Synthesis and properties of the strained alkene perfluorobicyclo[2.2.0]hex-1(4)-ENE

Article abstract of DOI:10.1021/jo502456h

The title fluoroalkene has been generated by dehalogenation of dibromide and diiodide precursors and trapped in situ. retro-Diels-Alder reaction of its adduct with N-benzylpyrrole has made the alkene available in high yield and purity. In sharp contrast t

Full text of DOI:10.1021/jo502456h

Article
pubs.acs.org/joc
Synthesis and Properties of the Strained Alkene
Perfluorobicyclo[2.2.0]hex-1(4)-ene
†
†
†
†
‡
§
Christopher P. Junk, Yigang He, Yin Zhang, Joshua R. Smith, Rolf Gleiter, Steven R. Kass,
∥
,†
Jerry P. Jasinski, and David M. Lemal*
†
Department of Chemistry, Dartmouth College, Hanover, New Hampshire 03755, United States
‡
Organisch Chemisches Institut, Universitat
̈
Heidelberg, Im Neuenheimer Feld 270, D-69120 Heidelberg, Germany
§
Department of Chemistry, University of Minnesota, 207 Pleasant Street Southeast, Minneapolis, Minnesota 55455, United States
∥Department of Chemistry, Keene State College, Keene, New Hampshire 03435, United States
S Supporting Information
*
ABSTRACT: The title fluoroalkene has been generated by
dehalogenation of dibromide and diiodide precursors and
trapped in situ. retro-Diels−Alder reaction of its adduct with
N-benzylpyrrole has made the alkene available in high yield and
purity. In sharp contrast to its extremely labile hydrocarbon counterpart, the fluoroalkene is very stable yet highly reactive. Its
characterization includes its electron affinity, photoelectron spectrum, and the previously reported structure determination by
electron diffraction.
INTRODUCTION
Our synthesis and study of the title fluorocarbon (1) was
inspired in part by the fascinating work of Wiberg’s group on
Scheme 1
■
1
the highly strained parent hydrocarbon (2). Like its parent, 1
displays striking reactivity, but as a fluorocarbon, its behavior is
radically different. Experimental investigations, mostly brief, of
2
−5
the synthesis, nature, and chemistry of 1 have been reported,5
as well as theoretical studies of its structure and behavior. −8
Here we flesh out a larger portion of the story of this
remarkable compound. Because hydrocarbon 2 dimerizes and
polymerizes at temperatures below 0 °C, we were prepared at
the outset for 1 to be very labile.
Scheme 2
RESULTS AND DISCUSSION
■
The synthesis of 1 begins with transformation of hexafluor-
9
10
obenzene via its Dewar isomer (3) and dibromide 4 into the
2
octafluorodibromide 5 (Scheme 1). Selective replacement of
the bridgehead fluorines of 3 presumably occurs by way of
pentadienylic cations. In the fluorination of 4, vigorous
mechanical stirring and relatively high dilution are important
for minimizing oligomer formation.
Initial evidence of the transient existence of alkene 1
appeared when dibromide 5 was treated with methyllithium
in the presence of furan, producing the Diels−Alder adduct 6.
Omission of furan did not allow detection of 1, however, as it
reacted with the methyllithium to give alkenes 7 and 8 (Scheme
fluoride ion to give the very strained alkene 9. Attack on 9 in
S 2′ fashion by methyllithium yields 7 and by bromide ion
N
yields 8 with considerable relief of strain. The finding that rapid
addition of excess methyllithium gives 7 exclusively is
consistent with this interpretation. Assisted by ultrasound,
zinc dust also generated alkene 1 from 5, as shown by the
formation of furan adduct 6 (Scheme 4). Again over-reduction
2
3
). A likely pathway for this over-reduction is shown in Scheme
, where addition of methyllithium to 1 results in elimination of
Received: October 27, 2014
©
XXXX American Chemical Society
A
DOI: 10.1021/jo502456h
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 3
Scheme 7
exclusion of moisture, because water adds readily to the alkene,
and attempts to scale it up were unpromising. Thus, a different
approach for isolation of the alkene was sought.
It seemed possible that a Diels−Alder adduct of 1 could be
found that would undergo thermal retro-reaction to regenerate
the alkene under conditions that would allow its isolation. A
suitable diene should possess these characteristics: (1) sufficient
nucleophilicity to trap the electron-deficient alkene efficiently,
occurred, however, with the formation of an almost equal
amount of 10, a likely mechanism for which appears in Scheme
(
2) high thermodynamic stability to favor retro-Diels−Alder
reaction, and (3) low volatility to permit easy separation from
. The first candidate to be tried was 9-methoxyanthracene
5
.
1
Scheme 4
13
(
12), prepared from anthrone. This diene reacted with 1
formed by mercury reduction of diiodide 11, giving adduct 13
Scheme 8). The 19F NMR spectrum of 13 comprised two
(
overlapping AA′BB′ patterns (“front” and “back”) with the
same set of coupling constants.
Scheme 8
Scheme 5
14
Computer simulation with gNMR confirmed the structure
assignment and revealed the F−F coupling constants (Figure
1
). The endo−endo F−F coupling constant could not be
We hoped that diiodide 11, with its relatively weak C−I
bonds, would permit generation of 1 under sufficiently mild
conditions that it would survive. Synthesis of 11 could not be
accomplished like that of 5, though, as the fluorination step
wreaked havoc with the C−I bonds. Fortunately, the bromines
in 5 could be replaced with iodines by UV irradiation of a
solution of the dibromide and KI in an acetonitrile/ether
measured directly from the spectrum because of the plane of
symmetry that makes endo-coupled fluorines chemical-shift
equivalent. Each half of each AA′BB′ pattern consists of two
pseudotriplets, the splitting of which depends on the value of
JFF(endo) (and disappears when that value is small). The X-ray
crystal structure of 13 was determined, and it reveals F−F endo
distances of 2.50 and 2.52 Å, well within the sum of the atomic
radii (2.94 Å). The 68 Hz coupling arises primarily, if not
exclusively, from through-space interaction. When 13 was
heated, it did undergo the desired retro-Diels−Alder reaction,
but not until the temperature exceeded 180 °C.
1
1
mixture (3:1) (Scheme 6). We dubbed this halogen
Scheme 6
The next diene to be tried as the source of a precursor for
alkene 1 was a pyrrole. Pyrroles offer the dual advantages of
high nucleophilicity to favor the trapping reaction and robust
aromaticity to promote the retro-Diels−Alder reaction. N-
15
Benzylpyrrole (14), chosen for its nonvolatility, trapped
alkene 1 efficiently, yielding adduct 15. The reaction was run
using diiodide 11 and mercury in acetonitrile with ultrasound,
but it was conducted most conveniently with dibromide 5 and
zinc or a zinc/copper couple (Scheme 9). The nicely
crystalline adduct 15 (mp 99−100 °C) proved to be an
excellent precursor for 1.
metathesis with iodide ion a “photo-Finkelstein reaction”.
12
Unlike its dark S 2 namesake, the photochemical trans-
N
formation proceeds via electron transfer. Diiodide 11 was
sufficiently reactive that even mercury or silver in acetonitrile
reduced it to 1 with assistance from ultrasound, and the alkene
3
2
was trapped with a variety of reagents.
