Electronic Tuning of Host-Guest Interactions within the Cavities of Fluorophore-Appended Calix[4]arenes

A series of fluorescent calix[4]arene scaffolds bearing electron-rich carbazole moiety conjugated at the lower rim have been prepared. Studies of the fluorescence quenching in the presence of the N-methyl pyridinium guest revealed that the electronic properties of the distal phenolic ring play a major role in the host–guest complexation. In particular, placing an electron-donating piperidine fragment at that ring significantly increased the host–guest interactions, while introducing the same fragment into the proximal phenolic ring weakened the fluorescence response. These results suggest that the dominant interactions between the guest and calixarene cavity involve the oxygen-depleted fluorophore-bearing aromatic ring and not the more electron-rich unsubstituted phenolic fragments.


Introduction
Complexation of cationic organic guests within the electron-rich cavity of calix [4]arene (calixarene) compounds has been at the heart of the host-guest complexation chemistry for several decades [1]. In particular, multiple calixarene scaffolds have been investigated in much detail with regard to the complexation of various pyridinium salts and their derivatives [2]. In addition to common calixarene hosts, these studies involved calixarene scaffolds adapting various conformations [3], and scaffolds containing two calixarene cavities ( Figure 1) [4]. In the great majority of the studies, 1 H NMR spectroscopy was the method of choice to determine the strength of the host-guest complexation [5], with the technique typically requiring relatively high concentrations. Surprisingly, to our knowledge, studies on common electronic effects on this complexation reaction have not been reported. While the introduction of electron-donating or -accepting substituents in the calixarene aromatic rings can be viewed as a judicious, albeit synthetically challenging, route to study these effects, the data analysis can be skewed by the conformational changes of the host molecule with regard to the guest cation. It is generally accepted that π interactions (cation-π and/or π-stacking) play an important role in the overall complexation of N-alkyl pyridinium salts within the calixarene hosts [6]. With nearly all studied calix [4]arene hosts having four alkyl ether groups at the lower rim, the average C4v conic structure of the cavity is not optimized for such interactions [7][8][9][10][11][12][13][14][15][16]. Naturally, π interactions would be maximized if a pair of the opposite aromatic groups adopted a parallel disposition, where two opposite aromatic groups are parallel to each other in a C 2v symmetrical conformation (1d, flattened cone), which for the symmetrically substituted calixarenes can be observed only at low temperatures [16][17][18]. A straightforward way to achieve such an arrangement is the selective 1,3-lower rim dialkylation or acylation of the phenolic oxygens, leading to the protected phenolic moieties adopting a parallel geometry. Alternatively, a replacement of an oxygen atom at the lower rim with a non-polar hydrocarbyl group also results in the oxygen-depleted (formerly) phenolic fragment becoming aligned with the opposite phenolic ring, as can be deduced from the available structural data (1e) [19,20]. Interestingly, although these parallel aromatic rings are pre-arranged to participate in π interactions, nolic fragment becoming aligned with the opposite phenolic ring, as can be deduced from the available structural data (1e) [19,20]. Interestingly, although these parallel aromatic rings are pre-arranged to participate in π interactions, they are less electron rich than the remaining unsubstituted phenolic rings which can adopt a similar arrangement by sacrificing the stabilizing hydrogen bonding between the OH groups. Distinguishing between the two different binding modes could be aided by studying electronic effects of appropriate substituents on the cation complexation. Surprisingly, no such studies have been reported to the best of our knowledge. While studying the chemosensory properties of oxygen-depleted 5,5′-Bicalixarene scaffolds (1c) bearing an alkyne function at the lower rim [21], we discovered that these compounds show strong NMR and fluorescence response upon the complexation of N-methyl pyridinium cation (2) [20,22]. Attachment of electron-donating fluorophores at the termini of the bicalixarene fragment expectedly increased the host-guest complexation properties of the scaffolds [23]. Yet, this observation alone does not provide compelling evidence for the π interactions with the oxygen-depleted part of the calixarene moiety. Here, we present our studies of model calixarene compounds that support the notion of the π interactions between the parallel opposing aromatic rings and N-methyl pyridinium cation playing major role in the host-guest complexation ( Figure 2). While studying the chemosensory properties of oxygen-depleted 5,5 -Bicalixarene scaffolds (1c) bearing an alkyne function at the lower rim [21], we discovered that these compounds show strong NMR and fluorescence response upon the complexation of Nmethyl pyridinium cation (2) [20,22]. Attachment of electron-donating fluorophores at the termini of the bicalixarene fragment expectedly increased the host-guest complexation properties of the scaffolds [23]. Yet, this observation alone does not provide compelling evidence for the π interactions with the oxygen-depleted part of the calixarene moiety. Here, we present our studies of model calixarene compounds that support the notion of the π interactions between the parallel opposing aromatic rings and N-methyl pyridinium cation playing major role in the host-guest complexation ( Figure 2).

