Simple acyclic molecules containing a single charge-assisted O-H group can recognize anions in acetonitrile?:?water mixtures

Article abstract of DOI:10.1039/d1ob00282a

Hydroxypyridinium and hydroxyquinolinium compounds containing acidic O-H groups attached to a cationic aromatic scaffold were synthesized,i.e. N-methyl-3-hydroxypyridinium (1⁺) andN-methyl-8-hydroxyquinolinium (2⁺). These very simple

Full text of DOI:10.1039/d1ob00282a

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Simple acyclic molecules containing a single
charge-assisted O–H group can recognize anions
in acetonitrile : water mixtures†
Cite this: Org. Biomol. Chem., 2021,
9, 2794
1
Rosemary J. Goodwin,
Mitchell T. Blyth,
Alfred K. K. Fung, Leesa M. Smith,
Philip L. Norcott, Sara Tanovic,
Michelle L. Coote * and Nicholas G. White
*
Hydroxypyridinium and hydroxyquinolinium compounds containing acidic O–H groups attached to a cat-
+
ionic aromatic scaffold were synthesized, i.e. N-methyl-3-hydroxypyridinium (1 ) and N-methyl-8-hydro-
+
xyquinolinium (2 ). These very simple compounds are capable of binding to chloride very strongly in
CD
3
CN and with moderate strength in 9 : 1 CD
3
CN : D
2
O. Comparison with known association constants
+
+
reveals that 1 and 2 bind chloride in CD
3
CN or CD
3
CN : D O with comparable affinities to receptors
2
containing significantly more hydrogen bond donors and/or higher positive charges. Crystal structures of
both compounds with coordinating anions were obtained, and feature short O–H⋯anion hydrogen
bonds. A receptor containing two hydroxyquinolinium groups was also prepared. While the low solubility
of this compound caused difficulties, we were able to demonstrate chloride binding in a competitive 1 : 1
Received 15th February 2021,
Accepted 8th March 2021
DOI: 10.1039/d1ob00282a
3 3
CD CN : CD OD solvent mixture. Addition of sulfate to this compound results in the formation of a crys-
rsc.li/obc
tallographically-characterised solid state anion coordination polymer.
gnition, while using O–H hydrogen bonds in conjunction with
a Lewis acidic boron centre can give selective fluoride binding
Introduction
1
3
The development of anion recognition and its application in a in THF : water mixtures. As the field of anion supramolecular
range of fields has been an ongoing research target within chemistry has expanded beyond simple anion recognition to
1
,2
1
4,15
supramolecular chemistry.
bound anions through N–H⋯anion interaction, although
Many of the initial receptors
applications,
in a range of applications including self-assembly,
transport,
so O–H⋯anion interactions have been used
16,17
anion
3
1
8–20
21–23
more recently C–H⋯anion hydrogen bonding, as well as
and organocatalysis.
4
5
halogen bonding and other sigma-hole interactions have
been developed as recognition motifs. While O–H⋯anion
interactions have been studied intermittently for several
decades, they have received relatively little attention, particu-
larly given their potential ease of synthesis and strong hydro-
It would appear that the development of strong O–
H⋯anion binding motifs would facilitate further applications
in these fields. Given that even relatively simple neutral O–H
containing molecules can be surprisingly potent anion
8,9,24,25
receptors,
it is likely that cationic hosts containing
6
gen bond character.
strong O–H hydrogen bond donors may be able to achieve
high anion affinities in competitive solvent media, although
very few such receptors have been prepared. Alfonso has
reported di- and tricationic O–H containing receptors that can
Early studies revealed that simple O–H-containing mole-
cules such as sugars, catechol and alizarin can bind anions in
polar organic solvents.
strated that incorporating O–H hydrogen bond donors into
more complex scaffolds such as foldamers
organic frameworks
7
–9
More recently, it has been demon-
26
function in acetonitrile/methanol mixtures, while we have
1
0,11
or metal
recently reported simple cationic acyclic receptors containing
amide and phenol hydrogen bond donors that can function in
12
can achieve impressive guest reco-
2
7
9
: 1 acetonitrile : water. In the current work, we describe the
anion recognition properties of a family of simple cationic
molecules that contain very strong O–H hydrogen bond donors
Research School of Chemistry, Australian National University, Canberra, ACT,
Australia. E-mail: michelle.coote@anu.edu.au, nicholas.white@anu.edu.au
http://rsc.anu.edu.au/~mcoote/, http://www.nwhitegroup.com
(
Fig. 1). These molecules can be prepared in one step from
†
Electronic supplementary information (ESI) available: Additional synthetic
commercially-available and cheap 3-hydroxypridine or 8-hydro-
xyquinoline. While they are prone to deprotonation by basic
anions, we show that they are able to bind anions in
acetonitrile : water mixtures.
details, characterization data, details of anion binding, X-ray crystallography, pK
a
measurements and computational studies. CCDC 2050536–2050549. For ESI and
crystallographic data in CIF or other electronic format see DOI: 10.1039/
d1ob00282a
2794 | Org. Biomol. Chem., 2021, 19, 2794–2803
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
reaction resulting from O-alkylation reactions: one of these,
4
·Br could be isolated from the reaction mixture and was
characterized crystallographically and by NMR spectroscopy
Scheme 1 and ESI†). A “tidier” synthesis of 3·Br may well be
(
2
possible starting from 8-methoxyquinoline and bis(bromo-
methyl)benzene followed by demethylation, however given that
the starting materials used in our synthesis are cheap, and
purification relatively simple, this was not investigated.
Anion exchange. We exchanged the potentially coordinating
2
halide anions in 1·I, 2·I and 3·Br for non-coordinating hexa-
fluorophosphate or tetraphenylborate anions by precipitation
Fig. 1 Cationic molecules containing strong O–H hydrogen bond
donors used in this work.
in good yields (86–93%). We prepared the hexafluoro-
2
+
phosphate salt of 3 and found that it was best to add a few
drops of 1% HPF6(aq) to avoid the products having an orange
colour, which we attribute to traces of deprotonated species
Results
Synthesis
(
see later).
In the case of 1 , we could not isolate the PF salt as this
+
−
6
+
+
2+
+
+
+
Synthesis of 1 , 2 and 3 . Compounds 1 and 2 were pre-
pared as their iodide salts by alkylation of commercially-avail-
able 3-hydroxypyridine or 8-hydroxyquinoline using methyl
did not precipitate from water. In the case of 2 , it was possible
to isolate a crystalline solid when a concentrated aqueous solu-
tion of hot 2·I was mixed with aqueous NH PF and allowed to
4
6
2
8,29
+
iodide.
