Photoelectron spectroscopy and theoretical studies of anion-π interactions: binding strength and anion specificity.

PCCP View Article Online

Published on 16 December 2014. Downloaded by University of California - San Francisco on 18/12/2014 13:34:17.

PAPER

Cite this: DOI: 10.1039/c4cp04687k

View Journal

Photoelectron spectroscopy and theoretical studies of anion–p interactions: binding strength and anion specificity† Jian Zhang,ab Bin Zhou,a Zhen-Rong Sun*a and Xue-Bin Wang*b Proposed in theory and then their existence confirmed, anion–p interactions have been recognized as new and important non-covalent binding forces. Despite extensive theoretical studies, numerous crystal structural identifications, and a plethora of solution phase investigations, anion–p interaction strengths that are free from complications of condensed-phase environments have not been directly measured in the gas phase. Herein we present a joint photoelectron spectroscopic and theoretical study on this subject, in which tetraoxacalix[2]arene[2]triazine 1, an electron-deficient and cavity self-tunable macrocyclic, was used as a charge-neutral molecular host to probe its interactions with a series of anions with distinctly different shapes and charge states (spherical halides Cl , Br , I , linear thiocyanate SCN , trigonal planar nitrate NO3 , pyramidic iodate IO3 , and tetrahedral sulfate SO42 ). The binding energies of the resultant gaseous 1 : 1 complexes (1Cl , 1Br , 1I , 1SCN , 1NO3 , 1IO3 and 1SO42 ) were

directly

measured

experimentally,

exhibiting

substantial

non-covalent

interactions

with

pronounced anion-specific effects. The binding strengths of Cl , NO3 , IO3 with 1 are found to be strongest among all singly charged anions, amounting to ca. 30 kcal mol 1, but only about 40% of that between 1 and SO42 . Quantum chemical calculations reveal that all the anions reside in the center of the cavity of 1 with an anion–p binding motif in the complexes’ optimized structures, where 1 is seen to be able to self-regulate its cavity structure to accommodate anions of different geometries and threedimensional shapes. Electron density surface and charge distribution analyses further support anion–p Received 15th October 2014, Accepted 4th December 2014

binding formation. The calculated binding energies of the anions and 1 nicely reproduce the experimentally estimated electron binding energy increase. This work illustrates that size-selective photoelectron

DOI: 10.1039/c4cp04687k

spectroscopy combined with theoretical calculations represents a powerful technique to probe anion–p interactions and has potential to provide quantitative guest–host molecular binding strengths and unravel

www.rsc.org/pccp

fundamental insights in specific anion recognitions.

Introduction Non-covalent interactions are ubiquitous in nature and play a dominant role in many areas at the forefront of modern chemistry,1 among which hydrogen bonding, electrostatic, hydrophobic, p–p stacking, and cation–p interactions and van der Waals forces are well recognized.2–4 The omnipresence of anions and aromatic molecules in biological systems and material sciences, and the importance of anion recognition in various biological functions have stimulated intensive research a

State Key Laboratory of Precision Spectroscopy, East China Normal University, Shanghai 200062, China. E-mail: [email protected] b Physical Sciences Division, Pacific Northwest National Laboratory, 902 Battelle Boulevard, P. O. Box 999, MS K8-88, Richland, Washington 99352, USA. E-mail: [email protected] † Electronic supplementary information (ESI) available. See DOI: 10.1039/ c4cp04687k

This journal is © the Owner Societies 2014

activities focusing on intermolecular interactions between anions and electron-deficient arenes. Four kinds of interactions, i.e., C–H hydrogen bonding, weakly covalent s interaction, strongly covalent or Meisenheimer complex, and the non-covalent anion–p interaction, have been identified and proposed.5–14 The anion–p interaction, in close analogy to the widely studied cation–p interaction,3 has particularly received a lot of attention as a new supramolecular bond.15–19 Earlier reports,5,6 in particular three computational studies published in 2002,20–22 suggested the existence of strong attractive binding motifs between anions and highly p-acidic aromatic systems, in which anions lie above the planes of p systems, and interactions are primarily dominated by electrostatic interactions between anions and positive quadrupole moments/anion-induced polarizations of electron-deficient arenes. Since then, more theoretical studies have been devoted to understanding the specifics of different anions and the nature of p systems (such as perfluorobenzene,

Phys. Chem. Chem. Phys.

View Article Online


Paper

PCCP

1,3,5-triazine, 1,3,5-tricyanobenzene, and 1,2,4,5-tetracyanobenzene) in anion–p interactions,10,23–27 and to explore cooperative interactions between anion–p and other non-covalent interactions, as well as solvent-driven structural changes.28–33 X-ray crystal structure survey of the Cambridge Structure Database (CSD) revealed the existence of anion–p interactions,9,10,17,19,34–36 and binding strengths have been evaluated in solutions.37–42 The relevant anion–p interactions in biological systems and evidence for its biological functions have been recognized recently as well.43–45 Anion–p interactions have also been exploited in fluoride ion sensing46 and forming anion/naphthalenediimide charge-transfer complexes.47,48 Despite these significant progresses, indisputable evidence for the pure and exclusive anion–p attraction is still rare. The existence of anion–p interactions in solids and crystals is complicated by possible multiple binding motifs, counter ion, H-bonding, etc. Therefore theoretical structural criteria and electron density surface are proposed to help identify true anion–p bonding,9,10,19 in which anions should be positioned right above the centroid of electron deficient p molecules or groups; and there should be a very small electron density, little mixing of orbitals, and a small charge transfer between anions and p molecules. In contrast to numerous theoretical studies,20–33 extensive X-ray structure identification/verification9,10,17,19,34–36 and a plethora of solution phase binding strength measurements,37–42 only a handful of gas phase studies have been reported in probing anion–substituted benzene complexes using high pressure mass spectrometry methods5,6 and vibrational spectroscopy.7,8 In all cases, the C–H anion hydrogen bond motif is favored over anion–p attractions, including the most fluorinated C6F5H.8 To our knowledge, no gas phase spectroscopic work has been reported in studying anion–p system complexes where the anion–p attraction is a dominant binding force, so intrinsic anion–p binding properties have been largely derived from calculations to date. In 2013, Wang and his co-workers reported an elegant experiment,49 in which tetraoxacalix[2]arene[2]triazine 1 (Scheme 1), a member of a new generation of macrocyclic host molecules of the heteroatom bridged calix(hetero)arenes,50–53 can form 1 : 1