Results and Discussion
Although the presence of the electron-donating fluorophores at the termini of the biphenyl chain in 1c increases the fluorescence response to the host-guest interactions with 2, there is no evidence for this chain being involved in the π interactions. Because

Results and Discussion
Although the presence of the electron-donating fluorophores at the termini of the biphenyl chain in 1c increases the fluorescence response to the host-guest interactions with 2, there is no evidence for this chain being involved in the π interactions. Because the adjacent free phenolic rings are more electron rich, they potentially can provide stronger π stabilization to the cationic aromatic guest. As stated above, such strong stabilization would come at the cost of breaking hydrogen bonding between the phenolic groups at the lower rim. To establish the pair of the opposing aromatic rings being responsible for the π interactions with 2, we decided to directly compare its complexation within the cavities of the substituted mono calixarene hosts 3 ( Figure 2).
We hypothesized that if the fluorophore-appended ring A is involved in the π interactions, the substituents in ring C should have major effect on the complexation of 2. On the other hand, if the phenolic rings B and D are the main contributors to the π interactions, substitution in ring B will cause some change in the fluorescence response. To verify this hypothesis, we developed synthetic protocols toward unsymmetrically substituted hosts 3a-d (Schemes 1-3). Moreover, although we earlier reported the synthesis of the parent compound 3a (Φ = 0.17) [23], we have now modified the procedure to obtain the desired compound in only three steps (Scheme 1).

