Stability and Characteristics of the Halogen Bonding Interaction in an

Subscriber access provided by GAZI UNIV

Article

Stability and Characteristic of the Halogen Bonding Interaction in Anion-Anion Complex: A Computational Chemistry Study Guimin Wang, Zhaoqiang Chen, Zhijian Xu, Jinan Wang, Yang Yang, Tingting Cai, Jiye Shi, and Weiliang Zhu J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.5b08139 • Publication Date (Web): 06 Jan 2016 Downloaded from http://pubs.acs.org on January 18, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry B is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Stability and Characteristic of the Halogen Bonding Interaction

in

Anion-Anion

Complex:

A

Computational Chemistry Study Guimin Wang, †,# Zhaoqiang Chen, †,# Zhijian Xu, *, †,⊥ Jinan Wang,† Yang Yang, † Tingting Cai, † Jiye Shi, *, § Weiliang Zhu*,† †

CAS Key Laboratory of Receptor Research, Drug Discovery and Design Center, Shanghai

Institute of Materia Medica, Chinese Academy of Sciences, Shanghai, 201203, China §

UCB Biopharma SPRL, Chemin du Foriest, Braine-l'Alleud, Belgium

⊥

State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese

Academy of Sciences, Shanghai, 201203, China *To whom correspondence should be addressed. Phone: +86-21-50806600-1201 (Z.X.), +32-23862861

(J.S.),

+86-21-50805020

(W.Z.),

Fax:

+86-21-50807088

(W.Z.),

E-mail:

[email protected] (Z.X.), [email protected] (J.S.), [email protected] (W.Z.). #

These authors contributed equally to the work.

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 44

ABSTRACT: Halogen bonding is the noncovalent interaction between the positively charged σhole of organohalogens and nucleophiles. In reality, both the organohalogen and nucleophile could be deprotonated to form anions, which may lead to the vanishing of the σ-hole and possible repulsion between the two anions. However, our database survey in this study revealed that there are halogen bonding-like interaction between two anions. Quantum mechanics calculations with small model complexes composed of halobenzoates and propiolate indicated that the anion-anion halogen bonding is unstable in vacuum but attractive in solvents. Impressively, the QM optimized halogen bonding distance between the two anions is shorter than that in neutral system, indicating possibly stronger halogen bonding interaction, which is verified by the calculated binding energies. Furthermore, natural bond orbital and quantum theory of atoms in molecules analyses also suggested stronger anion-anion halogen bonding than the neutral. Energy decomposition by symmetry adapted perturbation theory revealed that the strong binding might be attributed to large induction energy. The calculations on 4 protein-ligand complexes from PDB by QM/MM method demonstrated that the anion-anion halogen bonding could contribute to the ligands’ binding affinity up to ~3 kcal/mol. Therefore, anion-anion halogen bonding is stable and applicable in reality.

KEYWORDS: Halogen bonding; anion; symmetry adapted perturbation theory; natural bond orbital; quantum theory of atoms in molecules; QM/MM calculation; drug design


2

Page 3 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Noncovalent interactions play a key role in chemistry, biochemistry, material science and drug discovery.1-4 As a burgeoning group of noncovalent interactions, σ-hole bonding has attracted more and more interest during the past decades.5 The atoms of groups IV-VII can simultaneously contain an area with negative and another area with positive electrostatic potential (ESP) while covalent bonded.6 The positive ESP area, which is called σ-hole, originates from the anisotropic charge distribution of the atoms.6 The noncovalent attractive interaction between the σ-hole and a nucleophile (O, N, S, P) is named σ-hole bonding.6-7 As the first and most-widely studied example of σ-hole bonding, halogen bonding, parallelizing to hydrogen bonding, refers to the interaction between halogens and Lewis base.8-13 Besides its strength, another property of halogen bonding that is comparable to hydrogen bonding is its directionality.6, 14-17 Lone pair electrons of halogen bonding acceptor always prefer to approach the σ-hole, the most positive region on the surface of halogen which is along the extension of covalent bond, causing an ideal halogen bonding interaction to be linear.6,