complexes with various anions in the gas phase, in solution, and in the solid state. Due to the p-electron-deficient V-shaped cleft composed of two triazine rings, 1 has been shown to behave as a cavity self-tunable versatile electron-neutral macrocyclic host that accommodates various anions with different shapes.49,50 Thus, this set of complex clusters presents a rare, real anion–p chemical system, and provides excellent opportunities for researchers to study specific anion effects in the guest–host chemistry, as well as to characterize this unique selfregulated cavity in anion recognitions beyond conventional planar arene host molecules. Two theoretical studies were subsequently reported to obtain optimized cluster structures and binding interactions of halides with 154 and a similar host of bis(tetraoxacalix[2]arene[2]triazine) with three binding V-shaped cavities.55 However, no gas phase binding energies are available to directly compare with those calculations. Negative ion photoelectron spectroscopy (NIPES) is often used to obtain reliable estimates of cluster X S energies (X and S stand for an anion and a neutral solvent molecule, respectively), since the difference of electron binding energy (EBE) of X S and EBE of X , i.e. DEBE, is generally a good measure of this quantity (Fig. S1, ESI†), given that the interactions in X S are much weaker than those in the ionic complex and are typically small.56 NIPES has been applied to estimate anion binding energies with a variety of neutral molecules, e.g., water,57 organic acids,58 and polyols,59,60 and should also be an ideal technique to study the gas phase characterization of anion–p complexes. Here we report a NIPES study of anion–p interactions using 1 as charge-neutral molecular probe to interact with seven anions (X ) with distinctly different shapes and charge states (X = spherical halides Cl , Br , I , linear thiocyanate SCN , trigonal planar nitrate NO3 , pyramidic iodate IO3 and tetrahedral sulfate SO42 ). First-principles calculations were conducted on the closed-shell anion complexes 1X and the corresponding neutral or singly charged (X = sulfate) radicals 1X . The optimized structures and electron density analyses indicate unequivocally the existence of anion–p binding motifs in these clusters, and the anion–p attractions are found to contribute significantly to the total interaction energies. The good agreement between the calculated and experimentally measured EBEs lends appreciable credence to the complex structures. This study, to the best of our knowledge, accounts for the first spectroscopic probe on anion–p interactions that directly yields gas-phase anion–p interaction strengths and quantifies anion specific effects.

Experimental methods

Scheme 1

Structure of dichloro-substituted tetraoxacalix[2]arene[2]triazine 1.


The NIPES instrument consists of an electrospray ionization source, a cryogenically controlled 3D Paul trap, a modified Wiley–McLaren time-of-flight (TOF) mass spectrometer, and a magnetic-bottle TOF photoelectron analyzer.61 The desired anion–p complexes were readily produced by spraying ca. 0.1 mM aqueous acetonitrile solutions of tetraoxacalix[2]arene[2]triazine 1 mixed with the respective sodium anion salts to afford formation of


View Article Online


PCCP

the 1 : 1 anion–p complexes 1Cl , 1Br , 1I , 1SCN , 1NO3 , 1IO3 and 1SO42 . The generated anions were guided by two radio-frequency-only quadrupoles and a 901 bender into a 3D cryogenic ion trap, where they were cooled by collisions with a cold buffer gas (20% H2 and 80% He) for 20–100 ms, before being pulsed out into the extraction zone of the TOF mass spectrometer with a 10 Hz repetition rate. For each NIPES experiment, the desired anion complexes were first mass-selected and decelerated before being detached by a laser beam (355 nm, 266 nm from a Nd:YAG laser, 193 nm, 157 nm from an excimer laser) in the interaction zone of the magnetic bottle TOF photoelectron analyzer. In all cases, the laser was operated at a 20 Hz repetition rate with the ion beam off at alternating laser shots, affording shot-by-shot background subtraction. The photoelectrons were collected with nearly 100% efficiency by the magnetic bottle, and analyzed by a 5.2 m long electron flight tube. The TOF photoelectron spectra were converted to kinetic energy spectra, calibrated by the known spectra of I and Cu(CN)2 . The EBE spectra were obtained by subtracting the kinetic energy spectra from the detachment photon energies, and had a resolution (DE/kinetic energy) of B2% or 20 meV full width at half maximum (FWHM) for 1 eV kinetic energy electrons. The vertical detachment energy (VDE) was measured from the first resolved spectral maximum assigned to the transition from the ground state of the anion to the ground state of the neutral.

Computational methodology Theoretical calculations were employed to obtain the optimized geometries and calculate VDEs and binding energies (BEs) of the 1X (X = Cl , Br , I , SCN , NO3 , IO3 , SO42 ) complexes. oB97XD62,63 geometrical optimizations were carried out with the 6-311++G(d,p) basis set64 for C, H, O, N, Cl, Br, S atoms, and the aug-cc-pVTZ-PP basis set with the Stuttgart– ¨ln MCDHF RSC ECP (28 core electrons)65 for the I atom. All Ko basis sets were obtained from the EMSL Basis Set Exchange.66 The suitability of using the recently developed oB97XD functional with long-range correlations has been proved in the studies of bis(tetraoxacalix[2]arene[2]triazine)halide complexes.55 All geometries of the anionic complexes were converged without any symmetry constraints. Frequency calculations were also performed at the same level of theory to ensure that the optimized structures are the true minima and to obtain the zero-point energy (ZPE) corrections needed for the binding energy calculations. Theoretical VDE values, defined as the energy differences between the anionic and the corresponding neutral complexes both at the optimized anionic geometries, were obtained from single point energy oB97XD and M06-2X67–69 calculations. The interaction energies, i.e. BEs between the anions and 1, were computed from single point energy oB97XD calculations as well, including ZPE corrections and the basis set superposition error (BSSE) corrections estimated using the counterpoise method of Boys and Bernardi.70 Natural bond orbital (NBO) calculations were used to compute natural atomic charges used in the charge analysis, and charges derived from the electrostatic potential using the Merz–Kollman–Singh (MK) method were carried out as well for


