Enantioselective Anion Recognition by Chiral Halogen-Bonding [2]Rotaxanes

Article abstract of DOI:10.1021/jacs.7b06144

The application of chiral interlocked host molecules for discrimination of guest enantiomers has been largely overlooked, which is surprising given their unique three-dimensional binding cavities capable of guest encapsulation. Herein, we combined the stringent linear geometric interaction constraints of halogen bonding (XB), the noncovalent interaction between an electrophilic halogen atom and a Lewis base, with highly preorganized and conformationally restricted chiral cavities of [2]rotaxanes to achieve enantioselective anion recognition. Representing the first detailed investigation of the use of chiral XB rotaxanes for this purpose, extensive ¹H NMR binding studies and molecular dynamics (MD) simulation experiments revealed that the chiral rotaxane cavity significantly enhances enantiodiscrimination compared to the non-interlocked free axle and macrocycle components. Furthermore, by examining the enantioselectivities of a family of structurally similar XB [2]rotaxanes containing different combinations of chiral and achiral macrocycle and axle components, the dominant influence of the chiral macrocycle in our rotaxane design for determining the effectiveness of chiral discrimination is demonstrated. MD simulations reveal the crucial geometric roles played by the XB interactions in orientating the bound enantiomeric anion guests for chiral selectivity, as well as the critical importance of the anions' hydration shells in governing binding affinity and enantiodiscrimination.

Full text of DOI:10.1021/jacs.7b06144

Subscriber access provided by UNIV LAVAL
Article
Enantioselective Anion Recognition by Chiral Halogen Bonding [2]Rotaxanes
Jason Y. C. Lim, Igor Marques, Vítor Félix, and Paul D Beer
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b06144 • Publication Date (Web): 04 Aug 2017
Downloaded from http://pubs.acs.org on August 4, 2017
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 free 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 accessible to all
readers and 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.
Journal of the American Chemical Society 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 16
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Enantioselective Anion Recognition by Chiral Halogen Bonding
[2]Rotaxanes
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Jason Y. C. Lima, Igor Marquesb,c, Vítor Félixb,c and Paul D. Beera,*
a Chemistry Research Laboratory, Department of Chemistry, University of Oxford, Mansfield Road, Oxford, OX1
3TA (UK). Email: paul.beer@chem.ox.ac.uk. b Department of Chemistry, CICECO – Aveiro Institute of Materials,
University of Aveiro, 3810-193 Aveiro, Portugal. c Department of Medical Sciences, iBiMED – Institute of Biomed-
icine, University of Aveiro, 3810-193 Aveiro, Portugal.
ABSTRACT: The application of chiral interlocked host molecules for discrimination of guest enantiomers has been
largely overlooked, which is surprising given their unique three-dimensional binding cavities capable of guest encap-
sulation. Herein, we combined the stringent linear geometric interaction constraints of halogen bonding (XB), the non-
covalent interaction between an electrophilic halogen atom and a Lewis base, with highly preorganized and conforma-
tionally restricted chiral cavities of [2]rotaxanes to achieve enantioselective anion recognition. Representing the first
1
detailed investigation of the use of chiral XB rotaxanes for this purpose, extensive H NMR binding studies and mo-
lecular dynamics (MD) simulation experiments revealed that the chiral rotaxane cavity significantly enhances enantio-
discrimination compared to the non-interlocked free axle and macrocycle components. Furthermore, by examining the
enantioselectivities of a family of structurally similar XB [2]rotaxanes containing different combinations of chiral and
achiral macrocycle and axle components, the dominant influence of the chiral macrocycle in our rotaxane design for
determining the effectiveness of chiral discrimination is demonstrated. MD simulations reveal the crucial geometric
roles played by the XB interactions in orientating the bound enantiomeric anion guests for chiral selectivity, as well as
the critical importance of the anions’ hydration shells in governing binding affinity and enantio-discrimination.
INTRODUCTION
discrimination when judiciously decorated with chiral
functionalities. Although MIMs have been successfully
exploited for the recognition of charged guest species,17–
Stimulated by the exceptional levels of enantioselec-
tivity found in biotic systems, supramolecular chiral
recognition, the ability of a chiral host molecule to selec-
tively bind one guest enantiomer over the other, has
witnessed a surge in interest in recent years, driving ad-
vances in analytical chemistry,1–3 drug analysis,4 asym-
metric synthesis and catalysis.5–9 In nature, guest enanti-
omer selectivity occurs via structurally well-defined
binding pockets capable of encapsulating the target
guest enantiomer in three-dimensional chiral cavities
with exquisite complementarity.10 Indeed, the im-
portance of guest encapsulation in the chiral discrimina-
tion process has been demonstrated by the high levels of
selectivity attained by abiotic receptors possessing struc-
turally-enclosed chiral cavities in the form of macrocy-
cles11–15 and cryptand-like strapped calixpyrroles.16 In
this respect, the preorganized cavities of mechanically-
interlocked molecules (MIM) such as rotaxanes and cat-
enanes offer unique opportunities for enantioselective
19
in particular anions,20–22 their potential for chiral guest
recognition has been largely overlooked,23 which may be
a consequence of the challenging synthesis of chiral
MIMs. However, recent years have witnessed significant
advancements in their synthetic methodologies 24–30 most
notably via metal template protocols.31–35 Rare examples
of chiral binding studies involving MIMs to date include
a qualitative investigation with a racemic mechanically
bonded-chiral rotaxane36 and recently, a chiral catenane
which exhibits modest discrimination between cationic
chiral diamines.34 To the best of our knowledge, chiral
recognition by rotaxane host systems has yet to be stud-
ied in detail.
Amongst the non-covalent interactions employed in
solution-phase chiral discrimination, halogen bonding
(XB), the highly directional interaction resulting from the
sigma hole of an electron-deficient halogen atom and a
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 16
Lewis base,37–39 remains virtually unexploited.40,41 We
containing either a chiral macrocycle and achiral axle
1
2
3
4
5
6
7
8
have recently demonstrated the enhanced enantioselec-
tive recognition properties of chiral anions by acyclic
chiral XB host molecules compared to their hydrogen
bonding (HB) analogues.42,43 Herein, we describe the in-
tegration of XB-donor groups into a chiral macrocycle,44
and by using an active metal template approach, into
unprecedented chiral XB [2]rotaxane host structural
frameworks. Importantly, we demonstrate a XB rotaxane
containing both chiral axle and macrocycle units exhibits
notably enhanced chiral discrimination properties com-
pared to its free components. By comparing the enanti-
oselective anion binding properties of structurally-
related XB [2]rotaxanes containing different combina-
tions of chiral and achiral macrocycle and axle compo-
nents (Chart 1), we also show that the chiral macrocycle
unit was the main contributor to the overall enantiose-
lective recognition process. Furthermore, molecular dy-
namics (MD) simulations carried out in wet organic me-
dia highlight the crucial geometric roles played by XB
interactions and the importance of enantiomer anion
substrate hydration in determining the experimentally
observed enantioselectivities.
