Weak Donor-/Strong Acceptor-Linked Anthracenyl π-Conjugates as Solvato(fluoro)chromophore and AEEgens: Contrast between Nitro and Cyano Functionality.

Steady development on photophysical behaviors for a variety of organic fluorophores inspired us to generate anthracene-based fluorescent molecules with a strong acceptor and a significantly weak donor through a π-spacer. Such molecules are found to have substantial twisted conformational orientations in the solid state and enhanced apolar character because of the attachment of tolyl or mesityl group with an anthracenyl core. Upon exposure to a variety of solvents, strong solvatochromism was noticed for 4-nitro compound (84 nm red shift) in contrast to the cyano analogue (18 nm red shift). Both these probes were highly emissive in apolar solvents while nitro-analogue, in particular, could discriminate the solvents of E T(30) (a measure of microscopic solvent polarity) ranging from 31 to 37. Thus, 4-nitro compounds can be successfully employed to detect and differentiate the apolar solvents. On the contrary, the 2-nitro analogue is almost nonemissive for the same range of solvents perhaps because of favorable excited-state intramolecular proton-transfer process. The fundamental understanding of solvatochromic properties through the formation of twisted intramolecular charge-transfer (TICT) state is experimentally analyzed by synthesizing and studying the π-conjugates linked to only benzene in place of nitro or cyanobenzene, which exhibits no solvatochromism and that helped finding the possible emission, originated from the locally excited state. Moreover, the molecular structures for these compounds are determined by the single-crystal X-ray diffraction studies to examine the change in emission properties with molecular packing and alignment in the aggregated state. The measurement of dihedral angles between the substituents and anthracenyl core was helpful in finding the possible extent of electronic conjugations within the system to decipher both solvatochromism and aggregation enhanced emission (AEE)-behavior. The cyano analogue exhibited prominent AEE-behavior, whereas nitro analogues showed the aggregation-caused quenching effect. The reason behind such dissimilarity in solvatochromism and AEE-behavior between cyano- and nitro-linked anthracenyl π-conjugates are also addressed through experimental outcomes.

Introduction

Environment-sensitive organic fluorophores are recognized as functional materials because of their potential applications in chemosensing, monitoring biological events, and finding unusual photophysical behaviors.[1] The presence of strong electron push–pull substituents to a fluorophore is a common practice to access solvatochromic fluorophores[2] through the formation of intramolecular charge-transfer (ICT) state.[3] The fluorophores with a combination of amine as a strong donor and nitro/cyano/carbonyl as a strong electron acceptor are familiar in the literature.[1−4] The push–pull chromophores based on π-deficient nitrogen heterocycles are also well-known for their strong solvatochromic behaviors.[5] Typically, nitro functionality is known to quench the fluorescence (Fl.) intensity with its direct attachment to the fluorophore because of the efficient intersystem crossing[6a] and the photodecomposition of nitro compounds through intra- or inter-molecular H-abstractions.[6b] Despite these common tendencies of nitroaromatics, many fluorophores were identified to exhibit excellent quantum yield in aprotic medium.[6,7] With an aim of finding unusual photophysical behavior, Konishi explored the solvatochromic fluorophores A–C (Figure ), attached with a relatively weak donor and strong acceptor.[8] The nitroaromatics A and B were reported to be strong fluorescent compounds where B, in particular, displayed dual Fl.[9] Further, a comparative study on the impact between cyano and nitro functionality toward solvatochromism for a biphenyl system with a strong (amine) donor was known in the literature studies[4a,1b] where the cyano group offered strong emission and solvent-dependence in comparison to nitro.[1b]

Figure 1

Reported weak donor and strong acceptor systems.

In the reported π-extended fluorene system (B, Figure ), cyano derivatives offered very high quantum yield with no Fl. solvatochromism, whereas nitro exhibited unique solvatochromism with poor quantum yield.[9a] Notably, all these reported fluorophores A–C (Figure ), were fluorescent only in polar solvents. However, investigations on such weak donor and strong acceptor systems (Figure ) are significantly limited in the literature studies. With this background, we create and explore weak donor- and strong acceptor-linked anthracene-based fluorophores that are synthesized through a straightforward, inexpensive, and effective strategy. Notably, anthracene is historically renowned as a valuable skeleton[10] and solvatochromic fluorophore.[11] On the basis of our recent research on π-conjugates,[12] we have herein designed anthracene-based extensive π-conjugates where toluene or mesitylene serves as a substantially weak donor and nitro or cyano as a strong acceptor. Comparative photophysical studies for nitro (1a–b, Scheme ) and cyano (2) analogues are performed to examine their Fl. solvatochromic properties. In contrast to reported fluorophores A–C (Figure ), the anthracenyl fluorophores, described herein, are strongly emissive in apolar solvents. A range of fluorescent colors was highly prominent in the naked eye under a UV lamp (365 nm) for different solvents of ET(30) ranging from 31 to 37 and thus, could be discriminated using this probe. Besides, our interest is to study the aggregation enhanced emission (AEE) properties for these molecules although the flat anthracene skeleton prefers to quench the Fl. intensity in the aggregated state.[10] In the literature studies, anthracene was coupled with tetraphenylethylene to show strong aggregation-induced emission (AIE) behavior[13a] and distyrylanthracene was identified as AIE active molecules with potential applications.[13b] Recently, we reported 9- or 10-aryl-substituted anthracenyl π-conjugates, and the AIE-properties were observed in the presence of heteroaromatic moiety.[12a] Particularly, cyano group-containing AIE molecules were very well studied.[13c] Thus, in this paper, the AIE-behaviors for all these molecules are examined and compared between cyano- and nitro-substituted π-conjugates. The effort has been made to understand the origin of the differences in solvatochromism and the AEE-behavior between nitro and cyano compounds. Isomeric nitro analogue and a simple π-conjugate without nitro or cyano linkage are also synthesized and inspected the photophysical properties to support and explain our findings. The molecular structures for most of these molecules are determined by single-crystal X-ray diffraction (XRD) studies, mainly to find the possible positions of the groups and the torsion angles/bond distances. The molecular packings originated from diverse intermolecular forces (supramolecular interactions) for such crystalline substances are also investigated to interpret the emission properties in the aggregated state.

