The synthesis of rigid polycyclic structures for the study of diatropic or steric effects of a phenyl ring on CF bond

Article abstract of DOI:10.1021/jo402154f

Polycyclic compounds 1a-c were synthesized to study the diatropic effects of a flanking phenyl ring on nearby CH and CF bonds. ¹⁹F NMR spectra of 1b and 1c were strongly deshielded compared with those of the ring-opened compounds 3b, 7b, and 7c

Full text of DOI:10.1021/jo402154f

Note
pubs.acs.org/joc
The Synthesis of Rigid Polycyclic Structures for the Study of
Diatropic or Steric Effects of a Phenyl Ring on CF Bond
Yung-Yu Chang, I-Ting Ho, Tse-Lok Ho,* and Wen-Sheng Chung*
Department of Applied Chemistry, National Chiao-Tung University, Hsinchu 30050, Taiwan ROC
*
S Supporting Information
ABSTRACT: Polycyclic compounds 1a−c were synthesized to
study the diatropic effects of a flanking phenyl ring on nearby CH
1
9
and CF bonds. F NMR spectra of 1b and 1c were strongly
deshielded compared with those of the ring-opened compounds 3b,
7
b, and 7c. DMol3 calculations on 1a−c provided quantitative
bond lengths and torsional angles to support the conclusion that
1
9
the downfield shifts in the F NMR spectra are mainly due to steric
interactions between the CF bonds and the π clouds of the phenyl
ring(s).
nter- and intramolecular CH−π and CF−π interactions play
interactions between the CF bond and the π electrons of
alkenes were explained by Lectka to be due to steric hindrance,
7
I
important roles in host−guest chemistry, molecular
1
assembly, and the folding of proteins and polynucleotides.
The CH−π interaction is a weak force (0.5−2.5 kcal/mol)
anchimeric assistance, and the repulsive interactions from
overlap of lone-pair electrons on fluorine with the π electrons
on the olefin. Furthermore, they reported that the fluorine was
little perturbed by anisotropic effects from the π electrons of
the alkene, and only a slight upfield shift relative to theoretical
2a
that is usually difficult to measure directly using molecules with
flexible conformations. Therefore, the measurement of this
weak interaction in molecules with intramolecular folding and
2
7
unfolding has intrigued chemists for decades. Recently,
values of fluorine chemical shifts was observed. We report here
3
a
Tsuzuki reported an excellent review that summarizes
recently reported gas-phase measurements and high-level ab
initio calculations of the CH−π interactions. Ab initio
calculations show that the major source of attraction in the
CH−π interactions is the dispersion interaction, while the
contribution from electrostatic interactions is small. On the
the synthesis of the rigid polycyclic compounds 1a−c and
subsequent NMR spectral studies and theoretical calculations
to measure the diatropic effect of a phenyl ring on sterically
close aryl CH (1a) and CF bonds (1b and 1c).
Tandem Diels−Alder reactions followed by sequential
photocyclization reactions were the two key steps in our
3b
other hand, Nishio et al. surveyed and analyzed literature
results relevant to the CH−π interactions in crystal packing and
conformations based on data reported in crystallographic
databases (CSD and PDB). Moreover, as the organofluoride
compounds are getting more popular in medicinal chemistry,
the need to study CF−π interactions in organic and biological
molecules has markedly increased. In order to study such
CF−π interactions efficiently, one needs to have a special
molecular design in which the CF bond is pointing toward
8a
synthesis of the rigid polycyclic structures 1a−c. One of us
8b
and Winkler have provided excellent reviews of tandem
Diels−Alder reactions demonstrating that they are exception-
ally powerful methods in the synthesis of intricate polycarbo-
8,9
cycles.
Iodine-induced photocyclization of stilbene to
10
phenanthrene has also been well-studied. The CH and CF
bonds in the series of polycyclic frameworks 1a−c can be
arranged in a very close proximity to the aromatic ring, and
1
19
4
hence, H and/or F NMR spectroscopy can be used to
estimate the strength of the interactions between them and the
π-cloud of the phenyl ring. Moreover, theoretical calculations
were carried out on the polycyclic frameworks 1a−c to obtain
optimized geometries, bond distances, and torsional angles.
Synthesis of 1a. Compound 1a was obtained through a two-
step sequence that started by refluxing tetraphenylcyclopenta9-
dienone (2a) with 1,5-cyclooctadiene to afford compound 3a
conformationally rigid π bonds or aromatic rings.
In previous literature, the typical CF−π interactions between
fluoromethane and benzene/hexafluorobenzene were only
theoretically studied. For example, the very weak attractive
interaction between the fluorine of HF and hexafluorobenzene
5a
was measured on the basis of theoretical calculations.
