remaining mother liquor was concentrated
under reduced pressure and gave 10 as a brownish oil (3.4 g, 74%).
This crude product was used for the next step without further
purification.
2.3.14. 1,2-Bis(5-biphenyl-2-phenylethynylthien-3-yl)
hexafluorocyclopentene (2a)
Prepared by a method similar to that used for 1a, 2a was
obtained as yellow crystals (46%). 1H NMR (CD3COCD3, ı):
7.20–7.54 (m, 20H, Ar H), 7.57 (s, 2H, H4), 7.60–7.75 (m, 8H, Ar
Scheme 3. Synthesis of 11. Reagents and conditions: (i) NaH, toluene, Ar, 60 ◦C, 1.5 h; (ii) KOH (6% in EtOH), 15 h; (iii) Ac2O, 70 ◦C, 30 min.

180
J. Cai et al. / Journal of Photochemistry and Photobiology A: Chemistry 212 (2010) 176–182
2.4.2. (E)-2-(1-(2,5-Dimethyfurl-3-yl)ethylidene)-3-
isopropylidenesuccinic anhydride (11)
of the integrated areas for the peaks representing both isomers and
was calculated as a percentage. Note that irradiation of 1a with UV
light near the photostationary state did produce an unidentified
photoproduct which was not reversible upon exposure to visi-
ble light. Irreversible photoproducts resulting from the diatropic
rearrangement of the closed-ring isomer of other dithienylethenes
have been reported [31,32]. The observed differences between the
UV–vis spectra obtained by HPLC for 1b and the photoproduct are
consistent with these previous reports. As a result, prolonged irra-
diation with UV light should be avoided. The photoproduct was not
observed during irradiation times used for quantum yield determi-
nations. The integrated area for the photoproduct was included in
our sum and was <5% of the total integrated area. The retention
times for the unidentified photoproduct, 1a, 1b, 2a, 2b, 3a, and 3b
were 7.0, 7.4, 12.1, 7.7, 11.0, 20.9, and 27.0 min, respectively.
Compound 11, commonly known as Aberchrome 540, was syn-
thesized since it is no longer commercially available. Adapted from
a previously published procedure [26], a solution of 10 (3.30 g,
11.8 mmol) in acetic anhydride (34.2 mL, 362 mmol) was stirred at
70 ◦C for 30 min and concentrated under reduced pressure. Purifi-
cation of the crude product by column chromatography (1:8 ethyl
acetate/hexanes) and recrystallization (ethyl acetate/hexane) gave
pure 11 as orange crystals (1.0 g, 33%, mp 124.0–124.5 ◦C). 1H NMR
(CDCl3, ı): 1.35 (s, 3H, CH3), 1.99 (s, 3H, CH3), 2.24 (s, 3H, CH3), 2.33
(s, 3H, CH3), 2.57 (s, 3H, CH3), 5.92 (d, J = 0.6 Hz, 1H, Ar H).
2.5. Quantum yield and photoconversion studies
The cyclization and cycloreversion quantum yields of 1–3 in
n-hexane were determined using 11 in n-hexane as a chemical
actinometer. The procedures used to measure the photon rate or
moles of photons absorbed by the irradiated solution per unit
time have been well described [26–28]. Briefly, the absorbance
of the solution containing 11 was matched to a solution contain-
ing the dithienylethene (i.e., 1–3). The absorbances were matched
at 313 and 530 nm for the cyclization and cycloreversion quan-
tum yields, respectively. The matched solutions were irradiated
with either UV or visible light for a known period of time under
identical experimental conditions. After irradiation, the absorbance
of each solution was measured immediately. From the change in
absorbance for the solution containing 11, the photon rate (NAh␯/t)
was calculated using Eq. (1) when not all the incident light is
absorbed [29],
2.6. X-ray crystallography
Structure solution was carried out using the SHELX97 suite of
programs [33] and the WinGX graphical interface [34]. Initial solu-
tions were obtained by direct methods and refined by successive
least-squares cycles. The fluorinated cyclopentene ring in the struc-
ture of 2a was disordered.