14

Exchange-repulsion has been stated to be the dominated term in the directionality of halogen bonding instead of the electrostatic, the largest attractive term of halogen bonding interaction energy, by symmetry adapted perturbation theory (SAPT) calculation.17-19 Due to its considerable strength and novel characteristics, halogen bonding has become another important and exploitable noncovalent interaction, especially, in crystal stacking, protein-ligand interaction and drug discovery and development processes.3, 20-21 It is known that the strength of halogen bonding is tunable.22-23 Organohalogens with electron-withdrawing or positively charged substituents have larger and more positive σ-hole on the surface of halogen atoms, which is prone to form stronger halogen bonding interaction.20, 24


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 44

In addition, the halogen bonding involved negatively charged Lewis base as acceptor is proved to be much stronger than neutral Lewis base.5, 25 A typical example is that the halogen bonding interaction energy between pentafluoroiodobenzene and chloride ion was predicted to be -23.80 kcal/mol at MP2/aug-cc-pVDZ level, while the energies involved NH3 and H2O are only -5.69 and -3.65 kcal/mol, respectively, for the same halogen donor.26 Negatively charged halogen donors have been demonstrated to have no obvious σ-hole, leading its halogen bonding interaction with neutral Lewis bases to become metastable or unstable in vacuum.20 However, protein environment could stabilize the interaction between the negatively charged donors and the neutral Lewis bases, resulting in similar geometric and energetic characteristics as the conventional halogen bonding.20 As mentioned above, either the acceptor25 or the donor20 of halogen bonding is possibly negatively charged to form a stable non-covalent interaction in reality. However, to the best of our knowledge, no study has been reported to investigate whether the interaction is attractive or repulsive if both the halogen bonding donor and the acceptor are negatively charged (called anion-anion halogen bonding interaction hereinafter). Cavallo et al. carried out systematic investigation of the Cambridge Structure Database (CSD) and verified that halogen bonding could drive the self-assembly of the oxyanions.27-28 We analyzed their survey results and noticed that there are anion-anion halogen bonding patterns, which motivated us to further explore the strength and characteristics of the anion-anion halogen bonding interaction. Herein, we focus on the intriguing noncovalent interaction between negatively charged organohalogens and negatively charged Lewis bases. Firstly, we searched the CSD to identify crystal structures containing anion-anion halogen bonding interaction patterns. Protein Data


4

Page 5 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Bank (PDB) was also surveyed to demonstrate whether anion-anion halogen bonding exists between biomacromolecule and its ligand. Then, quantum chemistry calculations were performed on both model systems and protein structures with anion-anion halogen bonding interaction. To get better comprehension of the underlying mechanisms, SAPT calculations were employed to partition the interaction energy into chemically meaningful terms. Natural bond orbital (NBO) theory and quantum theory of atoms in molecules (QTAIM) were also utilized to characterize the anion-anion halogen bonding. In view of the diverse dielectric in protein environment, the anion-anion halogen bonding in solvents with different dielectric constants was also investigated. The results revealed that the anion-anion halogen bonding is unstable in vacuum but is comparable to, or even stronger than, neutral halogen bonding within the environment with large dielectric constant.


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 44

Methods Database Survey for anion-anion halogen bonding Crystal structures with no disorder and errors as well as R-factor smaller than 0.1 in CSD (version 5.36 + 1 update) were surveyed by following criteria: the X···O distance is less than the sums of their respective van der Waals radii, i.e., d(Cl···O) < 3.27 Å, d(Br···O) < 3.37 Å, d(I···O) < 3.50 Å; the C-X···O angle is larger than 140°;20, 29-31 and the molecules involved halogen or/and oxygen are negatively charged. Protein structures in the PDB (July 2015 release) were surveyed by the similar criteria, while the halogen donor is negatively charged ligand, always containing carboxyl, and the oxygen comes from protein residues Glu and Asp. We only considered X-ray crystal structures at resolutions of 3.0 Å or better and high quality structures determined by solution NMR. Then, the structures with typical anion-anion halogen bonding interaction patterns were chosen for quantum chemistry calculation. The PROPKA (V 3.1) was applied to predict the protonation status of the ligands and the related protein residues.32-33 Electrostatic Potential V(r) The electrostatic potential V(r) that the electrons and nuclei of a molecule create at any point r in the surrounding space can be calculated by eq 1.