Alkylation of 1 proceeds quantitatively, while a
cool. SCXRD studies of one of these crystals showed that it was
+
1
13
side-product is produced during the alkylation of 2 . However,
this is not troublesome as 2·I precipitates from the acetonitrile
reaction solution (see ESI† for an optimized preparation of 2·I
on a multi-gram scale).
2·PF
6
, and H and C NMR analyses of the bulk material were
1
9
30
satisfactory. However, quantitative F NMR spectroscopy
revealed a lower amount of PF
−
6
than expected (approximately
one half) suggesting that anion exchange was incomplete and
the material was a mixture of the I− and PF6− salt. Mixing a
2
+
The bis-hydroxyquinolinium compound 3 was synthesized
from commercially available 8-hydroxyquinoline and 1,2-bis
4
solution of this material in acetone with AgBF in acetone
(
bromomethyl)benzene (Scheme 1). The reaction of ten equiva-
resulted in the formation of cloudy precipitate (presumably
AgI), supporting this hypothesis.
lents of hydroxyquinoline with bis(bromomethyl)benzene in
refluxing acetonitrile gave a yellow precipitate in approximately
5
−
+
Given that we could not satisfactorily isolate PF
6
salts of 1
1
+
5% yield, which H NMR spectroscopy showed was 3·Br of
2
or 2 , these compounds were instead isolated as the tetra-
∼
90% purity (see ESI† for more details). Purification by
column chromatography eluting with a basic DCM/MeOH/
NEt mixture and then precipitating using HBr gave pure
phenylborate salts by precipitation from methanol. Integration
of the H NMR spectrum confirmed that complete anion
1
3
(aq)
exchange had been achieved in these cases, and integration
1
9
30
3
2
·Br . It appears that several side-products form during the
of the quantitative F NMR spectrum of 3·(PF ) indicated
6
2
complete conversion to the hexafluorophosphate salt.
+
2+
Deprotonation. We initially attempted to convert 2 and 3
to their PF6− salts using a relatively old bottle of NH PF . This
4
6
resulted in deep orange-coloured solutions or precipitates; an
intense orange-red colour is known to result from deproto-
nated hydroxyquinolinium groups, which display significant
2
8
4 6
solvatochromic effects. Analysis of this bottle of NH PF
1
9
using F NMR spectroscopy revealed that it contained a sig-
−
nificant amount of F , which presumably acted as a base to
partially deprotonate the compounds resulting in the deeply-
coloured zwitterionic form. Adding small amounts of triethyl-
amine to solutions of the compounds produced similar
−
H
colours, and we were able to obtain single crystals of 3 ·PF
6
suitable for X-ray diffraction studies in this way. X-ray crystallo-
−
graphy reveals a very short O–H⋯O distance, with a distance
3
1
of 2.414(2) Å between the two oxygen atoms. This hydrogen
bond links the structure into a one-dimensional hydrogen
bonded chain (Fig. 2).
To investigate this further, we attempted to measure the
+
pK
a
of 3·Br
2
in water to allow comparison with that of 2 ,
Scheme 1 Synthesis of 3·Br
crystal structure of cation 4 .
2
, structure of side product 4·Br, and X-ray
+
33
which is 6.8. Unfortunately precipitation meant we were
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 2794–2803 | 2795

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 2 X-ray crystal structure of 3− ·PF
H
(PF
−
anions and some C–H
6
6
hydrogen atoms are omitted for clarity). Note: PLATON-SQUEEZE was
32
used (see ESI† for details).
1
Fig. 3 Partial H NMR spectra of 1·BPh
4
upon addition of TBA·Cl
(
CD CN, 600 MHz, 298 K). The 11 spectra represent 0–2.0 equivalents
3
−
of Cl at intervals of 0.2 equivalents.
4
too high to be determined by NMR titration experiments (>10
M
−
1
). Addition of sulfate resulted in precipitation.
Given the very strong binding in CD CN, we next studied
3
binding in the highly competitive mixed organic/aqueous
system 9 : 1 CD CN : D O. Interestingly, chloride caused small
3
2
downfield shifts in C–H protons adjacent to the O–H group,
while sulfate caused upfield shifts. We do not know the reason
for this upfield peak movement, although it does not appear
Scheme 2 Synthesis of 3− ·PF
H
.
6
3
7
to be due to deprotonation. In this competitive solvent
mixture, binding was significantly weaker than in pure aceto-
+
2
+
nitrile, but moderate binding was still observed by 1 , with
a
unable to do so quantitatively, but qualitatively, the pK of 3
appears to be broadly similar to that of 2 , suggesting the
acidity of each O–H group is relatively independent of the
other. Due to the low solubility of 3 ·PF in polar solvents,
6
it was possible to isolate it as a crystalline solid by adding one
+
+
very weak binding observed by 2 (Table 1). The stronger
+
+
binding observed for 1 compared to 2 is not unexpected as
+
+
3
4
−H
1 is known to have a more acidic O–H group than 2 (the pK
a
3
8
33
values in water are 5.0 and 6.8, respectively).
X-ray crystal structures. We were able to obtain several
equivalent of triethylamine to a methanol solution of 3·(PF )
6
1
2
+
+
crystal structures of 1 and 2 with various anions: the struc-
tures of 1·I, 2·Cl and 2 ·SO are presented in Fig. 4, while
additional structures (1·BPh , 2·PF ) are discussed in the ESI.†
(
Scheme 2). NMR spectroscopy, including integration of the H
1
9
31
2
4
NMR and quantitative F NMR spectrum against a standard
confirmed that 3−H·PF was isolated selectively.
4
6
6
As shown, H⋯anion distances are short ranging from as little
+
+
40
Anion binding of 1 and 2
as 64% of the sum of the van der Waals radii (ΣvdW) in the
structure of 2
2
·SO
4
to 71% of ΣvdW in 2·Cl, to 76–80% of ΣvdW
NMR titrations. A preliminary qualitative screen of anion
binding was conducted by adding one equivalent of anion to
+
1
.0 mM solutions 2 in either CH
3
CN or 9 : 1 CH
3
CN : H
2
O.
−
−
Table 1 Association constants (M−1) for 1·BPh
a
This revealed that H PO caused precipitation, while OAc
4
and 2·BPh
4
with anions
2
4
caused deprotonation in both solvent systems and so these
2−
Receptor, solvent
Cl−
SO
4
anions were not studied quantitatively (Fig. S25 and S26†).
