Anion receptor chemistry: Highlights from 2016

Coordination Chemistry Reviews xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Coordination Chemistry Reviews journal homepage: www.elsevier.com/locate/ccr

Review

Anion receptor chemistry: Highlights from 2016 Philip A. Gale a,⇑, Ethan N.W. Howe a, Xin Wu a, Michael J. Spooner b a b

School of Chemistry, The University of Sydney, NSW 2006, Australia Chemistry, University of Southampton, Southampton SO17 1BJ, UK

a r t i c l e

i n f o

Article history: Received 10 October 2017 Received in revised form 5 February 2018 Accepted 5 February 2018 Available online xxxx

a b s t r a c t This review covers recent advances in anion receptor chemistry from 2016, including developments in self-assembly, sensing, anion separation, transport, catalysis and fundamental advances in anion recognition systems. Ó 2018 Elsevier B.V. All rights reserved.

Dedicated to Professor Paul D. Beer on the occasion of his 60th birthday. Keywords: Anion complexation Anion coordination Anion directed Supramolecular

Contents 1. 2.

3.

4. 5. 6. 7. 8. 9.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fundamental studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Hydrogen-bond based receptors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Halogen/chalcogen-bond based receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Aromatic and hydrophobic anion acceptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Ion-pair receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Photo-switchable anion receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6. Multivalent cationic systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metal–organic structures and anion-directed self-assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Metal–organic structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Anion-directed self-assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transmembrane anion transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Catalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anion separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

Abbreviations: AcO
, acetate; CD, circular dichroism; CyD, cyclodextrin; CPP, cell penetrating peptide; DFT, density functional theory; DMF, dimethylformamide; DMSO, dimethylsulfoxide; EPR, electron paramagnetic resonance; G, guest; H, host; HPLC, high performance liquid chromatography; ITC, isothermal titration calorimetry; Ka, association constant; MD, molecular dynamics; MS, mass spectrometry; NMR, nuclear magnetic resonance; TBA, tetrabutylammonium. ⇑ Corresponding author. E-mail address: [email protected] (P.A. Gale). https://doi.org/10.1016/j.ccr.2018.02.005 0010-8545/Ó 2018 Elsevier B.V. All rights reserved.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

2

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

1. Introduction This article marks the 20th year of coverage of anion receptor chemistry by this series of reviews. In the first review, published in Coordination Chemistry Reviews in 2000 (and covering highlights in anion coordination and anion directed assembly in 1997 and 1998) [1], it was noted that ‘‘Anion coordination has received little attention over the last 30 years when compared to that devoted to the coordination chemistry of cations.” Now as we look back over the last 50 years we see that this has changed and that many real world applications for anion complexation have emerged [2]. Although this review has moved between publications over the last two decades [1,3–11]; – it is appropriate for it return to Coordination Chemistry Reviews to mark both this milestone and the special issue devoted to Coordination Chemistry in Australia. 2. Fundamental studies This section covers the development of new receptors and understanding the fundamentals of anion interactions, ranging from classic hydrogen-bond systems to more exotic halogen and chalcogen bond-based receptors. In 2016, we have seen examples of anion receptors that advance our understanding of aromatic,

Fig. 3. Representation of the solid-state supramolecular network in the pyridine hydrochloride complex of b-HCH 2b [the b-sheet like ribbons (contact in orange and blue) and the 2D structure are shown in detail, while the formation of the 3D aggregate is suggested by the contacts (in red) involving the hydrogen atom at the 4-position of the pyridine rings]. Atom colors: white, H; gray, C; blue, N; green, Cl; yellow, Br. Reproduced with permission from Grosu et al. [13]. Copyright 2016 The Royal Society of Chemistry.

hydrophobic and ion-pair interactions, and the development of photo-switchable receptors and multivalent systems. 2.1. Hydrogen-bond based receptors

Fig. 1. Structure of tripodal CH hydrogen-bonding receptors 1a(PF6)3 and 1b(PF6)3.

Fig. 2. Diastereomers of hexachlorocyclohexane (HCH) 2a–e.

A number of groups have exploited CH hydrogen bonds for the recognition of anions. Amendola and co-workers [12] reported the CH hydrogen bonding receptor 1a(PF6)3 based on a tripodal 1,3,5trimethylbenzene scaffold and incorporating imidazolium and 2,3,4,5-tetrafluorobenzene moieties as CH hydrogen-bond donors (see Fig. 1). Proton NMR titrations in CD3CN and CD3CN/DMSO-d6 9:1 demonstrated that the receptor possessed a strong affinity for chloride with log Ka  6 and in the more competitive CD3CN/ D2O 4:1 binding chloride with a log Ka of 3.4. Analogous experiments using the non-fluorinated analog 1b(PF6)3 showed no shift in the ortho-benzyl protons and a much-reduced log Ka of 2.2. This, and upfield shifts of all signals in 19F NMR titrations of 1a3+ with TBACl in CD3CN due to polarization of the aromatic CAF bonds, revealed the contribution of the CH hydrogen-bonds from the 2,3,4,5-tetrafluorobenzene motifs to anion complexation. Grosu et al. [13] have investigated the anion recognition properties of the diastereomers of hexachlorocyclohexane (HCH) (Fig. 2). For example, in the solid state, b-HCH 2b formed a 1:2 complex with pyridinium chloride and pyridinium bromide (Fig. 3). b-HCH was able to bind an anion on each face of the ring through six CH  anion contacts. The pyridine rings connect to three anions through two CH  X
interactions and NH+  X
contacts, allowing the assembly of a 2D network of supramolecular ribbons. A second structure with pyridinium bromide was also observed (Fig. 4), comprising 20% of crystals obtained. This 1:1:1 complex of b-HCH/PyHBr/Py formed columns of b-HCH Br
complexes, linked into 2D sheets by halogen-bonding between CACl bonds on the ring. CH interactions between the anion and pyridine–pyridinium units sharing a proton link these sheets in the greater 3D structure. Proton NMR titrations however, indicated a 1:1 binding mode in solution with binding constants ranging
1 from 1400 M
1 for HSO
for Cl
(CD3CN). 4 to 2200 M

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

3

Fig. 4. Representation of the solid-state bonding network of b-HCH 2b with pyridine hydrobromide (pyridinium–pyridine units were omitted for clarity). Atom colors: white, H; gray, C; blue, N; green, Cl; yellow, Br. Reproduced with permission from Grosu et al. [13]. Copyright 2016 The Royal Society of Chemistry. Fig. 6. Structures of sodium and chloride ions coordinated to one or two molecules of 3. Atom colors: white, H; gray, C; cyan, F; purple, Na; green, Cl. Reproduced with permission from O’Hagan et al. [14]. Copyright 2016 American Chemical Society.

Fig. 5. Schematic demonstrating the Janus face polarity of all-cis 1,2,3,4,5,6hexafluorocyclohexane 3.

The binding of anions and cations by all-cis 1,2,3,4,5,6-hexafluor ocyclohexane 3 has been studied by O’Hagan and co-workers [14] using electrospray ionization to generate complexes in the gas phase. The large dipole of 3 (6.2 D) conferred a two-sided ‘Janus’ polarity, giving the fluorinated face the ability to coordinate to cations, while the hydrogen face is able to bind anions (Fig. 5). The computed energetics of the formation of 1:1 and 2:1 complexes between 3 and both Cl
and Na+ indicated that the binding enthalpies for the binding of the first and second molecule of 3 were comparatively very large, with the binding to chloride among the strongest in the literature. Only minor changes observed in the bond lengths in the core of the molecule supported the assertion that the interactions were generally electrostatic in origin. Infrared multiple photon dissociation spectra obtained for the generated complexes were in agreement with the calculated spectra, demonstrating characteristic shifts in the vibrational frequencies on binding of the ions. Structures of sodium and chloride coordinated to 3 are shown in Fig. 6. Aromatic CH hydrogen-bonds have been exploited by Kim and co-workers [15] in their octulene-based anion receptor 4 (Fig. 7). The saddle-shaped coronoid ring forms a pore with a diameter of

Fig. 7. Structure of octulene receptor 4, showing eight CH hydrogen-bond donors.

5.5 Å with eight CH hydrogen-bond donors oriented toward the center of the cycle. Proton NMR titration in C6D6 (1% CDCl3) with TBACl revealed an association constant of 2.2  104 M
1, fitted to a 1:1 binding model. DFT studies indicated that the cavity is highly preorganized for the binding of chloride, with models suggesting a distortion energy penalty of just 2 kJ mol
1 on binding to the anion. Sindelar and co-workers have continued their work on CH hydrogen-bond donor anion receptors [16]. In 2016 they reported the synthesis of propanediurea-based molecular clips 5a and 5b (Fig. 8) [17]. NMR titration data and DFT calculations suggested that the compounds were able to weakly bind halides via CH hydrogen-bonding interactions on the convex face of the clips (Fig. 8). Binding affinities for the halides were correlated with the solvation energy of the anions. The authors suggested that similar interactions could affect the synthesis of structurally-related

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

4

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 8. (a) Structure and binding conformation of propanediurea-based molecular clips 5a and 5b. (b) Calculated structure of the Cl
complex of 5b. Non-interacting hydrogen atoms in (b) have been omitted for clarity. Atom colors: white, H; black, C; blue, N; red, O; green, Cl.

Fig. 11. Structure of aryl-triazole receptors 8a–c.

Fig. 9. Structure of cyclo[4]carbazole receptor 6 for I
%.

Charge-assisted CH hydrogen-bonding from 2,4,5trimethylimidazolium groups was exploited by Caballero, Molina and colleagues [19] in their series of receptors 7a–d for oxoanions. meta-Substituted phenyl derivative 7a and 2,7-napthyl derivative
2
7d were able to recognize HP2O3
7 , H2PO4 and SO4 over halides, 1 nitrate and other anions as determined by H NMR titration experiments in 5% CD3OD in CD3CN, while the para-substituted 7b and 2
pyridyl 7c bound only HP2O3
7 and SO4 (Fig. 10). The authors suggested that the additional aromatic CH hydrogen-bond donors in 7a and 7d pointing into the cavity from the central aromatic ring may explain the additional interaction with H2PO
4 ; these protons are not present in 7b and 7c. The larger spacers in 7b and 7d 2 caused an inversion of selectivity for HP2O3
7 over SO4, instead of 2
the higher binding constants for SO4 observed for 7a and 7c. Wang, and co-workers [20] compared the hydrogen-bond donor ability of pyridinium-based receptor 8b with the parent receptor 8a (Fig. 11). Proton NMR titrations in acetone-d6 indicated the formation of strong pyridyl and triazole CH hydrogen bonding interactions with halides, with Ka for the cationic alkylated pyridinium 8b two orders of magnitude higher than neutral 8a. Analogous oligomer 8c, containing additional pyridinium and triazole units, yielded a further order of magnitude improvement in binding affinity over 8b in 6:96 D2O/pyridine-d3 (the solvent mixture was chosen to avoid aggregation of 8c). Upfield shifts of the phenyl protons of 8c indicated p-stacking of the terminal phenyl groups, suggesting induced folding of the compound induced by anions as observed in the calculated structure of the chloride complex and further evidenced by 2D NOESY experiments.

Fig. 10. Structure of charge-assisted 2,4,5-trimethylimidazolium receptors 7a–d for oxoanions.

cucurbiturils, in which anion association with the convex face of the intermediate oligoglycourils templates the curvature of the final product. A further example of a macrocyclic CH hydrogen-bonding based receptor was reported by Huang and co-workers (Fig. 9) [18]. Cyclo [4]carbazole receptor 6 forms an almost planar structure with an internal cavity diameter of 4.00 Å, slightly smaller than I
. UV– Vis experiments in CHCl3 indicated a selective interaction with I
over many other anions including other halides, with a binding constant of 82 ± 9 M
1. Downfield shifts of the internal aromatic protons supported a CH hydrogen-bonding interaction with iodide in CDCl3.

Fig. 12. Structure of fused polynorborane receptors 9a and 9b, plus flexible and rigid guests azelate and 2,6-naphthalate.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

5

Fig. 13. Structure of formylated dipyrromethane phosphate receptors 10a–c.

The steric bulk around an anion binding site has been used by Robson and Pfeffer [21] to achieve selective binding on the basis of guest flexibility. They synthesized fused polynorborane dicarboxylate receptors 9a and 9b, with 9b containing a central 4methoxybenzene blocking unit to disfavor ‘end-to-end’ linear binding. Proton NMR titrations (DMSO-d6) showed that 9a and 9b bind azelate equally strongly (log Ka = 4.3 and 4.5 respectively) (Fig. 12). The binding of 2,6-naphthalate by receptor 9b however (log Ka = 2.8), was an order of magnitude weaker than for 9a (log Ka = 3.7), as the guest is blocked from forming strong linear hydrogen bonds to the binding sites. Thus, 9b demonstrated a 50-fold selectivity for the more flexible azelate over the more rigid guest. Sessler and co-workers [22] have described two new formylated dipyrromethane receptors, extending their series of tetrakisreceptors with ferrocene linked 10c and adding hexakis-receptor 10b (Fig. 13). The anion binding properties of the hexakiscompound 10b were investigated by UV–Vis titration, showing high affinity for dihydrogen phosphate and pyrophosphate with
little interaction with HSO
or NO
4 , Cl 3 , in line with previously reported members of the family. Compound 10b demonstrated a 100-fold improvement in dihydrogen phosphate affinity over previously reported tetrakis-receptor 10a (Ka = 1.4  105, 2.1  103 M
1 respectively, UV–Vis titration, 3% CH3OH in CHCl3). DOSY and single-crystal X-ray diffraction studies provided evidence in support of a conformational switch from a trans- to a cisconformation of the binding moieties was necessary upon binding, with 10b able to adopt a more pre-organized structure that could partially overcome this barrier. The electrochemical properties of ferrocene receptor 10c were studied by cyclic voltammetry (10% DMF in CH3CN, 0.1 M TBAPF6 as supporting electrolyte) with the compound exhibiting an increasingly irreversible oxidation wave with an increasing anodic peak current with increasing concentration of H2PO
4. A number of groups have developed large macrocyclic hosts to enhance anion affinity and selectivity. Squaramides have attracted attention recently as highly effective receptors for anions [23–25]. Jolliffe and co-workers [26] reported a series of macrocyclic squaramide compounds that are potent receptors for sulfate (Fig. 14). Both 11a and 11b bound sulfate with binding constants >104 M
1 (1H NMR titration, 0.5% H2O in DMSO-d6) with 11b showing supe

rior selectivity for sulfate over H2PO
4 , AcO and Cl compared to 11a. Soluble analogs 11c and 11d were synthesized to quantitatively investigate the binding in more competitive media. Compound 11d still exhibited Ka for >104 M
1 (1H NMR titration, 33% H2O in DMSO-d6) for sulfate, which was attributed to the precise match of the binding pocket size to minimize exposure to solvent. Compound 11d also demonstrated remarkable selectivity for SO2
4 over CrO2
exceeding that of naturally occurring sulfate binding 4 2
protein (SBP) (Ka(SO2
4 )/Ka(CrO4 ) > 60 for 11d = 2.5 for SBP). Lee and co-workers [27] reported the synthesis of two new naphthobipyrrole-based macrocyclic receptors 12a and 12b (Fig. 15). UV–Vis titrations showed the compounds have high affin6 ity for tetrahedral anions H2PO
4 (both 12a and 12b, Ka = 1.4  10 and 6.3  106 M
1 respectively, CH3CN) and HSO
(12a only, K = 4 a 3.7  105 M
1, CH3CN) with excellent selectivity over small spher-

Fig. 14. Structure of macrocyclic squaramide receptors 11a–d for sulfate.

ical anions Cl
and Br
(no interaction). Proton NMR experiments in DMSO-d6 showed that binding of H2PO
4 in DMSO-d6 had slow complexation/decomplexation kinetics on the NMR timescale, with evidence consistent with deprotonation of bound H2PO
4 forming complexes with HPO2
4 as has been observed previously [28]. Designing very highly selective anion receptors is still challenging. A macrocyclic bisurea with a 2,20 -binaphthalene 13 was synthesized by Kondo and co-workers [29] using a TBACl-templated methodology to disfavor the formation of oligo-ureas. The crystal structure of the compound showed its high degree of preorganization for the binding of Cl
(Fig. 16). UV–Vis titrations (CH3CN) revealed a binding affinity for Cl
of 1  107 M
1, with a similar affinity for AcO
and strong interactions with other basic anions 6
1 such as F
(4  106 M
1) and H2PO
). The receptor’s 4 (5  10 M selectivity for chloride over basic anions was increased by an order of magnitude in aqueous CH3CN (5% H2O/CH3CN) and it retained its strong recognition properties in this solvent when alkali metal cations replaced TBA cations in titration experiments. Bao and co-workers reported a pair of macrocyclic receptors, 14a containing an imine moiety and its reduced amine analog 14b (Fig. 17) [30]. UV–Vis titration data in DMSO demonstrated strong binding interactions with F
, AcO
, and H2PO
4 , with no measurable interaction with various other anions. The binding constants for 14a and 14b showed a remarkable shift in the selec
tivity for F
over H2PO
4 , increasing from a 7.9:1 preference for F for imine 14a to 78.5:1 for amine 14b. This was attributed to an increased affinity for spherical anions such as F
afforded by the additional flexibility provided by the amine linkage and an increased number of NH donors.

Fig. 15. Structure of macrocyclic naphthobipyrrole receptors 12a and 12b.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

6

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 17. Structures of macrocyclic receptors 14a and 14b.