3. Results and discussion
Dithienylethenes 1a–3a all contain rigid phenylethynyl sub-
stituents at the C2 position (C␣) of both thiophene rings, the
reactive carbons that participate in electrocyclic ring-closing and
ring-opening reactions. The presence of potentially reactive alkyne
substituents does present some limitations to the relative sub-
stitution order on thiophene. For this reason, our regioselective
approach introduces functional groups at C␣ꢀ of thiophene prior
to the introduction of alkynyl substituents at C␣. The synthetic
methodology presented here is based on the conversion of 2,3,5-
tribromothiophene (4) to 2,4-dibromothiophene (5) (Scheme 2).
Palladium-catalyzed Suzuki cross-coupling reactions were used
for the regioselective substitution at C2 of the ␣ꢀ-halogenated
thiophene ring systems [23]. The availability of boronic acid deriva-
tives is dependent on the conversion of halogenated reagents
to the corresponding borates. We explored the possibility of
forming a borate at C2 of 5, however, reaction with an addi-
tional equivalent of n-BuLi followed by quenching with water
showed no evidence for lithiation at C2 (C␣ꢀ). Further, we observed
the formation of a 2-lithio-3,5-dibromothiophene intermediate,
primarily due to bromine rearrangements accompanied by hydro-
gen abstraction. This base-catalyzed halogen dance reaction has
been reported for other ␣-halothiophene ring systems [35,36].
Therefore, phenylboronic acid (6a), biphenylboronic acid (6b),
and 4-dodecylphenylboronic acid (6c) [22] were employed in
palladium-catalyzed Suzuki cross-coupling reactions with 5. We
observed regioselective coupling of 5 at C2 (C␣ꢀ) to produce
compounds 7a, 7b, and 7c in modest yields (72, 79, and 60%,
respectively). A two-phase catalytic method was utilized for fur-
ther bromination of 7a–7c. It has been shown that 0.1–10 mol%
of 70% perchloric acid can act as a phase-transfer catalyst in
the electrophilic substitution of heterocycles such as thiophene
[24]. The electrophilic bromination of 7a–7c was achieved in high
yields (>70%) and regioselectively at C␣ to produce 8a–8c. The
regioselective conversion of the 2,4-disubstituted thiophene to the
2,3,5-trisubstituted analog is illustrated in Fig. 1 by comparing the
aromatic regions of the 1H NMR spectra of 7c (a) and 8c (b). It is
important to note that no side-products were found at either C␤ꢀ
of thiophene, or the benzylic carbon of 7c. Conditions for the well-
known palladium-catalyzed Sonogashira cross-coupling were used
ꢀ
ꢁ
NAhꢁ
t
log T0
(1 − T0)
log Tt
(1 − Tt)
V
=
−
×
(1)
˚
11 × ε473 × l × t
where T = 10−A, A is the absorbance at 473 nm before irradiation,
At is the absorbance at 473 nm after irradiation, V is the irradiated
volume (L), ˚11 is the quantum yield for the cyclization of 11 at
the irradiation wavelength used (˚11 = 0.21 [30]), ε473 is the molar
absorptivity of 11 at 473 nm (ε473 = 8189 M−1 cm−1 [30]), l is the
optical pathlength of the quartz cell, and t is the irradiation time.
To determine the cyclization quantum yield, the change in
absorbance for the dithienylethene solution before and after irradi-
ation with UV light was measured at the wavelength of maximum
absorption (ꢀmax) in the visible region. From this change in
absorbance and rearranging Eq. (1), the quantum yield (˚) was
calculated using Eq. (2),
ꢀ
ꢁ
log T0
(1 − T0)
log Tt
(1 − Tt)
V
˚ =
−
×
(2)
(NAhꢁ/t) × εmax × l × t
where T0 is the transmittance at ꢀmax before irradiation, Tt is the
transmittance at ꢀmax after irradiation, εmax is the molar absorp-
tivity at ꢀmax in the visible region. The cycloreversion quantum
yield was determined in a similar manner. Before the solutions
containing 11 and the dithienylethene were matched, they were
irradiated with UV light (ca. 10–30 min) to their photostationary
state. The absorbance of each solution was measured following irra-
diation with visible light and the photon rate was calculated using
˚11 = 0.10 [29]. In order to determine εmax for the closed-ring iso-
mer of the dithienylethene, the concentration of the closed-ring
isomer at the photostationary state required calculation. This was
achieved by multiplying the photoconversion (see below) by the
concentration of the dithienylethene solution prior to irradiation
with UV light.
