Counterion influence on the N–I–N halogen bond

Counterions influence three-center halogen bonds differently than coordination bonds of transition metals.


Introduction
Counterions play important roles in chemistry, for example in the modulation of organocatalysis, 1,2 chemical reactions' energy proles, 3,4 metal cluster aromaticity, 5 spectroscopic properties of organic compounds, 6 surfactant micellization, 7 membrane permeation, 8 gelation, 9 supramolecular assemblies, 10-12 crystal structure, 13 and chromatographic separation. 14 A conceivable inuence of counterions on the strength and geometry of halogen bonding has, however, so far barely received any attention. [15][16][17] Halogen bonding has recently been dened, 18 and has attracted vast interest due to its immense potential to provide a novel tool for supramolecular chemistry, 19,20 crystal engineering, 21,22 anion recognition, 23,24 organic synthesis, 25 structural biology, 26 medicinal chemistry and chemical biology, 27 for instance.
Three-center halogen bonds, 15 sometimes also termed coordinative halogen bonds, 28,29 are formed by the simultaneous interaction of an electrophilic halogen (X) with two electron donor functionalities (D). The three interacting atoms may be arranged in an asymmetric [D-X/D] + , or in a symmetric [D/ X/D] + geometry. Of these, the former is best described by a double well, whereas the latter by a single well energy potential. 30 The asymmetric arrangement that encompasses a shorter and stronger covalent bond (D d+ -X), and a weaker and longer halogen bond (D dÀ /X) is more polar than the one possessing a symmetric geometry and, accordingly, a symmetric charge distribution (D dÀ /X d+ / dÀ D). Strong, asymmetric coordination of a counterion may favor the formation of the asymmetric, more polar geometry over the symmetric, less polar one. The asymmetry of the analogous three-center [N-H-N] + hydrogen bond 31 has been partly attributed to the polarizing effect of the counterion. 32 Limbach et al. reported the size and charge delocalization dependent effect of counterions on the symmetry of low barrier hydrogen bonds. 33 Whether a counterion is capable of desymmetrizing the [N-I-N] + halogen bond depends on the intrinsic energetic preference of the halogen bond for the symmetric geometry 30 as well as on the properties of the counterion.
Early solid state 14 N nuclear quadrupole resonance (NQR) and infrared studies suggested 34 that the N-Br bonds of [bis(pyridine)bromine] + are equivalent for its hexauorophosphate salt, but nonequivalent for its perchlorate and tribromide salts. The loss of centrosymmetry in the latter two may be interpreted as a consequence of crystal packing forces in combination with the different coordination strength of various counterions. Alternatively, it may be explained by the highly comparable energies of the symmetric and the asymmetric geometries. However, both interpretations are debatable. Different N-Br bond lengths, 2.120(3)Å and 2.156(3)Å, were reported for the analogous [bis(quinuclidine) bromine] + tetrauoroborate complex in the solid state. 35 For the analogous iodine centered pyridine salts no analogous counterion effects have yet been reported. Such [N-X-N] + complexes are common synthetic reagents, 36-41 whose reactivity is remarkably sensitive to minute structural 42 and environmental changes. 41 In contrast to the detailed knowledge collected on the inuence of environmental factors on the closely related threecenter hydrogen bonds, 32 similar effects on halogen bonds have so far been assessed to a much lesser extent. As part of our ongoing investigation of the three-center halogen bond, 30,43 the inuence of the counterion on the strength and geometry of the interaction in solution and in the solid state is addressed herein.

Results and discussion
[Bis(pyridine)iodine] + complexes 44 and their structurally closely related, geometrically restrained 1,2-bis(pyridin-2-ylethynyl) benzene analogues provide versatile models for the investigation of the structure and properties of three-center halogen bonds, 15,30,38,39,43,45,46 and were therefore utilized in this study. In order to evaluate the counterion's ability to modulate the geometry of the iodine centered, three-center halogen bond, a series of anions of varying size, charge distribution and coordination strength were explored. Besides the spherical, weakly coordinating anions 12 BF 4 À , ClO 4 À , and PF 6 À that have previously been suggested 34,35 to inuence the geometry of the related [bis(pyridine)bromine] + complex, the weakly coordinating SbF 6 À , the moderately coordinating TfO À and TsO À , and the small and strongly coordinating 10,12 NO 3 À and CF 3 CO 2 À were assessed (Fig. 1). To ensure tight coordination, 45 dichloromethane was chosen as solvent for the solution studies. By scavenging the counterion with the bambusuril Bn 12 BU [6], 47,48 the "counterion-free" [N-I-N] + complex was also investigated. As the three-center halogen bond was proposed to be essentially analogous to the coordination bond of silver(I) complexes, 28,29 and as iodine(I) resembles silver(I) in its ionic radius (I + : 1.33Å, Ag + : 1.29Å) and in the linear geometry of its bis-coordination complex, we discuss the properties of the [N-I-N] + complexes 1-I to 8-I in comparison to the corresponding [N-Ag-N] + species 1-Ag to 8-Ag (Fig. 1). For the preparation of [(1,2-bis(pyridin-2-ylethynyl)benzene) iodine] + tetrauoroborate (10), triate (11), and nitrate (12) and their mono-deuterated analogues (10-d to 12-d), published synthetic routes 30 were followed. See ESI † for details.