Kawahara also indicated that a weak interaction between
fluoromethane and hexafluorobenzene was observed; further-
10
5b
in 71% yield (Scheme 1). Iodine-catalyzed photocyclization
more, it is worth noting that this was an attractive force.
Nowak, however, studied CF−π interactions by measuring the
effect of fluorine on the reactivity of a proximal double bond
Received: October 4, 2013
Published: December 10, 2013
6
using a polyfluorinated indacene system. The through-space
©
2013 American Chemical Society
12790
dx.doi.org/10.1021/jo402154f | J. Org. Chem. 2013, 78, 12790−12794

The Journal of Organic Chemistry
Note
a
11
Scheme 1. Synthesis and X-ray Crystal Structure of 1a
acetic acid to afford 2,2′-difluorobenzil (6) in 72% yield.
Subsequent aldol condensation of 6 with dibenzyl ketone under
alkaline conditions gave the difluorinated substrate 2b in 35%
yield. After tandem Diels−Alder reactions of 2b with 1,5-
cyclooctadiene, we obtained precursor 3b in 68% yield. The
photocyclization of 3b with a catalytic amount of iodine was
conducted by irradiation in a Rayonet photoreactor (λmax = 250
nm) at room temperature (Scheme 3). The photocyclization
Scheme 3. Products of the Photocyclization of 3b
a
Reagents and conditions: (a) 1,5-cyclooctadiene, reflux, 4 days, 71%;
(
b) I , toluene, Rayonet, 300 nm, 58%.
2
of 3a in toluene afforded the target polycyclic compound 1a in
8% yield. Because the target compounds 1a and 3a have
5
similar polarities (R = 0.5 in hexane eluent), it was difficult for
f
us to obtain pure 1a even after repeated column chromatog-
1
raphy. The H NMR spectrum of the purified sample of 1a still
showed about 10% of the starting material 3a; however, their
signals can be easily discerned through chemical shift analysis.
In particular, a new doublet signal around 8.5 ppm appeared,
which is regarded as one of the characteristic protons of the
phenanthrene ring in 1a. Moreover, the signals of the aromatic
protons of 1a showed a significant downfield shift compared
with those of 3a (Figure S18 in the Supporting Information).
Furthermore, the formation of 1a could be easily recognized by
GC−MS analysis, which showed a new peak with retention
1
reaction of 3b was monitored by H NMR spectroscopy
(
2
Figure S17 in the Supporting Information). After irradiation at
50 nm for 10 min, new signals around δH = 8−9 ppm
emerged, while signals of the aromatic protons of the starting
compound 3b gradually decreased. The photocyclization
reaction of compound 3b was completed within 6 h of
irradiation at 254 nm.
Similar to the trouble we met in purifying the products of the
photocyclization of 3a, we also had difficulty in purifying the
products of the photocyclization of 3b because products 1a−c
have similar polarities. Even after separation by column
chromatography and recrystallization, the reaction mixture
time (r ) of 50.9 min (with m/z 462.5) as opposed to the peak
t
for the starting material 3a at r = 33.6 min (with m/z 464.6)
t
(
see Figure S20 in the SI). Finally, we were lucky to obtain a
single crystal of 1a by crystallization from a mixed solvent of
dichloromethane and ethanol (2:8 v/v). The single-crystal X-
ray structure of 1a (shown in Scheme 1) proved that it has a
rigid polycyclic structure.
Synthesis of 1b. Encouraged by the successful synthesis of the
polycyclic compound 1a, we then applied this methodology to
construct the fluoride analogue 1b. The synthesis of fluorinated
substrates 2b and 3b is shown in Scheme 2. The fluorinated
substrate 2b was obtained through a three-step synthesis. First,
the benzoin condensation of 2-fluorobenzene was conducted
using catalytic amount of sodium cyanide in ethanol to give the
analyzed by HPLC still showed three major peaks at r = 23.9,
t
2
4.8, and 25.9 min with an area ratio of 8:85:5 (Figure S21 in
the Supporting Information). The three products were
recognized by mass spectrometry as 1b (r = 23.9 min, m/z
t
4
4
98), 1c (r = 24.8 min, m/z 481), and 1a (r = 25.9 min, m/z
t
t
11
63) (see Scheme 3). Notably, photolysis of 3b afforded not
desired product, 2,2′-difluorobenzoin (5), in 72% yield.
only the prospective fluorine product 1b but also the
defluorinated and dehydrofluorinated side products 1a and 1c
(Figures S20 and S21 in the Supporting Information). The
structure of 1a was fully characterized by H and C NMR,
DEPT-135, HRMS, and single-crystal X-ray analysis (vide supra
and the Experimental Section). Photolysis of compound 3b in
Benzoin 5 was then oxidized by copper(II) acetate in 80%
a
Scheme 2. Synthesis of 2b and 3b
1
13
the presence of a catalytic amount of iodine may extrude F (g),
2
H (g), and HF(g) from the difluorodihydrophenanthrene
2
intermediates IIIa−c, leading to the formation of the
10a
aromatized products 1b, 1c, and 1a in 8%, 85%, and 5%
yield, respectively, based on HPLC area ratios (Scheme 3).
