Perturbation of1JC,FCoupling in Carbon-Fluorine Bonds on Coordination to Lewis Acids: A Structural, Spectroscopic, and Computational Study

Article abstract of DOI:10.1021/acs.inorgchem.0c02516

A lithiated m-terphenyl ligand bearing fluorine atoms at the ortho positions of the flanking aryl rings was synthesized and characterized using single crystal X-ray diffraction, variable-temperature multinuclear NMR spectroscopy, and computational methods. Changes in 1JC,F on coordination to lithium as a spectroscopic observable parametrizing the strength of the C-F···Li interaction are described, and a general, qualitative relationship between C-F bond lengths, Δ1JC,F values, and the extent of C-F bond activation as a result of Lewis acid coordination is proposed.

Full text of DOI:10.1021/acs.inorgchem.0c02516

pubs.acs.org/IC
Article
1
Perturbation of JC,F Coupling in Carbon−Fluorine Bonds on
Coordination to Lewis Acids: A Structural, Spectroscopic, and
Computational Study
Brighton A. Skeel, Michael A. Boreen, Trevor D. Lohrey, and John Arnold*
Cite This: https://dx.doi.org/10.1021/acs.inorgchem.0c02516
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: A lithiated m-terphenyl ligand bearing fluorine atoms
at the ortho positions of the flanking aryl rings was synthesized and
characterized using single crystal X-ray diffraction, variable-temper-
ature multinuclear NMR spectroscopy, and computational methods.
1
Changes in JC,F on coordination to lithium as a spectroscopic
observable parametrizing the strength of the C−F···Li interaction are
described, and a general, qualitative relationship between C−F bond
lengths, Δ1JC,F values, and the extent of C−F bond activation as a
result of Lewis acid coordination is proposed.
it has been reported that cation binding by fluorinated
INTRODUCTION
■
cavitands tends to result in decreased JC,F values,13−16 the
1
The unique chemical properties of fluorinated substituents
have made C−F bonds a common motif within agrochemical1
and pharmaceutical2 compounds, where they have been
successfully used to alter bioactivities and other target
properties. Among other effects, the incorporation of fluorine
in bioactive organic compounds has the ability to modulate
ligand binding affinities by way of introducing new electrostatic
or hydrogen-bonding effects absent in the nonfluorinated
analogues.3 In studying these compounds and their mecha-
nisms of action, it is desirable, though frequently exper-
imentally challenging, to quantify the strength of interactions
between the relevant C−F bonds and the Lewis acid (LA) sites
they may bind.4
The past few years have witnessed a flurry of publications
describing myriad related noncovalent interactions such as
agostic,5 π−π,6 cation−π,7 and dispersion force interactions.8,9
More than minor contributors to molecular geometry and
electronic structure, these relatively weak interactions often
play important roles in determining the structures of transition
metal complexes,10 biologically relevant macromolecules,11 and
host−guest complexes.12
underlying mechanism causing these changes has not yet been
addressed, so the generality of this phenomenon remains an
important question.
To elucidate the cause of changes in 1JC,F values, we sought a
diamagnetic, noncavitand system in which a strong C−F···LA
interaction could be prepared. To this end, we targeted the
lithium complex of a substituted m-terphenyl ligand. Reported
solid-state structures of unsolvated m-terphenyl lithium species
frequently show close Carene···Li or C−H···Li contacts,17−19
suggesting that a fluorine-substituted analogue might exhibit
persistent C−F···Li interactions.
We envisioned the ligand TerphF (TerphF = 2,6-bis(2,6-
difluoro-4-trimethylsilylphenyl)-4-tert-butylphenyl) with the
intent that the trimethylsilyl and tert-butyl groups would
serve both to provide solubility in hydrocarbon solvents and as
convenient and distinguishable NMR “handles”.20 The syn-
thesis of a related semifluorinated m-terphenyl has been
reported,21 and we developed a similar synthetic approach for
our system.
In comparing C−F···LA interactions to more broadly
studied agostic interactions, in which C−H bonds coordinate
to metal centers,5 we wondered if the two might necessarily
share common spectroscopic signatures. Specifically, we
hypothesized that the characteristic decreases in the one-
Received: August 22, 2020
bond carbon−hydrogen NMR coupling constant (1JC,H
)
observed in agostic interactions as a result of three-center−
two-electron bonding should similarly manifest as changes in
1JC,F values in systems featuring C−F···LA interactions. While
https://dx.doi.org/10.1021/acs.inorgchem.0c02516
© XXXX American Chemical Society
Inorg. Chem. XXXX, XXX, XXX−XXX
A

Inorganic Chemistry
pubs.acs.org/IC
Article
a
Scheme 1. Convergent Synthesis of Semifluorinated m-Terphenyl Lithium Dimer 4
a
Synthesis of 1 from 2,6-dibromo-4-tert-butylaniline (left). The protected, semifluorinated m-terphenyl 2 is synthesized via a double Negishi cross-
coupling utilizing 2,6-difluoro-4-trimethylsilylphenylzinc chloride and 1 (top). Deprotection of 2 with I2 yields the corresponding terphenyl iodide,
3 (right), which is lithiated to yield 4 (bottom).