Paper

comparison. All the calculations were performed with the Gaussian 09 program.71

Photoelectron spectra of anion–p complexes and experimentally measured binding strengths between the anions and 1 Fig. 1 shows the low temperature photoelectron spectra of 1 : 1 clusters of 1 with the halides (Cl , Br , I ) and the pseudo halide of SCN obtained at 20 K with 193 nm photons (6.424 eV). The spectrum of 1Cl shows one spectral band peaked at EBE = 5.05 eV, i.e. of VDE = 5.05 eV, followed by an intense rising tail at EBE 4 6 eV. Careful examination reveals a very tiny peak near the baseline with EBE = 3.61 eV (indicated by an arrow in Fig. 1a), which is assigned due to photodetachment of the free Cl ions by the second photons. The free Cl ions were generated in a dissociation process of the 1Cl complex by the first photons in our experiments. The nature of this two-photon process for Cl was confirmed via photon flux studies. This two-photon-generated free anion peak is also observed for other p-halides or pseudo halide complexes (Fig. 1b–d), which provides an internal calibration to show the VDE increase of the anion upon clustering with this electron-deficient cavity 1. The VDE difference of 1Cl (5.05 eV) and Cl (3.61 eV)72 yields a VDE increase of 1.44 eV (Table 1), which can be regarded as a rough estimate of the interaction energy between 1 and Cl . Similarly, photoelectron spectra of 1Br , 1I , and 1SCN (shown in Fig. 1b–d, respectively) each reveal a spectral pattern akin to that of the respective free anion, but with a significant blue-shift in binding energy, followed by a rising tail near 193 nm photon limit. The two dominant peaks at EBE = 4.55 and 5.00 eV in the 1Br spectrum give a gap of 0.45 eV, which is nearly identical to the spin–orbital splitting of 0.457 eV for the Br atom (2P3/2, and 2P1/2),73 as also evidenced by comparing to the two small peaks at EBEs of 3.36 and 3.82 eV for Br , produced via photo-fragmentation of the 1Br complex. The nearly identical intensity ratio and the energy gap for the two peaks within each doublet for 1Br and Br strongly suggest that the atomic Br electronic configuration remains intact in complex 1Br , leading to the conclusion that Br is not covalently bonded to the p molecule, but forms non-covalent interactions with 1. The amount of binding energy as estimated using DEBE is ca. 1.19 eV (Table 1). The photoelectron spectrum of the 1I complex (Fig. 1c) is dominated by two sharp peaks at EBE = 4.00 and 4.93 eV. This 0.93 eV separation is almost equal to the spin–orbital splitting of iodine of 0.943 eV,73 again suggesting a non-covalent interaction between 1 and I in the 1I complex. In this spectrum, the first transition from the ground state of I to the 2P3/2 state of I at EBE = 3.06 eV is discernible, but the peak that corresponds to the formation of 2P1/2 at EBE = 4.00 eV is buried underneath the first strong peak of 1I . DEBE = 0.94 eV indicates a considerably weaker interaction between I and 1, in comparison to that between


View Article Online

Paper

PCCP


The 193 nm photoelectron spectra of 1NO3 and 1IO3 (Fig. 2) show weak features with VDE = 5.25 and 5.90 eV for the nitrate and iodate clusters, respectively, followed by strong rising tails at the higher binding energy end. At 157 nm, these weak bands are better defined, and the features at higher binding energies 46 eV are resolved. Compared to the VDEs of 3.94 eV for NO3 ,72 and 4.77 eV for IO3 ,57 the binding strengths of 1 with NO3 and IO3 are estimated to be 1.31 and 1.13 eV, respectively (Table 1). It is interesting to note that the planar trigonal nitrate anion, albeit its bigger size and diffusive charge distribution, can bind with 1 as strongly as Cl . Considering the significant stabilization on electron binding energy that 1 has brought to various halides, pseudo halide, and polyatomic oxoanions (the weakest stabilization still amounts to more than 0.9 eV for I and SCN , Table 1), it is very tempting to see if 1 can also stabilize the SO42 sulfate dianion, which, as an isolated dianion, is highly unstable with a predicted 1.6 eV negative EBE.74 We successfully produced the 1 : 1 complex of 1SO42 in the gas phase, and its spectra at 355, 266 and 193 nm are shown in Fig. 3. In the 355 nm spectrum, only one peak of VDE B1.7 eV is resolved. At 266 and 193 nm, additional spectral bands at high EBE are exhibited (i.e. those centered at 2.8, 3.8, and B5 eV that correspond to the transitions from the ground electronic state of 1SO42 to the excited states of 1SO4 ). The characteristic cut-off for slow electrons (i.e. those electrons with high EBEs close to the photon energies are suppressed) in each spectrum, a hallmark for photodetachment of multiply charged anions due to the universal existence of the repulsive Coulomb barrier (RCB),75,76 provides undisputable evidence of the doubly charged nature of the complex in Fig. 3. An enormous EBE increase of ca. 3.3 eV (75 kcal mol 1) is found from the lowest EBE feature of 1SO42 and the calculated 1.6 eV of free SO42 , which amounts to about 2.5 times stronger than the strongest EBE stabilization for the singly charged anions, i.e. Cl , NO3 and 1 (Table 1).

Theoretical results and discussion

Fig. 1 Low-temperature (20 K) negative ion photoelectron spectra of 1Cl (a), 1Br (b), 1I (c) and 1SCN (d) at 193 nm. The blue arrows indicate electron binding energies of the respective free anions. The increase in electron binding energy of each anion upon associating with 1 provides a rough estimate of the anion–p interaction strength.

Cl and 1 by ca. 0.5 eV (11.5 kcal mol 1). A similar DEBE shift of 0.97 eV is observed for SCN upon clustering with 1, as indicated by the sharp peak at EBE = 4.51 eV for the complex and the small free SCN peak at 3.54 eV.


We carried out first-principles calculations on these cluster complexes to obtain converged optimized structures. The primary motivation to perform these calculations is to identify the dominant binding motifs and unravel significant anion specific effects that have been observed in the NIPES experiments. Among all these clusters, the crystal structures of 1SCN and 1NO3 are known,49 and provide good references to benchmark the gas-phase structures. We further calculated theoretical VDEs and anion–p interaction in direct comparison with the experimentally measured EBE blue-shifts. The electron density, natural bond orbital (NBO) and Merz–Kollman–Singh (MK) charge distribution analysis for each complex were also conducted to confirm the formation and to better understand the nature of the anion–p interaction. Optimized structures of anion–p complexes The structures of 1X (X = Cl , Br , I , SCN , NO3 , IO3 and SO42 ) complexes were optimized using the oB97XD functional


View Article Online

PCCP

Paper


Table 1 Experimental and theoretical VDEs of the complexes, free anions, the VDE increase of each complex relative to the corresponding free anion, and the binding energy (BE) of the anion and p molecule 1 with the BSSE corrected values in parentheses (all unit in eV)

Cmpd

Expt.

