Quantifying Through-Space Substituent Effects

Article abstract of DOI:10.1002/anie.202006943

The description of substituents as electron donating or withdrawing leads to a perceived dominance of through-bond influences. The situation is compounded by the challenge of separating through-bond and through-space contributions. Here, we probe the experimental significance of through-space substituent effects in molecular interactions and reaction kinetics. Conformational equilibrium constants were transposed onto the Hammett substituent constant scale revealing dominant through-space substituent effects that cannot be described in classic terms. For example, NO₂ groups positioned over a biaryl bond exhibited similar influences as resonant electron donors. Meanwhile, the electro-enhancing influence of OMe/OH groups could be switched off or inverted by conformational twisting. 267 conformational equilibrium constants measured across eleven solvents were found to be better predictors of reaction kinetics than calculated electrostatic potentials, suggesting utility in other contexts and for benchmarking theoretical solvation models.

Full text of DOI:10.1002/anie.202006943

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Quantifying Through-Space Substituent Effects
Authors: Rebecca J. Burns, Ioulia K. Mati, Kamila B. Muchowska,
Catherine Adam, and Scott Lee Cockroft
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202006943
Link to VoR: https://doi.org/10.1002/anie.202006943

10.1002/anie.202006943
Angewandte Chemie International Edition
RESEARCH ARTICLE
Quantifying Through-Space Substituent Effects
Rebecca J. Burns, Ioulia K. Mati, Kamila B. Muchowska, Catherine Adam and Scott L. Cockroft*[a]
[a]
R. J. Burns, Dr I. K. Mati, Dr K. B. Muchowska, Dr C. Adam, Prof. S. L. Cockroft
EaStCHEM School of Chemistry
University of Edinburgh
Joseph Black Building, David Brewster Road, Edinburgh, EH9 3FJ, United Kingdom.
E-mail: scott.cockroft@ed.ac.uk
Supporting information for this article is given via a link at the end of the document.
Abstract: The description of substituents as electron donating or
withdrawing leads to perceived dominance of through-bond
been described by substituent constants such as F, I and F.[2,
5b, 9]
a
Such was the proliferation of studies aiming to complement
influences. The situation is compounded by the challenge of
separating through-bond and through-space contributions. Here, we
probe the experimental significance of through-space substituent
effects in molecular interactions and reaction kinetics.
Conformational equilibrium constants were transposed onto the
Hammett substituent constant scale revealing dominant through-
space substituent effects that cannot be described in classic terms.
For example, NO2 groups positioned over a biaryl bond exhibited
similar influences as resonant electron donors. Meanwhile, the
electro-enhancing influence of OMe/OH groups could be switched
off or inverted by conformational twisting. 267 conformational
equilibrium constants measured across eleven solvents were found
to be better predictors of reaction kinetics than calculated
electrostatic potentials, suggesting utility in other contexts and for
benchmarking theoretical solvation models.
Hammett’s seminal constants, that at least twenty different
dissected scales had been defined by the 1970s.[8, 9b] Swain, Leo
and Taft criticized many of these dissections for making
incorrect assumptions about the transferability of field effects,
pointing out that field effects are intrinsically spatially dependent
and likely to dominate over through-bond effects for distant
9b]
substituents.[2,
The Kirkwood-Westheimer model can be
credited as being one of the earliest methods for estimating the
10]
geometric influence of substituent-induced dipoles.[3a,
The
model can estimate substituent effects on acid dissociation
constants, but with imperfect applicability.[7a, 9c, 11] More recent
examinations of substituent effects on dissociation constants
have seemingly side-stepped the Kirkwood-Westheimer model
in favor of contemporary computational approaches.[12] Early
work by Topsom showed that through-bond and through-space
substituent effects could be dissected by deleting bonds
separating a substituent and a site of interest.[13] More recently,
Wheeler, Houk and Suresh have found that additive substituent
Introduction
field effects can largely account for the calculated electrostatic
14]
potentials of aromatic rings.[6a,
Such through-space models
Systematic variation of substituents is often exploited to tune or
rationalize chemical behavior and reaction mechanisms.[1]
Substituent effects are usually ranked using relative
are beginning to supersede earlier empirically derived models of
aromatic interactions,[15] and numerous investigations have
highlighted the importance of field effects in enzyme-catalyzed
reactions[16] and synthetic organic chemistry.[17] However,
contrasting with the success of empirically derived Hammett
substituent constants in accounting for reactivity in a wide range
of contexts, high-quality experimental data quantifying through-
space substituent effects are surprisingly limited.
Here we have used 25 molecular balances and 17 pyridine
derivatives to quantify the importance of through-space
substituent effects on molecular interactions and reaction
kinetics, respectively (Figure 1). The experimentally observed
equilibrium and kinetic constants were correlated with calculated
electrostatic potentials (ESPs) to examine the experimental
significance of through-space substituent effects (Figures 2, 3
and 6). Transposing the experimentally observed conformational
equilibrium constants onto the Hammett substituent constant
scale (Figure 2, Table 1) revealed the remarkable extent to
which through-space effects can dominate experimental
behaviour compared to more classically considered through-
bond contributions (Figures 1 and 3). The transferability of
Hammett substituent constants derived from conformational
equilibrium constants was examined across eleven different
solvents (Table 1, Figures 4 to 5), and their ability to account for
through-space influences on the kinetics of a simple model
reaction was assessed (Figure 6).
electronegativities
constants.[2] Although the quantification of substituent effects
has
long history,[3] it was Hammett who defined and
and
empirically
derived
substituent
a
established the transferability, and thus great utility of
quantitative m and p substituent constants determined from the
pKa values of benzoic acid derivatives.[2, 4] However, it was soon
realized that transferrable o constants could not be easily
defined due to interactions occurring between ortho
substituents.[2, 5] Similarly, the lexicon of classifying substituents
as either ‘electron withdrawing’ or ‘electron donating’
emphasizes bond connectivity and the through-bond
contributions of induction and resonance. As a result, there is an
unconscious tendency to overlook the significance of
electrostatic (field) contributions that occur through space.
