Does Halogen Bonding Promote Intersystem Crossing and

2.5 Å), while there are minimal enhancement beyond an XB distance of 3.0 Å . More SOC data points with XB distance between 2.5 and 3.0 Å at 0.1 Å ...
0 downloads 3 Views 541KB Size
Subscriber access provided by Kaohsiung Medical University

C: Plasmonics; Optical, Magnetic, and Hybrid Materials

Does Halogen Bonding Promote Intersystem Crossing and Phosphorescence in Benzaldehyde? Shi Jun Ang, Tsz Sian Chwee, and Ming Wah Wong J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.8b03253 • Publication Date (Web): 18 May 2018 Downloaded from http://pubs.acs.org on May 18, 2018

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 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 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.

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 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

The Journal of Physical Chemistry

Does

Halogen

Bonding

Promote

Intersystem

Crossing

and

Phosphorescence in Benzaldehyde? Shi Jun Ang1,2, Tsz Sian Chwee*2, Ming Wah Wong*1,3 1

NUS Graduate School for Integrative Sciences and Engineering, Centre for Life Sciences (CeLS),

#05-01, 28 Medical Drive, Singapore 117456. 2

Institute of High Performance Computing, 1 Fusionopolis Way, #16-16 Connexis North, Singapore

138632. 3

Department of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore 117543.

ABSTRACT: In the past decade, halogen bonding (XB) has been utilized extensively in the design of novel materials and new drugs. One of the emerging applications of XB is in devising organic roomtemperature phosphorescent materials. Several reports showed that the use of benzaldehyde-based phosphors with appropriate XB donors result in high phosphorescence quantum yields, due to enhanced intersystem crossing (kISC) and phosphorescence(kPH) rates. It is often advocated that a combination of factors, namely rigidification, heavy-atom effect and reduction of quenching by triplet dioxygen, are responsible for the enhancement. However, to what extent each factor contributes to the enhancement is unknown. In this study, we performed ab initio excited state calculations on two halogen bonding complexes, benzaldehyde···XF (X = Br and I), with varying XB distance to elucidate the effect of XB on kISC and kPH. Our results show that XB reduces the kISC of benzaldehyde and changes the character of T1 from which phosphorescence takes place. Hence, the generally accepted assumption that XB enhances spin-orbit coupling and kISC is oversimplified.

1. INTRODUCTION Halogen bonding (XB) has gained considerable experimental and theoretical interest due to its promising applicability to materials1 and drug design.2 The attractiveness of this non-covalent interaction stems from the fact that the halogen atom of an R-X moiety has the ability to interact with electron rich species via the σ-hole at the extension of the covalent bond, due to the unique anisotropic distribution of electron density surrounding it. In the theoretical front, the geometrical properties of XB have been investigated thoroughly using a myriad of quantum chemical methods. The most prominent properties examined are its linearity and tunability.3 Furthermore, popular energy

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 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 17

decomposition methods, such as Symmetry-Adapted Perturbation Theory (SAPT) and Energy Decomposition Analysis (EDA), have been employed to study the nature of the XB bonding.4-6 On the other hand, experimentalists have leveraged on the unique geometrical properties to produce XBbased porous materials, liquid crystals, polymers and gels, which are covered in recent extensive reviews on XB.3,7 More recently, the unique photophysical properties of XB assembled crystals have been discovered by several groups.8,9 Heavy atoms, such as Br and I, are thought to promote spinorbit coupling and thereby increase the rate of forbidden transitions such as intersystem crossing and phosphorescence.10–12 This rate enhancement effect is known as “heavy-atom effect”. To date, there are several experimental reports of XB crystals involving lone-pair13–16 and π17–21 XB-acceptors exhibiting phosphorescent properties. Bolton et. al.13 suggested a combination of three factors: (1) triplet producing aromatic aldehyde, (2) XB interactions to reduce non-radiative relaxation pathways in crystalline state, and (3) heavy-atom effect, which lead to high phosphorescence quantum yield for their studied systems. Following this landmark paper,13 several review articles on XB advocated the same point to a wider XB community.3,7 On the biological front, Fateminia et al. has synthesized nanocrystals which exhibit persistent phosphorescence in air and are suitable for bioimaging of breast cancer cells.22 The authors have attributed the enhancement of phosphorescence partly to the short Br···carbazole contacts. Heavy–atom effect is certainly not the only factor that contributes to experimentally observed phosphorescence. In a recent review, Baroncini et al. proposed three possible factors that can enhance phosphorescence in organic molecules.23 Firstly, through rigidification of the chromophore by encapsulating it in a matrix or via crystallization to reduce vibrational relaxations. Secondly, through introducing an internal or external heavy atom to promote spin-orbit coupling (SOC). Thirdly, to reduce the energy gap between S1 and the nearest triplet state. It is unclear to what extent each factor contributes to efficient phosphorescence in XB-assembled materials, though understanding their relative importance will facilitate the design of novel phosphorescence materials. Herein, we present a theoretical study on the effects of lone-pair type XB on intersystem crossing rates (kISC) and phosphorescence rates (kPH) of benzaldehyde, by studying the excited states of benzaldehyde···BrF and benzaldehyde···IF dimers at varying XB distances. This will allow us to probe the extent to which spin-orbit coupling is enhanced, and how relevant energy gaps are altered upon the formation of halogen bond towards the lone-pair of the carbonyl oxygen. It is important to note that halogen bonds involving non-carbonyl compounds may have different effects and we caution the generality of our findings.