When the diiodide was irradiated with UV light in methylene
chloride in the presence of copper powder as an iodine
scavenger, the 1 F NMR spectrum of the reaction mixture
revealed a singlet at δ ca. −100. Addition of furan caused the
singlet to disappear with concomitant growth of signals for
Diels−Alder adduct 6 (Scheme 7). This was the first successful
attempt to observe 1 directly. The reaction required strict
Like adduct 13, 15 has C symmetry, but its plane of
S
1
9
symmetry is perpendicular to that of 13. As a result, its
F
9
NMR spectrum (Figure 2) offers an interesting contrast with
that of the 9-methoxyanthracene adduct. The spectrum
comprises two AB quartets, with the lower field half of each
split into doublets by the endo−endo coupling, here shown
directly. A 7 Hz through-space coupling between the vinyl
B
DOI: 10.1021/jo502456h
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Figure 1. Experimental (A) and simulated (B) 19F NMR spectra (CDCl ) of 9-methoxyanthracene adduct 13.
3
Scheme 9
N-Benzylpyrrole adduct 15 underwent retro-Diels−Alder
reaction at a suitable rate at 140 °C, a temperature considerably
lower than that for adduct 13 (Scheme 10). It was heated under
Scheme 10
1
protons and neighboring exo fluorines, revealed by the H and
HF COSY spectra, facilitated complete assignment of the
fluorine spectrum. Simulation of that spectrum is also shown in
Figure 2. The smaller JFF(endo) for 15 as compared with that of
13 indicates a somewhat longer endo−endo F−F distance,
aspirator pressure in a flask connected to a cold trap (77 K)
with furan frozen on its walls. After methylene chloride was
added and the trap warmed to RT, the 19F spectrum of the
resulting solution showed only signals for furan adduct 6.
Repetition of the experiment with an empty cold trap gave a
presumably because the pyrrole “cap” on the fluorocarbon
moiety exerts less nonbonded repulsion on it than does the
anthracene “cap”. Consistent with this finding, the X-ray crystal
structure of 15 reveals F−F endo distances of 2.56 Å.
Figure 2. Experimental (A) and simulated (B) 19F NMR spectra (CDCl ) of N-benzylpyrrole adduct 15.
3
C
DOI: 10.1021/jo502456h
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
solution that showed a singlet at δ −99.1 in its 19F NMR
spectrum (−70 °C), a singlet that persisted at RT. Thus, alkene
heated together at 160 °C in a sealed tube. Furan adduct 6 was
obtained, but no 16 was present in the product (Scheme 13).
Therefore, 16 had been produced in the earlier pyrolysis
experiments solely via retro-Diels−Alder reaction of 15.
1
was shown to be an isolable and stable compound.
However, the yield of 1 was very low (<10%), and thick,
black tar remained behind in the reaction flask. In the hope that
dilution would improve yields by minimizing intermolecular
reactions, the pyrolysis was repeated in a similar fashion by
injecting a solution of 15 into a refluxing solvent. Of the variety
of inert, high-boiling point solvents tried, 1,2,4-trichloroben-
zene proved to be the best, affording 23−24% yields of furan
adduct 6.
Scheme 13
Analysis of the residual black tar revealed the presence of
compound 16, an isomer of the starting material with a lower
energy as a consequence of aromaticity and diminished ring
strain. Most of the starting adduct was being converted to 16
even in dilute solution. Attempted purification of 16 by
preparative GC yielded instead an isomer (18) with
considerable release of strain. This rearrangement probably
occurred via zwitterionic intermediate 17 (Scheme 11).
We speculated that the adduct of 1 with a pyrrole substituted
at positions 2 and 5 might undergo less isomerization upon
being heated than 15. Accordingly, N-benzyl-2,5-dimethylpyr-
16
role (20) was tried at RT as a trap for alkene 1 (Scheme 14).
Scheme 14
Scheme 11
Early in the reaction, 19F NMR indicated that adduct 21 was
the only product, but at higher conversion, peaks grew in for
the more stable isomer 22, in which the pyrrole ring was
substituted at position 3. Although it is possible that the 21 to
1
9
The F NMR spectrum of 18 at RT comprised four singlets
2
2 transformation proceeded intramolecularly by way of a
for vinyl fluorines, one singlet for the CF group distal to the
pyrrole ring, and an unresolved AB quartet for the proximal
CF . Recognizable as a quartet at 0 °C, this last resonance
2
[
2.2.2]propellane intermediate, retro-Diels−Alder reaction/
recombination seems more likely. In any event, it is striking
that the isomerization occurs at RT. Gas chromatography at
2
became a sharp singlet by 60 °C. Thus, VTNMR of 18 revealed
that rotation about either single bond or both nearest to the
pyrrole ring is sufficiently hindered to occur at an intermediate
rate on the NMR time scale at RT.
Formation of 16 might take place in either of two ways, via
intramolecular rearrangement of 15 or retro-Diels−Alder
reaction followed by electrophilic substitution on the pyrrole
ring; i.e., zwitterionic intermediate 19 could be formed via
either pathway (Scheme 12). To distinguish between the inter-
and intramolecular pathways, adduct 15 and excess furan were
2
50 °C transformed the mixture of 21 and 22 into an isomer
analogous to 18.
N-Methylpyrrole behaved like 20 under the same gentle
conditions, giving at first exclusively adduct 23 but eventually a
1
:1 mixture of 23 and 24. This was a surprising result, because
N-methylpyrrole is not very different in character from the N-
benzyl derivative. It is not known whether the intra- or
intermolecular pathway shown in Scheme 5 was followed by
adduct 23.
Scheme 12
It was now clear that the N-benzylpyrrole adduct 15 was the
precursor of choice for 1, but the problem of low yield
remained. To forestall the recombination of 1 and pyrrole 14
that was occurring at 140 °C even in rather dilute solution, flash
vacuum pyrolysis (FVP) was conducted with adduct 15. The
apparatus consisted of a flask joined to a Pyrex pyrolysis tube
connected in turn to a pair of U-traps, the first at 0 °C and the
second at 77 K. With the pyrolysis tube at 300 °C, compound
D
DOI: 10.1021/jo502456h
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
5 was sublimed from the flask at 30 mTorr. N-Benzylpyrrole
Article
1
structure (C ) was a reasonable alternative (Figure 3). The fact
2
v
that the 19F NMR spectrum of 1 was a narrow singlet even at
collected uncontaminated with fluorine-containing material in
the 0 °C trap, and alkene 1 appeared in the 77 K trap in 85%
yield as determined by NMR. Under these conditions, a small
1
7
amount of 25, the dimer of tetrafluoroallene, was formed by
electrocyclic ring opening of 1. The optimal temperature for
maximizing the yield of 1 and minimizing that of 25 was found
to be 275 °C.
Figure 3. Possible geometries for the skeleton of alkene 1.
−
70 °C was not inconsistent ith C2v geometry, as pyramidal
inversion of this structure should be very rapid. We believed
that the combination of ring strain and the effect of eight
fluorines might force the bridgehead carbons to pyramidalize.
Insofar as fluorine’s electronegativity results in enhanced p
18
character in the C−F bonds (Bent’s Rule), transmission of
Finally, a source of pure 1 was available, and it was already
apparent that the contrasts between the fluorocarbon and its
very labile hydrocarbon counterpart 2 would be profound.
When 2 was prepared by pyrolysis at 190 °C of tosylhydrazone
anion 26, ring-opened 1,2-dimethylenecyclobutane (27) was
obtained in yields as high as 75%, and 2 reacted with 27 to give
that effect to the bridgehead carbons would favor a change from
2
3
sp to sp hybridization.
Infrared absorption is forbidden for the CC stretching
vibration in D2h but allowed in C symmetry. Whereas the
2v
−1
Raman spectrum of 1 displayed a band a 1621 cm for the
CC stretching mode, there was no band at this position in
the infrared spectrum. The conclusion from these observations
that the molecule is flat with D2h symmetry was confirmed by
an electron diffraction study coupled with hybrid density
functional calculations conducted by Kenneth Hedberg at
1a
adduct 28 even at −30 °C (Scheme 15). No reaction took
place between 1 and its ring-opened isomer 25 at 80 °C.
