Systematic Investigation of Lipophilicity Modulation by Aliphatic Fluorination Motifs

Article abstract of DOI:10.1021/acs.jmedchem.9b01172

Optimization of compound lipophilicity is a key aspect of drug discovery. The aim of this work was to compare the lipophilicity modulations induced by 16 distinct known and novel fluoroalkyl motifs on three parent models. Fifty fluorinated compounds, with 28 novel experimental aliphatic loga ?P values, are involved in discussing various lipophilicity trends. As well as confirming known trends, a number of novel lipophilicity-reducing motifs are introduced. Tactics to reduce lipophilicity are discussed, such as "motif extensions" and "motif rearrangements", including with concomitant extension of the carbon chain, as well as one-and two-fluorine a deletions' within perfluoroalkyl groups. Quantum chemical loga ?P calculations (SMD-MN15) based on solvent-dependent three-dimensional (3D) conformational analysis gave excellent correlations with experimental values, superior to Cloga ?P predictions based on 2D structural motifs. The availability of a systematic collection of data based on a small number of parent molecules illustrates the relative lipophilicity modulations of aliphatic fluorination motifs.

Full text of DOI:10.1021/acs.jmedchem.9b01172

Subscriber access provided by AUT Library
Article
A Systematic Investigation of Lipophilicity
Modulation by Aliphatic Fluorination Motifs
Benjamin Jeffries, Zhong Wang, Hannah R. Felstead, Jean-Yves Le Questel,
James S. Scott, Elisabetta Chiarparin, Jerome Graton, and Bruno Linclau
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01172 • Publication Date (Web): 02 Jan 2020
Downloaded from pubs.acs.org on January 2, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A Systematic Investigation of Lipophilicity
Modulation by Aliphatic Fluorination Motifs
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
a
a
b
Benjamin Jeffries, Zhong Wang, Hannah R. Felstead, Jean-Yves Le Questel, James S.
c
c
b
a
Scott, Elisabetta Chiarparin, Jérôme Graton, * Bruno Linclau *
a
University of Southampton, School of Chemistry, Highfield, Southampton SO17 1BJ, UK
CEISAM UMR CNRS 6230, Faculté des Sciences et des Techniques, Université de Nantes
b
2
, rue de la Houssinière – BP 92208,44322 Nantes Cedex 3, France
c
Medicinal Chemistry, Oncology R&D, AstraZeneca, Cambridge CB40WG, UK
Corresponding Author: *Tel.: +44 (0)23 8059 3816, Fax: +44 (0)23 8059 3781, E-mail:
bruno.linclau@soton.ac.uk
Abstract.
Optimization of compound lipophilicity is a key aspect of drug discovery. The aim of this work
was to compare the lipophilicity modulations induced by 16 distinct known and novel
fluoroalkyl motifs on three parent models. Fifty fluorinated compounds, with 28 novel
experimental aliphatic logP values, are involved in discussing various lipophilicity trends. As
well as confirming known trends, a number of novel lipophilicity reducing motifs are
introduced. Tactics to reduce lipophilicity are discussed, such as “motif extensions” and “motif
rearrangements”, including with concomitant extension of the carbon chain, as well as one-
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and two-fluorine ‘deletions’ within perfluoroalkyl groups. Quantum chemical logP
calculations (SMD-MN15) based on solvent-dependent 3D conformational analysis gave
excellent correlations with experimental values, superior to ClogP predictions based on 2D
structural motifs. The availability of a systematic collection of data based on a small number
of parent molecules illustrates the relative lipophilicity modulations of aliphatic fluorination
motifs.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Introduction.
Fluorine introduction is a much-used operation in the drug development optimization process.1-
4
While any change in molecular structure will have an impact on a number of properties, this
is particularly pronounced for fluorination. This is due to the high electronegativity of fluorine,
resulting in a highly polar C–F bond with concomitant strong inductive effect and low
5
polarizability of the fluorine lone pairs, without leaving much of a steric footprint. The
efficiency of fluorine introduction for property optimization is exemplified by statistics on the
6
-8
proportion of commercial drugs, or drugs currently in clinical trials, that are fluorinated. The
vast advances that are continuing to be made in synthetic organofluorine chemistry9-17 have
accelerated its use and, with the emergence of late stage fluorination methodologies,18,19 this is
set to continue. The increased understanding of the stability of fluorinated motifs, especially
when heteroatoms are involved,20-22 will further enhance its successful application.
Lipophilicity is an important property that is affected by fluorination. While fluorine
introduction on aromatic rings generally leads to a lipophilicity increase, the influence of
2
,4
aliphatic fluorination on lipophilicity is more complex. This is best illustrated by CH CF
3
3
exchange, which can result in a lipophilicity increase or decrease, which can be in both cases
quite significant (see below). Given the need to keep lipophilicity within optimal boundaries
during optimization,23-29 the possibility to use aliphatic fluorination for lipophilicity reduction
2
ACS Paragon Plus Environment