Scheme 1

Synthesis of Anthracene-Based Fluorophores 1–4

Results and Discussion

Synthesis of the Fluorophores

Our recent research on an inexpensive synthesis of new arylated anthracenylphosphonates using simple but effective Friedel–Crafts arylation reactions made the accessibility of a variety of 9-arylanthracenyl phosphonates much easier.[11b] Thus, integrating an aryl group using expensive Pd/phosphine combinations is completely avoided. Being suitable precursors, these phosphonates are successfully utilized in the most inexpensive and efficient Horner–Wadsworth–Emmons reactions to afford all these conjugated fluorophores in ∼60–70% yield (Scheme ).

These compounds are characterized by 1H/13C NMR spectroscopy, and X-ray crystallographic studies were performed for all these compounds to confirm the trans-geometry unequivocally for such molecules. The trans-coupling could not be identified for some of these molecules as the aromatic proton signals were merged with the olefinic protons. In fact, the molecules 1a–b and 2 were synthesized initially and eventually other analogues 3 and 4 were synthesized to compare and study the emission properties of this class of molecules with different substituents.

Photophysical Studies with a Focus on Solvatochromic Properties

Predominantly, the absorption and Fl. behaviors for nitro compound 1a were studied at room temperature in air under the environment of a wide range of typical solvents varying from aprotic apolar (or slightly polar) [benzene, diethylether (DEE), heptane, hexane, pentane, toluene, mesitylene, o-xylene, p-xylene, 1,4-dioxane], aprotic polar [acetone, acetonitrile, tetrahydrofuran (THF)], polar protic (ethanol and water) to chlorinated solvents (carbon tetrachloride and chloroform). The broad signals on maximum absorption were with the range from λmax ≈ 396 to 403 nm, ascribed due to π–π* transitions (Figure S1a, Supporting Information). Such a negligible solvent effect toward the absorption spectrum for this molecule indicates that the ground electronic state for 1a remains unchanged on varying the solvent polarity. However, the compound 1a started glowing under a UV lamp (365 nm) majorly in aforementioned aprotic apolar to slightly polar solvents. The emission red shift with a change in the solvent polarity is significant. The Fl. λmax varying from 508 nm (in pentane) to 588 nm (in acetone), that is, 80 nm red shift with a change in the solvent polarity is noteworthy (Figure S2a, Supporting Information). Almost, no emission was observed in polar solvents such as ethanol, water, acetonitrile, and chloroform medium. With this solvatofluorochromic behavior of compound 1a, we were curious to replace toluene with mesitylene with an aim to deal with a compound having better accessibility, crystallinity, and Fl. intensity. Hence, our newly developed methodology could afford compounds 1b and the cyano analogue 2 conveniently (Scheme ). The photophysical behaviors for both these compounds (1b and 2) were examined in the above-mentioned solvents, and the corresponding data are summarized in Table (for compounds 1a and others, please see Tables S1–S3 in the Supporting Information). The absorption bands of compound 1b under different solvent polarity have appeared as a broad signal in the range of λmax = 401–407 nm (Figure a), reflecting a very weak red shift. Notably, all these compounds are poorly soluble in water. Thus, alike compound 1a, the sensitivity of absorption spectra on the solvent polarity for compound 1b is very weak because the local environment of the molecule in the ground and excited state remain unchanged during the very fast (10–15 s) absorption process.[14] In contrast, the emission spectrum of 1b displayed the solvatochromic effect very similar to 1a with the same range of solvents (Figure b). In the Fl. spectroscopy time scale, a molecule can relax and feel the confined environment in the ground and excited state and consequently, solvent effects on the Fl. for some fluorophores are quite common.[14] In slightly polar or apolar aprotic medium both 1a–b fluoresced with more intensity as shown in Figures b and 3, although the quantum yield (ΦF) for 1a was substantially lower (19%, Table S1) than the compound 1b (50%, Table ) in hexane.

Table 1

Spectrophotometric Data for Compound 1b

entry	solventa [ET(30)]	absorbance (λmax) (nm)	emission (λmax) (nm)	relative QY %b (absolute QY)c	Stokes shift (cm–1)
1	n-pentane (31)	401	510	54	5330
2	n-hexane (31)	402	512	50 (54)	5344
3	heptane (31.1)	402	514	50 (48)	5420
4	CCl4 (32.5)	406	535	36	5939
5	mesitylene (33.1)	404	536	30	6096
6	p-xylene (33.5)	404	540	26 (29)	6234
7	o-xylene (34.3)	404	544	26 (28)	6370
8	toluene (33.9)	404	545	22 (23)	6403
9	benzene (34.5)	404	549	20 (21)	6538
10	DEE (34.6)	401	546	17 (22)	6623
11	1,4-dioxane (36)	403	558	15 (16)	6893
12	THF (37.4)	403	568	2 (5)	7208
13	EtOAc (38.1)	402	566	2	7208
14	CHCl3 (39.1)	405
15	acetone (42.2)	402	594	negligible	8041
16	EtOH (51.9)	401
17	H2O (63.1)	407

In CHCl3; the compound was not fluorescent; in H2O and EtOH: clear solution was not obtained due to poor solubility and the Fl. could not be determined, likely because of the formation of highly stable excited state (TICT state).

Relative quantum yield in reference with quinine sulphate in 0.1 M H2SO4.

Absolute quantum yields were measured for few cases using integrating sphere and most of the cases the value deviates ±10%.[17]

Figure 2

(a) UV–vis and (b) Fl. spectra for compound 1b (10–5 M) under different solvent polarity (84 nm red shift).

Figure 3

Solvatochromic Fl. effect for the fluorophore 1b (10–5 M) (under UV lamp 365 nm); solvents (14) with observed (through naked eye) intense Fl. are only included here. (a) Pentane, (b) hexane, (c) heptane, (d) CCl4, (e) mesitylene, (f) p-xylene, (g) o-xylene, (h) DEE, (i) toluene, (j) benzene, (k) 1,4-dioxane, (l) THF, (m) acetone, and (n) water. A similar effect was also observed for the probe 1a.

(a) UV–vis and (b) Fl. spen class="Chemical">ctra for compound 1b (10–5 M) under different solvent polarity (84 nm red shift).

Relative quantum yield in reference with quinine sulphate in 0.1 M n class="Chemical">H2SO4.

