The effect of conformational isomerism on the optical properties of bis(8-oxyquinolato) diboron complexes with a 2,2′-biphenyl backbone

Mateusz Urban a, Patrycja Górka a, Krzysztof Nawara b, Krzysztof Woźniak c, Krzysztof Durka *a and Sergiusz Luliński *a
aFaculty of Chemistry, Warsaw University of Technology, Noakowskiego 3, 00-664 Warsaw, Poland. E-mail: kdurka@ch.pw.edu.pl; serek@ch.pw.edu.pl
bFaculty of Mathematics and Science, Cardinal Stefan Wyszyński University, Dewajtis 5, 01-815 Warsaw, Poland
cBiological and Chemical Research Centre, Department of Chemistry, University of Warsaw, Żwirki i Wigury 101, 02-089 Warsaw, Poland

Received 5th August 2018 , Accepted 11th September 2018

First published on 12th September 2018


A fluorescent bis(8-oxyquinolato) diborinic complex with a central 2,2′-biphenyl backbone 1 and its octafluoro analogue 2 were synthesized to study the optical–structural relationship of sterically encumbered molecules featuring close intramolecular π-stacking interactions involving chromophore units. The crystal structure of 1 revealed a unique π-stacked arrangement of two pendant phenyl groups and two 8-oxyquinolato ligands (Q) located in the inner part of the complex. Unlike 1, the closely related complex 2 features conformational isomerism, and two major forms, namely 2-syn and 2-anti, are observed in solution to a varying extent depending on the solvent polarity. Form 2-syn, a geometrical analogue of 1, is preferable in polar solutions, whereas its rotational isomer 2-anti featuring π-stacking interactions between the terminal phenyl group and Q ligand dominates in benzene and chloromethane solutions. The observed conformational equilibria strongly affect the optical properties of the system, specifically leading to a significant increase of the quantum yield of emission (from 22% in MeCN to 38% in benzene) accompanied by a bathochromic shift (Δλ = 10 nm) of absorption and hypsochromic shifts (Δλ = −8 nm) of emission spectra with decreasing solvent polarity. This effect was ascribed to the variation in frontier orbital distributions.


Introduction

Since the discovery of the electroluminescence properties of tris(8-oxyquinolinato)aluminium (AlqQ3),1,2 wide-ranging research has been conducted on the application of metal chelate complexes in optoelectronics. Organoboron compounds show analogous Lewis acidic behaviour to heavier Group 13 congeners and they are able to form complexes with various chelating ligands. In addition, they turned out to be more thermally and chemically resistant compared to aluminium, gallium or zinc complexes. Due to these fundamental properties, boron compounds seem to be ideal candidates for the design of fluorescent materials with target application in optoelectronic devices. Furthermore, the macroscopic properties of organoboron complexes such as absorption/emission wavelength, quantum yield of emission and charge transport capabilities can be properly modified for the desired functionality by altering the type of ligand and its substitution. Thus, much effort has been devoted to the synthesis and structural modifications of various chelate organic complexes. Recently, numerous examples of such systems were published including both discrete monomers3–13 and polymeric structures.14–19

Other important factors bearing a great impact on the macroscopic properties of luminescent systems are related to the molecular conformation and molecular environment effects. They include intermolecular interaction with solvent or with other molecules of the same type in concentrated solutions or in the solid state. In this sense, it is possible to tune the emission colour and quantum yield by means of conformational and solid-state structural diversification. Since most of the fluorophores suffer from aggregation-caused quenching (ACQ) due to π–π stacking interactions in the formed aggregates, much effort has been devoted to designing molecular structures which would preclude solid-state aggregation. Precise molecular engineering gave rise to fluorescent materials displaying aggregation-induced emission (AIE).20 Their emission is amplified due to the restriction of the molecular motions upon aggregation, while the sterically hindered substituents prevent molecules from engaging in π-stacking interactions with neighbouring chromophore units. Further studies on the structure–property relationships allowed for the conclusion on other important structural effects that may influence the fluorescence. For instance, it was found that efficient fluorescence in the solid state is favoured upon the formation of the so-called J-aggregates, which involve specific slip-type stacking of molecules.21–24

Furthermore, the luminescence properties can also depend on the molecular conformation. However, since conformational variability induces different molecular packing modes, the conformational and structural effects are co-dependent and, in most cases, cannot be simply separated. This was clearly demonstrated in the studies on luminescent organic polymorphs.25–33 Exceptionally, Zhang and co-workers in their study on the polymorphic forms of 4,4′-(thiazolo[5,4-d]-thiazole-2,5-diyl)bis(N,N-diphenylaniline) demonstrated the distinct influence of molecular conformation on the luminescence properties of a studied system.34 Specifically, the emission wavelength and its efficiency is associated with the dihedral angle between the π-electron molecular units affecting their conjugation. A similar effect was observed for unsymmetrically substituted 1,3-diaryl-β-diketones, where the fluorescence quantum yield strongly depends on the twisting angle along the central C–C(O) bond.35 In addition, a Schiff base compound displaying a different emission colour that strictly resulted from a different molecular conformation was recently presented.36 Regarding the solution studies, in most cases the observed photophysical properties are averaged over all conformers, which usually result from the lability of the pendant group, while the rigid chromophore unit hardly contributes. Even if a molecule adopts several stable conformations, the contribution of each individual conformation to the overall luminescence is usually not distinguishable as their abundance is not controllable. In this regard, studies on the conformation–luminescence dependence in solutions are rather unique.

In this work we propose a new strategy for tuning the luminescence properties of a compound through its conformational variability. It is based on the diversification of the through-space arrangement of two chromophore units. Recently, we have found that this can be realized in sterically hindered semi-rigid bis(chromophore) systems. Specifically, 9,10-dihydro-9,10-diboraanthracene bis(8-oxyquinolato) complexes (Scheme 1a) have improved photo- and electroluminescence with respect to simpler diarylboron 8-oxyquinolates, which is supposedly attributed to the specific conformational arrangement of the two Q chromophores.37 The presented complexes can adopt the bent conformation (apart from the symmetrical one) where two Q ligands interact by means of an intramolecular C–H⋯O hydrogen bond. Based on TD-DFT calculations, the HOMO–LUMO transition in this conformer was interpreted as ligand-to-ligand charge transfer (LLCT). This is in contrast to other quinolinato complexes including multichromophore systems, where a π–π* transition within the chromophore unit is typically observed.


image file: c8dt03197e-s1.tif
Scheme 1 Structures of bis(oxyquinolinato) complexes based on (a) diboraanthracene and (b) 2,2′-biphenyl cores.

In this work we have employed novel semi-rigid diboron bis(oxyquinolinato) complexes based on a central 2,2′-biphenyl core (Scheme 1b). We have observed that the specific combination of extensive π–π and CH–π interactions between the Q ligands and biphenyl units with a significant steric hindrance factor may lead to the stabilization of the conformational isomers which in turn strongly affects fluorescence quantum yield. The presented results clearly demonstrate the impact of relatively subtle structural effects on the optical properties of luminescent organic molecules.

Results and discussion

Synthesis

Compound 1 was prepared using a simple, two-step protocol shown in Scheme 2. At first, 2-bromobiphenyl was converted to the respective boronic ester 3via Br/Li exchange followed by the addition of B(OMe)3 and subsequent quenching with Me3SiCl. In the second step, 2,2′-dilithiobiphenyl was generated by the treatment of 2,2′-dibromobiphenyl with n-BuLi. Then it was treated with 3 followed by quenching of an ate-complex with Me3SiCl and addition of 8-hydroxyquinoline. A similar approach was used for the preparation of 2 starting with 2,2′-dibromotetrafluorobiphenyl. Compounds 1 and 2 were isolated in moderate to good yields (45% and 63%, respectively) as yellow crystalline solids. 1 is properly soluble only in chlorinated organic solvents, however, fluorination enhanced the solubility and 2 is soluble also in other common organic solvents such as acetone, toluene or DMSO. It should be noted that 1 is susceptible to hydrolytic cleavage with traces of water present in commercially available solvents (Fig. S10, ESI) leading to the restoration of 8-hydroxyquinoline and starting bis(borinic) acid. Therefore, its reliable spectroscopic characterization was only achieved using anhydrous solvents and the samples were prepared in a glove-box. However, the use of a perfluorinated biphenyl core in 2 greatly improved the hydrolytic stability of the studied system which can be ascribed to the higher Lewis acidity of the boron atoms (Fig. S11, ESI). A similar stabilizing effect of aryl ring fluorination was previously observed for other diarylborinic (N,O) chelate complexes.38
image file: c8dt03197e-s2.tif
Scheme 2 Synthesis of complexes 1 and 2. Conditions: i: (a) n-BuLi, THF, −78 °C, (b) B(OMe)3, (c) Me3SiCl; ii (X = H): n-BuLi, THF/Et2O, −78 °C; ii (X = F): n-BuLi, Et2O, −78 °C; iii (X = H): 3 (2 equiv.) + (1 equiv.) 2,2′-dilithiobiphenyl, THF/Et2O, −100 °C, (b) Me3SiCl; iii (X = F): 3 (2 equiv.) + (1 equiv.) 2,2′-dilithiotetrafluorobiphenyl, Et2O, −78 °C, (b) HCl/Et2O (2 M); iv: 8-HQ.

