Anion binding of short, flexible aryl triazole oligomers

Article abstract of DOI:10.1021/jo901966f

(Graph Presented) The flexible, electropositive cavity of linear 1,4-diaryl-1,2,3-triazole oligomers provides a suitable host for complexation of various anions. The binding affinities for various combinations of oligomer and anion were determined by

Full text of DOI:10.1021/jo901966f

pubs.acs.org/joc
Anion Binding of Short, Flexible Aryl Triazole Oligomers
ꢀ
Hemraj Juwarker, Jeremy M. Lenhardt, Jose C. Castillo, Emily Zhao, Sibi Krishnamurthy,
Ryan M. Jamiolkowski, Ki-Hyon Kim, and Stephen L. Craig*
Department of Chemistry, Duke University, 124 Science Drive, Durham, North Carolina 27708
stephen.craig@duke.edu
Received September 19, 2009
The flexible, electropositive cavity of linear 1,4-diaryl-1,2,3-triazole oligomers provides a suitable
host for complexation of various anions. The binding affinities for various combinations of oligomer
and anion were determined by 1H NMR titrations. Effective ionic radius is found to be a primary
determinant of the relative binding interactions of various guests, with small but measurable
deviations in the case of nonspherical anions. Solvent effects are significant, and the strength of
the binding interaction is found to depend directly on the donor ability of the solvent. A picture
emerges in which anion binding can be effectively interpreted in terms of a competition between two
solvation spheres: one provided by the solvent and a second dominated by a folded cavity lined with
electropositive 1,2,3-triazole CH protons. Implications for rigid macrocycles and other multivalent
hosts are discussed.
Introduction
in complementary interactions such as cation-π,1 π-π,2
nitrogen-halogen,3 donor-σ-acceptor,4 and hydrogen
bonds derived from C-H groups.5 While not traditionally
included in lists of potential hydrogen bond donors, suffi-
ciently polar C-H bonds interact with both anionic and
neutral heteroatoms in a manner that places them on the
weaker but still useful end of a continuum of hydrogen-bond
donors interacting with electron-rich partners.6
Recent reports demonstrate that the polarity of neutral 1,4-
disubstituted aryl-1,2,3-triazoles (dipole moment ∼5D) andthat
of the C5-H bond create an electropositive site that can function
as an effective hydrogen bond donor for anion binding.7-9
The collective manipulation of individually weak inter-
molecular interactions is central to a wide range of molecular
phenomena, including ligand-receptor interactions in bio-
logy, medicine, and sensors, polymer mechanical and trans-
port properties, and the intra- and intermolecular contacts
that guide the secondary structure of natural and synthetic
macromolecules. As a result of its strength and direction-
ality, hydrogen bonding from conventional donors (H-X,
where X=N, O, or F) is a well-recognized and dominant
interaction in many contexts, but it is increasingly clear that
other chemical moieties hold promise as important partners
(5) (a) Chabinyc, M. L.; Brauman, J. I. J. Am. Chem. Soc. 2000, 122,
8739–8745. (b) Castellano, R. K. Curr. Org. Chem. 2004, 8, 845–865.
(6) (a) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond in Structural
Chemistry and Biology; Oxford University Press: New York, 1999.
(7) (a) Li, Y.; Flood, A. H. Angew. Chem., Int. Ed. 2008, 47, 2649–2652.
(b) Li, Y.; Flood, A. H. J. Am. Chem. Soc. 2008, 130, 12111–12122.
(8) Juwarker, H.; Lenhardt, J. M.; Pham, D. M.; Craig, S. L. Angew.
Chem., Int. Ed. 2008, 47, 3740–3743.
(1) Ma, J. C.; Dougherty, D. A. Chem. Rev. 1997, 97, 1303–1324.
(2) (a) Meyer, E. A.; Castellano, R. K.; Diederich, F. Angew. Chem., Int.
Ed. 2003, 42, 1210–1250. (b) Paulini, R.; Muller, K.; Diederich, F. Angew.
Chem., Int. Ed. 2005, 44, 1788–1805.
(3) Walsh, P. L.; Ma, S. H.; Obst, U.; Rebek, J. J. Am. Chem. Soc. 1999,
121, 7973–7974.
(4) (a) Li, H. F.; Homan, E. A.; Lampkins, A. J.; Ghiviriga, I.; Castellano,
R. K. Org. Lett. 2005, 7, 443–446. (b) Lampkins, A. J.; Abdul-Rahim, O.; Li,
H. F.; Castellano, R. K. Org. Lett. 2005, 7, 4471–4474.
(9) Meudtner, R. M.; Hecht, S. Angew. Chem., Int. Ed. 2008, 47, 4926–
4930.
8924 J. Org. Chem. 2009, 74, 8924–8934
Published on Web 11/03/2009
DOI: 10.1021/jo901966f
r
2009 American Chemical Society

Juwarker et al.
JOCFeatured Article
Li and Flood synthesized a tetrameric aryl-1,2,3-triazole
macrocycle that demonstrated a high binding affinity (K=
105 M-1 in CD2Cl2) for chloride. A noteworthy feature of
the macrocyclic receptor is that it is devoid of conventional
H-X hydrogen-bond donors but rather interacts exclusively
via C-H chloride contacts.7 Juwarker et al. demonstrated
that when the same aryl triazole functionality is presented in
the form of a flexible oligomer, chloride binding induces a
pro-helical conformation in the oligomer, a folding pattern
that creates an electropositive cavity that is similar to but
lacks the preorganization of the macrocycle reported by Li
and Flood. The work by Juwarker et al. further demon-
strated that the strength of the interaction increases with the
generation of triazole-containing oligomer.8 Also around
the same time, Hecht et al. reported on the helicity inversion
of a pyridyl 1,2,3-triazole oligomer induced by binding to
achiral halides in highly polar solvents.9
The practical utility of 1,4-disubstituted-1,2,3-triazoles as
functional species in intra- and intermolecular interactions is
enhanced by the fact that they are readily accessible through
the Cu(I)-catalyzed coupling of azides and alkynes.10 While
triazoles are historically viewed as “stealth” linkages with
negligible independent function, these recent reports of
anion recognition11-14 build on a growing body of work
regarding the potential functionality of substituted triazoles.
For example, the size and dipole moment of 1,2,3-triazoles
make them interesting candidates for amide bond surro-
gates in both a medicinal and structural context.15 Arora and
co-workers have reported the contributions of triazoles to
the conformational preferences of mixed amide-triazole
oligomers.16
FIGURE 1. Oligo(aryl-1,2,3-triazoles) 1 and 2 depicted in their
inferred anion binding conformations. Cavity binding triazole
protons (Hc, Hh) and aryl protons (Ha, Hd) are labeled.
Given the utility of triazole CH-anion interactions in
these and other contexts, we therefore set out to establish the
structure-activity relationships that guide these interactions
in more detail. This manuscript extends our earlier report to
the interactions of short aryl-1,2,3-triazole oligomers with a
range of anionic guests. The use of flexible hosts of various
lengths provides an opportunity to evaluate the intrinsic
properties of triazole-anion interactions, thus providing a
baseline from which to evaluate the effects of size and shape
complementarity found in increasingly ordered receptors.