The photoconversion from the open-ring to the closed-ring iso-
mers at the photostationary state after irradiation with UV light was
determined for 1–3 from HPLC studies. The integrated area for the
peak representing the closed-ring isomer was divided by the sum

J. Cai et al. / Journal of Photochemistry and Photobiology A: Chemistry 212 (2010) 176–182
181
Fig. 2. Normalized absorption spectra of 2 in n-hexane prior to irradiation (a) and
after irradiation with UV or visible light in the following sequence: (b) 6 min of UV,
(c) 21 min of UV, (d) 15 s of visible, and (e) 80 s of visible. The inset is an expansion
of the 475–700 nm region. The absorption spectra recorded after irradiation were
normalized to the absorption spectrum prior to irradiation with light by dividing
the absorbance values of the former by the absorbance value of the latter at ꢀmax
(i.e., 326 nm).
Fig. 1. 1H NMR spectra showing regioselective substitutions on thiophene (aro-
matic regions only) for 7c (a), 8c (b), and 9c (c). The peaks labeled with asterisks
represent the following proton environments: *thiophene, **2-dodecylphenyl, and
***2-phenylethynyl.
to conjugate phenylacetylene to the bromo-substituted C2 posi-
tion (C␣) of 8a–8c [23]. Regioselective modification at C2 (C␣) was
clearly observed by 1H NMR (Fig. 1(c)). The final coupling of 9a–9c
with octafluorocyclopentene gave compounds 1a–3a in moderate
yields.
the cyclization quantum yield cannot exceed 0.5. Consequently,
the cyclization quantum yield is dependent on the ratio of the
conformations. When the ratio of the non-photoreactive paral-
lel conformation is increased, the quantum yield is expected to
decrease. Consistent with a previous report [12], the photoconver-
sion from 1a to 1b was 56% and the cyclization quantum yield was
0.14. The cyclization quantum yield for 2a and 3a were 2-fold and
ca. 4-fold lower than 1a, respectively. The large decreases in the
cyclization quantum yields strongly correlate with the lower con-
versions for 2a and 3a of 40 and 21%, respectively. The decrease
in conversion efficiencies is most likely due to an increased con-
centration of the parallel conformation prior to photoirradiation.
In particular, we suggest that intramolecular interactions between
the long alkyl chains of 3a will favor the parallel conformation, thus
lowering the cyclization quantum yield.
The introduction of phenylethynyl substituents at the reactive
carbons has been shown to enhance the efficiency of the cyclorever-
sion reaction relative to the cyclization reaction [12]. We observed
similar enhancements, in which cycloreversion quantum yields for
1b–3b were at least 3-fold larger than the cyclization quantum
yields.
X-ray crystallographic analysis was performed for 2a. As shown
in Fig. 3, 2a is packed in the crystal as the antiparallel conformer.
The distance between the reactive carbon atoms in 2a was deter-
mined to be 355 pm, which is short enough for photocyclization
to occur within the crystal [38]. However, given the large change
in molecular geometry upon isomerization, 2a was not expected
to undergo a crystal-to-crystal transformation. A similar finding
was observed for 1a where the distance between reactive carbon
Compounds 1a–3a were observed to undergo reversible pho-
toisomerization in n-hexane. Upon irradiation with UV light, a pale
yellow solution of 2a turned blue and an increase in absorbance
was observed at 591 nm with a concomitant decrease at 326 nm, as
shown in Fig. 2(b). This new absorption band in the visible region
represents the formation of the closed-ring isomer 2b. The wave-
lengths of maximum absorption for 2 were ca. 18 nm larger than
1 due to the extended conjugation length of the biphenyl deriva-
tive (Table 1). Further irradiation with UV light brought the system
to a photostationary state of the open-ring and closed-ring isomers
(Fig. 2(c)). In fact, the isosbestic point observed at 439 nm is a strong
indication that only two interconverting species are present in solu-
tion [37]. The photoisomerization of 2b was completely reversible
following irradiation with visible light (Fig. 2(d and e)). Similar
reversibility was also observed for 1b and 3b. In addition, no ther-
mal discoloration of 1b–3b was observed even after 24 h in the dark
at room temperature.