Crystal structures and isomerism

Single crystals of 1 were grown by evaporation of a solution in CHCl3. The molecular structure of 1 features a unique comb-like arrangement (syn) of aromatic rings locked by π-stacking interactions (Fig. 1a) of two terminal phenyl rings and two Q ligands located in the inner part of the complex. Q–Q stacking can be properly described as an aromatic donor–acceptor interaction:39 the distance between the centroids of electron-deficient pyridine and electron-rich phenolate rings of two adjacent Q ligands is equal to 3.859(2) Å whereas the distance between the central C9 atoms of these ligands is equal to 3.766(2) Å. It should be noted that the Q ligands are not strictly parallel to each other as the dihedral angle between the mean planes of boron-Q fused tricyclic systems is equal to 9.37(2)°. Furthermore, the distance between the centroid of the terminal Ph ring and the C9 atom of the Q ligand is 3.549(2) Å whereas the dihedral angle between the mean planes of both the cyclic systems is 22.21(2)°. Another important structural feature is a short C–H⋯π contact (2.370(2) Å) between the α-H atom of the Q ligand and one of the ring centroids of the central biphenyl core. As a result of π-stacking and C–H⋯π interaction the central biphenyl rings deviate from perpendicular alignment by about 10°. The supramolecular packing comprises two main structural motifs supported by weaker, van der Waals in nature, contacts with other molecules. The aromatic rings from neighbouring molecules interlock with each other to form a [110] molecular zipper (Fig. 1a, top-right). The chains are further linked through the dimeric C–H⋯O hydrogen bonds (dC⋯O = 3.323(2) Å) propagating in the [001] direction (Fig. 1a, bottom-right).
image file: c8dt03197e-f1.tif
Fig. 1 Molecular structure of (a) 1 and (b) 2 along with the presentation of the main structural motifs. Intramolecular π–π stacking, and C–H⋯π and C–H⋯F interactions have been marked as brown, red and yellow dashed lines, respectively. Hydrogen atoms are omitted for clarity. (c) Molecular overlay through the best fitting of the central biphenyl rings by means of the least squares method (colour coding: 1-magenta, 2-cyan).

Single crystals of 2 were grown by evaporation of a solution in benzene. Molecules adopt the anti conformation resulting from the rotation along the central C–C bond in the fluorinated biphenyl core (Fig. 1b). The benzene rings of a central biphenyl core are almost perpendicular to each other with a dihedral angle of 89.51(1)°. In the case of 2, a close intramolecular π-stacking occurs only between the terminal Ph rings and the Q ligands: the distance between the Ph centroid and the central C9 atom of the Q ligand is 3.373(1) Å. The dihedral angle between the mean planes of the discussed cyclic systems is 16.80(1)°. The anti conformer displays close to parallel alignment of the boron-bound benzene rings from the terminal biphenyl groups with the distance between the corresponding centroids being about 4.3 Å. Importantly, the anti conformer does not feature any intramolecular C–H⋯π interaction. Instead, a very close intramolecular C–H⋯F contact of only 2.256(2) Å (dC⋯F = 2.910(2) Å, aCHF = 125.24(2)°) occurs between the α-H atom of the Q ligand and the fluorine atom located at the 3-position of the biphenyl ring (sum of the vdW radii of H and F atoms is 2.67 Å).

The average B–N, B–O and B–C bond lengths are consistent with the values found in the literature for tetracoordinate organoboron (N,O) chelate complexes (Fig. S1, ESI).40 In both structures the boron–carbon bonds with the central biphenyl ring are longer than those with the external ones, which is due to the considerable intramolecular strain of the inner part of the molecule. Furthermore, the comparison between the syn- and anti-conformers of 1 and 2, respectively, reveals important differences in the B–N and B–O distances (Table 1). Specifically, the B–N bond length in 1 is shorter by about 0.02 Å, which is accompanied by the elongation of the B–O bond to a similar extent.41 Variations in the bond lengths originate from the different orientation of the Q moieties with respect to the biphenyl ring. To fully describe the conformational differences, let us consider the relative spatial arrangements of molecular fragments in terms of stereochemical isomerism.

Table 1 The comparison of the boron–heteroatom bond lengths in the studied systems. Note that molecule 1 is bisected by the C2 rotational axis, while the whole molecule 2 is asymmetric and values for both boron centres are provided
  d B–N d B–O d B–Cint d B–Cext
1 1.631(3) 1.529(3) 1.647(3) 1.627(3)
2 1.651(2)/1.650(2) 1.511(2)/1.508(2) 1.646(2)/1.636(2) 1.619(2)/1.625(2)


Due to the presence of two asymmetric boron centres three different stereoisomers can be expected including (1BR,2BR), (1BS,2BS) and an achiral meso-(1BR,2BS) isomer. Furthermore, high steric hindrance in the 2,2′-substituted central biphenyl system restricts the free rotation along the C–C single bond and thus molecules show optical activity due to biphenyl dissymmetry (atropisomerism) abbreviated as (bphR) or (bphS). Thus 3 stereoisomers can be distinguished: (bphR)-(1BR,2BR), (bphS)-(1BR,2BR) and meso-(bphR/S)-(1BR,2BS), along with the corresponding optical isomers ((bphS)-(1BS,2BS), (bphR)-(1BS,2BS)). Finally, taking into account the syn and anti conformational isomerism, there can be 6 energetically non-equivalent forms in total (10 with enantiomers included), wherein it should be kept in mind that the biphenyl C–C rotation causes the (bphR)/(bphS) isomerization. Fig. 2 summarizes the discussed stereochemical considerations and shows the relationships between all forms.


image file: c8dt03197e-f2.tif
Fig. 2 Stereochemical relations between all studied forms. The corresponding synanti conformers resulted from the C–C bond rotation are located at the same relative positions on the graphs.

Both compounds crystallize from racemic mixtures and thus adopt centrosymmetric space groups (1C2/c, 2P[1 with combining macron]). In the case of 1, the form syn-(bphR)-(1BR,2BR) along with its optical isomer were exclusively found both in the solid state and solution. The measured crystal structure of 2 obtained from the benzene solution is composed of anti-(bphS)-(1BS,2BS) and its enantiomer anti-(bphR)-(1BR,2BR). However, as indicated by PXRD diffraction analysis (Fig. 3), crystallization from more polar solvents favours the formation of a different crystal phase which, based on solution NMR-studies, can be ascribed to the syn-(bphS)-(1BR,2BR) rotamer. It seems that the abundance of a particular isomer depends on the stabilizing/destabilizing effects of competitive intramolecular forces that occur between the Q moiety and biphenyl ring. Specifically, in the case of syn isomers one of the biphenyl aromatic rings is engaged in CH(1)/CF(2) interaction with Q (pyridine/phenolate), while the second ring exposes the π-cloud pointing toward the other side of the Q moiety (phenolate/pyridine, accordingly) (Fig. 4a). The rotation of the molecule along the central biphenyl C–C bond changes the intramolecular contact scheme. In the anti conformers the intramolecular interaction with the biphenyl hydrogen (1) or fluorine (2) atom located in the vicinity of the boron group persists, while the second aromatic group from the biphenyl ring is too far to be involved in any specific interaction with the Q moiety. Instead, the interaction is formed with the terminal biphenyl group (Fig. 4b).


image file: c8dt03197e-f3.tif
Fig. 3 PXRD patterns obtained for compound 2 crystallized (by evaporation to dryness) from benzene and acetone and compared to the theoretical pattern generated from the single-crystal X-ray structure.

image file: c8dt03197e-f4.tif
Fig. 4 The presentation of intramolecular forces that affect the molecule stability of (a) 1 and (b) 2. The potentially stabilizing interactions are shown as red dashed lines, while the repulsive ones as coloured spheres. The experimentally confirmed forms are given in bold. Note that only one-half of the molecule has been depicted.