The size of the anions, as described by their effective ionic
radii, are found to be primary determinants of the strength of
binding by the flexible oligomers, with small but measurable
deviations in the case of nonspherical anions. Further, the
affinity of the receptors for a given anion is typically well
correlated with the downfield shifts of the 1,2,3-triazole CH
protons upon binding. This correlation provides a useful
method for deconvolving the contributions to multivalent
binding in the longer oligomers. An unexpected fluoride-
catalyzed proton exchange reaction is observed in d6-ace-
tone. Finally, solvent effects on the binding of chloride are
presented, and the CH-chloride interaction is found to
depend directly on the donor ability of the solvent. A picture
emerges in which anion binding can be effectively interpreted
in terms of a competition between two solvation spheres: one
provided by the solvent and a second dominated by a folded
cavity lined with 1,2,3-triazole CH protons.
(10) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B.
Angew. Chem., Int. Ed. 2002, 41, 2596–2599.
(11) For selected recent reports on hydrogen-bonding-based anion re-
ceptors, see: (a) Maeda, H. Eur. J. Org. Chem. 2007, 32, 5313–5325. (b)
Amendola, V.; Esteban-Gomez, D.; Fabbrizzi, L.; Licchelli, M. Acc. Chem.
Res. 2006, 39, 343–353. (c) Choi, K. H.; Hamilton, A. D. Coord. Chem. Rev.
2003, 240, 167–189. (d) Bondy, C. R.; Loeb, S. J. Coord. Chem. Rev. 2003,
240, 77–99. (e) Sessler, J. L.; Camiolo, S.; Gale, P. A. Coord. Chem. Rev. 2003,
240, 17–55. (f) Beer, P. D.; Gale, P. A. Angew. Chem., Int. Ed. 2001, 40, 486–
516. (g) Suk, J. M.; Jeong, K. S. J. Am. Chem. Soc. 2008, 130, 11868–11869.
(12) For selected recent reports on cationic anion receptors, see: (a)
Llinares, J. M.; Powell, D.; Bowman-James, K. Coord. Chem. Rev. 2003, 240,
57–75. (b) Beer, P. D.; Hayes, E. J. Coord. Chem. Rev. 2003, 240, 167–189. (c)
Beer, P. D. Acc. Chem. Res. 1998, 31, 71–80. (d) Ihm, H.; Yun, S.; Kim, H. G.;
Kim, J. K.; Kim, K. S. Org. Lett. 2002, 4, 2897–2900. (e) Mullen, K. M.;
Mercurio, J.; Serpell, C. J.; Beer, P. D. Angew. Chem., Int. Ed. 2009, 48, 4781–
4784.
(13) For selected recent reports of CH-anion interactions, see: (a)
Bryantsev, V. S.; Hay, B. P. J. Am. Chem. Soc. 2005, 127, 8282–8283. (b)
Bryantsev, V. S.; Hay, B. P. Org. Lett. 2005, 7, 5031–5034. (c) Fujimoto, C.;
Kusunose, Y.; Maeda, H. J. Org. Chem. 2006, 71, 2389–94. (d) Vega Iel, D.;
Gale, P. A.; Light, M. E.; Loeb, S. J. Chem. Commun. 2005, 39, 4913–5. (e)
Belcher, W. J.; Fabre, M.; Farhan, T.; Steed, J. W. Org. Biomol. Chem. 2006,
4, 781–786. (f) Ilioudis, C. A.; Tocher, D. A.; Steed, J. W. J. Am. Chem. Soc.
2004, 126, 12395–12402. (g) Zhu, S. S.; Staats, H.; Brandhorst, K.;
Grunenberg, J.; Gruppi, F.; Dalcanale, E.; Luetzen, A.; Rissanen, K.;
Schalley, C. A. Angew. Chem., Int. Ed. 2008, 47, 788–792.
Results and Discussion
When tetrabutylammonium salts of various anions are
added to d6-acetone solutions of 1a and 2 (Figure 1), down-
1
field shifts of the H NMR resonances of the 1,2,3-triazole
CH protons and inner cavity aryl protons are observed,
indicating a polarizing interaction in the oligomer cavity
that we have previously attributed to anion binding.8 Speci-
fically for oligomer 1a, addition of anion induces downfield
shifts of triazole protons Hc and inner cavity aryl protons
Ha. Similarly, for oligomer 2, titration with anion induces
downfield shifts of triazole protons Hc, Hh and inner cavity
aryl proton Ha, suggesting in all cases a binding mode similar
to that established previously for chloride and fluoride
(Figure 2).
(14) For a recent review of interactions between electron-deficient aro-
matic rings and anions, see: Berryman, O. B.; Johnson, D. W. Chem.
Commun. 2009, 3143–3153.
(15) (a) Kolb, H. C.; Sharpless, K. B. Drug Discovery Today 2003, 8,
1128–1137. (b) Bourne, Y.; Kolb, H. C.; Radic, Z.; Sharpless, K. B.; Taylor,
P.; Marchot, P. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 1449–1454. (c) Brik,
A.; Alexandratos, J.; Lin, Y. C.; Elder, J. H.; Olson, A. J.; Wlodawer, A.;
Goodsell, D. S.; Wong, C. H. ChemBioChem. 2005, 6, 1167–1169. (d) Bock,
V. D.; Speijer, D.; Hiemstra, H.; van Maarseveen, J. H. Org. Biomol. Chem.
2007, 5, 971–975. (e) Pokorski, J. K.; Jenkins, L. M. M.; Feng, H. Q.; Durell,
S. R.; Bai, Y. W.; Appella, D. H. Org. Lett. 2007, 9, 2381–2383.
(16) (a) Angelo, N. G.; Arora, P. S. J. Am. Chem. Soc. 2005, 127, 17134–
17135. (b) Angelo, N. G.; Arora, P. S. J. Org. Chem. 2007, 72, 7963–7967.
J. Org. Chem. Vol. 74, No. 23, 2009 8925

Juwarker et al.
JOCFeatured Article
FIGURE 2. Partial 1H NMR spectra (400 MHz, d6-acetone, 298 K) of oligo(aryl-1,2,3-triazoles) with and without Bu4NþCl-: (i) 1 mM 1a, (ii)
1 mM 1a þ 1 mM Bu4NþCl-, (iii) 1 mM 2, and (iv) 1 mM 2 þ 1 mM Bu4NþCl-. Downfield shifts of binding protons are denoted by dashed
lines. Similar effects of varying magnitude are observed for all of the ions reported.