V (r ) = ∑ A

ρ ( r') dr' ZA −∫ R A −r r' − r

(1)

Where, ZA is the charge on nucleus A, located at RA, and ρ(r) is the electronic density of the molecule. V(r) is a physical observable and can be determined experimentally by diffraction methods34-35 as well as computationally. Its sign depends upon whether the positive contribution


6

Page 7 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of the nuclei or the negative one of the electrons is dominant in the region. Following Bader et al.,36 we take the surface to be the 0.001 au (electrons/bohr3) contour of the molecule’s electronic density, which has the advantage of reflecting features specific to the molecule, such as lone pairs, π electrons, strained bonds, atomic anisotropies, etc. The potential computed on this surface is designated VS(r) and its most positive and most negative values are identified as VS,max and VS,min, respectively.37-39 In this study, the VS(r) was computed by Multiwfn (v 3.3.5)40 using the wave functions generated by Gaussian 09 suite of programs.41 Geometry Optimization and Interaction Energy Calculation The complexes were fully optimized using the M06-2X method with no constrains, which has been reported to perform well on the characterization of halogen bonding complexes.20,

42-47

Stuttgart/Dresden pseudopotential (SDD) was adopted for iodine atoms and 6-311++G(d,p) for the other atoms. All systems were confirmed to be minima on their potential surfaces with frequency calculation at the same level. Geometric parameters were defined in Figure 1. Among that, d and θ are the crucial parameters in the characterization of halogen bonding; φ determines the extent of the halogen σ-hole interacting with the lone pair of oxygen. The interaction energy of the complexes was assessed by eq 2. ∆ = − + + BSSE

(2)

Here, ∆ is the interaction energy, Ecom is the energy of the complex, Edonor and Eacpt are the energies of the halogen bonding donor and acceptor, respectively, and BSSE stands for the basis set superposition error corrections obtained by the counterpoise method.48


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 44

Figure 1. Schematic models of halogen bonding and the geometric parameters definition in this study. Apart

from

M06-2X/6-311++G(d,p)

method,

MP2/aug-cc-pVTZ

and

B3LYP/6-

311++G(d,p) methods were also used to calculate the interaction energies according to eq 2. The results are consistent with each other (Table S1, Supporting Information), which manifests the reliability of the M06-2X/6-311++G(d,p) method used in this work. The optimization, frequency and single-point energy calculation of the complexes in solvents were carried out with the conductor-like polarizable continuum model (CPCM).49 The interaction energies was estimated according to eq 3. ∆ = − _ + _

(3)

In eq 3, superscript S represents the energy in solvents; subscript A_bAB means energy of monomer A calculated with the whole basis functions of dimer AB, where atoms of monomer B were treated as ghost atoms. Accordingly, the interaction energy with solvent effect was corrected for the BSSE. All these calculations were performed with the Gaussian 09 suite of programs.41


8

Page 9 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


SAPT Methodology To produce some in-depth insight into the bonding interactions, SAPT calculations were carried out to decompose the interaction energy into chemically meaningful components.50-52 The SAPT interaction energy is given by eq 4. = ! + "# + + !

(4)

The four components, namely, electrostatic (Eelst), exchange-repulsion (Eexch), induction (Eind) and dispersion (Edisp) are calculated directly by a sum of perturbative energy correction terms. We performed SAPT calculation with the optimized geometries by Gaussian 09 suite of programs at the SAPT0/jun-cc-pVDZ level, while for bromine atoms, pseudopotentials with the jun-cc-pVDZ-PP basis set were used for relativistic effects. All SAPT calculations were performed with the Psi4 (beta 5) program using density fitting.53 NBO Methodology Charge distributions of the complexes were examined in terms of the natural bond orbital (NBO) analysis which has been widely used in noncovalent interaction study.54 The charge of halogen atoms in complex and its difference with that in monomer were given, together with the intermolecular charge transfer. According to the NBO theory, halogen bonding can be viewed as nucleophile interactions with the LUMO of halogenous group and electrophile interactions with the HOMO of electron donor.55 The second-order perturbation stabilization energies E(2) for the local orbital interaction can be evaluated as eq 5.