Fig. 16. Structural formula (a) and crystal structure (b) of 2,20 -binaphthalene-based macrocyclic bisurea 13Cl
complex. Non-interacting hydrogen atoms, solvent molecules and counter-ions in (b) have been omitted for clarity. Atom colors: white, H; black, C; blue, N; red, O; green, Cl.

Gunnlaugsson and co-workers [31] have developed the [2]catenanes 15a–c of 2,6-bis(1,2,3triazol4yl)pyridine (Fig. 18). The interlocked structure is self-templated by complementary CH hydrogen-bond interactions between the triazole and pyridine moieties, before undergoing olefin ring-closing metathesis (RCM) using Hoveyda-Grubbs second-generation catalyst with yields up to 50%. The single crystal X-ray structure of 15a confirmed the topology of a ‘‘Hopf link” self-templated [2]catenane, with CHN hydrogen-bond distances ranging from 3.415 Å to 3.700 Å. Macrocycles 16a–c were also isolated as side products from the RCM reactions, as confirmed by the single crystal X-ray structure of 16a. NMR studies of [2]catenanes 15a–c with the addition of TBA+ salts of Cl
and H2PO
4 resulted in significant downfield shift of both triazole-CH and amide–NH proton resonances; while the same studies with macrocycles 16a–c resulted in a downfield shift of the amide–NH but less a pronounced shift of the triazole–CH. The derived association constants for H2PO
4 indicated the preorganized interlocked [2]catenanes (log b of 15a = 1.74 and 15b = 2.04) exhibited stronger binding affinity compared to the macrocycles (log b of 16a = 1.12 and 16b = 0.87). Flood and co-workers [32] reported the first directly observed example of a dimer of hydrogensulfate in solution, the complex stabilized by self-complementary hydrogen bonding between the anions and encapsulation within a supramolecular cyanostar (17a) receptor (Fig. 19). The crystal structure of the complex of 17a with NH4HSO4 clearly shows the dimer with two 2.51 Å hydrogen bonds between the anions, two cyanostars encapsulating the dimer with twenty CH H-bond interactions and two TBA cations capping the ends of the complex. Proton NMR titration and ESIMS data showed a mixture of complexes present in solution (dichloromethane) including 2:1, 3:2 and 2:2 17a:HSO
4 . A characteristic AOH resonance for the dimer was observed in the NMR spectrum of the complex. The authors proposed that the receptor’s

Fig. 18. Structures of [2]catenanes 15a–c and macrocycles 16a–c, and the single Xray crystal structures of 15a and 16a. Solvent molecules have been omitted for clarity.

ability to exclude solvent from the complex and the weak CH hydrogen-bonding from the encapsulation allow the anion–anion hydrogen-bonds to make a significant contribution to the stabilization of the complex. In further work on the cyanostar class of receptors, Flood and co-workers [33] investigated the dependency of the binding affinity of these molecules on size-matching. To do this, bulky cyanostar 17b was synthesized (Fig. 20), with isopropylphenyl groups to block stacking of the macrocycles, and hence limit the receptor to forming 1:1 complexes with anions. Comparison of binding affinities of a number of anions in aprotic solvent system (1:1 CD3CN: CD2Cl2) demonstrated that the affinity for the anion was best cor-

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

7

Fig. 21. Structure of bis(ethynylaniline) receptors 18a and 18b for HS
.

Fig. 19. Crystal structure showing HSO
2 dimer stabilized within Flood’s cyanostar receptor 17a. Most hydrogen atoms and solvent molecules have been omitted for clarity. Atom colors: white, H; black, C; blue, N; red, O; yellow, P; green, Cl.

related to size-matching with the cyanostar cavity, while values in protic systems (2:3 CD3OD:CD2Cl2) are biased to toward those that interact less strongly with the protic solvent. The first reversible hydrogen-bonding receptor for the HS
anion was reported by Haley and co-workers [34]. Due to the similarity in size between the chloride and sulfide anions, bis (ethynylaniline)-based receptors 18a and 18b for Cl
were investigated as potential receptors for this anion (Fig. 21). NMR titrations with TBASH in 10% DMSO-d6/CD3CN revealed a log Ka of 3.7 and further experiments showed no interactions with H2S or modification of the scaffold by HS
. Pyridyl receptor 18b demonstrated slightly weaker binding (log Ka = 3.0) due to the lack of an additional aryl CH H-bond stabilization. This did however contribute to improved selectivity for HS
over Cl
(6-fold compared to 2.8-fold for 18a), as the pyridyl lone pair is able to accept a weak hydrogen-bond from the HS
anion (instead for the repulsive lone pair interaction with Cl
).

Fig. 20. Structure of cyanostar receptors 17a and 17b.

Granja and co-workers [35] reported an a,c-cyclic peptide 19

with an ability to recognize CO2
3 , NO3 and Cl (Fig. 22). The compound crystallized from CHCl3 with a complex structure including two distinct supramolecular units. The first aggregate was a tetramer of the folded peptides, interacting through b-sheet type hydrogen bonding arrays. Each molecule interacts through its concave face with the second unit, a hexameric aggregate of the peptide, which includes 4CO2
ions sequestered from the CHCl3 drying 3 agent. Each anion forms hydrogen bonding interactions with three separate peptide units. Proton NMR titrations and DOSY experiments suggested the formation of these hexameric aggregates in CDCl3 solution. The authors also demonstrated the peptide’s ability to facilitate receptor concentration dependent Cl
/NO
3 exchange in synthetic EYPC vesicles. Ujaque and co-workers [36] described the synthesis of two new positively charged tetrabenzimidazolium cavitands 20a and 20b and their ability to recognize anions at various sites on the receptors (Fig. 23). Proton NMR binding studies in DMSO-d6 were per

formed with a wide range of anions (Cl
, Br
, I
, ClO
4 , NO3 , CN , hexanoate, benzenesulfonate and p-toluenesulfonate). In the case of compound 20a, while all anions generally caused a downfield shift in the imidazolium proton at the rim of the molecule, spherical anions Cl
and Br
also caused a downfield shift of the phenyl CH protons on the arms of the cavity, with MD simulations suggesting these interactions take place on the side of the molecule and with the anions not entering the cavity. Conversely addition of the aromatic anions benzenesulfonate and p-toluenesulfonate resulted in upfield shifts of the phenylene protons, attributed to the anion locating inside the cavity stabilized by CH  p interactions (again supported by MD simulations). Compound 20b, which showed similar interactions with the phenylene protons, but has no acidic proton on the rim, thus demonstrated remarkable selectivity for the aromatic anions (Ka = 3720 M
1 for benzenesulfonate, >104 M
1 for p-toluenesulfonate) over the other anions tested (only other measurable Ka for Cl
and Br
of 6.9 and 13.2 M
1 respectively). Related resorcinarene structures have been explored by Beyeh et al. [37] who have reported that N-alkyl ammonium resorcinarene salts such as 21 can act as effective receptors for chloride, when forming a salt with large, non-spherical counter ions. These structures are able to bind four spherical halides in a circular array to form a pseudo-cavitand structure as shown in Fig. 24. Proton NMR competition and titration experiments with 214OTf
and a reference bisurea chloride receptor gave indirect evidence that 21 binds up to 4 equivalents of chloride in 10% DMSO-d6 in CDCl3, with very high affinity for the first two equivalents of anion, Ka on the order of 103 M
1 and lower affinity for the 4th equivalent as might be expected due to increasing repulsion between bound chloride anions.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

8

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 22. Structure of a,c-cyclic peptide 19, and its crystal structure showing (a) the packing of tetramers (center, shown in space-filling model) and hexamers (periphery, 2
shown in tube model); (b) and hexamers including four entrapped CO2
3 ions; (c) hydrogen-bond interactions between CO3 and 3 peptide units. Reproduced with permission from Granja et al. [35]. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

Fig. 24. Structure of N-alkyl ammonium resorcinarene salt 214Cl
.

Fig. 23. Structure of tetrabenzimidazole-based cavitands 20a and 20b.

Gregolinski et al. [38] reported a series of macrocycles generated by the condensation of 2,6-diformylpyridine and 1,2diaminocyclopentane. The imine macrocyclic products could be reduced with NaBH4 to their amine analogs which upon protonation formed large folded structures for the binding of multiple anions. Example structures include the 4+4 amine 22a which crystallizes with the inclusion of two chloride anions and the 8+8 product 22b which crystallized with the inclusion of two sulfate anions within the container and a further six sulfates bound to the exterior clefts (Fig. 25). A hyperbranched polymer incorporating isophthalamide anion binding sites was synthesized by Twyman and co-workers [39] (Fig. 26). Two examples were prepared with ratios of isophthalamide moieties to polymer monomers of 1:5 (23a) and 1:20 (23b), and their halide binding ability was studied by 1H NMR in CDCl3. The compounds demonstrated a flip of selectivity from the small hard halides F
and Cl
to larger soft halides Br
and I
compared to a free isophthalamide reference compound. The authors

attributed this to the ability of the larger anions to interact with the polymer’s p-system surrounding the anion binding sites. The secondary structure of aliphatic oligoureas, which fold into polarized 2,5-helices, was exploited by Guichard and co-workers for anion recognition (Fig. 27) [40]. Proton NMR titration of 24a with TBACl (0.5% DMSO-d6 in CD3CN) revealed evidence of strong binding of the anion through NH H-bond donors at the positive pole of the helix. Other signals were hardly modified, suggesting the anion caused little conformational change. Fitting of the change in chemical shift of the two terminal NH donors revealed a Ka of 1700–2900 M
1, while the introduction of a terminal indole group (24c) increased Ka by an order of magnitude. Increasing the chain length from hexamer (24a) to a nonamer (24b) also increased Ka in more competitive DMSO (for solubility reasons) from 89 to 140 M
1, which the authors attributed to either the increase of the helix macropole or possible rigidification of the binding site afforded by the longer chain. Dipyrrolylpyrimidines were investigated as anion binding scaffolds by Maeda and co-workers [41]. Binding of anions requires a switch between the coplanar geometries (determined in the solid state and by 1H NMR) shown in Fig. 28, with inversion of the pyrrole rings such that the NH H-bond donors and CH donor from the central pyrimidine converge to form the binding pocket. The binding affinities for Cl
, Br
and CH3CO
2 (UV–Vis, dichloromethane) were lower than for the analogous dipyrrolylbenzene owing to the increased inversion energy of the pyrroles. Despite the inversion energies being the greatest for 25a > 25b > 25c and

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

9

Fig. 25. Crystal structures of select examples of Gregolinski and Lisowski’s amine macrocycles, 4+4 amine 22a bound to two chloride anions (a) and 8+8 product 22b with the inclusion of two sulfate anions (shown in green).

Fig. 26. Synthesis of a hyperbranched polymers 23a and 23b incorporating isophthalamide binding units.

Fig. 27. (a) Structure of the oligoureas 24a–c and (b) schematic showing the folding into a 2,5-helix.

Fig. 28. Structure of the dipyrrolylpyrimidine anion receptors 25a–c and 26a–c for use in ion-paired assemblies.

26a > 26b > 26c, the binding affinities for all the anions followed the same trend (25a > 25b > 25c and 26a > 26b > 26c), indicating that the electron withdrawing effect of the R2 substituent was a more a significant factor. The authors demonstrated the scaffold’s potential application as a p-electronic anion for the formation of ion pairing assemblies, by demonstrating the formation of columnar assemblies of stacked ion pairs in the crystal structures of 25aTBACl and 25aTATACl (TATA: triazatriangulenium cation). Maeda and co-workers [42] have also reported a series of dipyrrolyldiketone boron complexes as anion binding p-electron systems (Fig. 29a).

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

10

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 29. (a) Structures of dipyrrolyldiketone boron p-electron systems 27, 28a–g and 29a–b, and (b) the single X-ray crystal structures (front and top views) of 28bCl
and 28b2Cl
complexes as TBA salts. The TBA molecules are omitted for clarity. Atom colors: white, H; pink, B; black, C; blue, N; red, O; yellow, F; green, Cl.

Fig. 30. Lysine tetramer peptides 30a–d for DNA binding incorporating guanidinocarbonylpyrrole (GCP) binding units.

UV–Vis titration studies with the addition of Cl
, Br
and CH3as TBA salts, performed in CH2Cl2 at 20 °C, showed enhanced binding of CH3CO
2 by the arylethynyl-substituted 28a (Ka 150  103 M
1) compared to the previous a-aryl-substituted 27 (Ka 27  103 M
1) due to less steric hindrance. The binding affinities for Cl
among the ethynyl-substituted receptors (29b > 28a  28c > 28b  28d > 29a) correlated well with the electronic and steric effects from the various substitutions. While the UV–Vis studies established the 1:1 binding mode, the authors performed 1H NMR studies of 28b in the presence of TBA-chloride in CD2Cl2 at
50 °C to unravel the 2:1 sandwich binding mode from the integrals of the slow-exchanging signals of 28bCl
and 28b2Cl
complexes, and derived the equilibrium constants for the overall b21 (2H + G ? H2G), and stepwise K11 (H + G ? HG), K21 (HG + H ? H2G), K21 ? 11 (H2G + G ? [HG]2), notably with a strong positive cooperativity for the sandwich 28b2Cl
complex. The thermodynamic parameters estimated by variable-temperature (
50 to
20 °C) NMR studies revealed the enthalpy favoured b21, K11 and K21 equilibria, and K11 and K21 ? 11 are entropy driven processes. X-ray crystallography studies demonstrated both 28bCl
and 28b2Cl
complexes in solid state (Fig. 29b). The p-electron systems bearing long alkyl chains 28e–g were evaluated as ordered ion-pairing assemblies with TBA-chloride and 4,8,23-trioctyl-4,8, 12-triazatrianglenium (TATAC8) chloride for semiconductive properties using flash-photolysis time-resolved microwave conductivity technique. CO
2

Schmuck and co-workers [43] investigated the thermodynamic properties of the binding of lysine-based tetrapeptides 30a–d to DNA to determine the effect of incorporation of guanidinocarbonylpyrrole (GCP) anion receptors (Fig. 30). ITC binding data in cacodylate buffer showed that addition of the GCP group afforded an increase in an order of magnitude in Ka of DNA binding for 30a–d (Ka on the order of 106 M
1) over a tetramer of lysine (Ka = 2.5  105 M
1), with little significant difference between the GCP-containing compounds. Analysis of the thermodynamic profiles revealed that for 30c and 30d, where the number of hydrogen bonding GCP residues exceeds the number of positively charged lysine residues, the enthalpy of binding increased by an order of magnitude (over that for 30a and 30d), whereas the entropy of binding switched from a favorable to an unfavorable contribution. The authors suggested that the enthalpic contribution of the hydrogen bonding interactions of the GCPs is much greater than the electrostatic contribution from the lysine residues, yet the formation of a tightly bound complex increases the entropic cost of the binding event, evidenced by the apparent entropy–enthalpy compensation. DLS and zeta-potential experiments demonstrated that the higher GCP content peptides 30c and 30d were better able to condense DNA aggregates into smaller units, allowing 30d to facilitate gene transfection in HeLa cells where the other compounds could not. Kass and co-workers [44] investigated the nature of anion binding by 1-, 2- and 3-armed thiourea receptors 31a–c (Fig. 31) by photoelectron spectroscopy and DFT calculations. These techniques allowed accurate characterization of the vertical and adiabatic electron detachment energies of the conjugate bases, and Cl
, OAc
and H2PO
4 complexes of 31a–c. The results demonstrated that for all the anion complexes in the gas phase, positive cooperativity occurs and affinity increases with the number of hydrogen-bond donors, in contrast to findings in DMSO solution. The authors concluded that cooperative binding is cation/solvent dependent. The same authors [45] also used these techniques to probe the binding of Cl
by flexible acyclic polyalcohols, and showed that systems with a higher number of hydrogen-bond donors bind more strongly in the gas phase, whereas the reverse is observed in CD3CN, attributed to poorer solvation of the larger, charge-diffuse complexes. Marcos and co-workers [46] studied the competition between proton transfer and hydrogen bonding interactions during the recognition of electrogenerated di-anionic dinitrobenzenes by dihomooxacalix[4]arene-based bidentate thioureas 32a and 32b (Fig. 32) using voltammetric measurements in CD3CN (0.1 M TBAPF6 as supporting electrolyte). The authors demonstrated that

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

11

Finally, in important work discussing methodology used frequently in anion receptor chemistry, Jurczak and co-workers [47] discussed the limitations of the Job plot method in studying host– guest stoichiometries in supramolecular systems. They showed, through in-depth discussion of a number of simulated host–guest systems and case studies from the literature, the importance of using the correct model for Ka determination in titration experiments, to avoid reporting inaccurate or physically irrelevant values. It was demonstrated how Job plot analyses may be uninformative or misleading in certain cases, particularly at low host concentration (Fig. 33) or where the analytical signals for different complexes are of a similar value. The authors suggest that analysis of the residual distribution in data fitting results from titrations offers a more convenient and more accurate method for establishing host–guest stoichiometry and verifying the fitted model. 2.2. Halogen/chalcogen-bond based receptors

Fig. 31. Structures of the 1-, 2- and 3- armed thioureas 31a–c.