Indeed, the dissection of through-bond and through-space
contributions to substituent effects has been
challenge.[6]
a long-term
Through-bond resonant contributions often dominate the
behavior of delocalized systems and are therefore relatively
easy to dissect from the combined inductive and field effect (e.g.
R from m/p or R− and R+ from + and −).[2, 4b, 7] In contrast,
inductive and field effects have historically been treated together
due to the difficulty in separating them in experimental
8]
systems.[7a,
Nonetheless, inductive and field effects,
specifically those occurring along the direction of a bond, have
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202006943
Angewandte Chemie International Edition
RESEARCH ARTICLE
prediction of solvent effects remains notoriously difficult.[18]
Therefore, for our experimental investigations in solution, we
instead sought model systems in which through-space effects
would dominate over those occurring through bonds. The
compounds shown in Figure
1
contain biphenyl and
4-phenylpyridine units that position substituents in similar
geometries, allowing comparisons to be drawn between
influences on interactions and reaction kinetics. We reasoned
that through-space effects would be most likely to dominate over
through-bond influences when polar substituents were
positioned ortho to the biaryl bond. Such frameworks also allow
systematic variation of the positioning of substituents and an
assessment of their influence on non-covalent interactions and
chemical reactions occurring at a remote location. We selected
uncharged substituents for this investigation to avoid counterion
and solubility issues during solvent-screening experiments.
Through-space substituent effects on conformational
equilibria
Synthetic molecular balances of the type shown in Figure 1A
were adopted to examine the through-space influence of
substituents on non-covalent interactions. Molecular balances
provide useful tools for such an investigation since substituent
effects perturb the position of a conformational equilibrium in a
quantifiable manner.[19] Indeed, variants of the balances shown
in Figure 1A have previously been used to probe substituent and
solvent effects in carbonyl-carbonyl interactions, and hydrogen
and chalcogen bonds.[20] Slow rotation of the formamide C-N
bond on the NMR timescale allows the integration of discrete 19F
NMR signals corresponding to each conformer. The integral
ratio corresponds to the conformational equilibrium constant, KX,
which constitutes a quantitative assessment of the substituent
effects on interactions occurring within in the balances. Negative
electrostatic potential over the X-substituted ring would repel the
electron-rich carbonyl oxygen and contribute towards
a
preference for the H-conformer in Figure 1. In contrast, positive
electrostatic potential over the X-substituted ring would help to
stabilize the O-conformer. In addition, we reasoned that an
apolar solvent such as benzene would provide the best
opportunity for the manifestation of through-space substituent
effects. Thus, the conformational equilibrium constants, KX for
the 25 molecular balances shown in Figure 1A were determined
in benzene-d6 (Table 1).
The Hammett-style relationship −log10(KX/KH), encodes the
electronic effects of the X substituent on the position of the
conformational equilibrium. These −log10(KX/KH) values are
accordingly positive for classically ‘electron-withdrawing’ para-
substituents (X = Br, CN, NO2, CF3) and negative for ‘electron-
donating’ para-substituents (X = OMe, NEt2). Such substituent
effects are reflected in the calculated electrostatic potentials
taken over the carbon positioned ipso to the conformationally
exchanging formamide (ESPispo, Figure 2A). These electrostatic
potentials correlate strongly with −log10(KX/KH) for all 25
molecular balances in benzene (R2 = 0.92 in Figure 2A).
Figure 1. A) Molecular balances (1−X) and B) pyridine derivatives (2−X) used
in the present investigation to quantify through-space substituent effects on
molecular interactions and reaction kinetics, respectively. The values listed
under the structures of substituents a to n are the Hammett constants, p(conf)
determined from conformational equilibrium constants measured in benzene-
d6 at 298 K (Table 1) using the correlation shown in Figure 2B. Errors in p(conf)
< ±0.08 (see section S4 in SI). Color coding matches the use in subsequent
figures.
Results and Discussion
Design of experimental systems for quantifying through-
space substituent effects
The experimentally determined conformational equilibrium
constants can be transposed onto the established p Hammett
substituent constant scale since these values are known for the
X substituents in the eleven control balances (Figure 2B and
Table 1).[2, 21] Hence, the correlation shown in Figure 2B can be
used to determine the Hammett constants for the more unusual
phenyl substituents a to n (p(conf)). The 3,5-dinitrophenyl
The separation of through-bond and through-space substituent
effects is known to be challenging. While computational methods
are extremely useful in facilitating the quantitative dissection of
substituent effects, they often model situations that are difficult,
or even impossible to probe experimentally (e.g. functional
groups held in specific spatial orientations, or functional group
dissections that do not exist). Similarly, the computational
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202006943
Angewandte Chemie International Edition
RESEARCH ARTICLE
sterically twist the OMe groups such that the oxygen lone pairs
no longer point over the adjacent phenyl ring. Accordingly, the
ESPipso value over the bridging phenyl ring in 1−b was similar to
of the control compound 1−H, which does not contain any OMe
substituents (63 vs. 70 kJ mol−1, Figure 3C).