Page 3 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
purposes is an attractive prospect. New fluorine-containing motifs continue to be introduced
that are suitable for this purpose, with recent examples being the (Ar)-SCF CH and (Ar)-
2
3
3
0
31
OCF CH groups by O’Hagan, and the –OCH (CF ) CH (n = 2,3) groups from our lab. In
2
3
2
2 n
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
addition, lipophilicities of some known motifs such as a 1,2-difluoroethyl group, which has
been demonstrated to be very efficient in lipophilicity reduction (see below),32,33 and the CF H
2
group, which has additional interest as ‘lipophilic hydrogen bond donor’,34,35 have been
investigated in detail. In this context, the availability of a toolbox of “fluorinated motifs” from
which the medicinal chemist can pick and choose for optimal effect would be of great practical
use, and the understanding of the lipophilicity impact of diverse fluorinated motifs, with or
without modifying the chain lengths involved, will be of great benefit in medicinal chemistry.
The process of understanding the influence of aliphatic fluorination on lipophilicity was
pioneered by the Müller/Carreira group (Figure 1, series A–C), by investigating a series of
3
6
compounds with increasing fluorine content, such as A1–A5. Using a combination of dipole
moment and fluorine hydrophobic surface considerations, they explained the decrease in
lipophilicity upon monofluorination by the C–F dipole introduction. Given that a
trifluoromethyl group has approximately the same dipole moment as a C–F group, but a larger
hydrophobic surface, a higher logP would be expected compared to monofluorination. The
dipole moment of a CF -group is slightly larger than that of a single C–F, but the expected
2
logP decrease is overridden by the logP increase caused by the larger hydrophobic surface
area. The lower lipophilicities of a vicinal over a geminal difluoro motif could nicely be
explained by the higher vector sum of the two C–F dipoles for a ca. 60° (gauche dihedral) angle
compared to a ca. 110° (bond) angle.32,37 Subsequently, through the synthesis of Gilenya®
derivatives with alkyl side chains of various lengths, Gilmour has further demonstrated the
lower logP of 1,2-difluoroethyl groups compared to their ethyl and trifluoroethyl
3
3
counterparts. Müller also pointed out that the lipophilicity increase resulting from aliphatic
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
chain extension is compensated when a CF group is simultaneously replaced by a vicinal
2
3
2
difluoro group (e.g. compare A3, A4 with B4–B6). Such observations are of great interest as
these represent liponeutral bulk increases.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Although their lipophilicity range is three orders of magnitude apart, the fluorinated propoxy
side chains in series C and the corresponding propanols (series D, Figure 1) display the same
38
lipophilicity trend, except that now the trifluorinated D4 shows a slightly increased logP
3
8
compared to propanol D1. For the ethanol series (E, Figure 1) and the proline ester series (F,
3
9
Figure 1), this logP increase for the trifluorinated analogues is even more pronounced, with
the difluorinated derivatives in series E and F being as lipophilic as their nonfluorinated
parents. In these cases, the decrease in polarizability of the oxygen lone pairs caused by the
inductive effect of the trifluoromethyl group has a logP increasing effect (similar to how
aromatic fluorination increases lipophilicity). In addition, with a C2-chain, there is always a
conformation in which a C–F dipole is antiperiplanar to the C–O dipole. Furthermore, the
lipophilicity range of series D–F (from -0.75 to +0.55 logP units) is much lower compared to
that of series C (+2.3 to +2.9 logD units). Hence, the relative contribution of the dipole is much
reduced, in contrast to the volume contribution, a consideration already recognized by Müller
through the lipophilicity comparison across the logP scale of a set of trifluoromethylated
3
6
compounds with their methylated analogues. Interestingly, Rafique et al. reported the logD
of a number of CF CH OAr/FCH CH OAr matched molecular pairs (not shown), with Ar
3
2
2
2
representing 4-pyridyl rings having a range of (non-conjugated) substituents, in which the
trifluoromethylated derivative is not always the most lipophilic one although all compounds
are within a narrow lipophilicity range.40
4
ACS Paragon Plus Environment

Page 5 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
logP
logP
logP
0.6 _
HO
(
Linclau et al)
Rf
Rf
3
3
3
.7 _
O
B1 (+3.7) Ind
Rf
Series D: Rf = C3
Series E: Rf = C2
O
CF3
.6 _
.5 _
N
H
3.2 _
N
H
F4 (+0.55) Pro
0.5 _
0.4 _
Series A: Rf = C3
Series B: Rf = C4
Series C: Rf = C3
(K. Müller et al)
HO
CF3
3.1 _
3.0 __
2.9 _
O
D4 (+0.41)
HO
CF3
(
K. Müller et al)
E4 (+0.36)
3.4 _
N
O Rf
0
0
.3 _
.2 _
D1 (+0.30)
O
3
3
.3 _
.2 _
Ind
HO
A1 (+3.3)
C1 (+2.9) Ind
Series F: Rf = C2
Kubyshkin, Budisa)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
(
2.8 _
2.7 _
C5 (+2.8) Ind
CF3
O
0.1 _
F2
CH
F2
C
O
O
3
3
.1 _
Ind
Ind
C4 (+2.7) Ind
D3 (+0.04)
A5 (+3.1)
A3 (+2.9)
CF3
B2 (+3.1) Ind
B3 (+3.0) Ind
0.0 __
F3 (+0.02) Pro
F1 (+0.00) Pro
O
C
HO
F2
CH
F2
F2
CH
.0 __
2.6 _
2.5 _
2.4 _
2.3 _
2.2 _
2.1 _
C3 (+2.6) Ind
-0.1 _
-0.2 _
-0.3 _
-0.4 _
C
O
F2
2.9 _
CH
F2
C2 (+2.5) Ind
Ind
O
F
F2
CH
HO
F
F2
C
F
HO
B4 (+2.8)
B5 (+2.8)
B6 (+2.8) Ind
2
2
.8 _
.7 _
A4 (+2.8) Ind
A2 (+2.8) Ind
F
F
D2 (-0.26)
F
E3 (-0.29)
E1 (-0.30)
B7 (+2.7)
F
C6 (+2.3) Ind
HO
B8 (+2.7)
F
O
F
O
F2 (-0.43) Pro
F
Ind
F
2
2
.6 _
.5 _
F
F
F
F
HO
A6 (+2.5) Ind
F
-0.75 _
E2 (-0.75)
Figure 1. A selection of relevant precedence of aliphatic fluorination and lipophilicity
featuring different fluorination patterns for given ‘parent’ (nonfluorinated) compounds A–F.
logP/D
F2
C
(
Linclau et al)
OH
Rf
Rf
Series G: Rf = C4
F3C
C
F2
OR
2
.6 _
H4
Series H: Rf = C4
O
H2
N
+2.5
2
.4 _
O
N
2.2 _
2
.0 __
G18
+1.98
F2
C
OR
1.8
1.8 _
+
+
F3C
C
F2
OH
H1
H3
1.6 _
1.6
F2
C
F2
C
1
1
.4 _ F3C
OH
1.30
C
F2
OR
G16
+
+
.2 _
.0 __
0.8 _
H2
+1.1
1
F2
C
G1
G13
+0.87
0.88
OH
OR
F2
C
C
F2
OH
0
.6 _
.4 _
0
G9
+
0.34
OH
F2
C
0.2 _
Figure 2. LogP/D values of derivatives obtained by internal polyfluorination of the parent
compounds G1 and H1.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Based on the observation that there is a large logP decrease when terminal pentafluorination is
changed to a CF motif at the penultimate chain position (Figure 2, compare G16 with G9),
2
our group established that this remains the case even with longer perfluoroalkyl moieties (e.g.
G18G13).31 Interestingly, the tetrafluorinated G13 has a similar lipophilicity as its
nonfluorinated parent G1. Very similar observations were made when such butoxy groups were
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
1
introduced in a druglike molecule, such as evenamide H1, a schizophrenia drug currently in
phase II clinical trials.41
The availability of measured lipophilicity data of matched molecular pairs (or matched
molecular series) is of great interest for qualitative evaluation about the relative effect of a
given motif on lipophilicity, and more importantly, as a collective body of data to extract trends
and gain understanding of the effects at work. The reporting of experimental logP data is very
valuable27,29 as inputs to validate/optimize computational logP prediction algorithms, or to
obtain correlations of their predictions with the experiment.
Here, we report experimental octanol-water shake flask logP values of 50 saturated fluorinated
aliphatic compounds involving, in total, 16 different aliphatic fluorination motifs introduced
on three “parent” alcohol models. Our choice for alkanols as model compounds is based on the
aforementioned interplay of the polar C–F bond with an electronegative functional group,
which affects its resulting lipophilicity modulation. Given that the magnitude of fluorine’s
lipophilicity modulating effect is also dependent on the lipophilicity of the nonfluorinated
parent compound, it is important that different motifs can be compared when substituted on the
same framework, hence the limited number of nonfluorinated ‘parents’. Two nonfluorinated
‘parents’ are linear alkanols (Figure 3): 1-butanol (BuOH, logP +0.88) and 1-pentanol (1-
PentOH, logP +1.51). Positional isomers which have the fluorination motif at different carbon
atoms are investigated systematically for the CHF and CF -motifs. For this purpose, 2-pentanol
2
(2-PentOH, logP +1.19) is also included as a parent substrate. Around half of the logP values
6
ACS Paragon Plus Environment