1Cl 1Br 1I 1SCN 1NO3 1IO3 1SO42

5.05 4.55 4.00 4.51 5.25 5.90 B1.7

0.05 0.05 0.02 0.02 0.10 0.10

oB97XD

M06-2X

5.03 4.64 4.16 4.53 5.53 5.91 1.08

5.18 4.75 4.10 4.63 5.94 6.10 1.36

Free anion 3.61a 3.36a 3.06a 3.54a 3.94a 4.77b 1.6c

DVDE (expt.)d

DVDE (theo.)e

BE (BSSE) f

1.44 1.19 0.94 0.97 1.31 1.13 3.3

1.42 1.28 1.10 0.99 1.59 1.14 B2.7

1.29 1.15 1.03 1.14 1.33 1.40 3.16

(1.24) (1.13) (1.00) (1.07) (1.22) (1.29) (2.94)

a Ref. 72. b Ref. 57. c Ref. 74. d DVDE (expt.) = VDE (1X ) VDE (X ) from experimental measurements. e DVDE (theo.) = VDE (1X ) VDE (X ) using the oB97XD predicted value for 1X . f BE = E(1X ) E(1) E(X ) at each species’ optimized geometry with ZPE corrections. The BSSE corrected values are in parentheses.

with the aug-cc-pVTZ-PP basis set for I and 6-311++G(d,p) for all other atoms. The top and side views of these structures are shown in Fig. 4 and Fig. S2 (their geometric coordinates and energies are given in the ESI†). For 1halide complexes, the atomic anions are positioned in the center of the V-shaped cleft of the two triazine rings with a large offset distance to the centroid of the p-electron deficient triazine ring, decreasing from 1.30, 1.23, 1.15 Å with a concomitant increase in the distance to the triazine plane, i.e., 3.57, 3.74, 3.97 Å, for halides Cl , Br , I , respectively. The linear anion SCN is positioned in the V-shaped cleft with an angle of 311 along the Cl–Cl axis. The distance of the S atom and N atom to their respective adjacent triazine ring planes are 3.45 and 2.93 Å with offset values of 0.52 and 0.60 Å. As for the triangular anion NO3 , one oxygen atom is positioned above one of the triazine rings with a distance of 2.91 Å to the plane and an offset of 0.92 Å. The other two oxygen atoms are parallel to the other triazine plane and form a weak s-interaction. Like the NO3 anion, the pyramid IO3 anion is also located in the middle of the V-shaped cleft with one oxygen atom above one triazine (with a distance of 2.81 Å to one triazine and an offset of 0.34 Å), and the other two oxygen atoms are parallel to the other triazine ring. As for the SO42 dianion, one oxygen atom is positioned above a triazine ring and another two parallel to the other ring. The fourth oxygen is pointing away from the electron deficient p system. The distance of the first oxygen to the triazine plane and the offset value are 2.70 Å and 0.71 Å, respectively. Depending on different guest anions accommodated in the cavity, the calculated distance between the two upper-rim C atoms in the triazine rings followed a decreased order of 9.15 Å (free host 1), 8.88 Å (1SCN ), 8.82 Å (1NO3 ), 8.81 Å (1IO3 ), 8.72 Å (1I ), 8.66 Å (1SO42 ), 8.62 Å (1Br ), to 8.56 Å (1Cl ), and concomitantly the dihedral angle of the two triazine rings decreased from 119.51 (free host 1) to 108.01 (1SCN ), 107.41 (1IO3 ), 107.21 (1NO3 ), 103.61 (1I ), 101.41 (1SO42 ), 100.31 (1Br ), and 98.31 (1Cl ). The obvious variations in shape and size of the cleft formed by two opposing triazine rings evidence 1 as a versatile and self-tunable anion host. The optimized geometric structures of 1SCN and 1NO3 are comparable with the crystallographic data.49 For the 1SCN complex, the distances to the triazine centroid and plane are 2.99 and 2.93 Å from N, and 3.49 and 3.45 Å from S in our


calculations, close to the respective crystallographic data of 3.050 and 3.041 Å, and 3.643 and 3.598 Å. The calculated distance between the two upper-rim C atoms in the triazine rings is 8.88 Å, compared to 9.055 Å in the crystal. For the 1NO3 complex, the calculated distances from the O atom to the triazine ring centroid and plane are 3.06 and 2.91 Å, and the distance between the two upper-rim C atoms in the triazine rings is 8.82 Å, which are in good accord with the respective values reported for the crystal, i.e., 3.084, 2.953 and 8.964 Å. Nevertheless, the calculated distances for the gas phase clusters are found to be slightly smaller than the respective values in the crystals, a fact that was noted in a previous theoretical report,55 and might be caused by crystal packing effects. Calculated vertical detachment energies (VDEs) and binding energies of the anion–p complexes Single point calculations using two different functionals, oB97XD and M06-2X with the basis set of aug-cc-pVTZ-PP for I and 6-311++G(d,p) for all other atoms, were performed on the optimized cluster structures, and the calculated VDEs are listed in Table 1. Both methods, in general, give similar results and reproduce experimental VDEs, but oB97XD slightly outperforms M06-2X in calculating VDEs for all the complexes except 1SO42 in comparison with experimental values (the largest deviation of ca. 0.25 eV); while for the doubly charged complex 1SO42 , the latter method yields a VDE that is ca. 0.4 eV smaller than the experimental value. The generally good agreement between the experiments and calculations on VDE (and therefore on DVDE) further validates the theoretical methodologies that are used in this study and lends appreciable credence for the optimized structures. The amount of VDE increase of each anion upon associating with 1 obtained from the experiments also agrees with the calculated 1X binding energy reasonably (Table 1). Both exhibit that sulfate has a far stronger interaction with 1 among all the anions investigated here, presumably due to the doubly charged nature of this dianion, and that NO3 , IO3 and Cl have strongest binding with 1 among all the singly charged anions. Our calculated binding energies of 1Cl and 1Br , i.e., 1.29 and 1.15 eV, compare well with those reported in a previous calculation, 1.30 and 1.22 eV.54 Large binding energies like these reported in this paper were also predicted


View Article Online


Paper

PCCP

Fig. 3 Low-temperature (20 K) negative ion photoelectron spectra of 1SO42 at 355 (a), 266 (b) and 193 nm (c). The blue arrow indicates the electron binding energy of the sulfate free dianion.74 The increase in the electron binding energy of sulfate upon associating with 1 provides a rough estimate of the sulfate–p interaction strength.