Much higher quantum yield for 1b in comparison to 1a could be perhaps due to restricted intramolecular motion for sterically crowded mesitylene group along with the presence of more methyl groups responsible for σ–π interaction, a recent study by Konishi on pyrene system where pyrene fluorophores showed enhanced Fl. on increasing the number of alkyl groups directly attached to the pyrene core.[15]

The fundamental understanding of the macroscopic property like dipole moment for the solvents is inadequate to investigate such solvatochromic behavior of this fluorophore. Hence, the concept of microscopic solvent polarity ET(30)[16] is quite realistic, from which the environment sensitivity at a microscopic level has been considered. The probe 1b emitted very intense bluish green in most apolar solvents like pentane, hexane, and heptane [almost equal dipole moment (μD = 0) and ET(30): 31], whereas CCl4 [μD = 0 and ET(30): 32.5] and mesitylene [μD = 0.047 and ET(30): 33.1] solution radiated discrete light green. Because of the close polarity parameters for p-xylene [μD = 0.0 and ET(30): 33.5], the probe also fluoresced in the same wavelength (λem: 535 nm) similar to CCl4 and mesitylene. The light green turned into yellowish green (λem: 544 nm) in case of o-xylene [μD = 0.62 and ET(30): 34.3] and DEE [μD = 1.3 and ET(30): 34.6] solutions. The probe showed a light yellow Fl. in toluene [μD = 0.43 and ET(30): 33.9] but the emitted color was completely changed in benzene [μD = 0.0 and ET(30): 34.5]. In relatively more polar solvent like 1,4-dioxane [μD = 0.45 and ET(30): 36], the probe emitted a specific yellow color and faint reddish emission was observed from THF solution [μD = 1.75 and ET(30): 37.4]. Thus, the normalized Fl. spectra (Figure S3, Supporting Information) indicated 84 nm red shift with the nitroaromatic 1b and 80 nm red shift with the 1a by increasing the polarity of the solvent. As evident from Figure , this naked eye visualization using these new probes looks to be suitable in distinguishing the apolar solvents in the range of ET(30) ≈ 31–37. So far, reported fluorophores with weak donor and strong acceptor systems were fluorescent only in the polar solvent, not in apolar solvents. The probes are highly soluble in common organic solvents except for water and methanol (poorly soluble) and highly stable in the solid and solution state, indicating no chemical reactions with the solvent. It is noteworthy that the fluorosolvatochromic property remained unchanged for the compounds 1a–b even after 6 months, indicating a highly photostable nature for these compounds in the solution state.

The quantum yields of the probe 1b in different solvents are tabulated in Table where n class="Chemical">pentane, hexane, and heptane showed the maximum quantum yield ∼50–53%, a notable quantum yield for nitroaromatic compounds. Particularly, the quantum yield for weak donor and strong acceptor systems (−NO2) were reported to be 35% maximum in dimethylformamide.[9a] However, the quantum yield of 1b has appeared to be very low in polar solvents.

Next, of interest to us is the compound 2, cyano analogue of 1b. In an earlier report, the acceptor strength was compared with nitro and cyano groups for a given fluorophore and the effect was noteworthy.[4] Notably, 1-cyano-5-methylaminonaphthalene showed tremendous solvent-dependent properties,[1b] whereas no such effect is shown for the reported cyano analogue of B (fluorene system, Figure ). In fact, cyano analogues were less sensitive to solvent polarities in comparison to nitro for a strong donor/acceptor biphenyl system.[4a,4b] Thus, we were excited to see the photophysical properties of compound 2 in numerous solvents of various polarities. A structured absorbance band was observed for this compound with absorption maximum with the range of 397 (heptane) to 404 nm (water) reflecting an insignificant polarity effect on λmax (Figure S1b). This compound was fluorescent in apolar solvents [ET(30): 31–36] with maximum quantum yield in heptane (35%, Table S2) and exhibited only 18 nm red shift (Figure a) on polarity changes, which was not enough to discriminate the apolar solvents as observed for compounds 1a–b. Such moderate-to-weak changes could not be detected through the naked eye under the handy UV lamp (365 nm). Thus, the solvatochromic Fl. effect with the cyano analogue 2 was not promising. Meanwhile, the synthesis of compound 3 was planned to investigate whether the change in the position of a nitro compound makes any difference in the solvatochromic properties. However, this compound was somewhat fluorescent specifically in hexane but nonemissive in the visible region for all other solvents, exposed herein (see Figure S1c for absorption and Figure S2c for emission spectrum). Such nonfluorescent behavior for a different system such as 2-nitrodiphenylpolyenes was identified before.[6b] The excited-state intramolecular/intermolecular proton transfer for the molecule 3 might be responsible for such a property. However, all these π-conjugates consist of push–pull [donor−π–acceptor; (D−π–A)] systems with different strength and orientation. Furthermore, we were interested in the solvatochromic effect, if any, for the compound 4, a non push–pull system where nitro/cyanobenzene is replaced with only benzene. The compound 4 displayed the absorption maxima in the range of 392–402 nm (Figure S1d) and sharp emission spectra with λmax in the range 473–485 nm (Figure S2d) with a shift of 12 nm (Figure b) in various solvents of diverse polarity.

Figure 4

Normalized emission spectrum of (a) 2 (18 nm shift) and (b) 4 (12 nm shift) in different solvents. λex = 405 nm.

Normalized emission spen class="Chemical">ctrum of (a) 2 (18 nm shift) and (b) 4 (12 nm shift) in different solvents. λex = 405 nm.

This compound was highly fluorescent (see Table S3, Supporting Information) in most of the solvents, particularly in DEE where the quantum yield was found to be 71%. However, the solvatochromic effect was not encouraging as expected, but such a system was used to understand the nature of the excited state for this class of molecule.

Next, the effort was made to investigate the reason for this variance in the solvatochromic behavior between such nitro and cyano analogues. The solvent effect on emission is often quite complex and depends on many parameters.[14] On the basis of the literature studies[3] and our experimental observation, we anticipated the formation of internal charge-transfer (ICT) state for such systems. More appropriately, these molecules were presumed to have the rotation of groups on the excited state (twisted molecular structure) and hence the formation of twisted internal charge-transfer (TICT) state was expected. Wherever applicable, the solvatochromic effect primarily is known to originate from TICT excited state, and the change in the dipole moment between the ground and excited state is also a crucial factor.[3c] To find the ground-state geometrical conformation for these molecules, the molecular structures were determined using single-crystal XRD studies (Figure ), and the molecular information was significantly decisive for such systems to find the structural factors. Such studies disclose that the mesitylene ring (87.63°, nearly orthogonal) and cyano groups (44.46°) are significantly twisted from the anthracenyl and phenyl rings, respectively, for the compound 2 while the nitro analogue 1b has a comparatively much favorable situation for electronic conjugations between the donor and acceptor units with the torsion angle ∼76° between mesitylene/anthracene ring and nitro group is twisted out of ring by an angle of 13.13°.