Three factors must have played important roles in the
conformational distributions of 3b in its ground state (Scheme
3
): (1) intramolecular electronic repulsion between the two CF
a
bonds in 3b-I, (2) intramolecular steric hindrance between the
two CF bonds and the flanking phenyl rings in 3b-II, and (3)
intramolecular hydrogen-bonding interactions between CF and
Reagents and conditions: (a) cat. NaCN, EtOH, reflux, 3 h, 72%; (b)
Cu(OAc) , 80% AcOH, reflux, 2 h, 72%; (c) KOH, dibenzyl ketone,
EtOH, reflux, 1 h, 35%; (d) 1,5-cyclooctadiene, reflux, 24 h, 68%.
2
1
2791
dx.doi.org/10.1021/jo402154f | J. Org. Chem. 2013, 78, 12790−12794

The Journal of Organic Chemistry
Note
Table 1. 19F NMR Data for Organofluoride Compounds 1b,
1c, 3b, and 7−18
CH bonds in 3b-III. Among the three major conformations of
1
3,14
3b, the conformer 3b-III became the most stable one and
therefore led to the formation of compound 1c as the major
product (85% in HPLC ratio).
All attempts to separate the product mixtures failed, and we
could only collect a small amount of 1c by analytical HPLC
separation. After obtaining the fluorinated polycyclic com-
pound 1c and the mixture of 1b and 1c, we then took the
19
intended measurement on CF−π interaction using F NMR
1
9
spectroscopy. The F NMR spectra of 1b, 1c, and 3b showed
the fluorine peaks at −91.0, −92.6, and −111.3 ppm,
1
2
respectively, using hexafluorobenzene (δ = −162.2 ppm)
F
1
9
as an external standard. To our big surprise, the F NMR peaks
of 1b and 1c were downfield-shifted by 20.3 and 18.7 ppm,
respectively, compared with that of 3b (Figure S15 in the
Supporting Information); however, they were downfield-shifted
by 30.2 and 29.8 ppm compared with those of 1,8-
and 1-
respectively.
Furthermore, compound 3b is downfield-shifted by 3.8 ppm
1
3c
difluorophenanthrene 7b (δF = −121.2 ppm)
13a
fluorophenanthrene 7c (δ = −122.4 ppm),
F
13b
compared with 8 (δ = −115.1 ppm). Since the C−F bonds
F
of 1b and 1c are located very close to their phenyl rings, one
1
9
would have expected to see a significant upfield shift in their F
NMR resonances due to the diatropic shielding effect of the
phenyl rings. To our surprise, the fluorine chemical shifts of 1b
and 1c did not show any shielding effect of the phenyl ring but
instead showed strong deshielding (vide supra)!! Thus, the
diatropic shielding effect of phenyl rings did not seem to play
any role in determining the fluorine chemical shift of the C−F
bond, and other factors must have played more important roles
parallel to a nearby phenylcyclophane, was downfield-shifted by
only 2.3 ppm compared with that of fluorobenzene (18). It
14h
is important to note that our molecules 1b and 1c exhibited
19
downfield shifts of 30.2 and 29.8 ppm in their F NMR spectra
compared with those of 7b and 7c, respectively, which are by
far the largest downfield shift effects ever reported. Thus, a
1
9
nearby π cloud mainly exerts a steric effect on the F NMR
peak of a CF bond instead of the traditional diatropic effect and
therefore leads to strong downfield shifts of molecules 1b and
1c.
1
9
leading to the significant downfield shift of the F NMR peaks
of 1b and 1c.
It has been reported that the ring-current effect is relatively
The fluorine chemical shifts of compounds 1b and 1c were
1
9
1
14a,b
17
less important in F NMR than in H NMR,
whereas steric
also calculated by means of a published method, namely,
19
effects have a stronger influence on the chemical shift in
F
gauge including atomic orbitals (GIAO) combined with B3LYP
DFT using the cc-pVTZ basis set. The calculated fluorine
NMR. Even though the strong deshielding of the 19F NMR
peaks of 1b and 1c were opposite to what we originally
expected, the results are fully explainable by steric effects in
chemical shifts of compounds 1b (δ = −101.9 ppm) and 1c
F
(δ = −101.0 ppm) are upfield-shifted by 10.9 and 8.4 ppm
F
1
4a,b
fluorine NMR.