The DFT(PBE042/6-311++G(d,p)43,44) calculations show that in the case of 1 the most stable is the syn-(bphR)-(1BR,2BR) isomer, that is, the one exclusively found in the solution and solid state (Table 2). This is a result of relatively strong intramolecular C–H⋯π interactions between the pyridine moiety and central biphenyl ring (Fig. 4), supported by the C–H⋯O interaction between the phenolato moiety and the second biphenyl ring (dC⋯O = 2.781(2) Å, aCHO = 107.61(1)°). The flipping of quinolinato rings induces significant internal strain in the system and thus the isomer syn-(bphR)-(1BS,2BS) is less stable by 53 kJ mol−1. The phenolate oxygen atom points toward the π-cloud, while a close C–H⋯H–C contact (dH⋯H = 2.222 Å) occurs on the pyridine side. As a result, the dihedral angle between the biphenyl aromatic planes decreases to 65.1°, whereas the Q moieties slide alongside their planes leading to the weakening of their mutual stacking. In terms of thermodynamics, the rotation of the C–C bond is disadvantageous, too. Irrespective of the Q ligand ring orientation, all theoretically generated anti conformers display relatively short intramolecular C–H⋯H–C contacts leading to substantial molecular distortions. This is especially noticeable for anti-(bphS)-(1BR,2BR), which is less stable by 104.6 kJ mol−1 than the corresponding syn-(bphR)-(1BR,2BR) rotamer. Interestingly, both syn-(bphR)-(1BS,2BS) and anti-(bphS)-(1BS,2BS) rotamers are characterized by comparable stability, nonetheless, both being less stable by about 50 kJ mol−1 with respect to syn-(bphR)-(1BR,2BR). The achiral forms (synmeso and antimeso) are located in-between the aforementioned extremes in terms of their geometry and energy.

Table 2 The molecular energy of the isomers related to the most stable form in the series. All ΔE values are given in kJ mol−1. The values for the experimentally observed forms are given in bold. Level of theory: PBE0/6-311++G(d,p)
Conformation syn-(bphR) anti-(bphS)
(1BR,2BR) meso (1BS,2BS) (1BR,2BR) meso (1BS,2BS)
1 0.0 27.0 53.0 104.6 76.2 52.2
2 65.5 37.7 12.4 71.8 42.6 0.0


A substantially different intramolecular bond relationship is observed for compound 2. In this case all biphenyl hydrogen atoms are substituted with fluorine atoms, which, depending on the Q ligand orientation, can be involved either in stabilizing C–H⋯F or repulsive O⋯F intramolecular interactions. Even if the C–H⋯F interaction with organic fluorine is considered weak, it replaces the destabilizing C–H⋯H–C internal contact occurring in 1. It is also expected that close F⋯O contact with the phenolato ring will be strongly avoided, while the C–H⋯π interactions will be weaker with the electron-deficient perfluorinated biphenyl system.45–47 Theoretical calculations confirmed our expectations. The most stable form is anti-(bphS)-(1BS,2BS) along with its rotamer syn-(bphR)-(1BS,2BS) (ΔE = 12.4 kJ mol−1). The former owes its increased stability to the combination of two pairs of C–H⋯O and C–H⋯F stabilizing intramolecular interactions. In the syn rotamer the C–H⋯F contact is also present, however, the phenolate oxygen atom is now located in the region of π-electron density of the biphenyl ring. This causes some distortion (dihedral angle between the internal biphenyl aromatic planes decreases to 66.1°), which partially reduces the stabilizing effect of Q–Q stacking. Thus, both the considered rotamers of 2 are characterized by comparable energy. The flipping of Q ligands by 180° leads to considerable strain resulting from unfavourable C–H⋯H–C and O⋯F contacts in the anti and syn conformer with an energy increase by 71.8 kJ mol−1 and 65.5 kJ mol−1, respectively. According to our expectations, these conformers are not observed experimentally.

Because of the fact that the energy difference between the 2-syn-(bphR)- and 2-anti-(bphS)-(1BS,2BS) conformers is small, their appearance in solution or in the solid state may be affected by molecular environment effects such as temperature or intermolecular interaction with solvent or neighbouring molecules in the solid state. As a result of internal Q–Q stacking, the syn conformer seems to be more compact in comparison to the anti rotamer. Nevertheless, their volumes and surface areas are similar (Table S5, ESI). More important effects are related to the shape of the molecule, the charge distribution (see the electrostatic potential map in Fig. S18, ESI) and its capabilities for effective engagement in intermolecular contacts. The performed periodic DFT calculations in CRYSTAL09 for 1-syn and 2-anti and their corresponding theoretically generated fluorinated (1-anti) or hydrogen-substituted (2-syn) counterparts show that the molecules are, indeed, very well stabilized in their experimentally determined crystal environments. The calculations were performed with a full geometry and unit-cell optimisation procedure, preserving the crystal symmetry. The theoretical 2-syn as well as the 1-anti structures were generated from the experimental 1 and 2 structures, respectively, by simple H/F alternation with C–H/F bond length shortening/elongation. The structure 1-syn is energetically more stable than 1-anti by 61 kJ mol−1 (with respect to the single molecule), while 2-anti is more favourable than 2-syn by 66 kJ mol−1. Thus, it can be supposed that the preferential conformational selection should be regulated by the self-assembly process. Naturally, the crystal may adopt other symmetries that would facilitate the occurrence of more advantageous intermolecular interactions, eventually leading to the formation of a new crystal phase. Such a situation is observed for 2, where the PXRD pattern differs significantly depending on the crystallization conditions applied (Fig. 3). Based on the comparison with the PXRD pattern generated from the crystal structure of 2, crystallization from benzene leads to the formation of anti-(bphS)-(1BS,2BS), which is contaminated by a minor fraction presumably ascribed to the syn-(bphR)-(1BS,2BS) rotamer. This stays in agreement with solution NMR studies, where anti-(bphS)-(1BS,2BS) dominates in less-polar solutions. Conversely, the crystallization from acetone solution mainly produces the 2-syn conformer and presumably 2-anti as a minor fraction, as indicated by the analysis of the peak position and intensity on the corresponding PXRD patterns. In addition, DSC analysis revealed different thermal behaviour of materials obtained from the acetone and benzene solutions of 2 (Fig. 5). In the first case, two distinct exothermic peaks of first-order phase transitions are observed (Fig. 5, A, B). Considering the high energetic barrier for the synanti transformation, it can be assumed that both conformers are preserved in the solid state upon heating. Thus, the observed transitions are apparently related to polymorphic changes, i.e., crystal structure reorganization rather than synanti isomerization. In turn, the material enriched with the 2-anti conformer displays a second-order glass-type transition at ca. 123 °C (Fig. 5, C). It is interesting to note that the 2-syn phase melts at a higher temperature (T = 297.1 °C) than the 2-anti phase (T = 293.0 °C), which is also accompanied by the higher fusion enthalpy for the former one (−29.4 kJ mol−1vs. −23.1 kJ mol−1). It can be concluded that the crystal lattice of 2-syn is energetically more favourable than 2-anti, however, the abundance of a particular fraction seems to be regulated by the crystallization conditions.


image file: c8dt03197e-f5.tif
Fig. 5 Thermal behaviour (DSC plots) of compound 2 obtained upon the evaporation of acetone (black curve) and benzene (blue curve) solutions.

Structural behaviour in solution

The 11B NMR spectra of 1–2 in CDCl3 showed single broad resonances at 11–12 ppm, typical of diarylborinic Q chelates3 thus confirming the presence of tetracoordinate boron atoms. However, the 1H NMR spectrum of 1 in CDCl3 shows some distinct features which point to its remarkable structural specificity. Firstly, two dd-type multiplets are observed at 4.78 (J = 5.2, 1.1 Hz) and 5.70 (J = 8.3, 5.2 Hz) ppm, i.e., far outside the range expected for aromatic protons. Based on the 1H,1H COSY and 1H,13C HSQC experiments, these signals were assigned to the protons at the α position and β to the nitrogen atom in the pyridine part of the Q ligand (A and B, respectively in Fig. 6a), typically observed in Ph2BQ at 8.6 and 7.6 ppm, respectively. The very large upfield shift of ca. 4 ppm for the former proton clearly indicates its close C–H⋯π contact with one of the benzene rings of biphenyl moieties (Fig. 4a top-left). This lends strong support to the view that the molecular structure 1-syn-(bphR)-(1BR,2BR) (+enantiomer) determined by X-ray diffraction (Fig. 1a) and identified as the most stable by theoretical calculations is essentially retained in solution. However, complex 1 exhibits dynamic character as one of the resonances of the pendant 2-biphenylyl group is very broad at room temperature and becomes narrower at 50 °C. This indicates significant rotational restrictions along the C–C bond linking two aromatic rings of the terminal 2-biphenylyl substituent.
image file: c8dt03197e-f6.tif
Fig. 6 1H NMR spectra of (a) 1 and (b) 2, and (c) the 19F NMR spectrum of 2 in CDCl3. Signals assigned as the syn form are marked with blue colour, anti form – purple and meso – green.