Results and Discussion
Although the presence of the electron-donating fluorophores at the termini of the biphenyl chain in 1c increases the fluorescence response to the host-guest interactions with 2, there is no evidence for this chain being involved in the π interactions. Because the adjacent free phenolic rings are more electron rich, they potentially can provide stronger π stabilization to the cationic aromatic guest. As stated above, such strong stabilization would come at the cost of breaking hydrogen bonding between the phenolic groups at the lower rim. To establish the pair of the opposing aromatic rings being responsible for the π interactions with 2, we decided to directly compare its complexation within the cavities of the substituted mono calixarene hosts 3 ( Figure 2).
We hypothesized that if the fluorophore-appended ring A is involved in the π interactions, the substituents in ring C should have major effect on the complexation of 2.
On the other hand, if the phenolic rings B and D are the main contributors to the π interactions, substitution in ring B will cause some change in the fluorescence response. To verify this hypothesis, we developed synthetic protocols toward unsymmetrically substituted hosts 3a-d (Schemes 1-3). Moreover, although we earlier reported the synthesis of the parent compound 3a (Φ = 0.17) [23], we have now modified the procedure to obtain the desired compound in only three steps (Scheme 1). To prepare compounds 3b and 3c, bearing, at the C ring, the electron-donating piperidine group and electron-withdrawing cyano group, respectively, the corresponding bromo-derivative 7 was prepared in three steps [23]. Reacting compound 7 with piperidine under the Buchwald-Hartwig amination conditions afforded the amino derivative 8 which was converted to the triflate 9. The Sonogashira coupling with the carbazole alkyne gave 3b in a 14% overall yield and quantum yield of 30% (Φ = 0.30) (Scheme 2A) [24]. For 3c (Φ = 0.14), compound 7 was converted to the cyano derivative 10 via the Rosenmund-von Braun reaction with CuCN, followed by the similar protocols for the installation of the carbazole group at the lower rim (Scheme 2B). To prepare compound 3d, the selective protection of the phenolic groups on rings A, D, and C was performed  With these fluorescent calixarenes in hand, we moved to explore their complexation properties toward 2. As expected, addition of 2 to a 10 µM solution of a calixarene in 1,2-dichloroethane (DCE) resulted in the fluorescence decrease ( Figure 4). Titration of the solutions of 3 with 1-10 equiv. of 2 allowed measurements of binding constants (Table 1), which were in the same range reported for calix [4]arene receptors from the UV measurements in chloroform at similar concentrations [10]. The overall numbers (~4000-6000 M −1 ) are higher than Kass obtained by the 1 H NMR technique (Kass = 162 ± 13 M −1 for 3c) at To prepare compounds 3b and 3c, bearing, at the C ring, the electron-donating piperidine group and electron-withdrawing cyano group, respectively, the corresponding bromoderivative 7 was prepared in three steps [23]. Reacting compound 7 with piperidine under the Buchwald-Hartwig amination conditions afforded the amino derivative 8 which was converted to the triflate 9. The Sonogashira coupling with the carbazole alkyne gave 3b in a 14% overall yield and quantum yield of 30% (Φ = 0.30) (Scheme 2A) [24].
For 3c (Φ = 0.14), compound 7 was converted to the cyano derivative 10 via the Rosenmund-von Braun reaction with CuCN, followed by the similar protocols for the installation of the carbazole group at the lower rim (Scheme 2B). To prepare compound 3d, the selective protection of the phenolic groups on rings A, D, and C was performed followed by the bromination at the para-to the OH position of the remaining unprotected ring B (compound 12) [25,26]. The removal of the benzylic groups and selective triflation of the intermediate 13 produced the triflate 14 which was reacted with 4. Finally, the obtained compound 15 was converted to 3d (Φ = 0.39) via Buchwald-Hartwig amination with piperidine (Scheme 3) [27]. All new compounds were fully characterized by the multinuclear NMR spectroscopy and HRMS. All compounds 3a-d exhibit strong fluorescence upon irradiation with the UV light ( Figure 3).
With these fluorescent calixarenes in hand, we moved to explore their complexation properties toward 2. As expected, addition of 2 to a 10 µM solution of a calixarene in 1,2-dichloroethane (DCE) resulted in the fluorescence decrease ( Figure 4). Titration of the solutions of 3 with 1-10 equiv. of 2 allowed measurements of binding constants (Table 1), which were in the same range reported for calix [4]arene receptors from the UV measurements in chloroform at similar concentrations [10]. The overall numbers (~4000-6000 M −1 ) are higher than K ass obtained by the 1 H NMR technique (K ass = 162 ± 13 M −1 for 3c) at higher concentrations. The latter compares well with the literature data for NMP cation complexation obtained by the 1 H NMR technique for calixarene hosts with a single cavity [9,10]. Importantly, the most significant drop in the fluorescence intensity was observed for compound 3b, bearing an electron-rich piperidine moiety at the ring C opposite to the fluorophore unit. Calixarene 3a, unsubstituted at the upper rim, showed a weaker response while 3c, possessing an electron withdrawing cyano group, was the least responsive among these three compounds. Interestingly, calixarene 3d showed the weakest response to the presence of the cation 2 despite having an electron-donating substituent at the upper rim ( Figure 5, Table 1). These results suggest that the complexation of 2 within the calixarene cavity is directed by the π interactions with the aromatic rings A and C. With the electron-donating piperidine at ring C, the complexation is enhanced, while with the electron-withdrawing CN at ring C, the complexation is weakened compared with the parent 3a. On the other hand, an electron-donating piperidine unit at ring B weakens cation complexation presumably due to repulsive steric interactions between the piperidine and 2 (Supplementary Materials, Figure S3). Thus, the preset parallel alignment of rings A and C appears more important in the cation complexation within the calixarene cavity over higher electron density in rings B and D, which are prevented from maximizing their π interactions due to hydrogen bonding at the lower rim.  With these fluorescent calixarenes in hand, we moved to explore their complexation properties toward 2. As expected, addition of 2 to a 10 µM solution of a calixarene in 1,2-dichloroethane (DCE) resulted in the fluorescence decrease ( Figure 4). Titration of the solutions of 3 with 1-10 equiv. of 2 allowed measurements of binding constants (Table 1), which were in the same range reported for calix [4]arene receptors from the UV measurements in chloroform at similar concentrations [10]. The overall numbers (~4000-6000 M −1 ) are higher than Kass obtained by the 1 H NMR technique (Kass = 162 ± 13 M −1 for 3c) at     The fluorescence analysis was further corroborated by the 1 H NMR studies of complexation of 2 by 3a-d. Unlike the fluorescence quenching which would likely depend on the guest orientation within the cavity, the chemical shifts of the host's protons should only reflect the strength of the host-guest interactions. At 5 mM concentrations, the 1:1 mixtures of 2 with 3a or 3b showed significant upfield shift for the N-CH 3 group and aromatic protons ( Figure 6, Table 2). On the other hand, the same signals were only slightly shifted in the case of 3c and 3d, testifying to weaker host-guest interactions. Higher sensitivity of the aromatic protons in 2 suggests that the aromatic ring is likely partly immersed into the calixarene cavity with ensuing π-π interactions [10].
only reflect the strength of the host-guest interactions. At 5 mM concentrations, the 1:1 mixtures of 2 with 3a or 3b showed significant upfield shift for the N-CH3 group and aromatic protons ( Figure 6, Table 2). On the other hand, the same signals were only slightly shifted in the case of 3c and 3d, testifying to weaker host-guest interactions. Higher sensitivity of the aromatic protons in 2 suggests that the aromatic ring is likely partly immersed into the calixarene cavity with ensuing π-π interactions [10].