2. METHODOLOGY AND COMPUTATIONAL DETAILS Excited state optimizations of benzaldehyde were first carried out with the EOM-CCSD24,25 method together with the standard 6-31G* basis set. Frequency calculations were performed on the optimized

ACS Paragon Plus Environment

2

Page 3 of 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

The Journal of Physical Chemistry

geometries to ensure that the excited state structures of benzaldehyde are local minima. These structures were further refined at a higher level, i.e. EOM-CCSD/aug-cc-pVDZ. Ground state optimizations and frequency calculations of BrF and IF were carried out using CCSD/aug-cc-pVDZ(pp for Br and I) level of theory. The optimized bond lengths of BrF and IF are 1.793 and 1.944 Å, respectively. Upon obtaining the ground and excited state structures of the monomers, the optimized structures were brought together manually with the halogen bond directing towards the carbonyl oxygen of benzaldehyde at XB distances between 2.5 to 4.0 Å as depicted in Figure 1. To ensure that the range of XB distance studied is meaningful, we performed full geometry optimizations for benzaldehyde···BrF and benzaldehyde···IF complexes at M06-2X/aug-cc-pVTZ(pp for Br and I) level of theory as recommended by Kozuch and Martin.26 The optimized XB lengths are 2.40 and 2.51 Å, respectively, which give a lower bound on a meaningful range of XB distance before entering the repulsive region. In addition, domain-based local pair natural orbital coupledcluster theory with singles, doubles and iterative triples excitations,27 DLPNO-CCSD(T)/aug-ccpVTZ(-pp) scans along the XB coordinate (with fixed CCSD/aug-cc-pVDZ(-pp) monomer geometries) support the M06-2X XB lengths at the PES minima (Figure S1). Furthermore, we have carried out Cambridge Structural Database (version 5.36, Update November 2016) search using Conquest28 to investigate the number of non-covalent contacts between aromatic carbonyl oxygen and halogen (X = Br and I) and their structural parameters in experimentally determined crystal structures. In the CSD searches, we require the XB lengths to be shorter than the sum of van der Waal radii (3.37 and 3.50 Å for O···Br and O···I, respectively) and the XB angle, i.e. O···X-R, to be between 165 and 180˚. The shortest XB distances found are 2.846 and 2.765 Å for O···Br (99 unique contacts) and O···I (58 unique contacts), respectively. It is not surprising that these XB lengths are longer than that of benzaldehyde···BrF and benzaldehyde···IF due to other noncovalent interactions present in the crystals. Hence, both theory and crystal structure survey indicate that the range of XB distances studied is indeed relevant and realistic. The search results are tabulated in Table S1. EOM-CCSD/aug-cc-pVDZ(-pp for Br and I) single-point calculations were performed to obtain the excited state energies of benzaldehyde and benzaldehyde···XF structures. All EOM-CCSD and CCSD calculations were carried out with Q-Chem 5.0.29 On the other hand, ORCA (Version 4.0.1)30 was used for DLPNO-CCSD(T) calculations with “TightPNO” threshold, recommended for non-covalent interactions. It is important to note that based on benchmarking study by Schreiber et al.,31 EOMCCSD systematically overestimates singlet excitation energies involving valence transition of small organic molecules and nucleobases by 0.5 eV when benchmarked against theoretical best estimates. However, the excited state energies and ordering of benzaldehyde at various geometries are found to agree well with high-level second-order extended multi-configuration quasi-degenerate perturbation

ACS Paragon Plus Environment

3

The Journal of Physical Chemistry 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 17

theory (XMCQDPT2) based on a theoretical study of acetophenone by Huix-Rotllant.32 We did not consider the use of Time-Dependent Density Functional Theory (TD-DFT) as a viable choice because it is well established that TD-DFT triplet state energies have absolute errors significantly larger than their singlet counterparts,33 although coulomb-attenuated functional coupled with Tamm-Dancoff approximation (TDA) improves the triplet instability problem.34 Moreover, our EOM-CCSD/aug-ccpVDZ excitation energy for 11(ππ*) is in closer agreement with the experimental value,35 while the TDA-TD-ωB97X-D/6-311+G* counterpart36 overestimates by 0.71 eV (Table 1). On the other hand, both levels of theory agree well on the excitation energy to 21(ππ*). In order to study the effects of solvation on the excitation energies, and to compare with the experimental results of Kim and Dunietz et al,36 EOM-CCSD/aug-cc-pVDZ calculations with C-PCM solvation model were performed using the dielectric constant of chloroform (4.71) at the ground state geometry of benzaldehyde. The inclusion of solvation effect does not alter the π → π* excitation energies significantly, while the n → π* counterparts are increased from 0.13 to 0.15 eV. This is due to the significant decrease in dipole moment from 3.87 D (ground state) to 1.24 D (1(nπ*) state) and 1.14 D (3(nπ*) state) as reported by Molina and Merchán, which makes n → π* transition energetically higher in solution.