Scheme 15
5
Oregon State University. It is bow-tie-shaped, with r(C −C )
1
4
=
1.376(14) Å, r(C −C ) = 1.530(3) Å, and r(C −C ) =
1
2
2
3
1
.627(5) Å.
The electron deficiency of 1 is reflected in its gas phase
electron affinity, 1.3 ± 0.2 eV. The electron-accepting LUMO is
depicted in Figure 4. For the radical anion 1a, the proton
Hydrocarbon 2 dimerizes to give 29 (a transient intermediate
⧧
1
that ring opens) with a ΔH of only 11.5 kcal/mol, but its
fluorinated counterpart 1 does not form dimer 30 at any
temperature. We assumed that nonbonded repulsions between
fluorines in 30 were primarily responsible for the contrast, but a
6
thorough analysis by Borden based on density functional
theory has revealed a somewhat greater contribution to the
calculated difference of 44.6 kcal/mol in ΔH for the two
dimerizations: the π bond in 1 was found to be 16 kcal/mol
stronger than that in 2. Approximately 25% of this difference
was ascribed to stabilization of the π bond of 1 by donation of
its electrons into C−F σ* orbitals and 75% to destabilization of
the broken π bond. The latter effect arises because eclipsing a
pyramidalized carbon radical with a C−X bond is unfavorable
for X = F relative to X = H.
19
Figure 4. LUMO of alkene 1 (B3LYP/6-311G**+ level of theory).
affinity is <348 ± 2 kcal/mol, as it is less basic than the acetate
ion. These results are in accord with M06-2X/maug-cc-
pVT(+d)Z//M06-2X/aug-cc-pVDZ predictions of EA(1) =
1.67 eV and PA(1a) = 329.4 kcal/mol. Together with the
ionization potential of the hydrogen atom (313.6 kcal/mol)
and the computed electron and proton affinities, these findings
suggest that D[C−H] = 54 kcal/mol for the new bond in 31
(Scheme 16).
Fluorocarbon 1 is a liquid at RT with a narrow liquid range:
1
9
mp 19−20 °C, bp 52.5 °C. Its F NMR spectrum (CDCl )
3
1
3
consists of a sharp singlet at δ −99.1, and the C spectrum
CDCl ) comprises a signal at δ 121.6 for the four CF carbons
(
3
2
and one at δ 175.8 for the bridgehead carbons. Hydrocarbon 2
is a flat molecule with D2h symmetry, but for 1, a roof-shaped
E
DOI: 10.1021/jo502456h
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 16
benzylpyrrole gave an isomeric adduct (16) with a low yield
of 1. Rearrangement of the adducts of 1 with other pyrroles
occurred even at RT. It was found that isomerization of 15 took
place via retro-Diels−Alder reaction followed by recombination,
and flash vacuum pyrolysis made it possible to preclude the
latter process. This technique made alkene 1 available in high
yield and purity.
The He(I) photoelectron (PE) spectrum of 1 shows two
broad features between 12 and 14.5 eV. The first feature is
composed of a shoulder at Iv,1 = 12.2 eV belonging to a broad
The alkene is very stable toward reaction with itself, in sharp
contrast to its extremely labile parent hydrocarbon. Electron
diffraction measurements and hybrid density functional
Gaussian-like band superimposed upon a narrow band at Iv,2
=
5
calculations confirmed the indication from vibrational spectra
12.7 eV. The second broad feature centered at 13.6 eV
that it possesses D symmetry. Gas phase studies have revealed
2
h
originates from at least three ionization processes. With regard
to the first feature, the Gaussian-like band is typical for an
ionization process from σ orbitals, whereas the second,
narrower band is often found in ionization processes from π
orbitals. The empirical assignment of the second band to the
ionization from the π orbital of 1 is further underlined by the
its photoelectron spectrum, its electron affinity, and an upper
limit for the proton affinity of its radical anion.
EXPERIMENTAL SECTION
■
The 19F NMR spectra were recorded at 282.2 or 470.3 MHz on a 300
9
20
19
first PE band (π) of perfluoro-Δ -octalin (32) at I = 13.0 eV.
The slightly higher value of the π ionization energy found for
or 500 MHz spectrometer. F chemical shifts are reported on the δ
v
scale (parts per million from internal standard trichlorofluoromethane,
1
upfield negative). H NMR spectra were recorded at 300 or 500 MHz
3
2 is due to the stronger inductive effect of two octafluoro
with tetramethylsilane as an internal standard. 13C NMR spectra were
recorded at 75.4 MHz on the 300 MHz spectrometer. Melting points
were measured in open capillary tubes and were not corrected.
Ultrasonic irradiation was performed in a Bronsicator Ultrasonic B-22-
bridges of 32 as compared to that of the two tetrafluoro bridges
of 1. The PE spectra of 1 and 32 appear in the Supporting
Information.
4
cleaning bath. Preparative gas chromatography was performed with a
1
thermal conductivity detector. The column was 10 ft × / in., 10%
8
OV 101 on Chromosorb-P acid washed, and the standard program was
as follows: detection at 200 °C, injection at 200 °C, column
temperature as noted in text. All analytical GC was conducted using a
25 m methyl silicone capillary column and flame ionization detector.
The standard program was as follows: carrier pressure, 25 psi;
detector, 200 °C; injector, 150 °C; temp , 50 °C; temp , 250 °C;
To assign the first two ionization processes of 1, we make use
of Koopmans’ approximation, which correlates the measured
vertical ionization energies with the calculated negative value of
21
1
2
time , 5 min; 50 °C and then increasing at a rate of 15 °C/min to 250
1
the orbital energies of the ground state (Iv,j = −ε ). This
j
°
C. Photochemistry was conducted using a 450 W Canrad-Hanovia
supports the empirical assignment of σ above π for 1. Both
MOs are shown in Figure 5, the HOMO encompassing the
strained carbon skeleton. Their ionization energies lie far above
those of ethylene (10.6 eV) and tetrafluoroethylene (10.52 eV)
that are nearly identical as a consequence of the perfluoro
medium-pressure mercury lamp, with filters used as noted in the text.
Caution: Eye protection is essential when using this intense source of UV
radiation. For photoreaction mixtures that needed to remain at room
temperature, cooling was achieved with a stream of tap water directed
onto the outside of the reaction vessel. The runoff was collected in a
plastic tub beneath the photochemical apparatus.
2
2
effect.
Solvents used in this work were reagent grade. Acetonitrile was
distilled from calcium hydride and tetrahydrofuran from potassium/
benzophenone. Zinc dust was purified using the method of Shriner
SUMMARY
■
Strained alkene 1 has been generated at RT by reduction of
dibromide (5), and under milder conditions diiodide (11)
precursors, and trapped in situ. Photolysis of the diiodide
permitted the first direct observation of the alkene. Pyrolysis in
solution of the Diels−Alder adduct 15 of 1 with N-
23
and Neumann. The zinc/copper couple was generated by first
grinding together a mixture of finely powdered CuSO ·5H O (2 g)
4
2
and zinc dust (2 g) in a 10 mL vial. The vial was then covered with a
rubber septum that was pierced by a needle, and 0.5 mL of DMF was
added while the contents were being stirred to catalyze the reduction
1
9
Figure 5. HOMO (left) and NHOMO (right) orbitals of alkene 1 (B3LYP/6-311G**+ level of theory).