Page 7 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
are new, and the precedent is included to have as large a data set as possible to facilitate
discussion and allow comparison between different motifs, and positioning of motifs vs a polar
alcohol group.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Results.
Lipophilicity results.
1
9
The direct determination of logP values was conveniently achieved by a F NMR based
method developed by our group, which is suitable for measuring the octanol/water partition
31
coefficients P of (fluorinated) non-UV active substrates. The use of an internal standard as a
mixture with the compound of interest obviates the need for accurate weight/partition
volume/sample volume measurements. Potential quantitative integration issues due to the
solvent-dependence of the relaxation time are minimized by selecting a sufficiently long
relaxation delay for the NMR experiments (D1-setting).
The results of the lipophilicity measurements are shown in Figure 3, together with previously
measured values of analogues. The difference in lipophilicity with the nonfluorinated parent is
indicated.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Summary table with new logP data, together with previously reported values for
completion.31,38,42 Color coding: blue, monofluorinated; green, gem-difluorinated; red, a
(
single) CF group; purple, multifluorinated (fluorination at more than one carbon atom). The
3
difference in logP value with the corresponding nonfluorinated parent compounds G1, I1, J1
(depicted in black) is given (See Figure S1 for a list of all logP values).
Discussion
8
ACS Paragon Plus Environment

Page 9 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The discussion of the data is organized by motif. First, mono- and geminal difluorination
will be discussed, then vicinal fluorination motifs, followed by skipped motifs. A number of
compound series will be discussed, and finally correlations will be shown with a number of
logP prediction methods.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Monofluorination and geminal difluorination.
The data for all monofluorinated and geminal difluorinated analogues is shown in Figure 4.
As expected, monofluorination of acyclic alkanols results in a systematic decrease in
lipophilicity when compared to their respective parent compound. This decrease is clearly
dependent on the relative position of the fluorination and the alcohol functional group (cf. the
series G3,4,5; J2,3,4,7,8,9, and I3,4,9,11). The β-monofluorinated analogues G5, J7–9 and
I11 all have the highest logP values within their respective families. This is explained by the
proximity of the fluorination site to the hydroxyl group, resulting in a higher reduction in
polarizability of the oxygen lone pairs, as well as a possible conformation where the C–O/C–
F dipoles counteract. Monofluorination at the methyl group (J9) causes a slightly lower
reduction in lipophilicity compared to internal monofluorination (J7/J8). Unfortunately, due
1
9
to partial F NMR signal overlap in the aqueous solvent, the lipophilicity of the individual
diastereomers J7 and J8 could not be determined (the stereoisomers were not separable),
although qualitatively the values appear to be very similar (the average value is given, see SI
section 4.3.2). An increasing number of methylene groups between the OH and F groups leads
to a further reduction in lipophilicity. For I3 and I4, the decrease in lipophilicity is quite
substantial (ΔlogPI1-I3,4 –0.99). Compounds J3 and J4 were obtained in diastereomerically pure
form, and the lower lipophilicity of the syn-diastereomer J3 can be explained by the
comparatively larger stabilization of its most polar conformation (featuring parallel C–O and
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
C–F bonds) in water compared to that of J4 (and the opposite for the less polar conformers in
octanol).38
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Systematic mono- and difluorinated lipophilicity series of the parent compounds
1-butanol (G1), 2-pentanol (J1) and 1-pentanol (I1). Color coding: blue, monofluorinated;
green, gem-difluorinated.
All the geminal difluorinated alkanols are more lipophilic than their corresponding
37
monofluorinated congeners, which conforms to Müller’s ‘rule’ that the influence of the larger
CF -dipole moment (compared to C–F) is counterbalanced by an increased hydrophobic
2
surface. For the compounds in Figure 4, the increase from the monofluorinated congeners is
about 0.15-0.20 logP units, except for the vicinal difluorohydrins where the difference is up to
0
.40 logP units. This can be explained by the larger polarizability reduction of the alcohol lone
pairs by a CF -group, as well as by the increased probability that the C–O dipole is
2
antiperiplanar with a C–F dipole (compared to monofluorination). Nevertheless, all gem-
1
0
ACS Paragon Plus Environment