−
2−
4
4
Instead we focused on Cl and SO4
and studied guest 1·BPh , CD CN
>10
>10
Precipitation
Precipitation
4
4
3
3
1
2·BPh
, CD
CN
binding using quantitative H NMR titration experiments.
Addition of chloride anion to either 1 or 2 in CD CN resulted
3
+
+
4 3
1·BPh , 9 : 1 CD CN : D
2
O
2
65(3)
13(1)
77(8)
Negligible
in significant downfield shifts of hydrogen atoms adjacent to 2·BPh , 9 : 1 CD CN : D O
4
3
3
5
the O–H group (Fig. 3 and Fig. S29–S34†). Fitting the peak
a
Anions added as TBA salts, 1 : 1 binding constants calculated using
Bindfit, the asymptotic error is provided at the 95% confidence
3
6
movement to a 1 : 1 binding isotherm using Bindfit, revealed
that association of chloride with both of these receptors was interval in parentheses.
36
39
2796 | Org. Biomol. Chem., 2021, 19, 2794–2803
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
Interestingly, when there are no coordinating anions available
in the structure of 3·(PF ) the receptor adopts an anti con-
6
2
figuration where the two O–H groups point in opposite direc-
4
0
tions and form short hydrogen bonds (66, 71% ΣvdW ) to
acetonitrile solvent molecules. Another structure of 3·(PF6)2
was obtained from a different solvent mixture and has a
similar structure (see ESI†). In the structure of 3·Br
configuration of the cation is also observed, presumably to
facilitate hydrogen bonding to each anion. When 3·(PF was
2
, an anti
6 2
)
−
crystallised in the presence of one equivalent of either Cl or
−
Br , crystals of 3·Cl·PF and 3·Br·PF6 were obtained (Fig. 5c
6
and d). In these structures, the receptor adopts a syn configur-
ation with both O–H groups forming relatively short hydrogen
bonds to the halide anion. Hydrogen bonds are slightly
shorter as ΣvdW with chloride than with bromide. Very small
crystals of 3·SO
4
were obtained, which required the use of syn-
4
3
chrotron radiation. In this case, the molecule adopts an anti
2
+
configuration and short hydrogen bonds between 3 and
sulfate (64, 65% ΣvdW) result in the formation of a one-dimen-
sional hydrogen bonded polymer assembled by anion
4
4–46
coordination.
Two methanol solvent molecules also hydro-
gen bond to the sulfate anion, although these hydrogen bonds
are longer (70, 74% ΣvdW).
Fig. 4 X-ray crystal structures of (a) 1·I, (b) 2·Cl, and (c) 2
2
·SO
A range
is given for 1·I as this structure has Z’ = 5 so there are several different
4
.
4
0
H⋯acceptor distances for hydrogen bonds are given as ΣvdW
.
Colourimetric “anion” sensing by 2 , 3 and 3−H+
+
2+
−
2 4
H⋯I distances. Positional disorder of the anion in 2 ·SO is omitted for
As noted above, pale yellow solutions of 2·BPh take on a
4
clarity.
deep orange-red colour in the presence of OAc− (Fig. S25
+
and S26†), which we attribute to deprotonation of 2 to
form the zwitterionic form, which is a known chromo-
phore. A similar phenomenon is observed when OAc is
in 2·I. In the structures of 1·BPh and 2·PF , no O–H⋯anion
interactions shorter than ΣvdW are observed, consistent with
the non-coordinating nature of these anions. We note that the
28
−
4
6
2+
added to 3
3
(
(Fig. S27†). Additionally, red solutions of
−H
−
·PF6 partially decolourise upon addition of HSO4
Fig. S28†), which we attribute to partial protonation of 3
by the acidic HSO
7
1% of ΣvdW in 2·Cl is similar to the 72% Σ
observed in the
−H+
vdW
4
1,42
previously-reported structure of 1·Cl.
−
47
4
a
anion (pK = 2.0). While these col-
ourimetric responses suggest an anion sensing response
and similar responses are sometimes reported in the litera-
NMR titrations. 3·(PF ) suffers from poor solubility, and is ture as such ), this is not really an accurate description of
2
+
Anion binding of 3
(
6
6 2
not appreciably soluble in a range of polar organic solvents, what is occurring, and the compounds are instead simply
including acetonitrile and acetonitrile/water mixtures. We were responding to base/acid.
therefore not able to conduct binding studies in the same
+
+
+
a
pK measurements of 2
solvent mixture used for 1 and 2 . Instead, anion recognition
1
was studied using H NMR titration experiments in competi- We investigated the effect of solvent and anion on the pK
a
+
+
tive 1 : 1 CD
3 3
CN : CD OD. As was the case for 1 and 2 , a pre- values of these highly acidic O–H groups. Unfortunately, the
−
2+
liminary screen (Fig. S23†) revealed that H PO caused pre- low solubility of 3 prohibited the determination of quantitat-
2
4
−
+
cipitation, while OAc caused deprotonation and so these ive pK
anions were not studied quantitatively. Solvent effects. We conducted measurements in 9 : 1
Addition of chloride to 3 resulted in small but signifi- MeOH : H O and DMF to complement the known value in
a
measurements and so we focused attention on 2 .
2
+
2
3
3
cant downfield shifts in the C–H proton resonances close to
2
H O. Unsurprisingly, moving towards less polar solvents
+
the hydroxy groups and the peak movement was fitted to a increases the pK
a
of 2 , with this value increasing from 6.8 in
: 1 binding isotherm using Bindfit. This revealed that water, to 7.5 in MeOH : H O and 9.5 in DMF (Table 2).
3
6
1
2
chloride bound with moderate strength in this polar solvent
Anion effects. We were interested to see whether the anion
−
1
2+
+
mixture [K
cipitation and so we were unable to calculate an association this we prepared 2·Cl in good yield by precipitation from
constant. acetone (see Experimental section). We reasoned that a favour-
X-ray crystal structures. We were able to obtain several able O–H⋯Cl hydrogen bond might favour the protonated
a a
= 76(2) M ]. Addition of sulfate to 3 caused pre- present affected the pK of the O–H group in 2 . To investigate
−
2
+
+
crystal structures of 3 , which are shown in Fig. 5. form of 2 and thus increase the pKa of 2·Cl relative to
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 2794–2803 | 2797

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 5 X-ray crystal structures of (a) 3·(PF
6
)
2
, (b) 3·Br
2
, (c) 3·Cl·PF
6
, (d) 3·Br·PF
6
and (e) 3·SO
4
. H⋯acceptor distances for hydrogen bonds are given as
40
−
Σ
vdW
.