The cyclization and cycloreversion quantum yields of 1–3 were
also determined in n-hexane. An equilibrium mixture of two con-
formers exists for dithienylethenes in solution. The thienyl rings
have either mirror symmetry (parallel conformation) or C2 symme-
try (antiparallel conformation). The conrotatory photocyclization
reaction can proceed only from the antiparallel conformation [3].
In general, the ratio of the two conformers is 1:1, therefore,
Table 1
Absorption properties and photochromic reactivity of dithienylethenes (DTE) in n-hexane.a
ꢀmax (nm)b
εmax (104 M−1 cm−1
)
˚
o→c
DTE
ꢀmax (nm)b
εmax (104 M−1 cm−1
)
˚
c→o
Conversion (%)e
c
d
DTE
1a
2a
3a
308
326
312
4.27 0.14
8.01 0.35
5.76 0.23
0.14 0.01
0.070 0.002
0.036 0.002
1b
2b
3b
574
591
580
1.17 0.04
2.67 0.12
3.86 0.09
0.44 0.02
0.23 0.03
0.13 0.01
56
40
21
a
The error is the standard deviation for the mean taken from a minimum of three independent measurements.
The error is 1 nm.
Cyclization quantum yield.
Cycloreversion quantum yield.
The error is 2%.
b
c
d
e

182
J. Cai et al. / Journal of Photochemistry and Photobiology A: Chemistry 212 (2010) 176–182
ate Student Research Awards (USRA). The authors also thank B.T.
Sterenberg for helpful discussions.
References
[1] H. Dürr, H. Bouas-Laurent, Photochromism. Molecules and Systems, Elsevier
Science, New York, 1990, 1–14.
[2] K. Uchida, M. Irie, Photochromism of Diarylethene Derivatives, in: W. Horspool,
F. Lenci (Eds.), CRC Handbook of Organic Photochemistry and Photobiology, vol.
35, CRC Press, London, 2004, pp. 1–14.
[3] M. Irie, Chem. Rev. 100 (2000) 1685–1716.
[4] S. Kobatake, S. Takami, H. Muto, T. Ishikawa, M. Irie, Nature 446 (2007) 778–781.
[5] S.H. Chen, H.M.P. Chen, Y. Geng, S.D. Jacobs, K.L. Marshall, T.N. Blanton, Adv.
Mater. 15 (2003) 1061–1065.
[6] C. Denekamp, B.L. Feringa, Adv. Mater. 10 (1998) 1080–1082.
[7] M. Frigoli, C. Welch, G.H. Mehl, J. Am. Chem. Soc. 126 (2004) 15382–15383.
[8] T. Ikeda, J. Mater. Chem. 13 (2003) 2037–2057.
Fig. 3. ORTEP diagram showing the molecular structure of 2a. Hydrogen atoms have
been omitted for clarity.
[9] C. Kim, K.L. Marshall, J.U. Wallace, S.H. Chen, J. Mater. Chem. 18 (2008)
5592–5598.
[10] R.P. Lemieux, Soft Matter 1 (2005) 348–354.
[11] K.E. Maly, M.D. Wand, R.P. Lemieux, J. Am. Chem. Soc. 124 (2002) 7898–7899.
[12] K. Morimitsu, S. Kobatake, M. Irie, Tetrahedron Lett. 45 (2004) 1155–1158.
[13] K. Morimitsu, S. Kobatake, M. Irie, Mol. Cryst. Liq. Cryst. 431 (2005) 451–454.
[14] K. Yumoto, M. Irie, K. Matsuda, Org. Lett. 10 (2008) 2051–2054.
[15] J.G. Calvert, J.N. Pitts Jr., Photochemistry, Wiley & Sons, New York, 1966.
[16] D.D. Perrin, W.L. Armarego, Purification of Laboratory Chemicals, Elsevier Sci-
ence, New York, 1988.
atoms was 344 pm [12]. In addition, we suggest that intramolec-
ular ␲–␲ interactions between the biphenyl and phenylethynyl
substituents in the crystalline state may also suppress this
reactivity.