Counterion coordination in solution
Ion coordination in solution was evaluated by acquiring the translational diffusion coefficient of the positively and negatively charged species of complexes 1-9 (Table 1), using 1 H and 19 F NMR. Tight anion coordination to [bis(pyridine)iodine] + was indicated by the comparable translational diffusion rates of the two ions in the same solution, for each studied system. A signicant difference between the diffusion coefficients of the anionic and cationic species was detected for 9-I. This indicates that [bis(pyridine)iodine] + and BF 4 À of 9-I do not form a tight ion pair. In line with our expectations, the diffusion rate of the trapped BF 4 À of 9-I is lower than that of 1-I, and matches the diffusion rate of the anion scavenging agent Bn 12 BU[6] (D ¼ 5.8 Â 10 À10 m 2 s À1 ) that was added to the solution. The slower diffusion rate of [bis(pyridine)iodine] + of 9-I than that of 1-I is well explained by the increased solvation of charged species as compared to neutral ones in apolar solvents, 49 and by an increased viscosity of the solution upon addition of Bn 12 BU [6]. Diffusion NMR reveals strong coordination of the counterions of the silver centered complexes 1-Ag to 8-Ag as well. [bis(pyridine)iodine] + BF 4 À (1-I), ClO 4 À (2-I), PF 6 À (3-I), SbF 6 À (4-I), TfO À (5-I), TsO À (6-I), NO 3 À (7-I), and CF 3 CO 2 À (8-I), and their geometrically restrained [(1,2-bis(pyridin-2-ylethynyl)benzene) iodine] + BF 4 À (10-I), TfO À (11-I) and NO 3 À (12-I) analogues. The BF 4 À of 9-I was scavenged with dodecabenzylbambus [6]uril (Bn 12 BU [6]). 47 The spectroscopic data of 1-I to 8-I were compared to those of the corresponding silver(I) complexes (X ¼ Ag) 1-Ag to 8-Ag. 15 N NMR is a sensitive tool for investigation of halogen bond formation, 30,43,45,50 and is commonly used for the assessment of solvent effects, protonation and metal coordination of nitrogenous Lewis bases. [51][52][53] The 15 N NMR chemical shis of compounds 1-9 were detected using 1 H, 15 N-HMBC experiments, 54 providing a single signal for each compound ( Table 1). Formation of the [N-I-N] + complex is accompanied by a comparable, larger than 100 ppm chemical shi change (pyridine, d 15N ¼ À67 ppm (ref. 30)) for 1-I to 9-I. The variation of the 15 N NMR chemical shis of 1-Ag to 8-Ag indicates that, in contrast to iodine(I), silver(I) interacts with anions in a coordination-strength dependent manner. Solvent and counterion coordination to silver(I) complexes has previously been reported 52 to have a major inuence on pyridine 15  The complex 8-Ag shows the most deshielded 15 N NMR chemical shi among the silver(I) complexes, which is in line with CF 3 CO 2 À being the strongest coordinating anion of the series. A smaller absolute coordination shi, 41-45 ppm, of the silver(I) complexes possessing strongly coordinating counterions, such as TsO À , CF 3 CO 2 À and NO 3 À , is explained by an efficient charge transfer from the anion to silver(I) through orbital overlap. This decreases the positive charge of silver(I) and consequently the charge transfer from the pyridine nitrogens to silver(I), yielding smaller 15 N NMR coordination shis. The small, #1 ppm, 15 N chemical shi variation of 1-I to 9-I indicates the absence of an analogous direct charge transfer interaction of the counterion to the iodine(I) of [bis(pyridine)iodine] + . Hence, strong direct coordination of a counterion to iodine(I) by orbital overlap does not take place; rather, this interaction remains electrostatic. Slightly lower 15 N NMR chemical shis were observed for 10-I (À165.5 ppm), 11-I (À165.0 ppm) and 12-I (À163.3 ppm), reecting weakened N-I bonds, as a consequence of the steric restraint introduced by the 1,2-diethynylbenzene backbone. 