Remarkably, removing the capping methyl groups from
compound 1−b to give the dihydroxyphenyl compound 1−c
further shifted p(conf) from +0.03 to +0.54 (Figure 3C, right). The
2,6-dihydroxyphenyl substituent in compound 1−c, thus displays
a
similar electronic influence as the strongly electron-
withdrawing trifluoromethyl group in control compound 1−CF3
substituent in compound 1−e (purple, Figure 3B), which
positions two nitro groups meta to the biphenyl bond, was found
to have the most positive p(conf) = +1.00. Such behavior would
Figure 2. A) Correlation between the calculated electrostatic potential in the
position indicated (ESPipso) and the conformational equilibrium constants
determined in benzene-d6 at 298 K for the 1−X series of 25 molecular
balances shown in Figure 1A. ESPs were calculated using B3LYP/6-31G* on
the 0.002 electron/Bohr3 isosurface. Electrostatic potentials determined using
isolated (proton-capped) X-substituents (i.e. without through-bond
contributions) also correlated highly with the experimental data (R2 = 0.89,
Figure S1C). B) Correlation between known p Hammett substituent constants
and conformational equilibrium constants of balances 1−X determined in
benzene-d6 at 298 K. Errors in −log10(KX/KH) are < ±0.08 (section S4 in SI).
classically be described as more electron-withdrawing than the
directly connected nitro substituent in compound 1−NO2 (p(conf)
= +0.90). Strikingly, moving the two nitro groups such that they
are positioned over the biaryl bond in compound 1−d, results in
a very large change in the determined Hammett substituent
constant (p(conf) = −0.17, orange in Figure 3B). Part of this
difference in p(conf) can be attributed to the addition of a tert-
butyl group, but the p = −0.20 for a tert-butyl group,[2] which is
even further diminished by the intermediary phenyl ring, makes
a minor contribution to the total change in p(conf) of −1.17
between 1−d and 1−e.
Calculated electrostatic potentials (ESPs) provide insight
into the large changes in substituent effects upon repositioning
the nitro groups (Figures 3A−B). There are large differences in
the ESP values taken over the ring bridging between the X
substituent and the formyl group due to geometric differences in
the through-space influences of the nitro-groups (ESPipso = −66
vs. −16 kJ mol−1, Figure 3B). The major difference here arises
from the electron-rich oxygen atoms of nitro groups being
positioned over the bridging ring in compound 1−d. Indeed,
despite the large electrostatic change arising from differences in
the proximity of the nitro oxygen atoms, these ESPipso values
remain excellent predictors of the conformational preferences of
the balances (orange and purple points in Figure 2).
Strong through-space substituent effects are not limited to
the nitro group. Given a p(conf) = −0.85, it might be tempting to
describe the 2,6-dimethoxyphenyl substituent in 1−a as being
even more ‘electron donating’ than the amino group in 1−NEt2
(p(conf) = −0.66, red in Figure 3C). Such striking through-space
electronic influences likely account for the prevalence of
proximally positioned OR and NR groups in ligands widely
exploited in catalysis (e.g. SPhos, SagePhos, BI-DIME).[22]
Interestingly, the strong electro-enhancing influence of the 2,6-
dimethoxy groups was found to be completely switched off in
compound 1−b (p(conf) = +0.03, green in Figure 3C). This effect
arises from the installation of ortho tert-butyl groups, which
Figure 3. A) Calculated electrostatic potential slice showing electro-enhanced
(−ve) and electro-attenuated (+ve) regions in space surrounding
nitrobenzene. B) Experimentally determined Hammett substituent constants
p(conf) quantified using the conformational preferences of series 1−X
demonstrate switching from electro-enhancing to electro-attenuating behavior
upon changing the orientation of a nitro group. C) The strongly electro-
enhancing behavior of methoxy groups (left) can be switched off via
conformational twist induced by adjacent tert-Bu groups (center). In contrast,
hydroxyl groups in the same position exert strong electro-attenuating
a
a
influence (right). Electrostatic potentials are scaled from −100 kJ mol−1 (red) to
+100 kJ mol−1 (blue). Indicated electrostatic potential values correspond to
ESPipso as defined in Figure 2A at the positions indicated with arrows.
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202006943
Angewandte Chemie International Edition
RESEARCH ARTICLE
(p(conf) = +0.47). The electrostatic slices in Figure 3C show that
the effect arises due to the electron-poor protons of the OH
groups being positioned over the bridging phenyl ring in
compound 1−c.
to the control compounds 1−H, 1−Me, and 1−Ph. Such
findings also rules out electronic changes arising from ‘steric
pressure’.[24]
Our observations are in accord with Wheeler and Houk’s
proposal that through-space field effects, and not through-bond
polarization of the aryl -system, are the major determinants of
non-covalent interactions involving aromatic rings.[14a-c] Such a
hypothesis is supported by NMR spectroscopic analysis of the
molecular balances depicted in Figure 1. Substituent-induced
changes in the electron density of the central biphenyl ring
would be anticipated to manifest as chemical shift changes.[25]
However, in stark contrast with the strong electrostatic
correlations in Figures 2A and S1C, no correlation is observed
between the chemical shifts of the protons on the central
biphenyl ring and the conformational preferences of the
balances (Table S4). Moreover, application of Wheeler and
Houk’s deletion and proton-capping approach[14] reveals that the
ESPipso values utilized in Figures 2A and S1C correlate strongly
with the electrostatic potential determined for the isolated
(proton-capped) X-substituents at the point in space occupied by
the formyl oxygen in the O conformer of each balance (R2 = 0.94,
Figure S1D). Similarly, the same through-space electrostatic
potentials, in which through-bond contributions are definitively
absent, also correlated well with the experimental
conformational equilibrium constants (R2 = 0.89, Figure S1C).
Having convinced ourselves of the dominance of through-
space substituent effects over through-bond polarization of the
aryl -system, we next sought to examine whether such
dominance is manifested in other contexts; namely solvent and
reactivity influences.
Across the series of 1−a to 1−c, ESPipso on the bridging ring
changes by 78 kJ mol−1 and p(conf) changes by ~1.4 units
(Figure 3C and 2A). Notably, this trend runs counter to
expectations based on traditional through-bond considerations;
the sums of the Hammett constants for the substituents bonded
to the terminal phenyl ring equal −0.54, −0.74, and −0.94 for
balances 1−a, 1−b, and 1−c, respectively. Specifically, 1−c,
which contains the strongest through-bond electron donors
behaves like a strong ‘electron withdrawing’ group, whereas 1−a,
which contains the weakest through-bond electron donors is the
only compound of this trio that actually behaves like an ‘electron
donor’.[2],[23]
The observation of substituent effects running counter to
through-bond expectations, combined with the ability to switch
such substituent effects on or off via conformational change, is
consistent with through-space field effects playing a dominant
role in governing the conformational preferences of the balances
shown in Figure 1A. This hypothesis is supported by several
additional observations:
i) Hammett constants determined in the mono-ortho nitro and
methoxy compounds 1−f and 1−g were approximately half
that of the corresponding di-ortho-substituted analogs 1−d
and 1−a (p(conf) = −0.11 and −0.37 vs. −0.17 and −0.85).