The 1H and 19F NMR spectra of 2 in CDCl3 show three sets of signals in a 1[thin space (1/6-em)]:[thin space (1/6-em)]0.4[thin space (1/6-em)]:[thin space (1/6-em)]0.06 ratio. However, it is important to state that elemental analysis proves the purity of the compound and thus points to the higher structural complexity of 2 with respect to 1. Two major sets of signals can be assigned to the two most abundant conformers which do not interconvert rapidly on the NMR time scale even at higher temperatures (up to 100 °C in DMSO). The minor one from these two conformers features one strongly upfield shifted doublet at 5.33 ppm (J = 7.6 Hz). However, in this case the resonance exhibits a larger coupling constant, corresponding to the H atom in the phenolate ring at the ortho position with respect to the oxygen atom (2-F in Fig. 6b). Based on the analogy to 1, the upfield shift was induced by C–H⋯π contact in a similar comb-like structure, due to the location of the proton in the shielding cone of the perfluorinated aromatic ring. In the theoretically calculated 2-syn-(bphR)-(1BS,2BS) geometry, one can notice analogous positioning of 1-B (Fig. 4a, top-left) and 2-F (Fig. 4b, top-right) protons with respect to the biphenyl core as in the case of 1. Notably, a doublet (J = 5.2 Hz) at 8.38 ppm assigned to proton 2-A is only slightly shifted upfield with respect to an analogous signal in Ph2BQ. A similar situation was observed for the third set of signals (doublet at 5.59 ppm, J = 7.5 Hz), indicating the presence of another conformer featuring the syn geometry. However, it shows 8 signals in 19F NMR, indicating the lowered symmetry of a molecule with respect to the more abundant syn and anti conformers (Fig. 6c). We suppose that this set of signals originates from the 2-synmeso form (the third most stable one according to the calculations shown earlier). The most abundant conformer in the CDCl3 solution of 2 was identified as 2-anti-(bphS)-(1BS,2BS) (+enantiomer) as it lacks such a close C–H⋯π contact (Fig. 4b, bottom-right), thus signals of the corresponding protons (A, B, F) are not shifted upfield. Finally, we would like to remind that according to our calculations both chiral rotamers of 2 are characterized by comparable stability thus supporting their equilibration in solutions.

The geometrical differences between the syn and anti forms lead to different charge distributions within the corresponding molecules (Fig. S18, ESI). In the anti conformers both quinolinato ligands point toward opposite directions, thus except for meso conformers, the overall dipole moment of the molecule is close to zero. In contrast, in the case of syn conformers both quinolinato moieties are parallel and adopt the same spatial orientation, inducing a considerable dipole moment of the molecule (Table 3, Fig. S17, ESI). Therefore, the conformational equilibrium in 2 was studied in more detail using various solvents (Table 4). We have found that in benzene there is a strong preference for the anti conformer whereas in polar solvents such as DMSO, acetone, acetonitrile and methanol the syn conformer predominates. Overall, rough correlations persist between Ksyn/anti and common solvent polarity parameters such as the dielectric constant ε and the permanent dipole moment μ (Fig. S22, ESI). However, the solvent polarity should be considered as its overall solvation ability, which in general also involves directional inductive and dispersion, electron-pair donor and acceptor (EPD, EPA) and solvophobic interactions. Therefore, for a more meaningful description we used the empirical parameter ENT originally proposed by Reichardt.48ENT is a dimensionless figure calculated on the basis of the solvatochromic characteristics of selected pyridinium N-phenolate betaine dyes. We found that a fairly good linear correlation between ln[thin space (1/6-em)]Ksyn/anti and ENT was achieved when MeOD was not taken into account (R2 = 0.94) (Fig. 7). We suppose that deviation from the linear trend observed for methanol is due to its specificity as it is the only protic solvent within the studied series. Thus, it seems that highly directional H-bonding interactions (supposedly with the phenolato oxygen atom) may strongly contribute to the stabilization of various forms of 2.


image file: c8dt03197e-f7.tif
Fig. 7 syn/anti equilibrium in 2 as a function of the empirical solvent polarity parameter ENT. The fit of the linear function by the least squares method is presented by a black solid line. Using a dashed black line, regression without the CD3OD point was depicted.
Table 3 Comparison of the dipole moments (D) of the different forms of 1 and 2 derived from DFT calculations. Experimentally observed forms are given in bold
  syn-(bphR) anti-(bphS)
(1BR,2BR) meso (1BS,2BS) (1BR,2BR) meso (1BS,2BS)
1 5.91 8.93 9.11 3.84 7.10 2.46
2 8.26 12.20 11.76 1.58 5.98 0.33


Table 4 Equlibrium between the syn and anti conformers of 2 as a function of various solvent polarity parameters
Solvent K syn/anti ε μ/D ln(Ksyn/anti) E NT[thin space (1/6-em)]b
a K syn/anti was determined by the division of the 19F NMR signal integrals at −133 pm (syn) and −138 pm (anti) in CDCl3 or integrals of corresponding signals in other solvents. b Definition and numerical data for a number of solvents can be found in Reichardt's review.48
C6D6 0.11 2.3 0.00 −2.17 0.111
CDCl3 0.40 4.8 1.15 −0.92 0.259
CD2Cl2 1.15 8.9 1.55 0.14 0.309
(CD3)2CO 1.89 20.7 2.85 0.63 0.355
(CD3)2SO 1.99 46.5 3.90 0.69 0.444
CD3CN 3.18 35.9 3.45 1.16 0.460
CD3OD 2.22 32.7 1.66 0.80 0.762


In addition, the effect of temperature on the behaviour of 2 was investigated. In benzene and CDCl3 the syn/anti equilibrium was not strongly affected by warming to 65 and 50 °C, respectively (Fig. S58–61, Fig. S66–69, ESI). In DMSO, a higher temperature (100 °C) could be applied which resulted in the lowering of Ksyn/anti (1.36@100 °C vs. 1.99@25 °C) (Fig. S62–65 ESI). Moreover, the 2-synmeso is present to a significant extent (ca. 15%) under these conditions. We suppose that the appearance of the asymmetric form results from the Q ligand rearrangement at elevated temperature originating from the dynamic character of the B–N dative interaction.40 However, the nature of this process, as well as the synanti transformation mechanism is not clear. According to our calculation, the C–C bond rotation requires a huge amount of energy (300–350 kJ mol−1, the exact value is uncertain due to the convergence problems). Thus, it is supposed that the B–N interaction has to be temporary broken during the rotation, whereas its rearrangement and resulting molecular conformation depend on its overall thermodynamic stability rather than kinetic effects.

Optical properties

The optical properties of 1–2 in solution were investigated by UV-Vis absorption and photoluminescence spectroscopy under ambient conditions and the data are provided in Table 5. A plot of the recorded spectra is depicted in Fig. 8.
image file: c8dt03197e-f8.tif
Fig. 8 (a) Overlay of the UV-Vis absorption spectra of 2 in various solvents and (b) the overlay of the normalized emission spectra of 2 in various solvents.
Table 5 Photophysical properties of 1 and 2
Compound/solvent λ max/nm ε/M−1 cm−1 λ em/nm Δ/cm−1 Φ /% τ/ns
a Excited at the absorption maximum at the longest wavelength c = 1–2 × 10−5 M; standard: Coumarin 153 in EtOH, c = 4 × 10−6 M (rt). b Reliable data cannot be obtained due to observed sample degradation.
Ph2BQ/DCM 395 3700 511 5750 28
1/DCM 394 4150 515 5960 19 b
1/benzene 397 3900 b
1/acetone 394 3750 b
2/benzene 401 5700 504 5100 38 32.1
2/DCM 398 5400 506 5360 35 35.1
2/acetone 392 4850 512 5980 23 30.3
2/MeCN 392 4800 512 5980 22 32.7
2/MeOH 392 4050


Complex 1 has long-wave absorption and emission bands at 394 nm and 515 nm (in DCM), respectively (Table 5). Thus, a slight hypsochromic shift of the absorption and bathochromic shift of the emission band relative to the reference compound Ph2BQ (λmax = 395 nm, λem = 511 nm) are observed. In the case of 2, the effect of solvent variation on the optical properties was investigated. A slight negative solvatochromism, i.e., a hypsochromic shift of the absorption band was observed with increased solvent polarity (λmax = 401 nm in benzene vs. 392 nm in MeCN and acetone) whereas the reverse trend was found for emission spectra (λmax = 504 nm in benzene vs. 512 nm in MeCN and acetone). This results in larger Stokes shifts in polar solvents. Also a decrease of molar absorption coefficient was observed with the increase of solvent polarity.

Compound 1 is a moderately good emitter as its fluorescence quantum yield is 0.19 in DCM solution. The emission quantum yields in other solvents were not measured as they could not be reliably determined due to the hydrolytic instability of the complex. This was done for the more stable complex 2. It is important to note that the emission quantum yields for 2 vary depending on the solvent. In benzene Φf = 0.38, i.e., it is relatively high (for comparison, Φ = 0.28 for Ph2BQ). As the polarity of the solvent increases the quantum yield systematically decreases to 0.22 in MeCN (Table 5). We suppose that these changes may reflect different proportions of conformers in the used solvents as indicated by NMR data. It is remarkable that a solution rich in conformer 2-anti, whose formation is strongly preferred in benzene, exhibits the highest quantum yield. In contrast, the conformer 2-syn seems to be a significantly less efficient emitter.