The structure of 3·(PF
6
)
2
has Z’ = 2, only part of the unit cell is shown for clarity. Most solvent molecules and the PF
6
anions are omitted for clarity.
Table 2 pK
estimated 95% confidence interval of the mean
a 4
a
values of 2·Cl and 2·BPh in different solvents, with the able to obtain evidence of a difference in pK s between 2·Cl
a
4
and 2·BPh in DMF, which is perhaps somewhat surprising.
More details of these experiments and the rationale behind
them are provided in the ESI.†
Solvent
pK
a
for 2·Cl
a 4
pK for 2·BPh
3
3
H
2
O
6.8
7.4 ± 0.1
9.5 ± 0.2
—
7.5 ± 0.1
9.5 ± 0.1
b
9
: 1 MeOH : H
2
O
c
DMF
Computational studies and discussion
a
±2
standard errors (SE) of the mean, which were estimated as
pffiffiffi
b
SE ¼ σ= n. Determined by potentiometric titrations using NaOH as
+
+
c
Compounds 1 and 2 each appear to have only one significant
hydrogen bond donor, i.e. the O–H group. In the crystal struc-
base. Determined by potentiometric titrations using DBU as base.
tures, these groups are the only ones that form short hydrogen
which cannot form a strong hydrogen bond, and so bonds, although contacts slightly shorter than the sum of the
4
8
2
·BPh
4
,
a
measured the pK s of these compounds in both 9 : 1 vdW radii are also observed on some occasions between the
4
9
MeOH : H O and DMF. Unsurprisingly, in 9 : 1 MeOH : H O, no anion and the C–H group adjacent to the O–H group. To
2
2
difference in pK
a
is observed between 2·Cl and 2·BPh
4
, which investigate the energetic contributions of these possible inter-
+
we attribute to negligible hydrogen bonding between 2 and actions, we conducted density functional theory calculations at
−
50
Cl in this very competitive solvent mixture. We were also not the M06-2X/6-311+G(d,p) level of theory. We minimised the
2798 | Org. Biomol. Chem., 2021, 19, 2794–2803
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
+
+
structures of 1·Cl and 2·Cl using an SMD continuum solvent
1 and 2 also bind chloride more strongly than similarly
5
1
model to account for solvent effects, and then used Natural acidic molecules containing more O–H donors: for example,
Bond Orbital (NBO) analysis to calculate the relative contri- 4-nitrocatechol which has two highly acidic O–H groups (pK
butions of the possible interactions (see the ESI† for full 6.7,
a
=
5
4
−1
−
9
K
a
= 4400 M for Cl in CD
While the electrostatic attraction for anions in 1 and 2 is
3
CN).
+
+
details). These calculations reveal that 88–90% of the binding
−
energy is accounted for by O–H⋯Cl interactions, i.e. that presumably important, it is clearly not the only factor. For
other contributions are minimal and this single hydrogen example, 1 displays similar chloride binding affinities in 9 : 1
+
bond is responsible for the vast majority of the binding CD CN : D O to charged receptors containing a pyridinium
3
2
−
1
(
Tables S13–S15†).
group and two phenolic O–H and two amide groups (65 M
Despite the fact that 1 and 2 each contain only one signifi- for 1 vs. 61–170 M ).
cant hydrogen bond donor, they are able to bind anions between 1 and Cl in 9 : 1 CD CN : D O (65 M ) is within one
strongly, with association constants for chloride >10 M in order of magnitude of those reported for a di-cationic bis-imi-
+
+
+
−1 27,55
Indeed, the association constant
+
−
−1
3
2
4
−1
2
+
acetonitrile, and respectable values obtained in competitive dazolium macrocycle 4 (structure provided in Fig. 6, K
a
163
and
−
1
56
57
9
: 1 acetonitrile : water mixtures. To put this in context, we
M
),
tetra-cationic tetra-imidazolium macrocycles
have compared these association constants with known recep- tetra-imidazolium catenanes (binding constants and structures
−
56
tors that have been reported to bind Cl in acetonitrile or 9 : 1 for these compounds are provided in Fig. S69†).
+
+
acetonitrile : water. Unsurprisingly, 1 and 2 bind chloride
To investigate the effects of the various contributions to
more strongly than neutral molecules containing a single O–H guest binding, we used Energy Decomposition Analysis based
group. Interestingly, this is true even for molecules with simi- on Absolutely Localised Molecular Orbitals within a conti-
5
8
larly acidic O–H groups, suggesting that electrostatic attraction nuum solvent model [ALMO-EDA(solv)], to tease apart the
is important as well as the acidity of the O–H hydrogen bond. respective contributions to binding energies. It should be
+
52
For example, 4-nitrophenol has a very similar pK
a
to 2 (7.1
noted that these calculations isolate the contributions to the
3
3
+
compared with 6.8 for 2 ), but binds chloride with an associ- solution phase electronic energy and do not consider entropy,
ation constant of only 560 M in CD CN. It is notable that which would be expected to have a relatively consistent contri-
3
−
1
53
Fig. 6 Structures of molecules investigated computationally, their association constants for chloride, and ALMO-EDA(solv.) contributions to calcu-
lated interaction energies. The overall interaction energies are given to the left of the chart. ΔEPrep is preparation energy, ΔEElec is the electrostatic
component, ΔEPauli is the Pauli repulsion energy, ΔDisp is the dispersion energy, ΔESolv is the solvation energy contribution, ΔEPol is the polarisation
contribution and ΔECT is the charge transfer energy.
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 2794–2803 | 2799

View Article Online
Paper
Organic & Biomolecular Chemistry
bution as the same anion and solvent are being compared in Nevertheless, we suggest that future receptors containing well-
all cases. The calculations suggest that in acetonitrile binding designed charged O–H groups would be worthy targets of
+
+
2+
of chloride to any of 1 , 2 or the macrocycle 4 is significantly further investigation.
stronger than binding to nitrocatechol, which is itself signifi-
cantly stronger than chloride binding to nitrophenol (as
expected based on measured association constants). In 9 : 1
Experimental
2
+
2
MeCN : H O, binding of chloride to macrocycle 4 is more
+
Characterisation data, details of instrumentation, anion
binding experiments, pK measurements, X-ray crystallography
and computational studies are provided in the ESI.†
favourable than binding to 1 , but not by a huge amount.
As would be expected and as shown graphically in Fig. 6,
a
2
+
dicationic 4 has a significantly more favourable electrostatic
−
1
+
+
component (−758 kJ mol ) than mono-cationic 1 and 2
−
1
General remarks
(
−471 to −477 kJ mol ) or neutral nitrocatechol and nitro-
phenol (−214 to −316 kJ mol ). This favourability is largely The reaction of 3-hydroxypyridine and methyl iodide in n-pro-
−
1
2
9,59
counteracted however, by a significantly higher solvation panol afforded 1·I quantitatively, as previously reported.