RESULTS AND DISCUSSION
■
Synthesis and Solid-State Structure of a Semi-
fluorinated m-Terphenyl Lithium. The triazene-protected
semifluorinated m-terphenyl 2 (TerphF-N2-NC4H8) was
synthesized from the corresponding protected dibromide 1
by way of a palladium catalyzed double Negishi cross-coupling
using 2 equiv of 2,6-difluoro-4-(trimethylsilyl)phenylzinc
chloride (Scheme 1). From this protected terphenyl, semi-
fluorinated m-terphenyl iodide 3 (TerphF-I) was prepared in
74% yield by refluxing with 2 equiv of I2 in degassed 1,2-
dichloroethane for 16 h. Lithiation of the resultant aryl iodide
with n-BuLi in hexane furnished 4 ([TerphF-Li]2) in 82% yield
1
(95% pure; by H NMR the only impurity is the protonated
arene, TerphF-H, 5).
Figure 1. Solid-state structure of 4 with 50% probability ellipsoids.
Hydrogen atoms, tert-butyl and trimethylsilyl groups, and lattice
solvent molecules are omitted for clarity. Selected bond lengths (Å):
C−Li range 2.156(5)−2.170(7); F−Li range 1.925(4)−1.945(5);
average C−Ffree 1.356(2); average C−FLi bound 1.381(1). Average τ4
and τ4′ for lithium: 0.67 and 0.66, respectively.
Colorless single crystals of 4 suitable for X-ray diffraction
were obtained from a saturated n-hexane solution at −40 °C.
Like other base-free m-terphenyl lithium species,17−19 4 is
dimeric in the solid-state, with two lithium atoms bridging the
central carbons of the terphenyl units and C−Li distances
ranging from 2.156(6) to 2.170(7) Å (Figure 1). The solid-
state structure of the TerphF-Li dimer displays C−F···Li close
contacts, with F−Li distances ranging from 1.925(4) to
1.945(5) Å. Each lithium center adopts a coordination
geometry intermediate between square planar and tetrahedral
but is best described as loosely tetrahedral (average τ4 and τ4′
for lithium: 0.67 and 0.66, respectively) and is coordinated by
two fluorines, one from each terphenyl moiety of the dimer.
Reflecting their coordination to a metal center, each of these
C−F bonds is elongated by a small but significant distance (av
0.025 Å) relative to their corresponding unbound C−F bonds,
a phenomenon that has been predicted computationally22 and
observed in other systems with C−F···Li interactions evident
in the solid-state.15,16
in which both ligands, and both halves of each ligand, are
equivalent (i.e., effective D2h symmetry). Unexpectedly, no
signals were observed in the 19F NMR spectrum of 4 in C6D6
at room temperature. However, upon cooling a solution of 4 in
toluene-d8 to −10 °C, two signals in the 470 MHz 19F NMR
spectrum appeared at −113.58 and −135.63 ppm correspond-
ing to the free and lithium-coordinated fluorines, respectively.
This implies that the absence of a 19F NMR signal at room
temperature is due to peak coalescence over a 22 ppm range,
leading to a significant loss of signal intensity. Further cooling
7
to −60 °C revealed a 32 Hz scalar coupling to Li (I = 3/2,
92.4% natural abundance, Figure S24) as observed in the peak
at −137.91 ppm in the 19F NMR spectrum, directly indicating
F−Li interactions in solution. The effects of these interactions
are also apparent in the 1H NMR spectrum of 4 in toluene-d8:
below −20 °C, the resonances corresponding to the two
Variable-Temperature NMR Studies. The 1H NMR
spectrum of 4 in C6D6 at room temperature exhibits four
resonances, suggesting a time-averaged solution-state structure
B
https://dx.doi.org/10.1021/acs.inorgchem.0c02516
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
1
Figure 2. Variable-temperature 500 MHz H spectra of the aromatic region of 4 in toluene-d8 between 22 and −70 °C.
protons of each flanking aryl ring are resolved as distinct
features (Figure 2), suggesting that motion of the flanking aryl
rings becomes slow on the NMR time scale and that [TerphF-
Li]2 adopts an effective D2 geometry in solution at low
temperature. Eyring analysis of the peak widths23 of the 19F
signals at different temperatures suggests that this rotation
barrier is enthalpically dominated (ΔH⧧ = 13.0( 0.8) kcal/
mol), with a small, positive entropic contribution of 1.7( 0.1)
cal/mol K (Figure S1). This places an upper limit on the
strength of the C−F···Li interaction of 13.0( 0.8) kcal/mol.
1
At or below −60 °C in the H NMR spectrum, the four
flanking aryl protons on the exterior of the TerphF-Li dimer
appear as a doublet at 6.93 ppm with a three-bond proton−
Figure 3. Portion of the experimental 13C{1H} NMR spectrum of
[TerphF-Li]2 at −70 °C showing the carbons bound to fluorines
(black trace) and corresponding numerical simulation as generated
using Spinach24 (blue trace). Simulation parameters (see Figure S2
for atom labeling scheme): 3JHa,Fa, 8.8 Hz; 1JFa,Ca, 251.0 Hz; 4JFa,Cb, 7.0
fluorine coupling constant (3JH,F) of 8.8 Hz. The flanking aryl
protons in the interior of the dimer appear as a “virtual triplet”
at 6.66 ppm, with a 3JH,F value of 12.2 Hz. The virtual coupling
observed in these protons results from coupling to magneti-
cally inequivalent fluorine atoms that engage in strong (ca. 35
Hz, as obtained by globally fitting the low temperature 1H and
13C{1H} NMR spectra; see Figure S3 and Figure 3) through-
space coupling across individual terphenyl units of the dimer,
giving rise to an AA′XX′ splitting pattern. Such through-space
coupling indicates that 4 maintains its dimeric solid-state
structure in hydrocarbon solution. These second-order effects
also manifest in the −70 °C 13C{1H} NMR spectrum of 4 in
toluene-d8: the carbons of the C−F···Li unit appear as a
doublet of second-order multiplets shifted slightly (Δδ =
−0.68 ppm) from their free values in the exterior of the dimer
(Figure 3).