Fig. 32. Structure of the dihomocalixarene-based bidentate thioureas 32a and 32b and the family of dinitrobenzenes from which di-anions were electrogenerated.

for tert-butyl substituted 32a, the generated dianions were bound reversibly through hydrogen bonding to the diureas. For the more acidic phenyl substituted 32b, the voltammetric peaks associated with the formation of the ortho- and meta-dianions lose their reversibility with increasing urea concentration. In the case of the para-substituted dianion titrated with 32b, the peaks remained reversible and merged into a 2-electron wave. The mechanism of the competition between hydrogen bonding and proton transfer in the system was considered and rate constants for the proton transfer step obtained, indicating that for the more acidic receptor 32b, the fastest proton transfer occurs with the meta-dianion (k  25 M
1 s
1), over that of the ortho-dianion (k  5 M
1 s
1) while the para-dianion showed no proton transfer in these conditions.

Halogen bonding is becoming an increasingly useful tool in the design of anion receptors. Beer and co-workers demonstrated its utility in their report of a chiral S-BINOL ((S)-1,1-bi-2-naphthol) based receptor 33a for chiral anions (Fig. 34) [48]. Proton NMR titrations in 1% D2O in CD3CN showed that 33a was selective for (R)-NBoc–leucine, (S)-NBoc–tryptophan and (S)-BINOL–PO4 over their respective enantiomers, with enantioselectivities of 1.5–1.7. Binding constants were all of the order of 103–104 M
1. Contrastingly, hydrogen bonding receptor 33b had stability constants with all the chiral anions of the order of 102 M
1, with no preferred enantiomer. MD simulations and DFT calculations demonstrated that in the anion complexes of X-bonding analogs 33a, the halogen-bonding interactions are highly linear, while those in the hydrogen bonding analog 33b deviate significantly, which the authors suggest could help explain the differences in enantioselectivity. This, coupled with the apparently high degree of covalency of the interactions in the halogen-bonding analog, likely leads to the generally higher binding affinities. Beer and co-workers have also demonstrated a halogenbonding rotaxane 34 capable of recognizing iodide in water [49] (Fig. 35). Proton NMR titrations showed the rotaxane is capable of binding halides in water with a preference for I
(Ka = 6300 M
1 ) > Br
(1020 M
1) > Cl
(190 M
1). The affinities were much greater than those measured in control experiments with the halogen-bonding axle in combination with an acyclic model of the hydrogen-bonding macrocycle, demonstrating the importance of the interlocked molecule in creating a complementary binding site. MD simulations suggested a significant shielding of the bound anion from solvent molecules by the bulky interlocked molecule, and that selectivity for iodide was driven by thermodynamic effects. Beer and co-workers have also reported the strengthening of halogen-bonding interactions in a structurally-related series of rotaxanes by charge-assistance through methylation of the iodotriazole moiety [50]. Beer and co-workers have also exploited halogen bonding to improve the anion-induced shuttling behavior of a two station rotaxane 35X (Fig. 36) [51]. Upon exchange of the counter-anion
from PF
6 to Cl , shifts in the characteristic protons for the naphthalene diimide (NDI) and N-methyltriazole stations on the axle indicated that a change in conformation was taking place. Comparison of these shifts with values obtained in model one-station rotaxanes (100% occupancy) and of the naked axles in the presence of each anion (0% occupancy) the degree of occupancy on each station could be quantified in CDCl3 by 1H NMR. For hydrogenbonding analog 35H, although some shuttling to the NDI station was observed (0% for 35HCl
to 24% for 35HPF
6 ), the macrocycle demonstrated a preferred occupancy on the triazole station in the presence of both Cl
and PF
6 . In contrast, 35X exhibited a switch in

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

12

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 33. Simulated job plots for theoretical 1:2 host–guest interactions, at varying K1:1 and K1:2 ratios. K1:1 = 1000 M
1, YHG = 1, YHG₂ = 2 (Y = analytical signal), xmax = simulated plot maximum, column 1 is concentration of host. Note at low concentration (column 5, row 2) the Job Plot resembles that for a 1:1 interaction (xmax  0.5) despite a high K1:2, far from the expected (xmax  0.33) for a 1:2 stoichiometry. Reproduced with permission from Jurczak et al. [47]. Copyright 2016 American Chemical Society.

Fig. 34. S-BINOL based halogen bonding receptors 33a and 33b for chiral anions.

preference from the triazole station in the presence of Cl
to the NDI station with PF6 (NDI station occupancy 8% for 35XCl
to 62% for 35XPF
6 ). The same group has also reported a halogen bonding [3]-rotaxane 36 [52] based on similar moieties, which is able to recognize NO
via a pincer-like mechanism, in 3 which the two macrocycles shuttle from peripheral NDI stations to the two central iodo-triazolium stations to form a 1:1 sandwich complex.

Finally Taylor et al. have demonstrated the recognition of anions using a chalcogen bond donor [53]. These show similarities to halogen bonds, in that they originate from an area of positive electrostatic potential on an electron-deficient group 16 element (which is attracted to a Lewis base) and they are highly directional. The authors studied the anion binding properties of a series of 2,5 diaryltellurophanes 37a–d by UV–Vis titration in a variety of solvents. Only 37d displayed any appreciable interaction with anions

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

13

Fig. 35. Structure of halogen-bonding rotaxane receptor 34 for recognition of iodide in water.

under these conditions. DFT calculations determined that the magnitude of the r-hole in 37a and 37d were similar, and that the binding of Cl
was significantly assisted by an anion–p interaction with the electron deficient perfluorophenyl substituent. Bidentate receptor 38 demonstrated an order of magnitude improvement in Ka for Cl
binding in THF, and showed an increased affinity for Br
, I
, BzO
and NO
3 over 37d. DFT calculations indicated that the preferred binding conformation with Cl
involved two close Te  Cl contacts, with a bond angle of 170°, close to the ideal 180° (See Fig. 37).

possess moderate affinity (372–160 M
1). The general trend is smaller anionic guests led to larger entropic penalties, hence weaker binding. Quantum and MD simulations of the complexes



between 46 and ClO
4 , I and Cl , revealed that ClO4 and I retained about half of their solvation shells of water molecules when bound within the hydrophobic cavity of 46, while Cl
kept most of its hydration shell in the host compared with in the bulk. Therefore,
larger anions such as Cl3CCO
2 and ClO4 bind with minimal cocomplexing waters, and smaller anions are solvated by entropically costly water molecules.

2.3. Aromatic and hydrophobic anion acceptors

2.4. Ion-pair receptors

Anion–p interactions are governed by the combination of electrostatic effect and dispersion interactions. Aromatic rings containing electron-withdrawing groups such as cyano- and trifluoromethyl- will induce a large positive quadrupole moment (Qzz), perpendicular to the plane of the ring, affording an electron-deficient aromatic p-acceptor. Albrecht and co-workers [54] performed quantum–chemical calculations at the secondorder level of Moller–Plesset perturbation theory (MP2) on a series of CF3-substituted p-acceptors (39–43) interacting with fluoride, chloride and bromide anions. The Qzz and molecular polarizability () parameters increased with the increment of CF3 substituents on the aromatic ring, resulting in enhanced binding energies (DE) to the anions. Depending on the substitution on the aromatic and the polarizability of the anions, Meisenheimer r-complexes, single and double hydrogen bonding, and anion-p interactions, either directly above the centroid of the aromatic ring (g6-type) or side-on (g2-type) were observed from the model studies of 39– 43, as well as the crystal structure investigations on compounds such as 44 and 45 (Fig. 38). The recognition of hydrated anions within a large, well-defined binding pocket is very much dependent on the degree of hydration and the preferred hydration geometry. To address this, Gibb coworkers [55] examined the binding of monovalent anions (using sodium salts) to the hydrophobic cavity of cavitand 46 (Fig. 39) using ITC (50 mM sodium phosphate buffer, pH 11.5) and 1H NMR (10 mM sodium phosphate buffer, pH 11.3) titration studies. The binding anions which gave reliable data for both ITC and NMR are the soft, weakly hydrated anions, and were consistently enthalpically favored with entropic penalties. For instance, Cl3CCO
2 (6227 M
1) is the strongest binding anion, and I
(17 M
1) is the

weakest, and many tetrahedral anions such as ReO
4 , IO4 and ClO4

Rissanen and co-workers extended from the previous work on the bisurea based benzo[15]-crown-5 and benzo[18]-crown-6 ion-pair ditopic receptors 47a and 47b [56,57], and reported two uranyl salophen based ditopic receptors 48a and 48b with the same crown ether macrocycles (Fig. 40) [58]. Comprehensive crystallographic studies of receptors 48a and 48b with various alkali and ammonium halide salts demonstrated three interaction motifs, (I) separated ion-pair, (II) contact ion-pair, and (III) stacked packing of the receptors either with or without the ion-pair salts. More recently, the authors have reported the larger dibenzo[21]crown-7 49a and dibenzo[24]-crown-8 49b, and this time bearing two bisurea units [59]. Proton NMR titration studies of the free receptors 49a and 49b and in the presence of equimolar Rb+ and


Cs+ (as BPh
4 salts) with TBA-anion (Cl , Br and I ) salts in CDCl3/ DMSO-d6 (4:1, v/v) showed distinct positive heterotopic cooperativity. For instance, the association constants for 49aRb+ Cl
(78,000 M
1) versus 49a Cl
(354 M
1), and 49bRb+ Cl
(95,500 M
1) versus 49b Cl
(258 M
1) are more than two orders of magnitude greater. (Note that the association constants of free receptors are derived from 1:1 fitting, while 49aRb+ and 49bRb+ are from 1:2 fitting) Overall, the rubidium complexes of both 49a and 49b gave stronger binding to the halide anions compared to cesium, due to the stronger positive charge of the smaller Rb+. Mass spectrometry studies of the ion-pair complexes indicated receptor 49a shows better selectivity toward RbCl and CsCl in gas phase, consistent with the NMR results in solution. This selectivity is conferred by conformational differences of the ion-pair complexes, as demonstrated from the crystal structures in solid states (Fig. 40), 49a has an open conformation due to the asymmetry of 21-crown-7, while 49b has a folded conformation hence creating a stronger, more encapsulating anion binding site. In

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

14

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 36. Structure of multi-stationed halogen-bonding rotaxanes that show anion shuttling behavior.

addition, the crystal structure of 49bBa2+ 2Cl
was also reported to demonstrate the tritopic complexation of two anions and one cation, as were suggested from the NMR and MS studies. Receptors capable of extracting lithium ion-pairs are rare. Sessler and co-workers [60] have reported a hemispherand-strapped calix[4]pyrrole 50 as a selective receptor for LiCl over other chloride salts (NaCl, KCl and RbCl), evident from qualitative 1H NMR

studies in CD2Cl2/CD3OD (9:1, v/v). Five single crystal X-ray structures of the ion-pair complexes of 50 with LiCl, LiBr, LiI, LiNO2 and LiNO3 demonstrated the complexation of Li+ residing on one side (off-center) of the hemispherand cavity and anions are hydrogenbonded to the NH groups of calix[4]pyrrole. In addition, the LiCl complex of 50 revealed a water molecule as a solvent-bridge (Fig. 41) between Li+ and Cl
, while LiBr, LiI, LiNO2 and LiNO3 are

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 37. Structure of the chalcogen bonding aryltellurophane receptors 37a–d and 38.

15

encapsulated as tight contact ion-pairs, with the Li+  A
distances of 3.040, 3.064, 2.077 and 2.609 Å, respectively. The resulting association constants (Ka, M
1) from 1H NMR studies in CD2Cl2/CD3OD (9:1, v/v) of LiCl (45 M
1), LiBr (74 M
1), LiNO2 (108 M
1), and LiNO3 (263 M
1), revealed the selectivity trend. Due to the high level of nitrite anions present in the radioactive tank waste in the U.S., the authors devoted additional efforts in the recognition of LiNO2. Similar 1H NMR studies of 50 from titration of TBAnitrite and LiClO4, demonstrated the recognition of NO
2 ensues only in the presence of prebound Li+. Remarkably, ion-pair receptor 50 is the first example shown capable of extracting LiNO2 under both solid–liquid (CD2Cl2) and liquid–liquid (CDCl3 and D2O) conditions, with high selectivity over NaNO2 and KNO2. More recently, the authors also developed a chromogenic calix[4]arene-calix[4] pyrrole hybrid system with an indane substituent, functioning as a colorimetric sensor with AND logic gate modulation for CsF, CsCl and CsNO3 specifically [61].

Fig. 38. (a) Structures of CF3-substituted p-acceptors 39–43 and the energetic optimized structures in the presence of F
, Cl
, and Br
. The quadruple moment (Qzz) and molecular polarizability (hai) parameters, and the binding energies (DE) were calculated at the MP2/6-311++G**//MP2/6-31+G* level of theory, if not otherwise stated. (b and c) The single crystal X-ray structures of 44 and 45. Atom colors: white, H; black, C; blue, N; yellow, F; green, Cl.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

16

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 39. Structure of cavitand 46.

2.5. Photo-switchable anion receptors Vlatkovic´ and co-workers [62] continued from their previous work on chiral bisurea receptor 51 [63] that can photochemically and thermally switch between the three isomers (Fig. 42). For the first time, a dynamic interconvertible enantioselective binding has been demonstrated with chiral bisurea 51 in the binding of chiral binol phosphates. From 1H NMR titration studies in DMSOd6/0.5% water, (P,P)-cis-51 binds stronger to (R)-52
(415 M
1) than (S)-52
(100 M
1) by a factor of 4.2, while (M,M)-cis-51 favors

(S)-52
(55 M
1) to (R)-52
(17 M
1) by a factor of 3.2, and titration of (P,P)-trans-51 gave poor binding (<20 M
1). DFT optimized structures of [(P,P)-cis-54 (S)-55]
and [(P,P)-cis-54 (R)-55]
complexes revealed the naphthyl rings in (S)-52
are parallel with the plane of phenylurea substituents of (P,P)-cis-51, while (R)-52
are orthogonal. Hence there is less steric hindrance, resulting in higher stability and rationalizing the enantioselectivity of (P,P)cis-51 to (R)-52
. Comparably, the DFT results of [(M,M)-cis-51 (S)-52]
and [(M,M)-cis-51 (R)-52]
demonstrated the spatial complementarity with (S)-52
. Azobenzene is one of the most commonly used chemical motifs in photo-switchable systems. However, the cis isomer of this system has poor stability, and can easily undergo thermal- and photo-conversion to the thermodynamically stable trans isomer. Rananaware and co-workers [64] demonstrated the modulation of the stability of cis isomer in a naphthalenediimide (NDI) substituted azobenzene 53 through anion–p interactions. Irradiation studies of trans-53 in DMF at 366 nm, in the presence of TBA–fluoride salt, monitored using UV–Vis, fluorescence, EPR, 1H NMR (in acetone-d6) and 19F NMR (in acetone-d6) spectroscopy techniques unanimously confirmed the two stages of [NDI–F–NDI]
radical and [NDI-2F-NDI]2
dianionic complexes formation (Fig. 43).

Among F
, Cl
, Br
, I
, CH3CO
2 and H2PO4 anions, only F resulted in complexation. The cis-azo-NDI [53 2F]2
complex is stable and does not revert back to the trans isomer when irradiated at 500 nm or heated at 80 °C for 120 min. Addition of oxidizing agent NOBF4

Fig. 40. (a) Structures of benzo[15]-crown-5 and benzo[18]-crown-6 ditopic ion-pair receptors 47a–49b, and dibenzo[21]-crown-7 and dibenzo[24]-crown-8 tritopic +
2+ receptors 49a and 49b. The single crystal X-ray structures of the 49a2Rb+ CO2
2Cl
(d) complexes. Solvent molecules and most 3 (b), 49bRb Cl (c), and 49bBa hydrogen atoms are omitted for clarity. Atom colors: white, H; black, C; blue, N; red, O; green, Cl; purple, Rb; dark green, Ba.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

17

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 41. (a) Structure of hemispherand-strapped calix[4]pyrrole ion-pair receptor 50. (b) Single crystal X-ray structures of the LiClH2O (left) and LiNO2 (right) complexes of 50. Solvent molecules and non-acidic hydrogen atoms are omitted for clarity. Atom colors: white, H; purple, Li; black, C; blue, N; red, O; green, Cl.

2

O

P

O

60 °C

O

+ 1

+ 1 3K

31

(P,P)-cis-51

2

°C 1 +

3K

2 (P,P)-trans-51

O

(S)

+ 1 3K

(M,M)-cis-51

3K 1 +

O O

+ 1

O

36

20

31

nm

2n

2 2

+ 1

P

m

2

O O

(R)

m

+ 1

1 +

O

3K 1 +

5n

1 +

2

3K 1 +

P

O O

(R/S)-52 chiral binol phosphate

Fig. 42. Structure of chiral bisurea 51, and schematic illustration of the isomerization between three states and the enantioselective binding of chiral binol phosphate 52 between the cis isomers.

will recover trans-azo-NDI 53, demonstrating the reversible nature of this system. 2.6. Multivalent cationic systems Smith and co-workers [65] report the use of self-assembled multivalent (SAMul) cationic ligands (Fig. 44) for the recognition of two nanoscale biological polyanions, DNA and heparin via electrostatic interactions. The SAMul micelles are made up of amphiphilic molecules of the palmitoyl C16 hydrophobe bearing various amine ligands with different cationic charges, N,N-di-(3aminopropyl)-N-methylamine (DAPMA, +2) 54, spermidine (SPD, +2) 55, and spermine (SPM, +3) 56. The binding of DNA was evaluated using an ethidium bromide displacement assay monitored with fluorescence spectrometry, and a Mallard Blue competition

assay monitored using UV–Vis spectrometry was employed to study the binding of heparin. The highly charged C16-SPM 56 emerged as the best ligand for DNA binding, and C16-SPD 55 is the most effective heparin binder. The highly complex thermodynamic parameters of multiple simultaneous processes which include the binding of polyanion and SAMul aggregation in the presence of DNA or heparin were elucidated from ITC measurements. The thermodynamic data and molecular simulation studies rationalized the DNA/heparin binding selectivity arises from the intrinsic shape-persistence of DNA resulting in significant entropic penalty, while the adaptability of the structurally more flexible heparin can offset the entropic cost of binding. The authors also recently reported another class of SAMul based on a dipeptide cationic D/L-lysine-glycine palmitoyl amphiphiles 57a and 57b with enantioselectivity in DNA binding [66].