Again, the negative sign of p(conf) for the ortho-nitrophenyl
substituent runs counter to the traditional expectation that a
nitro group is strongly electron withdrawing.
ii) Positioning an amino group over the biaryl bond (1−n) gave
p(conf) = +0.11, with the through-space + charge of the NH
protons apparently overcoming through-bond resonant donor
ability of the nitrogen (Figure S25).
iii) The meta-methoxyphenyl substituent in 1−k (p(conf) = −0.31)
exhibited an electro-enhancing influence even though meta-
methoxy groups normally have net electron-withdrawing
character (p = +0.12).[2] Surprisingly, this effect is not
Solvent effects on through-space substituent effects
The success of Hammett substituent constants for rationalizing
electronic effects can be attributed to their established
transferability. Hammett constants are often applicable in
contexts far beyond the original defining system (pKa values of
benzoic acids in aqueous solution), even though the electronic
effects of substituents can be modulated by the solvent.[20a, 21c]
observed for the meta-nitrophenyl substituent (1−j). However, The fluorophenyl group contained within the molecular balances
ESP slices reveal that the bridging phenyl ring is located in
the electro-attenuated region when bonded to a meta-
nitrophenyl group, but in the electro-enhanced region for the
meta-methoxyphenyl example (Figures S21-S22).
employed in the present study (Figure 1) enable systematic
screening of the solvent influences on substituent effects. Hence,
the
−log10(KX/KH) values for 267 substituent/solvent
combinations were determined using 19F NMR spectroscopy
spanning 25 molecular balances in eleven different solvents
(Table 1). Many of the −log10(KX/KH) values of the simple control
balances appear to be relatively insensitive to the solvent (the
iv) The para-substituted phenyl rings of balances 1−h and 1−i
project their NO2 and OMe substituents along the same axis
as the simple control balances 1−NO2 and 1−OMe.
Accordingly, the signs and magnitude of their respective
p(conf) values (+0.72 and −0.15) are consistent with their
substituent effects being projected in the same direction, but
over a greater distance than the aforementioned control
balances (p(conf) = +0.72 vs. +0.90, and −0.15 vs. −0.25).
v) Apolar methyl and ethyl groups positioned ortho to the
biphenyl bond (1−l and 1−m) gave similar substituent effects
gray points in Figure
4 lie close to the 1:1 line). The
−log10(KX/KH) values for the most polar substituent/solvent
combinations deviate most strongly from the 1:1 line. For
example, outliers in dimethylsulfoxide (DMSO) include
compound 1−a (red point), and those containing good H-bond
donors (ringed in Figure 4C).
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202006943
Angewandte Chemie International Edition
RESEARCH ARTICLE
Figure 4. Effect of increasing solvent polarity on the conformational equilibrium constants −log10(KX/KH) determined at 298 K for the 1−X series of molecular
balances shown in Figure 1. Data for eleven solvents are reported in Table 1.
Table 1. Relative conformational equilibrium constants −log10(KX/KH) determined using the molecular balances shown in Figure 1 in eleven solvents at 298 K. The
−log10(KX/KH) values determined in C6D6 were transposed onto the standard Hammett p scale using the calibration graph shown in Figure 2B, and the resulting

p(conf) values listed under the structures shown in Figure 1B.
−log10(KX/KH)
MeCN-d3
Compound
C6D6
DMSO-d6
Acetone-d6
EtOAc[a]
THF-d8
CDCl3
DCM-d2
EtOHa
MeOH-d4
Diethyl
ether[a]
1−NEt2
1−OMe
1−H
−0.12
−0.01
0.00
−0.01
+0.07
0.00
−0.12
+0.02a
0.00
−0.19
+0.03
0.00
−0.21[a]
−0.03 [a]
0.00[a]
+0.01
−0.04
+0.05
0.00
−0.08
+0.06
0.00
−0.05[a]
+0.04[a]
0.00[a]
+0.07
−0.12
+0.02
0.00
−0.04
+0.05[a]
0.00
−0.01
+0.03
+0.14
+0.13
n.r.[d]
−0.23
−0.03
0.00
+0.04
+0.04
+0.12
+0.13
+0.17
+0.19
+0.26
+0.31
1−Me
−0.02
+0.03
+0.13
+0.11
+0.13
+0.15
+0.03
+0.07
−0.03
+0.05
+0.15
+0.16
+0.24
+0.15
+0.20[a]
+0.23[a]
+0.05
+0.07
+0.19
+0.22
+0.19
+0.27
+0.30
+0.35
−0.02
+0.03
+0.14
+0.11
+0.11
+0.19
+0.11
+0.14
−0.05
+0.08
+0.20
+0.21
+0.21
+0.28
+0.38
+0.44
−0.02
+0.05
+0.14
+0.15
+0.17
+0.16
+0.21
+0.25
−0.03
+0.08
+0.19
+0.26
+0.25
+0.39
+0.42
+0.54
1−Ph
+0.04[a]
+0.15
+0.04
1−F[b]
+0.14
1−Br
+0.20[a]
+0.17
+0.16[a]
+0.17
1−COCH3
1−CF3
1−CN
+0.22
+0.22
n.r.