3
4
1
Hz; JHb,Fb, 12.2 Hz; JFb,Ca, 7.0 Hz; JFb,Cb, 226.4 Hz; JFb,Fc, 35.0 Hz.
δCa: 159.85 ppm. δCb: 159.17 ppm.
Computational Studies of C−F Bond Coordination
within a Semifluorinated m-Terphenyl Lithium Dimer.
1
Having observed a perturbation in JC,F, we next sought to
investigate the system computationally so as to better
understand the physical basis for the observed spectroscopic
parameters. To this end, we studied a truncated variant of 4
(denoted as 4′, see Figure 4) in which the trimethylsilyl and
tert-butyl groups were replaced with hydrogens to minimize
computational cost while maintaining a system that repro-
duced the essential physics of the C−F···Li interaction.
Geometry optimization at the BP86 level with a def2-TZVP
basis set for all atoms (this functional and basis set were
selected for their demonstrated utility in satisfactorily
describing relevant interactions38) reproduced the geometric
parameters of the C−F···Li unit well, with an average absolute
error in bond length of 0.008 Å across the C−F, F−Li, Li−Li,
1
As a result of strong C−F···Li interactions in solution, JC,F
for the lithium-coordinated C−F bonds (226.4 Hz) on the
interior of the dimer is 25 Hz smaller than that for the free C−
F bonds of the dimer (251.0 Hz) at low temperature. At room
1
temperature, this is observed as an average JC,F of 238.7 Hz,
which is significantly lower than the 1JC,F values in 2, 3, and 5,
which span the narrow range 250.9−252.0 Hz.
C
https://dx.doi.org/10.1021/acs.inorgchem.0c02516
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 5. Graphical depiction of the Fermi contact (FC) and
1
paramagnetic spin−orbit (PSO) components of JC,F for the free
Figure 4. Depiction of structure 4′ as optimized at the BP86/def2-
TZVP level. Carbon is shown as gray, fluorine as green, hydrogen as
white, and lithium as purple.
(light blue) and lithium-coordinated (dark blue) C−F bonds in 4′ as
calculated at the BP86/pcsSeg-2 level of theory.
1
graphical depiction). The most important contributors to JC,F
and C−Li distances. Critically, the observed C−F bond
elongation upon coordination to lithium was observed in the
model (Table 1), and the average optimized F−Li distance of
in 4′ are the paramagnetic spin−orbit (PSO) and Fermi terms,
with all other terms either being vanishingly small or relatively
constant (<5 Hz difference) between the free and bound C−F
bonds. Detailed descriptions of the PSO and Fermi coupling
mechanisms are available elsewhere,25 and it is sufficient here
to state that the PSO coupling mechanism involves
interactions between occupied and energetically accessible
unoccupied orbitals and is frequently a reporter of π-bonding
character, whereas the Fermi mechanism couples spins through
σ-bonds with nonzero s-orbital character.
Table 1. Comparison of Various Experimentally Observed
a
Parameters in 4 with the Average Values Calculated for 4′
experimental (4)
calculated (4′)
bound
free
bound
free
C−Fav dist (Å)
1JC,F (Hz)
1.381(2)
226(1)
159.17
−137.91
ca. 35
1.356(4)
251(1)
159.85
−113.58
n/a
1.385(1)
−235(3)
165.66
−134.00
40
1.358(1)
−244(1)
166.39
−107.12
n/a
The coordinated C−F bonds in 4′ have average Fermi
δ
δ
13C (ppm)
19F (ppm)
1
components of JC,F higher in magnitude (more negative,
1
opposite to the overall change in JC,F) by an average of 12.5
JF,F (Hz)
1JF,Li (Hz)
Hz relative to their free counterparts, suggesting that, upon
coordination to a Lewis acid, the atomic s-character of the C−
F bond increases slightly. This is a result of two competing
effects. While coordination to a Lewis acid partially depletes
the bonding electron density of the C−F bond and thus acts to
reduce the magnitude of the J-coupling value (bonding
electron density being a critical mediator of coupling), the
slight elongation of the C−F bond on coordination raises the
magnitude of the Fermi term and is the dominant effect in this
case. One chemically plausible explanation for the latter
phenomenon, which is reflected in the computational data, is
that, upon C−F bond elongation, the system takes on more
[Caryl+][F−] character. In this limiting resonance contributor,
electron density is drawn from a carbon centered orbital of
predominantly p-character (on the basis of its lower ionization
energy) and shifted to fluorine to give a fully occupied set of s-
and p-orbitals. This depletion of electron density at carbon is
in turn compensated for by donation of additional electron
density from neighboring carbon atoms, which acts to preserve
the valence of the carbon bound to fluorine and distribute the
positive charge over multiple nuclei. The net result of this is an
increase in overall electron density on carbon and fluorine and
a concomitant increase in s-electron density at both carbon
and fluorine, resulting in an increase in the Fermi term for the
bond. We acknowledge that this analysis is convoluted slightly
in that LA binding also draws electron density away from the
fluorine, but evidently this has less effect on the Fermi term
than the previously mentioned factors.
32(1)
n/a
42(3)
n/a
a
Values in parentheses are standard errors for bond lengths (ESDs for
crystallographically determined metrics, and population standard
deviations where enough parameters were available from computed
structures) or uncertainties for NMR J-couplings (the reciprocal of
acquisition time for experimentally determined values, and population
standard deviations where enough parameters were available from
computed NMR parameters).
1.946 Å closely matched that observed in the solid-state
structure of 4 (1.935 Å) (see Table S4 for a detailed
comparison of structure metrics between 4 and 4′).
