Unraveling the Role of Alkyl F on CH-π Interactions and Uncovering the Tipping Point for Fluorophobicity

Article abstract of DOI:10.1021/acs.joc.5b01072

Although fluorine often plays an influential role in molecular recognition, little is known about the effect of aliphatic fluorine on the CH-π interaction in solution. A series of molecular balances were synthesized that contain fluorinated and nonfluorin

Full text of DOI:10.1021/acs.joc.5b01072

Note
pubs.acs.org/joc
Unraveling the Role of Alkyl F on CH−π Interactions and Uncovering
the Tipping Point for Fluorophobicity
Mark R. Ams,*,† Michael Fields,†,∥ Timothy Grabnic,†,∥ Benjamin G. Janesko,‡ Matthias Zeller,§
Rose Sheridan,†,∥ and Amanda Shay†,∥
†Department of Chemistry, Allegheny College, 520 North Main Street, Meadville, Pennsylvania 16335-3902, United States
‡Department of Chemistry, Texas Christian University, 2800 Souh University Drive, Fort Worth, Texas 76109, United States
§Department of Chemistry, Youngstown State University, One University Plaza, Youngstown, Ohio 44555, United States
S
* Supporting Information
ABSTRACT: Although fluorine often plays an influential role in molecular recognition, little is known about the effect of
aliphatic fluorine on the CH−π interaction in solution. A series of molecular balances were synthesized that contain fluorinated
and nonfluorinated alkyl groups. Our findings indicate that fluorine’s polarizing ability does enhance CH−π binding and depends
on molecular orientation. Surprisingly, when the terminal end of the alkyl group is completely fluorinated, the balance tips
toward fluorophobicity and assumes an unusual constrained conformation.
he ability to correctly predict the interactions occurring
attached to aromatics,9,10 systems that have been thoroughly
Tamong molecules in solution without engaging in studied.11 However, much less is known about how sp3-
centered aliphatic fluorine influences CH−π attraction. Few
studies report solution-phase experimental free energies for
aliphatic fluorine interacting with electron-rich π systems like
those of aromatic amino acids.12,13 Many important drugs
contain aliphatic fluorine,14 while many more are being
intensely sought after by synthetic chemists,15 underscoring
the need for accurate predictive binding models. This is
emphasized by Ho,1 Zhu,16 and others,17 who argue that such
data is needed to validate QM and MM calculations as well as
to treat molecular modeling issues involving halogens.
laboratory experimentation is an unrealized goal in modern
chemistry, in part due to the enormous challenge of properly
accounting for the many forces that act upon molecules in
dynamic systems. Ideally, the behavior of complicated processes
such as protein folding and drug−receptor binding is predicted
by applying the proper algorithms derived from experimental
data.1 Current computational approaches, however, are limited
by the lack of experimental data derived from the forces that
drive these processes. One of the specific information gaps that
contributes to the inaccuracy of current algorithms is the
experimentally elusive data occurring in noncovalent inter-
actions.2 Among the weakest of these interactions is the CH−π
interaction, which is worth just under 1 kcal/mol in energy but
is often very influential in molecular recognition.3
While some basic principles of the CH−π interaction are
now understood,4 the rapid emergence of fluorine (F) in the
past decade as a structural design feature to facilitate binding
has uncovered a largely unexplored area in which these two
entities intersect. Fluorine is now represented in at least 20% of
drugs on the market today5 and is increasingly engineered into
the design of oligopeptides,6 proteins,7 and small-molecule
inhibitors.8 Many of these involve fluorine atoms directly
We employed the Wilcox molecular torsion balance18 as a
precise tool to systematically quantify the relative Gibbs free
energies of fluorinated alkyl groups attracted to π systems. As
with related balances that measure nonbonding interaction
strengths,19 the Wilcox balance scaffold is structurally designed
to measure the CH−π attraction by using two separated
conformational states (folded and unfolded, Figure 1) that are
both directly observable and quantifiable by NMR spectroscopy
at ambient temperature. In the folded conformation, the R
Received: May 13, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.joc.5b01072
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
α
β
Figure 1. Folded and unfolded conformational states of molecular balance 2 (R = CH2 CH3) highlighting the β-carbon CH−π interaction. The
barrier to rotation (red arrow) is approximately 16 kcal/mol,18a allowing each isomer to be observable and quantified by H NMR.
1
group is situated directly above the bottom π shelf and is
subject to CH− attractions. Any such attractions between R
and the vertical aromatic ring are canceled due to an equivalent
one in the unfolded state, thereby isolating those occurring with
the bottom shelf. R groups of interest can be easily tailored
during synthesis, allowing us to survey aliphatic R groups
containing fluorine atoms.
To study the influence of fluorine systematically, balances 1−
8 containing aliphatic R groups with varying amounts of
fluorine were synthesized and characterized by X-ray
crystallography, and their folded/unfolded ratios were
determined by 1H NMR spectroscopy in CDCl3 solvent
(Table 1).20 Compound 1, in which R = methyl, was used as
Table 1. Folding of Balances 1−8
a
b
balance
R
K (folded/unfolded)
ΔG°fold (kcal/mol)
1
2
3
4
5
6
7
8
αCH3
1.07
1.79
2.01
2.00
1.48
2.09
2.35
1.19
−0.04
−0.35
−0.41
−0.41
−0.23
−0.44
−0.50
−0.10
αCH2 CH3
β
αCH2 CH2F
β
αCH2 CHF2
β
Figure 2. (a) Optimal orthogonal binding orientation for CHX3 with
benzene.22 (b) X-ray crystal structures of balances 2−4, 6 and 7 in the
folded conformation. Balance 5 did not crystallize. Disorder in crystals
2 and 4 is omitted with only the major moieties shown (95.5(4)%
occupancy in 2, 0.761(3) occupancy in 4). See the Supporting
Information for details.