2
+
−1
2+
−1
+
penalty for 4 (636 kJ mol for 4 , 329–330 kJ mol for 1
2·I was prepared on a multi-gram scale by a slight modification
+
−1
28
and 2 , 95–123 kJ mol for the neutral molecules). A more of a previously reported procedure (see ESI†). The synthesis
detailed data breakdown, including a similar analysis for 1
and 4 in 9 : 1 CH
+
H
of DBU ·PF from DBU is described in the ESI.†
6
2
+
3
CN : H
2
O is provided in the ESI.†
1·BPh
4
. A solution of 1·I (0.237 g, 1.00 mmol) in MeOH
(0.411 g,
Clearly recognition of chloride by more complex anion (10 mL) was added to a solution of NaBPh
4
receptors is complicated by preorganisation effects, i.e. it is 1.20 mmol) in MeOH (10 mL) resulting in the immediate for-
non-trivial to design a receptor where all donors can interact mation of a white precipitate. The suspension was allowed to
with the anion without significant reorganisation. stand for 30 minutes, and then filtered. The resulting white
+
+
Nevertheless, it is remarkable that 1 and 2 can effect chloride microcrystalline solid was washed with MeOH (3 × 5 mL) and
recognition in aqueous acetonitrile using a single charge- then thoroughly air-dried to give 1·BPh . Yield: 0.367 g
CN): 8.15 (br. s, 1H), 8.10 (d, J
4
1
assisted O–H group, and that the strength of this is broadly (0.856 mmol, 86%). H NMR (CD
3
comparable to receptors with a much higher charge and/or far = 7.8 Hz, 1H), 7.88 (dd, J = 8.8, 1.8 Hz, 1H), 7.73–7.78 (m, 1H),
more hydrogen bond donors. Calculations suggest that the 7.24–7.30 (m, 8H), 6.99 (dd, J = 7.4, 7.4 Hz, 8H), 6.84 (t, J = 7.4
1
3
combination of electrostatic attraction for the anion, coupled Hz, 4H). C NMR (d
6
-DMSO): 162.6–164.1 (m), 156.7, 136.3,
with an acidic hydrogen bond donor, and a relatively low sol- 135.5, 133.7, 130.8, 128.2, 125.3, 121.5, 47.8. HRESI-MS (pos.):
+
+
6 6
vation energy penalty drive the favourable binding of chloride 110.0608, calc. for 1 [C H NO] = 110.0606 Da. HRESI-MS
+
+
−
−
to 1 and 2 . Taken together, our results suggest that receptors (neg.): 319.1661, calc. for BPh
with few binding sites may bind guests more strongly than 2·BPh . A solution of 2·I (0.718 g, 2.50 mmol) in MeOH
would otherwise be expected, due to their lower solvation (25 mL) was added to a solution of NaBPh
4
[C24H20B] = 319.1658 Da.
4
4
(1.03 g,
3.00 mmol) in MeOH (25 mL) giving a pale yellow solution.
penalties.
After about 15 seconds, pale crystals began to form, and the
reaction was left to stand at room temperature for an hour,
and then in a freezer for an hour. After this time, the pale
yellow crystals were isolated by filtration, washed with MeOH
Conclusions
We have investigated the anion recognition capability of the (3 × 5 mL), and thoroughly air-dried to give 2·BPh
4
. Yield:
-DMSO): 9.16 (d, J = 5.7
H hydrogen bond donor. These molecules display strong chlor- Hz, 1H), 9.05 (d, J = 8.1 Hz, 1H), 7.95 (dd, J = 8.4, 8.4 Hz, 1H),
+
+
1
simple molecules 1 and 2 , which contain a single strong O– 1.04 g (2.17 mmol, 87%). H NMR (d
6
4
−1
ide binding in CD
binding in 9 : 1 CD
These association constants are much higher than neutral
compounds containing similarly acidic O–H groups, and are 135.7, 132.0, 130.6, 129.5, 125.5, 121.73, 121.70, 120.7, 119.7,
3
CN (K
a
> 10 M ), and moderate chloride 7.75–7.80 (m, 2H), 7.51–7.56 (m, 1H), 7.12–7.21 (m, 8H), 6.92
−
1
3
CN : D
2 a
O (K = 65 and 13 M , respectively). (dd, J = 7.4, 7.2 Hz, 8H), 6.78 (t, J = 7.2 Hz, 4H), 4.80 (s, 3H).
1
3
C NMR (d -DMSO) 162.8–164.2 (m), 151.2, 150.0, 147.1,
6
+
+
comparable to highly charged receptors. Computational 51.5. HRESI-MS (pos.): 160.0764, calc. for 2 [C10
H10NO] =
−
studies suggest this is due to a favourable balance of electro- 160.0762 Da. HRESI-MS (neg.): 319.1663, calc. for BPh
4
−
static attraction with a relatively low degree of solvation.
[C24
H
20B] = 319.1658 Da. ESI-MS (pos.): 639.3, calc. for
+
+
While it is clear that these types of molecules are prone to [2
2
4 2 2
·BPh ] i.e. [C44H40BN O ] = 639.3 Da.
deprotonation, it seems likely that hosts incorporating two or
2·Cl. A solution of 2·BPh (0.240 g, 0.500 mmol) in acetone
4
more hydroxypyridinium or hydroxyquinolinium groups would (10 mL) was added to a solution of TBA·Cl (0.167 g,
be very potent anion receptors. Unfortunately our first 0.600 mmol) in acetone (10 mL) resulting in the immediate
2
+
attempts to investigate such systems in the form of 3 were formation of a pale yellow precipitate. The suspension was
hampered by this compound’s low solubility. This compound swirled briefly, left to stand for 30 minutes and then the solid
2+
is also apparently poorly preorganised such that 3 has to was isolated by filtration, washed with acetone (3 × 3 mL) and
rearrange from an anti to syn conformation to bind anions. thoroughly air-dried to give 2·Cl as a yellow powder. Yield:
2800 | Org. Biomol. Chem., 2021, 19, 2794–2803
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
1
0
9
8
.096 g (0.049 mmol, 98%). H NMR (d
6
-DMSO): 12.20 (s, 1H), with water (3 × 5 mL), air-dried and then dried thoroughly in
1
.30 (d, J = 5.6 Hz, 1H), 9.10 (d, J = 8.4 Hz, 1H), 8.00 (dd, J = vacuo (<10 mbar, 50 °C). Yield: 0.318 g (0.464 mmol, 93%). H
.4, 5.7 Hz, 1H), 7.75–7.82 (m, 3H), 4.83 (s, 3H). C NMR (d
1
3
6
-
6
NMR (d -DMSO): 11.90 (br. s, 2H), 9.43 (d, J = 5.4 Hz, 2H), 9.22
DMSO): 151.1, 150.2, 146.9, 131.8, 130.3, 129.4, 121.5, (d, J = 8.0 Hz, 2H), 8.11–8.14 (m, 2H), 7.87 (d, J = 7.7 Hz, 2H),
+
1
20.2, 119.7, 51.2. HRESI-MS (pos.): 160.0764, calc. for 2
7.79 (dd, J = 7.9, 7.9 Hz, 2H), 7.47 (d, J = 7.7 Hz, 2H), 7.13 (s,
+
13
[C
10
H
10NO] = 160.0762 Da.