The calculated 13C and 19F chemical shifts of 4′ (BP86/
pcSseg-2) are in good agreement with those observed in 4
(Table 1). In particular, the relative 19F and 13C chemical shift
differences between atoms in lithium bound and free C−F
bonds (Δδ 19F and Δδ 13C) for 4′ of 26.88 and 0.73 ppm,
respectively, closely match those observed in 4 at low
temperature (24.33 and 0.68 ppm). The through-space
19F−19F coupling and scalar 19F−7Li coupling observed in 4
are also replicated well in 4′, providing further assurance that
the computed NMR parameters of 4′ accurately approximate
the real spin system present in 4. Accordingly, we can
1
confidently discuss the JC,F values of 4′ as proxies for those
observed in 4.
The average 1JC,F value for the free C−F bonds (−243.5 Hz;
1
note that while experimental JC,F values are frequently
reported as absolute values, it is understood that the sign of
these values is generally negative, as is found for the computed
1JC,F values) in 4′ is calculated to be larger in magnitude than
that of the bound C−F bonds (−234.7 Hz; see Figure 5 for a
The increase in magnitude of the Fermi term is similarly
3
reported by JH,F, where the coupling strength between the
flanking aryl protons nearest the lithium and their associated
bound fluorines in 4, 12.2 Hz, is higher than that observed for
D
https://dx.doi.org/10.1021/acs.inorgchem.0c02516
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 6. Plot showing decreases in 1JC,F values in reported structurally characterized systems featuring static, nonexchanging C−F···LA interactions
in solution from their “free” values versus C−F bond length as determined by X-ray crystallography (blue points). The solid-state structures, NMR
parameters, and CCDC accession codes for each data point are tabulated in Table S1 and are labeled from 1 to 17 in the same manner they are in
this plot. Dashed contour lines show, for a given C−F bond length and decrease in 1JC,F, what portion of the C−F bond order is lost due specifically
to coordination (rounded to the nearest 0.01), as calculated using the [C6H5F···Li(OMe2)3]+ model system.
the opposite pairs, 8.8 Hz, due to increased atomic s-character
in the bonds mediating the coupling in the interior of the
Structural Database contains 383 reported structures contain-
ing aryl C−F···LA interactions. The following analysis pertains
to 17 structures that satisfy the following criteria, as well as
compound 4 reported in this work:
3
dimer (3JH,F range for 2, 3, and 5: 6.6−6.9 Hz). JH,F is
overwhelmingly dominated by the Fermi coupling relayed
through a C−H, C−C, and C−F bond, so the increase in s-
character of the C−F bond raises the Fermi term for the whole
system.
• In solution, the C−F···LA interaction is maintained, and
multisite exchange does not occur. This precludes most
structures with polyfluorinated aryl rings for which ring
rotation is not slow on the NMR time scale.
1
The observed reduction in magnitude of JC,F in the bound
C−F bonds of 4′ relative to their free counterparts is a result of
changes to the PSO part of the coupling value that are larger in
magnitude and opposite in sign to the change in the Fermi
component. Specifically, while the average PSO contribution to
1JC,F in the free C−F bond is negative (PSOiso: −9.8 Hz), the
same value is positive for coordinated C−F bonds (PSOiso: 6.3
• High resolution, interpretable 13C{1H} or 13C NMR
data for both the Lewis acid bound complex and the
“free” complex exists. For macrocycles that engage in
Lewis acid binding, the “free” complex refers to the
neutral macrocycle. For complexes where the Lewis acid
is bound by an anionic ligand, the “free” complex refers
to the protonated ligand or, where unavailable, the
ligand halide. This excludes paramagnetic complexes for
which 13C NMR line widths may exceed 1JC,F, as well as
diamagnetic complexes for which incomplete NMR data
are reported.
1
Hz). Reported isotropic PSO components of JC,F in typical
sp3-hybridized systems (where π-effects are small) are also
positive,26 suggesting that an increasingly positive PSO term
may be indicative of reduced carbon−fluorine π-bonding
character. While this trend has not yet been established for
carbon−fluorine bonds, similar relationships between the value
of the PSO term and multiple-bonding character have been
demonstrated for carbon−carbon bonds.27
We note in passing that although we have excluded systems
exhibiting multisite exchange from our analysis, the 1JC,F values
for a coordinated C−F bond undergoing exchange with free
C−F bonds can be well-approximated by understanding the
observed 1JC,F value as an average. By comparing the observed
coupling constant with the “free” coupling constant, one can
The sum of these considerations begins to suggest that the
1
absolute value of JC,F should necessarily decrease in systems
where Lewis acid binding perturbs a carbon−fluorine π-
interaction (as in aryl fluorides). This proposal, that
coordination of an aryl C−F bond to a Lewis acid should
generally result in a decrease in |1JC,F|, is explored in the
following section.
1
calculate the approximate JC,F value for a coordinated C−F
bond involved in exchange without requiring the exchange
process be slow on the NMR time scale and could in principle
extend this analysis to these more complex systems.