28 The propensity of the counterions of 1-I to 9-I to compete with pyridine for coordination to the empty p-orbital of I + was evaluated by analyses of the 15 N NMR chemical shis of the [bis(pyridine)iodine] + counterion and of the alternative [(pyridine)(counterion)iodine] pyridine geometries, predicted at the B3LYP level using the dichloromethane continuum solvent model (Table 2, see ESI † for details). The 15 N chemical shis of the [bis(pyridine)iodine] + counterion geometries of 1-I to 8-I (N B , Table 2) are predicted to be virtually counterion independent, and comparable in magnitude to those experimentally observed (Table 1). In contrast, considerably counterion dependent shis are predicted for the [(pyridine)(counterion)iodine] complexes (N A , Table 2), with higher shielding of the nitrogen of those encompassing weaker coordinating anions. This suggests that pyridine versus anion exchange does not take place. Computational thermochemical analysis predicts that formation of a [(pyridine)(counterion)iodine] complex is endothermic for complexes that possess weakly and moderately coordinating anions, yet somewhat exothermic for TsO À , NO 3 À , and CF 3 CO 2 À ( Table 2). It should be underlined that the applied thermochemical calculation does not take into account the electrostatic attraction that arises in the electrolyte formed by [bis(pyridine)iodine] + and its counterion in the solution. This attraction, which is also supposed to account for the variation in the experimental 15 N shis for 1-I to 8-I, is estimated to be 20-30 kJ mol À1 (see ESI † for details). With this correction, the DFT results are compatible with the [bis(pyridine)iodine] + complex, being stable for 1-I to

8-I.
Strong ion pairing is supported by the diffusion NMR data. This conclusion is further supported by detection of the [bis(pyridine)iodine] + complex by HR(ESI)MS for all studied systems (see ESI for details †), further corroborating that pyridine versus anion exchange does not take place in solution.
The 15 N NMR shi of 9-I is comparable to that of 1-I, suggesting that the presence or absence of a weakly coordinating counterion does not inuence the electron distribution of the [N-I-N] + halogen bond. The BF 4 À of 9-I was scavenged by addition of an excess (1.8 eq.) of Bn 12 BU [6] to the solution of 1-I with the efficient trapping of the counterion being conrmed by characteristic changes (a) in the 1 H NMR chemical shis of bambusuril 47 (Fig. 2)

Halogen bond symmetry in solution
Isotopic perturbation of equilibrium (IPE) NMR 56 is a technique capable of distinguishing between a static, symmetric geometry and a rapidly interconverting pair of asymmetric geometries in solution. It has been used successfully, for example, in the symmetry elucidation of [N-H/N] + hydrogen bonded 57,58 and [N/X/N] + halogen bonded 30,46 molecular systems, as well as of carbocations 59 and metal chelating complexes. 60 The main advantage of IPE is that it succeeds even when the rapidly interconverting species cannot be "frozen out" due to a shallow energy barrier between them. It requires the analysis of an isotopologue mixture of a compound, which is oen a mixture of its selectively deuterated and nondeuterated analogues. In such solutions, two sets of NMR signals are observed due to the secondary isotope effect on the vibrational frequencies. 61 The chemical shi difference of the signal originating from the deuterated and that of the nondeuterated molecule, most commonly detected at 13 C NMR frequency, is the secondary isotope effect (IE) (eqn (1)), where n is the number of bonds between the deuterium and the observed nucleus, whilst C (D) and C (H) are the 13 C NMR chemical shis of the deuterated and nondeuterated isotopologues. Although the magnitude of the isotope effect is dependent on the distance of the detected nucleus from the position of isotopic substitution, a sizeable n D eq. need not be restricted to a small n. The observed isotope effect, n D obs , consists of intrinsic, n D 0 , and equilibrium, n D eq. , isotope effects (eqn (2)).