F
DOI: 10.1021/jo502456h
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
of copper. An exothermic reaction ensued, and once the vial returned
to room temperature, the resulting black Zn/Cu couple was ground to
a fine wet powder. The Zn/Cu couple was then heated under vacuum
to an aqueous KOH trap to remove any fluorine from the effluent gas
before being vented into the hood. (Rapid mechanical stirring was
essential to ensure an optimal yield for this reaction. The PTFE
bearing does not require any grease for lubrication, thus ensuring that
proper stirring can be maintained for the duration of the reaction.)
(
150 °C, 30 mTorr) to remove the DMF and stored under nitrogen.
The X-ray crystal structure data for 15 were collected by J.P.J..
24
Using the Olex2 software package, the structure was determined
with SUPERFLIP using charge flipping and refined with the
ShelXL refinement package using least-squares minimization. R.G.
obtained the He(I) photoelectron spectra of 1 and 32, which were
recorded at room temperature. The calibration was done with Ar and
Xe. A resolution of ∼30 meV was achieved with the P Ar line.
Other gas phase experiments were conducted by S.R.K. with a dual cell
Fourier transform mass spectrometer that was described previously.
Low-energy electron ionization (1.6 eV) or the reduction of 1 by NO−
which was generated from nitrous oxide) was used to produce 1a.
The solution was cooled to −78 °C and flushed with N for 15 min. A
2
25
mixture of 30% fluorine in helium was then bubbled (5−6 bubbles per
26
19
second) through the solution until F NMR indicated that reaction
was complete, ∼4 h. This reaction proceeded via the partially
fluorinated intermediate 1,4-dibromohexafluorobicyclo[2.2.0]hex-2-
2
19
3/2
ene, which could also be seen by F NMR. A distinct pale yellow
color appeared upon completion of the fluorination. Nitrogen was
then bubbled through the stirred solution for 20 min to purge it of any
remaining fluorine gas. The trichlorofluoromethane was removed by
distillation for reuse (it was washed with saturated Na S O /H O to
2
7
(
2
2
3
2
This ion was isolated and then transferred to the second reaction
region in the instrument where it was cooled with a pulse of argon,
remove bromine, dried, and recycled), and the residual thick maroon
liquid was vacuum transferred at 60 mTorr to give a waxy colorless
2
8
19
reisolated, and allowed to react with selected reference reagents. In
some cases, reverse processes were also examined by allowing an anion
of interest generated by electron ionization to react with 1. The proton
and electron affinity data given in the Results and Discussion were
obtained in this way (see the Supporting Information for further
details). Computations on 1, 1a, and 31 were conducted by S.R.K.
solid (25.0 g, 51.7% yield for two steps). F NMR (CDCl ): δ
3
−103.8, −113.4 (AX q, J = 261 Hz, 8F). The chemical shifts match
2
those in the literature, but the reported J value is incorrect.
11-Oxa-7,7,8,8,9,9,10,10-octafluorotetracyclo[4.2.2.12,5.01,6]-
undec-3-ene (6). A mixture of 1,4-dibromooctafluorobicyclo[2.2.0]-
2
hexane (5, 0.398 g, 1.04 mmol), ether (3 mL), and furan (3 mL) was
contained in a nitrogen-flushed 25 mL round-bottom flask. The flask
was capped with a septum, and connection to a nitrogen bubbler was
maintained via a syringe. After the mixture was cooled to −84 °C in an
2
9
using Gaussian 09. M06-2X geometry optimizations and vibrational
3
0,31
frequencies were calculated with the aug-cc-pVDZ basis set,
and
32
M06-2X/maug-cc-pVT(+d)Z single-point energy determinations
were subsequently conducted. The resulting proton and electron
affinities and bond energies are given at 298 K.
ethyl acetate/liquid N bath, a solution of methyllithium (2.25 mmol)
2
in 4 mL of ether, also cooled to −84 °C, was added dropwise with
stirring via a gastight syringe. The bath was allowed to warm slowly to
RT. Product was washed with water, and the aqueous phase was
extracted with ether. The combined ether phase was dried over
Hexafluoro Dewar Benzene (3). Caution: Hexafluoro Dewar
benzene is explosive and has been known to detonate capriciously.
Protective clothing (heavy jacket, leather gloves, and face shield) are
recommended for handling it. A cylindrical quartz vessel (5 L) fitted with
a PTFE valve was evacuated (30 mTorr). The vessel was then sealed
and removed from the vacuum line, and the bottom of the vessel was
placed in an ice bath. A rubber septum was attached to the inlet valve
extension, the valve opened, and hexafluorobenzene (6.0 mL, 9.6 g, 52
mmol) injected. The valve was closed and the vessel then irradiated at
MgSO , filtered, and distilled at atmospheric pressure, leaving a pale
4
yellow oil. This oil was dissolved in a small amount of ethyl acetate and
flushed through a 6 in. column of silica gel. Slow removal of the
solvent deposited white crystals of 6. Alternatively, the product could
be purified by sublimation (10 mTorr, 75 °C bath, 0 °C coldfinger).
1
9
Mp: 124.5−126 °C. Yield: 0.234 g (77%). F NMR (CDCl ): δ
3
2
54 nm for 4−6 days in a cylindrical photoreactor lined with ten 25 W
−113.2, −114.8 (AB q, J = 224 Hz, 4F), −115.3, −121.7 (AB q, J =
1
germicidal lamps. The resulting Dewar benzene was vacuum
228 Hz, 4F), J
= 48 Hz. H NMR (CDCl ): δ 6.4 (d, J = 7.5 Hz,
endo
3
HF
transferred into a 100 mL U-trap containing 10 mL of n-pentane at
vinyl, 2H), 5.3 (s, bridgehead, 2H). IR: 1336, 1315, 1290, 1262, 1228,
−
1
+
7
7 K. This trap was then surrounded by a wire cage and allowed to
1211, 1186, 1170, 1115, 1096, 1014 cm . MS: m/z 292 (M ), 68
+
warm slowly to room temperature. The resulting pentane solution was
(base, C H O ). HRMS calcd for C H F O, 292.0134; found,
4
4
10
4 8
checked by 19F NMR to determine yield by integration. Yields varied
292.0124.
19
from 5.8 to 7.6 g (from 60 to 80%). The product’s F NMR spectrum
matched the literature values.
1,4-Dimethylhexafluorobicyclo[2.2.0]hex-2-ene (7) and 1-
Bromo-4-methylhexafluorobicyclo[2.2.0]hex-2-ene (8). A sol-
ution of 1,4-dibromooctafluorobicyclo[2.2.0]hexane (5, 0.398 g, 1.04
mmol) in ether (6 mL) contained in a nitrogen-flushed, septum-
capped 25 mL round-bottom flask was cooled to −84 °C with an ethyl
9
2
1
,4-Dibromotetrafluorobicyclo[2.2.0]hexa-2,5-diene (4).
Caution: Compound 4 has been known to detonate, so this reaction
was conducted behind a blast shield. Protective clothing (heavy jacket,
leather gloves, and face shield) are recommended when handling it.
Hexafluoro Dewar benzene (23.4 g, 0.126 mol) in 150 mL of n-
pentane was added dropwise to a stirred suspension of aluminum
bromide (50.5 g, 0.189 mol) in 150 mL of n-pentane at 0 °C over a
period of 1 h, and stirring was continued for an additional 1 h at 0 °C.