General Relation between Δ1JC,F and C−F Bond
1
Activation. Having established that the reduced JC,F value
Candidate data sets include a diverse array of Lewis acidic
sites, including multiple group 1 and 2 elements (lith-
ium,13,28,29 sodium,13,30,31 potassium,13,32 rubidium,33 cesi-
um,16,34 and calcium35), as well as several transition elements
(zirconium,29,36 hafnium,29 and palladium37). The aggregate
data qualitatively agree with the statement that stronger C−F···
of the lithium-coordinated C−F bonds of 4 was the sum of two
opposing effects, it was still unclear whether or not this
decrease in coupling constant on C−F bond coordination to
Lewis acidic sites was a phenomenon to be generally
anticipated. To further assess this possibility, we conducted a
1
1
survey of reported JC,F values for systems with clear aryl C−
LA interactions lead to lower JC,F values: a simple plot of
1
F···LA interactions. At the time of writing, the Cambridge
observed JC,F versus coordinated C−F bond length shows a
E
https://dx.doi.org/10.1021/acs.inorgchem.0c02516
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
1
downward trend which, when fit to an empirical linear model
Lewis acid, and what change in JC,F this corresponded to.
These results are plotted as contours showing a Mayer bond
order decrease (rounded to the nearest 0.01) resulting directly
from Lewis acid binding in Figure 6 alongside the data
obtained from the literature survey for comparison.
1
(R2 = 0.64), suggests that JC,F values decrease at a rate of
about 6 Hz per 0.01 Å elongation (Figure S4).
1JC,F values depend on both the specific ligand as well as its
coordination state, however, so we next studied the relation-
ship between coordinated C−F bond length and the change in
1JC,F value from the “free” ligand value upon coordination
(Δ1JC,F) as a means of providing some level of referencing.
Plotting Δ1JC,F versus C−F bond length (dC−F) yields little
initial insight, as there is no clear correlation between dC−F and
The results of this calculation show a clear trend: longer
Lewis acid coordinated C−F bonds with large reductions in
1JC,F show the greatest extent of activation or, equivalently, the
strongest C−F···LA interactions. Previous work has frequently
discussed the strength of C−F···LA interactions in the context
of C−F bond elongations, but this work further asserts a
1
change in JC,F (Figure 6). This observation suggested that a
more complex relationship between Δ1JC,F, dC−F, and the
strength of a C−F···LA interaction existed. To elucidate this
relationship, we undertook a systematic computational study,
1
relation between JC,F and the extent of C−F bond activation.
An additional critical finding of this analysis is that “long” C−F
bonds do not necessarily imply greater activation relative to
shorter analogues, and that, for a given C−F bond length,
varying degrees of activation are possible. A simple
demonstration of this statement is found in comparing 4
with analogous semifluorinated m-terphenyl silylium ion
complex [2,6-bis(2,6-difluorophenyl)phenyl]dimethylsilylium
tetrakis(pentafluorophenyl)borate (herein referred to as
[TerphF′-SiMe2][B(C6F5)4]), where a formally cationic silicon
center acts as the Lewis acidic site.21 The Lewis acid bound
C−F bonds in [TerphF′-SiMe2]+ are on average significantly
longer than those in 4 (1.416 Å vs 1.381 Å), which, when
considered alone, suggests that the C−F bonds of [TerphF′-
SiMe2]+ are more activated. The C−F···LA interaction energies
(as gauged by activation energies for ring rotation) in these
two species, however, suggest the opposite: for 4, this value is
approximately 13 kcal/mol, whereas for [TerphF′-SiMe2]+ it
was calculated (the process being fast on the NMR time scale
down to 233 K) to be 4.5 kcal/mol. Discussion of Δ1JC,F values
for each system provides a more complete picture: for
[TerphF′-SiMe2]+, Δ1JC,F is 16 Hz (1JC,F for TerphF′-SiMe2H
utilizing as a model system embodying a representative Caryl
−
F···LA interaction the fluorobenzene-tris(dimethyl ether)
lithium cation ([C6H5F···Li(OMe2)3]+) (Figure 7). This
Figure 7. Depiction of the [C6H5F···Li(OMe2)3]+ system utilized to
1
study changes in JC,F as a function of dC−F and dF−Li. Geometry
optimizations followed by NMR calculations at the BP86/pcsSeg-2
level were performed for C−F bond distances between 1.34 and 1.40
Å in increments of 0.01 Å, and for F···Li distances of 2.0, 2.2, 2.4, 2.6,
is 247 Hz, while JC,F for [TerphF′-SiMe2]+, where only one
1
1
3.0, 3.5, and 5.0 Å with a fixed C−F−Li angle of 135° to yield JC,F
signal is observed for all fluorinated carbons, is 239 Hz;
values and C−F bond Mayer bond orders. Carbon is shown as gray,
1
assuming Lewis acid coordination has little effect on JC,F for
noncoordinated C−F bonds (as observed in 4), we calculate
the JC,F value for the silylium-coordinated C−F bonds to be
fluorine as green, hydrogen as white, lithium as purple, and oxygen as
red.
1
231 Hz), whereas for 4 it is 25 Hz. Qualitatively weighing these
values alongside the C−F bond lengths in the context of the
broad trends presented in Figure 6 suggests instead that the
coordinated C−F bonds in 4 and [TerphF′-SiMe2]+ are both
quite activated, probably to similar extents, in spite of their
differing metrical parameters. This suggests that utilizing C−F
bond lengths as the sole descriptor of C−F···LA interactions is
ill-advised, and that a more complete description of these
interactions should also include a discussion of NMR or
computational data, or a combination of all three.
This point is made more obvious by considering the metric
of C−F bond lengths in a general sense: Analysis of the nearly
29 000 structures in the CSD containing uncoordinated aryl
C−F bonds shows that while each of the 18 structures
analyzed in Figure 6 has a C−F bond length longer than the
average aryl C−F bond (1.347 Å), two fall within one standard
deviation of the average (1.365 Å), and 12 fall within two
standard deviations (1.383 Å), including 4. In general, there is
a significant overlap between reported Lewis acid coordinated
aryl C−F bond lengths and free aryl C−F bond lengths, with
the numeric spread for both data sets being comparable in
magnitude to the differences between each set’s average,
further complicating the use of C−F bond length as a reliable
parameter (Figure 8). This could presumably be ameliorated
by referencing coordinated C−F bond lengths to the same
system was chosen for its chemical simplicity, and there is
little reason to expect that the use of another Lewis acidic
center would significantly change the outcome of this study; it
need only be true that the model system incorporate an aryl
C−F bond acting as a Lewis base and a site with a vacant
orbital acting as a Lewis acid.