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

18

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

C8H17 O

N

O

O

N

O

C8H17 O

N

O

N

N

C8H17

O

N

O

C8H17

53

Stable cis-azo-NDI radical

cis-azo-NDI

Stable cis-azo-NDI

trans-azo-NDI Oxidation (NOBF4) Fig. 43. Structure of azo-NDI 53, and schematic illustration of the conversion between trans and cis isomers, with or without the presence of fluoride anion. Reproduced with permission from Rananaware et al. [64]. Copyright 2016 Nature Publishing Group.

Fig. 44. Structures of amphiphiles 54–57b for self-assembled multivalent (SAMul) binding of polyanions.

3. Metal–organic structures and anion-directed self-assembly 3.1. Metal–organic structures Coordination cages composed of metal ions and organic ligands have been utilized as anion receptors, where anions interact

through coordination to or electrostatic interaction with metal centers and can be further stabilized inside the cage by other non-covalent interactions. The size and shape of the cavity of the cage can dictate anion binding selectivity. Pfeffer [67] and coworkers have developed a rigid Pd2L4 coordination cage containing pre-organized, endohedrally oriented hydrogen bonds as a high affinity receptor for octahedral coordination anions (Fig. 45) [67]. The polynorbornene ligand 58 contains an imide group as a hydrogen-bond donor at its center. In CD3CN, the ligand selfassembled with Pd2+ to form cage {Pd4584} with four centrally oriented amide hydrogen-bond donors to bind anions and the two Pd2+ ions providing additional electrostatic interactions with anions. The cage was found to bind [Pt(CN)6]2
with strong affinity (>104 M
1) in CD3CN. Interestingly, upon being encapsulated inside {Pd4584}, the originally Oh symmetric [Pt(CN)6]2
underwent desymmetrization to D4h symmetry as revealed by single crystal structure of the complex showing different PtAC bond lengths and IR spectrum of the complex showing two distinct nitrile stretching bands. This was rationalized as arising from two different stabilizing interactions (electrostatic interactions with Pd2+ and hydrogen bonds with amide NH group). Cage {Pd4584} could also bind other species such as neutral carbonyl complexes ([M(CO)6]; M = Cr, Mo, W) and the linear anion [Ag (CN)2]
with weaker affinities. Clever and co-workers synthesized a sterically overcrowded metal-coordination cage [Pd2594] that displayed interesting properties including conformational dynamics within the cavity and binding of anionic guests driven by dispersion interactions (Fig. 46) [68]. The ligand 59 bears a bulky adamantyl group giving

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

19

Fig. 45. (a) Structure of ligand 58. (b) Formation of {Pd4584} cage and its encapsulation of [Pt(CN)6]2
. (c) Crystal structure of the {[Pt(CN)6]@Pd2584} complex. Solvents, counterions, and most hydrogen atoms have been omitted for clarity. Atom colors: white, H; black, C; blue, N; red, O; deep cyan, Pd; gray, Pt. Adapted with permission from Pfeffer et al. [67]. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

Fig. 46. (a) Formation of the [Pd2594] cage and binding of anionic guests inside the cavity. (b) Structures of anionic guests investigated. Adapted with permission from Clever et al. [68]. Copyright 2016 The Royal Society of Chemistry.

rise to a bent ligand structure with the adamantyl flipped to one side of the tricyclic acridone plane. In the free ligand, the adamantyl group can flip back and forth rapidly between two sides of the acridone plane, and upon formation of the [Pd2594] cage the ligand flipping dynamics preserved but significantly slowed down.

Despite the overcrowding due to the adamantyl groups in the cavity, the cage can accommodate anionic guests 60a to 60i in CD3CN with high affinities of >104
105 M
1 observed for 60a–60e. In the cases of 60d and 60e, computational studies revealed the host– guest interactions to be dominated by dispersion interactions

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

20

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

+16 H2 N 6 R H2 N

O O Mo O O O Mo O R O O

R R 61a (R = H) 61b (R = F) 24 O

NH2

NH2 CH3CN

62a (M = FeII, R = H) 62b (M = ZnII, R = H) 62c (M = FeII, R = F) 62d (M = ZnII, R = F)

N

28 MII(CF3SO3)2

16 CF3SO3

Fig. 47. Formation of four MII8L6 supramolecular tubes consisting of Fe2+ or Zn2+ as corner metal ions and non-fluorinated or fluorinated Mo2–tetraamine ligands. Adapted with permission from Nitschke et al. [69]. Copyright 2016 American Chemical Society.

Fig. 48. Halide exchange-induced positional switching of a phenyl group within a dinuclear Au(I)–Sb(V) complex.

between the guest and the surrounding adamantyl groups. Binding of the anionic guests led to compression of the cage along the Pd– Pd axis and further deceleration of the ligand flipping dynamics. Nitschke and co-workers reported self-assembly and guest binding of four supramolecular cubes in which subtle structural differences led to different conformation dynamics and anion binding selectivity [69]. Four MII8L6 cubes 62a–62d were assembled by a metal ion (Fe2+ or Zn2+), a Mo2–tetraamine ligand (nonfluorinated 61a or fluorinated 61b) and 2-formylpyridine (Fig. 47). They have the same topology and similar sizes but different configurational and phenyl ring rotation dynamics. The flexible coordination sphere of Zn2+ allowed the Mo2–tetraamine unit in cube 62b to exist in different diastereomers that interconvert in a temperature-dependent manner. No diastereomer formation was observed in the Fe2+ cube 62a. Fluorine substitution in the ligands resulted in rapid rotation of the phenylene rings in 62c and 62d as opposed to the slow rotation in 62a and 62b. These structural differences were also found to have a pronounced effect on anion binding to the molybdenum centers. While 62a–62c had the halide binding affinity in the order of I
> Br
> Cl
> F
, 62d showed the exactly opposite trend. The authors attributed this to the greatest extend of hardening of Mo center in 62d favoring binding of the hard anion F
. Indeed cube 62d was found to have

the lowest HOMO energy and the largest HOMO/LUMO gap presumably due to subtle ligand dynamics and inductive effects. Gabbaï and co-workers demonstrated an elegant example of anion-modulated molecular switch, in which substitution of the anions bound to metal centers led to positional switches of another ligand within the complex (Fig. 48) [70]. The authors synthesized a dinuclear Au(I)–Sb(V) complex 63a which contains a hexacoordinated Sb(V) center in a distorted octahedral geometry and a tetra-coordinated Au(I) center in a distorted square-planar geometry. Treatment of 63a with KF in CH3OH led to formation of a new difluorinated complex 63b involving a dramatic structural change that the Sb(V)-bound phenyl group displaced by 90° to a coordination site originally occupied by Cl
. Complex 63b could be transformed to a mixed halide complex 63c with no change in the position of the phenyl group by treatment with 1 equivalent of tert-butyldimethylsilyl chloride (TBDMSCl), or reverted to 63a by adding another equivalent of TBDMSCl. Remarkably, the phenyl group could be migrated to a third position, i.e., to the Au(I) center. This was achieved by treating 63b with TIPF6 to create a vacant coordination site at Au(I) (63d), followed by treatment with [Bu4N][Ph3SiF2] to form a zwitterionic intermediate 63e, which converted to 63f over 24–48 h. Except for 63d and 63e, the abovementioned com-

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

21

Fig. 49. (a) Formation of a catanene 64, a trefoil knot 65 and a Solomon link 66 under different conditions. Adapted with permission from Trabolsi et al. [71]. Copyright 2016 American Chemical Society. (b) Crystal structure of the 652Br
complex viewed from two different angles. The axial ligands and most hydrogen atoms have been omitted for clarity. Atom colors: white, H; black, C; blue, N; red, O; desaturated azure, Zn; brown, Br.

plexes contain a relatively strong Au ? Sb dative bond. The bond was dramatically weakened in cationic 63d complex and nonexistent in zwitterionic 63e complex. The system represents a mechanical three-way switch with the phenyl group moving between three distinct positions. 3.2. Anion-directed self-assembly Anions are often found to template the formation of metal– organic assemblies. Trabolsi and co-workers reported anion-

templated formation and anion binding properties of several interlocked metal–organic structures assembled by Zn2+, 1,3diformylpyridine and a diamino-2,20 -bipyridine (Fig. 49) [71]. The presence of an anion template and the size and shape of the anion used determined the topology and distribution of the products. Without anion templation, the catanene 64 was the most thermodynamically stable product and formed exclusively at 90 °C in 1:1 CD3OD/D2O. At 50 °C, both catanene 64 and a trefoil knot 65 formed, while the presence of two equivalents of Br
favored the formation of 65. A third product, the Solomon link 66 formed only

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

22

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 50. Formation of different metal–organic assemblies by bis(3-pyridylmethyl)amine and Cu2+ in the presence of different anions. Note that the SiF
6 inside the cage was presumably generated by reaction of F
(from hydrolysis of BF
4 ) with the glass surface of the container.

Fig. 51. (a) Synthesis of two tetrahedral M4L6 cages 71a and 71b. (b) Crystal structure of 71a. Most hydrogen atoms and solvent molecules have been omitted for clarity. Atom colors: white, H; black, C; blue, N; red, O; orange, P; green, Cl; gray, Fe.

when templated by the bulkier anion triflate (OTf
). The authors found by computational studies that 66 had a larger cavity than 65 and thus could better accommodate the larger OTf
by multiple CH  F hydrogen bonds. In D2O, 65 was a good anion receptor binding up to two equivalents of anions with high affinity (log b2

ranging from 4 to 6) and strong positive cooperativity (K2/K1 > 1) was found for Br
, I
and SCN
. By mixing a simple bis(3-pyridylmethyl)amine (dmpa) ligand with the Cu(II) salts of different anions under acidic conditions, Wu and co-workers obtained three different metal–ligand

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 52. Polyoxometalate anion-templated formation of the {Mo24Fe12(EDTA)12} macrocycle and anion exchange within the macrocycle. Adapted with permission from Cronin et al. [74]. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

Fig. 53. Formation of a tennis-ball like capsule by 73 binding to Cl
in a 2:1 stoichiometry. Solvent molecules, PF
6 counterions and non-interacting hydrogen atoms have been omitted in the crystal structure of the complex. Hydrogen bonds are represented by red dashed lines and Ni(II)  Cl
Coulomb interactions by blue dashed lines. Atom colors: white, H; black, C; blue, N; dark green, Ni. Br
can also form a similar tennis-ball like complex with 73.

assemblies, i.e., metallocages 67a–67e containing an encapsulated anion, a complex double salt 68, and a coordination polymer 69 (Fig. 50) [72]. The authors proposed that the shape of anions and their abilities to undergo hydrogen-bond and coordination interactions led to different outcomes in the assembly. For instance, the strongest coordinating anion AcO
formed a Cu2(OAc)4 motif presumably too stable for the dmpa to replace AcO
ions and form metallocages. Nakazawa and co-workers reported two examples of halidebinding driven formation of tetrahedral coordination cages 71a and 71b (Fig. 51) [73]. The two M4L6 type cages were synthesized by reaction of ferrocene-containing ligand 70 with PtX2(PPh3)2 (X = Cl
or Br
) in chloroform and consist of four square-planar Pt(II) centers bridged by six ligands. It is remarkable that the vertex positions of these cages are occupied by square-planar metal fragments, given that the vertex angle of a tetrahedron is 109.5°. The cages encapsulated four halide ions below the vertexes stabilized by NH  X
hydrogen bonds, and in the solid state one chloroform molecule was found at the center. The importance of halide binding in the cage formation was demonstrated in the failure to produce the corresponding cage structure when PtI2(PPh3)2 (I
is

23

much more weakly binding compared with Cl
and Br
) or Pt (PPh3)4 were used in place of PtX2(PPh3)2 (X = Cl
or Br
) in the synthesis. Cronin and co-workers showed the applicability of aniontemplated synthesis to formation of inorganic macrocycles (Fig. 52) [74]. By mixing Na2MoO4, FeCl36H2O, and Na2EDTA in a 2:1:1 ratio in aqueous solutions, the host–guest complex 72a was formed that containing an Anderson-type anion [FeMo6O18(OH)6]3
({FeMo6}) entrapped by the macrocycle {Mo24Fe12(EDTA)12} via multiple hydrogen bonds. In this case, the anionic template {FeMo6} was formed in situ. The {FeMo6} template could be exchanged by a Keggin-type anion [PMo12O40]3
({PMo12}) or a Dawson-type anion [P2W18O62]3
{P2W18} to form complexes 72b and 72c, which could also form directly by adding the anionic templates to the reaction mixture. The macrocycle could serve as a confined reaction container to induce the formation of an unstable Dawson-type anion [Mo12O36(HPO3)2(H2O)6]4
({Mo12(HPO3)2}) entrapped within the macrocycle (Complex 72d). Altmann and Pöthig demonstrated that anions binding to two equivalents of saddle-shape ligand can lead to the assembly of tennis-ball like capsules (Fig. 53) [75]. The ligand used was a dinuclear NHC–complex 73, which was found to bind Cl
and Br
ions in CD3CN in a 2:1 73/anion stoichiometry but did not recognize F
and I
. The crystal structures of Cl
and Br
complexes showed a ‘‘tennis-ball” like arrangement, in which the halide ion was encapsulated by two 73 ions, and stabilized within the cavity via Coulomb interactions with four square-planar Ni(II) centers and CH  X
hydrogen bonds with four endo-CH2 hydrogen atoms that point toward the central cavity. Given the catalytic activity of NHC complexes, future applications of such guest-encapsulating systems can be envisaged. Anion binding can influence the relative stability of interconvertible host architectures and thereby dictate the direction of interconversion as demonstrated by Su and co-workers (Fig. 54) [76]. Treatment of ligand 74 with Pd2+ using BF
4 as the counterion in CH3CN produced both a dimeric interlocked cage [3BF4@Pd4748] (BF4)5 and a monomeric cage [Pd2744](BF4)4. The interlocked cage was a kinetic product and upon DMSO treatment transformed into the thermodynamically stable monomeric cage within a week. However, when the reaction was performed in DMF or DMSO using NO
3 as the counterion, the monomeric cage [Pd2744](NO3)3 first formed but transformed into the dimeric interlocked cage [3NO3@Pd4748](NO3)5 after heating for a day. The authors proposed that the stronger binding anion NO
3 reduced the free energy of the interlocked cage through enthalpic Pd2+  NO
3 ion–ion interactions and allosteric effects, reversing the relative thermodynamic stability of the monomeric and interlocked cages. The monomer-to-dimer conversion could also be induced by other halide ions.

4. Transmembrane anion transport Recently significant research efforts have been dedicated to using small molecules to facilitate transport of anionic species across lipid bilayer membranes. The anion transport process can proceed via a mobile carrier mechanism or a channel-forming mechanism. Potential applications of this area of research include therapeutic use of anion transporters to disrupt pH or ionic gradients to treat cancer, or to replace the function of fatty anion channels to treat ion channel diseases such as cystic fibrosis, as well as exploitation of these compounds as tools to selectivity permeabilize biological membranes to anions in biophysics research. Several new structural scaffolds have been developed for carrying anions across lipid bilayers. Examples include the ‘‘perenosin”

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

24

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 54. Anion-dependent formation and interconversion between monomeric and interlocked cages. Adapted with permission from Su et al. [76]. Copyright 2016 The Royal Society of Chemistry.

Fig. 55. Structures of anion transporters 75–78.

Fig. 57. Mechanisms of H+/OH
conductance induced by synthetic anions transporters. (a) Mechanism I: H+ transport via transporter deprotonation and reprotonation; (b) mechanism II: OH
transport via reversible binding and release of OH
ions; (c) mechanism III: Indirect H+ transport via transporting the deprotonated form of fatty acids. Without the anion transporter, fatty acids cannot complete a H+ transport cycle rapidly because of slow transmembrane translocation rate of its anionic form (the process shown with a red forbidden sign). Reproduced with permission from Gale et al. [85]. Copyright 2016 American Chemical Society.

Fig. 56. Structures of anion transporters 79a–81b.

class of compounds with prodigiosin-inspired structures and anticancer activity such as 75 [77], oligo(aryl-triazole)s such as 76 that were found to be selective for Cl
over HCO
3 [78] and 1,3-bis(ben zimidazol-2-yl)benzene-based highly effective transporters such as 77 (Fig. 55) [79]. Different from those hydrogen bond-based transporters 75–77, transporter 78 reported by Matile and

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

25

Fig. 60. (a) Structures of bis-thioureidodecalins 89a–89d. (b) Cartoon representation explaining the optimal transport rate observed for 89d with medium alkyl chain length. Adapted with permission from Davis et al. [90]. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

Fig. 58. Structures of anion transporters 82–87c.

