Improved complexation of paraquats with crown ether-based pyridyl cryptands

Article abstract of DOI:10.1002/hc.21406

The commonly used paraquat guest N,N′-dimethyl-4,4′-bipyridinium bis(hexafluorophosphate) (7) was targeted for structural alterations to improve the binding constants with crown ethers and cryptands. The association constant was improved by one order of magnitude (to K_a?=?1.00?×?10⁶?mol?L^?1) with the dibenzo-30-crown-10 pyridyl cryptand host 5 by changing the N-methyl groups to benzyl and the counterions from PF₆ to TFSI, that is, 11. Moreover, through addition of a p-bromobenzyloxy moiety on the 4-position of the pyridyl ring of cryptand 18, an association constant of K_a?=?5.35?×?10⁶?mol?L^?1 was achieved with paraquat 11.

Full text of DOI:10.1002/hc.21406

Received: 5 June 2017
DOI: 10.1002/hc.21406
Revised: 30 October 2017
|
R E S E A R C H A R T I C L E
Improved complexation of paraquats with crown ether-based
pyridyl cryptands
Terry L. Price Jr.
Carla Slebodnick
Harry W. Gibson
|
|
Department of Chemistry, Virginia Tech,
Blacksburg, VA, USA
Abstract
The commonly used paraquat guest N,N′-dimethyl-4,4′-bipyridinium
bis(hexafluorophosphate) (7) was targeted for structural alterations to improve the
binding constants with crown ethers and cryptands. The association constant was
improved by one order of magnitude (to Ka = 1.00 × 106 mol L−1) with the dibenzo-
30-crown-10 pyridyl cryptand host 5 by changing the N-methyl groups to benzyl and
the counterions from PF6 to TFSI, that is, 11. Moreover, through addition of a
p-bromobenzyloxy moiety on the 4-position of the pyridyl ring of cryptand 18, an
association constant of Ka = 5.35 × 106 mol L−1 was achieved with paraquat 11.
Correspondence
Harry W. Gibson, Department of Chemistry,
Virginia Tech, Blacksburg, VA, USA.
Email: hwgibson@vt.edu
Funding information
National Science Foundation of the USA,
Grant/Award Number: DMR 0704076,
CHE-1106899, CHE-1507553
displays reasonable association constants with various crown
ethers[2–5] and analogous cryptands.[4,6–11]
1
INTRODUCTION
|
In supramolecular chemistry, one of the most important pa-
rameters of a host-guest system is its association constant. By
defining how well a host-guest pair bind one another, the ex-
perimentalist can determine what tasks may be achieved with
the system. An example of this is the production of supramo-
lecular polymers whose theoretical degree of polymerization
(DP) may be calculated from the experimentally determined
association constant (Equation 1), Ka, and [H]o, the concen-
tration of a heteroditopic (AB) monomer; clearly, higher as-
sociation constants lead to higher DPs.[1]
Here, we investigate how alterations to the paraquat salt
affect association constants with cryptands 5 and 6 as well as
how accessing a less polar solvent can increase the association
constant. Alterations to the paraquat salt were investigated in
terms of the N-alkyl group and the counter anion. Focus was
directed toward the association constants of cryptand 5, given
that its precursor crown ether 2 is produced through a high
yielding templation process called the WPW process after its
three key developers (Wang, Pederson, Wessels).[6,11–13] And
diol 2 is converted to cryptand 5 in 89% yield via a template
process.[14]
2Ka[H]o
DP=
(1+4Ka[H]o)1∕2 −1
(1)
2
RESULTS AND DISCUSSION
if 4 Ka[H]o ≫ 1 then DP = (Ka [H]o)1/2
.
|
Ourgroupfocusedontheevolutionofdibenzocrownether-
based hosts to bind paraquat (N,N″-dialkyl-4,4′-bipyridinium
salt) guests more strongly. Figure 1 and Table 1 show asso-
ciation constants reported by the Gibson and Huang groups.
Through analysis of CPK models and crystal structures, host
molecules have been advanced to highly evolved crown-
based pyridyl cryptands, for example 5 and 6. However, little
improvement has been offered in way of the paraquat guest,
the standard being dimethyl paraquat as the PF6 salt (7) which
Given that the association constants for this study were ex-
pected to be relatively high, in excess of 1 × 104, ITC was
used for their estimation. However, as the literature-reported
values for 5•7 were obtained through ITC and 6•7 via an
NMR study, the ITC titrations of 5•7 and 6•7 were carried
out. Under similar ITC titration conditions (acetone, 23°C),
the Ka of 5•7 was 2.21 × 105 mol L−1, while that of 6•7 was
1.12 × 106 mol L−1, bringing the separation of the values
for the two cryptands to just under one order of magnitude.
Note that the bis(m-phenylene)-32-crown-10-based pyridyl
Contract grant sponsor: National Science Foundation.
Contract grant number: DMR 0704076, CHE-1106899, CHE-1507553.
|
Heteroatom Chemistry. 2017;e21406.
https://doi.org/10.1002/hc.21406
wileyonlinelibrary.com/journal/hc
© 2017 Wiley Periodicals, Inc.
1 of 10

2 of 10
PRICE Et al.
|
counterions have been well documented in the literature
to increase the solubility of organic salts in low dielec-
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
HO
OH HO
OH
2
1
O
O
tric solvents because they are less ion paired than halides
− [16,17]
HO
or PF6 ;
consequently the TFSI and tetrakis(perflu-
O
O
O
O
O
O
O
O
orophenyl)borate anions were chosen for investigation.
Regarding changes to the N-substituted group, the choice
of benzyl in 10-13 was made because a wide range of ben-
zyl building blocks is commercially available.