To map the electronic landscape of various C−F···LA
interaction strengths, we performed a series of constrained
geometry optimizations, holding the C−F−Li angle fixed at
135° and the F−Li distance at either 2.0, 2.2, 2.4, 2.6, 3.0, 3.5,
or 5.0 Å (these F−Li distances were chosen to afford Mayer
bond order decreases between 0.18 and 0.06, in increments of
approximately 0.02), with the shorter and longer distances
serving as representative cases of strong and weak C−F···LA
interactions, respectively. For each F−Li distance, the C−F
distance was varied between 1.34 and 1.40 Å in increments of
1
0.01 Å, and the JC,F value for each optimized structure was
calculated. Taking the calculated C−F bond Mayer bond order
as a metric for interaction strength, we compared the C−F
bond orders of each computed structure with the correspond-
ing values obtained for the neutral fluorobenzene molecule
with C−F distances similarly varied between 1.34 and 1.40 Å.
This allowed us to determine, for a given C−F bond length,
how activated a bond was as a direct result of interactions with a
F
https://dx.doi.org/10.1021/acs.inorgchem.0c02516
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
system suggests that, in general, decreases in |1JC,F| on aryl C−F
bond coordination to a Lewis acid are indicative of interactions
of at least moderate strength, and that, taken together, C−F
bond lengths and Δ1JC,F values can qualitatively describe the
strength of C−F···Lewis acid interactions more completely
than either metric could in isolation.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.inorgchem.0c02516.
■
sı
Experimental procedures, NMR data, crystallographic
data, and computational data (PDF)
Accession Codes
CCDC 2005694−2005695 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by
emailing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
Figure 8. Histogram showing the frequencies of all C−F bond lengths
for coordinated and free aryl C−F bonds reported in the CSD.
bond lengths in the absence of a Lewis acid, though
crystallographic characterization for these metal-free com-
pounds is less frequently reported and, in some cases (where
the relevant compound is not a solid, for instance), may not be
feasible.
AUTHOR INFORMATION
Corresponding Author
■
1
John Arnold − Department of Chemistry, University of
California, Berkeley, California 94720, United States;
Chemical Sciences Division, Lawrence Berkeley National
Laboratory, Berkeley, California 94720, United States;
orcid.org/0000-0001-9671-227X; Email: arnold@
berkeley.edu
Fewer issues arise when considering changes in JC,F on
coordination as a metric for describing the strength of an aryl
C−F···LA interaction. While it remains true that, for a given
1
decrease in JC,F, the precise strength of the interaction may
1
vary, a decrease in JC,F on coordination is strongly indicative
of the presence of a C−F···LA interaction in solution. This is
due to the fact that, over the chemically relevant range of C−F
bond lengths (between about 1.30 and 1.41 Å), longer,
Authors
Brighton A. Skeel − Department of Chemistry, University of
California, Berkeley, California 94720, United States
Michael A. Boreen − Department of Chemistry, University of
California, Berkeley, California 94720, United States;
Chemical Sciences Division, Lawrence Berkeley National
Laboratory, Berkeley, California 94720, United States;
orcid.org/0000-0001-7325-2717
Trevor D. Lohrey − Department of Chemistry, University of
California, Berkeley, California 94720, United States;
Chemical Sciences Division, Lawrence Berkeley National
Laboratory, Berkeley, California 94720, United States;
orcid.org/0000-0003-3568-7861
1
uncoordinated bonds are expected to yield higher JC,F values
(Figure S5, vide supra).
It is important to note that the converse of this statement is
not universally true, and that Lewis acid binding of aryl C−F
1
bonds cannot be expected to always cause reductions in JC,F
.
This is shown in Figure 6 where it is evident that, for long,
1
relatively weakly activated C−F bonds, the JC,F value is
expected to increase on coordination to a Lewis acidic site. To
the best of our knowledge, this phenomenon has not yet been
conclusively observed experimentally.
We conclude this section by stating that, much like dC−F
,
Δ1JC,F values are not a metric to be used in isolation but,
rather, are an additional parameter that may aid the practicing
chemist in describing C−F···LA interactions they may
encounter in their work. In conjunction with dC−F values,
Δ1JC,F values can offer more insight into the bonding within
Lewis acid coordinated C−F bonding motifs than either value
can alone.
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.inorgchem.0c02516
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
CONCLUSION
Notes
■
The authors declare no competing financial interest.