(Rotaxane 2), or an achiral macrocycle and chiral axle
(Rotaxane 3) were prepared for comparison. Each chiral
rotaxane contains four convergent XB-donor groups po-
sitioned around the central interlocked host cavity re-
gion of the rotaxane for binding carboxylate/ phosphate
motifs of chiral oxoanion guest species. The bulky chiral
groups of the guest oxoanions are thus designed to be
oriented towards the asymmetric motifs of the three di-
mensional rotaxane binding cavity to achieve optimal
enantioselectivity.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The active metal template (AMT) strategy51,52 used to
synthesise the target chiral XB rotaxanes involved axle
formation by copper(I)-catalysed azide-alkyne cycload-
dition (CuAAC) reactions through the cavity of a XB
macrocycle, by capitalising on the endo/exo-
conformational flexibility of the iodotriazole groups for
copper(I) N-ligation.53,54 As shown in Scheme 1A, to as-
semble the macrocycle components (4.XB and 5.XB),
direct CuAAC iodotriazole formation from either
BINOL- or biphenyl-bis(azide) with an aryl (pro-
to)alkyne,55 followed by deprotection of the methoxyme-
thyl acetal protecting group under acidic conditions af-
forded the bis-phenol macrocycle precursors. Cyclisation
was achieved by reacting equimolar quantities of the bis-
phenol precursors with triethylene glycol-bis(tosylate) in
the presence of cesium carbonate base to give the target
macrocycles. The ability of macrocycle 4.XB to coordi-
RESULTS AND DISCUSSION
Design and Syntheses of Chiral XB [2]Rotaxanes. Due
to its rigidity, stable chiral configuration and ease of
chemical functionalisation,45 the chiral (S)-1,1’-bi-2-
naphthol (BINOL) motif was incorporated into the de-
sign of the XB macrocycle component of [2]rotaxane 1
(Chart 1). A methylene spacer inserted between the (S)-
BINOL and iodotriazole XB-donor motifs allows suffi-
cient flexibility for strong anion binding by convergent
XB interactions, while maintaining the necessary rigidity
and close proximity to the chiral (S)-BINOL motif for
chiral discrimination.42 Rotaxane 1 also contains a chiral
3,5-bis-iodotriazole pyridinium axle component derived
from enantio-pure (S)-serine that threads through the
macrocycle cavity.46 As shown in Chart 1, hydrogen
bonding interactions between the pyridinium methyl
group of the axle and the polyether region of rotaxane
1’s macrocycle component is expected to lock both com-
ponents rigidly together,47–49 sandwiching the chiral ani-
on guest in the restricted chiral binding space between
them (yellow shaded region). The anion is expected to be
encapsulated by chiral motifs on both the axle and mac-
rocycle of the rotaxane, allowing them to contribute syn-
ergistically to enantio-discrimination.50 To further appre-
ciate the roles played by the individual axle and macro-
1
nate copper(I) was established by H NMR titration ex-
periments with [Cu(CH
ing Information).
3
CN)
4
]PF (Section S3.1, Support-
6
Thereafter, the neutral XB chiral rotaxanes 6, 7 and 8
were prepared using 1.0 equivalent of macrocycle (4.XB
or 5.XB) and [Cu(CH
3
CN)
4
]PF in the presence of 4.0 and
6
8.0 equivalents of pyridine-3,5-bis(iodoalkyne) 9 and the
appropriate axle azide precursor respectively under an
inert nitrogen atmosphere (Scheme 1B). As direct rotax-
ane formation via two CuAAC reactions in one pot with
a bis-alkyne axle precursor is unprecedented, the AMT
reaction for preparing rotaxane 6 was monitored by elec-
trospray ionisation mass spectrometry (ESI-MS), which
revealed that rotaxane assembly occurred in two steps
(Section S3.2, Supporting Information). Initially, a Cu-
AAC reaction between one of the iodoalkyne groups of 9
and the terminal azide axle precursor gave an asymmet-
rically-substituted pyridine terminal iodoalkyne inter-
mediate, which then reacted with a second azide axle
precursor catalysed by 4.XB-bound copper(I) through
the macrocycle cavity to afford the rotaxane product
(Scheme 1B). Complete consumption of the axle azide
cycle components for chiral discrimination by
a
[2]rotaxane, two structurally-similar rotaxane analogues
2
ACS Paragon Plus Environment

Page 3 of 16
Journal of the American Chemical Society
Chart 1. Structures of the chiral XB [2]rotaxanes and anions investigated.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthesis of chiral XB macrocycle and rotaxanes.
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 16
and iodoalkyne precursors was observed after 3 days, and accessible enantio-pure forms, as well as chiral
1
2
3
4
5
6
7
8
whereupon ESI mass spectrometric analyses of the re-
spective crude rotaxane reaction mixtures showed the
presence of rotaxanes 6, 7 and 8 at m/z 2838, 2582 and
2678 [M + H]+ respectively. Following de-complexation
of the copper(I) catalyst with basic aqueous ethylenedi-
amine-tetraacetate (EDTA), the rotaxanes were isolated
via chromatographic purification in yields of up to 48%.
Interestingly, the chiral (S)-serine-derived terminal azide
axle precursor afforded slightly higher rotaxane yields
(rotaxanes 6 and 8) compared to its achiral counterpart
containing an unfunctionalized propyl spacer (for rotax-
ane 7). This may suggest the involvement of secondary
non-covalent interactions (e.g. HB) between the azide
precursor and the macrocycle which facilitates the
threading process necessary for CuAAC-AMT rotaxane
formation.
BINOL-derived phosphates, which have found im-
portant applications in asymmetric catalysis.57
Figure 1A shows the perturbations of various proton
resonances of rotaxane 1 upon addition of the NBoc-(S)-
proline carboxylate anion, as a representative example.
Downfield shifts were seen for signals arising from the
(S)-BINOL protons H
as proton H on the terphenyl stopper units proximal to
the rotaxane binding cavity. Notably, negligible shifts
were seen for the inner pyridinium proton (H ) of the
a
and H
b
on the macrocycle, as well
9
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
axle flanked by iodotriazole groups, suggesting that true
to our rotaxane host design, the chiral anion is bound in
the space between the macrocyclic and axle components
as shown in Chart 1. Importantly, these observations
also suggest that the chiral units on the axle and macro-
cycle are both interacting with the anion bound in the
chiral space between them, and hence are both contrib-
uting synergistically to chiral discrimination.
Exclusive mono-methylation of the rotaxanes’ axle
pyridine moiety was then achieved by reaction of each
rotaxane with iodomethane to afford the desired cationic
chiral XB [2]rotaxanes, and anion exchange to their hex-
afluorophosphate salts afforded hosts 1, 2 and 3. Nota-
In the presence of the opposite guest enantiomer
(NBoc-(R)-proline carboxylate), significantly different
interactions were seen for the proton environments sur-
rounding the rotaxane cavity. Perhaps due to its unique
position immediately adjacent to the BINOL moiety on
the macrocycle component, the perturbations of the sig-
1
bly, the H NMR spectra of 1 and 3 showed distinct evi-
dence of increased conformational rigidity following
methylation of the axle pyridine group. For the neutral
1
rotaxane congeners 6 and 8, the H NMR spectra re-
nals arising from the diastereotopic protons H were
f
vealed dynamic behavior with the macrocyclic compo-
nents adopting various conformations around the chiral
axle in slow exchange on the NMR timescale. However,
axle methylation led to dramatic simplification of the
rotaxanes’ NMR spectra to give well-defined proton sig-
nal environments (see Figure S2-25, Supporting Infor-
mation). This indicated that the macrocycle components
of cationic rotaxanes 1 and 3 were now predominantly
locked around the axle, brought about by HB interac-
tions between the axle pyridinium group and the macro-
cycles’ polyether unit. Full synthetic details of all macro-
cycles, axle precursors and rotaxanes are provided in the
Supporting Information, together with unequivocal evi-
dence of the interlocked nature of the cationic rotaxanes
most sensitive to changes in the conformations and elec-
tronic environment of the binding site in the presence of
different anion guest enantiomers. For instance, as
shown in Figure 1B, the binding isotherm resulting from
the shifts of Hf1 plotted as a function of anion equivalents
achieved saturation more quickly for NBoc-(S)-proline
carboxylate (black dotted curve) compared to that of the
opposite guest enantiomer (red dotted curve). More
dramatically, both enantiomers of the BINOL-PO
perturbed the Hf1 signals of rotaxane 1 in different direc-
tions, with 10.0 molar equivalents of (S)-BINOL-PO -
eliciting a significant upfield shift of -0.13 ppm and the
equivalent amount of (R)-BINOL-PO -
giving a much
4
- anion
4
4
smaller downfield shift of only +0.012 ppm. As shown, the
highly preorganized and conformationally restricted
chiral binding space of rotaxane 1 is capable of effective-
ly discriminating between different guest enantiomers.
Non-linear regression analysis of the binding isotherms
obtained from monitoring shifts of appropriate proton
signals in the rotaxane binding cavity during the titra-
tions (see Figure S4-1, Supporting Information), using
the WinEQNMR2 software58 with a host-guest 1:2 stoi-
chiometric binding model, determined association con-
stants (K) as well as the corresponding magnitudes of
1
1-3 obtained by two-dimensional H-1H ROESY experi-
ments.