The fluorine chemical shifts of para- and
compared with the observed values (Table S10 in the
Supporting Information).
meta-substituted fluorobenzenes showed a reasonable correla-
tion with the resonance and inductive effects of the
DMol3 Calculations of the Conformations of 1a−c. In order to
rationalize the conformations of 1a−c, we also calculated their
geometry-optimized structures by the DMol3 molecular
1
4i−k
substituents;
however, because of the intramolecular steric
effect between these substituents and the adjacent fluorine
atom, the fluorine chemical shifts of ortho-substituted
15a,b
modeling method
and simulated in a CHCl environment,
3
1
4i−k
fluorobenzenes exhibit a poor correlation.
others
Dolbier and
in which the B3LYP functional with the double-numeric-quality
with polarization functions (DNP) basis set was used. The size
of the DNP basis set is comparable to that of the Gaussian 6-
31G** basis set, but DNP is more accurate than a Gaussian
1
3,14
reported that all sterically congested or hindered
organofluoride compounds are downfield-shifted in F NMR
19
compared with those that are sterically unhindered (see Table
1
9
15c
1
). For example, the F NMR peaks of cis-9−11 were
basis set of the same size. The optimized geometries of 1a−c
downfield-shifted by 3−6 ppm compared to those of trans-9−
are displayed and related data of these calculations are
summarized in Tables S1−S6 in the Supporting Information).
In the optimized geometries of 1a−c, the distances between the
tips of the C−H and C−F bonds to the center of a phenyl ring
were measured to be 2.737, 3.367, and 3.393 Å, respectively.
For most literature reports on CH−π interactions, the distances
between the tip of the C−H/C-F bond to the center of a
1
4a−c
1
1;
furthermore, the peaks of cis-13−15 were downfield-
14d−f
shifted by 1−8 ppm compared with those of trans-13−15.
19
The deshielding effect of a bulkier substituent on the F NMR
spectrum is even more obvious in compounds 12a−d, where
the 19F NMR peak of 12d shows the largest downfield-shift
effect when its 8-substituent changed from H to t-Bu (Δδ =
1
4a,b
2c,3b,16
2
7.5 ppm).
phenyl ring range from 2.9 to 3.5 Å (typically 3.05 Å).
7
In addition, Lectka and co-workers synthesized compounds
Even though the C−H bonds of phenanthrene on 1a are not
pointing toward its flanking phenyl rings, they still fit
conventional CH−π interaction characteristics. The C10−
C9−C8−C1 torsional angles, Φ, of 1a−c were calculated to be
20.5°, 28.8°, and 35.0°, respectively, suggesting the increase in
repulsive interaction between CH/CF bonds and a phenyl ring,
1
6a and 16b to investigate the intramolecular CF−π
interactions. The C−F bond of 16b is very close to a double
bond, causing a 16.0 ppm downfield shift of its F NMR
resonance compared with that of 16a. The F NMR peak of
19
7
19
14g
4-fluoro[2.2]paracyclophane (17), which has its CF bond
1
2792
dx.doi.org/10.1021/jo402154f | J. Org. Chem. 2013, 78, 12790−12794

The Journal of Organic Chemistry
Note
as one would expect from the overlap of the π system with the
fluorine or hydrogen atoms. As a result of the steric hindrance
between phenanthrene and the flanking phenyl rings in the
crystal structure of 1a, the conformations of phenanthrene and
the cyclic skeleton (C1−C33−C32−C22−C29−C36) in 1a are
nonplanar and a twisted boat, respectively. The distance
between the tip of the CH bond and the center of a phenyl ring
was determined to be 2.567 Å, and the torsional angle Φ of 1a
was shown to be 18.0°, so the Dmol3-calculated results for 1a
were very close to those of the crystal structure of 1a. We infer
that steric hindrance plays a pivotal role in the interactions
between the CF bond(s) and the phenyl ring(s) in 1b and 1c,
(300 MHz, CDCl
3
): δ
H
8.49 (d, J = 8.1 Hz, 2H), 7.38 (m, 3.0 Hz, 4H),
7
.34−7.19 (m, 8H), 7.05 (d, J = 8.1 Hz, 2H), 7.00−6.88 (m, 2H), 2.79
13
(
(
(
(
s, 4H), 2.04 (d, J = 9.3 Hz, 4H), 1.72 (d, J = 9.7 Hz, 4H). C NMR
75.4 MHz, CDCl ): δ 147.3 (Cq), 138.5 (Cq), 130.3 (Cq), 130.3
Cq), 128.8 (CH), 128.0 (CH), 126.0 (CH), 125.9 (CH), 124.5
CH), 124.0 (CH), 122.9 (CH), 56.7 (Cq), 48.2 (CH), 29.7 (CH ),
3
C
2
+
2
5.4 (CH ). EI-MS: m/z 463 ([M + H] ). HRMS (EI): m/z calcd for
2
C H 462.2348, found 462.2351.