αCH2 CF3
β
αCH(βCH3)2
αCH(βCH2F)2
αCH(βCF3)2
a
Ratio of solute as determined by NMR spectroscopy in CDCl3 at 298
b
1
K. The uncertainty in H NMR integration measurements has an
estimated error value of 0.02 kcal/mol.
The most surprising result from the series is that complete
fluorination of the β-carbon still results in folding. The ΔG°fold
for the trifluoromethyl derivatives 5 and 8 were −0.23 and
−0.10 kcal/mol, respectively. Since aliphatic fluorine repels π
clouds in solution,24 we therefore anticipated that 5 and 8
would either strongly disfavor folding or be forced to invoke a
longer range (and thus only weakly attractive) CH−π effect
from the α-carbon. However, the X-ray crystal structure of 8
(Figure 3) shows its electron-deficient β-carbon positioned ∼5°
from the center of the π shelf at a distance of 3.968 Å.
Closer inspection of the X-ray structure of balance 8 shows
that the −CF3 carbon atom hovers directly above the center of
the π shelf, where π-density is at its lowest, and that the
polarized C−F bonds of that group point themselves toward
the positively polarized benzenoid rim. One explanation for this
arrangement is the existence of a direct electrostatic attraction
between these polarized groups, similar to that described by
Wheeler and Houk (for substituted benzene with benzene25).
Using dispersion-corrected density functional theory (DFT-D)
in continuum chloroform, we sought to capture this apparent
switch in binding modes for 5 and 8 as compared to the CH−π
attraction occurring in 2−4, 6, and 7. The calculations, which
focused on obtaining the folding energies ΔEfold of balances 1−
8 and on isolated benzene−CHnF4−n complexes (for reference),
our control model given its low preference for either
conformation. Lengthening the aliphatic chain with one carbon
and transforming it to an ethyl group as shown in balance 2
permits it to participate in the CH−π attraction. This
lengthening yields an experimental folding energy (ΔG°fold
)
of −0.35 kcal/mol.
Reasoning that iterative substitution of Hs for fluorines on
the β-carbon would polarize the remaining hydrogens and
render them more acidic toward the bottom π shelf,21 we
searched for possible enhanced folding preferences for balances
3 and 4 and their isopropyl counterparts 6 and 7. However, the
fluorine substitutions had an overall modest effect on the
experimental folding ratios. Close examination of the X-ray
crystal structures of balances 2−4, 6, and 7 in their folded
conformations show that the expected CH−π binding mode
still operates and that the C−F bond(s) point away from the π
shelf. The optimal CH−π binding orientation for these alkyl
groups, where the H points orthogonally toward the aromatic
ring,22 is not achievable within the torsion balance framework
and results in only small net changes in folding when R is
fluorinated (Figure 2).23
B
DOI: 10.1021/acs.joc.5b01072
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
Figure 3. X-ray crystal structures and highlighted geometries. (a) Balance 8. (b) X-ray fragment of balance 8 (top-down view) showing the CF3−π
orientation. (c) Hexafluoro ester 8a. Inspection of the crystal lattices in each shows that there are no close intermolecular CF3−π contacts.
Figure 4. Gas-phase ωB97X-D/6-31G(d) DFT computed geometries and molecular density isosurfaces (0.01 electrons/Bohr3) colored by
molecular electrostatic potential. (a) Balance 8. (b) CF4−benzene complex. Surfaces are colored from electrostatic potential −0.15 au (green) to
+0.15 au (blue).
roughly correlate with the observed ΔG°fold’s and do capture an
electrostatic attraction for 8 (see the Supporting Informa-
tion).22,26 Figure 4 shows the anticipated regions of negative
charge density on the fluorines and that the electropositive
−CF3 carbon prefers to hover above an electron-rich benzene
carbon below it, much like a CF4−benzene complex does. The
computed ΔEfold for 8, however, is less negative than it is for
the typical CH−π interaction in 2 (−1.99 vs −2.16 kcal/mol in
chloroform). Calculations on balance 5 predict that the CF3−π
Chart 1. Comprehensive CSD Search Result for Crystals
Containing the RCF3 Group (R = Csp3) That Have the
Trifluorinated C Positioned in Any Orientation within a 5 Å
Sphere of the Center of a Benzenoid Ring (See Chart Inset
a
(a))
α
interaction competes weakly with an CH−π attraction (−2.7
versus −1.8 kcal/mol, suggesting that 22% of 5 would have the
αCH−π interaction at room temperature); this is stronger than
a typical nonhalogenated CH−π due to the adjacent polarizing
−CF3 substituent.27 Our persistent, but so far unsuccessful,
attempts to crystallize 5 are consistent with a competition
α
between nearly isoenergetic CH−π and CF3−π interactions.
α
Balance 8 does not have the CH−π option since that would
cause severe steric clashes between the isopropyl R group and
the ester linker (see the Supporting Information).
Given the small energy differences for 5 and 8, and the subtle
orientation variance between the computed geometry for 8 and
its X-ray structure, we used the crystallographic data to
determine the prevalence of such an electrostatic interaction
within other known crystal structures. We performed a
comprehensive Cambridge Structural Database28 (CSD) search
for aliphatic CF3−π interactions (Chart 1), generating 2147
matches for complexes containing an aliphatic CF3 group
positioned within a 5 Å proximity of a benzenoid center. Even
more noticeable is that balance 8’s CF3−π contact distance
places it among the top 53 shortest in the entire set (all 53 have
a C-to-centroid distance <4 Å). Closer inspection of the specific
CF3−π interactions in each of these 53 complexes further
a
Balance 8’s short 3.968 Å CF3−π contact distance places it among
the shortest 3.5% of the 2147 total occurrences (checkered bar). Inset
(b) shows the umbrella-like orientation above the benzenoid
(imaginary red plane running perpendicular to the ring and running
approximately through the Csp3−C bond and the Fs pointing toward
b
benzenoid periphery). Value rounded to the nearest first decimal.
revealed that the CF3 was positioned within 25° above the
aromatic center in every case and that the majority (42 out of
53) had the CF3 in an umbrella-like orientation hovering over
the ring (Chart 1, inset b). The C−βC−F dihedral angle is
α
observed just slightly reduced in 8 as the −CF3 group fans
C
DOI: 10.1021/acs.joc.5b01072
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
α
upward, compared to the other C−βCF3 (Figure 3) on the
mixture was filtered, and the remaining solution was concentrated
under vacuum and isolated by silica chromatography (1:4 ethyl
acetate/hexanes).
noninteracting portion of the ester arm. The crystal structure of
the hexafluoroester derivative 8a is provided for comparison
purposes, showing two CF3’s that do not interact with π
systems.