$2& = Δ() = *(

+$(,)& 2 -) − -(

(5)


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 44

Where qi is the donor orbital occupancy, εj and εi are diagonal elements (orbital energies), and F(i,j) is the off-diagonal NBO fock matrix element. All the NBO analysis was carried out by the use of the NBO program implemented in the Gaussian 09 suite at the same level as geometry optimization. QTAIM Methodology The QTAIM calculations were carried out to find the bond critical points (BCPs) between the halogen and the oxygen.56-59 The interactions were characterized by following properties at the BCPs: the electron density (ρ), its Laplacian (∇2ρ), the total electron energy density (H), and its two components, the potential electron energy density (V), the kinetic electron energy density (G), and the ratio of |λ1|/λ3, where λ1 is one of the negative curvature values and λ3 is the positive curvature value.22, 60-62 The relationships among some parameters mentioned above can be described as eq 6. 1 1 ∇ 2 = 23 + 4 = 3 + 5 4

(6)

We performed QTAIM analysis with the help of Multiwfn (v 3.3.5),40 using the wave functions generated by Gaussian 09 suite of programs. QM/MM Calculation The QM/MM calculation of the protein systems was performed by a two-layer ONIOM (our own N-layered integrated molecular orbital and molecular mechanics) method.63-67 Hydrogen atoms were firstly added to the ligands by LigPrep68 module in Schrödinger 2010 software suite.69 Only the protein monomers which form halogen bonding interaction with the ligands were kept for QM/MM calculations. The pKa values of ionizable residues in the proteins were calculated by the


10

Page 11 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


H++ Web site at the pH values of the crystallization and hydrogen atoms were added accordingly.70-71 The ligands and the corresponding protein backbones that form halogen bonding interactions were included in the QM region which was described at the B97D/6-31G(d) level with SDD applied for iodine atoms. The rest part of the proteins was left in the MM layer and described by the AMBER parm96 force field.72 The QM/MM optimization was carried out with no constrains in vacuum. The QM layer of the system was then extracted without waters to calculate the single point energy at the MP2/6-311++G(d,p) level (SDD for iodine atom). The interaction energies in vacuum and solvents between the ligands and the protein residues were assessed by eq 2 and eq 3, respectively. All these calculations were also performed with the Gaussian 09 suite of programs.


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 44

Results and Discussion Anion-Anion Halogen Bonding in CSD To provide strong crystallographic evidence of anion-anion halogen bonding interaction, a survey of the CSD was performed. There are 119 anion-anion halogen bonding interactions presented in 99 crystal structures, while the numbers of anion-anion halogen bonding interactions for chlorine, bromine and iodine are 79, 29 and 11, respectively (See Table S2 for details, Supporting Information). Therefore, the existence of anion-anion halogen bonding interaction in crystal structures indicates the significance of the computational study. Searching for Anion-Anion Halogen Bonding in PDB The PDB was surveyed to demonstrate the potential anion-anion halogen bonding interaction between biomacromolecules and their ligands. In total, 15 protein structures were found to contain 20 typical halogen bonding interactions, among which 11, 4 and 5 interactions are with chlorine, bromine and iodine atoms, respectively (See Table S3 for details, Supporting Information). We selected four crystal structures as the representatives of the typical anion-anion halogen bonding interaction patterns, viz., 2Y3473 containing Cl (Figure 2a), 4J2V74, 4JK474 and 2BXN75 containing I (Figures 2b-d). The ligand in 4JK4 is the same as that in 4J2V, but the halogen atom forming halogen bonding with the protein residue is different. All the ligands and the corresponding protein residues were predicted to be deprotonated at the pH values of the crystallization apart from the ligand in 4J2V which would deprotonate at pH 7.0 (Table S4 and S5, Supporting Information). We calculated the ESP mapped on the whole molecular surface (electron density ρ = 0.001 au). The previous σ-hole area has negative ESP, but less negative than the other region (Figure 3). In order to explore whether these anion-anion halogen bonding


12

Page 13 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


interactions in protein systems are attractive and could contribute to the ligand binding affinity, we then carried out computational chemistry studies on the four typical systems.