TABLE 1. Anion Binding Constants of Oligo(aryl-1,2,3-triazoles) 1 X- and 2 X- Obtained from 1H NMR Titration (1 mM, d6-acetone, 298 K) with
3
3
(Bu)4NþX-a
anion (X-)
K (1a X-) (M-1
)
Δδmax (ppm)
K (2 X-) (M-1
)
Δδ max(i) (ppm)
Δδ max(o) (ppm)
3
3
Cl-
Br-
I-
PhCO2
HSO4
1260 (30)
470 (30)
43 (1)
1150 (70)
160 (30)
35 (3)
2.03 (0.02)
1.45 (0.05)
1.23 (0.02)
1.77 (0.02)
0.61 (0.03)
0.64 (0.02)
0.08 (0.02)
1.2 (0.4) ꢀ 104
1.1 (0.2) ꢀ 104
1.4 (0.2) ꢀ 103
4.3 (0.2) ꢀ 103
2.7 (0.6) ꢀ 103
910 (50)
1.63 (0.03)
1.52 (0.01)
1.21 (0.01)
1.34 (0.06)
0.67 (0.01)
0.63 (0.01)
0.30 (0.07)
1.08 (0.03)
1.11 (0.01)
0.91 (0.01)
0.76 (0.04)
0.67 (0.02)
0.60 (0.01)
0.32 (0.07)
-
-
-
NO3
PF6
-
1.7 (0.6)
4 (2)
aΔδmax (ppm) = calculated maximum change (Δ) in chemical shift (δ) of 1,2,3-triazole proton Hc upon anion binding. Subscripts denote inner (i) and
outer (o) 1,2,3-triazole protons for oligomer 2 (Hc and Hh). The titration data were fit by Benesi-Hildebrand, Scatchard, and nonlinear regression
methods.18 All three methods gave similar results, and the individual fits are provided in the Supporting Information. For convenience and clarity, the
average binding constants and variation due to choice of method (in parentheses) are presented here. None of the conclusions of this work rely on the
choice of fitting method. Absolute uncertainties arise from uncertainties in the concentrations and the possibility of subtle aggregation and/or dielectric
effects on chemical shift as a function of salt concentration. These absolute uncertainties are estimated to be less than 50% in K and less than 5% in
Δδmax. Binding constants reported for oligomer 2 are the average of binding constants derived from fits of both the inner and outer triazole protons.
In all cases, a 1:1 binding stoichiometry is determined by
Job’s Method of Continuous Variation.17 The anion-binding
strength of oligo(aryl-1,2,3-triazoles) 1a and 2 were deter-
mined in d6-acetone from titrations with seven different
anions of varying size, geometries, and basicities (Table 1).
Trends in the binding affinities of each oligomer are dis-
cussed sequentially, below, followed by comparisons be-
tween the two.
from geometric complementarity between host and guest. In
the case of benzoate in particular, the distribution of negative
charge is not isotropic but is concentrated in the plane of the
carboxylate group and away from the formally uncharged
phenyl group (Figure 5). Etter has pointed out that this
electrostatic distribution favors anti hydrogen bonding pat-
terns that are well met in this case by host 1a (Figure 5),19 a
feature that has been exploited in oxoanion receptors pre-
viously20 and which reinforces the hypothesis of a geometric
origin to the deviation in Figure 4. The concentrated charge
in benzoate relative to the spherical halides is also related to
its high β value (2.01), where β reflects the hydrogen-bonding
proton acceptor ability relative to hydrogen ion as obtained
from Kamlet-Taft parameters.21 We do find a rather weak
correlation between binding affinity and β (data not shown),
although we note that β and ionic radius will share an
intrinsic correlation that complicates an efficient separation
of the two effects. Couching the apparent intrinsic selectivity
of 1a for benzoate in terms of geometric complementarity
seems, therefore, to be more appropriate.
Oligo(aryl-1,2,3-triazole) 1a X-. Oligomer 1a displays a
3
1:1 binding stoichiometry with all of the studied anions
(Figure 3). The relative anion binding strengths of 1a in d6-
acetone (Table 1) are
Cl- ∼ C6H5CO2 > Br- > HSO4 > I- ≈ NO3 - > PF6
-
-
-
It is evident that as the size of the anion increases, the
binding strength decreases, a correlation that is consistent
with the expected electrostatic nature of the ion binding
interaction (Figure 4). Small deviations are observed, how-
ever, in the case of benzoate (a tighter binder than expected
from its size) and nitrate (a weaker binder than expected
from its size). The magnitude of these deviations is rather
small (∼1 kcal mol-1), but they indicate a subtle contribution
Li and Flood have reported halide binding studies of a
related acyclic oligo(aryl-1,2,3-triazole) that displays size
€
(19) Gorbitz, C. H.; Etter, M. C. J. Am. Chem. Soc. 1992, 114, 627–631.
(20) (a) Smith, P. J.; Reddington, M. V.; Wilcox, C. S. Tetrahedron Lett.
1992, 41, 6085–6088. (b) Fan, E.; van Arman, S. A.; Kincaid, S.; Hamilton,
A. D. J. Am. Chem. Soc. 1993, 115, 369–370.
(17) Job, P. Ann. Chim. 1928, 9, 113–203.
(18) Connors, K. A., Binding Constants; John Wiley & Sons: New York,
1987; pp 128-136.
(21) Marcus, Y. Ion Properties; Marcel Dekker: New York, 1997.
8926 J. Org. Chem. Vol. 74, No. 23, 2009

Juwarker et al.
JOCFeatured Article
FIGURE 5. Electrostatic potential maps (Spartan) of benzoate
(left) and an aryl triazole trimer (right) showing the complementa-
rity of the anisotropic electrostatic distributions, each of which is
concentrated in the plane. Red denotes regions of negative charge,
and blue denotes regions of positive charge.
selectivity for the larger Br- and I- over Cl- and F- in
CD2Cl2, a trend that differs from that found here for 1a in d6-
acetone as a solvent.7b The selectivity observed in Li and
Flood’s linear oligomers further differs from that reported
by the same authors for an aryl-1,2,3-triazole macrocycle in
which the fixed cavity favors Cl- and Br-. Together, these
differences indicate that anion selectivity is not an intrinsic
property of aryl-1,2,3-triazole functional groups but rather a
confluence of geometric and solvation effects, which we
explore further below.
FIGURE 3. Job’s plots of 1a X- displaying 1:1 binding stoichio-
3
metries with the Bu4Nþ salts of Cl-, Br-, I-, and HSO4 (d6-
-
-
acetone, 298 K). PhCO2-, NO3-, and PF6 also display similar
binding stoichiometries and are omitted for clarity but included in
the Supporting Information. Gaussian fit of data points was
performed using Origin 7.0. Inner triazole proton Hc displays a
1:1 binding stoichiometry with all measured X-. Δδ (ppm) =
Differences in anion binding strengths reflect differences
between electrostatic interactions (ion-dipole, ion-induced
dipole) and geometric distortions of the host-guest sys-
tem.22 Electrostatic effects are reflected by the magnitude
of the induced downfield shifts in the CH proton resonances
of the fully bound hosts (Δδmax), and we find for 1a a
reasonable correlation between Δδmax and the anion binding
constant K (Figure 6). Chloride, for example, is the best guest
and induces the largest difference in CH chemical shift
between the fully complexed and uncomplexed species
1
change in H chemical shift of monitored protons Hc. X1 = mole
fraction of oligomer 1a. Xanion = mole fraction of anion during data
collection.