Electron density and charge analysis of the anion–p clusters

Fig. 2 Low-temperature (20 K) negative ion photoelectron spectra of 1NO3 (a, b) and 1IO3 (c, d) at 193 nm and 157 nm. The blue arrows indicate electron binding energies of the respective free anions from ref. 57 and 72. The increase in electron binding energy of each anion upon associating with 1 provides a rough estimate of the anion–p interaction strength.

theoretically for induction-driven anion–p interactions in other electron-deficient arenes.32 It is also interesting to note that solution phase measurements reached the same conclusion, that is, NO3 and Cl have the largest association constants with 1 among various singly charged anions studied there.49 The general agreement among the experimentally measured VDE increase, calculated clustering binding energy, and solution phase association constant illustrates that NIPES is, indeed, a powerful technique to probe anion–p interactions in the gas phase without the complications of condensed-phase environments, and can provide quantitative anion–p interaction strength.


The maximum electron density between the two bonded entities, rmax, provides a quantitative measure of the degree of covalency.10 Covalent bonds generally have rmax values 40.1 e Å 3. Table 2 shows the renderings of electron density surfaces (plotted by the Multiwfn 2.2.1 software77 using the oB97XD calculated results) and lists the maxima of the electron density for the 1X complexes between the triazine rings and the anions; whereas the corresponding electron density analyses focusing on hydrogen bonds between aryl C–H and the anion are provided in the ESI† (Table S3). The electron densities between the triazine rings in 1 and all the anions studied here are very small, no bigger than 0.1 e Å 3. Therefore, it can be confirmed that the interaction between the anions and triazines is non-covalent and dominated by anion–p interactions. It is worth noting that the electron densities for the H-bonds are all larger than the anion–p interactions, due to the partially covalent nature of the hydrogen bond.78–80 Natural bond orbital (NBO) and Merz–Kollman–Singh (MK) charge distributions of the anionic 1X and neutral 1X complexes are listed in Table 3. NBO charge distributions


View Article Online

Paper


PCCP

Fig. 4 Top (left) and side view (right) of oB97XD/6-311++G(d,p) for C, H, O, Cl, N, S and aug-cc-pVTZ-PP for I geometries of 1Cl , 1SCN , 1NO3 , 1IO3 , and 1SO42 structures. The optimized structures for 1Br and 1I are shown in Fig. S2 (ESI†).

indicate that more than 90% of the extra charge resides on the anion, and less than 10% is transferred onto the p molecule 1 in all anionic complexes (Table 3), while the MK charge distributions suggest approximate 80% of the charge resides on the anion. This is another characteristic trait exhibiting significant anion–p interactions between 1 and the anions studied here. The photo-detached electrons are also found from the respective anions, as confirmed by the NBO and MK


charge distributions between the anion complexes and the corresponding neutral complexes, as well as by the photoelectron spectra (Fig. 1–3) which show similar spectral patterns to those of the free anions. Our finding is in accordance with the previous prediction that the charge transfer is very small in anion–p complexes.10 But what disagrees with the previous prediction is that the charge transfer in our experiments is bigger than they should


View Article Online

Paper

PCCP


Table 2 Electron density of the complexes with rmax listed in e/a03 (e Å 3). The density is plotted in the symmetry plane cut along Cl–Cl. The second electron density of 1SCN consists of two pieces focusing on the N end-p and S end-p interactions, respectively, displayed in the section cut along N(S) and the centroid of one triazine close to the N(S) end

be for ‘pure anion–p interactions’. Both anion–p interactions and H-bonds contribute to the charge transfer from anions to 1. It should be pointed out that the two hydrogen atoms in the two benzene rings are very close to the anion. For example, the average distances for anion Cl , Br , and I to the hydrogen atoms are 2.38, 2.57 and 2.83 Å, respectively, which are near enough to form C–H halide hydrogen bonds. For polyatomic oxyanions NO3 , IO3 and SO42 , two oxygen atoms are close to the aryl H atoms with an average distance of 2.13, 2.13 and


1.92 Å, respectively. It can be confirmed that H bonds are formed in all of these six complexes. The formation of H-bonds results in some additional charge transfer between the anions and 1. In 1SCN , the distances of N H–C and S H–C are 2.32 and 2.74 Å, respectively, too long to form significant C–H anion H bonds, which is consistent with the fact that the charge transfer in this complex is the smallest (NBO analyses in Table 3). This observation is also in accord with the recent study by Wang and Wang,49 which claimed


View Article Online

PCCP

Paper

Table 3 The natural bond orbital and Merz–Kollman–Singh (in parentheses) charge distribution of the anionic 1X and neutral (or one less charged) 1X complexes both calculated at the anions’ geometries

1X

1X


1 Cl /Cl Br /Br I /I SCN /SCN NO3 /NO3 IO3 /IO3 SO42 /SO4

that the 1SCN interactions.