To summarize, we have synthesized a new, semifluorinated m-
terphenyl lithium featuring close C−F···Li contacts in the
solid-state, and we have shown by multinuclear NMR
spectroscopy that these interactions persist in solution. Eyring
analysis allowed us to establish an upper limit on the enthalpy
of the C−F···Li interaction of 13( 0.8) kcal/mol. Computa-
tional analysis of the NMR J-couplings within the C−F···Li
ACKNOWLEDGMENTS
■
This work was supported by the Director, Office of Science,
Office of Basic Energy Sciences, Division of Chemical Sciences,
Geosciences, and Biosciences Heavy Element Chemistry
Program of the U.S. Department of Energy (DOE) at LBNL
under Contract DE-AC02-05CH11231. The Advanced Light
Source is supported by the Director, Office of Science, Office
of Basic Energy Sciences, of the U.S. Department of Energy
under Contract DE-AC02-05CH11231. B.A.S. acknowledges
1
unit suggest that the characteristic decrease in JC,F is due
predominantly to changes in the extent of π-delocalization of
the fluorine lone pairs into the arene π-system. In-depth
computational analysis of the [C6H5F···Li(OMe2)3]+ model
G
https://dx.doi.org/10.1021/acs.inorgchem.0c02516
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
(19) Blundell, T. J.; Hastings, F. R.; Gridley, B. M.; Moxey, G. J.;
Lewis, W.; Blake, A. J.; Kays, D. L. Ligand Influences on Homoleptic
Group 12 m-Terphenyl Complexes. Dalton Trans 2014, 43, 14257−
14264.
(20) Boreen, M. A.; Parker, B. F.; Lohrey, T. D.; Arnold, J. A
Homoleptic Uranium(III) Tris(aryl) Complex. J. Am. Chem. Soc.
2016, 138, 15865−15868.
the Goldwater Foundation for a scholarship. M.A.B. acknowl-
edges support from the National Science Foundation Graduate
Research Fellowship Program under Grant DGE 1106400.
T.D.L. thanks the U.S. DOE Integrated University Program for
a graduate research fellowship. We thank the UC Berkeley
College of Chemistry’s NMR facility for resources provided
and the staff, in particular Dr. Hasan Celik for assistance and
helpful comments on this manuscript. The instruments in
CoC-NMR are supported in part by NIH S10OD024998. We
thank Dr. Simon J. Teat for training and guidance throughout
our crystallography experiments at the ALS.
(21) Romanato, P.; Duttwyler, S.; Linden, A.; Baldridge, K. K.;
Siegel, J. S. Intramolecular Halogen Stabilization of Silylium Ions
Directs Gearing Dynamics. J. Am. Chem. Soc. 2010, 132, 7828−7829.
(22) Buschmann, H. J.; Hermann, J.; Kaupp, M.; Plenio, H. The
Coordination Chemistry of the CF Group of Fluorocarbons:
Thermodynamic Data and Ab Initio Calculations on CF-Metal Ion
Interactions. Chem. - Eur. J. 1999, 5, 2566−2572.
REFERENCES
■
(23) Gasparro, F. P.; Kolodny, N. H. NMR Determination of the
Rotational Barrier in N,N-Dimethylacetamide. A Physical Chemistry
Experiment. J. Chem. Educ. 1977, 54, 258−258.
(1) Fujiwara, T.; O’Hagan, D. Successful Fluorine-Containing
Herbicide Agrochemicals. J. Fluorine Chem. 2014, 167, 16−29.
(2) Meanwell, N. A. Fluorine and Fluorinated Motifs in the Design
and Application of Bioisosteres for Drug Design. J. Med. Chem. 2018,
61, 5822−5880.
(24) Hogben, H. J.; Krzystyniak, M.; Charnock, G. T. P.; Hore, P. J.;
Kuprov, I. Spinach − A Software Library for Simulation of Spin
Dynamics in Large Spin Systems. J. Magn. Reson. 2011, 208, 179−194.
(3) Purser, S.; Moore, P. R.; Swallow, S.; Gouverneur, V. Fluorine in
Medicinal Chemistry. Chem. Soc. Rev. 2008, 37, 320−330.
(4) Zhou, P.; Zou, J.; Tian, F.; Shang, Z. Fluorine Bonding  How
Does It Work In Protein−Ligand Interactions? J. Chem. Inf. Model.
2009, 49, 2344−2355.
̈
(25) Cremer, D.; Grafenstein, J. Calculation and Analysis of NMR
Spin−Spin Coupling Constants. Phys. Chem. Chem. Phys. 2007, 9,
2791−2816.
́
(26) Gryff-Keller, A.; Szczecinski, P. An Efficient DFT Method of
Predicting the One-, Two- and Three-Bond Indirect Spin−Spin
Coupling Constants Involving a Fluorine Nucleus in Fluoroalkanes.
RSC Adv. 2016, 6, 82783−82792.
(5) Brookhart, M.; Green, M. L. H.; Parkin, G. Agostic Interactions
in Transition Metal Compounds. Proc. Natl. Acad. Sci. U. S. A. 2007,
104, 6908−6914.
(6) Salonen, L. M.; Ellermann, M.; Diederich, F. Aromatic Rings in
Chemical and Biological Recognition: Energetics and Structures.
Angew. Chem., Int. Ed. 2011, 50, 4808−4842.
̈
(27) Grafenstein, J.; Kraka, E.; Cremer, D. Investigation of the π
Character of a C−C Bond with the Help of the Diamagnetic and
Paramagnetic Spin−Orbit Term of the NMR Spin−Spin Coupling
Constant. J. Phys. Chem. A 2004, 108, 4520−4535.
(7) Dougherty, D. A. The Cation−π Interaction. Acc. Chem. Res.
2013, 46, 885−893.
(28) Lutz, M.; Haukka, M.; Pakkanen, T. A.; McPartlin, M.; Gade, L.
H. Lithium Cation Solvation in the Ligand Periphery of an Ortho-
Fluorophenyl Substituted Tripodal Triaminostannate. Inorg. Chim.
Acta 2003, 345, 185−189.
(8) Wagner, J. P.; Schreiner, P. R. London Dispersion in Molecular
ChemistryReconsidering Steric Effects. Angew. Chem., Int. Ed.
2015, 54, 12274−12296.