Chiral Recognition Studies. The chiral anion discrimi-
1
nation properties of rotaxane 1 was studied using H
NMR titration experiments, where tetrabutylammonium
(TBA) salts of chiral anions56 (Chart 1) were added as
separate enantiomers in increasing quantities to individ-
ual solutions of the rotaxane in d
6
-acetone/ D O 98:2
2
(v/v). Derivatives of naturally-occurring amino acids
chiral selectivity (ξ = K
S/ K
R) for each chiral anion pair
were chosen for study due to their biological importance
4
ACS Paragon Plus Environment

Page 5 of 16
Journal of the American Chemical Society
reported in Table 1. Notably, the second association con-
stants (K
) were much smaller in magnitude (< 10 M-1)
compared to the first (K ) for all anion pairs studied,
To appreciate the importance of the rotaxane’s chiral
1
2
3
4
5
6
7
8
2
interlocked cavity for enantio-discrimination and the
roles played by its XB-donor iodotriazole motifs, analo-
gous 1H NMR binding studies were performed using the
free axle and macrocycle components as well. For free
chiral axle 10 (Figure 2A), the most significant perturba-
tions seen upon NBoc-(S)-proline carboxylate addition
were the downfield shifts of the internal pyridinium
1
suggesting that one chiral anion guest is predominantly
bound within the rotaxane chiral binding cavity, while
the second was weakly associated with the bound com-
plex, probably at its periphery. Hence, the values of K
1
were solely used to determine chiral selectivity as only
the first anion bound is held within the host cavity and
therefore properly influenced by its chiral environment.
Amongst the four chiral anion pairs investigated, the
most impressive and best enantioselectivities were seen
9
proton H signals. Although the amide linkages adjacent
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
to the stopper units of the axle are potentially capable of
hydrogen bonding with the anion, negligible shifts of the
proximal stopper proton H were seen, unlike those of
3
for NBoc-proline carboxylate and BINOL-PO -
4
, each pos-
rotaxane 1, clearly suggesting that the chiral anions were
mainly bound by the central pyridinium-3,5-
bis(iodotriazole) motif of the free axle. The dominance of
the charge-assisted XB interactions for carboxylate bind-
ing over competing HB from the amide groups also
gives further credence to the proposed anion binding
mode of rotaxane 1, where the oxoanion groups of the
chiral guests are mainly held by the iodotriazoles at the
core of the rotaxane to orientate their asymmetric por-
tions for interaction with the binding site chiral func-
tionalities (vide infra). In addition, from the small shifts
sessing rigid chiral units. The slightly improved discrim-
ination seen for NBoc-tryptophan carboxylate compared
to NBoc-leucine may imply the contribution of steric
effects in chiral selectivity.59,60
of the H
4
and H resonances around the free axle’s chiral
5
centres, the anions’ binding location appears to limit the
stereo-specific host-guest interactions with the bulky
asymmetric units responsible for chiral discrimination.
Unsurprisingly, WinEQNMR258 determination of the
association constants of axle 10 with the various chiral
anion pairs reveal generally greatly diminished enanti-
oselectivity (Table 1) compared with rotaxane 1.
1H NMR chiral discrimination studies were per-
formed using dicationic macrocycle 11.XB (synthesis
described in Section 1, Supporting Information) instead
of its neutral congener 4.XB, as the latter did not show
appreciable binding affinities for any of the chiral anions
in d
6
-acetone/ D O 98:2 v/v. Titrations of the NBoc-(S)-
2
proline carboxylate oxoanion with free dicationic macro-
cycle 11.XB elicited distinct upfield shifts of the macro-
cycle (S)-BINOL aromatic proton signals H
a
, H
b
, H
c
and
Hd
(Figure 2B). For comparison, we also synthesized the
dicationic (proto)triazolium halogen-free macrocycle
analogue 11.HB (Section 1, Supporting Information) and
studied its chiral discrimination properties in the same
solvent medium. The addition of NBoc-(S)-proline car-
boxylate gave negligible shifts of the aromatic BINOL
proton resonances on macrocycle 11.HB despite clear
evidence of anion binding which brought about notable
Figure 1. (A) Partial 1H NMR spectra of rotaxane 1 with 0.0,
1.0, 2.0 and 5.0 equivalents of NBoc-(S)-proline carboxylate;
(B) Changes in the chemical shift of proton Hf1 with increas-
ing quantities of both enantiomers of NBoc-proline-COO-
perturbations of H
f
and H (Fig. S4-8, Supporting Infor-
g
-
and BINOL-PO
4
([rotaxane 1] = 1.0 mM, d
6
-acetone/ D O
2
mation). This indicates that the stringent linearity of the
XB-carboxylate interactions was instrumental in direct-
ing the chiral anion guest for interaction with macrocy-
98:2, T = 298 K).
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
cle 11.XB’s (S)-BINOL group for chiral discrimination.
Page 6 of 16
1
2
3
4
This difference is reflected in their respective magni-
tudes of chiral selectivity for NBoc-proline carboxylate,
with 11.XB (K
S
/ K = 0.70 ± 0.01) proving superior to
R
11.HB (K / K = 0.92 ± 0.05) (see Table S3-1, Supporting
S
R
5
Information for enantioselectivity of 11.HB towards all
chiral anion pairs). However from Table 1, the enantio-
selectivities of 11.XB still pales in comparison to those of
rotaxane 1, most notably for the conformationally high-
6
7
8
9
ly-rigid NBoc-proline carboxylate and BINOL-PO -
anion
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
enantiomers. Undoubtedly, this arises from the more
complete encapsulation of the chiral anion guest possible
with the three-dimensional interlocked binding cavity of
rotaxane 1, allowing a greater number of stereoselective
interactions to take place for more effective chiral dis-
crimination (vide infra).
After establishing the superiority of rotaxane 1 for
chiral anion discrimination compared to its free axle and
macrocycle components, we subsequently compared its
chiral selectivity with rotaxanes 2 and 3, possessing one
chiral and one achiral component each, via analogous 1H
NMR titration experiments to determine the importance
of each component for chiral discrimination. For both
rotaxanes 2 and 3, the presence of oxoanions caused
shifts of the BINOL and biphenyl macrocyclic aromatic
protons, respectively, as well as the alkyl proton envi-
ronments on the axle flanking the anion binding cavity
(Figures S4.2 and S4.4, Supporting Information). Owing
to the strong charge-assisted XB-interactions arising
from the common pyridinium-bis(iodotriazole) motif in
the rotaxanes’ central region, all anions were bound in
the chiral binding space between the macrocycle and
axle components likely in a similar manner as rotaxane 1
shown in Chart 1.
Figure 2. Partial 1H NMR spectra of (A) free axle 10 and (B)
macrocycle 11.XB with 0.0 and 5.0 equivalents of NBoc-
proline-COO- ([host] = 1.0 mM, d
6
-acetone/ D O 98:2, T = 298
2
K).