·Br . 8-Hydroxyquinoline (2.90 g, 20.0 mmol) was dissolved 136.5, 132.4, 130.7, 128.5, 126.7, 122.0, 121.1, 120.3, 63.7.
in acetonitrile (50 mL) and the solution heated to reflux. Solid NMR (d -DMSO): −70.2 (d, J = 711 Hz). P NMR (d -DMSO):
6
4H), 6.55 (s, 4H). C NMR (d -DMSO): 151.8, 149.0, 148.4,
1
9
3
2
F
3
1
6
6
1
,2-bis(bromomethyl)benzene (0.528 g, 2.00 mmol) was added −144.2 (sept., J = 711 Hz). HRESI-MS (pos.): 197.0840, calc. for
2
+
2+
and the solution heated at reflux for 24 hours under a nitrogen
atmosphere, during which time the colourless solution turned calc. for [3
yellow and a yellow precipitate developed. The reaction was (neg.): 145.0, calc. for PF
concentrated under reduced pressure and acetonitrile (20 mL)
3
i.e. [C26
H
20
N
2
O
2
]
= 197.0835 Da. ESI-MS (pos.): 931.4,
−
H
+
+
·PF ] i.e. [C H N O F P] = 931.3 Da. ESI-MS
6 52 42 4 4 6
2
−
6
= 145.0 Da.
6 6 2
3 ·PF . 3·(PF ) (0.127 g, 0.200 mmol) was dissolved in
-
H
was added, the suspension was sonicated for 15 minutes, MeOH (20 mL) with gentle heating. NEt3 (28 μL, 20 mg,
refluxed for 15 minutes and then cooled to room temperature. 0.20 mmol) was added to the hot yellow solution resulting in a
It was filtered, and the resulting yellow powder washed with change to a deep red colour. Within a few seconds, a bright
1
acetonitrile (3 × 5 mL) and air-dried. Analysis by H NMR spec- orange precipitate developed. The suspension was cooled to
troscopy indicated that the compound was approximately 90% room temperature and then left to stand for 10 minutes. It was
pure at this point, and the yield at this point was 0.615 g then filtered to isolate a bright orange precipitate, which was
1
(
∼55%, see Fig. S7† for H NMR spectrum of this material).
washed with MeOH (3 × 3 mL), Et O (3 × 3 mL), air-dried and
2
The crude product was dry-loaded onto silica (∼10 mL) and then dried thoroughly in vacuo. Yield: 0.106 g (0.196 mmol,
1
6
added to the top of a column of silica (∼25 mL) in 9 : 1 98%). H NMR (d -DMSO): 9.15 (d, J = 5.5 Hz, 2H), 8.97 (d, J =
DCM : MeOH. Approximately 250 mL of 9 : 1 DCM : MeOH was 8.3 Hz, 2H), 7.86–7.90 (m, 2H), 7.48 (dd, J = 7.9, 7.9 Hz, 2H),
eluted to remove an orange impurity, and then the eluent was 7.30 (d. J = 7.6 Hz, 2H), 7.18 (s, 4H), 6.99 (d, J = 7.9 Hz, 2H).
1
3
switched to 88 : 10 : 2 DCM : MeOH : NEt
3 6
, which caused the 6.80 (s, 4H). C NMR (d -DMSO): 154.8, 149.5, 148.0, 136.9,
1
9
yellow compound at the top of the column to turn purple. The 132.9, 131.1, 129.6, 127.3, 121.2, 120.9, 116.1, 62.6. F NMR
3
1
column was eluted with 200 mL of this solvent to remove a (d
6
-DMSO): −70.2 (d, J = 711 Hz). P NMR (d -DMSO): −144.2
6
−H
small purple band and then the eluent switched to 83 : 15 : 2 (sept., J = 711 Hz). HRESI-MS (pos.): 393.1609, calc. for 3 i.e.
+
DCM : MeOH : NEt (∼250 mL) to remove a large purple band. [C H N O ] = 393.1603 Da. ESI-MS (neg.): 144.8, calc. for
3
26 21
2
2
−
This was concentrated in vacuo to give a purple solid. This was PF
dissolved in 9 : 1 DCM : MeOH (20 mL) to give a deep purple
solution, to which was added dropwise conc. HBr(aq). Addition
6
= 145.0 Da.
of HBr initially caused a colour change from deep purple to Conflicts of interest
bright red-orange and the development of a precipitate.
Continued addition of HBr(aq) caused the colour to change to a
There are no conflicts to declare.
bright yellow colour, at which point addition was stopped
(
0.3–0.4 mL HBr(aq) was added in total). The yellow suspension
Acknowledgements
was left to stand for 30 minutes, and then filtered to give a
lemon-yellow powder, which was washed with 9 : 1
DCM : MeOH (3 × 5 mL), DCM (3 × 5 mL), and then dried in
vacuo. Yield of 3·Br : 0.368 g (0.664 mmol, 33%). H NMR (d -
DMSO): 11.88 (br. s, 2H), 9.45 (d, J = 5.7 Hz, 2H), 9.23 (d, J = 8.3
Hz, 2H), 8.10–8.15 (m, 2H), 7.87 (d, J = 7.9 Hz, 2H), 7.79 (dd, J =
7
4
1
1
We thank the Australian Research Council for financial
support (Australian Government Research Training Program
Scholarships to R. J. G. and M. T. B., FL170100041 to M. L. C.,
DE170100200 to N. G. W.). M. L. C. acknowledges generous
allocations of supercomputing time on the National Facility of
the Australian National Computational Infrastructure. We
thank Dr Naomi Haworth for preliminary computational calcu-
lations, Dr Stephanie Boer for useful discussions, and
Dr Michael Gardiner for assistance with synchrotron X-ray
crystallography studies. Parts of this work were conducted
1
2
6
.9, 7.9 Hz, 2H), 7.48 (d, J = 7.9 Hz, 2H), 7.13 (s, 4H), 6.56 (s,
1
3
H). C NMR (d
6
-DMSO): 151.8, 148.8, 148.4, 136.4, 132.4,
30.7, 128.4, 126.7, 122.1, 121.2, 120.2, 63.6. HRESI-MS (pos.):
2
+
2+
97.0840, calc. for 3 i.e. [C H N O ] = 197.0835 Da. ESI-MS
26
+
20 2 2
+
(
pos.): 473.1, calc. for [3·Br] i.e. [C26
This reaction to produce 3·Br also produced 4·Br, which we
were able to isolate in small quantities (see ESI†).