Page 11 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
difluorinated compounds investigated in this study still have a lower lipophilicity in
comparison to their respective nonfluorinated parent compound. As with the monofluorinated
derivatives, their lipophilicity progressively decreases with fluorination further from the
alcohol group. Indeed, the β-difluorination analogues G12, J10, J11 and I17 all have the
highest logP values within their respective families. In the 2-PentOH family, just like for the
monofluorination, the internal β-difluorination J10 causes a larger logP decrease compared to
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
its terminal counterpart J11. For matched compound pairs featuring internal CF and terminal
2
3
2
CF H groups, Müller already noted that there was little difference in their logP values. In our
2
data set involving longer aliphatic chains, we show that this remains true only when CF H
2
compounds are compared with compounds having CF groups at the penultimate carbon atom
2
(cf. I7 and I8 for the PentOH family, J5 and J6 for the 2-PentOH family, and G7 and G9 for
the BuOH family). For CF groups further away from the final carbon atom the difference is
2
much more significant, because of the influence of the polar alcohol group. In all cases,
compounds having a RCH CF H group are slightly less lipophilic than their congeners with a
2
2
RCF CH group (compare G7 and G9; J5 and J6; and I7 and I8).
2
3
In longer chains such in 1-pentanol, the increase in lipophilicity in comparing C–F with CF2
is roughly compensated when moving one bond further away from the alcohol group. For
example, 3-fluoropentan-1-ol I9 has a similar logP as 4,4-difluoropentan-1-ol I8.
Vicinal fluorination series.
Two series featuring vicinal fluorination motifs at the terminal position of the alkyl chain are
shown in Figure 5. In general, the lipophilicity increases with increasing number of fluorine
atoms present in the motif, and it is notable that only the compounds with the fully fluorinated
pentafluoroethyl motif (G16, I19) have a larger lipophilicity than their nonfluorinated parents.
The introduction of vicinal difluorination causes significant reductions in lipophilicity (–1.00
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
32
logP unit for G2, –1.40 logP units for I2). This is in keeping with Müller’s analysis, and with
®
Gilmours’s Gilenya examples (there was a 1.7 logP unit decrease between the respective
®
33
CH –CH and CHF–CH F Gilenya analogues). The differences in logP tend to slightly
2
3
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
increase between compounds having 2 vs 3, 3 vs 4 and 4 vs 5 fluorines present. For motifs with
the same number of fluorines, their internal arrangement has an impact on the logP. For the
vicinal trifluorination motifs, which have been considered by Müller in the context of polarity
43
prediction, this impact is small (logP 0.02-0.06, and reverses between BuOH (G6/G8) and
PentOH (I5/I6)), but there is a larger difference between the two vicinal tetrafluorinated motifs
(logP 0.25, I13/I15).
In all cases, for series where the fluorine count at a particular carbon is increasing whilst
keeping that at the other carbon of the motif constant (e.g. consider I2I5I15 or
I6I13I19), the logP increases. This is further elaborated in the SI (Figures S2-S3).
1
2
ACS Paragon Plus Environment

Page 13 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Systematic exploration of all C H F motifs.
x (5-x)
2
The vicinal fluorination motifs in G2/I2, G6/I6, G8/I5, and G11/I13 feature different
conformations upon rotation of the bond between the fluorinated carbons and, given the
polarity of the various C–F bonds present, it was of interest to explore the impact of their
conformations on the lipophilicity. Hence, conformational analysis in water and octanol
medium was carried out for the corresponding butanol derivatives (Table 1).
a
Table 1. Calculated dipole moments and conformational distribution of selected vicinal
fluorinated butanols in the octanol and the water phase. The repartition of the g(+), g(–) and t
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
conformers (defined by the H(F)-C –C -C dihedral angle) is given. For full details, see Tables
4
3
2
S8-S11.
b
NMR
analysis
% antiperiplanar)
Compound
g(+)
g(–)
t
c
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(logP)
td
sf
(D)
sf
(D)
sf
(D)