Figure 2. Crystal structures (cyan) and optimized structures (yellow) of the anion-anion halogen bonding between protein and its ligand for (a) 2Y34; (b) 4J2V; (c) 4JK4, and (d) 2BXN. Ligands and the corresponding residues and waters are presented as balls and sticks with the backbones of protein are shown as cartoon. All the atoms in crystal are colored in cyan and for the optimized structures, carbon atoms are in yellow with oxygen atoms in red, nitrogen atoms in blue, chlorine atoms in green, bromine atoms in brown and iodine atoms in purple.


13


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 44

Figure 3. Electrostatic potentials mapped on the molecular surface (electron density ρ = 0.001 au) of ligands in: (a) 2Y34, (b) 4J2V & 4JK4, (c) 2BXN. Model Complexes of Anion-Anion Halogen Bonding Interaction and Their Geometries and Energetics In order to eliminate other intermolecular interactions to the utmost extent, model systems were designed to explore the stability and characteristic of the potential anion-anion halogen bonding interaction, which are composed of 4-halobenzoate and propiolate as the halogen bonding donor and acceptor, respectively (Figure 1, Figure S1 in Supporting Information). Quantum chemistry optimization were carried out on the anion-anion (-OOC-phX···-OOC-C≡CH), neutral-neutral


14

Page 15 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(HOOC-phX···HOOC-C≡CH) and neutral-anion (HOOC-phX···-OOC-C≡CH) halogen bonding complexes in vacuum. The geometric and energetic properties of the optimized complexes are presented in Table 1. For the neutral-neutral (4-halobenzoic acid-propiolic acid) halogen bonding complexes, all the optimized structures displayed typical halogen bonding geometry with the θ close to 180° and the d around 3.1 angstrom (Table 1). As expected, the angle of φ does not deviate far from 120°.24 The interaction energy (∆E) are all negative, indicating the attractive interaction between the halogen bonding donor and acceptor. The strength of the interaction increases from chlorine to iodine. As we can see, the overall free energy of the interaction are more positive than the enthalpy. In other words, the entropy significantly affect the formation of complexes, which is consistent with the experimental data acquired through isothermal calorimetric titrations.76 For the neutral-anion (4-halobenzoic acid-propiolate) halogen bonding complexes, the interaction has similar angles (θ and φ) with the neutral-neutral systems (Table 1). Nevertheless, the d are 0.3-0.4 angstrom shorter than the corresponding neutral-neutral systems and the interaction energy (∆E) is obviously much stronger, which is in accordance with previous studies.5, 25-26 For the anion-anion (4-halobenzoate-propiolate) systems, the designed complexes with the chlorobenzoate and bromobenzoate as the donors could not attain the classical halogen bonding geometry when optimized in vacuum (Table 1), suggesting the anion-anion halogen bondings are unstable. We fixed the geometry (viz., d = 3.10 Å, θ = 180° and φ = 120°) of the complexes and calculated the single point energy. Indeed, the ∆E are positive, implying a repulsive interaction (Table 1), which were confirmed by the MP2 and B3LYP calculations for the 4-bromobenzoate-


15


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 44

propiolate complex (Table S1, Supporting Information). Optimization of the 4-iodobenzoate with propiolate resulted in a halogen bonding geometry, while the ∆E is still positive. This result is expected that two negatively charged molecules are mutually repulsive. We should note that the φ of the organoiodine system is 160.8° (Table 1, Figure S1 in Supporting Information). Namely, the most positive area on the surface of atom iodine does not exactly face the lone pair of the oxygen, but the area between the two lone pairs, which is in favor of weakening the electrostatic repulsion. Accordingly, the anion-anion halogen bonding seems to be unstable in vacuum. Table 1. Geometric and Energetic Data of the Optimized Halogen Bonding Complexes in Vacuum Complex HOOC-phX···HOOC-C≡CH HOOC-phX···-OOC-C≡CH -