36
30
X-ray Crystal Data for (1a) ·CH Cl . C H Cl , M = 1472.73,
monoclinic, a = 16.037(3) Å, b = 11.554(2) Å, c = 21.481(4) Å, α =
90°, β = 109.402(4)°, γ = 90°, V = 3754.0(13) Å , space group P1
3
2
2
109 92
2
3
̅
, Z =
−3
2
, calculated density 1.303 Mg/m , crystal dimensions: 0.52 mm ×
0
.50 mm × 0.08 mm, T = 200(2) K, λ(Mo Kα) = 0.71073 Å, μ = 0.142
−1
1
9
mm , 24974 reflections collected, 6543 independent reflections (R
int
leading to their strong deshielding in the F NMR spectra
compared with those without such a sterically hindered
environment.
2
2
=
0.0981), 501 parameters refined on F , R = 0.0673, wR (F ) =
1 2
2
−3
0
.1621 (all data), goodness of fit on F = 1.018, Δρ = 0.342 e Å .
max
CCDC 963498 contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from the
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
In conclusion, we have designed and synthesized a series of
rigid polycyclic structures 1a−c where the key steps of the
synthesis involves (1) tandem Diels−Alder reactions and (2)
the photocyclization followed by extrusion of F , H , and HF
Synthesis of 2b. To a solution of 6 (0.49 g, 1.99 mmol) and KOH
(0.06 g, 1.07 mmol) in EtOH (15 mL) was added 1,3-diphenylacetone
2
2
from the difluorodihydro- phenanthrene intermediates IIIa−c,
1
9
(0.42 g, 1.99 mmol), and the mixture was refluxed for 30 min and then
respectively. According to the observed F NMR of 1b and 1c,
their chemical shift were downfield shifted by 30.2 and 29.8
ppm when compared to those of the 7b and 7c, respectively.
cooled to room temperature. The solvent was removed under reduced
pressure, and the residue was partitioned between H O (30 mL) and
2
CH Cl (50 × 3 mL). The combined organic layers were dried over
1
9
2
2
Even though the strong deshielding of the F NMR of 1c was
opposite to what we originally expected, the results are fully
explainable by “steric effect” on fluorine NMR. The torsional
angles of 1a−c increased from 20.5° in 1a, to 28.8° in 1b, and
anhydrous MgSO and evaporated. The residue was recrystallized from
4
EtOH to afford the product 2b as a dark-red solid in 35% yield. Mp
1
1
60−161 °C. H NMR (CDCl , 300 MHz): δ 7.35−7.29 (m, 5H),
3
H
1
3
6.98−6.86 (m, 4H). C NMR (CDCl , 75.4 MHz): δ 199.4 (Cq),
159.3 (d, J = 248 Hz, Cq), 149.3 (Cq), 130.6 (d, J = 8.4 Hz, CH),
3
C
3
5.0° in 1c, suggesting the increase in repulsive interaction
between C−H/C-F bonds and a phenyl ring. We conclude that
steric hindrance must have played a pivotal role in the
interaction between the CF bond and the phenyl ring in 1c,
leading to its strong deshield in F NMR compared to those
without such a steric hindered environment.
130.4 (Cq), 130.2 (CH), 129.4 (CH), 128.1 (CH), 127.8 (CH), 126.6
(CH), 123.9 (CH), 121.4 (d, J = 17 Hz, Cq), 115.7 (d, J = 22 Hz,
+
CH). EI-MS: m/z 420.1 (M ). HRMS (EI): m/z calcd for C H F O
420.1326, found 420.1333.
Synthesis of 3a. Compound 2a (2.50 g, 6.51 mmol) in 1,5-
cyclooctadiene (10 mL) was heated at reflux for 4 days. After cooling
29 18
2
1
9
9
to room temperature, the suspension was filtered to afford the product
EXPERIMENTAL SECTION
General Methods. Column chromatography was performed on
0−230 or 230−400 mesh silica gel; thin-layer chromatography
TLC) was performed on aluminum plates coated with silica gel 60
1
3
a as a white solid (2.14 g, 71%). Mp 305−306 °C. H NMR (300
■
MHz, CDCl
): δ 7.23 (d, J = 7.4 Hz, 4H), 7.00 (t, J = 7.5 Hz, 4H),
3 H
7
6.89 (t, J = 7.2 Hz, 2H), 6.72−6.56 (m, 6H), 6.51 (dd, J = 7.6, 1.7 Hz,
1
3
(
4H), 2.88 (s, 4H), 1.95 (d, J = 9.1 Hz, 4H), 1.54 (s, 4H). C NMR
(75.4 MHz, CDCl ): δ 144.8 (Cq), 142.2 (Cq), 141.4 (Cq), 130.6
(CH), 128.4 (CH), 127.0 (CH), 126.0 (CH), 124.9 (CH), 124.2
F254. Melting points were determined with a melting-point apparatus
3
C
1
and are uncorrected. H NMR spectra were measured with a 300 MHz
+
spectrometer with the residual solvent peak (usually CHCl or
(CH), 56.6 (Cq), 45.8 (CH), 25.01 (CH
2
). EI-MS: m/z 464.4 (M ).