pe
pc
pc
(
d
t (D)
(D)
F2
C
F
F
H
h
H
F
F
F
F
F
F
F
F
F
C
F
OH
2
F
H
H
G11 (+0.66)
R
R
R
Oct: 3.13
26%
35%
4.25
4.45
26%g
35%g
4.25
4.45
47%
30%
1.88
2.08
-0.60
Wat: 3.73
F
F
F
H
3JH3-H4 (oct) 3.3 Hz (41%)
3JH3-H4 (wat) 2.8 Hz (31%)
H
F
F
H
H
F
F
H
F
F
F
H
C
F
OH
H
2
R
R
R
G8 (+0.29)
3J F3-H4 (oct) 7.4 Hz (33%)
3J F3-H4 (wat) 9.0 Hz (50%)
Oct: 3.48
41%
31%
3.56
3.58
33%
50%
4.20
4.58
25%
19%
2.42
2.58
-0.42
Wat: 3.90
H
H
F
F2
C
3JH4-F3 (wat) 13.2 Hz (39%)
3JH4-F3 (oct): J-values could not be
F
F
F
F
F
F
F
F
F
OH
G6 (+0.27)
H
H
H
H
R
R
R
determined
Oct: 3.69
Wat: 4.15
24%
22%
2.76
2.93
24%g
22%g
2.76
2.93
52%
55%
4.56
5.14
-0.46
H
H
F
F
h
H
H
F
F
H
F
F
H
H
F
F
OH
H
H
G2 (-0.12)
R
R
R
Oct: 3.74
Wat: 4.04
38%
36%
3.12
3.95
0.0%
0.0%
2.24
2.34
62%
64%
4.10
4.09
-0.30
aCalculated at the SMD/MN15/aug-cc-pVTZ//MN15/cc-pVTZ level of theory in water and octanol medium. bt-
c
d
Conformer shown. Experimental logP. Weighted by the relative populations of each conformer:  =oct–wat (t
t
e
f
=
total). Sum of all conformers with a given H(F)-C
4
-C
3
-C
2
dihedral angle. Weighted dipole moment of the
g
conformers with the H(F)-C
4
-C
3
-C
2
dihedral angle as shown: 
s
(s = subset). Structure with equivalence of the
h
g(+) and g(-) conformations. Not investigated.
In all cases, the “total” molecular dipole moment  (weighed across all conformations) in the
t
octanol phase is lower than that in the water phase (cf.  ), which reflects the expected higher
t
populations of apolar conformers in the octanol phase due to its lower dielectric constant. This
difference is clearly significant for the tetrafluorinated G11.
1
4
ACS Paragon Plus Environment

Page 15 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The data were then further refined to provide insight into the contribution of the fluorinated
motif: the fluorohydrin conformations are grouped according to the three possible C3-C4
rotational minima. Their respective population in both the octanol and water phases, and the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
corresponding weighted dipole moment  of these subsets of conformers are given. The
s
magnitude of  is determined by the two main contributors, the fluorinated motif and the
s
alcohol group, and the contribution of the former is easily seen. For example, for G11, all C–
F dipole moments in the t-conformers are opposed so the  (t) value is much lower than that
s
of the (equivalent) g conformations, where only two C–F bonds are antiperiplanar. For G8, the
g(+)- and t-conformations each have two C–F dipoles opposed, but their corresponding s
values (for a given phase) are different, which is due to a different relative average orientation
of the C–O–H dipoles. Still, both values are lower than that of the  (g(–)) value, as there the
s
three C–F dipoles point into the same direction (vector sum ~2.16(C–F)). For a given motif
conformation, the  values are typically lower for the octanol phase, which is attributed to the
s
aforementioned conformational differences across the C1-C2 and C2-C3 bonds (higher
population of apolar conformations in the octanol phase).
The conformational population data show that there is considerable difference in the influence
of the medium on the conformational profile of the fluorinated motifs. For the compounds
bearing two fluorines at the C-terminal position, G11 and G8, the motif conformer populations
in water are very different from those in octanol. For G11, the population of the more apolar
trans conformation significantly increases going from water to octanol (from 30% to 47%, a
ca. 60% increase). Similarly for G8, the populations of the two less polar g(+) and t
conformations both increase by around 30% each. Conversely, with only one fluorine at the C-
terminal position (G6 and G2), there is minimal influence of the medium on the motif
conformer population. Hence, for these compounds, despite the overall conformational profile
in the two phases being clearly different (as shown by the  values), the different solvent
t
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
polarities do not influence the conformational behavior of these fluorination motifs: regardless
the medium, for G2 there is 36-38% of g(+), 62-64% of t, and 0% g(–) population (similar
consideration for G6). In other words: the dielectric constant of the medium has no impact on
the conformational distribution of the motif, despite the different dipole moments of these
conformations (e.g. compare G6(t) and G6(g)). This could be explained by additional factors
that stabilize conformations, over and above dipole stabilization by the dielectric constant of
the medium, such as steric effects, and by stereoelectronic effects such as the fluorine gauche
effect, which is a stabilizing C–H*C–F hyperconjugation. The latter was investigated
further.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The energies of hyperconjugations operating in these motifs can be estimated through NBO
analysis by calculating the charge transfer from the corresponding bonding to the antibonding
orbitals, E(2)*, reported in Table S15. As expected, the stabilization by C–H*C–F
hyperconjugation is larger than a C–C*C–F hyperconjugation, which in turn is larger than a
C–F*C–F hyperconjugation. Furthermore, the calculations show that the C–H*C–F
hyperconjugation energy roughly decreases with the extent of additional fluorination in the
order G2G6G8G11. Hence, the g conformations of G11 are stabilized by one weak C–
-1
-1
H*C–F (ca. 14 kJ mol ) and C–C*C–F hyperconjugation (ca. 6 kJ mol ). For G8, the g(–)
-1
conformation exhibits two C–H*C–F hyperconjugations (ca. 17 kJ mol ), and the t-
-1
conformation only one (ca. 17 kJ mol ). The g(+) and again the g(–)-conformations have a
-1
weaker C–C mediated hyperconjugation (ca. 9 kJ mol ). For G6, the t-conformer has one C-
-1
-1
H (ca. 21 kJ mol ) and one C–C (ca. 8 kJ mol ) mediated hyperconjugation, while its g(+)
-1
conformer only has one C-H mediated one (ca. 20 kJ mol ). Finally, compound G2 only
occupies conformations that have either two C-H mediated hyperconjugations (ca. 21 kJ mol-1
-1
in g(+)), or one C-H and one C-C mediated hyperconjugations (ca. 21 and 12 kJ mol in t).
The latter is the most populated, which may be due to the lower amount of steric congestion.
1
6
ACS Paragon Plus Environment