O
O
4
O
O
O
O
O
O
O
O
O
3
O
O
O
O
O
O
O
O
Table 3 lists the solubilities of paraquats 7-13 in acetone
and DCM. Due to the high costs associated with tetrakis(per-
fluorophenyl)borate anion, only small amounts of 9 and 12
were synthesized, and no acetone solubility testing was con-
ducted. A gain in acetone solubility of 13-fold was seen by
simply going from PF6 to TFSI counterions, 7 vs 8. Similarly,
by changing from an N-methyl to an N-benzyl group, a sol-
ubility increase of >4-fold was observed, 7 vs 10. In terms
of DCM solubility 9, 11, and 12 dissolved in DCM in mea-
surable amounts, while 8 was only very, very sparingly solu-
ble. Unfortunately perflurorobenzyl salt 13 was insoluble in
DCM.
O
O
N
O
O
N
O
O
O
6
O
O
O
O
O
O
O
O
O
O
O
O
5
O
O
O
O
O
O
O
O
O
F
F
F
F
F
F
F
F
F
N+
N+
P-
F
P-
F
F
7
FIGURE 1 Various hosts investigated by the Gibson and Huang
groups with dimethyl paraquat PF6 (7)
Table 4 compares association constants for the various
host-guest pairs in acetone and DCM. Additionally, thermo-
dynamic values are given for each titration. The first notice-
able items in Table 4 are the deviations of 5•7 and 6•7 from
the literature values. It is suspected that the deviation of 5•7
is a result of differences in data gathering. The reported Ka
value of 5•7 was derived from a 33-point titration,[6] while
values reported here were obtained using a 60-point titration
curve; as the isotherm is fit to a mathematical construct, a
more defined curve produces a better fit. As for the deviation
of 6•7 from the literature values, this is attributed to differ-
ences in experimental methodologies; the reported value for
6•7 was obtained from a competitive NMR study,[7] while the
results reported here within were obtained via ITC.
TABLE 1 Association constants of hosts 1-6 with 7 in acetone at
room temperature
Ka
Complex
(mol L−1
)
1•7[4b]
5.7 × 102
1.1 × 103
6.4 × 103
6.1 × 104
1.0 × 105
5.0 × 106
2•7[2]
3•7[7]
4•7[9]
5•7 (ITC)[6]
6•7 (NMR)[7]
As seen in Table 4, improvements for host 5 were incre-
mental with 8, 10, and 11 vs 7, and 5•7 in acetone at 25°C
has an association constant of 2.21 × 105 mol L−1; however,
changing PF6 to TFSI, the N-methyl to N-benzyl, and by alter-
ing the titration solvent to DCM resulted in an improvement
in Ka of 5•11 to 1.00 × 106 mol L−1. Teasing out the effect of
each change, we can see the effects of switching from acetone
to DCM by looking at 5•11 in Table 4 for acetone and DCM.
Ka did not change within experimental error, and ∆H and ∆S
also remained the same within experimental error. As for the
effects of changing to N-benzyl groups, we see in the titra-
tion of 5•8 compared to 5•11, Ka increased 2.9-fold. In terms
of ∆H and ∆S, the benzyl group leads to a more favorable
enthalpy term that offsets a less favorable entropy term. The
change in ∆H for 5•8 compared to 5•11, −14.7 vs −18.1 kcal/
mol, is likely attributable to the increased positive character of
the N-benzylic protons relative to the N-methyl protons.
In comparison with PF6, TFSI counterions consistently
yielded higher association constants in complexes with
cryptand 6 suffers from a low cyclization yield to form the
precursor diol (cyclization yield to diester 43%),[15] while a
very high yielding, template synthesis provides the analo-
gous precursor to cryptand 5 in very high yield (cyclization
yield to diester 93%).[6]
Table 2 provides a listing of paraquat guests investi-
gated. Considering that solvents with higher dielectric con-
stants lead to lower association constants, the paraquat salt
design was focused toward solubility in relatively nonpolar
solvents such as dichloromethane (DCM). Additionally,
hosts such as 1-6 have low solubilities in solvents such as
acetone. Salt solubilities are highly influenced by anion/
cation combination, and paraquat salts such as 7 are typi-
cally soluble only in relatively polar aprotic solvents such
as acetone. Changes in anion and increasing the organic
character of the cation took priority. The bis(trifluoro-
methylsulfonyl)imide (TFSI) anion and large borate-based

PRICE Et al.