Fig. 59. (a) Structures of phenylthiosemicarbazones 88a–88c. (b) Mechanism of pH switchability of 88a–88c in anion transport.

co-workers (Fig. 55) [80] bound and transported anions via chalcogen bond interactions between the two electron-deficient sulfur atoms and an anion, thus expanding the range of intermolecular interactions applicable to facilitation of membrane transport. Functionalized anion transporters with biologically useful fluorescent or stimuli–responsive properties have been developed that have enabled monitoring or control of the spatiotemporal localization and distribution of anion transporters in cells (Fig. 56). Gale and co-workers have functionalized (thio)ureas with a naphthalimide fluorophore to develop a series of fluorescent anion transporters including 79a and 79b, and studied their cellular localization using fluorescence microscopy imaging [81]. Compound 79a was the most active anion transporter in vesicle assays among the series, but it failed to show cytotoxicity, probably due to

Fig. 61. Structures of anion transporters 90 and 91.

localization in specific vesicles as shown by fluorescence imaging. Compound 79b, by contrast, could localize homogeneously in the cells and induced apoptosis in A549 cells. Jeong and co-workers synthesized carbohydrate-containing urea compounds such as 80a to confer enzyme responsive properties on anion transporters [82]. The hydrophilic carbohydrate unit in compound 80a prevented this compound from facilitating anion transport, but could be cleaved by Aspergillus oryzae b-galactosidase to release the more lipophilic 80b as an active anion transporter. Another interesting example of stimuli–responsive anion transporters are selenocontaining bisureas developed by Liu and co-workers such as compound 81a [83]. This compound aggregated to form nanoparticles in water and was thus incapable of anion transport. Thiolcontaining species such as cysteine could reduce 81a from nanoparticles to release an active anion transporter 81b.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

26

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 62. Structures of shape-persistent macrocycles 92a–92d developed by Gong and co-workers, and their self-assembly to form nanotubule with different ion conducting properties. Intramolecular hydrogen bonds are represented by red dashed lines. Adapted with permission from Gong et al. [93]. Copyright 2016 American Chemical Society.

Fig. 63. Structure of channel forming compound 93 and cartoon representation of its self-assembly to form anion conducting channels which induce can apoptosis in cells by transporting Cl
into cells. Adapted with permission from Talukdar et al. [94]. Copyright 2016 American Chemical Society.

Synthetic anion transporters are often found to induce dissipation of transmembrane pH gradients, presumably via H+ or OH
transport coupled with the movement of the anion present in the medium. In cellular assays, many synthetic anion transporters like the natural product prodigiosin can disrupt pH gradients in cells potentially leading to toxicity. Gale and co-workers have conducted mechanistic studies of H+/OH
transport facilitated by anion transporters and also attempted to develop compounds selective for Cl
over H+/OH
to minimize the side effect of pH gradient disruption (Figs. 57 and 58) [84,85]. They found that prodigiosin behaved differently from most synthetic anion transporters in that it facilitates simultaneous transport (cotransport) of H+ and Cl
but cannot transport either H+ or Cl
alone (uniport). Most anion transporters, by contrast can facilitate H+ or OH
uniport in addition to Cl
uniport. Three mechanisms have been proposed for H+/OH
transport activity, i.e., H+ transport by transporter deprotonation and reprotonation (Mechanism I, Fig. 57a), OH
transport by the transporter binding and transporting OH
ions (Mechanism II, Fig. 57b), and indirect H+ transport by the transporter binding and transporting deprotonated fatty acids (Mechanism III, Fig. 57c, free fatty acids are present in both synthetic and biological membranes). Mechanism I was found to be

Fig. 64. Structures of channel-forming molecules 94a–95.

the dominant mechanism for highly acidic compounds such as 82 (pKa determined to be 8.9 in 9:1 v:v acetonitrile–water) that can easily deprotonate as required by this mechanism. For transporters that are not strong acids but good anion receptors such as 83, mechanism III dominated with even traces of free fatty acids and has been supported by the observation of fatty acid dependence in H+/OH
transport facilitated by anion transporters. A similar mechanism had been proposed by biologists for the fatty acid-dependent proton transport function of uncoupling proteins in mitochondria [85]. Mechanism II appears to be responsible for

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

observed H+/OH
uniport activity of the halogen bond donor 84 (initially studied as anion transporters by Matile and co-workers [86]) in the absence of fatty acids, in which case the other two mechanisms are not possible [85]. Initially, multivalent anion transporters with modestly acidic hydrogen-bond donors including 85 developed by the Gale group and 86 by the Davis group were identified to be highly Cl
> H+/OH
selective transporters in vesicle models [84]. However, the Cl
> H+/OH
selectivity dramatically reduced when fatty acids (1 mol% with respect to lipid) were added to the vesicles, because of an enhancement of the fatty acid-dependent proton transport pathway (Mechanism III) [85]. Subsequently, Gale and co-workers reported that strapped calixpyrroles 87a–87c demonstrated high Cl
> H+/OH
selectivity, which is, importantly, unaffected (87a and 87b) or weakly affected (87c) by the presence of fatty acids [87]. Remarkably, these compounds were also capable of facilitating transmembrane F
transport despite the high dehydration cost of F
. Compound 87a with the smallest cavity among the series was found to be a F
> Cl
selective transporter while 87c with a larger cavity was selective for the larger Cl
. Gale and co-workers have also developed prodigiosinmimicking compounds that facilitate the cotransport of H+ and Cl
as an ion pair (the overall process is non-electrogenic, i.e., does

27

not involve charge translocation across membranes), but do not transport anions or protons via uniport mechanisms (uniport of ions would lead to charge translocation across membranes). As a result of this electrically silent mechanism, such compounds can alter pH gradients but do not directly affect membrane potential. The non-electrogenic compounds studied are phenylthiosemicarbazones 88a–88c which function as pH-switchable anion transporters (Fig. 59) [88]. Under neutral and basic conditions, these compounds exist in neutral forms whose anion binding sites that are ‘‘locked” by a six membered ring intramolecular hydrogenbond inhibiting anion binding and transport. Protonation of the imine nitrogen atom under acidic conditions breaks the intramolecular hydrogen-bond, allowing these compounds to bind +

Cl
and NO
3 ions, and thereby facilitate H /Cl cotransport or Cl /

NO3 exchange. The compounds cannot facilitate Cl uniport presumably due to the inability of protonated forms to translocate across the membrane. Quesada and co-workers also performed a QSAR study on substitution effects on tambjamine anion transporters, where a parabolic dependence was found between the transporter lipophilicity and the anion transport efficacy with the medium range of lipophilicity giving the best transport rate [89]. Many anion transporters have been decorated with alkyl substituents to improve the lipophilicity properties for efficient anion

Fig. 65. (a) Structures of cell-penetrating peptide (CPP) activators 96a–97b and the fluorescently labeled CPP 98. (b) Interactions of CPPs with parallel ion pair–p activator 96a, antiparallel io  n-pair  p activator 96b and fluorinated activator 97a.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

28

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

transport. To rationalize the effect of alkyl substituents on anion transport efficacy, Davis and co-workers have performed experimental and computational studies of anion transport properties of bis-thioureidodecalins 89a–89f without and with alkyl substituents of increasing lengths (Fig. 60) [90]. Measurement of Cl
/ NO
3 exchange rate in vesicles demonstrated a parabolic relationship between the transport rate and the alkyl chain length, with optimal performance found for 89d. Molecular dynamics (MD) simulations revealed that for the transporters with shorter alkyl chains, the polar regions of the free transporter and the chloride complex orientated toward the interface leading to strong polar interactions with water or lipid phosphate head groups, which inhibits their translocation into the hydrophobic core of lipid bilayers. Longer alkyl chains pulled the polar region of the free transporter and the chloride complex, which reduced the abovementioned polar interactions and potentially enabled rapid anion transport. The decrease in anion transport rate when the alkyl chains are too long was explained by slow rotation of the free transporter and the complex as the longer alkyl chains leads to stronger aligning of these species with other lipid molecules in the membrane. To evaluate the potential of synthetic anion transporters in performing the function of CFTR channel in live cells, Davis, Sheppard and co-workers studied cellular anion transport activity of several synthetic anionophores based on the steroid, transdecalin and cyclohexane-tripodal scaffolds (Fig. 61) [91]. To assess anion transport activity in live cells, the authors devised a sensitive yellow fluorescent protein (YFP) assay, in which Cl
/ I
exchange across the plasma membrane facilitated by the tested compound leads to fluorescence quenching of YFP inside Fischer rat thyroid (FRT) cells as a result of entry of externally added I
. trans-Decalin-based bisurea 90 turned out to be the most active anion transporter in FRT cells, and importantly this compound was non-toxic to all tested cell lines. The superior performance of 90 in the FRT cell assay resulted in part from its excellent deliverability when added to the cells as DMSO solutions, whereas many highly lipophilic compounds such as 91 showed a high transport activity in vesicle assays but failed to function in cells due to poor deliverability when added in DMSO. To deliver 91 to cells, Kros and co-workers developed a drugdelivery system based on peptide recognition-driven membrane fusion [92]. Compound 91 was pre-incorporated into lipopeptides containing an amphiphilic peptide. The peptide can dimerize with a complementary peptide installed on targeted giant unilamellar vesicles (GUVs) or FRT cells, and this process promotes efficient membrane fusion between the lipopeptide and the targeted system. Compound 91 was thus delivered from lipopeptides to GUVs and FRT cells where it facilitated anion transport as monitored by fluorescence assays. Apart from the abovementioned anion carriers, several membrane transporters functioning via channel-forming mechanisms have been reported. Ion conducting nanotubules assembled by helical stacking of shape persistent macrocycle 92a–92d were reported by Gong and co-workers (Fig. 62) [93]. Interestingly, different inward pointing groups resulted in different ion transport properties. Facilitation of H+ transport was observed for 92a (X = H) and 92c (X = F), but not for 92b (X = CH3) and 92d (X = NH2). All compounds could facilitate Cl
transport with activity following the sequence of 92d > 92b > 92c > 92a. The authors proposed that the conduction of Cl
but rejection of H+ for 92d is due to the presence of electropositive protons at the amino group. The compounds also facilitated K+ transport, with 92A being more K+ > Cl
selective than 92d. Electrostatic potential maps of these compounds showed that 92d and 92b had less electronegative cavities than 92a and 92c, consistent with the abovementioned ion transport preferences of different compounds.

In contrast to macrocycles 92a–92d that are cation-selective, anion-selective channels have been developed by Talukdar [94], Kempf and Schmitzer [95] and Yang [96]. Talukdar and coworkers synthesized a series of compounds (93 and other compounds with different alkyl chain lengths) containing two vicinal diols groups attached to a rigid 1,3-diethynylbenzene core (Fig. 63) [94]. Molecular modeling suggested that such compounds could dimerize via hydrogen bonds between diol moieties and the resultant dimer could further stack to form channels with an internal diameter of 6 Å that could allow anions to pass through. Vesicle and planar lipid bilayer assays confirmed the anion transport ability of the compounds, with no significant cation transport detected. Compound 93 could transport Cl
into cells, leading to an increase of intracellular Cl
concentration and eventually apoptosis of HeLa cells. An umbrella-shaped molecule 94a and an umbrella-dimer 94b were reported by Kempf and Schmitzer to form transmembrane pores that facilitated anion transport with 94a showing antimicrobial activity toward both Gram-positive and Gram-negative bacteria (Fig. 64) [95]. Yang and co-workers have developed a structural simple isophthalamide 95 as a channel-forming small molecule that facilitated Cl
and HCO
3 transport in lipid bilayers and could restore Cl
conductance in monolayers of mutated epithelial cells with dysfunctional chloride channels (Fig. 64) [96].

Fig. 66. Structures of thiourea-based catalysts 82, 99–101, and the schematic illustrations (A–C) for three different reactions in this study.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Matile and co-workers have continued their research into arginine-rich cell-penetrating peptides (CPPs), which can mysteriously cross cell membranes despite their polycationic nature [97]. They have previously shown that lipophilic counterions can function as CPP activators because they assist CPPs to move across lipid bilayers by forming tightly bound, yet dynamic ion pair complexes with guanidinium groups of CPPs [98]. In this work, two new classes of CPP activators have been discovered: (i) carboxylates with a polarized push–pull p-system (e.g., 96a and 96b) capable of ion-pair  p interactions with CPPs; and (ii) fluorinated carboxylates (e.g., 97a and 97b) with a potential of fluorous self-assembly

29

giving polyanionic aggregates that could interact strongly with polycationic CPPs (Fig. 65). The active activators 96a and 97a could facilitate uptake of a fluorescently labeled CPP 98 into HeLa Kyoto cells as shown by fluorescence microscopy. No activity in cells was found for 96b with an antiparallel ion-pair  p-surface dipole arrangement and 97b with a more basic carboxylate group that allowed aggregation in its neutral, protonated form without ionpairing with CPPs. 5. Catalysis An important contribution from anion receptor chemistry is the utilization of metal-free anion binding receptor to direct catalysis, and the use of thiourea-based receptors gathered the most attention for actual application. For instance, N,N0 -bis(3,5-bis(trifluoro methyl)-phenyl)thiourea 82 (also discussed in the previous transmembrane anion transport section), is a commercially available catalyst known as the Schreiner’s thiourea catalyst, used for a wide range of organic chemical transformations. Fan and Kass [99] further explored this simple thiourea system by introducing charged moieties (Fig. 66, receptors 100 and 101) for electrostatic enhancement of the hydrogen-bond catalyst. The catalytic activity of 100 and 101 was evaluated with the Friedel–Crafts alkylation of trans-b-nitrostyrene with N-methylindole in CDCl3 (Scheme A), along with the unsubstituted thiourea 99 and Schreiner’s thiourea 82 with 10 mol% loading. Thiourea 99 is a poor catalyst, resulting in a meager 1.4-fold increase in the second-order rate constant with a half-life (t1/2) of 1100 h, and comparatively, Schreiner’s thiourea 82 is a better hydrogen-bond donor, contributing to a better catalytic effect achieving a t1/2 of 29 h. Remarkably, both the positively charged thiourea 100 and 101 are more active than 82, with enhanced t1/2 by 7-fold (4.5 h) and 400-fold (0.071 h)

Fig. 67. (a) Structures of chiral thiourea catalyst 102 and the anion-abstraction catalysis of a-chloroether 103 alkylated with silyl ketene acetals 104. (b) Structures of chiral bis-thiourea catalysts 106a and 106b, and its monomeric analogs 107a and 107b.

Fig. 68. (a) Structures of chiral thiourea catalysts 108 and 109 for enantioselective synthesis of cyclic a-aminophosphonates. (b) Schematic illustration of the asymmetric hydrogenation of isoquinolines catalyzed by a mixture of rhodium and ferrocene-based thiourea chiral phosphine 110.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

30

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

respectively. Similarly, trends of catalytic efficiencies (82 < 100

101) were observed in the Diels–Alder reaction between cyclopentadiene and methyl vinyl ketone (Scheme B), and the ring-opening aminolysis of styrene oxide with aniline (Scheme C). These results validated the strategy of introducing charge moieties to thioureabased systems to electrostatically enhance the catalytic activity. The concept of anion binding asymmetric catalysis using classical hydrogen-bond donor receptors was pioneered by Jacobsen. Recently, he and his co-workers have performed detailed mechanistic studies using the alkylation of a-chloroether 103 with silyl ketene acetals 104 catalyzed by chiral thiourea 102 via chloride-abstraction as the model reaction (Fig. 67) [100], to understand the prevalent limitations of anion binding catalysis, such as (1) the need of high catalyst loadings (5–20 mol%), (2) long reaction times (24 h), and (3) requirement of dilute reaction conditions limiting to low throughput. The observed change in the kinetic order, along with supportive data from X-ray crystallography, 2D NOESY and DFT studies, indicated that the catalysts rest as ineffective aggregated homodimers at high concentrations, while at low loadings, two monomers of thiourea 102 are required to form a 2:1 catalyst/substrate transition state to activate the catalysis. To overcome these limitations, Jacobsen and coworkers developed the bis-thiourea systems with ideal linkages that would favor the preferred a cooperative Cl
binding mode via both thiourea units for activation of a-chloroether 103 [101]. Using the same model reaction to compare with the activity of

catalyst 102, the monomeric analogs 107a and 107b gave similar reactivity and slight decrease in enantioselectivity compared to 102, and bis-thiourea 106b with the shorter linkage also exhibited similar activity to the mono-thioureas. On the other hand, bisthiourea 106a with the longer five-atom spacer improved the efficiency of the reaction to afford 105a (96% yield, 92% ee) with excellent yield and enantioselectivity, at a much lower loading of 0.1 mol% (100-fold lower), shorter reaction time of 3 h (8 quickly), and higher concentration of substrates at 0.5 M (5 more concentrated). Altogether, this has demonstrated the cooperative effect of a bis-thiourea with an optimal spacer to efficiently mediate the anion-abstraction catalysis. The importance of a-aminophosphonates is highlighted by its extensive use in medicinal and pharmaceutical sciences; while many have reported the catalytic enantioselective synthesis of acyclic a-aminophosphonates, a strategy for cyclic derivatives remains elusive until recently. Choudhury and Mukherjee [102] employed anion-binding directed catalysis for the elusive enantioselective synthesis of cyclic a-aminophosphonates, in the form of silyl phosphites addition to the N-acyl-activated isoquinolines (Fig. 68a). The authors investigated a series of chiral anion receptors bearing the commonly used hydrogen-bonding motifs (such as urea, thiourea and squaramide); the thiourea-based catalysts 108 and 109 (at 10 mol% loading) emerged with the best enantiomeric ratios of 96:4 and 94.5:5.5, and 95% conversion for both. Additional studies with different substrates of isoquinolines