0.073 0.069 0.063 0.038 0.063 0.055 0.162

( ( ( ( ( ( (

X 0.170) 0.164) 0.179) 0.216) 0.249) 0.234) 0.595)

0.927 0.931 0.937 0.962 0.937 0.945 1.838

( ( ( ( ( ( (

complex is exclusively formed by anion–p

Comparison of H-bonding and anion–p interactions in the 1anion complexes The 1X complexes in our experiment are conceivably composed of two main interactions, that is, anion–p interactions and aryl C–H anion hydrogen bonds. Which is more important for 1 in binding the anion, and what is the relationship between the two interactions? Here we divide 1 in each 1X complex into two subsystems, one consisting of two facing benzenes (2) and the other is two 6-chloro-1,3,5-triazine-2,4diols (3) with their geometries taken out from the optimized complex structures. The two hydroxy substitutions (with both OHs opposite to A with a A C–O–H dihedral angle of 1801, where A = Cl/Br/I, or S/N in SCN or O in oxyanions, and the O–H bond length of 0.9632 Å) on triazine are set up to intend to keep the hydrogen atoms away from the anion to avoid H-bond contributions in the subsystem 3. We then computed the VDE and NBO charge distribution of 2X and 3X subsystems, and compared these with the 1X complexes to discuss the anion–p and C–H anion H bond effects. VDE increases by anion–p and C–H anion H bonds. We calculated VDE increase of each anion upon association with two reference subsystems. The DVDE is defined as the VDE of the subsystem complex (2X or 3X ) relative to the VDE of the isolated anion X . The DVDEs in both subsystems are listed in Table 4. In all cases, the anion–p interaction (3X ) increases the electron binding energy less than the H-bond does, suggesting that the hydrogen bonding contribution might overcome the anion–p interaction in these complexes.

Table 4 The VDE increase (in eV) and NBO charge transfer (|e|) in two subsystems emphasizing the anion–p (3) and aryl C–H anion hydrogen bonding interactions (2)

3X (anion–p)

2X (H-bond)

Anion

DVDE [NBO]

DVDE [NBO]

Cl Br I SCN NO3 IO3 SO42

0.21 0.21 0.18 0.13 0.48 0.27 1.23

0.66 0.60 0.52 0.40 0.82 0.48 1.53

[0.009] [0.009] [0.007] [0.015] [0.026] [0.016] [0.128]


[0.055] [0.049] [0.041] [0.025] [0.040] [0.030] [0.085]

X

1 0.830) 0.836) 0.821) 0.784) 0.751) 0.766) 1.405)

+0.003 0.005 0.008 0.003 0.013 0.002 0.082

( 0.081) ( 0.075) ( 0.082) ( 0.086) (+0.000) (+0.048) ( 0.325)

0.003 +0.005 +0.008 +0.003 +0.013 +0.002 0.918

(+0.081) (+0.075) (+0.082) (+0.086) (+0.000) ( 0.048) ( 0.675)

NBO charge transfer induced by anion–p interactions and C–H hydrogen bonding. Table 4 also lists the charge transfer from the guest anion to the hosts 3 and 2 due to pure anion–p and C–H hydrogen bonding interactions, respectively. It can be seen that for the singly charged anions’ series, H-bond interaction induces significantly more charge transfer than that due to anion–p interactions, especially for atomic halides. Among all H-bond induced charge transfers, SCN has the smallest value. This is because SCN is a linear anion located diagonally in the V-shaped cleft of the two triazine rings, which means S and N are far away from the aryl H atoms, so that the H-bond is relatively weak in this complex, as discussed above. However, for doubly charged anion complexes 1SO42 , the charge transfer due to anion–p interaction in 3SO42 is much bigger than that for any other 3X , even bigger than that in 2SO42 due to H-bond interactions. This indicates the anion–p interaction is very sensitive to the anion charge state, which explains the extremely strong interactions of sulfate with 1.

Conclusion and summary The anion–p interactions of tetraoxacalix[2]arene[2]triazine 1 with Cl , Br , I , SCN , NO3 , IO3 and SO42 have been investigated both experimentally and theoretically. Gas phase negative ion photoelectron spectra of 1Cl , 1Br , 1I , 1SCN , 1NO3 , 1IO3 , and 1SO42 have been obtained at laser wavelengths of 355, 266, 193, and 157 nm. The VDEs of these complexes are calculated with the oB97XD and M06-2X methods. The optimized geometries show that all of the anions are located in the center of the V-shaped cleft of the two triazine rings. The electron density between the anion and triazine rings is very small, which confirms that the 1X interaction is non-covalent, and significantly originates from anion–p interactions. NBO and MK charge distribution analysis indicates there is charge transfer from the anions to 1. Further analysis by dividing 1 into two subsystems suggests that charge transfer is mainly from the H-bond contribution in the singly charged anion clusters and significantly from the anion–p interaction in the doubly charged complex. However, the anion–p interaction, albeit significant, contributes less than the hydrogen bonding in all cases to the total binding energy of the complexes. Both experimentally measured DVDE and computationally calculated total interaction strength indicate that Cl , NO3 and IO3 can bind very strongly with tetraoxacalix[2]arene[2]triazine, in good


View Article Online


Paper

accordance with previous studies. The association energy of 1 and sulfate is enormous and stabilizes the electrons on the sulfate dianion by as much as 3.3 eV. This work quantifies the pronounced anion specific effects in anion–p interactions as well as showcases the potential of applying NIPES to probe noncovalent intermolecular interactions.

Conflicts of interest The authors declare no competing financial interest.

Acknowledgements We thank Prof. Mei-Xiang Wang (Tsinghua University) and Prof. Haibo Yang (East China Normal University) for providing us the tetraoxacalix[2]arene[2]triazine sample. The NIPES research at PNNL was supported by the U. S. Department of Energy (DOE), Office of Science, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences & Biosciences (X.-B.W.), and was performed at EMSL, a national scientific user facility sponsored by DOE’s Office of Biological and Environmental Research and located at Pacific Northwest National Laboratory. The theoretical study performed at ECNU was supported by the National Natural Science Foundation of China (No. 11474096).

References 1 J.-M. Lehn, Supramol. Chem., VCH Verlag. mbH, Weinheim, Germany, 1995. 2 G. A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford University Press, New York, 1997. 3 (a) J. C. Ma and D. A. Dougherty, Chem. Rev., 1997, 97, 1303; (b) D. A. Dougherty, Acc. Chem. Res., 2013, 46, 885; (c) D. A. Dougherty, Science, 1996, 271, 163. 4 (a) H.-J. Schneider, Angew. Chem., Int. Ed., 2009, 48, 3924; (b) L. M. Salonen, M. Ellermann and F. Diederich, Angew. Chem., Int. Ed., 2011, 50, 4808. 5 K. Hiraoka, S. Mizuse and S. Yamabe, J. Phys. Chem., 1987, 91, 5294. 6 G. J. C. Paul and P. Kebarle, J. Am. Chem. Soc., 1991, 113, 1148. 7 Z. M. Loh, R. L. Wilson, D. A. Wild, E. J. Bieske and A. Zehnacker, J. Chem. Phys., 2003, 119, 9559. 8 H. Schneider, K. M. Vogelhuber, F. Schinle and J. M. Weber, J. Am. Chem. Soc., 2007, 129, 13022. 9 O. B. Berryman, V. S. Bryantsev, D. P. Stay, D. W. Johnson and B. P. Hay, J. Am. Chem. Soc., 2007, 129, 48. 10 B. P. Hay and V. S. Bryantsev, Chem. Commun., 2008, 2417. 11 M. Giese, M. Albrecht, T. Krappitz, M. Peters, V. Gossen, G. Raabe, A. Valkonen and K. Rissanen, Chem. Commun., 2012, 48, 9983. 12 P. Gamez, T. J. Mooibroek, S. J. Teat and J. Reedijk, Acc. Chem. Res., 2007, 40, 435.