Model Procedure for Suzuki-Coupling Reaction (1−8). Under
an argon atmosphere, a flame-dried 4 mL screw cap glass vial equipped
with a magnetic stir bar and septum was charged with the
corresponding benzoate ester (0.41 mmol), 2-methyl-8-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl)-6,12-dihydro-5,11-
methanodibenzo[b,f ][1,5]diazocine (150 mg, 0.41 mmol), Pd(OAc)2
(4.4 mg, 0.02 mmol), and SPhos (16.3 mg, 0.04 mmol). The vial was
then capped and purged with argon. Acetonitrile (0.82 mL) was
added, followed by 2.0 M K2CO3 (0.41 mL) via syringe. The reaction
mixture was stirred for 3 h at 90 °C, cooled to room temperature, and
filtered through a Celite pad, eluting with ethyl acetate. The solution
was washed with water and brine and dried over MgSO4. The organic
layer was filtered through a Celite pad, concentrated under vacuum,
and isolated by silica chromatography (1:4 ethyl acetate/hexanes).
NMR and HRMS Characterization of New Compounds. 2-
Fluoroethan-2-yl 2-bromo-3-methylbenzoate (3a): yellow oil; yield
The CSD data reveal that the geometry of balance 8 is more
readily explained as the consequence of the inherent balance
framework than that of an electrostatic attraction. If the CF3−
benzenoid attraction was dominant, we would expect a distinct
peak in the population of Chart 1 at the ∼3.9 Å distance.
Rather, the fanning out of the −CF3 group and its unusually
short contact distance to the benzenoid center are the outcome
of pushing the two groups together and maximizing the contact
surface area between the tetrahedral −CF3 set and the flat π
surface underneath it. 19F NMR data support that this is the
folded conformation of 8 in CDCl3 as well, a likely
consequence of fluorophobic effects. In other nonfluorinated
solvents such CD2Cl2, DMSO-d6, and C6D6, the ΔG°fold of 8 is
virtually unchanged (−0.14, −0.11, and −0.15, respectively).
Neither these solvents nor balance 8 were miscible in
perfluorinated solvents, preventing our initial efforts to test
for fluorophilic effects.
To summarize, our systematic study quantitatively dissects
the influence of aliphatic fluorine on the CH−π attraction in
solution. We found that fluorine’s effect on the CH−π interplay
is influential but also depends on the CH−π binding
orientation. Perfluorination of the terminal end of an alkyl
group tips the balance toward solvent fluorophobicity, inducing
unusually close contacts with the π shelf. The observed stability
trends for 1−8 cannot be explained as strictly electrostatic29 but
are also the result of a delicate balance occurring between
conformational constraints, flexibility of the alkyl groups, and
solvophobic contributions. These data can be valuable in
calibrating current theoretical constructs for predicting the
affinity and activity of alkyl and perfluorinated groups toward π
molecular surfaces. As the fields of molecular recognition,
medicinal chemistry, and materials research increasingly involve
fluorine, information about these F-steering interactions
provide insights into the formulation of more effective catalysts,
surfactants, F-containing inhibitors, and other biomolecules.
1
450 mg (70%); H NMR (CDCl3) δ 7.46−7.44 (m, 1H), 7.26 (m,
1H), 7.17 (m, 1H), 4.71 (td, J = 2.6 and 1.4 Hz, 1H), 4.59 (td, J = 2.6
and 1.4 Hz, 1H), 4.54 (td, J = 2.6 and 1.4 Hz, 1H), 4.47 (td, J = 2.6
and 1.4 Hz, 1H), 2.38 (s, 3H); 13C NMR (CDCl3) δ 166.9, 139.8,
133.5, 128.2, 127.0, 123.2, 82.1, 80.4, 64.5, 64.3, 23.8; HRMS (ESI/
TOF) m/z [M + H]+ calcd for C10H11BrFO2 260.9926, found
260.9929.
2,2-Difluoroethan-2-yl 2-bromo-3-methylbenzoate (4a): yellow
oil; yield 540 mg (73%); 1H NMR (CDCl3) δ 7.49 (m, 1H), 7.32 (m,
1H), 7.20 (m, 1H), 5.93−6.21 (tt, J = 52 and 4.0 Hz, 1H), 4.49 (td, J =
16 and 4 Hz, 2H), 2.42 (s, 3H); 13C NMR (CDCl3) δ 166.1, 140.1,
133.9, 132.5, 128.5, 127.0, 123.5, 115.2, 112.8, 110.4, 63.6, 63.3, 63.0,
23.8; HRMS (ESI/TOF) m/z [M + Na]+ calcd for C10H9BrF2NaO2
300.9652, found 300.9658.
2,2,2-Trifluoroethan-2-yl 2-bromo-3-methylbenzoate (5a): yellow
1
oil; yield 574 mg (77%); H NMR (CDCl3) δ 7.55−7.57 (m, 1H),
7.38−7.41 (m, 1H), 7.25−7.29 (m, 1H), 4.66−4.73 (q, J = 8.0 Hz,
2H), 2.46 (s, 3H); 13C NMR (CDCl3) δ 165.4, 140.4, 134.4, 131.9,
128.8, 127.3, 127.2, 124.6, 121.8, 119.0, 61.8, 61.4, 61.1, 60.7, 24.0;
HRMS (ESI/TOF) m/z calcd for C10H8BrF3O2 295.9657, found
295.9660.