-
(Δδmax=2.03 ppm) compared to the weak binding of PF6
(Δδmax=0.08 ppm), with the other anions in between. The
larger, spherical halides Br- and I- bind more weakly than
Cl- and have lower induced chemical shifts (Δδmax of 1.45
and 1.23 ppm, respectively). This correlation between mag-
nitude of chemical shift and binding strength is similar to
that which forms the basis of the Guttman acceptor num-
ber,23 and similar correlations have been reported pre-
viously, for example, in a recent study by Jeong and co-
workers.24 For the triazole oligomers, a respectable linear
free energy relationship is observed (Figure 6). It is of interest
to note that benzoate fits the chemical shift trend quite well
(Δδmax = 1.8 ppm, K = 103 M-1), in comparison to its
deviation with respect to ionic radius (Figure 4). The good
fit of benzoate with respect to induced chemical shift
(Figure 6) further supports that its higher affinity relative
(22) Bandyopadhyay, I.; Raghavachari, K.; Flood, A. H. Chem-
PhysChem 2009, 10, 2535–2540.
(23) (a) Mayer, U.; Gutmann, V.; Gerger, W. Monatsh. Chem. 1975, 106,
1235–1257. (b) Mayer, U. Coord. Chem. Rev. 1976, 21, 159–179. (c) Gutmann,
V. The Donor-Acceptor Approach to Molecular Interactions; Plenum Press:
New York, 1978.
(24) Chang, K.-J.; Moon, D.; Lah, M. S.; Jeong, K.-S. Angew. Chem., Int.
Ed. 2005, 44, 7926–7929.
FIGURE 4. Correlation of anion binding strength to ionic radius
for oligomer 1a. Ionic radii obtained from Goldschmidt and Pauling
crystal data correlations as compiled by Marcus.21 Error bars reflect
the uncertainty associated with choice of binding isotherm.
J. Org. Chem. Vol. 74, No. 23, 2009 8927

Juwarker et al.
JOCFeatured Article
1
FIGURE 6. Correlation of calculated chemical shift change (Δδmax) with anion binding strength (K) of oligomer 1a. Data obtained by H
NMR titration of 1a (1 mM; d6-acetone, 298 K) and calculation of Δδmax performed by nonlinear regression curve fitting of 1H NMR data by
Origin 7.0. Δδmax is the calculated induced chemical shift of triazole proton Hc (ppm) when bound to anion. Error bars denote the uncertainty
associated with the choice of isotherm fitting method.
FIGURE 7. Job’s plots of 2 X- (d6-acetone, 298 K) displaying 1:1 binding stoichiometries with the Bu4Nþ salts of Cl-, Br-, I- and HSO4
.
-
3
Benzoate, NO3- and PF6- also display 1:1 binding but are omitted for clarity and included in the Supporting Information. X2 = mole fraction
of oligomer 2. Xanion = Mole fraction of anion. Δδ is the induced change in 1H NMR chemical shift of triazole protons Hc upon anion binding
(ppm).
to ionic radius is due to stronger electrostatic interactions
between the negative charge of the guest and partial positive
charge distribution in the host, rather than an effect due to,
for example, interactions between the host and the aromatic
ring of the guest. The only apparent exception to the
correlation between binding strength and induced chemical
shift is hydrogen sulfate (K=140 M-1, Δδmax=0.62 ppm),
which has a lower Δδmax and higher K than iodide (K=48
aggregates at higher concentrations. Such aggregation is
evident by 1H NMR through extreme broadening and
eventual coalescence of the aromatic protons. At the lower
concentrations, Job’s plots confirmed a 1:1 binding stoichio-
metry for all anions with oligomer 2 (Figure 7).
1H NMR titration with the corresponding Bu4Nþ salts
and subsequent nonlinear regression curve fitting enabled
the quantitative determination of binding affinities, which
are reported in Table 1. The relative anion binding affinities
of 2 were determined to be
M
-1, Δδmax = 1.16 ppm). We note that weak binding
interactions are potentially more susceptible to systematic
uncertainty, for example, due to minor contributions from
aggregation, and that erroneously low Δδmax will lead to
erroneously high K, exaggerating apparent differences. Even
if both the precision and accuracy of the measured values
were certain, however, the magnitude of the deviations are
sufficiently small (<1.5 kcal mol-1 per ion from the values
expected from the remainder of the series) that they defy a
meaningful explanation.
Oligo(aryl-1,2,3-triazole) 2. We next consider the extent to
which these same trends in binding extend to the longer
oligomer 2. The concentration of oligomer 2 was kept at 1
mM in order to overcome the formation of π-stacked
-
-
Cl > Br- > C6H5CO2- > HSO4 > I- ≈ NO3- > PF6
-
The trend in anion binding observed for 2 is similar to that
observed for 1a, in that larger anions lead to weaker binding.
As seen for 1a, a reasonable correlation between induced
chemical shift (Δδmax) and binding affinity K is found
(Figure 8). The magnitude of the induced chemical shift
provides an interesting insight into the nature of these multi-
valent interactions. When the total chemical shifts (e.g., in
Hc þ Hh for 2 versus Hc only in 1a) are considered, the binding
data versus induced chemical shift for 1a and 2 collapse
reasonably well onto a single master curve (Figure 8, right).
8928 J. Org. Chem. Vol. 74, No. 23, 2009

Juwarker et al.
JOCFeatured Article
FIGURE 8. (Left) Correlation of the sum of the calculated chemical shift changes (Δδmax) of the outer and innter triazole protons upon anion
binding with anion binding strength (K) of oligomer 2. Data obtained by 1H NMR titrations of 2 with Bu4NþX- (1 mM; d6-acetone, 298 K) and
calculation of Δδmax obtained by nonlinear regression curve fitting of 1H NMR data by Origin 7.0. (Right) Correlation of total chemical shift
change of triazole protons upon anion binding with anion binding strength for oliomers 1a and 2. Error bars denote the uncertainty associated
with the choice of isotherm fitting method.