Figure 1. Illustration of the geometrical parameters of benzaldehyde···XF dimers studied.

Table 1. Experimental and theoretical excitation energies (eV) of benzaldehyde. All excitation energies are vertical excitation energies unless otherwise stated. excited state 1

(nπ*)

expt (gas) a

expt (solution) b

ωB97X-D/6311+G* b

EOM-CCSD/aug-ccpVDZ (gas) c

EOM-CCSD/aug-ccpVDZ (solution) c,d

3.34 (0-0)

-

3.92

3.92 (3.61, adiabatic)

4.05

1

4.51

4.40

5.11

4.79

4.77

1

5.34

5.00

5.68

5.68

5.60

(nπ*)

3.12 (0-0)

-

-

3.55 (3.26, adiabatic)

3.70

13(ππ*)

3.3 (0-0)

-

-

3.56 (3.21, adiabatic)

3.57

1 (ππ*) 2 (ππ*) 3

a

Experimental gas-phase values from Molina and Merchán (Ref. 35) and references therein. Experimental (10-5 M of benzaldehyde in chloroform solvent) and TDA-TD-ωB97X-D (in the gas phase) values from Kim and Dunietz et al. (Ref. 36). c From present study. d C-PCM solvation calculations with chloroform as solvent. b

ACS Paragon Plus Environment

4

Page 5 of 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

The Journal of Physical Chemistry

As spin-orbit coupling (SOC) is well-known to be less sensitive to the method employed,37 we evaluated the SOC matrix elements with more economical TDDFT perturbatively, using an effective one-electron Breit-Pauli (BP) operator shown in Equation 1.   

!" 



  =       ∙ ̂ (1)



2   



where e is the electronic charge, me is the mass of electron, c is the speed of light,  is the effective 

nuclear change,  = # × %# is the orbital angular momentum operator of electron i with respect to

nucleus α, and ̂ is the spin angular momentum operator. The SOC term between a particular singlet

state &' and a triplet manifold () was evaluated by summing the three matrix elements with the equation presented below in Equation 2:





 0()2 30 + 01&' 0  0()5 30 + 01&' 0  0()6230 (2) * = /01&' 0    + ,-. 





BP spin-orbit matrix elements (SOMEs) were calculated using PySOC38 which can be interfaced with Gaussian 09 easily. We note that PySOC provides a Python interface based on MolSOC39 written in Fortran90. TD-DFT calculations were performed with M06-HF/aug-cc-pVDZ (Sapporo-2012-DZP with diffuse functions40 for I) level of theory using Gaussian 09 Rev. D01.41 All electron basis set was used because the effective nuclear charges available in PySOC are not parametrized with ECPs.42 It is noteworthy that the benchmarking work of PySOC showed that SOMEs evaluated with TD-DFT are not sensitive to basis set size and agree well with high-level multiconfigurational methods. Intersystem crossing rates (kISC) were calculated with Marcus equation for thermally activated process described in Equation 3, where ∆E is the adiabatic energy difference between Tk and Sj excited states (i.e. E[Tk (at optimized geometry in the Tk electronic state)] – E[Sj (at optimized geometry in the Sj excited state)]) and λ is the reorganization energy (i.e. E[Sj(Tk geometry)] − E[Sj(Sj geometry)]). T is temperature (in K) and room temperature (298.15 K) is used for all kISC calculations. 78 9&' → () ; =



< 7@ (A 4A7@ (

The phosphorescence rates (kPH) from a particular triplet state Tk to S0 were calculated with the perturbative expression in Equation 4, where ν is the frequency of the emitted photon in the process of

L M phosphorescence; H- and Hare the vertical energy gaps between Tk and Sn and between S0 and Tm

respectively; μO,L and μP,M are transition dipole moments between singlet and triplet states

respectively. Apart from SOC matrix elements, which were obtained from TD-DFT, the vertical

ACS Paragon Plus Environment

5

The Journal of Physical Chemistry 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 17

energy gaps and the transition dipole moments were obtained from EOM-CCSD computations. The summations in the equation run through the first three singlet and first three triplet excited states localized on benzaldehyde. 2 TUVW

7QR (() → &5 ) =  SXY Z [  \∑L

d