4. Conclusions
This work provides a regioselective approach for the preparation
of dithienylethenes containing phenylethynyl substituents at the
reactive carbons, and various functional groups at the periphery.
The prepared dithienylethenes undergo reversible photoisomer-
ization in solution and the closed-ring isomers are thermally stable.
Compound 3a has been specifically designed to improve the lipid
complementarity of these photochromic molecules in liposomes,
through the integration of long alkyl chains. Given the large changes
in molecular geometry expected for 3a, this molecule may be used
in the development of photoresponsive liposomes with reversible
photocontrols over membrane stability. In fact, preliminary results
suggest that the photoisomerization of 1a–3a in liposomes is
reversible. While these results are encouraging, further studies are
currently underway to fully characterize the photoisomerization
and inclusion of these photochromic molecules in liposomes, and
to evaluate the effect of their photoisomerization on membrane
stability.
[17] L. Brandsma, S.F. Vasilevsky, H.D. Verkruijsse, Application of Transition Metal
Catalysts in Organic Synthesis, Springer-Verlag, New York, 1999.
[18] M. D’Auria, A. De Mico, F. D’Onofrio, G. Piancatelli, J. Chem. Soc., Perkin Trans.
1 (1987) 1777–1780.
[19] D.L. Ladd, P.B. Harrsch, L.I. Kruse, J. Org. Chem. 53 (1988) 417–420.
[20] H. Dorn, J.M. Rodezno, B. Brunnhöfer, E. Rivard, J.A. Massey, I. Manners, Macro-
molecules 36 (2003) 291–297.
[21] O. Haba, T. Hayakawa, M. Ueda, H. Kawaguchi, T. Kawazoe, React. Funct. Polym.
37 (1998) 163–168.
[22] F. Dötz, J.D. Brand, S. Ito, L. Gherghel, K. Müllen, J. Am. Chem. Soc. 122 (2000)
7707–7717.
[23] R. Pereira, B. Iglesias, A.R. de Lera, Tetrahedron 57 (2001) 7871–7881.
[24] Y. Goldberg, H. Alper, J. Org. Chem. 58 (1993) 3072–3075.
[25] S.L. Gilat, S.H. Kawai, J.-M. Lehn, Chem.-Eur. J. 1 (1995) 275–284.
[26] P.J. Darcy, H.G. Heller, P.J. Strydom, J. Whittall, J. Chem. Soc., Perkin Trans. 1
(1981) 202–205.
[27] H.J. Kuhn, S.E. Braslavsky, R. Schmidt, Pure Appl. Chem. 76 (2004) 2105–2146.
[28] M. Montalti, A. Credi, L. Prodi, M.T. Gandolfi, Handbook of Photochemistry, CRC
Press, Boca Raton, 2006.
[29] A.P. Glaze, H.G. Heller, J. Whittall, J. Chem. Soc., Perkin Trans. 2 (1992) 591–594.
[30] H.G. Heller, J.R. Langan, J. Chem. Soc., Perkin Trans. 2 (1981) 341–343.
[31] A. Peters, N.R. Branda, Adv. Mater. Opt. Electron. 10 (2000) 245–249.
[32] M. Irie, T. Lifka, K. Uchida, S. Kobatake, Y. Shindo, Chem. Commun. (1999)
747–748.
Acknowledgements
[33] G.M. Sheldrick, Acta Crystallogr., Sect. A 64 (2008) 112–122.
[34] L.J. Farrugia, J. Appl. Crystallogr. 32 (1999) 837–838.
[35] H. Fröhlich, W. Kalt, J. Org. Chem. 55 (1990) 2993–2995.
[36] S. Kano, Y. Yuasa, T. Yokomatsu, S. Shibuya, Heterocycles 20 (1983) 2035–2037.
[37] D.C. Harris, Quantitative Chemical Analysis, W.H. Freeman, New York, 1991.
[38] S. Kobatake, K. Uchida, E. Tsuchida, M. Irie, Chem. Commun. (2002) 2804–2805.
We are grateful to the Natural Sciences and Engineering
Research Council (NSERC) of Canada, the Saskatchewan Health
Research Foundation (SHRF), and the University of Regina for
support of this work. A.F. and V.B. thank NSERC for Undergradu-