The rst component, n D 0 , is the direct consequence of isotope substitution, and is essentially temperature independent. The second component, n D eq. , manifests only in systems that are involved in a dynamic exchange process. Due to its dependence on the equilibrium constant of an exchange process (eqn (3)), n D eq. is temperature dependent.
where K is the equilibrium constant and D is the chemical shi difference of the exchanging species, in this study the chemical shi difference of a selected nucleus in the N d+ -I, and the N dÀ /I d+ forms of pyridine. Accordingly, static ([N/I/N] + ) and dynamic ([N-I/N] + % [N/I-N] + ) halogen bonding geometries are distinguishable in solution by observation of the temperature dependence of their 2 H-induced 13 C NMR isotope effects. 30,56 We applied IPE to evaluate whether the geometry of the threecenter [N-I-N] + halogen bond of [bis(pyridine)iodine] +14,30,46 could be inuenced in solution by counterion coordination. Previously this bond was demonstrated to be static and symmetric in solution in a complex encompassing TfO À as counterion. 30 By acquisition of 13 C { 1 H, 2 H} NMR spectra of isotopologue mixtures of 1-I to 9-I dissolved in CD 2 Cl 2 at various temperatures, the temperature dependence of the isotope shis was determined. 62 Similar to previous studies, 30,45,46,63 the C3 position of the pyridine of 1-I to 9-I showed the largest temperature dependence and provided data with the highest squared correlation coefficients (Fig. 3). The overall smaller temperature dependence of the IEs of the studied [bis(pyridine) iodine] + complexes (Table 3) as compared to that of pyridine pyridinium triate (5-H), which was previously shown to exist as a rapidly interconverting tautomeric mixture, 30,45,46 indicates that 1-I to 9-I are present as static, symmetric species in solution.
Here it should be noted that the nonzero temperature dependence of the intrinsic isotope effects is due to the temperature dependence of the polarity of the solvent, 64 the polarity-alteration of the solvent modulates the charge distribution of pyridine. 65 An efficient electron density transfer between the nitrogen lone pair and adjacent bonds, inuencing the magnitude of the observed isotope effects as well as the amine basicity, has been previously described. 66 Through-space polarization via dipolar interaction has also been reported previously for ethers, for example. 67 The static symmetric [N/I/N] + geometry of 1-I to 9-I is corroborated by the primary dependency of the magnitude of their IEs on the distance of the observed carbon from the position of 1 H-to-2 H substitution, in contrast to those of [bis(pyridine)hydrogen] + triate 5-H (Table 3)   complexes may inuence their synthetic applicability, 36,41,69 encouraging for further studies. The three-center halogen bond is slightly destabilized by the introduction of a geometrical restraint enforcing a somewhat longer than optimal nitrogen-nitrogen distance for the [N/I/N] + bond, thereby decreasing the overlap of the lled nonbonding orbital of the pyridine nitrogens with the empty p-orbital of iodine(I). 30,45 Consequently, optimal N-I bond lengths are obtainable for such a system only upon an energetically penalized adjustment of its covalent backbone. To evaluate whether the counterion is capable of inuencing a weakened three-center halogen bond, [(1,2-bis(pyridin-2-ylethynyl)benzene)iodine] + complexes (Fig. 1) encompassing the weakly coordinating tetrauoroborate (10-I), the moderately coordinating triate (11-I), and the strongly coordinating nitrate (12-I) were assessed using IPE NMR. The overall lower temperature dependence of the IEs of 10-I to 12-I as compared to 11-H (Table 3), and the primary dependence of the temperature coefficients of their IEs on the position of the 1 H-to-2 H substitution indicate that the [N/I/N] + halogen bond of these systems is also static and symmetric. Hence, counterion coordination is incapable of introducing asymmetry into even a slightly weakened three-center halogen bond in solution.