which contains unfunctionalized achiral propyl spacer
units, dramatically reduced its anion binding affinity
compared to rotaxanes 1 and 3, further emphasizing the
important roles of these HB motifs for anion binding. In
fact, the reduction in anion affinity was so significant
that the binding studies of rotaxane 2 had to be per-
formed in a less-competitive solvent mixture of ace-
tone/D O 99:1 (v/v) to achieve appreciable proton shifts
2
A closer inspection of the anion affinities of all XB
chiral host systems studied in Table 1 revealed that the
general anion binding strengths decreased in the order
of axle 10 > rotaxane 1 ≈ rotaxane 3 > macrocycle 11.XB >
rotaxane 2. Despite having four XB donor groups, the
rotaxanes bound the chiral oxoanions surprisingly more
weakly than free chiral axle 10 possessing just two iodo-
triazole units likely due to steric hindrance limiting ani-
on access to the rotaxanes’ spatially more restricted
binding cavities (vide infra). Interestingly, rotaxanes 1
and 3, which possess HB-donor groups on their axle
components (i.e. NHBoc and amide), were able to bind
the anions generally more strongly than even dicationic
macrocycle 11.XB, suggesting the important influence of
these secondary non-covalent interactions for anion as-
sociation (vide infra). In stark contrast, the absence of
these HB motifs on the axle component of rotaxane 2,
for reliable values of association constants to be ob-
tained. However, although all three rotaxanes investi-
gated possess the same structural skeleton, their respec-
tive anion binding cavities are very different from one
another due to the various combinations of chiral and
achiral components. The important consequence of
which is that each XB rotaxane host exhibits a unique
chiral anion guest preference. For instance, NBoc-(S)-
tryptophan carboxylate is bound most strongly by rotax-
ane 1, while 2 and 3 bind NBoc-(R)-proline carboxylate
with the highest affinity. It is nevertheless noteworthy
that the NBoc-proline carboxylate enantiomers exhibit
consistently strong anion binding for all chiral XB hosts,
with smaller variations in binding affinity compared to
the other chiral anion pairs. Likely due to their smaller
size compared to all the other anion pairs investigated,
they are expected to encounter less steric repulsions with
6
ACS Paragon Plus Environment

Page 7 of 16
Journal of the American Chemical Society
Table 1. Anion association constants (K/ M-1) for chiral anion pairs and the magnitudes of chiral selectivity (ξ) by
rotaxanes 1, 2, 3, free axle 10 and macrocycle 11.XB.a
1
2
3
4
5
6
7
8
9
K
S
K
R
ξ
K
S
K
R
ξ
Rotaxane 1 b
1511 ± 131
2886 ± 203
1465 ± 52
Rotaxane 2 c
471 ± 38
(NBoc-Leu)-
(NBoc-Trp)-
(NBoc-Pro)-
2450 ± 154
4990 ± 196
4298 ± 125
547 ± 16
1.62 ± 0.17
1.73 ± 0.14
2.93 ± 0.13
0.47 ± 0.04
138 ± 11
471 ± 37
1719 ± 59
354 ± 22
0.29 ± 0.03
0.62 ± 0.06
0.66 ± 0.04
0.56 ± 0.06
760 ± 45
2586 ± 110
631 ± 48
-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(BINOL-PO
(NBoc-Leu)-
(NBoc-Trp)-
(NBoc-Pro)-
4
)
1170 ± 99
Rotaxane 3
Axle 10 d
K
K
1
2
= 2100 ± 147
= 62 ± 5
K
K
1
2
= 2244 ± 99
= 49 ± 3
ξ1
ξ1
ξ1
ξ1
= 0.94 ± 0.08
= 0.85 ± 0.07
= 0.90 ± 0.03
= 1.19 ± 0.06
K
K
1
2
= 6027 ± 321
= 166 ± 9
K
K
1
2
= 4630 ± 86
= 142 ± 11
ξ
ξ
ξ
ξ
1
1
1
1
= 1.30 ± 0.07
= 1.41 ± 0.10
= 1.22 ± 0.04
= 0.82 ± 0.04
K
K
K
K
K
K
1
2
= 1905 ± 107
= 34 ± 2
K
K
K
K
K
K
1
2
= 2253 ± 138
= 32 ± 3
K
K
K
K
K
K
1
2
= 2581 ± 91
= 56 ± 2
K
K
K
K
K
K
1
2
= 1836 ± 117
= 95 ± 6
1
2
= 4031 ± 100
= 55 ± 2
1
2
= 4468 ± 131
= 55 ± 2
1
2
= 6193 ± 89
= 33 ± 1
1
2
= 5057 ± 128
= 39 ± 1
-
(BINOL-PO
4
)
1
2
= 4159 ± 100
= 86 ± 3
1
2
= 3491 ± 148
= 93 ± 4
1
2
= 4861 ± 122
= 24 ± 1
1
2
= 5942 ± 270
= 17 ± 1
Macrocycle 11.XB e
635 ± 31
(NBoc-Leu)-
(NBoc-Trp)-
(NBoc-Pro)-
423 ± 3
0.67 ± 0.03
0.66 ± 0.03
0.70 ± 0.01
1.58 ± 0.11
space left blank
324 ± 6
488 ± 21
3495 ± 64
854 ± 43
4961 ± 47
-
(BINOL-PO
4
)
540 ± 24
a Values of K calculated using the WinEQNMR2 software58 using a host-guest 1:2 binding model in d
itoring the perturbation of proton Hf1 for all hosts unless otherwise stated (see Section S4, Supporting Information for more
details); ξ = K / K
only for the first binding event; anions added as TBA salts; [host] = 1.0 mM, T = 298 K. b Although rotaxane 1
bound chiral anions with host-guest 1:2 binding stoichiometry, K
enantiomers of (BINOL-PO
99:1 due to weak binding. With the exception of (BINOL-PO
was monitored instead; values of K fit using a host-guest 1:1 binding model, with no evidence of a second binding event oc-
curring. d Internal pyridinium proton H monitored. e Hf1 monitored, K < 10 M-1.
6
-acetone/ D
2
O 98:2 mon-
S
R
2
< 10 M-1 for all cases. Axle proton H
2
monitored for both
-acetone/ D O
c
4
)- due to small shifts of Hf1 seen for the (R)-guest. Titrations were performed in d
6
2
4
)- for which Hf1 gave substantial shifts, axle pyridinium proton H
1
1
2
the bulky chiral groups decorating the hosts’ binding
cavities, allowing deeper penetration and hence stronger
binding.
only small reductions in selectivity, with NBoc-leucine
carboxylate being a notable exception. However, substi-
tuting the chiral macrocycle component in rotaxane 1 for
an achiral counterpart in rotaxane 3 resulted in expect-
edly poor discrimination between both enantiomers for
all chiral anion pairs investigated. In fact, the general
enantioselectivity for rotaxane 3 was found to be even
poorer than that of axle 10, with the bulky and rigid
Figure 3 summarizes the guest enantiomer prefer-
ence and magnitudes of enantioselectivity by all chiral
XB host molecules from Table 1 visually, which reveals
further insights regarding the chiral discrimination be-
havior of the rotaxanes. Regarding the magnitudes of
chiral selectivity, given that only one rotaxane compo-
nent surrounding the anion binding site is chiral in ro-
taxanes 2 and 3, fewer enantio-specific host-guest inter-
actions with the bound chiral anion can be expected
compared to rotaxane 1, resulting in generally poorer
enantioselectivity. Interestingly, replacing the chiral axle
of rotaxane 1 with an achiral axle in 2 generally afforded
BINOL-PO -
enantiomers being most effectively discrim-
4
inated with a modest preference of only 1.19 ± 0.06 for
the (S)-enantiomer. This suggests that the macrocycle
components of our XB rotaxane design exert dominating
influences on their effectiveness in chiral discrimination,
which may be partly due to the greater rigidity and larg-
er steric bulk of the macrocycle’s chiral (S)-BINOL group
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Bar charts summarizing the enantiomer binding preference and magnitudes of chiral selectivity aspects of chiral
discrimination by rotaxanes 1, 2 and 3, as well as those for their chiral non-interlocked macrocycle and axle components. All
amino acid derivatives used are NBoc-protected.
as compared to the point-chiral (S)-serine- derived
asymmetric units of the axle.61
each chiral rotaxane anion binding cavity gives rise to
different contributions of attractive and repulsive non-
covalent interactions to anion binding. Hence, it is the
interplay of these multiple interactions that result in sub-
tle energy differences between the various rotaxane-
anion diastereomeric complexes, which govern enanti-
omer preference in each case, making it difficult to pre-
dict their behavior by extrapolation from the free chiral
axle and macrocycle host molecules alone. Nevertheless,
all point-chiral amino acid derivatives were found to
exhibit the same chiral selectivity pattern with the host
While the macrocycle component appears to domi-
nate the magnitude of chiral selectivity, its influences on
the rotaxanes’ enantiomer binding preferences are less
obvious. For instance, although it appears that the mac-
rocycle component of rotaxane 2, being the only chiral
component, is driving a generally similar preference for
binding the (R)-anion enantiomers as the free macrocycle
11.XB, the selectivity patterns observed are not entirely
identical, as exemplified with BINOL-PO
-. Rotaxane 3,
4
molecules (e.g. all (S)-selective for rotaxane 1). However
bearing only the chiral axle, surprisingly exhibits the
opposite enantiomer binding preference to that of the
chiral axle 10. Although these two examples seem to
suggest that the influence of the macrocycle outweighs
that of the axle in enantiomer binding preference as well,
rotaxane 1 displays the same preference as chiral axle 10,
contrary to the expected macrocycle-dominated behav-
ior. As previously noted, the structural uniqueness of
-
interestingly, the chiral BINOL-PO
4
enantiomers gener-
ally exhibited the opposite selectivities as the amino acid
derivatives (vide infra).
Molecular Modeling Studies. MD simulations per-
formed in solution offered further insights on the chiral
recognition behavior of rotaxanes 1-3, axle 10 and mac-
8
ACS Paragon Plus Environment

Page 9 of 16
Journal of the American Chemical Society
rocycle 11.XB. As the enantio-discrimination of the
XB and HB interactions from both interlocked compo-
1
NBoc-proline carboxylate enantiomers is largely repre-
sentative of the various amino acid-derived chiral guests
studied (Figure 3), both enantiomers were used as the
model guest species for this theoretical investigation.