3 of 10
|
TABLE 2 Paraquat salts examined in this study
N
N
N
N
Counterion
−
PF6
7
8
9
10
11
12
(CF3SO2)2N−
(C6F5)4B−
TABLE 3 Solubilities (mmol L−1) of paraquats 7-13
increase (5•10 vs 5•11). In various other paraquat PF6 (7
and vinylogs) complexes with 5 and similar cryptands, the
paraquat dication sits directly in the cavity of the cryptand
nearly orthogonal to the pyridyl arm and does not penetrate
the space between the two ethyleneoxy segments.[6,10,18,19]
In a direct comparison of crystal structures of 5•7[6] to
5•8, there are several notable differences. First, in 5•8
(Figure 2), instead of sitting between the pyridyl arm and
crown ether portion of the cryptand, the paraquat has in-
serted itself in such a way that one methyl group lies di-
rectly inside the crown ether segment of the ring. This new
orientation leads to five hydrogen bonds at less than 3 Å
with that methyl group; additionally, two protons on the
paraquat ring provide four interactions that are less than
Compound
Acetone
DCM
7
8
56.2
749
Unobservable
<0.04
9
Not taken
261
0.49
Unobservable
6.26
3.60
Unobservable
10
11
12
13
1.08 × 103
Not taken
Not taken
5 (Table 4). Dimethyl paraquat displayed a 32% boost in
its association constant by switching from PF6 (in 5•7)
to TFSI (in 5•8), while dibenzyl paraquat enjoyed a 28%
TABLE 4 ITC results for cryptands 5 and 6 with paraquats 7, 8, 10, and 11
Solvent
Ka (mol L−1
)
∆G (kcal/mol)
∆H (kcal/mol)
∆S (cal/mol K)
5●7 (lit.)[6]
5●7
5●8
5●10
5●11
5●11
5●13
Acetone
Acetone
Acetone
Acetone
Acetone
DCM
1.00 × 105
−6.82
−13.0
−20.7
2.21 × 105 ( 0.22 × 105)
2.86 × 105 ( 0.17 × 105)
6.49 × 105 ( 0.54 × 105)
8.34 × 105 ( 0.49 × 105)
1.00 × 106 ( 0.13 × 106)
8.47 × 104 ( 0.39 × 104)
5.0 × 106
−7.29 ( 0.73)
−7.44 ( 0.44)
−7.93 ( 0.66)
−8.08 ( 0.48)
−8.18 ( 1.06)
−6.72 ( 0.31)
N/A
−14.8 ( 0.2)
−14.7 ( 0.1)
−18.9 ( 0.1)
−18.1 ( 0.1)
−18.1 ( 0.2)
−12.9 ( 0.2)
N/A
−25.3 ( 2.5)
−24.2 ( 1.5)
−36.7 ( 3.1)
−33.7 ( 2.0)
−33.3 ( 4.3)
−20.7 ( 1.0)
N/A
Acetone
Acetone
6●7 (lit. via
NMR)[7]
6●7
6●8
6●11
Acetone
Acetone
DCM
1.12 × 106 ( 0.21 × 106)
2.22 × 106 ( 0.36 × 106)
1.41 × 106 ( 0.23 × 106)
−8.25 ( 1.55)
−8.66 ( 1.40)
−8.39 ( 1.37)
−18.8 ( 0.4)
−17.5 ( 0.2)
−13.8 ( 0.1)
−35.4 ( 6.7)
−29.7 ( 4.8)
−18.2 ( 3.0)

4 of 10
PRICE Et al.
|
(A)
(B)
FIGURE 2 Crystal structure of 5•8
grown from a chloroform:acetone 1:1 (v:v)
mixture by liquid-liquid diffusion of diethyl
ether; nonparaquat hydrogen atoms and
impurities have been removed for clarity; A,
top view; B, side view; C, hydrogen bonding
to the p-ethyleneoxy chain (counterions
removed for clarity); D, hydrogen bonding
to the m-ethyleneoxy chain (counterions
removed for clarity); E, hydrogen bonding
to acetone (solvent), pyridine and ester
group (counterions removed for clarity);
F, planes of stacked aromatic rings shown
with centroids of stacked rings and plane
inclinations indicated. Hydrogen bond
parameters: C–O distances (Å), C–H–O
distances (Å), C–H–O angles (deg) A:
3.291, 2.560, 133.96; B: 3.818, 3.116,
129.64; C: 3.284, 2.339, 161.30; D: 3.403,
2.783, 121.72; E: 3.601, 2.826, 136.44;
F: 3.264, 2.332, 156.38; G: 3.370, 2.828,
115.61; H: 3.866, 3.024, 144.66; I: 3.004,
2.327, 127.76; J: 3.140, 2.607, 115.86;
K: 3.275, 2.426, 148.72; L: 3.248, 2.692,
117.95; M: 3.435, 3.066, 104.95. Face-
to-face π-stacking parameters: centroid-
centroid distance (Å): N) 3.586; O) 3.782;
ring plane/ring plane inclinations (deg): (i)
7.68°; (ii) 1.63°
(C)
(E)
(D)
(F)
3 Å in spacing. The second methyl group of the complex
sits directly outside the cryptand’s cavity and is hydrogen-
bonded to the oxygen of the TFSI counterion. In total, the
highly dispersed negative charge over the TFSI counterion
and multiple oxygen atoms allow for the anion to play a
greater role in interacting with the paraquat cation via hy-
drogen bonding. The result is a complex that contains more
and stronger hydrogen bonds, leading to a higher associa-
tion constant.
Switching to the TFSI counterion offers a near doubling
of the association constant for 6•8 (Ka = 2.22 × 106) relative
to 6•7. The complex of 6 with the benzyl paraquat 11 gave a
slight decrease in binding relative to 6•7, with lower ∆H and
∆S. Surprisingly, the association constant for 6•11 in DCM
at 25°C, 1.41 × 106 mol L−1, was less than the complex of
6•8 in acetone at 25°C, 2.22 × 106 mol L−1; this may reflect
solvation energies of the components in some manner.
Encouraged by the results for 5•11 in DCM, we spec-
ulated that the benzyl group of 11 could be used to fur-
ther increase the association constant with a cryptand that
also contained a benzyl group, because there would be an
increased opportunity for π-stacking. To test this hypothesis,
new cryptand 18 was prepared as shown in Scheme 1; the use
of the diisopropyl ester in the alkylation step essentially pre-
vented N-alkylation via steric constraints, whereas the diethyl
ester produced a lower yield of 15, which was contaminated
with ~20% of N-alkylated product. The formation of 18 by a
new templation method now results in a 74% yield.[14]
An ITC study (Figure 3) proved part of the hypothesis
correct; a fivefold increase in the association constant was
observed for 18•11 (Ka = 5.35 × 106 mol L−1) compared
to 5•11. The belief that this would be driven by π-stacking,
however, was possibly incorrect. Figure 4 shows the crystal
structure of 18•11. Unfortunately, only two of the four crys-
tallographically unique TFSI anions could be located. The
remaining two TFSI anions are in a channel and are too heav-
ily disordered to locate the individual atoms. Although this
affects the overall quality of the structure and thus the uncer-
tainties in both lengths and angles, the hosts and guests are
clearly identified in the structure and provide useful informa-
tion about the binding interactions. The 4-bromobenzyl rings
link two adjacent cryptands via interaction of their 2 protons

PRICE Et al.