Page 17 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Hence, we qualitatively suggest that the larger F-gauche stabilization in G2 would make it the
dominant force in both phases, while for G11 the smaller fluorine gauche stabilization would
be of the same order of magnitude as the interactions that destabilize polar conformations in
apolar phases, resulting in a ‘chameleon’ character in which the molecule adopts polar or apolar
conformations depending on the dielectric constant of the medium. Nevertheless, crystal
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4-
structures containing -CF CF H groups show that this motif was present as a g conformation.
2
2
4
7
The calculations of the motif conformations in G8 and G11 correspond reasonably well to
the relative energies of the respective conformations for 1,1,2-trifluoro and 1,1,2,2-
tetrafluoroethane as calculated with MP2/6-31+G*.48
A correlation of these calculated populations with the solution-phase populations through a
(qualitative) J-value NMR analysis (see SI section 8 for full details) was possible for G8. Using
3
(
octanol-saturated) D O and (water-saturated) octanol as solvents, the JH3-H4 value measured in
2
the octanol phase (3.3 Hz) is larger compared to the value in the water phase (2.8 Hz), and vice
versa for the 3JF3-H4 values (7.4 vs 9.0 Hz). This is consistent with the population differences of
conformers containing antiperiplanar H3-H4 and F3-H4 dihedral angles (see Table 1, last
column). For G6, the coupling constants in water-saturated octanol could not be determined
due to signal overlap.
‘Skipped’ fluorination.
A small number of ‘skipped’ fluorination motifs, with a methylene group separating the
fluorinated carbons, are shown in Figure 6. In all cases the motif contains a trifluoromethyl
group.
While in both cases skipped tetrafluorination leads to a lower logP than skipped
pentafluorination, a key difference between the butanol and the pentanol series is their relative
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
position with the nonfluorinated parent: the pentafluorinated G15 has a higher logP than
butanol G1, while I14 has a much lower lipophilicity than I1. This can be explained by the
proximity of the motif to the alcohol group (cf. Figure 4: ,-difluorination leads to a much
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
higher logP than ,-difluorination, see Figure S4 for further analysis).
Conformational analysis in water and octanol medium of some skipped fluorination motifs,
calculated on the 1-pentanol chain are given in Table 2. Following the analysis for the vicinal
motifs, the conformations are grouped according to the three possible C3-C4 rotational minima,
hence the data refer to the conformational profile of the skipped fluorination motif.
logP
1.6 _
1.5 _
1.4 _
1.3 _
I1 (+1.51)
OH
1.2 _
F2
C
I14 (+1.12) F3C
I10 (+0.84) F3C
1
.1 _
OH
OH
F3C
C
F2
OH
OH
G15 (+1.04)
G1 (+0.88)
1.0 __
0.9 _
F
0
.8 _
0.7 _
.6 _
0
F3C
OH
G10 (+0.60)
F
Figure 6. Skipped fluorination motifs.
a
Table 2. Calculated dipole moments and conformational distribution of the skipped
fluorinated pentanols I10 and I14 in the octanol and the water phase, next to the
monofluorinated I9. The repartition of the g(+), g(-) and t conformers (defined by the F C-C -
3
4
C -C or H C-C -C -C dihedral angle) is given. For full details see Tables S12-S14.
3
2
3
4
3
2
Compoundb
NMR
g(+)
g(–)
t
1
8
ACS Paragon Plus Environment

Page 19 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
c
d
sf
(D)
sf
sf
(
logP)

D)
t
analysis
(% antiperiplanar)
pe
pc
pc
d
(
(D)
(D)
t (D)
H
H
CF3
3JH4-F3 (wat) 15.6 (37%)
3JH4-F3 (oct) 15.5 (39%)
F2
C
F
F
F
F
F
F
F3C
OH
I14 (+1.12)
H
CF3
F3C
H
H
H
R
R
R
Oct: 3.81
26%
21%
2.87
3.06
26%g
21%g
2.87
3.06
47%
57%
4.86
4.92
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-
0.30
H O: 4.11
2
H
H
CF3
F
3JH3-H4R (wat) 9.0 Hz (97%)
3JH3-H4S (wat) 2.7 Hz (2%)
3JF3-H4R (wat) 16.5 Hz (1%)
3JF3-H4S (wat) 33.6 Hz (97%)
H
H
F
H
F
H
F
F3C
OH
CF3
F3C
H
HS
HR
I10 (+0.84)
R
R
R
Oct: 4.45
2%
1%
2.45
2.16
3%
2%
3.44
3.58
94%
97%
4.54
4.93
-0.42
H O: 4.87
2
H
H
CH3
F
h
H
H
F
H
F
H
H
F
OH
CH3
H C
H
H
3
I9 (+0.75)
R
R
R
Oct: 2.58
13%
27%
2.88
3.32
21%
16%
2.96
3.29
56%
56%
2.39
3.02
-
0.57
H O: 3.15
2
aCalculated at the SMD/MN15/aug-cc-pVTZ//MN15/cc-pVTZ level of theory in water and octanol medium. bt-
c
d
Conformer shown. Experimental logP. Weighted by the relative populations of each conformer:  =oct–wat (t
t
e
f
=
total). Sum of all conformers with a given H(F)-C
4
-C
3
-C
2
dihedral angle. Weighted dipole moment of the
g
conformers with the H
3
(F
3
)C-C
4
-C
3
-C
2
dihedral angle as shown: 
s
(s = subset). Structure with equivalence of
h
the g(+) and g(-) conformations. Not investigated.
In line with the results described for the vicinal fluorination motifs, a lower averaged molecular
dipole moment is found in the octanol phase than in the water phase for these three
fluoropentanol derivatives. The grouping of the conformations along the three possible motif
conformational minima reveals interesting differences between I10 and I14. There is a broad
population distribution for the skipped pentafluorinated motif, while the opposite is observed
for the skipped tetrafluorination motif, for which the t-conformer is by far the most stable
conformer. The destabilization of the t-conformer of I14 is most likely due to the linear 1,3-
4
9
repulsions between the CF and CF groups, although t-I10 also features such an interaction.
2
3
As a separate point of interest, the conformational profile of I10 was compared with that of I9,
in which the trifluoromethyl group is replaced with a methyl group. Again, there is much more
conformational disorder to a level close to that of I14. The estimated hyperconjugation
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
-
1
contributions (Table S16) for I9 (C–H4*C–F3 ca 25 kJ mol and C4–C5*C–F ca 14 kJ
-1
-1
mol ) are much higher than those calculated for I14 (C–H4*C–F3 ca 20 kJ mol and 
C–
-1
C*C–F ca 10 kJ mol ). Hence, the electronic stabilization of the g-conformers is higher for
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
I9, and is now able to compete with the steric destabilization between the methyl and R-groups.
For I10, the populations of the t-conformers is only slightly reduced in the octanol phase,
despite their weighted dipole moment being much higher than that of the g-conformers. For I9,
for which all conformer subsets have similar averaged dipole moments in water, only the g-
conformers show variation in population between the solvents.
1
19
H and F NMR analysis (see SI section 8 for a detailed discussion) was possible for I10 and
I14, and is fully consistent with the calculated conformational profile. For I10, the dominant
anti-conformation of the skipped tetrafluorinated motif in I10 is clearly demonstrated. In water,
the very different vicinal couplings of H3 to the two diastereotopic H4 protons, which are
averaged over all conformations, indicate a biased conformational profile, with the 9.0 Hz
coupling relating to an antiperiplanar arrangement between H3 and what must be the H4(R)
proton (cf. the corresponding Newman projection in Table 2), and a gauche arrangement
between H3 and the H4(S) proton. Hence, this H4(S) proton must then predominantly be
antiperiplanar with the fluorine at C3, which is indeed borne out with the large vicinal H-F
coupling. Unfortunately, due to signal overlap, the corresponding J-values in water-saturated
octanol could not be obtained. For I14, the 3JH4-F3 values in the octanol and water phases are
essentially the same (within the error limit), which is what was expected given their very
similar (averaged) antiperiplanar disposition between the H4 and F3 nuclei.
Fluorinated motif extension (no change in carbon framework)
2
0
ACS Paragon Plus Environment