The initial host-guest binding scenarios were generated
in the gas-phase via geometry optimizations with the
CAM-B3LYP functional using TeraChem,62–67 before be-
ing investigated in solution with MD simulations per-
formed using the AMBER16 software.68–70 The quantum
mechanics and classical force field calculations were per-
formed taking advantage of GPU acceleration in both
application suites.
nents. Notably, several water molecules bound to the
host by XB and HB interactions were found to actively
participate in anion binding by forming bridging HB
interactions with the carboxylate anion held in the rotax-
ane’s binding cavity (see Table S5-1, Supporting Infor-
mation). For instance, in the most frequent binding
structure (Figure 4A), one oxygen atom of the anion’s
carboxylate group participates in direct XB interaction
with the axle’s central pyridinium-3,5-bis(iodotriazole)
motif, while the other forms a HB network with water
molecules held together by a XB interaction with the
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
macrocycle iodotriazole unit (C-I···OH ), as well as by
2
the NBoc and amide units on the axle. In the second
most common binding arrangement (Figure 4B), the ani-
on’s carboxylate group is held by two direct XB interac-
tions from each iodotriazole on the macrocycle compo-
nent, and one from the axle. The spatial disposition of
the anion brings it near the chiral units flanking the
binding site, resulting in the signal perturbations from
the rotaxane’s axle and macrocycle components seen
during the 1H NMR titrations (Figure 1).
During the initial DFT geometry optimizations, two
starting binding scenarios were generated for each
NBoc-proline carboxylate guest enantiomer in each
host’s binding site (called S
A
, S
B, R
A
and R ), differing in
B
the orientation of the anion’s chiral pyrrolidine ring, as
detailed in Section S5 of the Supporting Information.
Due to the size of rotaxanes 1-3, the stopper groups of
their axle components were initially capped with tert-
butyl groups, affording the corresponding pseudo-
rotaxanes. The NBoc-proline enantiomers were then po-
sitioned in each rotaxane’s chiral interlocked binding
pocket, held by at least two putative XB interactions.
Using these DFT-optimized host-guest complexes (Fig-
ures S5-1 to S5-3, Supporting Information), the entire
rotaxane host structures were then restored by re-
attaching the bulky stopper units for the subsequent MD
simulations. Thereafter, each initial host-guest complex
was immersed in periodic cubic boxes composed of suit-
able number of water and acetone molecules in accord-
Owing to the spatial and geometric constraints of
rotaxane 1’s chiral binding cavity, MD simulations re-
vealed a generally less favorable association with NBoc-
(R)-proline carboxylate. Simulations starting from the R
B
initial binding scenario (Section S5, Supporting Infor-
mation) resulted in systematic decomplexation of the
anion along the five MD independent runs. Slightly
more preferential binding was observed for the R start-
A
ing scenario, with the most frequent binding arrange-
ment (Figure 4C) showing one weak and sporadically-
broken direct halogen bond from the rotaxane’s macro-
1
ance with the H NMR titrations (Table 1). The rotax-
anes’ components were described with the General
AMBER Force Field (GAFF)71,72 and the XB interactions
were modelled using the extra point of charge ap-
proach.73 The remaining force field parameters and sim-
ulation details are provided in Section S5.3, Supporting
Information. Five independent 20 ns MD runs were per-
formed for each initial binding arrangement, and the
subsequent analyses of the MD production data is based
on the most frequent binding arrangements obtained by
cluster analysis of the concatenated trajectory files as
detailed in Section S5.3, Supporting Information.
cycle component, which is assisted by
a
(N–
H)host···OH
2
···(O=C)guest bridge from the rotaxane axle’s
NBoc unit (average dimensions and frequencies of the
XB interactions are collected in Tables S5-2 and S5-3 in
the Supporting Information for the anions and water
molecules respectively). However, for a significant peri-
od of sampling, the anion was also found outside the
rotaxane binding pocket, interacting peripherally only
with the axle’s amide NH group. This dynamical behav-
ior suggests a significantly weaker interaction between
rotaxane 1 and the NBoc-(R)-proline carboxylate than
with the (S)-enantiomer, as illustrated in Fig. S5-4 (Sup-
porting Information), where the ratio of the host-guest
XB interaction frequency are plotted for both anion en-
antiomers. These structural findings stress the im-
portance of the XB interactions on the anion enantiose-
lective recognition of 1, corroborating the binding pref-
erences determined by 1H NMR titrations.
Rotaxane 1 exhibits distinct binding scenarios with
each enantiomer of NBoc-proline carboxylate. With
NBoc-(S)-proline two main binding arrangements were
obtained independently of the initial scenarios S
A
or S .
B
The anion is held within rotaxane 1’s chiral binding
pocket in the space between the axle and macrocycle
components throughout all simulation runs, bound by
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. (A) The most common binding arrangement of rotaxane 1 with NBoc-(S)-proline carboxylate, with the binding cavi-
ty expanded; (B) View of the second most common arrangement of NBoc-(S)-proline carboxylate within the rotaxane binding
cavity; (C) Most common binding arrangement of rotaxane 1 with the (R)-enantiomer. The XB interactions from the iodine
atoms are indicated by purple dashes, while HB interactions between water molecules are shown in orange and between wa-
ter molecules and the axle amide binding sites in pink. The stopper groups are not shown in (B) and (C) for clarity.
The modeling studies also provided clear insights on
the observed superiority of rotaxane 1 for chiral recogni-
tion compared with free axle 10 and macrocycle 11.XB.
From the two initial DFT-optimised binding scenarios
both hydrated NBoc-proline carboxylate guest enantio-
mers to be tightly bound to the axle’s central pyridini-
um-3,5-bis(iodotriazole) motif via direct XB interactions
(Figure 5A), consistent with the signal perturbations
1
for each anion enantiomer with macrocycle 11.XB (S
A
, S
B,
seen during the H NMR titrations in Figure 2A, and
R
A
and R ), and the single binding arrangement for each
B
accounts for its strong anion affinities (Table 1). Conse-
quently, both anion enantiomers are able to interact only
indirectly with the axle’s point-chiral amide binding
units through water bridges, which is in particular evi-
dent with NBoc-(R)-proline carboxylate, justifying the
axle’s general poor enantioselectivity. Furthermore, the
enantiomer with axle 10 (Figures S5-5 and S5-6 in the
Supporting Information), MD simulations were per-
formed using the same approach as rotaxanes 1-3 result-
ing in the representative anion binding complex struc-
tures shown in Figure 5. For axle 10, the lack of sterics
conferred by the macrocycle unit of rotaxane 1 allows
10
ACS Paragon Plus Environment

Page 11 of 16
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. (A) Illustrative binding scenarios of axle 10 (stoppers removed for clarity) with NBoc-(S)-proline (left) and NBoc-(R)-
proline (right); (B) Most frequent binding scenarios of free macrocycle 11.XB with NBoc-(R)-proline carboxylate (left) and the
opposite enantiomer (middle and right respectively). XB and HB interactions are depicted as dashed lines as described in Fig-
ure 4.
XB interactions to NBoc-(S)-proline carboxylate are
maintained throughout the sampling time, while those
to the (R)-guest enantiomer are sporadically interrupted,
with one halogen bond being occasionally established
with water molecules (Tables S5-4 and S5-5 in the Sup-
porting Information). This manifests in the observed
modest chiral selectivity for NBoc-(S)-proline carbox-
ylate.
zolium group and an indirect association to the second
iodotriazolium bridged by a water molecule (Table S5-5,
Supporting Information). Notably, this XB coordination
arrangement also allows the anion to adopt a binding
conformation oriented far away from the macrocycle (S)-
BINOL unit. For instance, the second most common
binding arrangement of NBoc-(S)-proline carboxylate
involves the entire anion being held above the achiral
portions of 11.XB (Figure 5B, right). Compared to the
restricted binding cavity of rotaxane 1, which forces both
anion enantiomers to adopt very different binding ar-
rangements, the poorly preorganized binding space of
macrocycle 11.XB allows the guest enantiomers to adopt
a greater range of spatially and geometrically more flex-
ible binding conformations. As a consequence, opposite
enantiomers can display very similar binding arrange-
ments resulting in the much poorer chiral discrimination
of 11.XB compared with rotaxane 1 (see Table 1).