5 of 10
|
SCHEME 1 Synthesis of new cryptand 18: (i) iPrOH/H2SO4 reflux 36 h, 86%; (ii) p-BrC6H4CH2Br/K2CO3/acetone, reflux 8 h, 87%; (iii)
KOH/H2O/THF, rt, 12 h, 99%; (iv) SOCl2/cat. DMF, reflux 4 h, 100%; (v) DCM pseudo-high dilution via syringe pump addition, then 3 days at rt,
46%
with ether oxygen atoms of the cryptands; in Figure 4b, the
hydrogen bonds labeled A and B have bond distances H–O
(C–H–O angles) of 2.532 (154.80°) and 2.855 (158.91°),
respectively.
The increase in association constant for 18•11 may be
partly attributed to the substituent affect on the basicity
of the pyridyl nitrogen atom. Unique to 18•11 over 5•11
is the interaction of TFSI with the host: in Figure 4c two
bonds (C and D) at 2.717 and 3.020 Å. The other two
oxygen atoms of TFSI hydrogen bond to the paraquat,
I, J, F, and G in Figure 4d. In regard to the electron re-
leasing effect of the 4-bromobenzyloxy group, it can be
seen in Figure 4g that the nitrogen-hydrogen bonds U
and V (3.092 and 2.276 Å, respectively) are strong when
compared to the nitrogen-hydrogen bond lengths of M
and L (3.066 and 2.692 Å, respectively) found in 5•8
(Figure 2e).
It was anticipated that the electron-deficient pentafluoro-
FIGURE 3 ITC titration of 11 into 18 in DCM at 25°C:
phenyl rings of 13 could greatly increase the acidity of the
Ka = 5.35 ( 0.36) × 106 M−1, ΔΓ = −9.17 ( 0.04) kcal/mol,
ΔH = −16.3 ( 0.6) kcal/mol, ΔS = −23.9 ( 2.2) cal/mol·deg
benzylic protons and thereby ultimately drive the binding
1
constant higher with cryptand 5. In the H NMR spectrum,
the benzylic protons of 13 appear 0.25 ppm downfield from
those of hydrogenated analog 11, indicating deshielding and
more acidic protons. However, by changing to fluorinated
rings, solubility in DCM was lost. From Table 4, it can be
seen that despite the benzylic protons in 13 being more
acidic, the binding constant decreased by an order of mag-
nitude compared to 11; ΔH decreased markedly from −18.1

6 of 10
PRICE Et al.
|
(A)
(B)
(C)
(D)
(E)
(F)
(G)
(H)
(I)

PRICE Et al.
7 of 10
|
FIGURE 4 Incomplete crystal structure of 18•11 grown by liquid-liquid diffusion of pentane into a 1:1 v:v mixture of
dichloromethane:acetone; hydrogens which are not hydrogen-bonded removed for clarity; A, side view; B, top view with cryptand-cryptand
hydrogen bonding; C, TFSI hydrogen bonding to cryptand; D, TFSI and water hydrogen bonding to paraquat; E, hydrogen bonding with m-
ethyleneoxy chain; F, hydrogen bonding with p-ethyleneoxy chain; G, hydrogen bonding at the pyridyl arm; hydrogen bond parameters: C–O
distances (Å), C–H–O distances (Å), C–H–O angles (deg) A: 3.417, 2.532, 154.80; B: 3.758, 2.855, 158.91; C: 3.566, 2.717, 144.13; D: 3.793,
3.020, 135.71; E: 3.260, 2.428, 146.08; F: 3.414, 2.616, 141.89; G: 3.149, 2.745, 104.97; H: 3.295, 2.597, 127.53; I: 3.418, 2.515, 158.62; J: 3.368,
2.467, 158.21; K: 3.430, 2.529, 158.30; L: 3.104, 2.469, 124.18; M: 3.239, 2.261, 169.92; N: 3.502, 2.693, 143.32; O: 3.145, 2.803, 102.23; P:
3.493, 2.745, 136.13; Q: 3.110, 2.573, 115.97; R: 3.141, 2.208, 166.30; S: 3.554, 3.050, 112.89; T: 3.300, 2.833, 111.26; U: 3.790, 3.092, 131.53;
V: 3.183, 2.276, 159.36; W: 3.059, 2.496, 117.87. Face-to-face π-stacking parameters: centroid-centroid distance (Å): X: 4.501; Y: 3.696; Z: 4.497;
AA: 3.713; ring plane/ring plane inclinations (deg): (i) 7.90°; (ii) 1.10°; (iii) 3.97°; (iv) 3.65°
to −12.9 kcal/mol, even though ∆S also decreased substan-
tially. Thus, the anticipated increased acidity of the benzylic
protons of 13 did not result in improved complexation.
Interestingly, a linear correlation was found between en-
tropy and enthalpy values (Figure 5). This type of correlation
has been observed and described in the literature in terms
of enthalpy-entropy compensation.[20–22] Correlations such
as these have recently driven arguments for a connection be-
tween entropy and internally stored energies,[23] hidden ter-
m(s) connecting enthalpy and entropy,[24] and the potential
for a fourth law of thermodynamics that explains intermolec-
ular binding processes.[20]
DCM actually decreased slightly compared to dimethyl para-
quat PF6 (7) in acetone. Although cryptand 6 benefited from
the change of PF6 to TFSI with dimethyl paraquat (7 vs 8), at
best only a doubling in Ka was observed. Inspection of several
paraquat cryptand combinations (7 vs 8, 10 vs 11) revealed
that TFSI-containing paraquats generally yielded higher as-
sociation constants than their PF6 counterparts, but gains were
modest. A dibenzo-30-crown-10-based pyridyl cryptand con-
taining a p-bromobenzyloxy group in the pyridyl ring (18) led
to Ka = 5.35 × 106 mol L−1 with dibenzyl paraquat TFSI (11).