PCCP

13 O. B. Berryman and D. W. Johnson, Chem. Commun., 2009, 3143. 14 L. M. Salonen, M. Ellermann and F. Diederich, Angew. Chem., Int. Ed., 2011, 50, 4808. 15 P. Gamez, Inorg. Chem. Front., 2014, 1, 35. 16 B. L. Schottel, H. T. Chifotides and K. R. Dunbar, Chem. Soc. Rev., 2008, 37, 68. 17 H. T. Chifotides and K. R. Dunbar, Acc. Chem. Res., 2013, 46, 894. 18 A. Frontera, P. Gamez, M. Mascal, T. J. Mooibroek and J. Reedijk, Angew. Chem., Int. Ed., 2011, 50, 9564. 19 B. P. Hay and R. Custelcean, Cryst. Growth Des., 2009, 9, 2539. 20 M. Mascal, A. Armstrong and D. Bartberger, J. Am. Chem. Soc., 2002, 124, 6274. 21 I. Alkorta, I. Rozas and J. Elguera, J. Am. Chem. Soc., 2002, 124, 8593. 22 D. Quinonero, C. Garau, C. Rotger, A. Frontera, P. Ballester, A. Costa and P. M. Deya, Angew. Chem., Int. Ed., 2002, 41, 3389. 23 D. Kim, P. Tarakeshwar and K. S. Kim, J. Phys. Chem. A, 2004, 108, 1250. 24 D. Quinonero, C. Garau, A. Frontera, P. Ballester, A. Costa and P. M. Deya, J. Phys. Chem. A, 2005, 109, 4632. ˜ onero, A. Frontera, D. Escudero, 25 C. Garau, D. Quin `, Chem. Phys. Lett., P. Ballester, A. Costa and P. M. Deya 2007, 438, 104. 26 M. Albrecht, C. Wessel, M. Groot, K. Rissanen and ¨chow, J. Am. Chem. Soc., 2008, 130, 4600. A. Lu õnero and P. M. Deya `, J. Phys. 27 X. Lucas, A. Frontera, D. Quin Chem. A, 2010, 114, 1926. õnero, A. Frontera, P. Ballester, A. Costa 28 C. Garau, D. Quin `, J. Phys. Chem. A, 2005, 109, 9341. and P. M. Deya 29 D. Kim, E. C. Lee, K. S. Kim and P. Tarakeshwar, J. Phys. Chem. A, 2007, 111, 7980. 30 M. Zaccheddu, C. Filippi and F. Buda, J. Phys. Chem. A, 2008, 112, 1627. 31 Y. Lu, Y. Liu, H. Li, X. Zhu, H. Liu and W. Zhu, J. Phys. Chem. A, 2012, 116, 2591. 32 D. Y. Kim, I. Geronimo, N. J. Singh, H. M. Lee and K. S. Kim, J. Chem. Theory Comput., 2012, 8, 274. 33 D. Y. Kim, N. J. Singh, J. W. Lee and K. S. Kim, J. Chem. Theory Comput., 2008, 4, 1162. 34 P. d. Hoog, P. Gamez, I. Mutikainen, U. Turpeinen and J. Reedijk, Angew. Chem., Int. Ed., 2004, 43, 5815. 35 T. J. Mooibroek, C. A. Black, P. Gamez and J. Reedijk, Cryst. Growth Des., 2008, 8, 1082. 36 Y. S. Rosokha, S. V. Lindeman, S. V. Rosokha and J. K. Kochi, Angew. Chem., Int. Ed., 2004, 43, 4650. ´s, C. Bazzicalupi, A. Bianchi, C. Giorgi, 37 P. Arranz-Mascaro ´rrez-Valero, R. LopezM.-L. Godino-Salido, M. D. Gutie ń and M. Savastano, J. Am. Chem. Soc., 2013, 135, 102. Garzo 38 O. B. Berryman, F. Hof, M. J. Hynes and D. W. Johnson, Chem. Commun., 2006, 506. 39 O. B. Berryman, A. C. Sather, B. P. Hay, J. S. Meisner and D. W. Johnson, J. Am. Chem. Soc., 2008, 130, 10895.


View Article Online


PCCP

40 G. Gil-Ramirez, E. C. Escudero-Adan, J. Benet-Buchholz and P. Ballester, Angew. Chem., Int. Ed., 2008, 47, 4114. 41 P. Ballester, Acc. Chem. Res., 2013, 46, 874. 42 H. T. Chifotides, B. L. Schottel and K. R. Dunbar, Angew. Chem., Int. Ed., 2010, 49, 7202. 43 C. Estarellas, A. Frontera, D. Quinonero and P. M. Deya, Angew. Chem., Int. Ed., 2011, 50, 415. 44 R. E. Dawson, A. Hennig, D. P. Weimann, D. Emery, V. Ravikumar, J. Montenegro, T. Takeuchi, S. Gabutti, M. Mayor, J. Mareda, C. A. Schalley and S. Matile, Nat. Chem., 2010, 2, 533. 45 A. V. Jentzsch, D. Emery, J. Mareda, P. Metrangolo, G. Resnati and S. Matile, Angew. Chem., Int. Ed., 2011, 50, 11675. 46 S. Guha and S. Saha, J. Am. Chem. Soc., 2010, 132, 17674. 47 S. Guha, F. S. Goodson, L. J. Corson and S. Saha, J. Am. Chem. Soc., 2012, 134, 13679. 48 S. Guha, F. S. Goodson, S. Roy, L. J. Corson, C. A. Gravenmier and S. Saha, J. Am. Chem. Soc., 2011, 133, 15256. 49 D.-X. Wang and M.-X. Wang, J. Am. Chem. Soc., 2013, 135, 892. 50 D.-X. Wang, Q.-Y. Zheng, Q.-Q. Wang and M.-X. Wang, Angew. Chem., Int. Ed., 2008, 47, 7485. 51 M.-X. Wang, Chem. Commun., 2008, 4541. 52 M.-X. Wang, Acc. Chem. Res., 2012, 45, 182. 53 M.-X. Wang and H.-B. Yang, J. Am. Chem. Soc., 2004, 126, 15412. 54 M. E. Alberto, G. Mazzone, N. Russo and E. Sicilia, Chem. Commun., 2010, 46, 5894. 55 X. Zheng, Z. Shuai and D. Wang, J. Phys. Chem. A, 2013, 117, 3844. 56 A. W. Castleman Jr. and K. H. Bowen Jr., J. Phys. Chem., 1996, 100, 12911. 57 H. Wen, G. L. Hou, S. M. Kathmann, M. Valiev and X. B. Wang, J. Chem. Phys., 2013, 138, 031101. 58 G. L. Hou, W. Lin, S. H. M. Deng, J. Zhang, W. Zheng, F. Paesani and X. B. Wang, J. Phys. Chem. Lett., 2013, 4, 779.