Fig. 69. (a) Structures of chiral triazole-based catalysts 111–115. (b) Reissert-type enantioselective dearomatization of isoquinoline catalyzed by 111–115.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

and silyl phosphites provided a possible mechanistic rationale for the decrease in enantioselectivity due to the poor chloride anion stabilization by the catalyst to form an ion-pairing transition state with N-acyl isoquinolinium (refer to Fig. 69, same transition state of N-acylated isoquinoline). Zhang and co-workers [103] continued on the use of ferrocene-based thiourea chiral phosphine 110 with rhodium to catalyze the asymmetric hydrogenation of isoquinolines and quinolones (Fig. 68b). The catalytic studies were performed using a mixture of rhodium metal precursor [Rh(COD)Cl]2 (0.5 mol%) and 110 (1 mol%), in the presence of a strong Brønsted acid (HCl), and under pressurized H2 environment (40 atm). Among the different solvents (i-PrOH, CH3OH, CH2Cl2, dioxane) investigated, the mixture of CH2Cl2/i-PrOH (2:1, v/v) at 25 °C gave the optimal condition for high conversion (>99%) and enantioselectivity (99% ee). Mechanistic insight was obtained using deuterium labeled NMR studies to demonstrate an enamine–iminium tautomeization equilibrium after the initial hydride addition. Mancheño and co-workers continued to develop anion binding catalysts based on CH hydrogen-bond donors, using chiral helical tetratriazole-based systems in the Reissert-type enantioselective dearomatization of N-heretoarenes. A large diverse library of triazole-based catalysts 111–115 were prepared [104], and catalytic studies with all the triazole catalysts were screened using the asymmetric dearomatization of isoquinoline (0.1 mM) in Et2O, with TrocCl (1 equiv) as the alkylating agent, silyl ketene acetal (2 equiv) as the nucleophile and a catalyst at 10 mol% loading. The background reaction without a catalyst gave 44% conversion with no enantioselectivity (i.e. enantiomeric ratio (er) = 50:50). Most tetratriazole-based catalysts gave good enantioselectivity, with 112 and 113a emerging as the best catalysts, with 69:31 er (84% conversion) and 68:32 er (79% conversion), respectively. Bistriazole 111 and hexatriazoles 114b and 115b led to no appreciable enantioinduction, thus confirming that tetratrizoles are more effective in the binding of chloride anion stabilizing the transition state (Fig. 69). Further studies of the best catalysts 112 and 113a on the influence of solvents revealed methyl tertbutylether (MTBE) enhanced the enantioselectivity with 75:25 er (79% conversion) and 78:22 er (74% conversion) respectively, but reaction in CH2Cl2, THF and toluene led to poor enantioinduction. Kinetic studies of the same reaction for quinoline and isoquinoline substrates with tetratriazole catalysts 112 and 113a in MTBE monitored using HPLC indicated 112 required longer period (3 h) for the formation of active catalytic species, however 113a is extremely fast (<5 min); suggesting 113a is more preorganized for the binding of Cl
. The authors also utilized tetratriazole 113a in asymmetric dearomatization of diazarenes and showed high regioselectivity and enantioselectivity of up to 96:4 er [105]. Anion-binding catalysts based on anion–p interactions on the p-acidic surfaces of naphthalenediimide (NDI) were recently introduced by Matile and co-workers. They have continued on the development of using NDIs to catalyze enolate addition to introolefins [106], as well as the remarkable asymmetric catalysis for a cascade of iminium chemistry, conjugate additions, and Henry reactions to afford cyclohexane rings with five stereocenters [107]. More recently, the authors explored the use of perylenediimide (PDI) with an expanded p-acidic surface and larger positive quadrupole moments (Qzz). PDIs 116–119 functioned as catalysts for the enamine addition to nitroolefins with a stereoselective stabilization of the transition state (TS), and affording the product with two stereocenters (Fig. 70) [108]. The catalytic activity was determined in C6D6 with 25 mM catalyst loading, and out of the six reported PDI catalysts, 116(S) provided the best enantioselectivity with 96% ee, outperforming all examples reported in the literature. Contrary to the NDIs, the reaction rates and selectivities of these PDIs decreased with increasing redox potentials from the sulfide-substituted PDIs to the sulfones.

31

Fig. 70. (a) Structures of PDI catalysts 116–119. (b) Asymmetric addition of enamine to nitroolefin, and schematic illustration of the transition state.

Circular dichroism (CD) spectroscopy indicated the M-helical twist of sulfone 118(S) with a negative exciton-coupled bisignate CD at 300 nm, whereas the CD spectra of sulfides were nearly silent. Additional studies concluded that deplanarization of the psurfaces inactivates the anion–p catalysts; however, deplanarization overcompensates activation by withdrawing substituents. Nome and co-workers [109] reported the use of water soluble, cationic pillar[5]arene 120 as catalyst for the hydrolysis of 2,4dinitrophenylphosphate (DNPP). The complexation of DNPP within the cavity of pillar[5]arene 120 was evident from 1H NMR titration study with DNPP-lutidinium salt in D2O and molecular dynamics simulation. The hydrolysis of DNPP was monitored by the formation of 2,4-dinitrophenolate (DNP) using UV–Vis spectroscopy at 400 nm, at pH 7.0 in a bis–tris buffered solution. Based on the observed first-order rate constants (kobs), the optimal catalyst

Fig. 71. Structure of pillar[5]arene catalyst 120 for the hydrolysis of DNPP to DNP.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

32

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

loading of 1 was 2 mM with 0.05 mM of DNPP substrate, with kobs increment by 10-fold. Catalytic studies over a pH range of 2–13 revealed relatively no change of kobs between pH 5–13, but drastically lagged at acidic pH, due to the acidity of DNPP (pKa 4.62) – and the association constant of dianionic DNPP is 5.4-fold greater than the monoanion (See Fig. 71). Supramolecular catalysis based on self-assembled metalcoordinated cages and other similar molecular capsules has been a cornerstone of supramolecular chemistry since its emergence. The relatively rigid confined spaces of molecular cages as reactive catalytic site is an important attribute in achieving selectivity and catalytic rate enhancements, as is often displayed by biological systems in Nature. It is very pleasing to see the use of molecular capsule as anion-binding catalytic system recently reported by Reek and co-workers [110], in the form of a M12L24 nanosphere containing 24 endohedral guanidinium anion-binding motifs (Fig. 72a). The Pd(II)- and Pt(II)-coordinated nanospheres 122 and 123 were obtained by treating the guanidinium-tethered bialkynylpyridine ligand 121 (2 equiv) with one equivalent of Pd(CH3CN)4(OTf)2 or Pt(CH3CN)4(OTf)2 in CD3CN, confirmed by the NMR

and cold-spray ionization mass spectroscopy. NMR titration studies of the free ligand 121 with TBA salts of benzoate and ptoluenesulfonate in CD3CN gave Ka of 1850 M
1 and 330 M
1 respectively. Addition of benzoate to nanosphere 122 led to disassembly and 123 showed 80% encapsulation with 12 equivalents of benzoate. On the contrary, studies with p-toluenesulfonate indicated full encapsulation, and competitive binding experiment estimated much higher binding affinities (>105 M
1) by the nanospheres with p-toluenesulfonate, presumably enhanced by the multitopic cooperative binding geometry. The binding affinity difference of sulfonate/benzoate by the nanospheres was exploited for firm encapsulation of gold catalyst TPPMSAuCl (TPPMS = triphe nylphosphinomonosulfonate) and weaker binding of carboxylate substrate of deprotonated acetylenic acid within the more stable sphere 123, for the Au+-catalyzed cyclization to give enol lactone, in the presence of triethylamine as base (Fig. 72b). Catalytic reaction studies monitor by NMR of the free catalyst (TPPMSAuCl) gave 44% conversion with a turnover frequency (TOFini) of 0.55 h
1. Remarkably, in the presence of sphere 123 (+NEt3) to encapsulate catalyst and substrate in proximity significantly enhanced the

Fig. 72. Schematic illustrations of (a) the self-assembled nanospheres 122 [(Pd12L24)(OTf)48] and 123 [(Pt12L24)(OTf)48] and the PM-3 geometry optimized structure of Pd12L24 nanosphere with the endohedral guanidinium groups shown in space-filling mode; (b) and the nanosphere/Au+-catalyzed cyclization of acetylenic acid. Atom colors: white, H; gray, C; blue, N; red, O; orange, Pd. Adapted with permission from Reek et al. [110]. Copyright 2016 Nature Publishing Group.

Fig. 73. Structure of cage 124 [Co8L12](BF4)16 for the catalysis of the Kemp elimination of benzisoxazole, and the single crystal X-ray structure of 124benzisoxazole complex. All tetrafluoroborate anions and most hydrogen atoms are omitted for clarity. Atom colors: white, H; black, C; blue, N; red, O; pink, Co.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

33

Fig. 74. Structures of free pentafoil knot 125 and metal-coordinated pentafoil knots [Fe5125](PF6)10 and [Zn5125](BF4)10.

Fig. 75. Schematic illustration of the allosteric regulation of catalysis with [Zn5125] (BF4)10 for the generation of trityl carbocation.

reaction with a high conversion of >95% and 5.75 h
1 TOFini, while TPPMSAuCl + NEt3 without 123 resulted in a drop to 19% conversion and 0.19 h
1 TOFini. Notably, the absence of the NEt3 base (i.e. 123 + TPPMSAuCl) led to poor reactivity (17% conversion and

0.14 h
1 TOFini), due to no encapsulation of the neutral acetylenic acid. The capability of base-triggered encapsulation of anionic carboxylate substrate by 123 was applied as a substrate-selective catalytic system for the conversion of acetylenic acid over an alcohol substrate, allenol. Another work that is also based on a metal-coordinated system reported by Ward and co-workers [111], is an octanuclear coordination cage 124 [Co8L12](BF4)16 for the catalysis of the Kemp elimination of benzisoxazole, undergoing a ring-opening reaction with hydroxide to afford the anionic product, 2-cyano-phenolate (Fig. 73). The inclusion of neutral benzisoxazole within the hydrophobic cavity of cage 124 has an association constant of Ka 4  103 M
1 in water, and the single crystal X-ray structure of 124benzisoxazole complex was obtained by simply treating the single crystals of cage 124 with neat liquid benzisoxazole. The reaction was monitored by 1H NMR spectroscopy in D2O (pD = pH + 0.4) aqueous base (pD 8.5–11.4) solutions at 25 °C. Under these conditions with excess catalyst 124 (1 mM) to substrate concentration (0.85 mM), the Kemp elimination was distinctively enhanced by a factor of 4500 at pD 10.2, based on the first order reaction rate in the presence of cage 124 (kcat) and the background reaction (kuncat). Over the studied pD range, the rate of the background reaction drops as with the decrease in pD, however the rate of the catalyzed reaction in the presence of cage 124 remained unaffected; as a result, the catalytic enhancement is considerably higher at pD 8.5, with a kcat/kuncat factor of 2  105. The authors also successfully demonstrated high catalytic turnover at pD 10.2 with no change in the catalytic activity from multiple additions of benzisoxazole (0.85 mM). In another experiment, 100 equivalents of benzisoxazole was added to a lower concentration of cage 124 (0.1 mM) at pD 9.9, and again showing high turnover with a kcat/kuncat factor of 8800. The remarkable catalytic activity of cage 124 originates from an unconventional mediation of two orthogonal supramolecular interactions, (1) hydrophobic binding of benzisoxazole in the cavity of cage 124 and (2) polar binding of hydroxide anions on the cage surface; and the effective release of the anionic product ensuing high catalytic turnover. Leigh and co-workers’ [112] seminal report on the development of a controllable anion-binding catalytic system, much like the fundamental function of enzymatic reactions regulated

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

34

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 76. F
sensing by 126, showing crystal structure of the 126F
complex, and fluorescent changes of 126 upon addition of 2.4 ppm of F
in 9/1 (v/v) water/DMSO containing 10 mM of cetyltrimethylammonium bromide at pH 4.8. Atom colors: black, C; yellow, F; purple, Sb. Fluorescence photographs were reproduced with permission from Gabbaï et al. [113]. Copyright 2016 American Chemical Society.

Fig. 77. (a) Cl
sensing via gel-to-sol transition of gelator 127. DFT-calculated structure of the 127F
complex is shown. Gel and solution photographs were reproduced with permission from Feng et al. [115]. Copyright 2016 The Royal Society of Chemistry. (b) Luminescent sensing of F
by 128 and 128–Zn gels. Atom colors: black, C; red, O; yellow, F; purple, I. Gel photographs were reproduced with permission from Gong et al. [116]. Copyright 2016 The Royal Society of Chemistry.

Fig. 78. Structure of dihydrazide ligand 129 for formation of lanthanide(III)helicates and the mechanism of AMP binding by [Eu21292(NO3)2(H2O)4]4+.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

Fig. 79. Structure of Ru(II) halogen and hydrogen bonding sensors 130a(PF6)2 and 130b(PF6)2 for phosphates.

by allosteric control in Nature, is one of the most notable advances in this research theme. The authors reported a molecular pentafoil knot 125 (Fig. 74), as a catalyst with allosteric regulation. The single crystal X-ray structure of [Fe5125Cl
] (PF6)6(BF4)3 confirmed the topology of the pentafoil knot. Halide anions bind to the iron(II)-coordinated pentafoil knot [Fe5125](PF6)10 within the central cavity through CH hydrogen-bond and long range metal cation electrostatic interactions, with extremely high association constants (Ka(Cl
)  2. 6  1010 M
1 and K a(Br
)  1.4  1010 M
1) in CD3CN. The halide anion binding constants of Zn(II)–pentafoil knot [Zn5125](BF4)10 are approximately an order of magnitude lower for both Cl
(2.9  109 M
1) and Br
(7.7  108 M
1). Both [Fe5125](PF6)10 and [Zn5125](BF4)10 can function as a catalyst independently for the hydrolysis of bromodiphenylmethane, or for the generation of catalytic species such as the Lewis acidic trityl carbocation (for catalysis of Michael addition or Diels–Alder reactions) by facilitating the cleavage of the carbon–halogen bond (Fig. 75). Due to the fast and reversible coordination kinetics of Zn(II), the Zn(II)–pentafoil knot can be allosterically regulated by using Na4EDTA to regenerate the demetallated knot 125, followed by treating with Zn(BF4)2 to initiate the catalysis. 6. Sensing Gabbaï and co-workers have continued developing F
sensors based on coordinatively unsaturated Sb(V) complexes (Fig. 76) [113]. Compared with their previous work [114], sensor 126+ has the advantage of long-wavelength emission and high fluorescence

35

quantum yield. In 9/1 (v/v) water/DMSO containing 10 mM of cetyltrimethylammonium bromide at pH 4.8, sensor 126+ bound F
with a high affinity (10,000 ± 500 M
1) and displayed an 8fold enhancement in perylene fluorescence in the 440–570 nm region upon formation of the 126–F complex, with no response


for Cl
, Br
, NO
3 , HCO3 , HSO4 , and H2PO4 . The authors successfully used 126 to determine sub-ppm levels of F
in tap water. Two examples of supramolecular organogel-based anion sensors have been reported (Fig. 77). Gelator 127 developed by Feng and co-workers underwent gel-sol transition upon halogen bond interactions with Cl
[115]. In acetone where no gelation occurred, 127 bound halide anions with affinities of 650 M
1 (Cl
), 390 M
1

(Br
), and 140 M
1 (I
), while the affinities for NO
3 , HSO4 , CN
1 were below 10 M . In 8:1 hexane/acetone, 127 formed an organogel consisting of fibril networks of 127, which collapsed selectively in the presence of 1 equivalence of TBACl giving a clear solution of 127. Other less weakly binding anions could not induce the gel-to-sol transition at the same concentration. In contrast to 127, organogels of 128 and its Zn2+ complex 128–Zn developed by Gong, Ning and co-workers did not disassemble, but showed emission color changes upon interaction with the anionic analyte [116]. Gelator 128 formed blue-emissive gels at 0.5% in DMF and upon coordination to one equivalence of Zn2+ formed coordination polymer gels that emitted slightly blue-shifted and weaker fluorescence. The addition of F
changed the emission colors of 128 and 128–Zn gels to yellow and blue green, respectively, presumably due to hydrogen-bond interactions between F
and hydrazide NH protons. Solution-state fluorescence studies revealed that the yellow fluorescence of 128–F gels originates from aggregationinduced emission while 128–F itself is non-emissive at low concentrations in DMF. Subramanian and colleagues [117] reported a series of lanthanide(III)helicates with unsaturated coordination spheres which act as luminescent sensors for adenosine monophosphate (AMP) in aqueous buffer. On reaction with Eu(NO3)3, TbCl3 and Tb(NO3)3, ligand 129 forms dinuclear helical complexes with the lanthanides, forming discrete units of [Eu21292(NO3)2(H2O)4]4+ (with 4 NO
3 counter anions) for the europium salts and [Tb21292(H2O)4]6+ (with 6Cl
or NO
3 counter ions dependent on the starting salt) for terbium. The sensitive luminescent properties of [Eu21292(NO3)2(H2O)4](NO3)4 led the authors to screen the compound with 100 equiv. of a wide array of anions in HEPES buffer. A significant change in luminescence was only observed in the presence of AMP, even over ATP, ADP and other phosphates, with 100 equiv inducing a 64-fold increase in the Eu(III) emission intensity at 614 nm (an increase in observed quantum yield from 1.8% to