1,3-Difluoroisopropan-2-yl 2-bromo-3-methylbenzoate (7a): yel-
low oil; yield 240 mg (36%); 1H NMR (CDCl3) δ 7.51−7.53 (m, 1H),
7.38−7.36 (m, 1H), 7.28−7.24 (m, 1H), 5.48 (tp, J = 20 and 4.0, 1H),
4.80−4.68 (dd, J = 48 and 4, 4H), 2.46 (s, 3H); 13C NMR (CDCl3)
δ166.3, 140.1, 133.8, 133.8, 133.1, 128.4, 128.3, 127.1, 127.1, 81.2,
81.1, 79.5, 79.4, 79.4, 71.8, 71.6, 71.6, 23.9, 23.9; HRMS (EI/TOF)
m/z calcd for C11H11BrF2O2 291.99105, found 291.99062.
1,1,1,3,3,3-Hexafluoroisopropan-2-yl 2-bromo-3-methylben-
EXPERIMENTAL SECTION
■
1H, 19F, and 13C NMR spectra were recorded using a 400 MHz
spectrometer. Flash column chromatography for benzoate ester
syntheses were performed using silica gel (60 μm). Flash column
chromatography for molecular balances was performed using an
automated column system. All hexanes and THF were distilled.
“SPhos” represents the molecule 2-dicyclohexylphosphino-2′,6′-
dimethoxlbiphenyl. NMR signals abbreviations are as follows: s =
singlet, d = doublet, t = triplet, q = quartet, p = pentet, and m =
multiplet. In order to ensure continuity among all NMR data, all
spectra were subjected to line fitting and Bernstein polynomial
baseline corrections, available in MestReNova 8.0 NMR software.
High-resolution mass spectra (HRMS) were recorded in the
electrospray ionization (ESI) or electron impact (EI) mode on a
TOF mass analyzer.
1
zoate (8a): white solid; yield 113 mg (14%); H NMR (CDCl3) δ
7.63−7.61 (m, 1H), 7.48−7.46 (m, 1H), 7.34−7.33 (m, 1H), 6.00
(sept, J = 8.0 Hz, 1 H), 2.50 (s, 3H); 13C NMR (CDCl3) δ 163.3,
140.8, 135.1, 130.0, 129.1, 127.2, 124.8, 124.5, 122.0, 119.2, 116.4,
68.2, 67.8, 67.5, 67.2, 66.8, 66.5, 66.1, 23.9; HRMS (EI/TOF) m/z
calcd for C11H7BrF6O2 363.95334, found 363.95394; mp 52−53 °C.
Ethyl 3-methyl-2-((11R)-8-methyl-6,12-dihydro-5,11-
methanodibenzo[b,f ][1,5]diazocin-2-yl)benzoate (2): clear oil;
1
yield 68 mg (35%); H NMR (CDCl3) δ 7.58/7.55 (d, J = 8.0 Hz,
1H), 7.35−7.30 (m, 1 H), 7.27−7.23 (m, 1H), 7.15−7.13 (m, 1 H),
7.60−6.93 (m, 1H), 6.69 (m, 2 H), 4.71−4.67 (m, 2H), 4.37−4.32 (m,
2H), 4.17−4.10 (m, 2H), 3.66−3.50/3.48−3.40 (m, 2H), 2.25/2.20 (s,
3 H), 2.13/2.04 (s, 3 H, 1.90/1.06), 0.96/0.05 (t, 3H); 13C NMR
(CDCl3) δ 169.5, 169.1, 146.9, 146.7, 145.1, 140.7, 140.5, 137.2, 137.0,
136.3, 136.1, 133.7, 133.5, 132.9, 132.8, 132.8, 132.8, 128.3, 128.2,
128.1, 127.2, 126.9, 125.1, 124.7, 124.3, 124.3, 67.3, 67.0, 60.8, 60.6,
59.2, 59.0, 58.5, 20.9, 13.8, 12.5; HRMS (ES/TOF) m/z [M + H]+
calcd for C26H27N2O2 399.2073, found 399.2067.
Model Procedure for Benzoate Ester Synthesis (1a−8a).
Under an argon atmosphere, a flame-dried 40 mL vial (fitted with cap,
septum, and stir bar) containing 2-bromo-3-methylbenzoic acid (537.6
mg, 2.5 mmol) in THF (5.0 mL) was sequentially charged with
DMAP (30.5 mg, 0.25 mmol) in THF (3.0 mL, distilled) and alkyl
alcohol (10.0 mmol) via syringe. DCC (515 mg, 2.5 mmol) in THF
(8.0 mL, distilled) was added via syringe, and the reaction mixture was
stirred for 5 min at 0 °C. The reaction mixture was then warmed to
room temperature and stirred for an additional 48 h. The reaction
2-Fluoroethyl 3-methyl-2-((11R)-8-methyl-6,12-dihydro-5,11-
methanodibenzo[b,f ][1,5]diazocin-2-yl)benzoate (3): yellow oil;
1
yield 50 mg (30%); H NMR (CDCl3) δ 7.65−7.63/7.60−7.58 (d, J
= 8.0 Hz, 1 H), 7.38−7.25 (m, 2 H), 7.17−6.97 (m, 4 H), 6.77−6.70
D
DOI: 10.1021/acs.joc.5b01072
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
(m, 2H), 4.72−4.68 (m, 2H), 4.42−4.12 (m, 4 H), 3.76−3.64/3.63−
3.51 (m, 2 H), 3.25/3.13 (m, 2H), 2.26/2.23 (s, 3 H), 2.14/2.04 (s, 3
H, 2.07/1.03); 13C NMR (CDCl3) δ 169.1, 168.6, 147.0, 146.8, 145.5,
145.2, 141.1, 140.9, 137.4, 137.2, 136.3, 135.9, 133.6, 133.5, 133.2,
132.0, 131.8, 128.3, 128.1, 127.8, 127.5, 127.2, 125.1, 124.9, 124.6,
124.5, 82.0, 81.0, 80.3, 79.3, 75.1, 67.3, 66.9, 63.8, 63.6, 63.4, 63.2,
59.1, 58.9, 58.5, 25.0, 20.9, 14.3; HRMS (ES/TOF) m/z [M + H]+
calcd for C26H26FN2O2 417.1978, found 417.1974.