Page 21 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
As indicated in the introduction, there are many reasons for fluorine introduction as part of the
drug development process. In the context of lipophilicity control, it can be of interest to have
knowledge whether the logP of a candidate that already contains fluorine can be further
reduced by introducing additional fluorines. The possibility of fluorinated motif extension for
that purpose, where fluorine is introduced at other carbons than those already fluorinated, is
explored in Figures 7 and 8 (further introduction of fluorine at an already fluorinated carbon is
expected to lead to a lipophilicity increase).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The extension of the monofluorination in I3 and I4 by a vicinal C–F group (Figure 7), leading
to I2, leads to a significant logP reduction. The same operation starting from the difluorinated
structures I7 and I8, leading to I5/I6, still leads to a logP decrease, albeit much smaller. The
extension of the monofluorination in I3 and I4 by two geminal C–F bonds, leading to I5/I6, is
either liponeutral or slightly logP enhancing. In contrast, the same operation starting from the
difluorinated structures I7 and I8, leading to I13, now leads to a significant logP increase.
Hence, extension with a single fluorine leads to an overall increase in polarity that outweighs
the introduction of hydrophobic surface, while the opposite is the case when extending with
two geminal C–F bonds.
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. F-motif extension of mono- and geminal difluorinated motifs (See Figure S7 for
BuOH).
The extension of a trifluoromethyl group is illustrated in Figure 8. Introducing a C–F group in
the vicinal position, leading to I15, is liponeutral, but a large logP reduction is observed when
the C–F group is added in the -position, leading to I10. The comparative operations with a
gem-difluoro motif lead to very different results: the vicinal pentafluorination (I19) leads to a
very large logP increase, while the skipped pentafluorinated I14 has a reduced logP. Hence,
comparing the lipophilicity of the vicinal and skipped motifs with the same number of fluorines
shows that the latter have a lower lipophilicity. This may be rationalized by C–F dipole
annihilation in the vicinal motif, and by the polarized C–H bonds of the CH -group within the
2
skipped fluorination motif (the chemical shift of CF CH CF - protons is 2.88 ppm, compared
3
2
2
to 2.15 ppm for a methylene group adjacent to a pentafluoroethyl group).
2
2
ACS Paragon Plus Environment

Page 23 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 8. Motif extension involving a trifluoromethyl group. (see Figure S8 for BuOH).
The “reverse” motif extension operation, i.e. extending a mono- or difluorinated motif (cf. I5,
I8, I9, I12) with a trifluoromethyl motif always leads to a logP increase which can be very
large (e.g. I8I19: +1.01 logP units) or very modest (e.g. I9I10: +0.09 logP units).
Fluorine motif reorganization (same numbers of fluorines, same carbon framework)
In motif reorganization, the fluorine atoms of a given motif are re-distributed on the same
3
2
carbon skeleton. The trend identified by Müller (Figure 1) regarding the lower lipophilicity
of vicinal difluorinated compounds compared to the corresponding geminal analogues is also
observed here (Figure 9). The vicinal difluorinated I2 has a much lower logP than I7/I8, with
a difference of up to -0.6 logP units for the latter. A similar observation is made for the butanol
compounds (compare G2 with G7/G9, see Figure S9).
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Interestingly, our data show that a similar operation for a trifluoromethyl group, i.e. changing
a “geminal” trifluorinated motif to the corresponding vicinal trifluorinated motifs, also results
in a significant decrease in lipophilicity (I16I5/I6), with logP differences of up to -0.7 units.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 9. Motif reorganizations (See Figure S9 for the BuOH series).
As a further observation, Müller’s data in Figure 1 show that converting a –CH –CF group to
2
3
a vicinal –CHF–CH F group leads to a reduction in lipophilicity. In proposing the vicinal
2
difluoroethylidene group as a bioisostere for a trifluoromethyl group, Gilmour has further
demonstrated that this reduction is substantial when modifying the aliphatic side chain of
®
Gilenya in this way, and showed that >1 logP unit reductions are possible in side chains of
3
3
varying lengths. The 1.11 logP decrease observed going from I16I2 fully agrees with
Gilmour’s data.
Fluorine ‘deletions’ from perfluoroalkyl groups
2
4
ACS Paragon Plus Environment