OOC-phX···-OOC-C≡CH

a

Values are in angstrom.

b

X Cl Br I Cl Br I Cl Br I

da 3.10 3.06 3.08 2.75 2.73 2.67 -g 3.01

θb 176.3 177.7 178.9 177.1 176.4 176.0 176.7

φc 114.6 123.1 123.8 112.4 115.1 118.7 160.8

∆Ed -0.82 -1.52 -2.73 -5.43 -8.62 -14.38 36.78h 34.60h 29.32

∆He 0.33 -0.43 -1.53 -4.32 -7.51 -13.18 30.50

∆Gf 7.42 6.88 6.15 4.15 1.43 -3.86 38.09

and cValues are in degree. d, e and fValues are in kcal/mol. gOptimization could not end in a typical

halogen bonding geometry. hSingle point energy when d = 3.10 Å, θ = 180° and φ = 120°.

SAPT-Based Energy Decomposition To gain more insight into the anion-anion halogen bonding in the model complexes, the interaction energy was decomposed by means of the SAPT method. In view of the similarity of halogen bonding, we just take organobromine system as an example. The total interaction energies based on the SAPT are -1.50 kcal/mol, -8.04 kcal/mol, and 34.93 kcal/mol for the neutral-neutral, neutral-anion and anion-anion complexes, respectively, which gives a high degree of consensus with the result by supermolecular approach (Table 1) and implies high reliability of the individual SAPT terms.


16

Page 17 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The decomposition results are illustrated in Figure 4 (Table S6, Supporting Information). Electrostatics is the largest attractive component for the neutral-neutral halogen bonding, while induction and dispersion also have significant contribution, leaving exchange-repulsion term being the only repulsion. In the neutral-anion complex, all the energy components are much stronger than that in the neutral-neutral complex. It is noteworthy that induction term is almost equal to the electrostatics for the neutral-anion complex which means significantly great charge transfer, while the proportion of dispersion decreases. According to Hill,77 this neutral-anion halogen bonding complex should be described as Mulliken inner complex to distinguish from the other weak halogen bonding complexes, which are classified Mulliken outer complex. For the anion-anion halogen bonding complex, it is the greatly repulsive electrostatics interaction that is account for the positive total interaction energy. Induction seems to be the main attractive term which is quite strong comparing with the neutral-neutral system. We also performed MP2/augcc-pVTZ and B3LYP/6-311++G(d,p) calculations, which supported that dispersion energy makes an important contribution to halogen bonding (Table S1, Supporting Information). Taking into account the shorter distance (3.01 Å) than that of neutral system (3.08 Å) in organoiodine systems, the anion-anion halogen bonding stands a chance of becoming attractive once the electrostatic repulsion weakens.


17


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 44

Figure 4. SAPT decomposition of the interaction energy for halogen bonding complexes. NBO Analysis As shown in Table 2, the charge on the halogen atoms turns more positive upon complexation in all cases including the anion-anion complexes. Obviously, there is a magnitude of charge transfer (QCT), in a range of 3-75 me (Table 2), from electron donor to the halogenous group. The contribution to the total interaction energy of charge-transfer is included in the induction term in SAPT decomposition. The amount of charge transfer in the neutral-anion complex is the greatest and that in the neutral-neutral complex is the least, which is consistent with the results of energy decomposition. According to the NBO results, the main intermolecular orbital interaction is the interaction between the occupied lone pair of oxygen atom and the formally empty antibonding orbital of C-X bond. The total second-order perturbation NBO stabilization energies E(2) of these orbital interactions are in the range of 0.85-13.73 kcal/mol (Table 2), which is close to the corresponding interaction energies in the neutral-neutral and neutral-anion complexes. The significant value of E(2) in the anion-anion complexes suggests the potential of stable anionanion halogen bonding again. Table 2. Calculated Charge and Second-Order Perturbation Stabilization Energies in Natural Bond Orbital Analysis Complex