ASSOCIATED CONTENT
* Supporting Information
■
S
1H and 13C NMR spectra for all newly reported compounds,
error analysis, NMR spectroscopy experiments, computational
details, and X-ray data for 1−4, 6−8, and 8a (CIF). The
Supporting Information is available free of charge on the ACS
Publications website at DOI: 10.1021/acs.joc.5b01072.
2,2-Difluoroethyl 3-methyl-2-((11R)-8-methyl-6,12-dihydro-5,11-
methanodibenzo[b,f ][1,5]diazocin-2-yl)benzoate (4): yellow oil;
yield 83 mg (50%); 1H NMR (CDCl3) δ 7.65−7.61 (m, 1 H),
7.40−7.28 (m, 2 H), 7.16−6.69 (m, 4 H), 6.76−6.67 (m, 2 H), 5.65−
5.37 (tt, 1 H), 4.71−4.66 (m, 2 H), 4.38−4.11 (4 H), 3.72−3.61/
3.55−3.44 (m, 2H, 0.76/0.66), 2.25/2.20 (s, 3 H), 2.12/2.03 (s, 3 H,
1.82/0.91); 13C NMR (CDCl3) δ 171.3, 168.2, 167.8, 147.0, 145.7,
145.0, 141.2, 137.5, 136.1, 134.1, 133.7, 131.0, 127.9, 127.7, 127.3,
125.2, 124.6, 114.7, 112.3, 109.9, 75.0, 67.4, 67.2, 66.9, 62.9, 62.7, 62.4,
62.1, 60.5, 59.3, 59.0, 58.5, 25.0, 20.9, 14.3; HRMS (ES/TOF) m/z
[M + H]+ calcd for C26H25F2N2O2 435.1884, found 435.1882.
2,2,2-Trifluoroethyl 3-methyl-2-((11R)-8-methyl-6,12-dihydro-
5,11-methanodibenzo[b,f ][1,5]diazocin-2-yl)benzoate (5): yellow
AUTHOR INFORMATION
Corresponding Author
*E-mail: mams@allegheny.edu.
■
Notes
∥(M.F., T.G., R.S., A.S.) Undergraduate student at Allegheny
College.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The research was supported by an award from Research
Corporation for Scientific Advancement, The Thomas Lord
Charitable Trust Fellowship, and Allegheny College. The X-ray
diffractometer was funded by NSF Grant Nos. 0087210 and
1337296, Ohio Board of Regents Grant No. CAP-491, and by
Youngstown State University. We are grateful for helpful
discussions with Dr. Brijesh Bhayana (Massachusetts General
Hospital), S. Shaun Murphree (Allegheny College), and Agusti
1
oil; yield 64 mg (37%); H NMR (CDCl3) δ 7.66−7.64/7.58−7.56
(d, J = 8 Hz, 1H), 7.39−6.63 (m, 8 H), 4.68−4.64 (m, 2H), 4.37−4.07
(m, 4 H), 3.92−3.82/3.45−3.36 (m, 2 H), 2.24/2.18 (s, 3 H), 2.13/
2.01 (s, 3 H, 3.18/2.15); 13C NMR (CDCl3) δ 171.3, 167.6, 167.0,
147.4, 147.2, 145.8, 145.2, 141.8, 141.4, 137.9, 137.6, 135.6, 135.4,
134.0, 133.8, 133.56, 130.8, 130.2, 128.3, 127.5, 127.3, 127.0, 125.2,
124.7, 67.4, 66.9, 60.6, 60.3, 60.0, 59.6, 59.2, 59.0, 58.6, 21.0, 20.8,
14.4; HRMS (ES/TOF) m/z [M + H]+ calcd for C26H24F3N2O2
453.1790, found 453.1792.
́
́
́
Lledo-Ponsati (University of Girona) and insight provided by
Dr. Allen D. Hunter (Youngstown State University).
1,3-Difluoropropan-2-yl 3-methyl-2-((11R)-8-methyl-6,12-dihy-
dro-5,11-ethanodibenzo[b,f ][1,5]diazocin-2-yl)benzoate (7): yellow
1
oil; yield 100 mg (59%); H NMR (CDCl3) δ 7.64−7.62/7.17−7.15
REFERENCES
■
(d, J = 8 Hz, 1 H), 7.40−7.27 (m, 3 H), 7.07−6.94 (m, 3 H), 6.78−
6.70 (m, 2 H), 5.18−5.05/4.87−4.75 (tp, J = 20, 4.0 Hz, 1 H), 4.39−
4.11 (m, 4 H), 3.70−3.32 (m, 4 H), 2.26/2.22 (s, 3 H), 2.10/2.04 (s, 3
H, 9.22/3.93); 13C NMR (CDCl3): δ 168.2, 147.2, 146.9, 145.4, 145.1,
141.1, 137.6, 136.4, 134.0, 133.6, 131.6, 131.3, 127.4, 125.2, 124.9,
124.7, 81.0, 80.2, 78.5, 70.1, 70.1, 69.9, 67.4, 67.0, 59.3, 59.0, 58.5,
20.9; HRMS (ES/TOF) m/z [M + H]+ calcd for C27H27F2N2O2
449.2041, found 449.2041.
(1) Scholfield, M. R.; Zanden, C. M. V.; Carter, M.; Ho, S. Protein Sci.
2013, 22, 139−152.