In all binding arrangements of NBoc-(S/R)-proline
carboxylate anions with free macrocycle 11.XB, the ani-
ons are invariably held a considerable distance above the
plane of the macrocycle by highly linear XB interactions
(see Table S5-4, Supporting Information). As shown in
Figure 5B (left and middle), the most common binding
conformations adopted by both (R)- and (S)-anion enan-
tiomers along the 200 ns simulation time involves their
rigid pyrrolidine rings held close to the chiral (S)-BINOL
motif on the macrocycle, differing slightly in their rela-
tive spatial separation and orientations. Each anion en-
antiomer is bound via only one carboxylate oxygen at-
om, comprising a direct XB interaction to an iodotria-
Analogous MD simulations on the binding of NBoc-
(S/R)-proline carboxylate enantiomers to rotaxanes 2 and
3 also shed light on their chiral discrimination behavior.
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
With rotaxane 2, bearing only a chiral macrocycle subu-
Page 12 of 16
1
2
3
4
5
6
nit, MD runs revealed that the (R)-carboxylate is more
strongly bonded to the rotaxane binding cavity com-
pared to the (S)-enantiomer, accounting for the rotax-
ane’s observed (R)-selectivity. In the most common bind-
ing scenario shown in Figure 6A (generated from the R
B
initial binding arrangement), the anion is bound by three
synergetic direct XB interactions between its carboxylate
unit and the iodotriazoles in the rotaxane binding site
(see Table S5-2, Supporting Information). Even in the
presence of reduced water content (acetone/ water 99:1
v/v), these XB interactions are susceptible to bridging by
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
water molecules to form (C-I)host···(OH
works, as depicted in Fig. S5-7, Supporting Information.
On the other hand, simulations starting from the R , S
and S optimized structures (Fig S5-2, Supporting In-
2
)2···(O=C)anion net-
A
A
B
formation) result in the fully hydrated anions systemati-
cally leaving rotaxane 2’s binding pocket during the first
10 ns of sampling time in the five MD runs. Notably,
replacing the chiral (S)-serine derived axle of rotaxane 1
with the propyl spacer in rotaxane 2 prevents the for-
mation of (N–H)host···OH ···(O=C)anion bridges, which aid
2
in anchoring the anion within the rotaxane’s cavity
(shown in Figure 4A). This loss translates to dramatically
reduced anion binding strength.
Replacing the bulky chiral (S)-BINOL group on ro-
taxane 1’s macrocycle component with the achiral bi-
phenyl spacer in rotaxane 3 considerably enlarges the
conformational possibilities for anion binding within the
rotaxane cavity. Consequently, the obtained MD binding
scenarios (Fig. S5-8, Supporting Information) of the host-
guest diastereomeric complexes were found to be highly
dependent on the initial DFT-optimized binding ar-
rangements (Fig S5-3, Supporting Information). The
most frequent binding conformations with NBoc-(S)-
proline carboxylate involve only two XB interactions,
with some HB assistance between the axle component
and the NBoc group of the anion (Figure 6B, top). Con-
trastingly, the (R)-guest is typically held by three XB
interactions, comprising of two from the macrocycle
component and one from the axle (see Figure 6B bottom
and Table S5-2, Supporting Information). Furthermore,
these XB binding interactions are often assisted by wa-
ter-bridged hydrogen bonds similar to rotaxane 1, ac-
counting for the appreciable and comparable anion affin-
ities exhibited by rotaxane 3 (Table 1). Generally, the
more frequent XB interactions observed for the NBoc-
(R)-proline carboxylate likely accounts for its selectivity
over the (S)-enantiomer. Nonetheless, it is noteworthy
that the preferential carboxylate binding by the XB do-
nors in the central portion of the rotaxane directs the
Figure 6. (A) Illustrative binding scenario of rotaxane 2 with
NBoc-(R)-proline; (B) Expanded views of the most frequent
binding arrangements of NBoc-(S/R)-proline carboxylate
with rotaxane 3 (see Fig S5-8, Supporting Information for
full views of all binding scenarios). All stoppers were omit-
ted for clarity. XB and HB interactions are depicted as
dashed lines as described in Figure 4.
orientation of anion coordination in each binding pock-
et. Due to the geometric flexibility of anion binding in
rotaxane 3’s binding pocket, the proline’s chiral pyrroli-
dine ring can be either oriented towards the achiral bi-
phenyl spacer on the macrocycle component (Figure S5-
8, Supporting Information), or held above the axle’s
point chiral-unit by poorly-directional HB interactions
12
ACS Paragon Plus Environment

Page 13 of 16
Journal of the American Chemical Society
showing significant deviations from linearity, bridged nents led to each rotaxane having unique enantiomer
1
by water molecules in some cases. These factors, coupled
with the axle’s conformationally-flexible point-chiral
groups, allows opposite anion enantiomers to adopt re-
markably similar binding arrangements (Figure 6B).
Contrastingly for rotaxanes 1 and 2, the XB-directed ani-
on binding geometries forces the prolines’ asymmetric
units into close proximity with the macrocycle compo-
nents’ rigid and sterically-bulky chiral (S)-BINOL
groups (Figures 4 and 6A). This likely accounts for the
observed dominance of the chiral macrocycle in the ro-
taxane structure for effective chiral discrimination to
occur (Figure 3).
binding preferences, which could not be easily predicted
a priori. Through MD simulations, the chiral selectivities
of each rotaxane were shown to arise from a complex
and subtle interplay of attractive and repulsive non-
covalent interactions occurring between the chiral guest
anions and the rotaxanes’ chiral binding cavities in three
dimensions, akin to natural enzymes.74–76 These in silico
studies also revealed the unexpected influence of the
hydration shell of each chiral oxoanion towards their
binding affinities and chiral selectivities. Representing
the first detailed study of the use of chiral [2]rotaxanes
for anion enantio-discrimination, our present findings
provide important insights on how this emerging class
of molecules function as chiral selectors, which will
stimulate further exploitations in catalysis,30,77–80 drug
delivery81 and nanotechnological82–84 applications.
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Finally, our modeling experiments also yielded val-
uable insights on the general opposite enantiomer pref-
erence observed for BINOL-PO -
compared to the amino
4
acid anion derivatives. Using rotaxane 1, MD simula-
tions based on the DFT optimized structures of the best-
fit single binding arrangement of (S/R)-BINOL-PO -
in
4
the rotaxane binding cavity (see Figure S5-9, Supporting
Information) were carried out, which confirmed the sta-
bility of both host-guest diastereomeric complexes along
the five independent 20 ns MD runs. Each anion enanti-
omer is bound, via a single oxygen atom, by two con-
vergent XB interactions of comparable dimensions ex-
clusively from the rotaxane axle component’s iodotria-
zoles (Figure S5-10, Supporting Information). Moreover,
ASSOCIATED CONTENT
Supporting Information
Detailed synthetic procedures for all novel compounds,
characterization data, binding isotherms for chiral anions,
additional structures and details from molecular modeling
are provided. The Supporting Information is available free
of charge on the ACS Publications website.
the (R)-BINOL-PO -
is further stabilized by favorable
4
π···π stacking interactions with the helicoidal (S)-BINOL
unit on the macrocycle (shown in Figure S5-10 bottom,
Supporting Information). Contrastingly, the relative spa-
tial disposition of the aromatic rings of the (S)-BINOL-
AUTHOR INFORMATION
Corresponding Author
* paul.beer@chem.ox.ac.uk
PO -
tions (Figure S5-10 top, Supporting Information), ac-
counting for the observed preference for (R)-BINOL-PO -
4
guest does not favor the formation of these interac-
Author Contributions
4
The manuscript was written through contributions of all
authors. / All authors have given approval to the final ver-
sion of the manuscript.
.