The individual interactions, insofar as they are accurately
reported by the induced chemical shifts, therefore are additive
and can be compared from one oligomer to another. For
example, the Δδmax differences associated with chloride bind-
ing in 1a and 2 (2.03 ppm for 1a versus 1.63 ppm for the inner
triazole proton and 1.08 for the outer triazole proton of 2) can
be used to infer further structural details of multivalent
binding.25 The fact that oligomer 2 has a larger binding
affinity for chloride (K > 104 M-1 for 2 versus K = 1280
M-1 for 1a) is not due to tighter interactions of chloride with
individual triazole CH protons; on the contrary, the indivi-
dual interactions appear to be weaker (the individual Δδmax
values for both the inner and outer triazole CH resonances are
smaller for 2 than for 1a). Instead, the weakening of individual
contacts is overcome by the increased number of triazole CH
donors.
binding mode is not optimal for hydrogen-bonding-type
interactions.19 Even if benzoate were able to maintain a
coplanar binding geometry with host 2, however, the non-
symmetrical distribution of negative charge in the carbo-
xylate would prevent simultaneous CH-anion interactions
between the effectively circularly symmetric host and un-
symmetric guest. This effect of uneven charge distribution is
similar to that observed in trends in the progressive aqueous
solvation of carboxylates. The addition of a first molecule of
water to acetate in vacuo, for example, is more exothermic
than the addition of a single molecule of water to chloride,
although the overall bulk aqueous solvation energy of
acetate is considerably lower than that of chloride.27 One
(and potentially a second) directional, hydrogen-bonding-
type interaction interacts strongly with the concentrated,
directional charge distribution of the carboxylate, but sub-
sequent solvating interactions have significantly less effect
relative to spherical anions of comparable size. Here, the
crescent-shaped host 1a effectively saturates the preferred,
“specific” binding region on benzoate, and increasing the
generation of the host generates rapidly diminishing returns
in binding constant, a picture that is supported by the
significant difference in induced Δδmax values for the inner
(1.31 ppm) and outer (0.74 ppm) triazole CH protons upon
binding to benzoate. Therefore, in terms of geometric com-
plementarity, the isotropic charge distribution on bromide is
a better fit for 2 than is the anisotropic charge distribution on
benzoate, while the reverse is true for 1a. Overall, the effect of
size/shape complementarity on binding in the linear oligo-
mers is much less than that in the rigid macrocycles reported
by Li and Flood, but these results suggest that the onset of
some geometric effects occurs in linear oligomers of compar-
able size, without the need for covalent conformational
restraints.
The major difference in binding trends between the two
oligomers is the relative binding strengths of bromide and
benzoate. In the smaller oligomer 1a, benzoate binds more
tightly than bromide (1150 versus 470 M-1), whereas in the
longer oligomer 2, bromide binds more tightly than benzoate
(1.1 ꢀ 104 versus 4.3 ꢀ 103 M-1). Notably, that difference is
consistent with the differences in Δδmax, which is smaller for
2 benzoate than 1a benzoate, implying a poorer fit of
3
3
benzoate within the pseudocircular cavity of 2. Oligomer 2
is longer than its counterpart, and in order to simultaneously
maximize CH-anion contacts and preserve planarity, it
must wrap around anions to form a cavity that is at least
partially “closed off” by the terminal aryl groups. It can
therefore be envisioned that in a fully “wrapped” oligomer 2,
benzoate must bind perpendicularly to the plane of the
triazoles, as the overall size of benzoate would disallow
planar entry into the binding cavity that is allowed by 1a.
The perpendicular binding of benzoate in the cavity of
calixpyrroles has been reported previously by Sessler,26 and
it is well-known from Etter’s seminal work that such a
Fluoride Binding. Fluoride is a common and important
target for anion binding, and its absence from the previous
discussion is likely conspicuous. Studying fluoride binding in
heterocyclic anion receptors is often problematic because of
(25) Badjic, J. D.; Nelson, A.; Cantrill, S. J.; Turnbull, W. B.; Stoddart, J.
F. Acc. Chem. Res. 2005, 38, 723–732.
(26) (a) Piatek, P. L.; Lynch, V. M.; Sessler, J. L. J. Am. Chem. Soc. 2004,
126, 16073–16076. (b) Sessler, J. L.; Barkey, N. M.; Pantos, G. D.; Lynch, V.
M. New J. Chem. 2007, 31, 646–654.
(27) Kelly, C. P; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B. 2006,
110, 16066–16081.
J. Org. Chem. Vol. 74, No. 23, 2009 8929

Juwarker et al.
JOCFeatured Article
FIGURE 9. Aggregation effects indicated by 1H NMR spectra upon addition of Bu4NþF- to a 1 mM solution of oligomer 2 in d6-acetone.
Addition of >1 equiv of F- results in severe peak broadening.
the inherent basicity of the fluoride anion (β=2.88, com-
pared to 1.67 for Cl-), and such is the case here. The fluoride
anion can form very stable hydrogen bonds, for example, in
polynuclear aggregates such as HF2-. Hydrofluoric acid is a
relatively weak acid (pKa = 3.2), and in urea-based and
pyrrole-based anion receptors, fluoride has been shown to
deprotonate the acidic NH hydrogen bond donor to form
two separate species, L- and HF2-, according to the reac-
tion28
HL þ 2F- f L- þ HF2
-
where HL is the protic fluoride receptor, a role filled here by
the triazoles (vide infra).
FIGURE 10. Effect of high fluoride concentration (55 mM, 11
equiv) on disappearance of the 1H NMR resonance of the triazole
In the case of oligomer 2, the use of Bu4NþF- complicates
quantitative study by inducing aggregation to a much greater
proton of 3 (5 mM) in d6-acetone. The effect is quite extreme, as seen
from the drastic reduction in intensity over time of the sharp 1,2,3-
triazole CH singlet that is originally at δ = 9.15 ppm.
extent than the other salts (Figure 9). Nonetheless, at low
1
concentrations of F- in d6-acetone, a reasonable H-1H
NOESY spectrum has previously been reported that indi-
of F- in d6-acetone reveals an increase of 1 Da in the parent
ion peak (from 456.21 to 457.21), consistent with a single
H/D exchange. The combination of proton transfer and the
apparent requirement of >1 equiv of F- for host deproto-
cates a 2 F- binding mode similar to that of 2 Cl-.8
3
3
Aggregation is less problematic for the shorter oligomer 1a
and diaryltriazole 3, but the addition of more than 1 equiv of
F- leads to deprotonation of the triazole CH protons
(Figures 10 and 11). For example, while there is no sub-
stantial change in the intensity of the triazole CH proton
resonance upon addition of 0.8 equiv of Bu4NþF- to 1a, the
addition of 3 equiv results in a pronounced loss of intensity
over a few hours (Figure 11). A similar effect is noticed upon
the addition of 1.2 equiv; however, the loss of the CH
resonances takes considerably longer. For both 3 and 1a
(Figure 12), the triazole CH proton resonance is not restored
when AgNO3 is added to precipitate F- as the silver salt,
indicating a chemical change in the triazoles.
-
nation supports an HF2 mediated pathway (with possible
contributions from H2OF- through adventitious water)
similar to that observed previously in ureas and calixpyr-
roles.28 We note that similar processes may contribute to the
aggregation of 2, which is more pronounced when >1 equiv
of Bu4NþF- is used (Figure 9). These behaviors present
obvious challenges to quantifying F- binding, resulting in its
omission from our binding studies.