4
EXPERIMENTAL
Measurements
|
4.1
|
3
CONCLUSIONS
|
1H NMR spectra were obtained on JEOL ECLIPSE-500,
BRUKER-500, and AGILENT-NMR-vnmrs400 spectrome-
ters; 13C NMR spectra were collected at 125, 125, and 101 MHz,
respectively, on these instruments. HR-MS were obtained using
an Agilent LC-ESI-TOF system with acetonitrile as solvent.
Reagents were purchased and used as received without further
purification unless otherwise stated. Compounds 2,[6] 5,[6] 6,[7]
7,[25] 10,[26] and 13[27] were prepared as described by the litera-
ture procedures; similar yields were achieved. Crystals of 5•8
were grown from an equimolar solution of chloroform:acetone
1:1 (v:v) by liquid-liquid diffusion of diethyl ether. Crystals of
18•11 were grown from an equimolar DCM:acetone 1:1 (v:v)
solution by liquid-liquid diffusion of pentane.
In conclusion, the alteration of the dimethyl paraquat PF6
motif (7) to a dibenzyl paraquat TFSI motif (11) resulted in
increased solubility of the paraquat salt in less polar solvents.
In the less polar solvent DCM, the association constant of
dibenzo-30-crown-10-based pyridyl cryptand 5 with dibenzyl
paraquat TFSI (11) increased by an order of magnitude over
the analogous dimethyl paraquat PF6 (7) complex in acetone.
However, the association constant of the 32-crown-10-based
cryptand 6 did not increase similarly; the Ka of 6 with 11 in
4.2
Solubility testing
|
A mixture of solvent and excess paraquat salt was made and
placed in a water bath at 25°C for 8 hour. A 5.00 mL of aliquot
was removed from the saturated solution using a volumetric
pipette and placed in a clean, tared vial. Solvent removal and
determination of the mass of the solid residue allowed calcu-
lation of the concentration of the saturated solution.
4.3
Sample ITC titration
|
The following is an example using 5•11 in DCM at 25°C.
Host 5 was loaded into the cell of the instrument at a concen-
tration of 0.160 mmol L−1, while a 250-μL ITC syringe was
FIGURE 5 Plot of entropies vs enthalpies for complexation of
cryptands 5 and 6 with paraquats 7, 8, 10, 11, and 13; values in acetone
or DCM from Table 4

8 of 10
PRICE Et al.
|
loaded with 11 at a concentration of 2.291 mmol L−1. The
instrument was set to high gain (high sensitivity). The titra-
tion was achieved through 60 injections of 4.167 μL every
180 seconds; a primary filter period of 1 second and a sec-
ondary filter period of 3 second were applied (filter period
switch time 60 seconds). A background titration used exactly
the same titration conditions with the exception that the solu-
tion of 5 in the cell was replaced with pure DCM. The heats
for the dilution experiment were subtracted from the heats for
the titration of 11 with 5. Analysis of the data was carried out
using software provided by the manufacturer; the first datum
point was ignored. A “One Set of Sites” model was used;
stoichiometries other than 1:1 provided unsatisfactory fits.
solvent was removed by rotary evaporation. The crude ma-
terial was triturated with acetonitrile and DCM, followed
by collection on a fritted glass filter where it was washed
with DCM and air-dried: 6.45 g (99%), mp 260 (dec);
lit. mp 260 (dec).[30] The precipitate (1.47 g, 2.95 mmol)
in 10 mL of water was mixed with a solution of LiTFSI
(2.11 g, 8.02 mmol) in 5 mL of water. The precipitate was
filtered, washed with water, and allowed to air-dry; this
provided a colorless solid: 2.6514 g (100%), recrystallized
1
from water-acetone mixture (×3), mp 112.4-113.3°C. H
NMR (500 MHz, DMSO-d6) δ 9.52 (d, J = 7 Hz, 4H),
8.75 (d, J = 7 Hz, 4H), 7.62 (d, J = 7 Hz, 4H), 7.52-7.41
(m, 6H), 5.96 (s, 4H). 13C NMR (126 MHz, DMSO-d6) δ
149.34 (s), 145.65 (s), 134.01 (s), 129.48 (s), 129.22 (s),
128.86 (s), 127.23 (s), 119.45 (q, J = 323 Hz), 63.54 (s) (9
peaks expected and 9 peaks found). High res ESI MS: calc.
for C28H26F12N5O8S4 [M + NH4]+: m/z 916.0467; found:
m/z 916.0503 (error 3.9 ppm).
4.4
Dimethyl paraquat TFSI (8)
|
Dimethyl paraquat diiodide[28] (2.65 g, 6.02 mmol) in
20 mL of water was added to a solution of LiTFSI (4.10 g,
15.6 mmol) in 10 mL of water. The resulting precipitate was
filtered and allowed to air-dry to provide a colorless solid:
4.32 g (96%), recrystallized from water and acetone three
4.7
Dibenzyl paraquat TPPB (12)
|
times, 2.88 g (64%), mp 125.9-127.9°C; lit. mp 130°C.[29] 1
H
General procedure 1 was used to produce an off-white
solid, 162.3 mg (98%), mp 240.6-242.3°C, using diben-
zyl paraquat dibromide (48.6 mg, 0.0975 mmol) and lith-
ium tetrakis(pentafluorophenyl)borate tri(ethyl etherate)
NMR (500 MHz, acetone-d6) δ 9.42 (s, 4H), 8.87 (s, 4H),
4.77 (s, 6H). 13C NMR (126 MHz, acetone-d6) δ 150.67 (s),
147.88 (s), 127.76 (s), 120.95 (q, J = 322 Hz), 49.46 (s) (5
signals expected and 5 signals found). High res ESI MS: calc.
for C16H14NaN4O8S4F12 [M + Na]+: m/z 768.9395; found:
m/z 768.9405 (error 1.3 ppm).