Figure 5

Molecular structure of compound 1b (left) and 2 (right) with important torsion angles (°); H’s are omitted for the clarity.

Molen class="Chemical">cular structure of compound 1b (left) and 2 (right) with important torsion angles (°); H’s are omitted for the clarity.

Thus, molecular planarity for nitro compound 1b is much more favorable than that of compound 2, resulting in a much better electronic conjugations and charge separation for 1b. Therefore, on excitation, nitro compounds will possibly have better charge separation within the molecule in comparison to cyano analogues and such dissimilarity can induce the change in the fluorosolvatochromic behavior. In fact, the less bulky nitro analogue 1a has even much better situation in terms of the co-planarity with a p-tolyl ring (torsion angle with anthracenyl ring ∼67°) (see Supporting Information Figure S4) and the nitro group (twisted only 7.88°). Further, the C=C bond length changes of the styryl linker are found to be interesting to compare ICT into the structures where C=C bond lengths follow the trend 1b (1.323 Å) < 3(1.327 Å) < 2 (1.330 Å) < 4(1.332 Å). This trend also specifies more prominent ICT for −nitro compounds in comparison to cyano. The reason behind this variance in the molecular structures for 1a–b and 2 is likely to be governed primarily by the strong electron-withdrawing effect of nitro group that assist the substituents attaining the proper alignments to adjust the electronic conjugations and other inter/intramolecular interactions. Further, the almost unchanged wavelength for the maximum absorption at different polarity indicated a negligible change in the ground-state conformations. Thus, these molecules perhaps formed a conformationally different excited energy state under different solvent medium and that brought the difference in the emission properties. The strong emission in apolar solvents [ET(30): 31–36] with higher quantum yield for 1a–b can be attributed due to the energetically and geometrically favorable radiative decay from the locally excited (LE) state of the molecule. Upon increasing the polarity of the solvents, the contributions from TICT states predominate and get stabilized. The red-shifted emission with poor quantum yield is expected from such a stabilized and geometrically different TICT state. On the other hand, the cyano compound 2 radiates mostly from the LE state and due to the lack of planarity between the donor and acceptor unit with the anthracenyl core, formation and contribution from ICT state are less significant. The crystal structure reveals a very strong intermolecular C–N···H interaction (2.744 Å) that can twist the cyano groups out of the plane by 44.46°. Such interactions can also be present in some of the solvent media (mostly polar), presumably responsible for lower quantum yield.

The crystal structure of compound 3 (Figure ) was very much helpful to find the possible H-abstraction by the nitro functionality. In fact the intramolecular distance between C–H···O (NO2) was found to be 2.412 Å and it appeared to be very strong bond (mostly covalent).[6b]

Figure 6

Molecular structure of compound 3 (left) and 4 (right) with important torsion angles (deg).

Molen class="Chemical">cular structure of compound 3 (left) and 4 (right) with important torsion angles (deg).

The −NO2 functionality also involves in intermolecular interactions with the hydrogen of the anthracenyl ring with the distance of 2.621 Å. Both these C–H···O (NO2) distances would favor a strong interaction, resulting intramolecular/intermolecular proton abstraction in the excited state,[6b] responsible for the nonfluorescent nature for most of the solvents.

Next, we were attentive to find out the difference in the dipole moment between excited and ground state, being an important parameter to decipher the solvatochromism. On the basis of the reported weak donor- and strong acceptor-linked fluorophore,[8,9] we expected a larger dipole moment for the excited state in comparison to the ground state for these molecules. In fact, significant increments of Stokes shift by changing apolar to polar medium also indicate that the emitting state’s dipole moment (μ) is more than the ground state.[18,19] Further, the differences in the dipole moment between the excited and ground state were calculated using density functional theory (DFT) and time-dependent (TD)-DFT-studies in the gaseous state and that showed a higher difference for 1b (1.67 D) than 2 (1.16 D) (see Table S4 for details). Subsequently, we also found the variations of Stokes shift (Figure S5) and maximum emission wavelength (λmax) (Figure S6) with ET(30) and the plots suggested a very good correlation for 1a–b. The Figure S5 slopes comprise 287 cm–1 (1a) < 301 cm–1 (1b) ≫ 41 cm–1 (2) > 9 cm–1 (4) and Figure S6 slopes include 8 nm (1a) < 9 nm (1b) ≫ 0.5 nm (2), which indicated the favored fluorosolvatochromism for compounds 1a–b in comparison to 2. Moreover, these results also dictate more pronounced TICT effect for 1a–b. Further, the dependency of quantum yields with ET(30) (Figure S7) indicated the lowering of the Fl. intensity being prominent for 1b. The compound 4 was considerably emissive in all of the solvents (Figure S7d). Thus, the fluorophores 1a–b with strong dipoles (becasue of the presence of nitro functionality) exhibited the excellent potential to become sensitive toward the solvents with ET(30) ranging from 31 to 37.4.

We have further examined the Fl. lifetime for these compounds in several solvents of diverse polarity (Table , Figure S8). Compounds 1a–b and 2 revealed a single exponential decay in hexane, heptane, and DEE where compound 2 has a comparatively higher Fl. decay time. These results also confirm that the emission originates from a single excited species. In toluene, a biexponential decay with components ∼80% (B1′) and ∼20% (B2′) was found for compounds 1a–b, whereas compound 2 showed a single exponential decay. By increasing the solvent polarity, compounds 1a–b showed a triexponential decay in acetone where two minor components (3–9%) contributed significant lifetime to the system (τ 1.16 and 4.34 ns), whereas compound 2 had only biexponential decay with a trace component (1%) of relatively long lifetime (τ 5.58 ns). With the increase in the polarity of the environment, the lifetime was reduced for both 1a–b and 2. It also supports the reason behind the lowering of quantum yield on increasing the solvent polarity. The other analogue 4 without any push–pull system displayed a single exponential decay in all of the solvents, which is very much consistent with our previous discussion on solvatochromism. The longer lifetime (2–3 ns) did not even decrease upon the enhancement of the solvent polarity for compound 4. Thus, for compound 4, only one excited state (expected to be LE) with a single excited species is presumably functional to emit the light while compounds 1a–b could emit from the LE state in apolar environment while TICT states would be gradually opened up on raising the polarity of the medium.