+0.17
+0.17
+0.27[a]
+0.31[a]
+0.27[a]
+0.32[a]
1−NO2
1−a
1−b
1−c
1−d
1−e
1−f
1−g
1−h
1−i
1−j
1−k
1−l
1−m
1−n
−0.18
+0.07
+0.21
+0.01
+0.34
+0.03
−0.04
+0.26
+0.02
+0.20
−0.03
0.00
−0.01
+0.03
−0.07
+0.02
+0.10
+0.04
−0.08
+0.17
+0.03
+0.21
+0.04
+0.02
−0.02
+0.01
−0.06
−0.04
+0.17
+0.03
+0.14
+0.11
0.00
+0.18
+0.10
+0.18
+0.04
+0.05
+0.04
+0.06
−0.03
+0.09
+0.19
+0.09
+0.25
+0.09
+0.02
+0.23
+0.07
+0.20
+0.08
+0.08
+0.10
+0.03
−0.07
+0.04
+0.18
+0.02
+0.20
+0.07
−0.02
+0.16
+0.02
+0.18
+0.05
+0.06
−0.01
+0.10
−0.01
+0.01
+0.17
+0.01
+0.13
+0.06
+0.02
+0.20
−0.03
+0.20
+0.03
−0.01
+0.03
+0.08
−0.19
−0.02
+0.33
+0.02
+0.31
+0.13
−0.09
+0.23
+0.08
+0.21
+0.01
−0.02
+0.01
+0.12
+0.03
−0.03
+0.15
+0.09
+0.21
+0.07
−0.01
+0.16
+0.03
+0.19
−0.02
−0.04
0.00
−0.06
n.s.[c]
+0.30
n.s.
−0.06
+0.06
+0.20
+0.03
+0.19
+0.02
+0.02
+0.20
+0.03
+0.22
+0.01
0.00
−0.27
+0.06
+0.29
+0.09
n.s.
+0.13
0.00
+0.21
+0.09
n.s.
+0.13
+0.11
−0.02
+0.03
n.s.
+0.10
−0.01
+0.18
n.r.
+0.20
+0.04
+0.04
+0.07
+0.05
+0.01
+0.09
−0.02
+0.11
+0.04
[a] Values obtained in non-deuterated solvent. [b] Hypothetical compound, KX = 1 due to symmetry. [c] n.s. = insufficient solubility. [d] n.r = distinct conformer
peaks not resolved by 19F or H NMR at 298K.
Hunter’s / hydrogen-bond model has found use in
accounting for the influence of solvents on the conformational
26]
preferences of molecular balances.[20b,
We used the same
model employed in this previous work to rationalize that the
conformational free energies G of the formyl balances would
be governed by: i) differences in the intramolecular interactions
between the O- and H-conformers (E), and ii) the change in
Boltzmann averaged hydrogen-bond donor and acceptor
constants between each conformer ( and ), as represented
in Figure 5A. The 267 experimental conformational free energies
(determined from G = −RTlnKX as the s and s solvent H-bond
constants were varied) were fitted to the equation in Figure 5A
(Section S7 in SI, Figure S40, R2 = 0.85). Pleasingly, the
dissected solvent-independent E values (Table S10) gave an
improved correlation against the electrostatic potential over the
ipso-carbon (ESPipso, Figure 5B, R2 = 0.95) compared to the
Figure 5. A) Energetic contributions to the difference in free energy between
two conformations of a molecular balance, G where solvophobic effects are
negligible. EO and EH correspond to the intramolecular interactions in the O-
and H-conformers respectively; , H, S, , H and S are the hydrogen-
bond donor () and acceptor constants () of the O- / H-conformers and the
solvent, respectively.21a, 27 B) Correlation of calculated electrostatic potentials
over the ipso-carbon ESPipso vs. the solvent-independent intramolecular
interaction energy difference E = EH − EO dissected using the same solvation
model.
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202006943
Angewandte Chemie International Edition
RESEARCH ARTICLE
prior correlation against equilibrium constants determined in
benzene-d6 (Figure 2A, R2 = 0.92). The improved correlation
reaffirms the dominance of electrostatics in determining the
conformational preferences of these molecular balances, while
also showing that a simple empirical solvation model can
partially account for the attenuating influence of the surrounding
solvent.
Through-space substituent effects on reaction kinetics
Having examined the transferability of the through-space
substituent effects in different solvents, we next sought to
determine the significance of such effects on reaction kinetics.
Inspired by previous work,[27] we selected the N-methylation of
substituted pyridines with methyl iodide as a model reaction
(Figure 1B). Such an investigation presented several
experimental challenges. Firstly, the synthesis of phenyl pyridyl
derivatives proved to be more challenging than the equivalent
biphenyl molecular balances, particularly for the sterically
hindered and electron-poor examples. Secondly, certain
substituents and some polar solvents could not be used since
they react with methyl iodide. Additionally, inert apolar solvents
did not offer good solubility across the full range of pyridine
derivatives and N-methylated iodide products that we sought to
examine. Eventually, we settled on acetone-d6 as our solvent of
choice for this part of our investigation. The rate constants kX of
N-methylation at 298 K were determined under pseudo-first
order conditions as the X-substituent was varied. 1H NMR
spectroscopy was used to monitor the relative integrals of the
pyridine derivative and its N-methylated product as the reaction
progressed (Section S11 in the SI).
The through-space substituent effects on the chemical
reactivity of compound series 2−X mirrored several important
trends observed in the molecular balance series 1−X:
i) The experimentally determined rate constants correlate with
electrostatic potentials as the X-substituent was varied
(Figure 6A, R2 = 0.85, Tables S14-S15). The most reactive
control compound was 2−OMe (50% completion in ~35
minutes), while 2−CN was the least reactive (40+ hours to
50% completion).
Figure 6. A) Relationship between calculated electrostatic potentials taken
over the nitrogen atom (ESPN) and the N-methylation of the 17 pyridine
derivatives shown in Figure 1B in acetone-d6 at 298 K. ESPs were calculated
using B3LYP/6-31G* on the 0.002 electron/Bohr3 isosurface. B) Correlation of
electrostatic potentials in X-substituted phenyl derivatives (ESPipso
) vs.
corresponding X-substituted pyridine derivatives (ESPN). C) Correlation of
conformational equilibrium constants measured in the 1−X balance series vs.
rate constants for the N-methylation of correspondingly substituted 2−X
pyridine derivatives, when both sets of measurements were performed in
acetone-d6. D) Improved correlations were found between rate constants
measured in acetone-d6 and conformational equilibrium constants measured in
five other solvents including tetrahydrofuran (R2 = 0.88 to 0.94, Figures S38-
S39). All experiments were performed at 298K.