Fig. 80. Structure of poly amine macrocycle 131 and the structures of the anionic indicators with which it forms assemblies.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

36

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

6.1%). Titration with AMP yielded an estimated log Ka of 3.8. The authors suggested that AMP has the correct size and shape to effectively bridge the two Eu(III) centers, sterically excluding water from the coordination sphere giving rise to the large increase in luminescent quantum yield, evidenced by DFT models (See Fig. 78). Selective phosphate sensors have also been investigated by Ghosh and co-workers [118], who reported Ru(II) based halogen bonding sensor 130a(PF6)3. The complex exhibits a weak, broad luminescence band at 585 nm after excitation at its metal to ligand charge transfer absorption maximum at 403 nm. This emission is selectively enhanced in the presence of 10 equiv. of H2PO
4 (17  enhancement) and HP2O3
(5) over a variety of other anions, 7

including halides and other oxo-anions such as NO
3 , HSO4 , HCO3
and AcO . This response is unaffected by the presence of competitive anions in competition assays. Photoluminescence titrations 4
1 gave Ka of 1.9  105 M
1 for H2PO
for HP2O3
4 and 5.6  10 M 7 (CH3CN). H-bonding analog 130b(PF6)3 exhibits similar sensing properties to 130a(PF6)3 in CH3CN, although the emission enhancement is weaker (Ka = 5.6  104 M
1 for H2PO
4 in CH3CN 3
and the complex is unable to sense H2PO
4 and HP2O7 in mixtures of 10% water or more in CH3CN. By contrast, 130a(PF6)3 is able to tolerate mixtures up to 20%. 1H and 31P NMR plus single-crystal XRD data was evidence that led the authors to suggest that 130a (PF6)3 binds phosphates through a single halogen bonding interaction (See Fig. 79). Bencini and colleagues [119] studied the adducts formed between a giant polyamine macrocycle 131 and a number of anionic indicators (Fig. 80). The macrocycle is able to exist in a wide range of protonated states, with H81318+ being the most prominent species in neutral solution and potentiometric titrations allowed the estimation of binding constants for the formation of 1:1 and 1:2 adducts in aqueous solution. In UV–Vis pH titrations the 2:1 adduct BCP2131 demonstrated a two pH unit shift to a lower apparent pKa of the indicator (4.1) for its yellow to violet color change compared to the free indicator (6.3), due to the stabilization of the deprotonated form by complexation by the macrocycle. BCP2131 was tested as a possible displacement sensor for anions in aqueous solution at pH 5.4 with a wide variety of halides and oxoanions including several phosphates. Only triphosphate and diphosphate exhibited detectable change in the UV–Vis spectrum of the complexed indicator, with the spectrum reverting to that resembling that of the uncomplexed dye at pH 5.4. Comparison of the log Ka for the formation of a 1:1 adduct of 131 and BCP (log Ka = 7.7), triphosphate (8.9) and diphosphate (6.9) demonstrated why only these anions were able to displace the dye, with triphosphate exhibiting a much greater effect than diphosphate. The effect was strong enough with triphosphate for the color change to be detectable by the naked eye with 5 equiv. of anions, which was not possible for diphosphate. Anzenbacher and co-workers [120] developed a new family of macrocyclic chiral fluorescent receptors 132a–132d bearing the arylethynyl-appended BINOL chiral group orthogonally positioned to the macrocycle with four amide–NH donors (Fig. 81a). The fluorogenic receptors were investigated as chemosensors for chiral carboxylate guests. Fluorescence spectroscopy studies in propionitrile with the titration of TBA-carboxylate salts resulted in fluorescence quenching for receptors 132a–132b, presumably due to photoinduced electron transfer (PET); whereas the more electron-rich fluorophores 132c–132d led to an increase in fluorescence intensity due to the decrease in molecular motions upon anion complexation. The resulting association constants derived from fluorescence titration studies with enantiomers of ibuprofen, ketoprofen, 2-phenylpropanoate, mandelate and phenylalanine, showed the enantiomeric fluorescence difference ratio (ef) to

quantify the chiral discrimination of the analyte enantiomers. Subsequently, linear discriminant analysis (LDA) successfully established the clustering and classification of the analytes to validate the use of fluorophores 132a–132d in an array-based chemosensor. Finally, quantitative analysis of unknown samples of various enantiomeric compositions of ibuprofen, ketoprofen, 2phenylpropanoate, mandelate and phenylalanine demonstrated that the chemosensors can identify mixtures of varying enantiomeric excess with an error of <1.6%. The authors also

Fig. 81. Structures of (a) the macrocyclic arylethynyl-BINOL-based chiral fluorescent receptors 132a–132d for determination of enantiomeric excess of carboxylates; (b) and the tri-serine tri-lactone based tris(thiourea) 133 and tris (sulfonamide) 134 chemosensors.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

37

Fig. 82. (a) Structure of carbazolo[1,2-a]-carbazole receptor 135 for the binding of dicarboxylate anions; (b) and the single crystal X-ray structure of 1352adipate 2:1 sandwich complex. All TBA cations, DMSO solvent molecules and most hydrogen atoms are omitted for clarity. Atom colors: white, H; black, C; blue, N; red, O. (c) Emission spectra (kex 370 nm) of 135 (2  10
5 M) with the titration of TBA–glutarate; inset: fluorescence changes of 135 upon addition of TBA–glutarate (4  10
5 M). Adapted with permission from Curiel et al. [122]. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

Curiel and co-workers [122] reported the carbazolo[1,2-a]carbazole receptor 135 as a molecular sensor responding to dicarboxylate anions. Initial screening with the addition of TBA–dicarboxylate salts to receptor 135, monitored by 1H NMR titration studies in DMSO-d6, showed the typical downfield shifts of the carbazole–NH, carbazole–CH and amide–NH protons from the titration of the larger carboxylate (succinate, glutarate and adipate), while oxalate and malonate only induced downfield shifts of carbazole’s NH and CH signals. The single crystal X-ray structure of 1352adipate (Fig. 82b) as well as the binding isotherms and Job plots from NMR titration studies with succinate, glutarate and adipate demonstrated the formation of 2:1 H2:G sandwich complex; and the derived stepwise association constants of succinate (K11 1300 M
1, K21 800 M
1), glutarate (K11 3600 M
1, K21 250 M
1) and adipate (K11 2700 M
1, K21 500 M
1), indicated strongest affinity toward glutarate. Additional titration studies of receptor 135 with TBA–glutarate in DMSO using UV–Vis spectrometry showed a bathochromic shift with a discrete new absorption band at 435 nm, with color change of the solution from colorless to pale yellow. Similarly, the fluorescence intensity of free 135 at 409–425 nm decreases with addition of TBA–glutarate, accompanied with the appearance of a new emission band at 485 nm (Fig. 82c). The competitive ratiometric response (I(485 nm)/I(425 nm)) of 135 between succinate, glutarate and adipate confirmed the preferential binding of glutarate anion in DMSO.

7. Anion separation

Fig. 83. Crystal structure of the metal organic framework (MOF) developed by Ghosh and co-workers showing the (a) asymmetric unit and (b) crystal packing. Atom colors: gray, C; blue, N; red, O; orange, S; green, Ni. Reproduced with permission from Ghosh et al. [123]. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

reported the tris(thiourea) 133 and tris(sulfonamide) 134 fluorescent sensors based on a tri-serine tri-lactone scaffold (Fig. 81b), for the discrimination of mono-, di- and tri-carboxylate anions. In addition, a simple two-sensor array of 133 and 134 was able to determine the concentration of citrate from an unprocessed urine sample with an error of <2% [121].

An important application of anion coordination chemistry is to remove anionic contaminants from the environment. Ghosh and co-workers developed a cationic metal–organic framework (MOF) that can absorb oxoanion pollutants from water (Fig. 83) [123]. The MOF consists of tris(4-(1H-imidazol-1-yl)phenyl)amine ligands and Ni2+ ions coordinated to the four ligands, one H2O molecule and one SO2
4 ion. Notably free (uncoordinated to metal centers) SO2
4 ions are also present inside the MOF cavities stabilized by non-covalent interactions. The MOF can selectively absorb
Cr2O2
and MnO
7 4 (a model for the radioactive TcO4 ) ions from aqueous solutions in the presence of competing anions ClO
4,

2

NO
3 , BF4 , and CF3SO3 . The uptake of Cr2O7 and MnO4 replaced the SO2
4 ions from the matrix and gave rise to visible color changes in the MOF.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

38

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

2
Fig. 84. (a) SO2
4 separation by crystallization with 136–SO4. (b) Crystal structures of 136–SO4 showing guanidinium  SO4 hydrogen bonding interactions (left) and stacking of 1362+ cations. Atom colors: white, H; black, C; blue, N; red, O; yellow, S. Reproduced with permission from Custelcean et al. [124]. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

Separation of the highly hydrophilic anion SO2
4 is challenging and normally the toxic Ba2+ is used to form the insoluble BaSO4. Custelcean and co-workers have developed an effective SO2
4 separation technique by crystallization of SO2
with a bis4 guanidinium receptor that can be in situ formed via imine condensations in aqueous solutions (Fig. 84a) [124]. The resultant 136–SO4 salt has an extremely low aqueous solubility determined to be 1.6  105 M
1 at 25 °C. Crystal structures of 136– SO4 demonstrated a [(SO4)2(H2O)4]4
4 cluster entrapped by stacks of 1362+ cations, and each SO2
formed 7 NH  O hydrogen 4 bonds with guanidinium groups, and 4 OH  O hydrogen bonds with water. The 1362+ cations stacked via C@N  Ph and H2N  C@N electrostatic interactions, which probably contributed to the low solubility of the crystal. This technique has been demonstrated to remove 99% of SO2
from seawater using 1.5 4 equivalents of 1362+. 8. Other applications Fig. 85. Operating mechanism of the thermo-electrochemical cell developed by Yamada and co-workers based on the I
/I
3 redox pair and the temperature sensitive a-CyD–I
3 interactions. Reproduced with permission from Yamada et al. [125]. Copyright 2016 American Chemical Society.

Yamada and co-workers have designed conceptually new thermo-electrochemical cells by harnessing supramolecular anion binding interactions (Fig. 85) [125]. The cells based on the I
/I
3

Fig. 86. Structure of ligand 137 and formation of [Er137(H2O)]+ and [(Er137)2F]+ (the latter shown as a DFT-optimized structure). Hydrogen atoms have been omitted. Atom colors: black, C; blue, N; red, O; yellow, F; green, Er.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

redox pair could convert heat into electric energy because of the temperature sensitive encapsulation of the oxidizing form I
3 by a-cyclodextrin (a-CyD). At the cold side of the cell, I
3 binding to a-CyD was favored which reduced the local concentration of I
3 shifting the equilibrium in the oxidation direction. At the hot side, I
3 dissociated from a-CyD leading to elevated local concentration
of I
3 and consequently a shift of the equilibrium toward I3 reduction to I
. Thus, the presence of a-CyD converted a temperature gradient into a concentration gradient of I
3 and eventually led to generation of an electric potential. Upconversion, an unusual luminescent phenomenon in which the light emitted is of higher energy than that absorbed, is difficult to achieve in small molecules. Charbonnière and co-workers reported the first example of room-temperature molecular upconversion in D2O based on a lanthanide-anion complex (Fig. 86) [126]. The [Er137(H2O)]+ complex self-assembled in the presence of F
to form a dimeric [(Er137)2F]+ complex with a F
ion sandwiched between two monomers. Both complexes can emit upconversion green light upon excitation by near infrared light in D2O, but [(Er137)2F]+ is 8 times more luminescent than [Er137 (H2O)]+. The authors proposed that an excited state energy transfer between two Er(III) centers occurred during the upconversion emission of [(Er137)2F]+. 9. Conclusions Over the last twenty years we have seen continuing advancement in the development of anion receptors and particularly recently real-world applications of these systems. In 2016, we have seen new receptor designs to enhance binding selectivity and bind anions in water. In addition, anion-directed self-assembly continues to deliver many fascinating architectures allowing access to more complex molecular topologies. Applications continue to drive developments in anion sensing and separation, while it is exciting to see anion complexation play a role in electrochemical cells and luminenscent materials. There are some notable advances in transmembrane anion transport and selective translocation of anions, and also in the use of more complex molecular cages and knots to regulate catalysis. We look forward to the next 20 years of anion receptor chemistry. Acknowledgments PAG thanks the University of Sydney and the Australian Research Council (DP180100612, DP170100118 and LE180100050) for funding. MJS thanks the University of Southampton and the EPSRC for a Doctoral Prize Award (EP/N509747/1). References [1] P.A. Gale, Coord. Chem. Rev. 199 (2000) 181–233. [2] N. Busschaert, C. Caltagirone, W. Van Rossom, P.A. Gale, Chem. Rev. 115 (2015) 8038–8155. [3] P.A. Gale, E.N.W. Howe, X. Wu, Chem 1 (2016) 351–422. [4] P.A. Gale, N. Busschaert, C.J.E. Haynes, L.E. Karagiannidis, I.L. Kirby, Chem. Soc. Rev. 43 (2014) 205–241. [5] M. Wenzel, J.R. Hiscock, P.A. Gale, Chem. Soc. Rev. 41 (2012) 480–520. [6] P.A. Gale, Chem. Soc. Rev. 39 (2010) 3746–3771. [7] C. Caltagirone, P.A. Gale, Chem. Soc. Rev. 38 (2009) 520–563. [8] P.A. Gale, S.E. Garcia-Garrido, J. Garric, Chem. Soc. Rev. 37 (2008) 151–190. [9] P.A. Gale, R. Quesada, Coord. Chem. Rev. 250 (2006) 3219–3244. [10] P.A. Gale, Coord. Chem. Rev. 240 (2003) 191–221. [11] P.A. Gale, Coord. Chem. Rev. 213 (2001) 79–128. [12] V. Amendola, G. Bergamaschi, M. Boiocchi, L. Legnani, E.L. Presti, A. Miljkovic, E. Monzani, F. Pancotti, Chem. Commun. 52 (2016) 10910–10913. [13] M.I. Rednic, R.A. Varga, A. Bende, I.G. Grosu, M. Micla˘usß, N.D. Ha˘dade, A. Terec, E. Bogdan, I. Grosu, Chem. Commun. 52 (2016) 12322–12325. [14] B.E. Ziegler, M. Lecours, R.A. Marta, J. Featherstone, E. Fillion, W.S. Hopkins, V. Steinmetz, N.S. Keddie, D. O’Hagan, T.B. McMahon, J. Am. Chem. Soc. 138 (2016) 7460–7463.