Counterion coordination in silico
For theoretical conrmation of the experimental ndings, the equilibrium geometries of 1-I to 9-I were calculated using density functional theory (DFT) employing the B3LYP exchange and correlation functional. [70][71][72][73] In the thermochemical calculations, the LANL08d and LANL08f 74 basis sets, in conjunction with the LANL2DZ 75,76 effective core potential, were used for I and Sb, respectively; the LANL2DZ 75,76 basis set was used for Ag; Pople's 6-311+G(d,p) 77-79 basis set was used for B, O, N, F and Cl; and Pople's 6-311G(d,p) basis set was used for the remaining atoms. For the estimation of chemical shieldings, single-point calculations at the geometries obtained were performed using the 6-311+G(d,p) basis set 80 for I, and Kutzelnigg's IGLO-III basis set 81 for the remaining atoms. Solvent effects were accounted for by the Polarizable Continuum Model (PCM), 82,83 with CH 2 Cl 2 as the solvent. All calculations were performed using the Gaussian09 program package. 84 The potential energy surfaces (PES) for all the complexes containing counterions are complicated and likely to show a number of local minima for the placement and orientation of the counterion. Care was taken to nd the most favourable conguration for each counterion; however, it can never be fully excluded that there are local minima with slightly lower energies than those presented. The PESs are shallow in the region of interest, so even if a slightly lower minimum geometry was missed this would not affect the conclusions of the thermochemical analysis, shown in Table 2. In good agreement with the spectroscopic data, DFT calculations predict symmetric [bis(pyridine)iodine] + geometry for complexes 1-I to 9-I. The minor, <0.2%, difference in the N-I bond lengths of 8-I (Table 4), for example, is likely insignicant. 85 The N-I bond length is virtually unaffected by the counterion, whereas the N-I-N angle shows some minor variation, yet remains overall linear. The latter deviation from complete linearity is likely due to weak hydrogen bonding of some of the counterions to the H-2 of the pyridines. This interaction has previously been noticed for structurally closely related complexes. 52,86,87 Most of the analogous silver complexes are also linear and symmetric; however, a slight asymmetry along with a signicant distortion from linearity of the N-Ag-N angle is predicted for 5-Ag, 7-Ag and 8-Ag. These complexes encompass small, moderately or strongly coordinating counterions (TfO À , NO 3 À and CF 3 CO 2 À ), and their geometry is likely altered due to the steric requirements of tight counterion coordination that yields neutral, T-shaped and slightly asymmetric species (Fig. 4). Neither small symmetric (BF 4 À , ClO 4 À , PF 6 À , SbF 6 À ) nor larger sterically hindered anions (TsO À ) are capable of direct coordination to silver(I) in these complexes. It should be noted that the corresponding iodine(I) centered complexes 7-I and 8-I do not form a comparable strong iodineoxygen bond, and thus remain linear and bis-coordinate.

Counterion coordination in the solid-state
Single crystals were obtained via slow diffusion of hexane into the dichloroethane solution of the studied complexes under stepwise cooling from 25 C to À20 C. X-ray crystallographic analyses veried that the iodine centered complexes 1-I to 7-I prefer a symmetric, linear, bis-coordinated N-I-N structure in the solid state (Table 4). The counterion has only a minor, <2%, inuence on the N-I bond lengths. A weak coordination of the anions and their consequent negligible effect on the geometry of the threecentered halogen bond in the solid state is in good agreement with the DFT predicted stable, symmetric [bis(pyridine)iodine] + solution geometries. The N-I bond lengths of these complexes are signicantly shorter than the sum of the van der Waals radii of the participating atoms (R XB ¼ 0.64, where R XB ¼ d NI /(I vdW + N vdW ) 88 ), well reecting the unusual strength of the halogen bond 29 of 1-I to 7-I. Importantly, the asymmetric coordination of the counterion (Fig. 5) of these complexes is unable to induce asymmetry in the three-center halogen bond. Whereas the [N/I/N] + interaction has a signicant covalent character, 15,28,30,43,45 the coordination of the counterions to [bis(pyridine)iodine] + is predominantly Coulombic and is weak. Unfortunately, single crystals for complex 8-I were not obtained, despite repeated attempts. Silver(I) centered complexes 1-Ag to 4-Ag that encompass weakly coordinating anions prefer linear, bis-coordinate N-Ag-N geometries in the solid state (Table 4). The interaction of [bis(pyridine)silver] + with these counterions is of ionic character, similar to that of the corresponding [bis(pyridine) iodine] + . In contrast, pseudo-tetracoordinate (3 + 1) dimers were