1
(286.0 mg, 0.3148 mmol) dissolved in 3 mL of water. H
NMR (400 MHz, acetone-d6) δ 9.63 (d, J = 7 Hz, 4H), 8.96
(d, J = 7 Hz, 4H), 7.75-7.58 (m, 4H), 7.57-7.43 (m, 6H),
6.22 (s, 4H). 13C NMR (101 MHz, acetone-d6) δ 150.67 (s),
148.15 (br d, J = 245 Hz), 146.09 (s, overlap), 142.71 (br d,
J = 683 Hz), 136.10 (br d, J = 254 Hz), 136.92 (s, overlap),
133.24 (s), 130.05 (s), 129.58 (s), 129.17 (s), 127.77 (s), 64.98
(s) (12 signals expected and 12 signals found). High res ESI
MS: calc. for C72H22F40NaN2B2 [M + Na]+: m/z 1719.1223;
found: m/z 1719.1303 (error 4.7 ppm).
4.5
Dimethyl paraquat TPFB (9)
|
Dimethyl paraquat diiodide[28] (63.1 mg, 0.143 mmol)
in 3 mL of water was added to a solution of lithium
tetrakis(pentafluorophenyl)borate
tri(ethyl
etherate)
(310.4 mg, 0.3417 mmol) in 3 mL of water. The resulting
precipitate was filtered and allowed to air-dry to provide a
colorless solid: 0.2171 g (98%), mp 204.6-208.5°C. 1H NMR
(400 MHz, acetone-d6) δ 9.51 (d, J = 7 Hz, 4H), 8.96 (d,
J = 7 Hz, 4H), 4.79 (s, 6H). 13C NMR (101 MHz, acetone-d6)
δ 149.81 (s), 147.08 (s, overlap), 148.13 (br d, J = 237 Hz,
overlap), 143.13 (br d, J = 773 Hz, overlap), 136.11 (br d,
J = 253 Hz), 136.90 (s, overlap), 126.90 (s), 48.64 (s) (8 sig-
nals expected and 8 signals found). High res ESI MS: calc.
for C36H14F20N2B [M − TPFB]+: m/z 865.0925; found: m/z
865.0890 (error 4.0 ppm).
4.8
N,N′-Bis[pentafluorobenzyl]-4,4′-
|
bipyridinium TFSI (13)
To a round bottom flask containing acetonitrile (20 mL)
were added 4,4′-dipyridyl (0.7161 g, 4.585 mmol) and
2,3,4,5,6-pentafluorobenzyl bromide (2.0 mL, 13 mmol)
under nitrogen. The reaction mixture was held at reflux for
21 hours, after which the solvent was removed by rotary
evaporation. The crude material was triturated with acetoni-
trile and DCM, followed by collection on a fritted glass fil-
ter where it was washed with DCM and air-dried. The solid
was dissolved in 5 mL of water and mixed with a solution of
LiTFSI (3.29 g, 11.5 mmol) in 5 mL of water. The precipitate
was filtered, washed with water, and allowed to air-dry: a
4.6
General procedure 1: dibenzyl
|
paraquat TFSI (11)
To a round bottom flask containing acetonitrile (100 mL)
were added 4,4′-dipyridyl (2.03 g, 13.0 mmol) and benzyl
bromide (5.0 mL, 42 mmol) under nitrogen. The reaction
mixture was held at reflux for 21 hours, after which the
1
colorless solid, 4.88 g (99%), mp 166.5-167.9°C. H NMR
(500 MHz, DMSO-d6) δ 9.38 (d, J = 7 Hz, 4H), 8.75 (d,
J = 7 Hz, 4H), 6.21 (s, 4H). 13C NMR (126 MHz, DMSO-d6)

PRICE Et al.
9 of 10
|
δ 150.13 (s), 146.70 (s), 146.09 (br d, J = 255 Hz), 142.25
(br d, J = 255 Hz), 137.71 (dt, J = 252, 13 Hz), 127.66 (s),
119.9 (q, J = 323 Hz), 107.87 (tt, J = 8, 4 Hz), 51.99 (s) (9
signals expected and 9 signals found). High res MS: calc. for
C24H13F10N2 [M − 2TFSI + H]+: m/z 260.0493; found: m/z
260.0492 (error 0.4 ppm).
chilled, and filtered to produce a colorless solid, 2.13 g (99%),
1
mp 185.7-187.7°C. H NMR (400 MHz, DMSO-d6) δ 7.80
(s, 2H), 7.62 (s, 2H), 7.46 (d, J = 8 Hz, 2H), 5.36 (s, 2H). 13
C
NMR (101 MHz, DMSO-d6) δ 163.17 (s), 162.12 (s), 146.66
(s), 131.94 (s), 128.36 (s), 126.81 (s), 118.32 (s), 110.75 (s),
66.17 (s) (9 peaks expected and 9 peaks found). HR ESI MS:
calc. for C14H11O5NBr [M + H]+: m/z 351.9815; found: m/z
351.9827 (error 3.4 ppm).
4.9
Isopropyl chelidamate (14)
|
A solution of isopropanol (200 mL, 2.61 mol), sulfuric
acid (0.20 mL, 3.8 mmol), and chelidamic acid[27] (2.48 g,
13.5 mmol) was held at reflux under nitrogen for 36 hours.