39

[15] M.A. Majewski, Y. Hong, T. Lis, J. Gregolin´ski, P.J. Chmielewski, J. Cybin´ska, D. Kim, M. Ste˛pien´, Angew. Chem. Int. Ed. 55 (2016) 14072–14076. [16] M.A. Yawer, V. Havel, V. Sindelar, Angew. Chem. Int. Ed. 54 (2015) 276–279. [17] T. Lizal, L. Ustrnul, M. Necas, V. Sindelar, J. Org. Chem. 81 (2016) 8906–8910. [18] H. Zhu, B. Shi, K. Chen, P. Wei, D. Xia, J.H. Mondal, F. Huang, Org. Lett. 18 (2016) 5054–5057. [19] P. Sabater, F. Zapata, A. Caballero, I. Fernández, C.R.D. Arellano, P. Molina, J. Am. Chem. Soc. 81 (2016) 3790–3798. [20] J. Shang, W. Zhao, X. Li, Y. Wang, H. Jiang, Chem. Commun. 52 (2016) 4505– 4508. [21] R.N. Robson, F.M. Pfeffer, Chem. Commun. 52 (2016) 8719–8721. [22] M.K. Deliomeroglu, V.M. Lynch, J.L. Sessler, Chem. Sci. 7 (2016) 3843–3850. [23] A. Sampedro, R. Villalonga-Planells, M. Vega, G. Ramis, S. Fernández de Mattos, P. Villalonga, A. Costa, C. Rotger, Bioconjugate Chem. 25 (2014) 1537– 1546. [24] B. Soberats, L. Martínez, E. Sanna, A. Sampedro, C. Rotger, A. Costa, Chem. Eur. J. 18 (2012) 7533–7542. [25] N. Busschaert, I.L. Kirby, S. Young, S.J. Coles, P.N. Horton, M.E. Light, P.A. Gale, Angew. Chem. Int. Ed. 51 (2012) 4426–4430. [26] L. Qin, A. Hartley, P. Turner, R.B.P. Elmes, K.A. Jolliffe, Chem. Sci. 7 (2016) 4563–4572. [27] A. Abebayehu, R. Dutta, S.-J. Kim, J.H. Lee, H. Hwang, C.-H. Lee, Eur. J. Org. Chem. 2016 (2016) 3959–3963. [28] P.A. Gale, J.R. Hiscock, S.J. Moore, C. Caltagirone, M.B. Hursthouse, M.E. Light, Chem. Asian J. 5 (2010) 555–561. [29] A. Satake, Y. Ishizawa, H. Katagiri, S.-I. Kondo, J. Org. Chem. 81 (2016) 9848– 9857. [30] F. Zhang, Y.-H. Zhou, X.-P. Bao, Supramol. Chem. 28 (2016) 305–313. [31] J.P. Byrne, S. Blasco, A.B. Aletti, G. Hessman, T. Gunnlaugsson, Angew. Chem. Int. Ed. 55 (2016) 8938–8943. [32] E.M. Fatila, E.B. Twum, A. Sengupta, M. Pink, J.A. Karty, K. Raghavachari, A.H. Flood, Angew. Chem. Int. Ed. 55 (2016) 14057–14062. [33] B. Qiao, J.R. Anderson, M. Pink, A.H. Flood, Chem. Commun. 52 (2016) 8683– 8686. [34] M.D. Hartle, R.J. Hansen, B.W. Tresca, S.S. Prakel, L.N. Zakharov, M.M. Haley, M.D. Pluth, D.W. Johnson, Angew. Chem. Int. Ed. 55 (2016) 11480–11484. [35] N. Rodríguez-Vázquez, M. Amorín, I. Alfonso, J.R. Granja, Angew. Chem. Int. Ed. 55 (2016) 4504–4508. [36] S. Ruiz-Botella, P. Vidossich, G. Ujaque, E. Peris, Chem. Eur. J. 22 (2016) 15800–15806. [37] N.K. Beyeh, F. Pan, S. Bhowmik, T. Mäkelä, R.H.A. Ras, K. Rissanen, Chem. Eur. J. 22 (2016) 1355–1361. [38] J. Gregolin´ski, K. S´lepokura, T. Pac´kowski, J. Panek, P. Stefanowicz, J. Lisowski, J. Org. Chem. 81 (2016) 5285–5294. [39] G. Mann, L.J. Twyman, P.A. Gale, Chem. Commun. 52 (2016) 6131–6133. [40] V. Diemer, L. Fischer, B. Kauffmann, G. Guichard, Chem. Eur. J. 22 (2016) 15684–15692. [41] Y. Haketa, Y. Tamura, N. Yasuda, H. Maeda, Org. Biomol. Chem. 14 (2016) 8035–8038. [42] R. Yamakado, T. Sakurai, W. Matsuda, S. Seki, N. Yasuda, S. Akine, H. Maeda, Chem. Eur. J. 22 (2016) 626–638. [43] M. Li, S. Schlesiger, S.K. Knauer, C. Schmuck, Org. Biomol. Chem. 14 (2016) 8800–8803. [44] E.V. Beletskiy, X.-B. Wang, S.R. Kass, J. Phys. Chem. A 120 (2016) 8309– 8316. [45] A. Shokri, X.-B. Wang, Y. Wang, G.A. O’Doherty, S.R. Kass, J. Phys. Chem. A 120 (2016) 1661–1668. [46] E. Martínez-González, F.J. González, J.R. Ascenso, P.M. Marcos, C. Frontana, J. Org. Chem. 81 (2016) 6329–6335. [47] F. Ulatowski, K. Da˛browa, T. Bałakier, J. Jurczak, J. Org. Chem. 81 (2016) 1746– 1756. [48] J.Y.C. Lim, I. Marques, L. Ferreira, V. Felix, P.D. Beer, Chem. Commun. 52 (2016) 5527–5530. [49] M.J. Langton, I. Marques, S.W. Robinson, V. Félix, P.D. Beer, Chem. Eur. J. 22 (2016) 185–192. [50] A.E. Hess, P.D. Beer, Org. Biomol. Chem. 14 (2016) 10193–10200. [51] T.A. Barendt, S.W. Robinson, P.D. Beer, Chem. Sci. 7 (2016) 5171–5180. [52] T.A. Barendt, A. Docker, I. Marques, V. Félix, P.D. Beer, Angew. Chem. Int. Ed. 55 (2016) 11069–11076. [53] G.E. Garrett, E.I. Carrera, D.S. Seferos, M.S. Taylor, Chem. Commun. 52 (2016) 9881–9884. [54] M. Albrecht, H. Yi, O. Köksal, G. Raabe, F. Pan, A. Valkonen, K. Rissanen, Chem. Eur. J. 22 (2016) 6956–6963. [55] P. Sokkalingam, J. Shraberg, S.W. Rick, B.C. Gibb, J. Am. Chem. Soc. 138 (2016) 48–51. [56] T. Mäkelä, E. Kalenius, K. Rissanen, Inorg. Chem. 54 (2015) 9154–9165. [57] T. Mäkelä, K. Rissanen, Dalton Trans. 45 (2016) 6481–6490. [58] T. Mäkelä, M.-E. Minkkinen, K. Rissanen, Inorg. Chem. 55 (2016) 1339–1346. [59] T. Mäkelä, A. Kiesilä, E. Kalenius, K. Rissanen, Chem. Eur. J. 22 (2016) 14264– 14272. [60] Q. He, Z. Zhang, J.T. Brewster, V.M. Lynch, S.K. Kim, J.L. Sessler, J. Am. Chem. Soc. 138 (2016) 9779–9782. [61] Y. Yeon, S. Leem, C. Wagen, V.M. Lynch, S.K. Kim, J.L. Sessler, Org. Lett. 18 (2016) 4396–4399. [62] M. Vlatkovic´, B.L. Feringa, S.J. Wezenberg, Angew. Chem. Int. Ed. 55 (2016) 1001–1004.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005

40

P.A. Gale et al. / Coordination Chemistry Reviews xxx (2018) xxx–xxx

[63] S.J. Wezenberg, M. Vlatkovic´, J.C.M. Kistemaker, B.L. Feringa, J. Am. Chem. Soc. 136 (2014) 16784–16787. [64] A. Rananaware, M. Samanta, R.S. Bhosale, M.A. Kobaisi, B. Roy, V. Bheemireddy, S.V. Bhosale, S. Bandyopadhyay, S.V. Bhosale, Sci. Rep. 6 (2016) 22928. [65] L.E. Fechner, B. Albanyan, V.M.P. Vieira, E. Laurini, P. Posocco, S. Pricl, D.K. Smith, Chem. Sci. 7 (2016) 4653–4659. [66] C.W. Chan, E. Laurini, P. Posocco, S. Pricl, D.K. Smith, Chem. Commun. 52 (2016) 10540–10543. [67] M.D. Johnstone, E.K. Schwarze, J. Ahrens, D. Schwarzer, J. Holstein, B. Dittrich, F.M. Pfeffer, G.H. Clever, Chem. Eur. J. 22 (2016) 10791–10795. [68] S. Löffler, J. Lübben, A. Wuttke, R.A. Mata, M. John, B. Dittrich, G.H. Clever, Chem. Sci. 7 (2016) 4676–7684. [69] W.J. Ramsay, F.J. Rizzuto, T.K. Ronson, K. Caprice, J.R. Nitschke, J. Am. Chem. Soc. 138 (2016) 7264–7267. [70] S. Sen, I.-S. Ke, F.P. Gabbaï, Inorg. Chem. 55 (2016) 9162–9172. [71] R.A. Bilbeisi, T. Prakasam, M. Lusi, R.E. Khoury, C. Platas-Iglesias, L.J. Charbonnière, J.-C. Olsen, M. Elhabiri, A. Trabolsi, Chem. Sci. 7 (2016) 2524– 2531. [72] J.-Y. Wu, M.-S. Zhong, M.-H. Chiang, D. Bhattacharya, Y.-W. Lee, L.-L. Lai, Chem. Eur. J. 22 (2016) 7238–7247. [73] M. Ito, M. Iseki, M. Itazaki, H. Nakazawa, Chem. Commun. 52 (2016) 7205– 7208. [74] W. Xuan, A.J. Surman, Q. Zheng, D.-L. Long, L. Cronin, Angew. Chem. Int. Ed. 55 (2016) 12703–12707. [75] P.J. Altmann, A. Pöthig, Chem. Commun. 52 (2016) 9089–9092. [76] Y.-H. Li, J.-J. Jiang, Y.-Z. Fan, Z.-W. Wei, C.-X. Chen, H.-J. Yu, S.-P. Zheng, D. Fenske, C.-Y. Su, M. Barboiu, Chem. Commun. 52 (2016) 8745–8748. [77] W.V. Rossom, D.J. Asby, A. Tavassoli, P.A. Gale, Org. Biomol. Chem. 14 (2016) 2645–2650. [78] S. Chen, S. Zhang, C. Bao, C. Wang, Q. Lin, L. Zhu, Chem. Commun. 52 (2016) 13132–13135. [79] C.-C. Peng, M.-J. Zhang, X.-X. Sun, X.-J. Cai, Y. Chen, W.-H. Chen, Org. Biomol. Chem. 14 (2016) 8232–8236. [80] S. Benz, M. Macchione, Q. Verolet, J. Mareda, N. Sakai, S. Matile, J. Am. Chem. Soc. 138 (2016) 9093–9096. [81] S.N. Berry, V. Soto-Cerrato, E.N.W. Howe, H.J. Clarke, I. Mistry, A. Tavassoli, Y.T. Chang, R. Pérez-Tomás, P.A. Gale, Chem. Sci. 7 (2016) 5069–5077. [82] Y.R. Choi, B. Lee, J. Park, W. Namkung, K.-S. Jeong, J. Am. Chem. Soc. 138 (2016) 15319–15322. [83] C. Lang, X. Zhang, Z. Dong, Q. Luo, S. Qiao, Z. Huang, X. Fan, J. Xu, J. Liu, Nanoscale 8 (2016) 2960–2966. [84] X. Wu, L.W. Judd, E.N.W. Howe, A.M. Withecombe, V. Soto-Cerrato, H. Li, N. Busschaert, H. Valkenier, R. Pérez-Tomás, D.N. Sheppard, Y.-B. Jiang, A.P. Davis, P.A. Gale, Chem 1 (2016) 127–146. [85] X. Wu, P.A. Gale, J. Am. Chem. Soc. 138 (2016) 16508–16514. [86] A. Vargas Jentzsch, D. Emery, J. Mareda, S.K. Nayak, P. Metrangolo, G. Resnati, N. Sakai, S. Matile, Nat. Commun. 3 (2012) 905. [87] H.J. Clarke, E.N.W. Howe, X. Wu, F. Sommer, M. Yano, M.E. Light, S. Kubik, P.A. Gale, J. Am. Chem. Soc. 138 (2016) 16515–16522. [88] E.N.W. Howe, N. Busschaert, X. Wu, S.N. Berry, J. Ho, M.E. Light, D.D. Czech, H.A. Klein, J.A. Kitchen, P.A. Gale, J. Am. Chem. Soc. 138 (2016) 8301–8308. [89] N.J. Knight, E. Hernando, C.J.E. Haynes, N. Busschaert, H.J. Clarke, K. Takimoto, M. García-Valverde, J.G. Frey, R. Quesada, P.A. Gale, Chem. Sci. 7 (2016) 1600– 1608. [90] S.J. Edwards, I. Marques, C.M. Dias, R.A. Tromans, N.R. Lees, V. Félix, H. Valkenier, Chem. Eur. J. 22 (2016) 2004–2011. [91] H. Li, H. Valkenier, L.W. Judd, P.R. Brotherhood, S. Hussain, J.A. Cooper, O. Jurcˇek, H.A. Sparkes, D.N. Sheppard, A.P. Davis, Nat. Chem. 8 (2016) 24–32. [92] N.L. Mora, A. Bahreman, H. Valkenier, H. Li, T.H. Sharp, D.N. Sheppard, A.P. Davis, A. Kros, Chem. Sci. 7 (2016) 1768–1772.

[93] X. Wei, G. Zhang, Y. Shen, Y. Zhong, R. Liu, N. Yang, F.Y. Al-mkhaizim, M.A. Kline, L. He, M. Li, Z.-L. Lu, Z. Shao, B. Gong, J. Am. Chem. Soc. 138 (2016) 2749–2754. [94] T. Saha, A. Gautam, A. Mukherjee, M. Lahiri, P. Talukdar, J. Am. Chem. Soc. 138 (2016) 16443–16451. [95] J. Kempf, A. Schmitzer, RSC Adv. 6 (2016) 42713–42719. [96] P.-Y. Liu, S.-T. Li, F.-F. Shen, W.-H. Ko, X.-Q. Yao, D. Yang, Chem. Commun. 52 (2016) 7380–7383. [97] N. Chuard, K. Fujisawa, P. Morelli, J. Saarbach, N. Winssinger, P. Metrangolo, G. Resnati, N. Sakai, S. Matile, J. Am. Chem. Soc. 138 (2016) 11264–11271. [98] N. Sakai, S. Matile, J. Am. Chem. Soc. 125 (2003) 14348–14356. [99] Y. Fan, S.R. Kass, Org. Lett. 18 (2016) 188–191. [100] D.D. Ford, D. Lehnherr, C.R. Kennedy, E.N. Jacobsen, J. Am. Chem. Soc. 138 (2016) 7860–7863. [101] C.R. Kennedy, D. Lehnherr, N.S. Rajapaksa, D.D. Ford, Y. Park, E.N. Jacobsen, J. Am. Chem. Soc. 138 (2016) 13525–13528. [102] A.R. Choudhury, S. Mukherjee, Chem. Sci. 7 (2016) 6940–6945. [103] J. Wen, R. Tan, S. Liu, Q. Zhao, X. Zhang, Chem. Sci. 7 (2016) 3047–3051. [104] M. Zurro, S. Asmus, J. Bamberger, S. Beckendorf, O.G. Mancheño, Chem. Eur. J. 22 (2016) 3785–3793. [105] T. Fischer, J. Bamberger, O.G. Mancheño, Org. Biomol. Chem. 14 (2016) 5794– 5802. [106] S.B.Y. Cotelle, A.-J. Avestro, T.R. Ward, N. Sakai, S. Matile, Angew. Chem. Int. Ed. 55 (2016) 4275–4279. [107] L. Liu, Y. Cotelle, A.-J. Avestro, N. Sakai, S. Matile, J. Am. Chem. Soc. 138 (2016) 7876–7879. [108] C. Wang, F.N. Miros, J. Mareda, N. Sakai, S. Matile, Angew. Chem. Int. Ed. 55 (2016) 14422–14426. [109] D.G. Liz, A.M. Manfredi, M. Medeiros, R. Montecinos, B. Gómez-González, L. Garcia-Rio, F. Nome, Chem. Commun. 52 (2016) 3167–3170. [110] Q.-Q. Wang, S. Gonell, S.H.A.M. Leenders, M. Dürr, I. Ivanovic´-Burmazovic´, J.N. H. Reek, Nat. Chem. 8 (2016) 225–230. [111] W. Cullen, M.C. Misuraca, C.A. Hunter, N.H. Williams, M.D. Ward, Nat. Chem. 8 (2016) 231–236. [112] V. Marcos, A.J. Stephens, J. Jaramillo-Garcia, A.L. Nussbaumer, S.L. Woltering, A. Valero, J.-F. Lemonnier, I.J. Vitorica-Yrezabal, D.A. Leigh, Science 352 (2016) 1555–1559. [113] M. Hirai, M. Myahkostupov, F.N. Castellano, F.P. Gabbaï, Organometallics 35 (2016) 1854–1860. [114] I.-S. Ke, M. Myahkostupov, F.N. Castellano, F.P. Gabbaï, J. Am. Chem. Soc. 134 (2012) 15309–15311. [115] Z.-X. Liu, Y. Sun, Y. Feng, H. Chen, Y.-M. He, Q.-H. Fan, Chem. Commun. 52 (2016) 2269–2272. [116] H. Mehdi, H. Pang, W. Gong, M.K. Dhinakaran, A. Wajahat, X. Kuang, G. Ning, Org. Biomol. Chem. 14 (2016) 5956–5964. [117] J. Sahoo, R. Arunachalam, P.S. Subramanian, E. Suresh, A. Valkonen, K. Rissanen, M. Albrecht, Angew. Chem. Int. Ed. 55 (2016) 9625–9629. [118] B. Chowdhury, S. Sinha, P. Ghosh, Chem. Eur. J. 22 (2016) 18051–18059. [119] F. Bartoli, A. Bencini, L. Conti, C. Giorgi, B. Valtancoli, P. Paoli, P. Rossi, N.L. Bris, R. Tripier, Org. Biomol. Chem. 14 (2016) 8309–8321. [120] A. Akdeniz, T. Minami, S. Watanabe, M. Yokoyama, T. Ema, P. Anzenbacher, Chem. Sci. 7 (2016) 2016–2022. [121] A. Akdeniz, M.G. Caglayan, P. Anzenbacher, Chem. Commun. 52 (2016) 1827– 1830. [122] M. Más-Montoya, D. Curiel, C. Ramírez de Arellano, A. Tárraga, P. Molina, Eur. J. Org. Chem. 2016 (2016) 3878–3883. [123] A.V. Desai, B. Manna, A. Karmakar, A. Sahu, S.K. Ghosh, Angew. Chem. Int. Ed. 55 (2016) 7811–7815. [124] R. Custelcean, N.J. Williams, C.A. Seipp, A.S. Ivanov, V.S. Bryantsev, Chem. Eur. J. 22 (2016) 1997–2003. [125] H. Zhou, T. Yamada, N. Kimizuka, J. Am. Chem. Soc. 138 (2016) 10502–10507. [126] A. Nonat, C.F. Chan, T. Liu, C. Platas-Iglesias, Z. Liu, W.-T. Wong, W.-K. Wong, K.-L. Wong, L.J. Charbonnière, Nat. Commun. 7 (2016) 11978.

Please cite this article in press as: P.A. Gale et al., Anion receptor chemistry: Highlights from 2016, Coord. Chem. Rev. (2018), https://doi.org/10.1016/j. ccr.2018.02.005