observed for 5-Ag and 6-Ag that possess the moderately coordinating counterions TfO À and TsO À (Fig. 5). X-ray analyses of [bis(pyridine)silver] + nitrate showed a pseudo-square planar N-Ag(NO 3 À ) 2 -N geometry (Fig. 6a, Table 4, 7-Ag), and a tetramer (dimer of dimers) with an extraordinary combination of a pseudo-tetrahedral (2 + 2) N-(Ag)Ag(NO 3 À )-N coordination and an exceptional pseudo-octahedral (5 + 1) N-(Ag)Ag(NO 3 À ) 2 -N coordination (Fig. 6b, Table 4, 7-Ag-2). Despite repeated attempts, single crystals for the [bis(pyridine)silver] + complex with the strongly coordinating counterion CF 3 CO 2 À (8-Ag) were not obtained. X-ray analyses of 7-Ag and 8-Ag were attempted earlier by White et al., 90 who reported the formation of binuclear species containing silver nitrate and pyridine in a 1 : 3, and silver triuoroacetate and pyridine in a 3 : 2, ratio. DFT predicts the tris-coordinate complexes of 7-Ag and 8-Ag to be energetically more favorable than their bis-coordinate analogues. The strong interaction of the counterion of complexes 5-Ag to 8-Ag with silver(I) manifests in specic, close interatomic contacts, i.e. d Ag-O ¼ 2.705Å (5-Ag, XRD), 2.545Å (6-Ag, XRD), 2.690Å (7-Ag, DFT), and 2.749Å (8-Ag, DFT). The counterion coordination to silver(I) in these species is strong (R coord ¼ 0.79-0.85, where Table 4 Computationally predicted and X-ray crystallographically determined N-X bond distances and N-X-N bond angles for the complexes 1-I to 9-I and 1-Ag to 8-Ag

Anion Structure
Computationally predicted distances and angles X-ray crystallographic distances and angles r(N-X) 1 (Å) r(N-X) 2 (Å) s(N-X-N) ( ) r(N-X) 1 (Å) r(N-X) 2 (Å) s(N-X-N) ( )  The DFT-predicted equilibrium geometries of complexes 8-Ag and 8-I are shown on the left and right, respectively. Whereas silver(I) is predicted to form T-shaped, tris-coordinate species with a strongly bound counterion, short iodine-oxygen contact is not seen for the corresponding iodine(I) complex, which prefers a linear, bis-coordinate N-I-N geometry.  29,88 ), yet it remains weaker as compared to the coordination of pyridines (R coord ¼ 0.67-0.69).

Conclusions
Counterions do not inuence the intrinsically preferred linear, centrosymmetric geometry of the [N/I/N] + halogen bond, either in solution or in the solid state. Neither does the scavenging of the counterion inuence the symmetry of the threecenter bond. In contrast, the weakly coordinating counterions BF 4 À and ClO 4 À were proposed 34,35 to be capable of desymmetrizing the [N/Br/N] + halogen bond of [bis(pyridine)bromine] + in the solid state. This observation may be explained by the lower energy gain upon formation of the [N/Br/N] + as compared to the [N/I/N] + halogen bond. 43 The three-center halogen bonds of 1-I to 8-I resemble the classical coordination bond of the corresponding [N/Ag/N] + complexes. 28 Whereas moderately and strongly coordinating counterions, e.g. TfO À , TsO À , NO 3 À and CF 3 CO 2 À , do not affect the [N/I/N] + halogen bond, they have a considerable inuence on the corresponding coordinative silver(I) complexes. Hence, the latter are altered from linear, bis-coordinate [bis(pyridine)silver] + counterion complexes to multi-coordinate [bis(pyridine)(counterion)silver] species. These counterions enforce nonlinear, and occasionally also asymmetric geometries, of the [N/Ag/N] + coordination bonds. This is most apparent in the solid state (Table 4, Fig. 5  and 6), and yet it takes place in solution (Fig. 4) as well.
Overall, the three-center [N/I/N] + and [N/Ag/N] + bonds behave similarly in the presence of weakly coordinating anions (1-I to 4-I, and 1-Ag to 4-Ag), but differently when moderately or strongly coordinating counterions are present (5-I to 8-I, and 5-Ag to 8-Ag). Our results suggest that care should be taken when the three-center halogen bond is discussed in terms of coordination bonds of transition metals. The [N/I/N] + bonds are best described as halogen bonds possessing a signicant charge transfer character, 43 which is present for all types of halogen bonds but yet may be of varying importance. [91][92][93] Despite some differences, 15 the three-center [D/X/D] + halogen bond resembles a "short-strong" [D/H/D] + hydrogen bond, 94 and may alternatively be termed a short, strong halogen bond.
[Bis(pyridine)iodine] + complexes are common synthetic reagents for halogenation and oxidation, 95-103 with Barluenga's reagent being the most famous one. 36,41 The latter was recently shown to exhibit a counterion dependent order-disorder phase transition in the solid state 13 and has received increasing attention for applications in parallel synthesis 104 and protein chemistry, 69 for example. The understanding of the structure and properties of these reagents is therefore of both fundamental and practical importance.