Solvent Effects. Molecular recognition in any solvated
environment depends on the nature of the interactions
between host and guest, but it also involves at least partial
desolvation of both the free host and the free guest in order to
facilitate the host:guest interactions.29 When ionic species
are present, these desolvation penalties can be enormous: in
the case of anionic guests, the desolvation penalty is espe-
cially severe in protic solvents. We anticipated that anion
(de)solvation would be a major factor in the present systems,
and we therefore investigated the solvent dependency of
We attribute the loss of the CH resonance to fluoride-
catalyzed proton/deuteron exchange between the aryl-1,2,3-
triazoles and d6-acetone, a hypothesis supported by repeti-
tion of the F-/AgNO3 sequence in protio acetone, after
which 3 is isolated and returned to deuterated solvent, and
the triazole CH resonance is found intact (Figure 12). In
addition, mass spectrometry of 3 after exposure to >1 equiv
1
chloride recognition by H NMR titration experiments of
1a with (Bu)4NþCl- in d6-acetone, CD2Cl2, CD3CN,
(28) (a) Boiocchi, M.; Del Boca, L.; Gomez, D. E.; Fabbrizzi, L.;
Licchelli, M.; Monzani, E. J. Am. Chem. Soc. 2004, 126, 16507–16514. (b)
Gu, R.; Depraetere, S.; Kotek, J.; Budka, J.; Wagner-Wysiecka, E. W.;
Biernat, J. F.; Dehaen, A. Org. Biomol. Chem. 2005, 3, 2921–2923. (c) Evans,
L. S.; Gale, P. A.; Light, M. E.; Quesada, R. New J. Chem. 2006, 30, 1019–
1025.
(29) (a) Lamb, M. L.; Jorgenson, W. L. Curr. Opin. Chem. Biol. 1997, 1,
449–457. (b) Sessler, J. L.; Gross, D. E.; Cho, W.-S.; Lynch, V. M.;
Schmidtchen, F. P.; Bates, G. W.; Light, M. E.; Gale, P. A. J. Am. Chem.
Soc. 2006, 128, 12281–12288.
8930 J. Org. Chem. Vol. 74, No. 23, 2009

Juwarker et al.
JOCFeatured Article
FIGURE 11. (Left) 1 mM 1a þ 0.8 equiv of Bu4NþF- shows no pronounced decrease in the intensity of the triazole CH proton. (Right) 1 mM
1a þ 3 equiv of Bu4NþF- shows a gradual disappearance of the triazole CH proton resonance.
FIGURE 12. 1H NMR spectra of 3 F- showing that triazole proton disappearance is reliant upon acetone being in the deuterated form. (a) In
3
1
protic acetone solvent, the triazole CH H NMR resonance (He) persists after 24 h with excess fluoride. (b) Upon switching to deuterated
solvent, the triazole CH 1H NMR resonance (He) completely disappears. (c) Excess fluoride in conjunction with deuterated acetone is necessary
to cause the triazole CH 1H NMR resonance to disappear, as 3 in d6-acetone without fluoride yields no change in the 1H NMR resonance of
triazole CH proton He. The slight downfield shift of the CH resonance in (a) relative to (c) is attributed to partial association with residual
nitrate. Mass spectral data (see Supporting Information) support the assignment of deuterated oligomer 3 (right) after fluoride treatment in d6-
acetone.
DMSO, CDCl3, and 1:1 d6-acetone:cosolvent mixtures. As
expected, the choice of solvent exerts a large effect on the
affinity of the aryl-1,2,3-triazole receptor for the anionic
guest (Table 2); the magnitude of the stability constants
varies by over 2 orders of magnitude across the series, and
1a shows the highest chloride binding constant in d6-acetone.
There is no quantitative correlation between binding con-
stant and either dielectric constant and dipole moment, but a
good correlation was found with the Gutmann acceptor
number (AN) of the solvents, which gives a quantitative
measure of the solvent’s ability to accept and/or donate
electron density, for example, through hydrogen-bonding-
type electrostatic interactions.23,30
Table 2 shows that as AN decreases, the anion binding
affinity K increases.32 The trend includes 1:1 solvent mixtures
of acetone with high AN solvents, for which the binding
strength is larger than that of the pure solvents themselves.
The exceptions to the trend are DMSO and the 1:1 acetone/
DMSO cosolvent. We note that DMSO is an excellent
acceptor of localized positive charge (e.g., that of a hydrogen
bond proton donor) and hypothesize that it competes with
chloride for the triazole CH donors. Support for this hypoth-
esis is found in the downfield shift of the triazole CH protons
in DMSO (9.642 ppm) relative to the other solvents
(8.42-9.35 ppm), consistent with the acidic nature of the
C4-H proton.33 Anion complexation, therefore, is dominated
by a competition between the electrostatic interactions bet-
ween solvent and anion versus those between the host and
anion. These trends readily explain the apparent discrepancy
in anion selectivity observed here for host 2 and reported
previously by Li and Flood for a related, acyclic tri-
azolophane.7b In our experiments, conducted in d6-acetone,
(30) For an overview on empirical solvent parameters, see: (a) Reichardt,
C. Solvents and Solvent Effects in Organic Chemistry; Wiley-VCH: Weinheim,
2003. For binary solvent parameters: (b) Marcus, Y. Solvent Mixtures-Proper-
ties and Selective Solvation; Marcel Dekker: New York, 2002. (c) Mancini, P. M.
E.; Terenzani, A.; Gasparri, M. G.; Vottero, L. R. J. Phys. Org. Chem. 1995, 8,
617–625. (d) Schmid, R.; Sapunov, V. N. Non-Formal Kinetics. In Search for
Chemical Reaction Pathways; Verlag Chemie: Weinheim, 1982. (e) Schmid, R.
Solvent Effects on Chemical Reactivity. In Handbook of Solvents; Wypych, G.,
Ed.; ChemTec Publishing: Toronto, 2001; Chapter 13 .
(32) A similar correlation between binding strength of anions and AN is
described in Beer, P. D.; Shade, M. Chem. Commun. 1997, 24, 2377–2378.
(33) Matulis, V. E.; Halauko, Y. S.; Ivashkevich, O. A.; Gaponik, P. N.
THEOCHEM 2009, 909, 19–24.
(31) (a) Kamlet, M. J.; Taft, R. W. J. Am. Chem. Soc. 1976, 98, 377–383.
(b) Kamlet, M. J.; Doherty, R. M.; Abboud, J. L.; Abraham, M. H.; Taft, R.
W. J. Pharm. Sci. 1986, 75, 338–349.
J. Org. Chem. Vol. 74, No. 23, 2009 8931

Juwarker et al.
JOCFeatured Article
TABLE 2. Effect of Solvent on the Chloride Binding Affinity of 1
distribution. There is no evidence from the trends that the
pseudocircular cavity presents a “special” binding arrange-
ment for appropriately sized ions.
acceptor
number (AN)b
solvent
acetone
1:1 acetone/DMSO
1:1 acetone/CD2Cl2
1:1 acetone/CD3CN
DMSO
1:1 acetone/CDCl3
CD2Cl2
CDCl3
K (M-1 a
)
Δδmax (ppm)
These results prompt an interesting question: if a tightly
enforced circular cavity is not particularly advantageous for
anion binding, why are the anion binding affinities of the
macrocyclic receptors reported by Li and Flood up to 5
orders of magnitude stronger than those of the related acyclic
oligomers?7b The entropic benefits of preorganization must
surely play a role, but the number of conformational degrees
of freedom that are lost upon anion binding is rather modest.