·(PF . A solution of 3·Br (0.277 g, 0.500 mmol) in 3 : 1
O : MeOH (30 mL) was added to a solution of NH PF
0.196 g, 1.20 mmol) in H O (10 mL) containing 10 drops of
22 2 2
H N O Br] = 473.1 Da.
2
60
43
using the MX1 and MX2 beamlines at the Australian
Synchrotron.
3
6
)
2
2
H
(
2
4
6
2
Notes and references
1
% HPF6(aq) resulting in the immediate formation of a precipi-
tate. The suspension was left to stand for 30 minutes and then
filtered to give a very pale yellow powder, which was washed
1 N. H. Evans and P. D. Beer, Angew. Chem., Int. Ed., 2014, 53,
11716–11754.
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 2794–2803 | 2801

View Article Online
Paper
Organic & Biomolecular Chemistry
L. Chen, S. N. Berry, X. Wu, E. N. W. Howe and P. A. Gale, 30 In the case of the PF −
salts, the F NMR spectrum was
19
2
6
Chem, 2020, 6, 61–141.
recorded in a solution containing a known amount of
2,2,2-trifluoroethanol. A delay time of 30 seconds was used
between scans to ensure complete relaxation of the fluorine
nuclei, see: (a) W. He, F. Du, Y. Wu, Y. Wang, X. Liu, H. Liu
and X. Zhao, J. Fluorine Chem., 2006, 127, 809–815;
(b) A. O. Mattes, D. Russell, E. Tishchenko, Y. Liu,
R. H. Cichewicz and S. J. Robinson, Concepts Magn. Reson.,
Part A, 2016, 45, e21422.
3
4
J. Cai and J. L. Sessler, Chem. Soc. Rev., 2014, 43, 6198–6213.
T. M. Beale, M. G. Chudzinski, M. G. Sarwar and
M. S. Taylor, Chem. Soc. Rev., 2013, 42, 1667–1680.
J. Y. C. Lim and P. D. Beer, Chem, 2018, 4, 731–783.
S. A. Boer, E. M. Foyle, C. M. Thomas and N. G. White,
Chem. Soc. Rev., 2019, 48, 2596–2614.
5
6
7
8
9
J. M. Coterón, F. Hacket and H.-J. Schneider, J. Org. Chem.,
1
996, 61, 1429–1435.
H. Miyaji and J. L. Sessler, Angew. Chem., Int. Ed., 2001, 40,
54–157.
31 A peak for the O–H⋯O− hydrogen atom appears to be
located halfway between the two oxygen atoms. In both
structures, this is a special position, so the hydrogen atom
was placed on this position, i.e. exactly halfway between the
two oxygen atoms 1.21 Å from each atom.
1
K. J. Winstanley, A. M. Sayer and D. K. Smith, Org. Biomol.
Chem., 2006, 4, 1760–1767.
1
1
1
1
1
1
1
1
1
1
2
0 J.-I. Kim, H. Juwarker, X. Liu, M. S. Lah and K.-S. Jeong, 32 A. L. Spek, Acta Crystallogr., Sect. C: Struct. Chem., 2015, 71,
Chem. Commun., 2010, 46, 764–766. 9–18.
1 M. J. Kim, H.-W. Lee, D. Moon and K.-S. Jeong, Org. Lett., 33 S. F. Mason, J. Chem. Soc., 1958, 674–685.
012, 14, 5042–5045. 34 We acidified a solution of 3·Br with HBr(aq) and then
2
2
2 K. L. Wong, G. L. Law, Y. Y. Yang and W. T. Wong, Adv.
Mater., 2006, 18, 1051–1054.
3 C. H. Chen and F. P. Gabbaï, Angew. Chem., Int. Ed., 2018,
titrated in NaOH(aq) while monitoring the pH of the solu-
tion using a pH meter. At a pH of approximately 5.5, a faint
orange colour began to appear, consistent with deprotona-
tion beginning to occur. By a pH of 6, precipitation
occurred and so we were not able to measure further.
5
7, 521–525.
4 N. Busschaert, C. Caltagirone, W. Van Rossom and
P. A. Gale, Chem. Rev., 2015, 115, 8038–8155.
5 X. Wu, A. M. Gilchrist and P. A. Gale, Chem, 2020, 6, 1296–
35 We were not possible to follow the O–H resonance in any of
the anion binding studies reported herein. Even when no
1
309.
6 N. G. White and M. J. MacLachlan, Chem. Sci., 2015, 6,
245–6249.
D O or MeOD was used in the solvent mixture, these peaks
2
+
were not well-resolved. In the case of 1 in CD
3
CN, we
+
6
could not see this proton resonance at all. In the case of 2
7 C.-F. Ng, H.-F. Chow, D. Kuck and T. C. W. Mak, Cryst.
Growth Des., 2017, 17, 2822–2827.
8 J. W. A. Harrell, M. L. Bergmeyer, P. Y. Zavalij and
J. T. Davis, Chem. Commun., 2010, 46, 3950–3952.
in CD CN, it was visible but very broad and weak and thus
3
not suitable for monitoring quantitative binding.
Qualitatively, it showed a significant movement ∼4 ppm, and
a similar trend to that observed for C–H proton resonances.
9 X.-D. Wang, S. Li, Y.-F. Ao, Q.-Q. Wang, Z.-T. Huang and 36 Bindfit, accessed at supramolecular.org.
D.-X. Wang, Org. Biomol. Chem., 2016, 14, 330–334.