These common patterns are consistent with some
transferability of through-space substituent effects upon varying
the X-substituent. Supporting this assertion, the calculated
electrostatic potentials taken over the ipso-carbon in molecular
balances (ESPipso
) was found to correlate strongly with
electrostatic potential of the pyridine nitrogen atom (ESPN)
(Figure 6B, R2 = 0.97). Despite the numerous commonalities
outlined above, the −log10(KX/KH) values determined using
molecular balances correlated surprisingly poorly with the
−log10(kX/kH) pyridine rate data when both sets of data were
determined in acetone-d6 (R2 = 0.66, Figure 6C). This suggests
that energetic influence of the solvent on conformational
preferences differs from those encountered in the transition state
of the N-methylation reaction. However, conformational
−log10(KX/KH) values determined in six of the eleven solvents
examined were found to correlate more strongly with
−log10(kX/kH) (R2 = 0.88 to 0.94, Figures 6D, S80 to S81) than
the electrostatic potential of the pyridine nitrogen calculated in
the gas phase (R2 = 0.85, Figure 6A). Hence, given that the
influence of the solvent on through-space substituent effects is
both complex and computationally challenging, the empirically
determined conformational −log10(KX/KH) values compiled in
Table 1 may prove to be useful in broader chemical contexts.
ii) The dimethoxyphenyl derivative 2−a (red, Figure 6A), which
positions two OMe groups over the pyridine ring reacted with
a similar rate to the most reactive control compound 2−OMe
(4-methoxypyridine).
iii) The ortho-nitrophenyl compound 2−f (blue, Figure 6A),
which positions
a nitro group in an electro-enhancing
position over the pyridine ring attained 50% completion in
3.5 hours vs. 5.5 hours for the electro-attenuating para- and
meta-substituted nitrophenyl compounds 2−h and 2−j.
iv) The electronically neutral dimethyl- and diethylphenyl
derivatives (2−m and 2−l) exhibited similar reactivity to
pyridine (2−H, 50% completion in 2 hours).
v) The dinitrophenyl derivative 2−e drops below the line of best
fit (purple in Figure 6A cf. 1−e in Figure 4C). This effect
presumably arises in both systems due to solvation of the
polar aromatic edge with an electro-enhancing H-bond
acceptor solvent.
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202006943
Angewandte Chemie International Edition
RESEARCH ARTICLE
Conclusion
Keywords: • Substituent effects • Noncovalent interactions •
Electrostatic interactions
In conclusion, we have used synthetic molecular balances and a
simple model reaction to quantify the importance of through-
space substituent effects on non-covalent interactions and
reaction kinetics, respectively (Figure 1). Experimentally
determined conformational equilibrium constants measured in
the molecular balances were transposed onto Hammett’s well-
known substituent constant scale (Table 1, Figure 2). The
determined Hammett constants bring to our attention both the
magnitude of the field effects, and the inadequacy of describing
substituent effects in terms of ‘electron donation’ and ‘electron
withdrawal’. For example, a 2,6-nitrophenyl group that positions
two nitro groups over a biaryl bond were found to have a net
electro-enhancing influence comparable to that of a directly
bonded OMe substituent, and contrasting with the classically
accepted electron-withdrawing nature of nitro groups (Figure 3B).
A more extreme manifestation of through-space effects was
observed with the 2,6-dimethoxylphenyl substituent, which was
found to be ~28% more electro-enhancing than a directly
bonded NEt2 group (Figure 3C, left). Remarkably, it was possible
to completely switch off the electro-enhancing behavior via a
sterically induced conformation change of the orientation of the
oxygen-lone pairs (Figure 3C, middle). Meanwhile, OH protons
pointed over the biaryl bond was found to exert a strong electro-
attenuating influence comparable to that a CF3 group (Figure 3C,
right). The switchable nature of these substituent effects
indicates these remarkable influences are manifested through
space (i.e. via electric fields) and not by through-bond electron
[1]
The Investigation of Organic Reactions and Their Mechanisms, (Ed.: H.
Maskill), Blackwell Publishing Ltd, Oxford, 2006.
[2]
[3]
C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165-195.
a) L. M. Stock, J. Chem. Educ. 1972, 49, 400; b) C. G. Derick, J. Am.
Chem. Soc. 1911, 33, 1162-1167; c) C. G. Derick, J. Am. Chem. Soc.
1911, 33, 1167-1181; d) C. G. Derick, J. Am. Chem. Soc. 1911, 33,
1181-1189.
[4]
[5]
a) L. P. Hammett, Chem. Rev. 1935, 16, 125-136; b) L. P. Hammett, J.
Am. Chem. Soc. 1937, 59, 96-103; c) L. P. Hammett, Trans. Faraday
Soc. 1938, 34, 156-165.
a) C. H. Suresh, P. Alexander, K. P. Vijayalakshmi, Saj, Phys. Chem.
Chem. Phys. 2008, 10, 6492-6499; b) C. G. Swain, S. H. Unger, N. R.
Rosenquist, M. S. Swain, J. Am. Chem. Soc. 1983, 105, 492-502; c) F.
B. Sayyed, C. H. Suresh, Tetrahedron Lett. 2009, 50, 7351-7354; d) C.
B. Santiago, A. Milo, M. S. Sigman, J. Am. Chem. Soc. 2016, 138,
13424-13430.
[6]
[7]
a) F. B. Sayyed, C. H. Suresh, G. S. R., J. Phys. Chem. A 2010, 114,
12330-12333; b) O. Exner, Z. Friedl, in Progress in Physical Organic
Chemistry, Vol. 19 (Ed.: R. W. Taft Jr), John Wiley & Sons, 1993, pp.