Table 2

Lifetime (ns) Fl. Parameters for All of the Compounds in Selected Solvents with Different Polaritya

comp.	hexane	heptane	DEE	toluene	acetonitrile	acetone
1a	2.25 (1)	2.25 (1)	2.36 (1)	0.47 (0.21), 2.46 (0.79)	negligible Fl.	0.2 (0.94), 1.16 (0.04), 4.34 (0.02)
1b	2.03 (1)	2.03 (1)	2.24 (1)	0.5 (0.18), 2.39 (0.82)	negligible Fl.	0.3 (0.88), 1.59 (0.09), 4.42 (0.03)
2	3.01 (1)	2.96 (1)	2.31 (1)	2.61 (1)	0.38 (0.99), 5.58 (0.01)	0.68 (0.99), 3.7 (0.01)
4	2.84 (1)	2.82 (1)	3.2 (1)	2.93 (1)	3.32 (1)	3.13 (1)

The relative weight component is given in the bracket. λex = 405 nm.

We believe that the LE and TICT states are very close in energy and the differences are roughly estimated to be ∼0.343 eV for 1a–b from the Fl. λmax at 510 nm in hexane and at 594 nm in acetone. Such a difference between LE and TICT state is found to be very small (0.08 eV) for 2 and (0.06 eV) for 4. Finally, on the basis of these experimental outcomes, we speculate that the Fl. emission of 1a–b occurs from the LE state for the apolar solvents, and on increasing the polarity of the solvent, the TICT state activates gradually to produce a red-shifted emission with poor quantum yield. The broad emission spectra also suggest the possible existence of both LE and TICT states together; however, no such prominent dual emission was noticed. In such cases, both the emissions are probably merged, leading to the broad emission. In this context, we have attempted to find the contribution of LE and CT states for the fluorophores in various solvents by considering the emission λmax = 476 nm for non-ICT compound 4, where Fl. originates from the state (LE) responsible for normal Fl.[20] A detailed scrutiny on the absorption spectra of compound 1b (see Figure S1′ in the Supporting Information) revealed a noticeable variation with solvent polarity starting with extremely apolar pentane to polar water. The absorption spectrum in pentane shows a single peak at ∼400 nm, which might be corresponding to the ground to LE transition. The absorption spectra in polar solvents appear to contain signatures of more than one transition. The appearance of a hump ∼440 nm in addition to the main absorption band at ∼400 nm, as the solvent polarity is gradually increased is probably indicative of a ground to CT transition. The steady-state emission spectra (Figure b) give an indication of dual emission in polar solvents. A simple deconstruction of the emission spectra can be achieved by a double Gaussian function of the form

The two Gaussians are centered at x1 nm and x2 nm representing, respectively, the LE and CT emission peaks, the values of which are obtained from the position of the main emission band in purely apolar (heptane) and polar (acetone). During fitting, the position of the LE peak is kept fixed at 474 nm [explanation with 4], whereas the position for the CT peak is kept floating. The position and width of the LE peak is ascertained from the emission band of compound 4 in heptane, where there is no ICT state. Following our expectations, no solvatochromism is observed in the emission spectra of compound 4 (Figure S2d). The fitting results (Table S8) show red shift of the CT emission peak as the solvent polarity is increased. In solvent with ET(30) 31.1, the contribution of the LE state to the emission spectrum is ∼50%, whereas in solvent with ET(30) > 34 emission is predominantly obtained from the CT state, with almost negligible LE contribution. The maximum Fl. shift induced by increasing solvent polarity observed in the compound is 84 nm. Assuming that the width of the LE state Gaussian component is not varying with solvent polarity, it is kept fixed while that of the CT state is floated during the fit. Table S8 gives a summary of the fitting parameters and Figure a shows the variation in the weightage of the LE state in the emission spectrum as the solvent polarity is varied progressively. Figure b shows the corrected emission spectra of the compound 2 in pentane and THF along with the fitted lines and that almost overlap the experimentally obtained spectra, validating our fittings to a larger extent.

Figure 7

(a) Contribution of the LE state with the change of ET(30). (b) Corrected emission spectrum with the fitted line.

(a) Contribution of the LE state with the change of ET(30). (b) Corren class="Chemical">cted emission spectrum with the fitted line.

Emission Studies in Heptane–Acetone and Heptane–Methanol Environments

We were interested to find the effect of Fl. intensity and the λmax change in a mixture of polar solvent (acetone) and apolar solvent (heptane). Upon incremental addition (see Figure a) of acetone to a heptane solution of 1b, a substantial bathochromic shift of 55 nm (till fraction of acetone fa = 25%) was observed in the emission spectra with a diminution in Fl. intensity. This observation can be attributed due to the shift in the excited state from higher energy LE to lower energy TICT state upon gradual addition of acetone due to the polarity orientation change of molecules in these two different solvents with diverse polarity. Although the above experiment was repeated with more polar protic solvent methanol (Figure b), a mild bathochromic shift of 7 nm was observed with drastic Fl. quenching even with 1% methanol (fm v/v %). In fact, 25% methanol completely quenched the Fl. intensity of the probe. As there is no such shift in the λmax, the strong supramolecular interaction between O-atom of nitro and −OH functionality has perhaps quenched the Fl. intensity. Thus, the presence of such polar solvents can be easily quantitatively estimated using this probe.

Figure 8

(a) Emission spectra of 1b (10–5 M in heptane) with increase in acetone fractions with shift from 505 to 560 nm. (b) Emission spectra of 1b (10–5 M) with increase in MeOH fractions with shift from 506 to 513 nm. λex = 405 nm.

(a) Emission spectra of 1b (10–5 M in n class="Chemical">heptane) with increase in acetone fractions with shift from 505 to 560 nm. (b) Emission spectra of 1b (10–5 M) with increase in MeOH fractions with shift from 506 to 513 nm. λex = 405 nm.