The mixture was poured over 200 mL of crushed ice in a separatory
funnel, and the pentane layer was gently washed with three 100 mL
portions of cold water until the orange color faded. After the mixture
had dried over magnesium sulfate, most of the pentane was removed
with a rotary evaporator behind a blast shield in an ice bath; the
residue was moved into a hood and evaporated using a water aspirator
acetate/liquid N bath. A solution of methyllithium (2.25 mmol) in 4
2
mL of ether, similarly cooled, was injected dropwise with stirring. The
mixture was allowed to warm slowly to RT and then washed with
water. The aqueous phase was extracted with ether, and the combined
ether phase was dried over MgSO , filtered, and evaporated to leave a
4
mixture of 7 and 8. 19F NMR (CDCl ) of 7: δ −112.0, −125.8 (AB q, J
3
1
= 220 Hz, 4F), −123.6 (s, vinyl, 2F). H NMR (CDCl ): δ 1.30 (s,
3
+
+
+
6H). MS: m/z 216 (M ), 201 (C F Me ), 117 (base, C F ). HRMS
6
6
5 3
19
calcd for C H F , 216.0374; found, 216.0370. F NMR (CDCl ) of 8:
8
6
6
3
δ −110.5, −114.8 (AB q, J = 208 Hz, 4F), −114.0, −122.2 (AX q, J =
1
205 Hz, 4F), −119.7 (s, vinyl, 1F), −123.5 (s, vinyl, 1F). H NMR
+
(
20 Torr) at 0 °C until bubbling ceased. Trichlorofluoromethane (500
(CDCl ): δ 1.42 (s, 3H). MS: m/z 280/282 (M ), 201 (base,
3
19
+
mL) was then added as the solvent for the next step. F NMR
CDCl ): δ −131.4 (s, 4F).
C F Me ). HRMS calcd for C H F Br, 279.9322; found, 279.9330.
6
6
7
3 6
(
Product composition was dependent upon the rate of addition of
methyllithium, as quick injection gave only 7 and polymeric products.
Gas chromatographic separation of 7 and 8 was not successful.
Debromination of 1,4-Dibromooctafluorobicyclo[2.2.0]-
hexane (5) with Zinc. Ether (3 mL), furan (3 mL), 5 (0.190 g,
0.493 mmol), and activated zinc (0.2 g, 3 mmol) were placed in a 25
mL round-bottom flask. Connected to a nitrogen bubbler, the flask
was rocked for 2 h with a modified Kugelrohr motor in an ultrasound
bath. After filtration, the product solution was washed with water, and
3
2
1
,4-Dibromooctafluorobicyclo[2.2.0]hexane (5). Caution:
Elemental fluorine, 30% in helium, is an extremely aggressive reagent
that must be handled with care in a well-ventilated hood. The
trichlorofluoromethane solution described above was placed in a 1 L
three-neck flask. The middle neck was fit with a mechanical stirrer
shaft mounted in a PTFE bearing. In another neck, a Pyrex tube that
reached almost to the bottom of the flask was used to admit fluorine-
containing gas. A takeoff adapter in the remaining neck was connected
G
DOI: 10.1021/jo502456h
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
the aqueous phase was extracted with ether. The combined ether
a flame-dried 10 mL flask were added diiodo compound 11 (30 mg,
6.3 × 10 mol), N-benzylpyrrole 14 (30 mg, 1.9 × 10 mol, 3 equiv),
4 mL of dry acetonitrile, and 1 mL of mercury. This mixture was
−5
−4
solution was dried with MgSO , filtered, and distilled to give a
4
colorless oil containing 11-oxa-7,7,8,8,9,9,10,10-octafluorotetracyclo-
2
,5 1,6
[
4.2.2.1 .0 ]undec-3-ene (6, 82 mg, 57%) and 11-oxa-3,3,4,4,6,6,7-
placed in a water-cooled ultrasonic bath for 5 h. The resulting yellow
2
,5 2,7
19
19
heptafluorotetracyclo[6.2.1.0. .0 ]undec-3-ene (10). F NMR
CDCl ) of 10: δ −106.7, −119.4 (AX q, J = 222 Hz, 2F), −110.5,
solution contained only the desired adduct as determined by
F
(
−
2
NMR. Workup involved adding 10 mL of water and extracting with 3
× 3 mL portions of pentane. The combined pentane layers were dried
3
116.9 (AB q, J = 219 Hz, 2F), −113.7, −114.5 (AB q, J = 228 Hz,
+
+
F), −182.7 (bridgehead). MS: m/z 274 (M , C H F O ), 223 (base,
over MgSO and evaporated to 1 mL under reduced pressure. A white
10
5
7
4
+
+
19
C H F O ), 68 (C H O ).
solid (15 mg, 62%) precipitated upon standing at 0 °C overnight.
F
9
4
5
4
4
2
1
,4-Diiodooctafluorobicyclo[2.2.0]hexane (11). Compound 5
NMR (CDCl ): δ −111.2, −114.2 (AB q, J = 227 Hz, 4F), −112.5,
3
1
(
4.0 g, 8.4 mmol) was placed in a cylindrical 50 mL Pyrex tube along
−118.3 (AB q, J = 217 Hz, 4F), J
= 52 Hz, J
= 7 Hz. H
F(exo)H
endo
with 45 mL of an acetonitrile/ether mixture [3:1 (v/v)]. Potassium
iodide (3.0 g, 18 mmol) was also added to this solution, producing a
light yellow suspension. The mixture was stirred magnetically and
cooled with a stream of water (18 °C) under nitrogen while being
irradiated by the 450 W Canrad-Hanovia lamp until reaction was
complete as determined by 19F NMR (10−15 h). At this point, the
mixture contained mostly the desired diiodo compound and was dark
brown in color. Workup involved washing with 3 × 50 mL portions of
water to remove the acetonitrile, and back-washing the aqueous phase
with 2 × 20 mL portions of ether. The combined ether layers were
washed with a saturated Na S O solution, then dried over MgSO ,
NMR (CDCl ): δ 7.26 (m, 5H), 6.35 (d, J = 7 Hz, vinyl, 2H), 4.31 (s,
3
bridgehead, 2H), 3.49 (s, benzylic, 2H). (Proton/fluorine assignments
were verified by an HMQC experiment.) 13C NMR (CDCl ): δ 137.5,
3
13
19
134.7, 129.4, 128.5, 127.8, 66.0, 49.5. C NMR (CDCl , F-
3
+
+
decoupled): δ 115.4, 114.1, 59.9. MS: m/e 381 (M ), 91 (C H ). Mp:
7
7
99.0−100.0 °C dec at 135−140 °C. Anal. Calcd for C H NF : C,
1
7
11
8
53.55; H, 2.91; N, 3.67. Found: C, 53.51; H, 2.90; N, 3.53. IR (KBr):
2860, 1499, 1459, 1331, 1313, 1218, 1166, 1093, 998, 914, 892, 864,
−1
827, 791, 768, 734, 702 cm .
11-Aza-11-benzyl-7,7,8,8,9,9,10,10-octafluoroquadricyclo-
[4.2.2.12 .0 ]undec-3-ene (15) from Dibromo Compound 5.
,5
1,6
2
2
3
4
and distilled to ∼25% of the original volume. The remaining solution
was vacuum transferred at 30 mTorr and then blown down to an oil
using a stream of nitrogen. The oil was purified by preparative GC
To a flame-dried 25 mL flask were added dibromo compound 5 (0.83
−3
g, 2.2 × 10 mol), 15 mL of dry acetonitrile, N-benzylpyrrole (0.35 g,
−3
2.2 × 10 mol, 1.0 equiv), and 1.0 g of zinc/copper couple. This
(
4
(
column temperature of 160 °C) to give pure 11 (2.2 g, 4.6 mmol,
mixture was placed in a water-cooled ultrasonic bath and rocked using
5% yield). The product’s 19F NMR spectrum and melting point
19
a modified Kugelrohr motor for 4 h. The F NMR spectrum of the
23
112.5−113.5 °C) matched the literature values. The UV−vis
resulting yellow solution showed complete conversion to desired
adduct 15. The acetonitrile layer was combined with 25 mL of water,
and this solution was then extracted with 3 × 10 mL portions of ether.