Theoretical calculations (DFT) revealed that the distribution of frontier orbitals changes upon rearrangement from the 2-syn to 2-anti conformer (Fig. 9). For the 2-syn form, both Q rings are equally involved in creating HOMO and LUMO orbitals. In contrast, in the 2-anti conformer the HOMO orbital spans across one of the Q ligands and partially over the second ligand. Conversely, the LUMO orbital is spread mostly onto the second Q ligand with some contribution from the first Q ligand (and also bridging biphenyl groups). Consequently, in the 2-syn conformation, π–π* electron transition is observed within the Q ligands, while for the 2-anti conformer the HOMO–LUMO transition can be rationalized in terms of ligand-to-ligand charge transfer (LLCT). The separation of molecular orbitals leads also to the slight increase of the HOMO molecular energy levels by 0.11 eV and a decrease of the LUMO by 0.16 eV. However, the origin of this effect is not clear. We suppose that it probably results from the fact that in the syn form the HOMO orbital spans over two chromophore units gaining some additional stability. In the case of the anti conformer, the LUMO is spread onto one Q ligand with some non-negligible contributions from the second ligand and the bridging biphenyl moiety. As a consequence, the HOMO–LUMO energy gap for the syn-conformer is higher, and thus a hypsochromic shift of the absorption maximum is observed.


image file: c8dt03197e-f9.tif
Fig. 9 Plot of the frontier molecular orbitals of 2-syn and 2-anti. Blue and red areas correspond to the negative and positive signs of the function, respectively, depicted with an isovalue of 0.2 e Å−3. Full data for all conformers are provided in Table S6, ESI.

The emission spectra of compound 2 are independent of the excitation wavelength. This suggests that both forms (syn and anti) emit in exactly overlapped spectral regions, or the observed emission results from the single rotameric form of compound 2. The existence of identical emission spectra for both forms is not supported by theoretical calculations of the HOMO and LUMO orbitals. In order to gain better insight into the nature of observed emission, we have measured the fluorescence lifetime decays (τ, Table 5). All measured fluorescence decays are monoexponential and the values range from 30 to 35 ns depending on the solvent environment. These data lend support to the view that emission originates from only one rotameric form – anti, while the syn form does not show fluorescence. The process of internal conversion for the syn form is most likely very fast due to the strong overlap of π orbitals, therefore no emission of this form can be observed. Fluorescence lifetimes, which do not differ substantially from one solvent to another support the view that the transition between both rotamers is rather slow as it is not observed during the lifetime of an excited state. If it were the case, shortening of the fluorescence lifetime should be observed due to the dynamic quenching of fluorescence. The emission spectra and the fluorescence lifetime decays characterize the anti rotamer, while the fluorescence quantum yield is determined for the mixture of rotameric forms (as both forms absorb light, but only one emits). Both forms (syn and anti) exist to a varying extent depending on the solvent. Therefore, the emission quantum yields are dependent on both: (a) the ratio of rotamers, and (b) the absorbance of both forms at the excitation wavelength. On the other hand, it should be noted that 1 – existing solely in the syn form – exhibits moderately good luminescence with a quantum yield of 0.19. However, substitution with multiple fluorine atoms in 2 induces significant electronic changes and therefore direct comparison should not be made.

Qualitatively, similar conclusions on the conformation–property relationship were derived from our recent studies of diboraanthracene complexes with bis-spiro structures bearing two 8-oxyquinolinato37 and also from Zhang's report on 2-(1-phenyl-1H-benzo[d]imidazole-2-yl)phenolate or 2-(2-pyridyl)phenolate ligands.49 It was found that diboraanthracene complexes are far better emitters with Φ values in the range of 40–60% than the corresponding monochromophore Ph2B(L) complexes, while their absorption and emission wavelengths are hardly affected. It was concluded that this emerged from LLCT excitation character attributed to the specific spatial orientation of the two chromophore units. However, the conformational effect was not distinguishable from solution studies, as fast equilibration between the two different forms was observed. In the present study, the interconversion between the two forms of 2 is strongly hampered, yet occurs to some extent. This allowed us to gain a clearer picture on the conformation–optical property relationships. The presented conclusions seem to be in line with those formulated by us and Zhang49 and indicate that optical properties may be enhanced by conformational effects favouring LLCT-type transitions. Last but not least, as stated by Zhang and co-workers, such a separation of frontier orbitals can also improve electron and hole transport capabilities of a material, which in turn would be beneficial for final device performance.

Conclusions

The fluorescence properties of organic compounds result not only from their chemical structures, but also from the molecular conformation and crystal packing effects. Since, due to the rigidity of the chromophore unit, most traditional fluorescent materials do not display any distinct conformation–luminescence dependence, we have proposed a different strategy, based on the diversification of intramolecular chromophore–chromophore spatial orientation. Our general considerations on the conformation–property relationship led us to the conclusion that it can be accomplished in sterically hindered semi-rigid systems. Consequently, two sterically congested diboron complexes with Q ligands were obtained and fully characterized. The steric hindrance at the boron atoms seems to be responsible for decreased stability of boron atom chelation in 1, manifested by gradual ligand abstraction in dilute solutions due to cleavage with traces of water. However, fluorination of the central 2,2′-biphenyl core resulted in an improvement of hydrolytic stability. Interestingly, this modification has also a significant impact on the structural behaviour of the studied system. Both experimental data and theoretical calculations indicate that compound 1 tends to exist as a comb-like conformer stabilized by multiple π⋯π and CH⋯π interactions. In the case of 2, the situation is more complicated as two major conformers equilibrate in solution to a varying extent depending on the kind of solvent and temperature. Thus, changes in the proportion of the conformational isomers of 2 should be taken into account as an important factor affecting the optical properties of the complex. Specifically, the frontier orbital distribution in the two major conformers of 2 is different and consistent with switching of the excitation pathway between intraligand (ILCT) and ligand-to-ligand charge transfer (LLCT) mechanisms. The presented results amply demonstrate the effect of conformational isomerism on the luminescence properties. It should be taken into account in designing more extended oligomeric and polymeric materials where controlling an intramolecular arrangement of building blocks (e.g. stacking of aromatic chromophores) may be crucial for fine tuning of the light emitting and charge transport properties.

Experimental section

General comment

Starting materials 2,2′-dibromobiphenyl, 2-bromobiphenyl, 8-hydroxyquinoline, trialkyl borates, n-BuLi (1.6 M and 10 M in hexane), and t-BuLi (1.7 M in pentane) were used as received from Aldrich Chemical Co. without any additional purification. 2,2′-Dibromooctafluorobiphenyl and Ph2BOEt were prepared according to the literature procedures.4,50 THF and Et2O were dried by heating to reflux with sodium or potassium and benzophenone, and distilled under argon. Reactions and manipulations involving air and moisture-sensitive reagents were carried out under an argon atmosphere.

Characterization data

1H, 11B, 13C, 19F NMR and 2D NMR spectra were recorded on Bruker Advance III 300 MHz and Agilent NMR 400 MHz DDR2 spectrometers. Variable temperature 1H and 19F NMR were recorded on a Varian 500 MHz spectrometer. 1H and 13C chemical shifts were referenced to TMS using known chemical shifts of solvent residual peaks. 11B and 19F NMR chemical shifts are given relative to BF3·Et2O and CFCl3, respectively. In the 13C NMR spectra the resonances of boron-bound carbon atoms were not observed in most cases as a result of their broadening by the quadrupolar boron nucleus. The used deuterated solvents were dried with 3 Å molecular sieves and samples were prepared under an argon atmosphere. Preparation of 1 and all VT NMR samples was conducted in a glove-box using Young tubes. HR-MS analyses were performed on a GCT Premier mass spectrometer equipped with an EI ion source and a Maldi SYNAPT G2-S HDMS spectrometer equipped with an ESI ion source.