Solvent was removed by rotary evaporation, and the crude
material was dissolved in chloroform, washed with 5%
Na2CO3 (×1), 2% NaHCO3 (×4), water (×2), saturated NaCl
(×1), and dried over sodium sulfate. Filtration and removal of
the solvent provided the desired product as a colorless solid:
4.12
4-(p-Bromobenzyloxy)pyridine-2,6-
|
dicarboxylic acid chloride (17)
4-(p-Bromobenzyloxy)pyridine-2,6-dicarboxylic acid (16,
0.5177 g, 1.470 mmol), thionyl chloride (5 mL, 0.07 mol)
and DMF (1 drop) were held at reflux for 4 hours with
magnetic stirring. Excess thionyl chloride was removed
by distillation to provide 0.5719 g (100%) of an off-white
solid with a yellow tint, which was used directly without
characterization.
3.11 g (86%), mp 128.7-129.8°C; lit. mp 145-146°C.[31] 1
H
NMR (500 MHz, CDCl3) δ 7.11 (s, 2H), 5.33-5.24 (m, 2H),
1.40 (d, J = 6 Hz, 12H). 13C NMR (126 MHz, CDCl3) δ
181.06 (s), 161.07 (s), 136.91 (s), 120.55 (s), 72.05 (s), 21.73
(s) (6 peaks expected and 6 peaks found). HR-MS: calc. for
C13H18O5N [M + H]+: m/z 268.1179; found: m/z 268.1193
(error 5.2 ppm).
4.13
Dibenzo-30-crown-10-based 4-(p-
|
bromobenzyloxy)pyridyl cryptand (18)
cis-Dibenzo-30-crown-10 diol[6] (2, 0.72256 g, 1.211 mmol)
and 4-(p-bromobenzyloxy)pyridine-2,6-dicarboxylic acid
chloride (17, 0.4719 g, 1.213 mmol) were freshly prepared
and dissolved separately in 50.0 mL of freshly distilled
DCM. The two solutions were loaded into plastic syringes
with HPLC tubing sealed to the syringe. Additions from the
syringes were made to a stirred solution of pyridine (3.0 mL,
37 mmol) in DCM (2.7 L, freshly distilled) via syringe
pump at 1.5 mL/h. After additions were complete, the
reaction mixture was allowed to stir for 3 days, after which
time solvent was removed by rotary evaporation, and the
resulting solid was dissolved in chloroform. The crude mix-
ture was washed with 1 mol L−1 HCl (×1) and water (×3).
Solvent was removed by rotary evaporation. The result-
ing residue was purified by column chromatography using
neutral alumina, eluting with chloroform:methanol (99:1);
4.10
Diisopropyl 4-(p-bromobenzyloxy)
|
pyridine-2,6-dicarboxylate (15)
A mixture of acetone (125 mL), isopropyl chelidamate
(14, 4.30 g, 16.1 mmol), p-bromobenzyl bromide (4.54 g,
18.2 mmol), and potassium carbonate (3.59 g, 26.0 mmol)
was held at reflux under nitrogen for 8 hours, filtered
through Celite p545®, and the solvent was removed by ro-
tary evaporation. The crude material was triturated with
hexanes, and the product was collected as a colorless solid:
6.14 g (87%), mp 123.5-124.6°C. 1H NMR (500 MHz,
CDCl3) δ 7.80 (s, 2H), 7.55 (d, J = 8 Hz, 2H), 7.32 (d,
J = 8 Hz, 2H), 5.29 (hept, J = 6 Hz, 2H), 5.17 (s, 2H), 1.43
(d, J = 6 Hz, 12H). 13C NMR (126 MHz, CDCl3) δ 166.23
(s), 164.06 (s), 150.72 (s), 133.84 (s), 132.04 (s), 129.36
(s), 122.80 (s), 114.27 (s), 70.23 (s), 69.91 (s), 21.81 (s)
(11 peaks expected and 11 peaks found); HR ESI MS: calc.
for C20H23O5NBr [M + H]+: m/z 436.0754; found: m/z
436.0760 (error 1 ppm).
1
0.5082 g (46%) of colorless solid, mp 188.8-192.7°C. H
NMR (400 MHz, CDCl3) δ 7.88 (s, 2H), 7.55 (d, J = 8 Hz,
2H), 7.32 (d, J = 8 Hz, 2H), 6.94-6.92 (m, 4H), 6.77 (d,
J = 9 Hz, 2H), 5.30 (s, 4H), 5.18 (s, 2H), 4.20-4.11 (m,
4H), 4.03-3.97 (m, 4H), 3.95-3.90 (m, 4H), 3.84-3.78 (m,
4H), 3.74 (m, 4H), 3.72-3.67 (m, 8H), 3.64 (m, 4H). 13C
NMR (126 MHz, CDCl3) δ 166.43 (s), 164.88 (s), 150.24
(s), 149.12 (s), 149.00 (s), 133.85 (s), 132.14 (s), 129.37 (s),
128.19 (s), 122.87 (s), 121.80 (s), 114.80 (s), 114.34 (s),
114.02 (s), 71.10 (s), 70.86 (s), 70.83 (s), 70.73 (s), 70.03
(s), 69.79 (s), 69.55 (s), 69.09 (s), 68.90 (s), 67.83 (s) (24
peaks expected and 24 peaks found). High res ESI MS: calc.
for C44H51NO15Br [M + H]+1: m/z 912.2437; found: m/z
912.2452 (error 1.6 ppm).
4.11
4-(p-Bromobenzyloxy)pyridine-2,6-
|
dicarboxylic acid (16)
A solution of diisopropyl 4-(p-bromobenzyloxy)pyridine-
2,6-dicarboxylate (15, 2.65 g, 6.07 mmol), 10% wt. KOH
(50 mL), and THF (50 mL) was stirred for 12 hour at room
temperature. THF was removed by rotary evaporation. The
remaining aqueous solution was carefully acidified to pH 1,

10 of 10
PRICE Et al.
|
ACKNOWLEDGMENTS
[14] T. L. Price Jr., H. R. Wessels, C. Slebodnick, H. W. Gibson, J.
Org. Chem. 2017, 82, 8117.