Paper

59 E. V. Beletskiy, J. C. Schmidt, X. B. Wang and S. R. Kass, J. Am. Chem. Soc., 2012, 134, 18534. 60 A. Shokri, S. H. M. Deng, X. B. Wang and S. R. Kass, Org. Chem. Front., 2014, 1, 54. 61 X.-B. Wang and L.-S. Wang, Rev. Sci. Instrum., 2008, 79, 073108. 62 J.-D. Chai and M. Head-Gordon, J. Chem. Phys., 2008, 128, 084106. 63 J.-D. Chai and M. Head-Gordon, Phys. Chem. Chem. Phys., 2008, 10, 6615. 64 R. Krishnan, J. S. Binkley, R. Seeger and J. A. Pople, J. Chem. Phys., 1980, 72, 650. 65 K. A. Peterson, D. Figgen, E. Goll, H. Stoll and M. Dolg, J. Chem. Phys., 2003, 119, 11113. 66 EMSL Basis Set Exchange: https://bse.pnl.gov/bse/portal. 67 Y. Zhao and D. G. Truhlar, J. Phys. Chem. A, 2008, 112, 1095. 68 Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215. 69 Y. Zhao and D. G. Truhlar, Acc. Chem. Res., 2008, 41, 157. 70 S. F. Boys and F. Bernardi, Mol. Phys., 1970, 19, 553. 71 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci and G. A. Petersson, et al., Gaussian 09, Gaussian, Inc., Wallingford CT, 2009. 72 J. C. Rienstra-Kiracofe, G. S. Tschumper, H. F. Schaefer III, S. Nandi and G. B. Ellison, Chem. Rev., 2002, 102, 231. 73 D. W. Arnold, S. E. Bradforth, E. H. Kim and D. M. Neumark, J. Chem. Phys., 1995, 102, 3493. 74 A. I. Boldyrev and J. Simons, J. Phys. Chem., 1994, 98, 2298. 75 L. S. Wang and X. B. Wang, J. Phys. Chem. A, 2000, 104, 1978. 76 X. B. Wang and L. S. Wang, Nature, 1999, 400, 245. 77 T. Lu, Revision 2.2.1, University of Science and Technology Beijing, Beijing, China, 2011. 78 J. J. Dannenberg, J. Mol. Struct., 2002, 615, 219. 79 J. Del Bene and J. A. Pople, J. Chem. Phys., 1970, 52, 4858. 80 P. Gilli, V. Bertolasi, V. Ferrati and G. Gilli, J. Am. Chem. Soc., 1994, 116, 909.


Communication: Remarkable electrophilicity of the oxalic acid monomer: an anion photoelectron spectroscopy and theoretical study.

Alanate anion, AlH4(-): photoelectron spectrum and computations.

Low-energy photoelectron imaging of HS2 anion.

Microsolvation of LiBO2 in water: anion photoelectron spectroscopy and ab initio calculations.

On the dissolution of lithium sulfate in water: anion photoelectron spectroscopy and density functional theory calculations.

Slow Photoelectron Velocity-Map Imaging Spectroscopy of the ortho-Hydroxyphenoxide Anion.

5,10-Diacylcalix[4]pyrroles: synthesis and anion binding studies.

Hydrogen bonding interaction between active methylene hydrogen atoms and an anion as a binding motif for anion recognition: experimental studies and theoretical rationalization.

Photoelectron spectroscopy of CoC2H2(-) and density functional study of Co(n)C2H2 (n = 1-3) anion and neutral clusters.

Low-lying electronic structure of EuH, EuOH, and EuO neutrals and anions determined by anion photoelectron spectroscopy and DFT calculations.

Preorganized anion traps for exploiting anion-π interactions: an experimental and computational study.

Photoelectron-photofragment coincidence studies of the tert-butoxide anion (CH3)3CO((-)), the carbanion isomer (CH3)2CH2COH((-)), and corresponding radicals.

Enolate chemistry with anion-π interactions.

Solid state anion-π interactions involving polyhalides.

Quantifying the role of anion-π interactions in anion-π catalysis.

A theoretical study of ternary indole-cation-anion complexes.

Enhancing the basicity of ionic liquids by tuning the cation-anion interaction strength and via the anion-tethered strategy.

Photoelectron angular distributions as probes of cluster anion structure: I(-)·(H2O)2 and I(-)·(CH3CN)2.

Designation and Exploration of Halide-Anion Recognition Based on Cooperative Noncovalent Interactions Including Hydrogen Bonds and Anion-π.

Biotin[6]uril Esters: Chloride-Selective Transmembrane Anion Carriers Employing C-H···Anion Interactions.

Anion-π and cation-π interactions on the same surface.

[C-H···anion] interactions mediate the templation and anion binding properties of topologically non-trivial metal-organic structures in aqueous solutions.

Theoretical model to investigate the alkyl chain and anion dependent interactions of gemini surfactant with bovine serum albumin.

Anion-Anion Bonding and Topology in Ternary Iridium Seleno-Stannides.