AEE-Studies

Inspired by the recent development of new small anthracenyl π-conjugate-based AIEgens,[21] AIE behaviors for these molecules are tested. Earlier, we found that the presence of heterocycle played a role to make such molecules AEE-active.[12a] However, it is not essential on well-explored distyrylanthracenyl-based AIEgens because of highly twisted molecular orientation in the aggregated state.[21] Hence, being sterically hindered, compounds 1b and 2 were specifically checked if they exhibit such AEE-behavior. Thus, to confirm a compound whether AEE active or not, Fl. studies were carried out by gradual addition of water (a bad solvent in terms of solubility) to acetonitrile solution of compounds 1b (Figure S9b) and 2 (Figure ). The selection of acetonitrile can be accounted for its water-miscible properties and relatively less Fl. intensity under this environment. Upon gradual addition of water, the molecules started forming aggregates in the solution and the corresponding absorption and emission spectra were systematically studied, where the Fl. intensity was boosted to only compound 2 as shown in Figure . The Fl. intensity started increasing when a fraction of water (fw, v/v %) is higher than 60% and reached to maximum if fw = 80%. Like many other cases, the attenuation of Fl. intensity occurs upon further increments of fw.[22] In the absorption spectra, the molecular form (00% water) of compound 2 exhibited an absorption maximum at 398 nm (Figure S9). We kept increasing the water content and measured both the absorption (Figure S9) and emission spectra (Figure ).

Figure 9

(a) Emission spectrum of 2 (10–5 M) with different fw in MeCN (b) plot showing Fl. intensity at different fw (v/v %).

(a) Emission spectrum of 2 (10–5 M) with different fw in n class="Chemical">MeCN (b) plot showing Fl. intensity at different fw (v/v %).

The aggregated form (80, 90, and 99%) exhibited the absorption maximum at 404 nm, with a mild red shift of 6 nm (Figure S9). In the emission spectra, the emission maximum of the molecular form was observed at 504 nm. On aggregation, the emission maximum was slightly red-shifted to 508 nm with ∼ninefold increment of Fl. intensity. Further, to support the AEE phenomenon, quantum yields (Φf) were estimated and that was found to be enhanced upon aggregate formation [Φf = 2.8% (fw = 0%); 26.3% (fw = 80%), Table S9]. However, such enhancement in the Fl. intensity was not observed for nitro analogues (1a–b) and compound 4. The nonfluorescent nature of these fluorophores in acetonitrile could partly be due to the relaxed excited state in such polar environment and also free ultrafast intramolecular rotation (in the range of picosecond) that quench the Fl. intensity.[21a] To find the polarity effect behind this Fl. intensity enhancement, we replace water with methanol, where we found the lowering of Fl. intensity till methanol fraction (fm) = 50% and almost negligible enhancement of Fl. intensity was noticed by increasing fm = 100% (Figure S10). The main origin for such AEE-behavior was established by various methods in the literature studies where restricted intramolecular motion plays a crucial role.[13,21] Therefore, we have measured the Fl. intensity in a viscous medium by mixing a different fraction of glycerol (fg) with a methanolic solution of compound 2. The Fl. intensity was found to be heightened upon increasing fg (Figure S11), which support the fact of Fl. enhancement for compound 2 due to the restricted intramolecular motion. The average particle size of the aggregates for fw = 80% was measured using dynamic light scattering (DLS, Figure S12) method and that appeared to be ∼155 nm. Thus, this particle size on aggregate formation is very much comparable to the earlier reported AIEgens.[13c]

Next, the difference in the AEE behavior between cyano and nitro analogues were attempted to address by analyzing the molecular structures for both 1b and 2. Before starting with the molecular structure analysis, we have taken the powder XRD data for the aggregate (fw = 80%) which had a few identical intense peaks as obtained from the crystalline sample (Figure S13). Thus, the aggregate may exhibit similar molecular interactions as obtained from the crystalline state. Consequently, the supramolecular interactions (Figure ) on molecular structure 2 are analyzed where the cyano group interacts with the peripheral hydrogen atom of neighboring molecules, specifically anthracenyl hydrogen. This C–H···N hydrogen bond distance 2.744 Å is quite strong and plays a substantial role in forming the supramolecular network that also may cause the twisting structure for this molecule upon aggregation. Even there are significant CH···π (2.709 and 2.871 Å) and CH···H (2.310 Å) interactions observed for this cyano molecule (four types of interactions).

Figure 10

Supramolecular interactions along with significant torsion angles (°) and distances (Å) for 2.

Supramolen class="Chemical">cular interactions along with significant torsion angles (°) and distances (Å) for 2.

Such a molecular network will also restrict the rotation around the vinyl moiety. In particular, the twisted molecular structure is proved to enhance the Fl. intensity by opening the radiative channels. The π···π interactions that bring the molecules to close are also suppressed for this aggregate. The measured dihedral angles also specified the twisted molecule structure. The nitro compound showed comparatively more supramolecular interactions (eleven types of interactions, Figure ) with better planarity. The nitro compound had CH···π (2.796 Å) and CH···H (2.198 Å) interactions along with many other interactions and interestingly, a four-membered (N–O···N–O···; 3.006 Å, apart from N–O···H; 2.61 and 2.712 Å) ring was formed (yellow circle in Figure ), and all these interactions possibly favor the relaxation of molecules to the ground state in a nonradiative pathway. Also, the twisted structure for 2 has higher centroid···centroid distance (8.596 Å, Figure ) in comparison to the nitro analogue 1b (8.033 Å; Figure ), suggesting a relatively disfavored situation for the intermolecular π···π interactions. Thus, apart from polarity induced Fl. quenching, compound 1b has relatively more intermolecular interactions that could also be responsible for being nonfluorescent in the aggregated state.

Figure 11

Supramolecular interactions along with significant torsion angles (°) and distances (Å) for compounds 1b.

Further, Fl. decay dynamics (Figure S14, Table S10) were studied for the compound 2 which relaxed from the excited state via biexponential manner where 99% molecules had lifetime 0.38 ns and 1% molecules are with 5.58 ns, causing weak fluorescent in acetonitrile (average lifetime 0.43 ns). By adding water to this acetonitrile solution (fw = 80%), the molecules began relaxing in two channels with lifetime 0.58 ns (for 24% molecules) and 1.96 ns (for 76% molecules). Thus, the average lifetime of the molecules was almost fourfold increased (1.63 ns) in fw = 80% and support the Fl. enhancement in the aggregated state.