We acknowledge and are thankful for support from the
National Science Foundation (DMR 0704076, CHE-
1106899, CHE-1507553).
[15] H. W. Gibson, D. S. Nagvekar, Can. J. Chem. 1997, 75, 1375.
[16] T. L. Greaves, C. J. Drummond, Chem. Soc. Rev. 2013, 42, 1096.
[17] T. L. Greaves, C. J. Drummond, Chem. Rev. 2015, 115, 11379.
[18] H. W. Gibson, H. Wang, C. Slebodnick, J. Merola, W. S. Kassel,
A. L. Rheingold, J. Org. Chem. 2007, 72, 3381.
ORCID
[19] X. Yan, P. Wei, M. Zhang, X. Chi, J. Liu, F. Huang, Org. Lett.
2011, 13, 6370.
Harry W. Gibson
http://orcid.org/0000-0001-9178-6691
[20] C. Piguet, Dalton Trans. 2011, 40, 8059.
[21] P. Gilli, V. Ferretti, G. Gilli, P. A. Borea, J. Phys. Chem. 1994, 98,
1515.
[22] A. Pan, T. Biswas, A. K. Rakshit, S. P. Moulik, J. Phys. Chem. B
2015, 119, 15876.
REFERENCES
[1] a) L. Brunsveld, B. J. B. Folmer, E. W. Meijer, R. P. Sijbesma,
Chem. Rev. 2001, 101, p. 4071; b) F. Huang, D. S. Nagvekar, X.
Zhou, H. W. Gibson, Macromolecules 2007, 40, 3561.
[2] C. He, Z. Shi, Q. Zhou, S. Li, N. Li, F. Huang, J. Org. Chem.
2008, 73, 5872.
[23] F. L. Lambert, H. S. Leff, J. Chem. Educ. 2009, 86, p. 94.
[24] E. B. Starikov, B. Nordén, Chem. Phys. Lett. 2012, 538, 118.
[25] B. Tang, H.-M. Yang, W.-J. Hu, M.-L. Ma, Y. A. Liu, J.-S. Li, B.
Jiang, K. Wen, Eur. J. Org. Chem. 2014, 2014, 6925.
[26] J.-F. Xu, Z. Huang, L. Chen, B. Qin, Q. Song, Z. Wang, X. Zhang,
ACS Macro Lett. 2015, 4, 1410.
[3] B. L. Allwood, H. Shahriari-Zavareh, J. F. Stoddart, D. J.
Williams, J. Chem. Soc. Chem. Comm. 1987, 1058.
[27] G. Howáth, C. Rusa, Z. Köntös, J. Gerencsér, P. Huszthy, Synth.
Commun. 1999, 29, 3719.
[4] a) Z. Niu, C. Slebodnick, D. Schoonover, H. Azurmendi, K.
Harich, H. W. Gibson, Org. Lett. 2011, 13, 3992. b) W. S. Bryant,
J. W. Jones, P. E. Mason, I. Guzei, A. L. Rheingold, F. R. Fronczek,
D. S. Nagvekar, H. W. Gibson, Org. Lett. 1999, 1, 1001.
[5] P. R. Ashton, D. Philp, M. V. Reddington, A. M. Z. Slawin, N.
Spencer, J. F. Stoddart, D. J. Williams, J. Chem. Soc. Chem.
Comm. 1991, 1680.
[28] L. Pescatori, A. Arduini, A. Pochini, A. Secchi, C. Massera, F.
Ugozzoli, Org. Biomol. Chem. 2009, 7, 3698.
[29] N. Jordão, L. Cabrita, F. Pina, L. C. Branco, Chem. Eur. J. 2014,
20, 3982.
[30] M. Albrecht, M. Yulikov, T. Kohn, G. Jeschke, J. Adams, A.
Schmidt, J. Mater. Chem. 2010, 20, 3025.
[6] A. M. P. Pederson, E. M. Ward, D. V. Schoonover, C. Slebodnick,
H. W. Gibson, J. Org. Chem. 2008, 73, 9094.
[31] J. B. Lamture, Z. H. Zhou, A. S. Kumar, T. G. Wensel, Inorg.
Chem. 1995, 34, 864.
[7] F. Huang, K. A. Switek, L. N. Zakharov, F. R. Fronczek, C.
Slebodnick, M. Lam, J. A. Golen, W. S. Bryant, P. E. Mason,
A. L. Rheingold, M. Ashraf-Khorassani, H. W. Gibson, J. Org.
Chem. 2005, 70, 3231.
SUPPORTING INFORMATION
[8] A. M. P. Pederson, R. C. Vetor, M. A. Rouser, F. Huang, C.
Slebodnick, D. V. Schoonover, H. W. Gibson, J. Org. Chem.
2008, 73, 5570.
Additional Supporting Information may be found online in
the supporting information tab for this article.
[9] F. Huang, H. W. Gibson, W. S. Bryant, D. S. Nagvekar, F. R.
Fronczek, J. Am. Chem. Soc. 2003, 125, 9367.
How to cite this article: Price TL, Slebodnick C,
Gibson HW. Improved complexation of paraquats
with crown ether-based pyridyl cryptands. Heteroatom
Chem. 2017;e21406. https://doi.org/10.1002/hc.21406
[10] M. Zhang, X. Yan, F. Huang, Z. Niu, H. W. Gibson, Acc. Chem.
Res. 2014, 47, 1995.
[11] A. M.-P. Pederson, T. L. Price Jr., C. Slebodnick, D. V.
Schoonover, H. W. Gibson, J. Org. Chem. 2017, 82, 8489.
[12] H. R. Wessels, H. W. Gibson, Tetrahedron 2016, 72, 396.
[13] H. W. Gibson, H. Wang, K. Bonrad, J. W. Jones, C. Slebodnick,
B. Habenicht, P. Lobue, Org. Biomol. Chem. 2005, 3, 2114.