3
13
DMSO-d ) as the internal standard. Natural-abundance C NMR
HRMS (EI): m/z calcd for C36 32 464.2504, found 464.2496.
H
6
spectra were recorded using pulse Fourier transform techniques with a
Synthesis of 3b. A solution of compound 2b (0.10 g, 0.238 mmol)
in 1,5-cyclooctadiene (0.37 mL) was heated at reflux for 24 h. After
cooling to room temperature, the suspension was filtered to afford the
19
300 MHz spectrometer operating at 75.4 MHz. F NMR spectra were
measured on a 470 MHz spectrometer with the solvent peak (C F ) as
6
6
12
1
an external standard (δF = −162.2 ppm). High-resolution mass
spectrometry (HRMS) was performed with a magnetic-sector-type
analyzer using the EI method. UV/vis spectra were recorded with a
spectrophotometer, and solvents were of HPLC grade. HPLC
experiments were recorded with a Nucleosil-5 C18 column (4.5 mm
product 3b as a white solid (8.30 mg, 68%). Mp 275−276 °C. H
NMR (300 MHz, CDCl ): δ 7.44 (d, J = 9.0 Hz, 2H), 7.20 (d, J = 9.0
3 H
Hz, 2H), 7.08−6.88 (m, 6H), 6.67−6.58 (m, 4H), 6.50 (t, J = 7.2 Hz,
2H), 6.40 (t, J = 8.9 Hz, 2H), 6.52−6.47 (m, 1H), 3.0 (d, J = 12.1 Hz,
1
3
2H), 2.86 (m, 2H), 2.04−1.88 (m, 4H), 1.68−1.45 (m, 4H).
C
×
250 mm), and solvents were of HPLC grade; the mobile phase was
NMR (75.4 MHz, CDCl
136.7 (dd, J = 6 Hz, J = 4 Hz, CH), 130.7 (CH), 124.9 (CH), 121.2
(dd, J = 10 Hz, J = 3 Hz, Cq), 116.4 (d, J = 22 Hz, CH). EI-MS: m/z
3 C
): δ 190.2 (Cq), 162.9 (d, J = 254 Hz, Cq),
9
0−100% (v/v) MeOH/H O. Compounds 5 and 6 were prepared
1
2
2
11
according to literature procedures.
1
2
+
Photocyclization of 3a to 1a. A mixture of compound 3a (0.10 g,
.215 mmol) and a catalytic amount of iodine (0.547 mg, 0.0022
500.2 (M ). HRMS (EI): m/z calcd for C36H F 500.2316, found
30 2
0
500.2305.
mmol) in THF (250 mL) was stirred at room temperature and
irradiated in a Rayonet photoreactor at 300 nm for 8 h. The solvent
was removed under reduced pressure. The residue was dissolved in
CH Cl , washed with a 10% aqueous solution of Na S O followed by
Photocyclization of 3b to 1a−c. A mixture of compound 3b
(0.1g, 0.199 mmol) and a catalytic amount of iodine (5.0 mg, 0.0197
mmol) in THF (100 mL) was stirred at room temperature and
irradiated at 250 nm in a Rayonet photoreactor for 8 h. The solvent
was removed under reduced pressure. The residue was dissolved in
CH Cl , washed with a 10% aqueous solution of Na S O followed by
2
2
2
2
3
a saturated aqueous solution of Na CO , and dried over anhydrous
2
3
MgSO , and then the solvent was evaporated under reduced pressure.
4
2
2
2
2
3
The resulting residue was purified by flash column chromatography
a saturated aqueous solution of Na CO , and dried over anhydrous
2 3
(
hexane, R = 0.5) to afford a mixture of 3a and 1a in which 1a was
MgSO , and then the solvent was evaporated under reduced pressure.