CONCLUSIONS
ACKNOWLEDGMENT
In summary, we have synthesized a family of novel XB
J.Y.C.L thanks the Agency for Science, Technology and Re-
search (A*STAR), Singapore, for postgraduate funding. The
theoretical studies were supported by project P2020-
PTDC/QEQ-SUP/4283/2014 as well as by CICECO – Aveiro
Institute of Materials (UID/CTM/50011/2013) and iBiMED –
Institute of Biomedicine (UID/BIM/04501/2013), financed by
National Funds through the FCT/MEC and, when applica-
ble, co-financed by FEDER through COMPETE, under the
PT2020 Partnership Agreement. I.M. thanks the FCT for the
PhD scholarship SFRH/BD/87520/2012.
chiral [2]rotaxanes using an active metal template ap-
proach, and studied their enantioselective anion recogni-
tion properties. The presence of a chiral macrocycle
component in the rotaxane host design was found to be
essential for enantioselective anion binding due to the
chiral anion guest orientations dictated by XB interac-
tions, allowing generally superior chiral discrimination
properties compared to the free axle and macrocycle
components. Contrastingly, the presence of a chiral axle
exerted relatively minor influence on the rotaxanes’ chi-
ral selectivity as a whole. Notably, the various combina-
tions of chiral and achiral macrocycle and axle compo-
REFERENCES
(1)
Jo, H. H.; Lin, C.-Y.; Anslyn, E. V. Acc. Chem. Res. 2014,
47 (7), 2212–2221.
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 16
(2)
(3)
Finn, M. G. Chirality 2002, 14 (7), 534–540.
Hembury, G. A.; Borovkov, V. V.; Inoue, Y. Chem. Rev.
(29) J. Cantrill, S.; C. T. Fyfe, M.; M. Heiss, A.; Fraser
1
2
Stoddart, J.; J. P. White, A.; J. Williams, D. Chem Commun 1999,
No. 13, 1251–1252.
2008, 108 (1), 1–73.
3
4
(4)
(5)
Izake, E. L. J. Pharm. Sci. 2007, 96 (7), 1659–1676.
Phipps, R. J.; Hamilton, G. L.; Toste, F. D. Nat Chem
(30) Cakmak, Y.; Erbas-Cakmak, S.; Leigh, D. A. J. Am.
Chem. Soc. 2016, 138 (6), 1749–1751.
5
6
2012, 4 (8), 603–614.
(31) Mitchell, D. K.; Sauvage, J.-P. Angew. Chem. Int. Ed.
Engl. 1988, 27 (7), 930–931.
(6)
Brak, K.; Jacobsen, E. N. Angew. Chem. Int. Ed. 2013, 52
7
8
9
(2), 534–561.
(32) Perret-Aebi, L.-E.; von Zelewsky, A.; Dietrich-
Buchecker, C.; Sauvage, J.-P. Angew. Chem. 2004, 116 (34), 4582–
4585.
(7)
518–533.
(8)
Mahlau, M.; List, B. Angew. Chem. Int. Ed. 2013, 52 (2),
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bauer, E. B. Chem Soc Rev 2012, 41 (8), 3153–3167.
(33) Bordoli, R. J.; Goldup, S. M. J. Am. Chem. Soc. 2014, 136
(13), 4817–4820.
(9)
Christmann, M.; Brase, S. Asymmetric Synthesis; Wiley-
VCH: Weinheim: Germany, 2008.
(34) Mitra, R.; Thiele, M.; Octa-Smolin, F.; Letzel, M. C.;
Niemeyer, J. Chem Commun 2016, 52 (35), 5977–5980.
(35) Glen, P. E.; O’Neill, J. A. T.; Lee, A.-L. Tetrahedron
2013, 69 (1), 57–68.
(10) Mezzetti, A.; Schrag, J. D.; Cheong, C. S.; Kazlauskas,
R. J. Chem. Biol. 2005, 12 (4), 427–437.
(11) Ema, T.; Okuda, K.; Watanabe, S.; Yamasaki, T.;
Minami, T.; Esipenko, N. A.; Anzenbacher, P. Org. Lett. 2014, 16
(5), 1302–1305.
(36) Hiratani, K.; Kaneyama, M.; Nagawa, Y.; Koyama, E.;
Kanesato, M. J. Am. Chem. Soc. 2004, 126 (42), 13568–13569.
(37) Cavallo, G.; Metrangolo, P.; Milani, R.; Pilati, T.;
Priimagi, A.; Resnati, G.; Terraneo, G. Chem. Rev. 2016, 116 (4),
2478–2601.
(12) Rossi, S.; Kyne, G. M.; Turner, D. L.; Wells, N. J.; Kil-
burn, J. D. Angew. Chem. Int. Ed. 2002, 41 (22), 4233–4236.
(13) Sessler, J. L.; Andrievsky, A.; Král, V.; Lynch, V. J. Am.
Chem. Soc. 1997, 119 (40), 9385–9392.
(38) Beale, T. M.; Chudzinski, M. G.; Sarwar, M. G.; Taylor,
M. S. Chem Soc Rev 2013, 42 (4), 1667–1680.
(14) Li, Z.-B.; Lin, J.; Pu, L. Angew. Chem. 2005, 117 (11),
1718–1721.
(39) Gilday, L. C.; Robinson, S. W.; Barendt, T. A.; Langton,
M. J.; Mullaney, B. R.; Beer, P. D. Chem. Rev. 2015, 115 (15),
7118–7195.
(15) Sansone, F.; Baldini, L.; Casnati, A.; Lazzarotto, M.;
Ugozzoli, F.; Ungaro, R. Proc. Natl. Acad. Sci. 2002, 99 (8), 4842–
4847.
(40) Kaasik, M.; Kaabel, S.; Kriis, K.; Järving, I.; Aav, R.;
Rissanen, K.; Kanger, T. Chem. – Eur. J. 2017, 23 (30), 7337–7344.
(41) Peluso, P.; Mamane, V.; Aubert, E.; Dessì, A.; Dal-
locchio, R.; Dore, A.; Pale, P.; Cossu, S. Enantioseparations 2016,
1467, 228–238.
(16) Miyaji, H.; Hong, S.-J.; Jeong, S.-D.; Yoon, D.-W.; Na,
H.-K.; Hong, J.; Ham, S.; Sessler, J. L.; Lee, C.-H. Angew. Chem.
Int. Ed. 2007, 46 (14), 2508–2511.
(17) Langton, M. J.; Beer, P. D. Acc. Chem. Res. 2014, 47 (7),
1935–1949.
(42) Lim, J. Y. C.; Marques, I.; Ferreira, L.; Felix, V.; Beer, P.
D. Chem Commun 2016, 52 (32), 5527–5530.
(18) Knighton, R. C.; Beer, P. D. Chem Commun 2014, 50
(13), 1540–1542.
(43) Borissov, A.; Lim, J. Y. C.; Brown, A.; Christensen, K.
E.; Thompson, A. L.; Smith, M. D.; Beer, P. D. Chem Commun
2017, 53 (16), 2483–2486.
(19) Romero, J. R.; Aragay, G.; Ballester, P. Chem Sci 2017, 8
(1), 491–498.
(20) Caballero, A.; Zapata, F.; Beer, P. D. Coord. Chem. Rev.
2013, 257 (17–18), 2434–2455.
(44) Kubik and co-workers have very recently reported the
synthesis of a chiral XB pseudopeptide based macrocycle which
binds halides in 2.5% water/DMSO: Mungalpara, D.; Stegmul-
ler, S.; Kubik, S. Chem Commun 2017, 53 (37), 5095–5098.
(45) Pu, L. Acc. Chem. Res. 2012, 45 (2), 150–163.
(46) The macrocycle and axle components of rotaxanes 1-3
are designed to be symmetric (i.e. identical groups on both
sides of a mirror plane passing through the center of each com-
ponent) to prevent the rotaxanes from exhibiting additional
mechanically bonded chirality.
(21) Brown, A.; Beer, P. D. Chem Commun 2016, 52 (56),
8645–8658.
(22) Lim, J. Y. C.; Marques, I.; Thompson, A. L.; Christen-
sen, K. E.; Félix, V.; Beer, P. D. J. Am. Chem. Soc. 2017, 139 (8),
3122–3133.
(23) Neal, E. A.; Goldup, S. M. Chem Commun 2014, 50 (40),
5128–5142.
(24) Ogoshi, T.; Yamafuji, D.; Aoki, T.; Kitajima, K.;
Yamagishi, T.; Hayashi, Y.; Kawauchi, S. Chem. – Eur. J. 2012, 18
(24), 7493–7500.
(47) Wisner, J. A.; Beer, P. D.; Drew, M. G. B. Angew. Chem.
Int. Ed. 2001, 40 (19), 3606–3609.