(2) Johnson, E. R.; Keinan, S.; Mori-Sanchez, P.; Contreras-Garcia, J.;
Cohen, A. J.; Yang, W. J. Am. Chem. Soc. 2010, 132, 6498−6506.
(3) (a) Umezawa, Y.; Tsuboyama, S.; Takahashi, H.; Uzawa, J.;
Nishio, M. Tetrahedron 1999, 55, 10047−10056. (b) Bazzicalupi, C.;
Dapporto, P. Struct. Chem. 2004, 15, 259−268. (c) Singh, N. J.; Min, S.
K.; Kim, D. Y.; Kim, K. S. J. Chem. Theory Comput. 2009, 5, 515−529.
(d) Vacas, T.; Corzana, F.; Jimenez-Oses, G.; Gonzalez, C.; Gomez, A.
M.; Bastida, A.; Revuelta, J.; Asensio, J. L. J. Am. Chem. Soc. 2010, 132,
12074−12090. (e) Ozawa, T.; Okazaki, K.; Kitaura, K. J. Comput.
Chem. 2011, 32, 2774−2782. (f) Asensio, J. L.; Arda, A.; Canada, F. J.;
Jimenez, -Barbero Acc. Chem. Res. 2013, 46, 946−954.
1,3-Hexafluoropropan-2-yl 3-methyl-2-((11R)-8-methyl-6,12-di-
hydro-5,11-methanodibenzo[b,f ][1,5]diazocin-2-yl)benzoate (8):
yellow oil; yield 73 mg (34%); 1H NMR (CDCl3) δ 7.73−7.71/
7.67−7.65 (d, J = 8.0 Hz, 1 H), 7.47−6.66 (m, 8 H), 5.76/5.50 (sept, J
= 8.0 Hz, 1 H), 4.73−4.64 (m, 2 H), 4.39−4.07 (m, 4 H), 2.26/2.22
(s, 3 H), 2.08/2.04 (s, 3 H, 0.94/0.79); 13C NMR (CDCl3) δ 183.4,
165.8, 165.0, 147.9, 147.4, 145.1, 144.7, 142.3, 141.7, 138.4, 138.3,
135.1, 134.9, 133.9, 133.8, 133.67, 128.6, 128.4, 128.4, 128.3, 128.0,
127.9, 127.8, 127.7, 127.4, 127.4, 127.3, 127.1, 125.1, 124.9, 124.2,
67.0, 58.7, 21.0, 20.9, 20.8; HRMS (ES/TOF) m/z [M + H]+ calcd for
C27H23F6N2O2 521.1664, found 521.1658.
(4) (a) Nishio, M.; Hirota, M. Tetrahedron 1989, 45, 7201.
(b) Nishio, M.; Umezawa, Y.; Hirota, M.; Takeuchi, Y. Tetrahedron
1995, 51, 8665. (c) Tewari, A. K.; Dubey, R. Bioorg. Med. Chem. 2008,
16, 126. (d) Nishio, M.; Umezawa, Y.; Honda, K.; Tsuboyama, S.;
Suezawa, H. CrystEngComm 2009, 11, 1757. (e) Takahashi, O.;
Kohno, Y.; Nishio, M. Chem. Rev. 2010, 110, 6049. (f) Nishio, M. Phys.
Chem. Chem. Phys. 2011, 13, 13873. (g) Zhang, Z. B.; Xia, B. Y.; Han,
C. Y.; Yu, Y. H.; Huang, F. H. Org. Lett. 2010, 12, 3285. (h) Zhang, Z.
B.; Yu, G. C.; Han, C. Y.; Liu, J. Y.; Ding, X.; Yu, Y. H.; Huang, F. H.
Org. Lett. 2011, 13, 4818. (i) Xue, M.; Yang, Y.; Chi, X. D.; Zhang, Z.
B.; Huang, F. H. Acc. Chem. Res. 2012, 45, 1294.
(5) (a) Chopra, D. Cryst. Growth Des. 2012, 12, 541−546. (b) Walker,
M. C.; Thuronyi, B. W.; Charkoudian, L. K.; Lowry, B.; Khosla, C.;
Chang, M. C. Y. Science 2013, 341, 1089−1094.
(6) Jackel, C.; Koksch, B. Eur. J. Org. Chem. 2005, 2005, 4483−4503.
(7) Jackel, C.; Salwiczek, M.; Koksch, B. Angew. Chem., Int. Ed. 2006,
45, 4198−4203.
Model Procedure for Quantification of Folded/Unfolded
Conformers in Solution. The ratio of folded to unfolded
conformation for each balance was determined by the integration of
1
both the top methyl singlets in the H NMR spectra. Due to the
anisotropic effect, the furthest upfield peak represents the unfolded top
methyl hydrogens. This is paired with the next closest upfield peak
(based on integration and temperature dependent NMR analysis). The
methyl peaks for the folded and unfolded conformations can be used
to determine the Gibbs free energy using the equation ΔG°folded
=
−RT ln(Kfolded) = −RT ln(folded/unfolded). The ratio of folded to
unfolded conformation for balance 8 was confirmed by 19F NMR in
CDCl3, CD2Cl2, DMSO-d6, and C6D6. The folded conformation
shows two signals for the chemically inequivalent −CF3 groups and
one signal representing both −CF3 groups for the unfolded
conformation. All NMR spectra have corrected baselines and were
fitted to a Gaussian (line fitting/deconvolution) function using
MNova 8 NMR software.
(8) (a) Nakagawa, Y.; Irie, K.; Yanagita, R. C.; Ohigashi, H.; Tsuda,
K. J. Am. Chem. Soc. 2005, 127, 5746−5747. (b) Leimbacher, M.;
Zhang, Y.; Mannocci, L.; Stravs, M.; Geppert, T.; Scheuermann, J.;
Schneider, G.; Neri, D. Chem. - Eur. J. 2012, 18, 7729−7737.