Synthesis

Dimethyl 2-biphenylylboronate (3). A solution of 2-bromobiphenyl (19.25 g, 83 mmol) in anhydrous Et2O (20 mL) was added dropwise over 0.5 h to a stirred solution of n-BuLi (9.0 mL, 90 mmol, 10 M in hexane) in THF (150 ml) at −78 °C. A yellow precipitate formed. After 0.5 h, the mixture was warmed to −50 °C, stirred for 0.5 h and then cooled to −78 °C. A solution of B(OMe)3 (10 mL, 90 mmol) in Et2O (10 mL) was added dropwise and stirred for 1 h. The mixture clarified. The mixture was warmed up to −20 °C, treated with Me3SiCl (12 mL, 95 mmol) and warmed to 40 °C. The mixture was stirred for 3 h and then the volatiles were removed under reduced pressure. Et2O (40 mL) was added and the obtained slurry was filtered under an argon atmosphere. The filter cake was washed three times with Et2O (3 × 30 mL) whereupon the combined filtrate was concentrated in vacuo. The crude product was purified by fractional distillation under reduced pressure to give 3 as a colourless liquid, b.p. 80–82 °C (10−3 Torr). Yield 11.6 g (62%). 1H NMR (300 MHz, CDCl3) δ 7.57–7.34 (m, 9H), 3.39 (s, 6H) ppm. 13C NMR (75 MHz, CDCl3) δ 145.1, 143.5, 132.4, 129.2, 128.8, 128.3, 128.1, 127.3, 126.7, 52.2 ppm. 11B NMR (96 MHz, CDCl3) 28.5 ppm. The obtained compound is highly susceptible to hydrolysis and should be stored and handled under an inert gas atmosphere. Therefore, HRMS and elemental analysis could not give reliable results due to very fast degradation of 3 to the corresponding boronic acid under standard analytical conditions.
Complex 1. A solution of 2,2′-dibromobiphenyl (0.31 g, 1.0 mmol) in anhydrous Et2O (5 ml) was added dropwise to a stirred solution of n-BuLi (1.25 ml, 2.0 mmol, 1.6 M in hexane) in a mixture of anh. THF (5 ml) and Et2O (5 ml) at −78 °C. After 1 h, the reaction mixture was cooled to −100 °C and the solution of 3 (0.45 g, 2.0 mmol) in anh. Et2O (2 ml) was added dropwise and stirred for 1 hour. The suspension was warmed to −80 °C and Me3SiCl (0.25 ml, 2.0 mmol) was added. The mixture was allowed to slowly reach r.t. and 8-hydroxyquinoline (0.29 g, 2.0 mmol) was added in one portion; the mixture was left stirring overnight. The yellow precipitate was filtered off, washed with Et2O and dried. Then it was washed with water, EtOH (99.8%), Et2O and hexane, and dried under vacuum to obtain pure 1 as a bright yellow powder, mp. 275–285 °C. Yield 0.35 g (45%). 1H NMR (500 MHz, CDCl3, 25 °C) δ = 8.07 (dd, J = 7.8, 1.5, 1H, G), 7.34 (td, J = 7.5, 1.4 Hz, 1H, H), 7.30 (t, J = 7.9 Hz, 1H, E), 7.26–7.19 (m, 1H, ArH), 7.22–7.15 (m, 2H, ArH), 7.01 (td, J = 7.4, 1.5 Hz, 1H, I), 6.90–6.84 (m, 1H, ArH), 6.75 (dd, J = 7.6, 0.8 Hz, 1H, F), 6.72 (dd, J = 7.6, 1.5, 1H, J), 6.62 (dd, J = 8.2, 1.1 Hz, 1H, C), 6.53 (dd, J = 8.2, 0.8 Hz, 1H, D), 6.45 (–d, J = 7.0 Hz, 2H, L), 6.32 (broad-tt, J = 7.4, 1.3 Hz, 1H, M), 6.21 (broad, 2H, K), 5.70 (dd, J = 8.3, 5.2 Hz, 1H, B), 4.78 (dd, J = 5.2, 1.1 Hz, 1H, A) ppm. 13C NMR (101 MHz, CDCl3) δ = 157.7, 146.4, 146.0, 144.3 (CA), 144.2, 136.9, 136.9, 135.3 (CC), 134.7 (CG), 130.8 (CJ), 130.7 (CE), 130.4, 128.0 (broad, CL), 126.8, 126.6, 126.5 (CI), 126.2, 126.1, 125.9 (broad), 124.8 (CM), 120.8 (CB), 112.0 (CD), 108.2 (CF) ppm. 11B NMR (96 MHz, CDCl3) 11.7 ppm. HRMS (ESI): C54H38B2N2O2 [M + Na]+ calcd: 791.3017. Found: 791.3011.
Complex 2. A solution of 2,2′-dibromooctafluorobiphenyl (0.46 g, 1.0 mmol) in Et2O (5 mL) was added dropwise to a stirred solution of n-BuLi (1.25 mL, 2.0 mmol, 1.6 M in hexane) in Et2O (10 ml) at −78 °C. After 2.5 h a solution of 3 (0.45 g, 2.0 mmol) in Et2O (2 mL) was added dropwise resulting in the formation of a white suspension. After 1.5 h HCl (2 M solution in Et2O 1.0 mL, 2.0 mmol) was added. The mixture was warmed to −40 °C, and then 8-hydroxyquinoline (0.29 g, 2.0 mmol) was added in one portion. After warming up to room temperature over 1 hour the reaction was stirred overnight. A pale yellow precipitate was filtered off, washed with Et2O and dried. The crude solid was washed with water, EtOH (99.8%), Et2O and hexane, and dried under vacuum to obtain pure 2 as a bright yellow powder, m.p. 275–303 °C. Yield 0.58 g (63%). 1H NMR (500 MHz, C6D6, 25 °C) δ = 8.33 (d, J = 7.6 Hz, 1H), 8.22 (dd, J = 5.2, 1.1 Hz, 1H, A), 7.12 (d, J = 8.0 Hz, 1H), 7.13–7.09 (m, 1H), 7.09–7.03 (m, 2H), 6.99 (dd, J = 8.3, 1.1 Hz, 1H, C), 6.50 (d, J = 7.9 Hz, 1H), 6.49 (dd, J = 8.2, 0.5 Hz, 1H), 6.42–6.37 (m, 1H), 6.35 (broad, 2H, M), 6.22 (dd, J = 8.3, 5.3 Hz, 1H, B) ppm. 19F NMR (470 MHz, C6D6, 25 °C) δ = −127.52 (dd, J = 24.5, 11.0 Hz), −137.20 (dd, J = 24.0, 11.0 Hz), −159.48 (dd, J = 24.5, 21.0 Hz), −160.69 (t, J = 23.0 Hz). 13C NMR (101 MHz, C6D6, 25 °C) δ = 157.7, 146.6, 144.7, 142.0, 141.8, 138.3 (CC), 137.9, 137.7(CA), 132.0, 129.8, 127.1, 126.4, 126.0, 124.7, 122.4 (CB), 112.4, 110.0 ppm. 11B NMR (96 MHz, CDCl3) δ = 11.1 ppm. 11B NMR (96 MHz, acetone-d6) δ = 11.0 ppm. HRMS (ESI): C54H30B2F8N2O2 [M + H]+ calcd: 913.2444. Found: 913.2439. Anal. calc. for C54H30B2F8N2O2: C, 71.08; H, 3.31; N, 3.07; found: C, 70.80 H 3.23 N 3.21. HRMS (ESI): C54H30B2F8N2O2 [M + H]+ calcd: 913.2444. Found: 913.2439.

Diphenylborinic 8-oxyquinolate (Ph2BQ)

To a solution of Ph2BOEt (0.40 g, 1.9 mmol) in anhydrous EtOH 8-hydroxyquinoline (0.28 g, 1.9 mmol) was added in one portion. The mixture was stirred at r.t. for 5 hours. A bright yellow precipitate was filtered off, washed with cold EtOH, hexane and dried. The product (0.51 g) was purified by recrystallization from the EtOH/CHCl3 mixture and dried under vacuum. Yield 0.42 g (71%). 1H NMR (300 MHz, CDCl3) δ = 8.59 (dd, J = 5.1, 1.1 Hz, 1H), 8.39 (dd, J = 8.3, 1.1 Hz, 1H), 7.67 (t, J = 8.0 Hz, 1H), 7.61 (dd, J = 8.3, 5.1 Hz, 1H), 7.52–7.40 (m, 4H), 7.35–7.21 (m, 7H), 7.19 (dd, J = 7.8, 0.7 Hz, 1H) ppm.

X-ray structural measurements and refinement details

The single crystals of 1 and 2 were obtained by slow evaporation of CHCl3 and benzene solutions, respectively. They were measured at 100 K on a SuperNova diffractometer equipped with an Atlas detector (Cu-Kα radiation, λ = 1.54184 Å). Data reduction and analysis were carried out with the CrysAlisPro program.51 All structures were solved by direct methods using SHELXS-9752 and refined using SHELXL-2014.53 Crystallographic Information Files (CIFs) have been deposited with the Cambridge Crystallographic Data Centre as supplementary publications no. 1857286 (1) and 1857287 (2).

Crystal data for 1: C54H38B2N2O2, Mr = 768.48 a.u.; monoclinic; C 2/c; a = 21.956(7) Å, b = 10.8039(9) Å, c = 18.984(9) Å, β = 119.63(4) °, V = 3914(3) Å3; dcalc = 1.304 g cm−3; μ = 0.877 mm−1; Z = 4; F(000) = 1440; number of collected/unique reflections (Rint = 5.67%) = 16[thin space (1/6-em)]851/3483, R[F]/wR[F] (I ≥ 3σ(I)) = 7.22%/18.32%, Δρ(min/max)res = −0.247/+0.307 e Å−3.