19
spectrum of this compound showed a λ at 276 nm (ε ∼ 602).
F
max
NMR (CDCl ): δ −90.8, −112.1 (AB q, J = 220 Hz, 8F). Analytical
3
GC retention time: 5.7 min.
The combined ether layers were dried over MgSO , and the ether was
4
Photochemical Generation of Alkene 1. To a flame-dried 10
removed under reduced pressure, leaving a brown sticky solid.
Recrystallization from ether at 0 °C gave large colorless crystals (0.60
g, 73% yield). An X-ray crystal structure was obtained from this crop of
crystals that confirmed the proposed structure. Crystallographic data
−5
mL quartz tube were added diiodide 11 (19 mg, 4.0 × 10 mol), 28
−4
μL of furan (26 mg, 3.9 × 10 mol, 9.7 equiv), 100 mg of copper
−3
powder (1.6 × 10 mol, 40 equiv, to intercept iodine), and 5 mL of
dry acetonitrile. This suspension was externally cooled with water,
magnetically stirred under nitrogen, and irradiated for 6 h using a
Vycor-filtered 450 W Canrad-Hanovia lamp. The only signals present
in the 19F NMR spectrum were those of furan adduct 6.
for C17
H11NF : M = 381.27, monoclinic, P2 /c, a = 12.154(3) Å, b =
8
r
1
3
9.965(3) Å, c = 12.740(3) Å, β = 101.872(19)°, V = 1510.1(7) Å , Z =
−1
3
4, T = 298 K, μ(Mo Kα) = 0.168 mm , Dcalc = 1.677 g/cm , 3622
reflections measured (5.332° ≤ 2θ ≤ 54.992°), 3465 unique (Rint
0.0481, Rsigma = 0.0938), which were used in all calculations. The final
was 0.0444 [I > 2σ(I)], and wR was 0.1583 (all data). CCDC =
=
To a flame-dried 10 mL quartz tube were added diiodide 11 (20
−5
mg, 4.2 × 10 mol), 2 mL of dry methylene chloride, and copper
R
1
2
−2
powder (900 mg, 1.4 × 10 mol, 330 equiv). This suspension was
irradiated as described above using a Corex-filtered 450 W Canrad-
Hanovia lamp. After 2 h, a singlet appeared at −100 ppm (not
referenced) representing alkene 1. Further irradiation continued to
yield more alkene, and then addition of furan to this reaction mixture
produced 19F NMR peaks corresponding to furan adduct 6.
1016151. The remaining brown ether layer was reduced in volume and
chilled to produce an additional 100 mg of a brown solid; its purity
was lower than that of the original product. When a very high purity is
desired, this compound can be sublimed at 75 °C under vacuum (30
mTorr) to give a white powder.
GC/MS analysis of this compound also showed a peak that eluted
7
,7,8,8,9,9,10,10-Octafluoro-2-methoxydibenzo[c,m]quadri-
before adduct 15 corresponding to olefin 1, formed via retro-Diels−
2,5 1,6
+
cyclo[4.2.2.2 .0 ]dodeca-3,11-diene (13). To a flame-dried 20
Alder reaction while in the injection port. MS: m/z 205 (C F ), 174
6
7
−4
+
+
+
+
+
mL flask were added diiodide 11 (230 mg, 4.8 × 10 mol), 9-
(C F ), 155 (C F , base), 124 (C F ), 93 (C F ), 69 (CF ).
5 6 5 5 4 4 3 3 3
−3
methoxyanthracene 12 (500 mg, 1.44 × 10 mol, 3.0 equiv), 4 mL of
dry acetonitrile, and 2 mL of mercury. This mixture was placed under
nitrogen in the ultrasonic bath, which was cooled with water passing
through a U-tube immersed in the bath. After 8 h, the 19F NMR
spectrum of the resulting yellow solution contained only peaks
corresponding to adduct 13. The reaction mixture was extracted with 5
Generation of Alkene 1 via retro-Diels−Alder Reaction of N-
Benzylpyrrole Adduct 15. To a flame-dried 10 mL round-bottom
flask equipped with a reflux condenser and a septum-covered side arm
was added 5 mL of 1,2,4-trichlorobenzene. The condenser was
connected to a U-trap, which was connected in turn to a water
aspirator. The trap was cooled to −196 °C, and 1 mL of furan was
added. With water running through the condenser, the pressure in the
system was reduced (30 Torr), and an oil bath was used to heat the
solvent to 150 °C. A solution of N-benzylpyrrole adduct 15 (0.10 g,
×
10 mL portions of benzene. The combined benzene layers were
then dried over MgSO , reduced in volume to 2 mL under reduced
4
pressure, and placed on a silica gel column. Elution with pentane gave
a yellow solid. At this point, adduct 13 was still contaminated with 9-
methoxyanthracene. Recrystallization from a 1:1 pentane/ether
mixture over the course of a week gave the desired product as a
colorless crystalline solid. An X-ray crystal structure was obtained from
−4
2.6 × 10 mol) in 3 mL of 1,2,4-trichlorobenzene was then injected
into the side arm in 0.5 mL portions over 10 min. The heating was
continued for an additional 30 min, at which time the cold trap was
disconnected from the apparatus and allowed to warm to room
these crystals, confirming the proposed structure. 19F NMR (CDCl ):
temperature. Hexafluorobenzene (20 μL, 1.7 × 10 mol) was then
−4
3
δ −109.3, −113.1 (AB q, J = 220 Hz, 4F), −109.6, −113.4 (AB q, J =
20 Hz, 4F), J
.72 (s, 1H), 3.99 (s, 3H). MS: m/z 253, 208 (C H O ), 193 (base,
added as an internal standard. Integration of the 19F NMR spectrum of
this solution determined the ratio of furan adduct 6 to hexafluor-
obenzene, thus revealing that the amount of alkene 1 generated was 13
mg (23% yield).
1
2
4
= 68 Hz. H NMR (CDCl ): δ 7.5−7.1 (m, 8H),
endo
3
+
15
12
+
C H O ), 165, 82. Mp: 166.0−167.0 °C.
1
4
1
9
1-Aza-11-benzyl-7,7,8,8,9,9,10,10-octafluoroquadricyclo-
Sealed Tube Reaction of N-Benzylpyrrole Adduct 15 with
Furan. To a flame-dried Pyrex tube (15 cm × 1 cm, 4 mm inside
2,5 1,6
[
4.2.2.1 .0 ]undec-3-ene (15) from Diiodo Compound 11. To
H
DOI: 10.1021/jo502456h
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
diameter) were added N-benzylpyrrole adduct 15 (70 mg, 1.8 × 10−4
mol), 3 mL of 1,2,4-trichlorobenzene, and 0.5 mL of furan. The tube
was cooled to −196 °C and sealed with a flame under reduced
pressure (30 mTorr). It was then placed in a 160 °C oil bath for 30
min. After it was cooled again to −196 °C, the top was broken off and
pressure. In a sublimer at 2.0 Torr and 55−60 °C, the remaining
yellow oil gave 157 mg of a white solid (38%). This solid was still a
1
9
mixture of the two isomers 23 and 24 as determined by F NMR, and
careful recrystallization from ether did nothing to enrich the mixture.