(9) For articles discussing crystal data in the Cambridge Structural
Databank and the Protein Data Bank for F in close proximity to π
E
DOI: 10.1021/acs.joc.5b01072
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
systems, see: (a) Matter, H.; Nazare, M.; Gussregen, S.; Will, D. W.;
Schreuder, H.; Bauer, A. Angew. Chem., Int. Ed. 2009, 49, 2911.
(b) Bissantz, C.; Kuhn, M.; Stahl, M. J. Med. Chem. 2010, 53, 5061.
(c) Sun, H. Cryst. Growth Des. 2012, 12, 5655−5662. (d) Chopra, D.
Cryst. Growth Des. 2012, 12, 541−546.
(10) (a) Muller, K.; Faeh, C.; Diederich, F. Science 2007, 317, 1881−
1886. (b) Fernandez, A.; Fraser, C.; Scott, L. R. Trends Biotechnol.
2012, 30, 1−7. (c) Vulpetti, A.; Dalvit, C. Drug Discovery Today 2012,
17, 890−897.
(11) (a) Kim, E.; Paliwal, S.; Wilcox, C. S. J. Am. Chem. Soc. 1998,
120, 11192−11193. (b) DerHovanessian, A.; Rablen, P. R.; Jain, A. J.
Phys. Chem. A 2000, 104, 6056−6061. (c) Gardarsson, H.; Schweizer,
W. B.; Trapp, N.; Diederich, F. Chem. - Eur. J. 2014, 20, 4608−4616.
(d) Garcia, A. M.; Determan, J. J.; Janesko, B. G. J. Phys. Chem. A
2014, 118, 3344−3350.
(12) (a) Saraogi, I.; Vijay, V. G.; Das, S.; Sekar, K.; Guru Row, T. N.
Cryst. Eng. 2003, 6, 69−77. (b) Jackel, C.; Salwiczek, M.; Koksch, B.
Angew. Chem., Int. Ed. 2006, 45, 4198−4203.
(13) Two related experimental studies reporting some aspects of
alkyl F−π strengths are have been published but contain notable
limitations: (a) Adams, H.; Cockroft, S. L.; Guardigli, C.; Hunter, C.
A.; Lawson, K. R.; Perkins, J.; Spey, S. E.; Urch, C. J.; Ford, R.
ChemBioChem 2004, 5, 657−665. The model system in this study is
limited by solubility issues, complexation, competing steric/electro-
static effects, and the geometry of the system. (b) Thomas, K. M.;
Naduthambi, D.; Zondlo, N. J. J. Am. Chem. Soc. 2006, 128, 2216−
2217 The model system in this study does not strictly isolate the
CH−π interaction..
(14) (a) Ismail, F. M. D. J. Fluorine Chem. 2002, 118, 27−33. (b)
Reference 8. (c) Purser, S.; Moore, P. R.; Swallow, S.; Gouverneur, V.
Chem. Soc. Rev. 2008, 37, 320−330. (d) Fabbri, L. M.; Calverley, P. M.
A.; Izquierdo-Alonso, J. L.; Bundschuh, D. S.; Brose, M.; Martinez, F.
J.; Rabe, K. F. Lancet 2009, 374, 695−703.
(15) For several recent advances in the synthesis of aliphatic fluorine
compounds, see: (a) Ritter, T. Nature 2010, 466, 447−448. (b) Furuya,
T.; Kamlet, A.; Ritter, T. Nature 2011, 473, 470−477. (c) Reference
5b.
(16) Lu, Y.; Wang, Y.; Zhu, W. Phys. Chem. Chem. Phys. 2010, 12,
4543−4551.
(17) (a) Raju, R. K.; Bloom, J. W.; An, Y.; Wheeler, S. E.
ChemPhysChem 2011, 12, 3116−3130. (b) Raju, R. K.; Bloom, J. W.
G.; An, Y.; Wheeler, S. E. ChemPhysChem 2011, 12, 3116−3130.
(18) (a) Paliwal, S.; Geib, S.; Wilcox, C. S. J. Am. Chem. Soc. 1994,
116, 4497−4498. (b) Kim, E.; Paliwal, S.; Wilcox, C. S. J. Am. Chem.
Soc. 1998, 120, 11192−11193. (c) Hof, F.; Scofield, D. M.; Schweizer,
W. B.; Diederich, F. Angew. Chem., Int. Ed. 2004, 43, 5056−5059.
(d) Fischer, F. R.; Schweizer, W. B.; Diederich, F. Chem. Commun.
2008, 4031−4033. (e) Bhayana, B.; Ams, M. R. J. Org. Chem. 2011, 76,
3594−3596. (f) Bhayana, B. J. Org. Chem. 2013, 78, 6758−6762.
(g) Yang, L.; Adam, C.; Nichol, G.; Cockroft, S. L. Nat. Chem. 2013, 5,
1006−1010. (h) Gardarsson, H.; Schweizer, W. B.; Trapp, N.;
Diederich, F. Chem. - Eur. J. 2014, 20, 4608−4616.
halogens, fluorine is too polarized to contain a σ hole and repels π
clouds. For further references, see: (b) Forni, A.; Pieraccini, S.;
Rendine, S.; Sironi, M. J. Comput. Chem. 2014, 35, 386−394. (c) Lu,
Y.; Zou, J.; Wang, Y.; Yu, Q. Chem. Phys. 2007, 334, 1−7.
(22) (a) Tsuzuki, S.; Honda, K.; Uchimaru, T.; Mikami, M.; Tanabe,
K. J. Am. Chem. Soc. 2000, 122, 3746−3753. (b) Tsuzuki, S.; Honda,
K.; Uchimaru, T.; Mikami, M.; Tanabe, K. J. Phys. Chem. A 2002, 106,
4423−4428.