Conclusions

Four anthracenyl π-conjugates linked with a weak donor and strong acceptor are conveniently synthesized in a conventional but effective synthetic route. All these molecules were unambiguously characterized by single-crystal XRD studies which were very much necessary to explain the observed dissimilarities of the molecules in both fluorosolvatochromic and AEE behaviors. The nitro analogues with ∼50% quantum yield in aliphatic alkanes (pentane, hexane, and heptane) were proved to be potential candidates for detecting the apolar solvents within ET(30): 31–37 while cyano analogues could not. It was found that these nitro compounds are highly photostable. The solvatochromic properties remained unchanged even after 6 months for the same set of solution. The difference in solvatochromic behavior between cyano and nitro analogues is explained by examining their geometrical pattern that depicts the extent of conjugation within the molecules. The comparative studies were also conducted by synthesizing anthracenyl π-conjugates attached with 2-nitro (3) or only benzyl (4) groups to understand the electronic effect on these molecules, which also helped us to understand the contribution of LE and ICT states as an origin of Fl. for these molecules. The dissimilarity in exhibiting AEE-behavior between cyano and nitro is demonstrated by comparing the intermolecular interactions in the aggregated state. Thus, such probes would be potential candidates to differentiate the apolar solvents and much smaller anthracenyl conjugates can also serve as AEEgens.

Experimental Section

General Consideration: Reagents

All experiments were carried out in hot air oven-dried glassware under nitrogen and argon atmosphere. Diethyl (anthracen-9-yl(hydroxy)methyl)phosphonate was prepared in our laboratory using the recently published procedure.[12b] All aldehydes were purchased from Aldrich and Alfa-Aesar. KOBu was purchased from Aldrich and used as received. THF was redistilled from sodium metal and benzophenone mixture. All other reagents were purchased from common suppliers and used without further purification. Column and flash chromatography were performed by using silica gel 100–200 and 230–400 mesh, respectively. Reactions were monitored by thin-layer chromatography (TLC) on 60 F254 plates precoated with silica gel (Merck & Co.) and were visualized by UV-light (∼365 nm). All photophysical studies were performed using stock solution (10–3 M in dioxane) of the fluorophore.

Analytical Methods

1H, 13C NMR spectra were recorded in CDCl3 solution using Bruker (400 and 700 MHz). The signals were referenced to tetramethylsilane and solvent used is deuterated chloroform (7.26 ppm in 1H, 77.16 ppm 13C). Chemical shifts are reported in ppm, multiplicities are indicated by s (singlet), d (doublet), t (triplet), and dd (doublet of a doublet). Elemental analyses were carried out on a CHN analyzer. The Fl. spectra were recorded on a spectrofluorimeter. The electronic absorption spectra were recorded with JASCO-650V UV–Vis scanning spectrophotometer. Electrospray ionization (ESI)–liquid chromatography–mass spectrometry (LCMS) was recorded in Shimadzu LCMS-2020. The X-ray quality crystals of the compounds were grown by slow diffusion of n-hexane over CH2Cl2 or n-hexane over ethyl acetate solution. Single-crystal X-ray data were collected on a Rigaku XtaLAB Pro 200 diffractometer using graphite monochromated Mo or Cu radiation. Data were collected and processed using CrysAlisPro (Rigaku Oxford Diffraction). The structures were solved by direct methods and refined by full-matrix least-squares method using standard procedures. Absorption corrections were done using Lorentz and polarization effects, where applicable. In general, all nonhydrogen atoms were refined anisotropically; hydrogen atoms were fixed by geometry or located by a difference Fourier map and refined isotropically. The complete crystallographic data for all of the compounds are tabulated in Table S11. All bond angles, bond or other distances, dihedral angles, and packing diagram are determined using Mercury 3.3 software. Powder XRD data were carried out on Rigaku XRD. Time-resolved measurements were performed by using time-correlated single-photon counting spectrometer (Edinburgh, OB920) with laser diode source (λexc = 405 nm). A dilute LUDOX solution in water was used to measure lamp profile. F900 decay analysis software was used to analysis the decay curves by using nonlinear least-squares iteration method. The quality of the fit was judged by the chi square (χ2) values. Solid-state quantum yields were calculated using SC-30 integrating sphere module on FS5 spectrofluorimeter (Edinburgh Instruments). DLS particle size analysis was carried out using a Zetasizer Nano S from Malvern Instruments at 25 °C. The DFT and TD-DFT studies were performed only for compounds 1b and 2 using CAM-B3LYP/6-311++g(d,p).

Compound 1a: (E)-9-(4-Nitrostyryl)-10-(p-tolyl)anthracene

Diethyl ((10-(p-tolyl)anthracen-9-yl)methyl)phosphonate (0.300 g, 0.716 mmol) was taken in a 25 mL round-bottom flask and was dissolved in dry THF under argon atmosphere. Potassium tert-butoxide (0.241 g, 2.150 mmol) was added to the above solution. The solution was allowed to stir for 2–3 min. After the generation of carbanion, 4-nitrobenzaldehyde (0.101 g, 0.071 mmol) was added to the above solution and was allowed to stir for 2 h. Completion of the reaction was monitored by TLC. The reaction mixture was quenched with water, washed with brine, and extracted with ethyl acetate (20 mL × 2). The resulting organic layer was dried over anhydrous sodium sulphate and concentrated. Compound 1a was purified by column chromatography using ethyl acetate and hexane (3% ethyl acetate in hexane). The product was obtained as reddish orange crystals with yield of 60% (0.178 g); mp 277–280 °C; IR (ν cm–1, in KBr): 2918, 1515, 1338, 969, 765; 1H NMR (400 MHz, CDCl3): δ 8.41–8.30 (m, 4H), 8.21 (d, J = 16.5 Hz, 1H), 7.85 (d, J = 8.7 Hz, 2H), 7.75 (d, J = 8.7 Hz, 2H), 7.54–7.47 (m, 2H), 7.46–7.30 (m, 6H), 7.09 (d, J = 16.5 Hz, 1H), 2.56 (s, 3H). 13C NMR (101 MHz, CDCl3): δ 147.2, 143.6, 138.1, 137.3, 135.6, 135.2, 131.2, 131.1, 130.3, 130.1, 129.3, 129.1, 127.6, 127.1, 125.6, 125.5, 125.1, 124.3, 21.4; ESI-MS: 416 [MH]+; Anal. Calcd for C29H21NO2: C, 83.83; H, 5.09; N, 3.37; O, 7.70. Found: C, 83.76; H, 5.14; N, 3.45; the molecular structure is determined using single crystal diffraction studies (CCDC number 1840758).