0 T. Saha, S. Dasari, D. Tewari, A. Prathap, K. M. Sureshan,
A. K. Bera, A. Mukherjee and P. Talukdar, J. Am. Chem. Soc.,
37 Sulfate is not appreciably basic, and indeed the conjugate
base of sulfate, HSO4−, appears to protonate 3
−H+
, the
(Fig. S28†). As shown in
2
+
deprotonated form of 3
3
Fig. S25,† addition of sulfate to 2 in CH CN causes a
+
2
014, 136, 14128–14135.
2
1 A. G. Schafer, J. M. Wieting, T. J. Fisher and A. E. Mattson,
Angew. Chem., Int. Ed., 2013, 52, 11321–11324.
2 K. M. Engle, L. Pfeifer, G. W. Pidgeon, G. T. Giuffredi,
purple colour (as opposed to the orange colour that results
from deprotonation), but it is unclear what the cause of
this is.
2
A. L. Thompson, R. S. Paton, J. M. Brown and 38 A. Albert and J. N. Phillips, J. Chem. Soc., 1956, 1294–1304.
V. Gouverneur, Chem. Sci., 2015, 6, 5293–5302. 39 P. Thordarson, Chem. Soc. Rev., 2011, 40, 1305–1323.
3 K. M. Diemoz, S. O. Wilson and A. K. Franz, Chem. – Eur. J., 40 S. Alvarez, Dalton Trans., 2013, 42, 8617–8636.
2
2
2
2
2
2
2
2
016, 22, 18349–18353.
41 M. Szafran, A. Komasa, A. Katrusiak, Z. Dega-Szafran and
P. Barczyński, J. Mol. Struct., 2007, 844–845, 102–114.
4 A. Shokri, J. Schmidt, X.-B. Wang and S. R. Kass, J. Am.
Chem. Soc., 2012, 134, 16944–16947.
5 A. Shokri, X.-B. Wang and S. R. Kass, J. Am. Chem. Soc.,
42 A structure of 2·Cl as its hydrate has been reported. In this
+
structure the O–H group of 2 hydrogen bonds to a water
−
2
013, 135, 9525–9530.
molecule, which then hydrogen bonds to a Cl anion, i.e.
+
6 E. Faggi, R. Porcar, M. Bolte, S. V. Luis, E. García-Verdugo
and I. Alfonso, J. Org. Chem., 2014, 79, 9141–9149.
7 M. Morshedi, S. A. Boer, M. Thomas and N. G. White,
Chem. – Asian J., 2019, 14, 1271–1277.
there are no direct H-bonds from 2 to the anion.
C. Romming and E. Uggerud, Acta Chem. Scand., Ser. B,
1983, 37, 791–795.
43 D. Aragao, J. Aishima, H. Cherukuvada, R. Clarken,
M. Clift, N. P. Cowieson, D. J. Ericsson, C. L. Gee,
S. Macedo, N. Mudie, S. Panjikar, J. R. Price, A. Riboldi-
Tunnicliffe, R. Rostan, R. Williamson and T. T. Caradoc-
Davies, J. Synchrotron Radiat., 2018, 25, 885–891.
8 J. P. Saxena, W. H. Stafford and W. L. Stafford, J. Chem.
Soc., 1959, 1579–1587.
9 S. L. Shapiro, K. Weinberg and L. Freedman, J. Am. Chem.
Soc., 1959, 81, 5140–5145.
2802 | Org. Biomol. Chem., 2021, 19, 2794–2803
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
4
4 M. W. Hosseini, R. Ruppert, P. Schaeffer, A. De Cian, 50 Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120,
N. Kyritsakas and J. Fischer, J. Chem. Soc., Chem. Commun.,
994, 2135–2136.
5 S. J. Brooks, P. A. Gale and M. E. Light, CrystEngComm,
005, 7, 586–591.
6 N. G. White and M. J. MacLachlan, Cryst. Growth Des.,
015, 15, 5629–5636.
215–241.
1
51 A. V. Marenich, C. J. Cramer and D. G. Truhlar, J. Phys.
Chem. B, 2009, 113, 6378–6396.
52 M. M. Fickling, A. Fischer, B. R. Mann, J. Packer and
J. Vaughan, J. Am. Chem. Soc., 1959, 81, 4226–4230.
53 D. K. Smith, Org. Biomol. Chem., 2003, 1, 3874–3877.
4
4
4
4
2
2
7 X. He, S. Hu, K. Liu, Y. Guo, J. Xu and S. Shao, Org. Lett., 54 R. Pizer and L. Babcock, Inorg. Chem., 1977, 16, 1677–1681.
2
006, 8, 333–336.
55 We note that while the halide binding properties are
similar, these receptors display signifcantly stronger sulfate
8 Measurement of the anion recognition properties of the
protonated form of DBU, the base we used to measure pKas
+
binding than 1 . See ref. 27.
in an organic solvent show it interacts relatively weakly 56 C. J. Serpell, A. Y. Park, C. V. Robinson and P. D. Beer,
−
1
with chloride (K
ger interaction of Cl with 2 over DBU to favour the pro- 57 W. W. H. Wong, M. S. Vickers, A. R. Cowley, R. L. Paul and
tonated form of the O–H receptor. See ESI† for further P. D. Beer, Org. Biomol. Chem., 2005, 3, 4201–4208.
details, and synthesis and X-ray crystal structure of 58 Y. Mao, M. Loipersberger, K. J. Kron, J. S. Derrick,
a
= 419 M ). Hence we expected the stron-
Chem. Commun., 2021, 57, 101–104.
−
+
H+
H
DBU ·PF .
C. J. Chang, S. M. Sharada and M. Head-Gordon, Chem.
Sci., 2021, 12, 1398–1414.
6
4
9 The shortest C–H⋯halide hydrogen bond observed is 96%
of ΣvdW in the structure of 3·Br
hydrogen bond distances range from 71–80% of ΣvdW).
Slightly shorter C–H⋯sulfate hydrogen bonds are seen in 60 N. P. Cowieson, D. Aragao, M. Clift, D. J. Ericsson, C. Gee,
2
(whereas O–H⋯halide 59 R. A. Lowe, D. Taylor, K. Chibale, A. Nelson and
S. P. Marsden, Bioorg. Med. Chem., 2020, 28, 115442.
the structures of 2
although these are significantly longer than the O–
H⋯sulfate hydrogen bonds (64–66% ΣvdW).
2
·SO
4
and 3·SO
4
(88 and 89% ΣvdW),
S. J. Harrop, N. Mudie, S. Panjikar, J. R. Price, A. Riboldi-
Tunnicliffe, R. Williamson and T. Caradoc-Davies,
J. Synchrotron Radiat., 2015, 22, 187–190.
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 2794–2803 | 2803