Compound 23. 19F NMR (CDCl ): δ −111.3, −114.4 (AB q, J = 226
3
−4
1
the solid was allowed to thaw. Hexafluorobenzene (30 μL, 2.6 × 10
Hz, 4F), −112.7, −119.9 (AB q, J = 226 Hz, 4F), J
= 53 Hz. H
endo
mol) was added as an internal standard. 19F NMR of the solution
showed only furan adduct 6 and hexafluorobenzene. Integration
revealed that 45 mg of 6 had been formed (85% yield).
NMR (CDCl ): δ 6.27 (d, J
= 6.0 Hz, vinyl, 2H), 4.26 (s,
3
F(exo)H
1
9
bridgehead, 2H), 2.10 (s, methyl, 3H). Compound 24. F NMR
(CDCl ): δ −109.7, −112.8 (AB q, J = 220 Hz, 4F), −112.3 (s, 4F).
3
1
-Benzyl-2-(2,2,3,3,5,5,6,6-octafluorobicyclo[2.2.0]hex-1-yl)-
Elemental analysis was performed on the mixture of isomers. Anal.
Calcd for C H F N: C, 43.28; H, 2.30. Found: C, 43.60; H, 2.24.
1
H-pyrrole (16). This compound was generated as a thick black oil
11
7 8
19
during the pyrolysis of 15 and was very difficult to purify. F NMR
CDCl ): δ −109.7, −112.8 (AB q, J = 226 Hz, 4F), −112.0 (s, 4F).
Octafluorobicyclo[2.2.0]hex-1(4)-ene (1). Adduct 15 (0.32 g,
0.83 mmol) was placed in a flame-dried 5 mL round-bottom flask,
which was connected to a pyrolysis apparatus consisting of a 15 mm
outside diameter Pyrex tube enclosed in a 55 cm-long tube furnace.
The outlet of the tube was connected to an ice-cold U-trap followed by
(
3
1
-Benzyl-2-(1-difluoromethylene-2,2,3,3,5,5-hexafluoro-
pent-4-enyl)-1H-pyrrole (18). A chloroform solution of impure
compound 16 was injected into a preparative gas chromatograph (250
°
C), and the peak corresponding to compound 18 was collected as a
a liquid N -cooled U-trap. The furnace was heated to 275 °C, and then
2
19
colorless liquid. F NMR (CDCl ): δ −68.4, −70.5, −72.2, −73.4
3
15 was allowed to sublime slowly into the hot zone under reduced
pressure (60 mTorr). N-Benzylpyrrole (14) was intercepted in the
(
2
(
5
vinyl Fs), −108.5 (CF , 2F), −111.8 (unresolved AB q, CF , 2F, J ∼
2
2
1
65 Hz). By 60 °C, this resonance became a sharp singlet. H NMR
CDCl ): δ 7.33 (m, 3H), 7.10 (m, 2H), 6.72 (s, 1H), 6.23 (bs, 2H),
.00 (s, 2H), 4.70 (dt, J = 22, 13 Hz, 1H). MS: m/e 381 (M ), 113
ice−water trap, and alkene 1 was collected in the liquid N trap. The
2
3
system was filled with N , and the liquid N trap was allowed to warm
2
2
+
to RT. Alkene 1 was obtained as a colorless liquid (0.154 g, 82%). Mp:
+
+
+
19
13
(
C F H , base), 91 (C H ), 65 (C H ). IR (neat): 1752 (CC),
19−20 °C. Bp: 52.5 °C. F NMR (CDCl ): δ −99.1 (s). C NMR
3
4
7
7
5
5
3
1
9
2
720 (CC), 1474, 1364, 1333, 1301, 1207, 1128, 1094, 1039, 977,
(CDCl ): δ 175.8 (CC), 121.6 (CF ). IR (vapor) 1312, 1234, 1184,
886, 827 cm . MS: m/z 205 (C F ), 174 (C F ), 155 (base, C F ),
6 7 5 6 5 5
3
2
−1
−1
+
+
+
44, 893, 865, 803, 720 cm . Anal. Calcd for C H F N: C, 53.55; H,
17
11
8
+
+
+
.91. Found: C, 53.34; H, 2.92.
1-Aza-11-benzyl-7,7,8,8,9,9,10,10-octafluoro-2,5-dimethyl-
124 (C F ), 93 (C F ), 69 (CF ). HRMS calcd for C F , 223.9880;
4 4 3 3 3 6 8
1
found, 223.9872. It can be stored in a refrigerator for long periods of
time without decomposing.
2
,5 1,6
quadricyclo[4.2.2.1 .0 ]undec-3-ene (21), 1-Benzyl-2,5-di-
methyl-3-(2,2,3,3,5,5,6,6-octafluorobicyclo[2.2.0]hex-1-yl)-1H-
pyrrole (22), and 1-Benzyl-3-(1-difluoromethylene-2,2,3,3,5,5-
hexafluoropent-4-enyl)-2,5-dimethyl-1H-pyrrole. To a flame-
dried 10 mL flask were added dibromo compound 5 (200 mg, 5.2
ASSOCIATED CONTENT
Supporting Information
■
*
S
−4
−3
×
10 mol), 2,5-dimethyl-N-benzylpyrrole (20, 0.40 g, 2.2 × 10
NMR and PE spectra, X-ray crystallographic information,
computational data, and gas phase experimental data. This
material is available free of charge via the Internet at http://
pubs.acs.org. CCDC 1016151 contains the supplementary
mol, 4 equiv), 4 mL of dry acetonitrile, and 0.5 g of activated zinc. This
mixture was placed in a water-cooled ultrasonic bath and rocked using
a modified Kugelrohr motor. A 19F NMR spectrum taken within the
first 2 h showed only starting material and compound 21. The 19
F
crystallographic data for C H N Cl (15). These data can be
NMR spectrum of the resulting bright yellow solution after 15 h
showed an approximately 50:50 mixture of the desired Diels−Alder
adduct (21) and the rearranged 3-substituted pyrrole (22), but the
spectral resolution was poor because of the number of overlapping
peaks. This solution was added to 15 mL of water, and the aqueous
layer was extracted with 3 × 10 mL of pentane. The combined pentane
17 12
3
obtained free of charge via http://www.ccdc.cam.ac.uk/conts/
retrieving.html or from the Cambridge Crystallographic Data
Centre, 12, Union Road, Cambridge CB2 1EZ, U.K., fax (+44)
1223-336-033, or e-mail deposit@ccdc.cam.ac.uk.
portions were dried over MgSO , and the solvent was removed under
4
AUTHOR INFORMATION
Corresponding Author
*E-mail: david.m.lemal@dartmouth.edu.
■
reduced pressure to give a brown oil. Upon attempted purification by
preparative gas chromatography at 250 °C, these compounds
isomerized to a 3-substituted pyrrole analogous to 18 as the only
isolable product. Compound 21. 19F NMR (CDCl ): δ −110.0,
3
Notes
−
113.6 (AB q, J = 226 Hz, 4F), −110.6, −115.4 (AB q, J = 226 Hz,
The authors declare no competing financial interest.
1
4
F), J
= 61 Hz. H NMR (CDCl ): δ 7.31 (m, 5H), 6.15 [d, vinyl,
endo 3
2
H, JF(exo)H = 7.4 Hz], 3.47 (s, benzyl, 2H), 1.56 (s, CH , 6H).