(9) Liptrot, D. J.; Power, P. P. London Dispersion Forces in
Sterically Crowded Inorganic and Organometallic Molecules. Nat.
Rev. Chem. 2017, 1, 1−12.
(29) Lee, W.-Y.; Liang, L.-C. Fluorinated Diarylamido Complexes of
Lithium, Zirconium, and Hafnium. Inorg. Chem. 2008, 47, 3298−
3306.
(30) Plenio, H.; Diodone, R. The Coordination Chemistry of
Fluorocarbons: Group-I and Group-II Metal Ion Complexes of Fluoro
Macrocycles. Chem. Ber. 1997, 130, 963−968.
(31) Plenio, H.; Hermann, J.; Diodone, R. The Coordination
Chemistry of Fluorocarbons: Difluoro-m-Cyclophane-Based Fluoroc-
ryptands and Their Group I and II Metal Ion Complexes. Inorg. Chem.
1997, 36, 5722−5729.
(10) Mahmudov, K. T.; Kopylovich, M. N.; Guedes da Silva, M. F.
C.; Pombeiro, A. J. L. Non-Covalent Interactions in the Synthesis of
Coordination Compounds: Recent Advances. Coord. Chem. Rev.
2017, 345, 54−72.
̌
́
(11) Cerny, J.; Hobza, P. Non-Covalent Interactions in Bio-
macromolecules. Phys. Chem. Chem. Phys. 2007, 9, 5291−5303.
(12) Ji, X.; Ahmed, M.; Long, L.; Khashab, N. M.; Huang, F.;
Sessler, J. L. Adhesive Supramolecular Polymeric Materials Con-
structed from Macrocycle-Based Host−Guest Interactions. Chem. Soc.
Rev. 2019, 48, 2682−2697.
(32) Takemura, H.; Kon, N.; Yasutake, M.; Kariyazono, H.;
Shinmyozu, T.; Inazu, T. A Potassium Complex of a Fluorine-
Containing Macrocyclic Cage Compound: Interactions between
Fluorine Atoms and Metal Ions. Angew. Chem., Int. Ed. 1999, 38,
959−961.
(13) Plenio, H.; Diodone, R. Covalently Bonded Fluorine as a σ-
Donor for Groups I and II Metal Ions in Partially Fluorinated
Macrocycles. J. Am. Chem. Soc. 1996, 118, 356−367.
(33) Takemura, H.; Iwanaga, T.; Shinmyozu, T. C−F···Rb+
Interaction in a Fluorinated Cage Compound Complex. Tetrahedron
Lett. 2006, 47, 8989−8991.
(14) Plenio, H.; Burth, D. 19F NMR Indicator for Protons and Metal
Ions with Direct Fluorine−Metal Interactions. J. Chem. Soc., Chem.
Commun. 1994, 19, 2297−2298.
(34) Plenio, H.; Diodone, R.; Badura, D. Synthesis and
Coordination Chemistry of Fluorine-Containing Cages. Angew.
Chem., Int. Ed. Engl. 1997, 36, 156−158.
(15) Takemura, H.; Nakashima, S.; Kon, N.; Yasutake, M.;
Shinmyozu, T.; Inazu, T. A Study of C−F···M+ Interaction: Metal
Complexes of Fluorine-Containing Cage Compounds. J. Am. Chem.
Soc. 2001, 123, 9293−9298.
(16) Takemura, H.; Kon, N.; Kotoku, M.; Nakashima, S.; Otsuka,
K.; Yasutake, M.; Shinmyozu, T.; Inazu, T. A Study of C−F···M+
Interaction: Alkali Metal Complexes of the Fluorine-Containing Cage
Compound. J. Org. Chem. 2001, 66, 2778−2783.
(17) Ruhlandt-Senge, K.; Ellison, J. J.; Wehmschulte, R. J.; Pauer, F.;
Power, P. P. Isolation and Structural Characterization of Unsolvated
Lithium Aryls. J. Am. Chem. Soc. 1993, 115, 11353−11357.
(18) Sharpe, H. R.; Geer, A. M.; Williams, H. E. L.; Blundell, T. J.;
Lewis, W.; Blake, A. J.; Kays, D. L. Cyclotrimerisation of Isocyanates
Catalysed by Low-Coordinate Mn(II) and Fe(II) m-Terphenyl
Complexes. Chem. Commun. 2017, 53, 937−940.
(35) Barrett, A. G. M.; Crimmin, M. R.; Hill, M. S.; Hitchcock, P. B.;
Lomas, S. L.; Mahon, M. F.; Procopiou, P. A. Carbodiimide Insertion
Reactions of Homoleptic Heavier Alkaline Earth Amides and
Phosphides. Dalton Trans. 2010, 39, 7393−7400.
(36) Memmler, H.; Walsh, K.; Gade, L. H.; Lauher, J. W. Tripodal
Amido Ligands Containing an “Active” Ligand Periphery. Inorg.
Chem. 1995, 34, 4062−4068.
(37) Scharf, A.; Goldberg, I.; Vigalok, A. Evidence for Metal−Ligand
Cooperation in a Pd−PNF Pincer-Catalyzed Cross-Coupling. J. Am.
Chem. Soc. 2013, 135, 967−970.
(38) Goerigk, L.; Grimme, S. A Thorough Benchmark of Density
Functional Methods for General Main Group Thermochemistry,
H
https://dx.doi.org/10.1021/acs.inorgchem.0c02516
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Kinetics, and Noncovalent Interactions. Phys. Chem. Chem. Phys.
2011, 13, 6670−6688.
I
https://dx.doi.org/10.1021/acs.inorgchem.0c02516
Inorg. Chem. XXXX, XXX, XXX−XXX