Because the orientation of the terminal aryl groups in 2 is
likely to be a minor contributor to binding, there are only six
aryl-triazole single bond torsions that are expected to be
restricted upon anion complexation. Each of these torsions
would be expected to be dominated by two fairly narrow
regions of values centered at 0° and 180° in the unbound
oligomer, and each would be limited to only one of those
ranges upon idealized anion binding. The entropic penalty
associated with that reduced conformational freedom is
seemingly inadequate to explain a difference in binding
affinity of several orders of magnitude between the macro-
cyclic and acyclic triazole oligomers.34 For comparison, the
complete loss of torsional entropy upon binding has been
calculated for similar rotors to be ∼11 eu.34a Even if all six
rotors were completely frozen upon anion binding (which
surely overestimates the actual entropic loss), then the total
cost in free energy of just over 4 kcal mol-1 is not sufficient to
account for the reported differences between the rigid macro-
cycles and acyclic oligomers.
A complementary effect could be that the macrocyclic
receptor preorganizes the triazoles not so that they are well-
positioned to maximize the interactions between the receptor
and the anionic guests; after all, the charge distribution in
most of the ions is not heavier within an equatorial plane.
Rather, the macrocycles might achieve higher affinity by
preorganizing the electropositive C-H end of the triazole
dipoles so that they create repulsive interactions in the un-
bound host (that is, whereas the unbound state of the acyclic
receptor can relax to a lower-energy conformation, the
shape-persistent macrocycle can not) and/or a cavity that
can only be filled at additional entropic and enthalpic
penalty by the surrounding solvent.35 In other words, macro-
cyclization might destabilize the unbound state of the host
rather than stabilizing the bound host:guest complex. This
type of contribution, while often less discussed than con-
tributions due to favorable preorganization, is a well-known
contributor to the macrocyclic effect and similar multivalent
binding processes.36 Support for the importance of destabi-
lization in the unbound host in these systems comes from
several sources. For example, we note that in the crystal
structure of the triazole oligomer 1b (Figure 13), the triazoles
1260 (30)
27 (1)
110 (3)
100 (3)
5.1 (0.4)
61 (6)
2.03 (0.02)
1.48 (0.03)
1.95 (0.01)
2.03 (0.01)
1.34 (0.08)
1.40 (0.08)
2.37 (0.04)
2.6 (0.1)
12.5
17.3
18.7
19.2
19.3
20.3
20.4
23.1
34 (1)
18 (1)
aBinding constants obtained from 1H NMR titration of oligomer 1 (1
mM, 298 K) with Bu4NþCl-. The titration data were fit by Benesi-
Hildebrand, Scatchard, and nonlinear regression methods.18 All three
methods gave similar results, and the individual fits are provided in the
Supporting Information. For convenience and clarity, the average
binding constants and variationdue to choice of method (in parentheses)
are presented here. None of the conclusions of this work rely on the
choice of fitting method. Absolute uncertainties arise from uncertainties
in the concentrations and the possibility of subtle aggregation and/or
dielectric effects on chemical shift as a function of salt concentration.
These absolute uncertainties are estimated to be less than 30% in K and
b
less than 5% in Δδmax. Acceptor numbers for pure solvents obtained
from Gutmann et al.23 For 1:1 solvent mixtures, AN is calculated from
Dimroth-Reichardt parameters ET.30 ET values for 1:1 solvent mixtures
were derived from linear solvation energy relationships of solvatochro-
mic parameters described in detail by Kamlet and Taft.31
chloride is found to be a better guest than bromide, whereas
Li and Flood observed that a similar acyclic receptor bound
bromide more tightly than chloride in CD2Cl2. The change in
anion selectivity can be attributed to differential solvation;
the smaller chloride ion is better solvated than bromide.
When the binding studies are conducted in CD2Cl2 versus d6-
acetone, therefore, the increased donor ability of the CD2Cl2
will stabilize the free chloride more than its bromide counter-
part. The magnitude of the selectivity for chloride over
fluoride in the macrocyclic receptors of Li and Flood is
similarly expected to have a significant solvent dependency,
as solvation energies of fluoride are much larger than those
of chloride.27,30a
Implications for Receptor Design. When considering anion
binding in the acyclic receptors 1a and 2, perhaps the most
significant conclusion is the lack of major surprises. The
identified trends in binding are largely valid across the entire
series of anion guests with only a few deviations, which are
typically of a fairly small magnitude. Anion binding varies
relatively smoothly with ion size, the number and strength of
individual triazole-anion contacts in the host, and the donor
ability of the solvent. Even in cases where deviations are
observed, they are relatively minor (generally ∼1 kcal mol-1
or less). These results suggest a model of anion binding in
which the host-guest interaction can be thought of in terms
of a transfer of anion from the solvation sphere provided by
the solvent to a partial solvation sphere provided by the aryl
triazole host. Any contributions from shape- or size-com-
plementary “molecular recognition” are relatively minor. As
shown in our prior work, for example, the host 2 can wrap
neatly around the chloride ion to form a pseudocircular
cavity for binding that is reminiscent of the macrocycles of
Li and Flood. The ability to do so, however, presents no
great advantage: iodide is incapable of fitting into the planar
binding site, but it (presumably) accesses another host
conformation that provides high-quality CH contacts
and results in a binding affinity only 1 order of magnitude
lower than that of chloride, despite its more diffuse charge
(34) (a) Mammen, M.; Shakhnovich, E. I.; Whitesides, G. M. J. Org.
Chem. 1998, 63, 3168–3175. (b) Schmidtchen, F. P. Coord. Chem. Rev. 2006,
250, 2918–2928. (c) Hunter, C. A. Angew. Chem., Int. Ed. 2004, 43, 5310–
5324. (d) Schneider, H.-J. Angew. Chem., Int. Ed. 2009, 48, 3924–3977.
(35) (a) Zhang, B.; Breslow, R. J. Am. Chem. Soc. 1993, 115, 9353–9354.
(b) Calderone, C. T.; Williams, D. H. J. Am. Chem. Soc. 2001, 123, 6262–
6267.
(36) (a) Mammen, M.; Choi, S.-K.; Whitesides, G. M. Angew. Chem., Int.
Ed. 1998, 37, 2754–2794. (b) Christensen, T.; Gooden, D. M.; Kung, J. E.;
Toone, E. J. J. Am. Chem. Soc. 2003, 125, 7357–7366. (c) Tobey, S. L.;
Anslyn, E. V. J. Am. Chem. Soc. 2003, 125, 10963–10970. (d) Kitov, P. I.;
Bundle, D. R. J. Am. Chem. Soc. 2003, 125, 16271–16284.
8932 J. Org. Chem. Vol. 74, No. 23, 2009

Juwarker et al.