Compound 1b: (E)-9-Mesityl-10-(4-nitrostyryl)anthracene

Compound 1b was synthesized in a fashion similar to compound 1a. Diethyl ((10-mesitylanthracen-9-yl)methyl)phosphonate (0.300 g, 0.672 mmol), potassium tert-butoxide (0.226 g, 2.015 mmol), and 4-nitrobenzaldehyde (0.101 g, 0.672 mmol) were used. The product was obtained as reddish orange crystals with yield of 65% (0.193 g); mp 270–274 °C; IR (cm–1, in KBr): 1592, 1517, 1456, 1339, 1113, 967, 858, 752; 1H NMR (400 MHz, CDCl3): δ 8.44–8.33 (m, 4H), 8.23 (d, J = 16.5 Hz, 1H), 8.12–8.06 (m, 1H), 7.85 (d, J = 8.8 Hz, 2H), 7.58–7.48 (m, 4H), 7.41–7.34 (m, 2H), 7.19–7.09 (m, 3H), 2.49 (s, 3H), 1.76 (s, 6H); 13C NMR (101 MHz, CDCl3): δ 147.1, 143.7, 137.5, 137.3, 136.8, 135.2, 134.5, 130.9, 130.5, 130.3, 129.6, 128.3, 127.0, 126.6, 125.8, 125.8, 125.5, 124.3, 21.3, 20.0; ESI-MS: 443 (MH+); Anal Calcd for C31H25NO2: C, 83.95; H, 5.68; N, 3.16; O, 7.21. Found: C, 83.85; H, 5.61; N, 3.09. The molecular structure is determined using single crystal XRD studies (CCDC number 1840755).

Compound 2: (E)-4-(2-(10-Mesitylanthracen-9-yl)vinyl)benzonitrile

Compound 2 was synthesized in a manner similar to compound 1b. Diethyl ((10-mesitylanthracen-9-yl)methyl)phosphonate (0.300 g, 0.672 mmol), potassium tert-butoxide (0.226 g, 2.015 mmol), and 4-formylbenzonitrile (0.088 g, 0.672 mmol) were used. The product was obtained as greenish yellow crystals with yield of 70% (0.199 g); mp 274–278 °C; IR (ν cm–1, in KBr): 3444, 2919, 2223, 1599, 1504, 1437, 1375, 1172, 1024, 971, 905, 846, 814, 776, 754; 1H NMR (400 MHz, CDCl3): δ 8.37 (d, J = 8.8 Hz, 2H), 8.16 (d, J = 16.5 Hz, 1H), 7.83–7.74 (m, 4H), 7.57–7.46 (m, 4H), 7.37 (ddd, J = 8.6, 6.5, 1.1 Hz, 2H), 7.13 (s, 2H), 7.07 (d, J = 16.5 Hz, 1H), 2.49 (s, 3H), 1.75 (s, 6H); 13C NMR (101 MHz, CDCl3): δ 141.73, 137.50, 137.29, 136.62, 135.61, 134.51, 132.70, 131.02, 129.55, 129.35, 128.33, 127.02, 126.58, 125.90, 125.69, 125.50, 119.01, 111.10, 21.26, 20.01; ESI-MS: 424 [MH]+; HR-MS (ESI) m/z: [M + Na]+ calcd for C31H25NNa, 446.1885; found, 446.1887. The molecular structure is determined using single crystal XRD studies (CCDC number 1840754).

Compound 3: (E)-9-Mesityl-10-(2-nitrostyryl)anthracene

Compound 3 was synthesized in a fashion similar to compound 1b. Diethyl ((10-mesitylanthracen-9-yl)methyl)phosphonate (0.300 g, 0.672 mmol), potassium tert-butoxide (0.226 g, 2.015 mmol), and 2-nitrobenzaldehyde (0.101 g, 0.672 mmol) were used. The product was obtained as yellowish orange crystals with yield of 58% (0.172 g); mp 169–173 °C; IR (ν cm–1, in KBr): 1608, 1521, 1438, 1335, 977, 851, 773, 729; 1H NMR (400 MHz, CDCl3): δ 8.53–8.46 (m, 2H), 8.16–8.07 (m, 2H), 8.00 (d, J = 16.3 Hz, 1H), 7.84–7.75 (m, 1H), 7.59–7.49 (m, 6H), 7.38 (ddd, J = 8.5, 6.5, 1.0 Hz, 2H), 7.13 (s, 2H), 2.49 (s, 3H), 1.76 (s, 6H); 13C NMR (101 MHz, CDCl3): δ 148.0, 137.6, 137.2, 136.6, 134.6, 133.4, 133.4, 132.6, 131.0, 130.6, 129.6, 129.5, 128.8, 128.4, 128.3, 126.5, 126.0, 125.7, 125.5, 124.9, 21.3, 20.0; ESI-MS: 443 (MH+); HR-MS (ESI) m/z: [M + H]+ calcd for C31H26NO2, 444.1964; found, 444.1964. The molecular structure is determined using single-crystal XRD studies (CCDC number 1840756).

Compound 4: (E)-9-Mesityl-10-styrylanthracene

Compound 4 was synthesized in an approach similar to compound 1b. Diethyl ((10-mesitylanthracen-9-yl)methyl)phosphonate (0.300 g, 0.672 mmol), potassium tert-butoxide (0.226 g, 2.015 mmol), and 4-nitrobenzaldehyde (0.071 g, 0.672 mmol) were used. The product was obtained as bright green crystals with yield of 74% (0.198 g); mp 218–222 °C; IR (ν cm–1, in KBr): 3452, 3024, 1435, 1375, 984, 850, 748, 694; 1H NMR (700 MHz, CDCl3): δ 8.48–8.43 (d, J = 8.8 Hz, 2H), 8.01 (d, J = 16.6 Hz, 1H), 7.73 (d, J = 7.5 Hz, 2H), 7.53 (d, J = 8.7 Hz, 2H), 7.50–7.44 (m, 4H), 7.38 (t, J = 7.4 Hz, 1H), 7.36–7.32 (m, 2H), 7.12 (s, 2H), 7.05 (d, J = 16.6 Hz, 1H), 2.48 (s, 3H), 1.76 (s, 6H); 13C NMR (176 MHz, CDCl3): δ 137.7, 137.6, 137.5, 137.3, 135.8, 134.9, 132.5, 129.8, 129.7, 129.0, 128.4, 128.1, 126.7, 126.6, 126.5, 125.5, 125.4, 125.3, 21.4, 20.2; ESI-MS: 399 (MH+); HR-MS (ESI) m/z: [M + H]+ calcd for C31H27, 399.2113; found, 399.2096. The molecular structure is determined using single crystal XRD studies (CCDC number 1840757).