259-294.
a) J. D. Roberts, W. T. Moreland Jr, J. Am. Chem. Soc. 1953, 75, 2167
- 2173; b) H. C. Brown, Y. Okamoto, J. Am. Chem. Soc. 1958, 80,
4979-4987; c) H. H. Jaffé, Chem. Rev. 1953, 53, 191-261.
M. Charton, in Progress in Physical Organic Chemistry, Vol. 13 (Ed.: R.
W. Taft), John Wiley & Sons, 1981, pp. 119-251.
[8]
[9]
a) O. Exner, M. Charton, V. Galkin, J. Phys. Org. Chem. 1999, 12, 289-
289; b) C. G. Swain, E. C. Lupton Jr, J. Am. Chem. Soc. 1968, 90,
4328-4337; c) C. F. Wilcox, C. Leung, J. Am. Chem. Soc. 1968, 90,
336-341.
[10] a) F. H. Westheimer, J. G. Kirkwood, J. Chem. Phys. 1938, 6, 513-517;
b) J. G. Kirkwood, F. H. Westheimer, J. Chem. Phys. 1938, 6, 506-512.
[11] a) C. L. Liotta, W. F. Fisher, G. H. Greene, B. L. Joyner, J. Am. Chem.
Soc. 1972, 94, 4891-4897; b) E. J. Grubbs, R. Fitzgerald, R. E. Phillips,
R. Petty, Tetrahedron 1971, 27, 935-944; c) R. Golden, L. M. Stock, J.
Am. Chem. Soc. 1972, 94, 3080-3088; d) S. Acevedo, K. Bowden, J.
Chem. Soc., Chem. Commun. 1977, 608-609.
donation or withdrawal.
A total of 267 substituent/solvent
combinations were determined for 25 molecular balances in
eleven different solvents to examine the context dependency of
through-space substituent effects (Table 1). As anticipated, polar
solvent were found to attenuate the magnitude of substituent
effects projected through space (Figures 4 and 5). Nonetheless,
the same substituent effects that governed the conformations of
the molecular balances were also found to govern the N-
methylation kinetics of correspondingly substituted pyridine
derivatives (Figure 6). Even though the substituents examined
were all neutral and uncharged, the through-space kinetic
influences were still large enough to be manifested in the polar
solvent acetone. Conformational equilibrium constants
measured in several solvents correlated better with the
experimental rate data than electrostatic potentials calculated in
the gas phase. This suggests that the data compiled in Table 1
may prove useful for rationalizing through-space substituent
effects in other contexts and chemical reactions. Similarly, the
simplicity of the model system combined with the high energetic
precision of the conformational equilibrium constants measured
across a range of solvents (equivalent of ±0.2 kJ mol−1) may also
prove useful for benchmarking emerging theoretical solvation
models.
[12] a) C.-T. Chen, J. S. Siegel, J. Am. Chem. Soc. 1994, 116, 5959-5960;
b) J. S. Padial, R. de Gelder, C. Fonseca Guerra, F. M. Bickelhaupt, J.
Mecinović, Chem. Eur. J. 2014, 20, 6268-6271; c) V. Bosmans, J.
Poater, R. Hammink, P. Tinnemans, F. M. Bickelhaupt, J. Mecinović, J.
Org. Chem. 2019, 84, 3632-3637; d) J. Simó Padial, J. Poater, D. T.
Nguyen, P. Tinnemans, F. M. Bickelhaupt, J. Mecinović, J. Org. Chem.
2017, 82, 9418-9424.
[13] a) R. D. Topsom, J. Am. Chem. Soc. 1981, 103, 39-44; b) S. Marriott, R.
D. Topsom, J. Am. Chem. Soc. 1984, 106, 7-10; c) S. Marriott, R. D.
Topsom, J. Chem. Soc., Perkin Trans. 2 1985, 697-698.
[14] a) S. E. Wheeler, K. N. Houk, J. Am. Chem. Soc. 2008, 130, 10854-
10855; b) S. E. Wheeler, K. N. Houk, J. Chem. Theory Comput. 2009, 5,
2301 - 2312; c) S. E. Wheeler, Acc. Chem. Res. 2013, 46, 1029-1038;
d) S. E. Wheeler, J. W. G. Bloom, J. Phys. Chem. A 2014, 118, 6133-
6147; e) F. B. Sayyed, C. H. Suresh, J. Phys. Chem. A 2011, 115,
5660-5664.
[15] a) F. Cozzi, M. Cinquini, R. Annunziata, T. Dwyer, J. S. Siegel, J. Am.
Chem. Soc. 1992, 114, 5729-5733; b) F. Cozzi, M. Cinquini, R.
Annuziata, J. S. Siegel, J. Am. Chem. Soc. 1993, 115, 5330-5331; c) F.
Cozzi, F. Ponzini, R. Annunziata, M. Cinquini, J. S. Siegel, Angew.
Chem. 2017, 129, 11405-11410; Angew. Chem. Int. Ed. 1995, 34,
1019-1020; d) C. A. Hunter, J. K. M. Sanders, J. Am. Chem. Soc. 1990,
112, 5525-5534; e) R. M. Parrish, C. D. Sherrill, J. Am. Chem. Soc.
2014, 136, 17386-17389.
Acknowledgements
We thank the EPSRC (RJB, IKM, KM EP/H021620/1), the
Leverhulme Trust (Philip Leverhulme Prize, SLC) and Pfizer Ltd
for funding (CA).
[16] a) A. Warshel, Acc. Chem. Res. 1981, 14, 284-290; b) A. Warshel, P. K.
Sharma, M. Kato, Xiang, Chem. Rev. 2006, 106, 3210-3235; c) S.
Shaik, D. Mandal, R. Ramanan, Nat. Chem. 2016, 8, 1091-1098; d) S.
D. Fried, S. G. Boxer, Ann. Rev. Biochem. 2017, 86, 387-415.
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202006943
Angewandte Chemie International Edition
RESEARCH ARTICLE
[17] a) Z. Wang, D. Danovich, R. Ramanan, S. Shaik, J. Am. Chem. Soc.