The resulting residue was purified by column chromatography
f
4
1
obtained in 58% yield based on H NMR peak ratios. Single crystals of
a were obtained from crystallization of the mixture of 3a and 1a using
1
(hexane, R = 0.5) and recrystallization. However, the reaction mixture
f
1
a mixed solvent of dichloromethane and ethanol (2:8 v/v). H NMR
analyzed by HPLC still showed three major peaks at r = 23.9, 24.8,
t
1
2793
dx.doi.org/10.1021/jo402154f | J. Org. Chem. 2013, 78, 12790−12794

The Journal of Organic Chemistry
Note
and 25.9 min with an area ratio of 8:85:5. The mixture of compounds
B.; Mallory, C. W. J. Am. Chem. Soc. 1972, 94, 6041−6048. (g) Lewis,
F. D.; Kurth, T. L.; Kalgutkar, R. S. Chem. Commun. 2001, 1372−1373.
(11) Romanov-Michailidis, F.; Besnard, C.; Alexakis, A. Org. Lett.
2012, 14, 4906−4909.
1
a−c was also confirmed by EI-MS and HRMS. For 1a: m/z 463 ([M
+
+
H] ); HRMS (EI) calcd for C H 462.2348, found 462.2354. For
36 30
+
1
c: m/z 481 ([M + H] ); HRMS (EI) calcd for C H F 480.2253,
36 29
+
found 480.2250. For 1b: m/z 498 (M ); HRMS (EI) calcd for
C H F 498.2159, found 498.2163. The 1b:1c:1a peak-area ratio was
(12) Stadlbauer, S.; Ohmori, K.; Hattori, F.; Suzuki, K. Chem.
Commun. 2012, 48, 8425−8427.
36
29 2
determined to be 8:85:5 by HPLC analysis.
(13) (a) Dewar, M. J. S.; Kelemen, J. J. Chem. Phys. 1968, 49, 499−
5
08. (b) Byrne, P. A.; Rajendran, K. V.; Muldoon, J.; Gilheany, D. G.
ASSOCIATED CONTENT
Supporting Information
Org. Biomol. Chem. 2012, 10, 3531−3537. (c) Luthe, G.; Ariese, F.;
■
Brinkman, U. A. Th. Chromatographia 2004, 59, 37−41.
*
S
(14) This unexpected deshielding was explained as due to the direct
Crystallographic data for 1a (CIF); H and 13C NMR spectra
and MS data for compounds 1a−c, 2b, 3a, 3b, 5, and 6; and
DMol3 calculation data for compounds 1a−c. This material is
available free of charge via the Internet at http://pubs.acs.org.
1
overlap of the van der Waals radii of the adjacent substituents and
fluorine, causing the restricted motion of electrons on the fluorine and
thus making the fluorine nucleus respond to the magnetic field as if the
electron density were lowered. See: (a) Dolbier, W. R. Guide to
Fluorine NMR for Organic Chemists; Wiley: New York, 2009.
(b) Gribble, G. W.; Keavy, D. J.; Olson, E. R.; Rae, I. D.; Staffa, A.;
Herr, T. E.; Ferraro, M. B.; Contreras, R. H. Magn. Reson. Chem. 1991,
AUTHOR INFORMATION
Corresponding Authors
E-mail: tselokho@yahoo.com.
E-mail: wschung@nctu.edu.tw.
■
*
*
2
9, 422−432. Also see: (c) Rae, I. D.; Staffa, A.; Diz, A. C.; Giribet, C.
G.; Ruiz de Azua, M. C.; Contreras, R. H. Aust. J. Chem. 1987, 40,
1
L.; Koroniak, H. J. Am. Chem. Soc. 1990, 112, 363−367. (e) Landelle,
G.; Turcotte-Savard, M.-O.; Angers, L.; Paquin, J.-F. Org. Lett. 2011,
́
923−1940. (d) Dolbier, W. R.; Gray, T. A.; Keaffaber, J. J.; Celewicz,
Notes
The authors declare no competing financial interest.
1
3, 1568−1571. (f) Uno, H.; Sakamoto, K.; Semba, F.; Suzuki, H. Bull.
Chem. Soc. Jpn. 1992, 65, 210−217. (g) Ernst, L.; Ibrom, K. Magn.
Reson. Chem. 1997, 35, 868−876. (h) Furuya, T.; Kaiser, H. M.; Ritter,
T. Angew. Chem. 2008, 120, 6082−6085. (i) Fifolt, M. J.; Sojka, S. A.;
Wolfe, R. A.; Hojnicki, D. S. J. Org. Chem. 1989, 54, 3019−3023.
(j) Taft, R. W.; Price, E.; Fox, I. R.; Lewis, I. C.; Anderson, K. K.;
Davis, G. T. J. Am. Chem. Soc. 1963, 85, 3146−3156. (k) Taft, R. W.;
Price, E.; Fox, I. R.; Lewis, I. C.; Anderson, K. K.; Davis, G. T. J. Am.
Chem. Soc. 1963, 85, 709−724.
ACKNOWLEDGMENTS
■
We thank the National Science Council (NSC) and the MOE
ATU Program of the Ministry of Education, Taiwan, Republic
of China, for financial support.