X Cl HOOC-phX···HOOC-C≡CH Br I Cl HOOC-phX··· OOC-C≡CH Br I Cl Br OOC-phX··· OOC-C≡CH I a

q(X)a ∆q(X)b 0.0437 0.0240 0.1154 0.0300 0.2396 0.0386 0.1290 0.1093 0.2072 0.1218 0.3323 0.1313 0.0214 0.0770 0.0887 0.0922 0.2250 0.1330

E(2)d QCTc 0.0036 0.85 0.0066 1.78 0.0120 3.20 0.0216 4.01 0.0390 7.80 0.0753 13.73 0.0072 1.21 0.0128 2.24 0.0152 4.01

Charge of halogen atoms in complex. bDifference in atomic charge on halogen atoms between the complex and monomer.

c

Charge transfer from halogen bonding acceptor to donor. All charges are given in au. dSecond-order perturbation stabilization


18

Page 19 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


energies of the natural bond orbital interaction between the occupied lone pair of O atom and the formally empty antibonding orbital of C-X bond.

Effect of Dielectric Constant on the Anion-Anion Halogen Bonding Taking account of the different dielectric environment of protein, we then optimized the model complexes in solvents with different dielectric constants using implicit solvation model. The geometrical parameters and the interaction energies are summarized in Table 3. It is evident that all the halogen bondings have similar values of θ (173-180º) and φ (113-128º), indicating that all the designed halogen bonding systems could eliminate other intermolecular interaction to the utmost. The optimized interaction distances in different solvents showed that the neutral-anion systems always have shorter interactions than the corresponding neutral-neutral systems, indicating stronger binding in the former systems that is in well agreement with previous studies.5, 25 To our surprise, the distances in the anion-anion halogen bonding systems are also shorter than that in the neutral-neutral systems except for organochlorine system. For instance, the distances between I and O in the 4-iodobenzoic acid-propiolic acid complexes in different solvent are 3.08 Å, while that in the corresponding anion-anion systems are 2.96 Å, suggesting the possibility of stronger anion-anion halogen bonding than the neutral one. The calculated binding energy (Table 3) revealed that the strength of halogen bonding always increases from chlorine, bromine to iodine, accompanied by the shortening of the bond distance when the systems have the same charging states and in the same solvent. As a result of the major contribution of electrostatic, the neutral-neutral halogen bonding complexes are slightly destabilized in different solvents with the halogen bond distances elongated and the interaction energies weaken a little apart from the halogen bonding involved


19


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 44

chlorine (Table 3 and Figure 5). Overall, the neutral-neutral halogen bonding shows little change, both geometries and energetics. In comparison with the neutral-neutral systems, the strength of the neutral-anion halogen bonding undergoes a significant drop in the solvents, which reflects the fact that electrostatic is a main contribution component to the whole binding energy as shown in Figure 4. When we move to the anion-anion halogen bonding complexes, quite different effect of the solvents was observed. The total repulsive interactions convert into attractive as the dielectric constant of the solvents increases (Table 3 and Figure 5). What is worth noting is that the attractive interaction between 4-iodobenzoate and propiolate is stronger than the neutral-neutral halogen bonding between 4-iodobenzoic acid and propiolic acid in DMSO and water. Therefore, solvents have a significant stabilizing effect on the anion-anion halogen bonding, possibly making the anion-anion complex more stable than the neutral-neutral halogen bonding system in real environment with large dielectric constant.


20

Page 21 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Interaction energy (∆E) of halogen bonding complexes optimized in vacuum (ε = 1) and a series of solvents (chloroform, ε = 4.71; acetone, ε = 20.49; DMSO, ε = 46.83; water, ε = 78.36).