2018, 140, 13350-13359; b) S. Shaik, R. Ramanan, D. Danovich, D.
Mandal, Chem. Soc. Rev. 2018, 47, 5125-5145; c) Y. Zhao, Y. Cotelle,
L. Liu, J. López-Andarias, A.-B. Bornhof, M. Akamatsu, N. Sakai, S.
Matile, Acc. Chem. Res. 2018, 51, 2255-2263; d) A. J. Neel, M. J.
Hilton, M. S. Sigman, F. D. Toste, Nature 2017, 543, 637-646; e) W. A.
Golding, R. Pearce-Higgins, R. J. Phipps, J. Am. Chem. Soc. 2018, 140,
13570-13574; f) W. A. Golding, R. J. Phipps, Chem. Sci. 2020, 11,
3022-3027; g) S. Shaik, S. P. de Visser, D. Kumar, J. Am. Chem. Soc.
2004, 126, 11746-11749; h) J. Joy, T. Stuyver, S. Shaik, J. Am. Chem.
Soc. 2020, 142, 3836-3850.
[18] J. J. Varghese, S. H. Mushrif, Reaction Chemistry & Engineering 2019,
4, 165-206.
[19] a) I. K. Mati, S. L. Cockroft, Chem. Soc. Rev. 2010, 39, 4195-4205; b) L.
Yang, C. Adam, G. S. Nichol, S. L. Cockroft, Nat. Chem. 2013, 5, 1006-
1010; c) B. W. Gung, B. U. Emenike, M. Lewis, K. Kirschbaum, Chem.
Eur. J. 2010, 16, 12357-12362; d) F. Diederich, F. H. Allen, P. A. Wood,
F. R. Fischer, Proc. Natl. Acad. Sci. USA 2008, 105, 17290-17294; e) J.
Luccarelli, I. M. Jones, S. Thompson, A. D. Hamilton, Org. Biomol.
Chem. 2017, 15, 9156-9163; f) J. Hwang, P. Li, E. C. Vik, I. Karki, K. D.
Shimizu, Org. Chem. Front. 2019, 6, 1266-1271 ; g) P. Li, E. C. Vik, J.
M. Maier, I. Karki, S. M. S. Strickland, J. M. Umana, M. D. Smith, P. J.
Pellechia, K. D. Shimizu, J. Am. Chem. Soc. 2019, 141, 12513-12517;
h) E. C. Vik, P. Li, P. J. Pellechia, K. D. Shimizu, J. Am. Chem. Soc.
2019, 141, 16579-16583; i) J. Hwang, P. Li, W. R. Carroll, M. D. Smith,
P. J. Pellechia, K. D. Shimizu, J. Am. Chem. Soc. 2014, 136, 14060-
14067; j) M. R. Ams, M. Fields, T. Grabnic, B. G. Janesko, M. Zeller, R.
Sheridan, A. Shay, J. Org. Chem. 2015, 80, 7764-7769; k) S. L. McKay,
B. Haptonstall, S. H. Gellman, J. Am. Chem. Soc. 2001, 123, 1244-
1245; l) R. Wechsel, M. Žabka, J. W. Ward, J. Clayden, J. Am. Chem.
Soc. 2018, 140, 3528-3531.
[20] a) K. B. Muchowska, C. Adam, I. K. Mati, S. L. Cockroft, J. Am. Chem.
Soc. 2013, 135, 9976 - 9979; b) I. K. Mati, C. Adam, S. L. Cockroft,
Chem. Sci. 2013, 4, 3965-3972; c) D. J. Pascoe, K. B. Ling, S. L.
Cockroft, J. Am. Chem. Soc. 2017, 139, 15160-15157; d) N. Dominelli-
Whiteley, J. J. Brown, K. B. Muchowska, I. K. Mati, C. Adam, T. A.
Hubbard, A. Elmi, A. J. Brown, I. A. W. Bell, S. L. Cockroft, Angew.
Chem. 2017, 129, 7766-7770; Angew. Chem., Int. Ed. 2017, 56, 7658-
7662. e) K. B. Muchowska, D. J. Pascoe, S. Borsley, I. V. Smolyar, I. K.
Mati, C. Adam, G. S. Nichol, K. B. Ling, S. L. Cockroft, Angew. Chem.
2020, DOI: 10.1002/ange.202005739; Angew. Int. Ed. 2020, DOI:
10.1002/anie.202005739.
[21] a) D. H. McDaniel, H. C. Brown, J. Org. Chem. 1958, 23, 420-427; b) A.
T. Shulgin, A. W. Baker, Nature 1958, 182, 1299-1299; c) T. J. A.
Corrie, L. T. Ball, C. A. Russell, G. C. Lloyd-Jones, J. Am. Chem. Soc.
2017, 139, 245-254.
[22] Y.-S. Jang, Ł. Woźniak, J. Pedroni, N. Cramer, Angew. Chem. 2018,
130, 13083-13087; Angew. Chem., Int. Ed. 2018, 57, 12901-12905.
[23] p values were used for substituents positioned ortho to the biphenyl
bond.
[24] D. J. Byron, G. W. Gray, R. C. Wilson, J. Chem. Soc. C 1966, 837-840.
[25] C. S. Wannere, P. von Ragué Schleyer, Org. Lett. 2003, 5, 605-608.
[26] a) C. A. Hunter, Angew. Chem. 2004, 116, 5424-5439; Angew. Chem.
Int. Ed. 2004, 43, 5310-5324; b) S. L. Cockroft, C. A. Hunter, Chem.
Commun. 2009, 3961-3963.
[27] C. A. Hunter, C. M. R. Low, J. G. Vinter, C. Zonta, J. Am. Chem. Soc
2003, 125, 9936-9937.
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202006943
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
Measurements of through-space substituent effects question classic descriptions of substituent effects.
9
This article is protected by copyright. All rights reserved.