REFERENCES
■
(
1) (a) Salonen, L. M.; Ellermann, M.; Diederich, F. Angew. Chem.,
(
15) (a) Delley, B. J. Chem. Phys. 1990, 92, 508−517. (b) Delley, B. J.
Int. Ed. 2011, 50, 4808−4842. (b) Jose, D.; Datta, A. J. Phys. Chem.
Lett. 2010, 1, 1363−1366. (c) Raghavender, U. S.; Aravinda, S.;
Shamala, N.; Kantharaju; Rai, R.; Balaram, P. J. Am. Chem. Soc. 2009,
Chem. Phys. 2000, 113, 7756−7764. (c) Benedek, N. A.; Snook, I. K.;
Latham, K.; Yarovsky, I. J. Chem. Phys. 2005, 122, No. 144102.
(
16) Mooibroek, T. J.; Gamez, P. CrystEngComm 2012, 14, 8462−
1
31, 15130−15132. (d) Zhao, R.; Matsumoto, S.; Akazome, M.;
Ogura, K. Tetrahedron 2002, 58, 10233−10241. (e) Brandl, M.; Weiss,
M. S.; Jabs, A.; Suhnel, J.; Hilgenfeld, R. J. Mol. Biol. 2001, 307, 357−
77. (f) Takahashi, H.; Tsuboyama, S.; Umezawa, Y.; Honda, K.;
Nishio, M. Tetrahedron 2000, 56, 6185−6191.
2) (a) Zhao, C.; Parrish, R. M.; Smith, M. D.; Pellechia, P. J.;
Sherrill, C. D.; Shimizu, K. D. J. Am. Chem. Soc. 2012, 134, 14306−
4309. (b) Nijamudheen, A.; Jose, D.; Shine, A.; Datta, A. J. Phys.
8
467.
(17) Saielli, G.; Bini, R.; Bagno, A. Theor. Chem. Acc. 2012, 131, 1140.
̈
3
(
1
Chem. Lett. 2012, 3, 1493−1496. (c) Carroll, W. R.; Zhao, C.; Smith,
M. D.; Pellechia, P. J.; Shimizu, K. D. Org. Lett. 2011, 13, 4320−4323.
(
3) (a) Tsuzuki, S. Annu. Rep. Prog. Chem., Sect. C: Phys. Chem. 2012,
1
08, 69−95. (b) Nishio, M.; Umezawa, Y.; Honda, K.; Tsuboyama, S.;
Suezawa, H. CrystEngComm 2009, 11, 1757−1788.
4) (a) Sun, H.; Tottempudi, U. K.; Mottishaw, J. D.; Basa, P. N.;
Putta, A.; Sykes, A. G. Cryst. Growth Des. 2012, 12, 5655−5662.
b) Chopra, D.; Guru Row, T. N. CrystEngComm 2011, 13, 2175−
186. (c) Hunter, L. Beilstein J. Org. Chem. 2010, 6, 1−14. (d) Bissantz,
G.; Kuhn, B.; Stahl, M. J. Med. Chem. 2010, 53, 5061−5084.
5) (a) Alkorta, I.; Rozas, I.; Elguero, J. J. Fluorine Chem. 2000, 100,
33−238. (b) Kawahara, S.-I.; Tsuzuki, S.; Uchimaru, T. J. Phys. Chem.
(
(
2
(
2
A 2004, 108, 6744−6749.
(6) Nowak, I. J. Fluorine Chem. 1999, 99, 59−66.
(7) Scerba, M. T.; Bloom, S.; Haselton, N.; Siegler, M.; Jaffe, J.;
Lectka, T. J. Org. Chem. 2012, 77, 1605−1609.
8) (a) Ho, T.-L. Tandem Reactions in Organic Synthesis; Wiley: New
York, 1992. (b) Winkler, J. D. Chem. Rev. 1996, 96, 167−176.
9) Akhtar, I. A.; Fray, G. I.; Yarrow, J. M. J. Chem. Soc. C 1968, 812−
15.
10) (a) Mallory, F. B.; Mallory, C. W. Org. React. 1984, 30, 1−456.
b) Mallory, F. B.; Wood, C. S.; Gordon, J. T. J. Am. Chem. Soc. 1962,
4, 4361−4362. (c) Mallory, F. B.; Gordon, J. T.; Wood, C. S. J. Am.
(
(
8
(
(
8
Chem. Soc. 1963, 85, 828−829. (d) Mallory, F. B.; Wood, C. S.;
Gordon, J. T. J. Am. Chem. Soc. 1964, 86, 3094−3102. (e) Wood, C.
S.; Mallory, F. B. J. Org. Chem. 1964, 29, 3373−3377. (f) Mallory, F.
1
2794
dx.doi.org/10.1021/jo402154f | J. Org. Chem. 2013, 78, 12790−12794