Crystal data for 2: C54H30B2F8N2O2, Mr = 912.42 a.u.; triclinic; P[1 with combining macron]; a = 11.4963(2) Å, b = 11.8442(2) Å, c = 15.9638(3) Å, α = 77.924(2), β = 76.456(2), γ = 81.249(2), V = 2054.29(7) Å3; dcalc = 1.475 g cm−3; μ = 0.963 mm−1; Z = 2; F(000) = 932; number of collected/unique reflections (Rint = 2.19%) = 30[thin space (1/6-em)]427/7905, R[F]/wR[F] (I ≥ 3σ(I)) = 3.46%/8.87%, Δρ(min/max)res = −0.209/+0.405 e Å−3.

Despite several attempts, we did not manage to obtain the single crystals of 2-syn suitable for X-ray analysis.

Powder diffraction

Powder X-ray diffraction (PXRD) analyses were carried out on a Bruker D8 Advance powder diffractometer equipped with a Cu radiation source (Cu Kα, λ = 1.54184 Å), a zero background sample holder and a LYNXEYE detector. Data were collected over a 2θ range of 3°–60° in the Bragg–Brentano (θ/θ) horizontal geometry (flat reflection mode) with a generator setting of 40 kV and 40 mA, step size of 0.04°, and an exposure time per step of 3 s.

DSC

Differential scanning calorimetric analysis was conducted by differential scanning calorimetry (DSC) (Model DSC 1, Mettler-Toledo) under the flow of nitrogen. The heating rate was 5 K min−1 from 25 °C to 320 °C in the first cycle and from 25 °C to 500 °C in the second cycle. Sample decomposes after melting.

Optical properties

The UV-Vis absorption spectra of 1, 2 and Ph2BQ (as the reference compound) were recorded using a Hitachi UV-2300II spectrometer. The emission spectra were recorded using a Hitachi F-7000 spectrofluorometer, equipped with a photomultiplier detector, calibrated using the Spectral Fluorescence Standard Kit certified by the BAM Federal Institute for Materials Research and Testing.54 The measurements were performed at room temperature, according to published procedures.55,56 Suprasil quartz cuvettes (10.00 mm) were used. 1.5 nm slits were used for absorption and 2.5 nm slits were used for emission spectra. To eliminate any background emission, the spectrum of pure solvent was subtracted from the sample spectra. QY were determined in diluted solutions (A < 0.1 for the longest wavelength band) by comparison with the known standard – Coumarin 153 in ethanol (c = 4 × 10−6 mol dm−3, QY = 0.38).57 The concentrations of complex solutions were in the range of 1–2 × 10−5 mol dm−3 (concentration was adjusted to reach similar absorbance to the absorbance of the reference solution at the excitation wavelength). To calculate the QY the following formula was used:
image file: c8dt03197e-t1.tif
where F is the relative integrated photon flux of the sample (x) and standard (st), A is the absorbance at the excitation wavelength, and n is the refractive index of the used solvents.
image file: c8dt03197e-t2.tif

Photon fluxes (F) were calculated by integration of the corrected spectra (Ic), obtained by the division of the intensity of the emission spectra by the spectral responsivity (s) at the corresponding wavelengths (λem). All measurements were carried out at room temperature.

The fluorescence lifetime measurements were acquired using a Becker-Hickl TSCPC module Simple-Tau 150. The emitted photons were collected using a multiwavelength detection system: a polychromator and a 16 channel TCSPC detector (PML-16, Becker-Hickl model), which registers 16 individual fluorescence decays from the 200 nm spectral range (12.5 nm per channel). For excitation the supercontinuum laser Fianium Whitelase Micro was used. The excitation wavelength (430 nm) was selected using a continuously variable bandpass filter from the Delta Optical Thin Film. The IRF signal measured for this setup has 0.130 ns FWHM. The data were collected using SPCM software. The analysis of fluorescence decays was performed using SPCImage 5.0 Software using an incomplete multiexponential model.

Theoretical calculations

The single-molecule optimisations were performed for all conformers of 1 and 2 at the PBE042/6-311++G(d,p)43,44 level of theory. The minima were confirmed by vibrational frequency calculations within the harmonic approximation (no imaginary frequencies). During the calculations no symmetry constraints were applied. The initial geometries were extracted from the crystal structures, then, if necessary, hydrogen/fluorine atoms were replaced by the fluorine/hydrogen atoms to generate new isomers. The optimised structures are presented in Fig. S13–16 in the ESI. Excited state geometries, along with the emission spectra (Fig. S21, ESI), were obtained using TD-DFT methods with the same basis set, starting with the geometries obtained from the ground state optimizations. The periodic DFT calculations were performed using CRYSTAL09.58 All energy computations within the CRYSTAL09 program package were performed at the DFT(B3LYP) level of theory with the POB triple-zeta valence + polarization basis set (TZVP).59–62 The results were corrected for dispersion63–65 and the basis set superposition error (BSSE).66

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financed by the National Science Centre (Narodowe Centrum Nauki, Grant No. UMO-2015/19/D/ST5/00735). The X-ray measurements were undertaken at the Crystallographic Unit of the Physical Chemistry Laboratory, Chemistry Department, University of Warsaw. We would like to thank Paulina H. Marek from the Warsaw University of Technology for performing PXRD measurements. The authors acknowledge The Interdisciplinary Centre of Mathematical and Computational Modelling in Warsaw (grant no. G33-14) for providing computer facilities (Gaussian09). Periodic DFT calculations within the Crystal09 package have been carried out at the Wroclaw Centre for Networking and Supercomputing (http://www.wcss.pl), grant no. 285.