3
19
ACKNOWLEDGMENTS
Compound 22. F NMR (CDCl ): δ −110.9, −113.0 (AB q, J = 220
Hz, 4F), −112.4 (s, 4F). Pyrolysis product: F NMR (CDCl ): δ
■
3
19
3
D.M.L. thanks the National Science Foundation (NSF) for
support of the work conducted at Dartmouth College. He is
also grateful to Maren Pink and Victor G. Young, Jr., of the X-
ray Crystallographic Laboratory at the University of Minnesota
for the crystal structure of 13 and to Richard Longmire for
obtaining the Raman spectrum of 1. S.R.K. thanks Katherine M.
Broadus for conducting the FTMS experiments, and R.G.
thanks R. Flatow for recording the PE spectra. J.P.J.
acknowledges the NSF (Grant 8818307) for funds to purchase
the X-ray diffractometer for determining the crystal structure of
−
72.9 (bs, 1F), −74.2 (s, 1F), −76.2 (s, 1F), −77.2 (bs, 1F), −109.1
1
(
s, 2F), −110.5 (s, 2F). H NMR (CDCl ): δ 7.30 (m, 3H), 6.83 (d,
3
2
1
H, J = 7.2 Hz), 5.88 (s, 1H), 5.02 (s, 2H), 4.64 (dt, J = 23, 14 Hz,
H), 2.12 (s, 3H), 2.02 (s, 3H). MS: m/z 409 (M ), 91 (base, C H ).
+
+
7
7
1
1-Aza-7,7,8,8,9,9,10,10-octafluoro-11-methylquadricyclo-
2 , 5 1 , 6
[
(
(
(
4.2.2.1 .0 ]undec-3-ene (23) and 1-Methyl-2-
2,2,3,3,5,5,6,6-octafluorobicyclo[2.2.0]hex-1-yl)-1H-pyrrole
24). To a flame-dried 25 mL flask were added dibromo compound 5
518 mg, 1.3 × 10 mol), N-methylpyrrole (147 mg, 1.8 × 10 mol,
.4 equiv), 8 mL of dry acetonitrile, and 0.4 g of activated zinc. This
−3
−3
1
mixture was placed in a water-cooled ultrasonic bath and rocked using
a modified Kugelrohr motor. A 19F NMR spectrum after 2 h showed
only starting material and compound 23. The spectrum after 15 h, just
as in the 2,5-dimethyl-N-benzylpyrrole case, showed a 50:50 mixture
of the desired Diels−Alder adduct 23 and its aromatized isomer 24.
This solution was added to 15 mL of water, and the aqueous layer was
extracted with 3 × 10 mL of pentane. The combined pentane portions
15.
REFERENCES
■
(1) (a) Wiberg, K. B.; Maturro, M. G.; Okarma, P. J.; Jason, M. E.;
Dailey, W. P.; Burgmaier, G. J.; Bailey, W. F.; Warner, P. Tetrahedron
1986, 42, 1895. (b) Wiberg, K. B.; Bailey, W. F.; Jason, M. E. J. Org.
Chem. 1974, 39, 3803. (c) Jason, M. E.; McGinnety, J. A.; Wiberg, K.
were dried over MgSO , and the pentane was removed under reduced
4
I
DOI: 10.1021/jo502456h
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
B. J. Am. Chem. Soc. 1974, 96, 6531. (d) Wiberg, K. B.; Burgmaier, G.
J.; Warner, P. J. J. Am. Chem. Soc. 1971, 93, 246. See also:
(30) (a) Zhao, Y.; Truhlar, D. G. J. Phys. Chem. A 2008, 112, 1095−
1099. (b) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−
241. (c) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157−167.
(31) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007−1023.
(
e) Casanova, J.; Rogers, H. R. J. Org. Chem. 1974, 39, 3803.
2) Zhang, Y.; Smith, J. R.; Lemal, D. M. J. Am. Chem. Soc. 1996, 118,
454.
3) He, Y.; Junk, C. J.; Cawley, J. J.; Lemal, D. M. J. Am. Chem. Soc.
003, 125, 5590.
(
9
(
2
(
32) Papajak, E.; Truhlar, D. G. J. Chem. Theory Comput. 2010, 6,
5
97−601.
(
4) He, Y.; Junk, C. J.; Lemal, D. M. Org. Lett. 2003, 5, 2135.
(
5) Richardson, A. D.; Hedberg, K.; Junk, C. J.; Lemal, D. M. J. Phys.
Chem. A 2003, 107, 3064.
6) Shelton, G. R.; Hrovat, D. A.; Wei, H.; Borden, W. T. J. Am.
Chem. Soc. 2006, 128, 12020.
(
(
(
(
7) Zdravkovski, Z. J. Mol. Struct.: THEOCHEM 2004, 684, 99.
8) Csaszar, A. G. J. Phys. Chem. A 2004, 108, 2002.
9) Barlow, M. G.; Haszeldine, R. N.; Morton, W. D.; Woodward, D.
R. J. Chem. Soc., Perkin Trans. 1 1972, 2170.
10) The dichloro analogue of 4 had been prepared earlier in
analogous fashion: Camaggi, G.; Gozzo, F. J. Chem. Soc. 1969, 489.
11) Bromoiodide is also a product of this reaction, but when ether is
(
(
present in the acetonitrile, diiodide 11 becomes the dominant product.
The reaction fails in ether alone.
(
(
(
12) Finkelstein, H. Chem. Ber. 1910, 43, 1528.
13) Pirkle, W. H. J. Org. Chem. 1983, 48, 2279.
14) Cherwell Scientific Publishing Ltd., The Magdalen Centre,
Oxford Science Park, Oxford OX44GA, England.
15) (a) Santaniello, E. Synthesis 1979, 617. (b) D’Silva, C.; Walker,
D. A. J. Org. Chem. 1998, 63, 6715.
(
(
(
(
(
(
(
(
16) Hamelin, J.; Klein, B.; Texier-Boullet, F. Synthesis 1986, 409.
17) Servis, K. L.; Roberts, J. D. J. Am. Chem. Soc. 1965, 87, 1339.
18) Bent, H. A. Chem. Rev. 1961, 61, 275.
19) Jaguar, version 7.0; Schrodinger, LLC: New York, 2007.
20) Kaprinidis, N. A.; Turro, N. J. Tetrahedron Lett. 1996, 37, 2373.
21) Koopmans, T. Physica 1934, 1, 104.
22) (a) Brundle, C. R.; Robin, M. B.; Kuebler, N. A.; Basch, H. J.
Am. Chem. Soc. 1972, 94, 1451. (b) Brundle, C. R.; Robin, M. B.;
Kuebler, N. A. J. Am. Chem. Soc. 1972, 94, 1466.
(
23) Shriner, R. L.; Neumann, F. W. Organic Syntheses; Wiley &
Sons: New York, 1955; Collect. Vol. 3, p 73.
24) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A. K.;
(
Puschmann, H. OLEX2: A complete structure solution, refinement
and analysis program. J. Appl. Crystallogr. 2009, 42, 339−341.
(
(
25) SUPERFLIP. J. Appl. Crystallogr. 2007, 40, 786−790.
26) Sheldrick, G. M. SHELXL. Acta Crystallogr. 2008, A64, 112−
122.
(
1
(
27) Merrill, G. N.; Zoller, U.; Reed, D. R.; Kass, S. R. J. Org. Chem.
999, 64, 7395−7406.
28) All of the gas phase electron and proton affinities come from:
Bartmess, J. E. NIST Chemistry WebBook, NIST Standard Reference
Database Number 6; Mallard, W. G., Lustrum, P. J., Eds.; National
Institute of Standards and Technology: Gaithersburg, MD, 20899
(http://webbook.nist.gov).
(
29) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09;
Gaussian, Inc.: Wallingford, CT, 2009.
J
DOI: 10.1021/jo502456h
J. Org. Chem. XXXX, XXX, XXX−XXX