(25) Shinoda, S.; Maeda, T.; Miyake, H.; Tsukube, H.
Supramol. Chem. 2011, 23 (3–4), 244–248.
(48) Wisner, J. A.; Beer, P. D.; Drew, M. G. B.; Sambrook,
M. R. J. Am. Chem. Soc. 2002, 124 (42), 12469–12476.
(49) White, N. G.; Serpell, C. J.; Beer, P. D. Cryst. Growth
Des. 2014, 14 (7), 3472–3479.
(26) Mohry, A.; Vögtle, F.; Nieger, M.; Hupfer, H. Chirality
2000, 12 (2), 76–83.
(27) Reuter, C.; Mohry, A.; Sobanski, A.; Vögtle, F. Chem. –
Eur. J. 2000, 6 (9), 1674–1682.
(50) Despite the symmetry of the axle component, only one
guest anion is expected to be predominantly bound in the ro-
taxane cavity due to the spatial constraints of the binding site
and size of respective chiral anion guest species.
(51) Aucagne, V.; Hänni, K. D.; Leigh, D. A.; Lusby, P. J.;
Walker, D. B. J. Am. Chem. Soc. 2006, 128 (7), 2186–2187.
(28) Schmidt, T.; Schmieder, R.; Müller, W. M.; Kiupel, B.;
Vögtle, F. Eur. J. Org. Chem. 1998, 1998 (9), 2003–2007.
14
ACS Paragon Plus Environment

Page 15 of 16
Journal of the American Chemical Society
(52) Aucagne, V.; Berná, J.; Crowley, J. D.; Goldup, S. M.;
Monard, G.; Nguyen, H.; Nguyen, H. T.; Omelyan, I.; Onufriev,
1
2
3
4
5
6
7
8
Hänni, K. D.; Leigh, D. A.; Lusby, P. J.; Ronaldson, V. E.;
Slawin, A. M. Z.; Viterisi, A.; Walker, D. B. J. Am. Chem. Soc.
2007, 129 (39), 11950–11963.
A.; Roe, D. R.; Roitberg, A.; Sagui, C.; Simmerling, C. L.; Swails,
J.; Walker, R. C.; Wang, J.; Wolf, R. M.; Wu, X.; Xiao, L.; York,
D. M.; Kollman, P. A. AMBER 2016; University of California:
San Francisco, 2016.
(53) Langton, M. J.; Xiong, Y.; Beer, P. D. Chem. – Eur. J.
2015, 21 (52), 18910–18914.
(69) Salomon-Ferrer, R.; Götz, A. W.; Poole, D.; Le Grand,
S.; Walker, R. C. J. Chem. Theory Comput. 2013, 9 (9), 3878–3888.
(70) Le Grand, S.; Götz, A. W.; Walker, R. C. Comput. Phys.
Commun. 2013, 184 (2), 374–380.
(54) Lim, J. Y. C.; Bunchuay, T.; Beer, P. D. Chem. – Eur. J.
2017, 23 (19), 4700–4707.
(55) Brotherton, W. S.; Clark, R. J.; Zhu, L. J. Org. Chem.
2012, 77 (15), 6443–6455.
9
(71) Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.;
Case, D. A. J Comput Chem 2004, 25 (9), 1157–1174.
(72) Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.;
Case, D. A. J. Comput. Chem. 2005, 26 (1), 114–114.
(73) Ibrahim, M. A. J Comput Chem 2011, 32 (12), 2564–2574.
(74) Lang, D. A.; Mannesse, M. L. M.; De Haas, G. H.; Ver-
heij, H. M.; Dijkstra, B. W. Eur. J. Biochem. 1998, 254 (2), 333–340.
(75) Morozov, A. N.; D’Cunha, C.; Alvarez, C. A.; Chat-
field, D. C. Biophys. J. 100 (4), 1066–1075.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(56) Zhou, E.; Zhang, J.; Dong, C. ARKIVOC 2014, v, 351–
364.
(57) Parmar, D.; Sugiono, E.; Raja, S.; Rueping, M. Chem.
Rev. 2014, 114 (18), 9047–9153.
(58) Hynes, M. J. J. Chem. Soc. Dalton Trans. 1993, No. 2,
311–312.
(59) Kim, S.-G.; Kim, K.-H.; Jung, J.; Shin, S. K.; Ahn, K. H.
J. Am. Chem. Soc. 2002, 124 (4), 591–596.
(60) Zhang, X. X.; Bradshaw, J. S.; Izatt, R. M. Chem. Rev.
1997, 97 (8), 3313–3362.
(76) Müller, T. A.; Zavodszky, M. I.; Feig, M.; Kuhn, L. A.;
Hausinger, R. P. Protein Sci. 2006, 15 (6), 1356–1368.
(77) Hoekman, S.; Kitching, M. O.; Leigh, D. A.; Papmeyer,
M.; Roke, D. J. Am. Chem. Soc. 2015, 137 (24), 7656–7659.
(78) Blanco, V.; Leigh, D. A.; Marcos, V.; Morales-Serna, J.
A.; Nussbaumer, A. L. J. Am. Chem. Soc. 2014, 136 (13), 4905–
4908.
(61) Interestingly, Sánchez and co-workers reported the
preponderance of axial chirality over point chirality in the for-
mation of homochiral supramolecular aggregates in Buendía, J.;
Greciano, E. E.; Sánchez L. J. Org. Chem., 2015, 80 (24), 12444–
12452.
(62) Ufimtsev, I. S.; Martinez, T. J. J. Chem. Theory Comput.
2009, 5 (10), 2619–2628.
(79) Tachibana, Y.; Kihara, N.; Takata, T. J. Am. Chem. Soc.
2004, 126 (11), 3438–3439.
(63) Titov, A. V.; Ufimtsev, I. S.; Luehr, N.; Martinez, T. J. J.
Chem. Theory Comput. 2013, 9 (1), 213–221.
(80) Galli, M.; Lewis, J. E. M.; Goldup, S. M. Angew. Chem.
Int. Ed. 2015, 54 (46), 13545–13549.
(64) Song, C.; Wang, L.-P.; Martínez, T. J. J. Chem. Theory
Comput. 2016, 12 (1), 92–106.
(81) Barat, R.; Legigan, T.; Tranoy-Opalinski, I.; Renoux, B.;
Peraudeau, E.; Clarhaut, J.; Poinot, P.; Fernandes, A. E.;
Aucagne, V.; Leigh, D. A.; Papot, S. Chem Sci 2015, 6 (4), 2608–
2613.
(65) Kästner, J.; Carr, J. M.; Keal, T. W.; Thiel, W.; Wander,
A.; Sherwood, P. J. Phys. Chem. A 2009, 113 (43), 11856–11865.
(66) Goumans, T. P. M.; Catlow, C. R. A.; Brown, W. A.;
Kästner, J.; Sherwood, P. Phys. Chem. Chem. Phys. 2009, 11 (26),
5431.
(82) Carlone, A.; Goldup, S. M.; Lebrasseur, N.; Leigh, D.
A.; Wilson, A. J. Am. Chem. Soc. 2012, 134 (20), 8321–8323.
(83) Lewandowski, B.; De Bo, G.; Ward, J. W.; Papmeyer,
M.; Kuschel, S.; Aldegunde, M. J.; Gramlich, P. M. E.; Heck-
mann, D.; Goldup, S. M.; D’Souza, D. M.; Fernandes, A. E.;
Leigh, D. A. Science 2013, 339 (6116), 189.
(67) Tomov, S.; Dongarra, J.; Baboulin, M. Parallel Comput.
2010, 36 (5–6), 232–240.
(68) Case, D. A.; Betz, R. M.; Botello-Smith, W.; Cerutti, D.
S.; T.E. Cheatham, I.; Darden, T. A.; Duke, R. E.; Giese, T. J.;
Gohlke, H.; Goetz, A. W.; Homeyer, N.; Izadi, S.; Janowski, P.;
Kaus, J.; Kovalenko, A.; Lee, T. S.; LeGrand, S.; Li, P.; Lin, C.;
Luchko, T.; Luo, R.; Madej, B.; Mermelstein, D.; Merz, K. M.;
(84) Yang, W.; Li, Y.; Zhang, J.; Chen, N.; Chen, S.; Liu, H.;
Li, Y. Small 2012, 8 (16), 2602–2607.
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 16
1
2
TOC Graphic
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
16
ACS Paragon Plus Environment