Table 3. Geometric and Energetic Data of the Optimized Halogen Bonding Complexes in Solvents with Different Dielectric Constants Solvent (ε)

Complex HOOC-phX··· HOOC-C≡CH

Chlorofor m (4.71)

HOOC-phX··· OOC-C≡CH -

OOC-phX··· OOC-C≡CH

-

HOOC-phX··· HOOC-C≡CH

Acetone (20.49)



-


DMSO (46.83)



-


Water (78.36)



-

X Cl Br I Cl Br I Cl Br I Cl Br I Cl Br I Cl Br I Cl Br I Cl Br I Cl Br I Cl Br I Cl Br I Cl Br I

da 3.12 3.10 3.08 2.92 2.89 2.85 -g 3.02 2.96 3.11i 3.11 3.08 3.01 2.94 2.90 3.17 3.01 2.96 3.11i 3.11 3.08 3.03 2.95 2.90 3.17 3.02 2.96 3.11i 3.11 3.08 3.04 2.96 2.91 3.17 3.01 2.96

θb 179.3 175.9 177.6 176.1 177.1 175.8 177.0 176.7 175.6 175.5 177.4 176.4 177.6 176.3 174.0 176.9 176.5 175.1 175.6 177.3 176.5 177.8 176.3 173.9 177.0 175.9 175.3 175.4 177.3 176.5 177.8 176.4 174.0 177.0 176.6

φc 113.6 117.7 122.0 126.0 125.8 119.1 129.4 127.4 114.7 117.2 121.6 126.1 127.2 122.1 124.4 128.0 125.6 114.7 117.2 121.5 126.1 127.7 122.0 124.2 127.2 125.2 114.7 117.1 121.5 126.1 127.8 122.3 124.2 127.9 125.4

∆Ed -0.93 -1.42 -2.49 -1.19 -2.56 -5.41 7.26h 5.96 3.55 -0.89 -1.34 -2.33 -0.80 -1.83 -4.03 1.21 0.38 -1.55 -0.89 -1.33 -2.31 -0.76 -1.73 -3.81 0.23 -0.55 -2.68 -0.89 -1.33 -2.30 -1.11 -2.06 -4.08 -0.43 -1.19 -2.96

∆He 0.29 -0.30 -1.26 -0.04 -1.37 -4.18 7.09 4.86 0.28 -0.19 -1.10 0.36 -0.62 -2.74 2.39 1.56 -0.22 0.28 -0.16 -1.05 0.41 -0.52 -2.52 1.42 0.63 -1.39 0.29 -0.17 -1.04 0.20 -0.71 -2.65 0.89 0.13 -1.49

∆Gf 7.19 6.82 6.69 7.82 7.48 4.69 15.37 13.05 6.95 7.00 6.38 8.02 7.82 4.88 10.55 10.06 8.78 6.97 7.25 7.00 8.05 7.88 5.18 9.69 8.61 6.78 6.95 7.04 7.01 7.90 7.76 4.25 9.27 8.68 7.61


21


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

a

Values are in angstrom.

b

Page 22 of 44

and cValues are in degree. d, e and fValues are in kcal/mol. gOptimization could not end in a typical

halogen bonding geometry. hSingle point energy when d = 3.10 Å, θ = 180° and φ = 120°. iKeep the propiolic acid being coplanar with the chlorine atom.

QTAIM-Based Analysis QTAIM has been successfully applied to study the properties of conventional and unconventional noncovalent interactions.58, 78-82 According to the QTAIM theory,56 BCPs contain a wealth of chemical information to characterize the interactions. Koch and Propelier83 summarized the eight criteria for C-H···O hydrogen bonding, which has been widely used, added or amended by others. For its parallelism to hydrogen bonding, halogen bonding is also characterized a lot with QTAIM.61-62, 84-86 Our calculations indicated that BCPs exist between the halogen and the oxygen for all the systems investigated in this work (Figure S2, Supporting Information). All the calculated parameters at the BCPs of X···O are listed in Table 4. For all systems, apart from the neutralanion complex involved iodine in vacuum, the calculated results are similar and could be concluded as following: the values of ρ are small (

Stability and Characteristics of the Halogen Bonding Interaction in an

Recommend Documents