(23) To illustrate the known influence of fluorination on the CH−π
interaction, we performed DFT-D and coupled cluster calculations for
gas-phase interaction energies between benzene and CH4, CH3F,
CH2F2, CHF3, and CF4 (see the Supporting Information). They
reproduce fluorination’s known large effect on CH−π interactions in
gas-phase benzene−CHnF4−n complexes. The coupled cluster binding
energies are −1.44, −2.49, −3.40, −4.24, and −1.89 kcal/mol,
respectively. The differences between these gas-phase interactions and
those seen for the balances are consistent with the fact that the
methane derivatives bind with an optimal orientation, while the R
groups under consideration are constrained by the structure of the
balance’s scaffold.
(24) (a) Vangala, V. R.; Nangia, A.; Lynch, V. M. Chem. Commun.
2002, 1304−1305. (b) DerHovanessian, A.; Doyon, J. B.; Jain, A.;
Rablen, P. R.; Sapse, A. Org. Lett. 1999, 1 (9), 1359−1362.
(c) Reimann, B.; Buchhold, K.; Vaupel, S.; Brutschy, B.; Havlas, Z.;
Spirko, V.; Hobza, P. J. Phys. Chem. A 2001, 105, 5560−5566.
(d) Neretin, I. S.; Lyssenko, K. A.; Antipin, M. Y.; Slovokhotov, Y. L.;
Boltalina, O. V.; Troshin, P. A.; Lukonin, A. Y.; Sidorov, L. N.; Taylor,
R. Angew. Chem. 2000, 112, 3411−3414.
(25) Wheeler, S. E.; Houk, K. N. J. Am. Chem. Soc. 2008, 130,
10854−10855.
(26) DFT-D calculations on balances 1−8 in continuum chloroform
solvent, starting from the experimental crystal structures, give
computed ΔEfold values of −1.51 (1), −2.16 (2), −1.98 (3), −1.21
(4), −2.70 (5), −2.66 (6), −3.27 (7), and −1.99 (8). Consistent with
the experimentally obtained ΔG°fold values, 7 and 8 have ΔEfold values
more and less negative (respectively) than balance 6. Balance 7 has the
most negative overall ΔEfold value, and 1 has ΔEfold less negative than
2. The predicted effect of fluorination on balances 2−4 is reduced or
even reversed relative to the benzene−CHnF4−n complexes in ref 23.
The computed ΔEfold thus seem to capture or overestimate the
conformational effects discussed in ref 23. Overall, the folding of
balances 2−4 appears to involve a subtle interplay of electrostatic
attraction, steric constraints, and other effects (solvophobic, entropic,
etc.) that are not captured by DFT-D calculations on single
conformations. As in our previous work, these effects make the
computed ΔEfold values substantially more negative than experimental
ΔG°fold values. See: Janesko, B. G.; Ams, M. R. Theor. Chem. Acc. 2014,
133, 1490.
(27) Several theoretical studies have been conducted on the effects of
fluorine and the CH−π interaction. For more information, we direct
the reader to: (a) Menapace, J. A.; Bernstein, E. R. J. Phys. Chem. 1987,
91, 2843−2848. (b) Reference 22b. (c) Kawahara, S.; Tsuzuki, S.;
Uchimaru, T. J. Phys. Chem. A 2004, 108, 6744−6749. (d) Reference
21. (e) Lu, Y.; Zou, J.; Wang, Y.; Yu, Q. Chem. Phys. 2007, 334, 1−7.
(f) Dey, R. C.; Seal, P.; Chakrabarti, S. J. Phys. Chem. A 2009, 113,
10113−10118.
(19) (a) Carroll, W. R.; Pellechia, P.; Shimizu, K. D. Org. Lett. 2008,
10 (16), 3547−3550. (b) Carroll, W. R.; Zhao, C.; Smith, M. D.;
Pellechia, P. J.; Shimizu, K. D. Org. Lett. 2011, 13 (16), 4320−4323.
(c) Zhao, C.; Parrish, R. M.; Smith, M. D.; Pellechia, P. J.; Sherrill, C.
D.; Shimizu, K. D. J. Am. Chem. Soc. 2012, 134 (35), 14306−14309.
(d) Li, P.; Zhao, C.; Smith, M. D.; Shimizu, K. D. J. Org. Chem. 2013,
78 (11), 5303−5313. (e) Zhao, C.; Smith, M. D.; Pellechia, P. J.;
Shimizu, K. D. Org. Lett. 2014, 16, 3520−3523.
(28) Groom, C. R.; Allen, F. H. Angew. Chem., Int. Ed. 2014, 53,
662−671.
(29) The Supporting Information shows the molecular electrostatic
potential computed for R groups 1−8. Hunter’s model for hydrogen
bonding interactions predicts a linear dependency for the electrostatic
minima of these and ΔG°fold, but in agreement with the Diederich
laboratory, we do not observe a close correlation. For details of
Hunter’s model, see: (a) Hunter, C. A. Angew. Chem., Int. Ed. 2004, 43,
5310−5324. (b) Cockroft, S. L.; Hunter, C. A. Chem. Commun. 2006,
3806−3808. (c) Cockroft, S. L.; Hunter, C. Chem. Commun. 2009,
3961−3963. For the Diederich studies, see: (d) Gardarsson, H.;
Schweizer, W. B.; Trapp, N.; Diederich, F. Chem. - Eur. J. 2014, 20,
4608−4616.
(20) (a) A explanation of how the folding ratios are quantified by 1H
NMR is provided in the Supporting Information, along with NMR
spectra, coalescence NMR experiments, and details of the DFT
calculations.. (b) The crystal structure of 7 contained two different
folded conformers with slightly different geometries, each showing
CH−π interactions. Only one congener of 7 is shown in the
manuscript..
(21) (a) Tsuzuki, S.; Honda, K.; Uchimaru, T.; Mikami, M.; Fujii, A.
J. Phys. Chem. A 2006, 110, 10163−10168. Unlike the heavier
F
DOI: 10.1021/acs.joc.5b01072
J. Org. Chem. XXXX, XXX, XXX−XXX