References

  1. C. W. Tang and S. A. VanSlyke, Appl. Phys. Lett., 1987, 51, 913–915 CrossRef CAS.
  2. C. W. Tang, S. A. VanSlyke and C. H. Chen, J. Appl. Phys., 1989, 65, 3610–3616 CrossRef CAS.
  3. G. Wesela-Bauman, P. Ciećwierz, K. Durka, S. Luliński, J. Serwatowski and K. Woźniak, Inorg. Chem., 2013, 52, 10846–10859 CrossRef CAS PubMed.
  4. M. Urban, K. Durka, P. Jankowski, J. Serwatowski and S. Luliński, J. Org. Chem., 2017, 82, 8234–8241 CrossRef CAS PubMed.
  5. D. Li, H. Zhang and Y. Wang, Chem. Soc. Rev., 2013, 42, 8416–8433 RSC.
  6. D. Frath, J. Massue, G. Ulrich and R. Ziessel, Angew. Chem., Int. Ed., 2014, 53, 2290–2310 CrossRef CAS PubMed.
  7. D. Frath, K. Benelhadj, M. Munch, J. Massue and G. Ulrich, J. Org. Chem., 2016, 81, 9658–9668 CrossRef CAS PubMed.
  8. M. Tanaka, S. Muraoka, Y. Matsui, E. Ohta, A. Sakai, T. Ogaki, Y. Yoshimoto, K. Mizuno and H. Ikeda, ChemPhotoChem, 2016, 1, 188–197 CrossRef.
  9. Y. Qi, Y. Wang, G. Ge, Z. Liu, Y. Yu and M. Xue, J. Mater. Chem. C, 2017, 5, 11030–11038 RSC.
  10. A. D'Aléo, M. H. Sazzad, D. H. Kim, E. Y. Choi, J. W. Wu, G. Canard, F. Fages, J.-C. Ribierre and C. Adachi, Chem. Commun., 2017, 53, 7003–7006 RSC.
  11. M. Gon, K. Tanaka and Y. Chujo, Angew. Chem., Int. Ed., 2018, 57, 6546–6551 CrossRef CAS PubMed.
  12. Y. Hisamune, T. Kim, K. Nishimura, M. Ishida, M. Toganoh, S. Mori, D. Kim and H. Furuta, Chem. – Eur. J., 2017, 24, 4628–4634 CrossRef PubMed.
  13. S. Kim, Y. Zhou, N. Tohnai, H. Nakatsuji, M. Matsusaki, M. Fujitsuka, M. Miyata and T. Majima, Chem. – Eur. J., 2017, 24, 636–645 CrossRef PubMed.
  14. J. Wang, B. Jin, N. Wang, T. Peng, X. Li, Y. Luo and S. Wang, Macromolecules, 2017, 50, 4629–4638 CrossRef CAS.
  15. A. Nagai, J. Miyake, K. Kokado, Y. Nagata and Y. Chujo, J. Am. Chem. Soc., 2008, 130, 15276–15278 CrossRef CAS PubMed.
  16. Y. Tokoro, A. Nagai, K. Kokado and Y. Chujo, Macromolecules, 2009, 42, 2988–2993 CrossRef CAS.
  17. H. Li and F. Jäkle, Angew. Chem., Int. Ed., 2009, 48, 2313–2316 CrossRef CAS PubMed.
  18. C. Dou, Z. Ding, Z. Zhang, Z. Xie, J. Liu and L. Wang, Angew. Chem., Int. Ed., 2014, 54, 3648–3652 CrossRef PubMed.
  19. R. Zhao, C. Dou, Z. Xie, J. Liu and L. Wang, Angew. Chem., Int. Ed., 2016, 55, 5313–5317 CrossRef CAS PubMed.
  20. J. Mei, N. L. C. Leung, R. T. K. Kwok, J. W. Y. Lam and B. Z. Tang, Chem. Rev., 2015, 115, 11718–11940 CrossRef CAS PubMed.
  21. S. Choi, J. Bouffard and Y. Kim, Chem. Sci., 2014, 5, 751–755 RSC.
  22. F. Würthner, T. E. Kaiser and C. R. Saha-Möller, Angew. Chem., Int. Ed., 2011, 50, 3376–3410 CrossRef PubMed.
  23. S. Herbst, B. Soberats, P. Leowanawat, M. Lehmann and F. Würthner, Angew. Chem., Int. Ed., 2017, 56, 2162–2165 CrossRef CAS PubMed.
  24. K. Cai, J. Xie and D. Zhao, J. Am. Chem. Soc., 2014, 136, 28–31 CrossRef CAS PubMed.
  25. J. D. Wuest, Nat. Chem., 2012, 4, 74–75 CrossRef CAS PubMed.
  26. D. Yan and D. G. Evans, Mater. Horiz., 2013, 1, 46–57 RSC.
  27. P.-Y. Gu, G. Liu, J. Zhao, N. Aratani, X. Ye, Y. Liu, H. Yamada, L. Nie, H. Zhang, J. Zhu, D.-S. Li and Q. Zhang, J. Mater. Chem. C, 2017, 5, 8869–8874 RSC.
  28. S.-J. Yoon, J. W. Chung, J. Gierschner, K. S. Kim, M.-G. Choi, D. Kim and S. Y. Park, J. Am. Chem. Soc., 2010, 132, 13675–13683 CrossRef CAS PubMed.
  29. G. Wesela-Bauman, S. Luliński, J. Serwatowski and K. Woźniak, Phys. Chem. Chem. Phys., 2014, 16, 22762–22774 RSC.
  30. Z. He, L. Zhang, J. Mei, T. Zhang, J. W. Y. Lam, Z. Shuai, Y. Q. Dong and B. Z. Tang, Chem. Mater., 2015, 27, 6601–6607 CrossRef CAS.
  31. W. Yang, C. Liu, S. Lu, J. Du, Q. Gao, R. Zhang, Y. Liu and C. Yang, J. Mater. Chem. C, 2018, 6, 290–298 RSC.
  32. J. Shi, S.-J. Yoon, L. Viani, S. Y. Park, B. Milián-Medina and J. Gierschner, Adv. Opt. Mater., 2017, 5, 1700340 CrossRef.
  33. Z. Zhang, X. Song, S. Wang, F. Li, H. Zhang, K. Ye and Y. Wang, J. Phys. Chem. Lett., 2016, 7, 1697–1702 CrossRef CAS PubMed.
  34. K. Wang, H. Zhang, S. Chen, G. Yang, J. Zhang, W. Tian, Z. Su and Y. Wang, Adv. Mater., 2014, 26, 6168–6173 CrossRef CAS PubMed.
  35. X. Cheng, F. Li, S. Han, Y. Zhang, C. Jiao, J. Wei, K. Ye, Y. Wang and H. Zhang, Sci. Rep., 2015, 5, 9140 CrossRef CAS PubMed.
  36. H. Liu, Z. Lu, B. Tang, Z. Zhang, Y. Wang and H. Zhang, Dyes Pigm., 2018, 149, 284–289 CrossRef CAS.
  37. K. Durka, I. Głowacki, S. Luliński, B. Łuszczyńska, J. Smętek, P. Szczepanik, J. Serwatowski, U. E. Wawrzyniak, G. Wesela-Bauman, E. Witkowska, G. Wiosna-Sałyga and K. Woźniak, J. Mater. Chem. C, 2015, 3, 1354–1364 RSC.
  38. G. Wesela-Bauman, M. Urban, S. Luliński, J. Serwatowski and K. Woźniak, Org. Biomol. Chem., 2015, 13, 3268–3279 RSC.
  39. C. R. Martinez and B. L. Iverson, Chem. Sci., 2012, 3, 2191–2201 RSC.
  40. The Cambridge Structural Database search resulted in 227 crystal structures of (N,O) boron chelate complexes. C. R. Groom, I. J. Bruno, M. P. Lightfoot and S. C. Ward, Acta Crystallogr., Sect. B: Struct. Sci., Cryst. Eng. Mater., 2016, 72, 171–179 CrossRef CAS PubMed.
  41. K. Durka, R. Kamiński, S. Luliński, J. Serwatowski and K. Woźniak, Phys. Chem. Chem. Phys., 2010, 12, 13126–13136 RSC.
  42. C. Adamo, J. Chem. Phys., 1999, 110, 6158–6170 CrossRef CAS.
  43. R. Seeger and J. A. Pople, J. Chem. Phys., 1980, 72, 650–654 CrossRef.
  44. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785–789 CrossRef CAS.
  45. J. I. Wu, F. G. Pühlhofer, P. von R. Schleyer, R. Puchta, B. Kiran, M. Mauksch, N. J. R. van, E. Hommes, I. Alkorta and J. Elguero, J. Phys. Chem. A, 2009, 113, 6789–6794 CrossRef CAS PubMed.
  46. T. Okazaki and K. K. Laali, Org. Biomol. Chem., 2006, 4, 3085–3095 RSC.
  47. N. Dias, K. Shimizu, P. Morgado, E. J. M. Filipe, J. N. Canongia Lopes and F. Vaca Chávez, J. Phys. Chem. B, 2014, 118, 5772–5780 CrossRef CAS PubMed.
  48. C. Reichardt, Chem. Rev., 1994, 94, 2319–2358 CrossRef CAS.
  49. Z. Zhang, Z. Zhang, K. Ye, J. Zhang, H. Zhang and Y. Wang, Dalton Trans., 2015, 44, 14436–14443 RSC.
  50. R. Filler, A. E. Fiebig and M. Y. Pelister, J. Org. Chem., 1980, 45, 1290–1295 CrossRef CAS.
  51. CrysAlis Pro Software, Oxford Diffraction Ltd., Oxford, U.K., 2010 Search PubMed.
  52. G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008, 64, 112–122 CrossRef CAS PubMed.
  53. G. M. Sheldrick, Acta Crystallogr., Sect. C: Struct. Chem., 2015, 71, 3–8 Search PubMed.
  54. D. Pfeifer, K. Hoffmann, A. Hoffmann, C. Monte and U. Resch-Genger, J. Fluoresc., 2006, 16, 581–587 CrossRef CAS PubMed.
  55. C. Würth, M. Grabolle, J. Pauli, M. Spieles and U. Resch-Genger, Nat. Protoc., 2013, 8, 1535–1550 CrossRef PubMed.
  56. U. Resch-Genger and K. Rurack, Pure Appl. Chem., 2013, 85, 2005–2013 CAS.
  57. A. M. Brouwer, Pure Appl. Chem., 2011, 83, 2213–2228 CAS.
  58. R. Dovesi, R. Orlando, B. Civalleri, C. Roetti, V. R. Saunders and C. M. Zicovich-Wilson, Z. Kristallogr., 2005, 220, 571–573 CAS.
  59. A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098–3100 CrossRef CAS.
  60. J. P. Perdew, Phys. Rev. B: Condens. Matter Mater. Phys., 1986, 33, 8822–8824 CrossRef.
  61. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785–789 CrossRef CAS.
  62. M. F. Peintinger, D. V. Oliveira and T. Bredow, J. Comput. Chem., 2013, 34, 451–459 CrossRef CAS PubMed.
  63. B. Civalleri, C. M. Zicovich-Wilson, L. Valenzano and P. Ugliengo, CrystEngComm, 2008, 10, 405–410 RSC.
  64. S. Grimme, J. Comput. Chem., 2006, 27, 1787–1799 CrossRef CAS PubMed.
  65. S. Grimme, J. Comput. Chem., 2004, 25, 1463–1473 CrossRef CAS PubMed.
  66. S. F. Boys and F. Bernardi, Mol. Phys., 1970, 19, 553–566 CrossRef CAS.

Footnote

Electronic supplementary information (ESI) available. CCDC 1857286 and 1857287. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c8dt03197e

This journal is © The Royal Society of Chemistry 2018