Page 25 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
36,38,39
As shown above in Figure 1, a CF H motif leads to a lower lipophilicity than a CF motif.
2
3
Hence, converting the widely used CF group into a CF H group can be used to achieve a
3
2
reduction in lipophilicity. Indeed, the same observation was made for I16 and also for I15
Figure 10). Such a lipophilicity reduction operation will be of even more interest when
working with larger perfluoroalkyl motifs. Our group had already shown that a significant logP
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
3
1
reduction can be thus obtained for a pentafluoroethyl group (compare I19 with I13), and this
observation is confirmed for the butanol series (G16G11). In both cases, the tetrafluorinated
compounds have a much lower lipophilicity than the nonfluorinated parents, which has
precedent in the literature.50
Figure 10. Single fluorine ‘deletions’ from perfluoroalkylated carbon motifs.
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Even for a nonafluorinated ‘ponytail’ as in I22, replacing a terminal C–F bond with a C–H
bond to give an octafluorinated motif as in I20 leads to a large logP reduction. Removing an
internal fluorine from a perfluoroalkyl group also leads to a logP reduction: compare G18 with
commercially available G17.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The ‘deletion’ of two geminal fluorines from a perfluorinated moiety also reduces the
lipophilicity, as expected from comparing a pentafluorinated with a trifluorinated compound
(
Figure 11). Very large logP reductions are achieved by replacing an internal CF group of a
2
perfluoroalkyl group by a CH group as shown by comparing G18 with G15 and I21 with I14.
2
Figure 11. Geminal fluorination ‘deletion’ effects.
2
6
ACS Paragon Plus Environment

Page 27 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Finally, as illustrated in Figure 2 above, this type of operation with the trifluoromethyl group
of a polyfluoroalkyl moiety also leads to a drastic lipophilicity reduction, and all the relevant
examples are gathered in Figure S10.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Chain extension
Extending the size of an aliphatic substituent is sometimes required to fill hydrophobic pockets
in binding sites. However, the addition of methyl groups, or the insertion of methylene groups,
usually leads to an increase in lipophilicity, which is often undesired. Hence, it is of interest to
have a panel of extensions available that do not lead to such a logP increase. As mentioned in
the introduction, Müller had identified that aliphatic chain extension is liponeutral when a CF2
3
2
group is simultaneously replaced by a vicinal difluoro group, and our group has shown that
extending a terminal polyfluorinated group by a (nonfluorinated) methyl group even leads to a
logP reduction (summarized in Figure S11).31
Figure 12. Chain extensions with motif rearrangements that result in logP reduction.
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Our data (Figure 12) show that chain extension with simultaneous modification of a CF H
2
group into a CHF-CH F group leads to a small logP reduction (cf. G7I2 and D3G2, Figure
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
12). In addition, our data also show that chain extension of a trifluoromethylated compound is
also possible with concomitant logP reduction using similar operations. Changing a C–F with
a fluoromethyl group reduces the logP: the logP of 3,3,3-trifluoropropanol D4 is higher than
that of 3,3,4-trifluorobutanol G6 ( 0.14 units) and that of 4,4,4-trifluorobutanol G14 is higher
than that of 4,4,5-trifluoropentanol I6 ( 0.33 units). In addition, carrying two fluorine atoms
forward with the extra methyl group also leads to a lipophilicity reduction: compare the logP’s
of D4 and G14 with those of G8 and I5 respectively. The latter has a logP reduction of 0.39
logP units.
We also observed that chain extension of a vicinal pentafluoroethyl group by insertion of a
methylene group leads to a lipophilicity reduction (Figure 13).
Figure 13. Chain extension of pentafluoroethyl motifs resulting in logP reduction.
Even more interesting is how extension of nonfluorinated chains can be achieved without logP
increase by exploiting fluorination. The scope is huge: from the summary table (Figure 3), it is
easily seen that 1-butanol G1 (logP +0.88) has a higher lipophilicity than many fluorinated 1-
2
8
ACS Paragon Plus Environment

Page 29 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
pentanol derivatives (namely I2–I11). Equally, 1-propanol D1 (logP +0.30) has a higher
lipophilicity than many fluorinated 1-butanols (G2–G8). Remarkably, given that the logP of
1-pentanol I1 (+1.51) is so much larger than that of 1-propanol D1 (+0.30), chain extension of
the latter by two carbon atoms having terminal vicinal difluorosubstitution, as in 4,5-
difluoropentanol I2 (logP +0.11), still leads to a lipophilicity reduction.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Further analysis of logP trends is given in the SI (Section 2.1)
Finally, lipophilicity trends for a vicinal trifluorinated motif were further investigated by
incorporation in the evenamide analogue K1 (Table 3). Because of the presence of the
ammonium group, logD values were measured at pH 7.4, this time using AstraZeneca’s routine
shake-flask method involving UV-detection for concentration determination. The change in
pKa(H) of the amine due to the fluorination of the remote alkoxy group is expected to be very
minimal and this factor was thus assumed to have a negligible influence on the logD of the
7
.4
evenamide analogues.
Table 3. Analysis of fluorinated motifs when incorporated in evenamide H1 and analogue K1.
Hu
Rat
Solubilityb
PPBc
hERGf
Structure
logD7.4a
Mics
Heps
(µM)
(%free)
(µM)
CLintd Clinte
+
2.331
779
23
47
49
28
29
<3
7
91
34
1.8
15
+
1.6
870
1.731
+
1000
38
7.6
9.6
+2.3
>1000
10
105
2
9
ACS Paragon Plus Environment