JOCFeatured Article
FIGURE 13. Partial crystal structure of oligomer 1b (oligoethyleneglycol units truncated for clarity; see Supporting Information for complete
structure), showing the alternating orientation of the triazole dipoles.
adopt a zigzag “anti” conformation that minimizes the internal
dipole of the molecule in a way that is obviously inaccessible to
the macrocycle. Arora and co-workers16 have previously noted
similar conformational preferences in oligomeric triazoles, and
calculations on the macrocyclic receptors reveals puckering of
the unbound (but not anion-complexed) planar macrocycle,
presumably to relieve hydrogen-hydrogen repulsions.22 The
conformation of the oligomer without anion is in marked
contrast to the crystal structure of a related oligomer:chloride
complex reported previously,8 in which the triazoles neatly
wrap around the anion partner in the expected geometry.
This qualitative structural evidence is supported by pre-
viously reported calculations from Li and Flood,7b in which
the “anti” arrangement of unrestricted triazoles is favored by
∼2 kcal mol-1 over the “syn” conformation that is preorga-
nized for anion binding. The ability of the unbound acyclic
host to “relax” into more stable, anti conformations might
therefore be expected to contribute up to 5-6 kcal mol-1 of
stability to the unbound state of 2 versus the macrocyclic
host (∼2 kcal mol-1 for each of three sequential triazole
pairs). The actual contribution due to such conformational
relaxations to the free energy of the unbound state might be
either higher or lower than this crude estimate, depending on
entropic penalties and interactions between nonadjacent
triazoles, but it demonstrates that enforced dipolar repulsion
in the unbound macrocycle is capable of making up the
difference between the observed differences in binding con-
stants and those expected on the basis of purely entropic
grounds. Finally, the induced chemical shifts, Δδmax, whose
correlation to the strength of individual interactions has been
discussed above, are greater for the acyclic rather than the
macrocyclic hosts, suggesting that, if anything, the flexibility
of the linear oligomers facilitates rather than inhibits favor-
able CH-anion contacts. While a quantitative breakdown
of the contributions from these competing mechanisms is not
within the scope of this work, this analysis may be useful in
the design of future triazole-based receptors.
interaction is well-correlated with the effective ionic radius of
the anion (smaller anions are bound more tightly), the total
induced chemical shifts of all interacting triazole CH protons
(greater shifts are associated with stronger interactions), and
the Gutmann number of the solvent (better hydrogen bond
donors lead to weaker binding). Increasing the number of
triazoles in the linear oligomers leads to a decrease in the
magnitude of the individual CH-anion contacts, but over-
comes that penalty through the multivalent presentation of
additional hydrogen-bond-type donors. Geometric comple-
mentarity appears to have a modest but measurable effect on
the binding of benzoate, but in general these results suggest a
model in which the host-guest interaction is dominated by
what can be thought of as a partial transfer of anion from one
amorphous solvation sphere provided by the solvent to
another provided by the aryl triazole host. The nonspecific
nature of the binding suggests that favorable contacts pro-
vided by the preorganization of a macrocyclic triazole-con-
taining host are unlikely to account for the increased affinity
observed in those receptors and, together with the magnitude
of the difference in binding constants between the rigid and
flexible hosts, suggests that a significant fraction of the
binding affinity of macrocyclic hosts arises from the desta-
bilization of their unbound state relative to that of the
flexible analogues. Finally, fluoride ion is observed to cata-
lyze proton/deuteron exchange between the host triazoles
and d6-acetone in a process that appears to be mediated by
the formation of HF2-. The proton transfer reactions are
inhibited by the presence of host that apparently sequesters
fluoride ion.
Experimental Section
The synthesis and characterization of aryl-1,2,3-triazole oli-
gomers 1a,b, 2, and 3 have been reported previously.8 1H NMR
titration experiments were carried out with the correspond-
ing tetrabutylammonium salts of bromide, hydrogen sulfate,
iodide, nitrate, benzoate, and hexafluorophosphate.
1H NMR Titrations. Host oligomer concentrations were kept
to 1 mM by weighing out appropriate amounts of host and
dissolving in either 500 μL or 1 mL of deuterated solvent. These
stock solutions were diluted to 1 mM concentrations in 2 mL
samples by pipet. Weighing an appropriate amount of the
respective tetrabutylammonium chloride salt and dissolving in
Conclusion
The 1:1 binding affinities of short, oligomeric aryl-1,2,3-
triazoles for various anions have been investigated in multi-
ple solvents. Over much of the series, the strength of the
J. Org. Chem. Vol. 74, No. 23, 2009 8933

Juwarker et al.
JOCFeatured Article
1 mL of the 1 mM host solution yielded chloride stock solutions.
Subsequent additions of anion from this stock solution enabled
the host oligomer concentration to remain constant throughout
the titration experiments, with the only possibleperturbation of
concentration coming from evaporation of deuterated solvent.
Anion was added sequentially to an initial 600 μL volume of the
host by micropipet, and 1H NMR of the solution was recorded
with chemical shifts of the 1,2,3-triazole CH proton recorded
versus d6-acetone (2.05 ppm).
in 1,2 dichloroethane, which was arbitrarily assigned the value
of 100. This empirical solvent parameter can be mathematically
described by the following relationship:
acceptor number ðANÞ ¼ ½δcorr =δcorrðEt3PO-SbCl5Þꢁ ꢀ 100
-
¼ δcorr ꢀ 2:348
The acceptor numbers are dimensionless numbers describing
the acceptor property of a given solvent relative to those of
SbCl5.
21,23a
Acceptor Number of Binary Solvent Mixtures. The Gutmann
acceptor number (AN) of a solvent gives a quantitative measure
of the solvent’s ability to accept negative charge (for example,
through hydrogen bonding). Acceptor numbers were derived by
Gutmann and co-workers as empirical quantitites for charac-
terizing the electrophilic properties of electron pair acceptor
solvents. These unitless numbers are obtained from the relative
31P NMR chemical shifts produced by the electrophilic actions
of acceptor solvents (A) in triethylphosphine oxide, described
below:
Acknowledgment. H.J. and J.L. contributed equally to
this work. We thank Duke University and NSF (CHE-06-
46670) for support. J.L. was supported by a NSF IGERT
and the Schering-Plough Fellowship from the ACS Divi-
sion of Organic Chemistry, and K.K. was supported by a
Gordon Research Fellowship from Duke University. The
participation of E.Z., J.C., and S.K. was supported through
an introductory research seminar at Duke: CHEM 26S.
NMR facilities were supported by NCBC Grant 2008-IDG-
1010.
The relative 31P NMR chemical shift values δcorr (n-hexane as
reference solvent, arbitrarily assigned 0) are related to those of
the 1:1 complex Et3PO-SbCl5 (δcorr(Et3PO-SbCl5)) dissolved
Supporting Information Available: Additional spectra, ti-
tration data, and binding constant fits. This material is available
free of charge via the Internet at http://pubs.acs.org.
8934 J. Org. Chem. Vol. 74, No. 23, 2009