Materials and Methods
The synthetic manipulations involving air-sensitive compounds were performed in a nitrogen-filled Innovative Technology or Vigor glove box. All solvents were degassed and stored under high-purity nitrogen and activated 4Å molecular sieves. All deuterated solvents were stored under high-purity nitrogen on 3Å molecular sieves. Commercially available reagents (Aldrich, Strem, and Fluka) were used as received. The NMR spectra were recorded on a Bruker Avance 400 MHz spectrometer. 1 H and 13 C NMR signals are reported in ppm downfield from TMS. All measurements were performed at 22 °C in CDCl3/CD2Cl2 unless stated otherwise. Mass spectra were recorded on a VG-Autospec M-250 instrument. UV and fluorescence spectra were recorded on a Vernier fluorescence/UV-Vis spectrophotometer and Hitachi F-2710 fluorescence spectrophotometer.
Synthesis of 5: A sample of 4.24 g (10.0 mmol) of calix [4]arene 4 and 0.64 g (11.8 mmol) of NaOCH3 was refluxed in 300 mL of CH3CN for 30 min to monodeprotonate the

Materials and Methods
The synthetic manipulations involving air-sensitive compounds were performed in a nitrogen-filled Innovative Technology or Vigor glove box. All solvents were degassed and stored under high-purity nitrogen and activated 4Å molecular sieves. All deuterated solvents were stored under high-purity nitrogen on 3Å molecular sieves. Commercially available reagents (Aldrich, Strem, and Fluka) were used as received. The NMR spectra were recorded on a Bruker Avance 400 MHz spectrometer. 1 H and 13 C NMR signals are reported in ppm downfield from TMS. All measurements were performed at 22 • C in CDCl 3 /CD 2 Cl 2 unless stated otherwise. Mass spectra were recorded on a VG-Autospec M-250 instrument. UV and fluorescence spectra were recorded on a Vernier fluorescence/UV-Vis spectrophotometer and Hitachi F-2710 fluorescence spectrophotometer.
for the planar cationic guest undergoing π interactions with only one pair of the calixarene aromatic rings which is not involved in the hydrogen bonding at the lower rim. Although more electron rich, this hydrogen bonding between the unsubstituted phenolic rings is likely too strong to make them available for the π interactions.
Supplementary Materials: Synthesis and characterization of all new compounds, UV-vis, fluorescence and NMR spectra, fluorescence and NMR complexation studies. This information can be downloaded at: https://www.mdpi.com/article/10.3390/molecules27175689/s1. Reference [28] has been cited the Supplementary Materials.