Design of Near-UV Photoluminescent Liquid-Crystalline Dimers: Roles of Fluorinated Aromatic Ring Position and Flexible Linker

Abstract

Near-ultraviolet photoluminescence liquid-crystalline molecules (PLLCs) have attracted attention for temperature-responsive photoluminescence (PL) modulation and ON/OFF sensing under external stimuli. We recently developed mesogenic dimers composed of two hexyloxy-substituted, fluorinated tolane-type cores linked by alkylene-1,n-dioxy chains that exhibited near-UV PL in the solid state. However, the formation of LC phases and the temperature range of the LC state were limited. To improve LC phase stability, in this study, we extended the flexible terminal chains and repositioned the fluorinated aromatic rings from the outer to the inner core positions. Accordingly, we synthesized mesogenic dimers with even-numbered alkylene-1,n-dioxy linkers (hexylene, octylene, and decylene) and outer- or inner-ring fluorination. Outer-ring fluorination led to high melting temperatures and stable crystalline phases with limited mesophase formation. In contrast, inner-ring fluorination induced nematic phases upon heating and cooling owing to zig-zag molecular structures that disrupted crystallinity. Photophysical studies confirmed near-UV PL in solution and solid states; however, the quantum yield of the solution PL was low (<0.01). In the solid state, the PL efficiencies and wavelengths were influenced by the fluorinated aromatic ring position and linker length. This study provides important molecular design criteria for developing stable LC materials with tunable near-UV luminescence for temperature-responsive optical devices.

1. Introduction

Materials that emit near-ultraviolet light [1,2] have attracted significant attention in the medical field as UV-light sources for sterilization and disinfection and for use in fluorescent biological probes [3]. In addition, these materials are expected to find applications in high-density recording media and near-UV light-emitting diodes [4,5]. In addition, sensing materials that exploit changes in the photoluminescence (PL) intensity or wavelength are of great interest for various applications. However, most of the sensing materials reported to date are inorganic or inorganic/polymer hybrid materials [6], and reports on organic near-UV PL molecules have been scarce [7]. To realize the aforementioned functions, developing luminescent materials with liquid-crystalline (LC) properties is considered important LC materials undergo reversible changes in their aggregated structures in response to external stimuli such as temperature and electric field. This feature enables the design of devices with functions unique to LC molecules, including temperature-responsive luminescence modulation (e.g., color or intensity modulation) and dynamic control over luminescence ON/OFF and luminescence direction under external fields. These functions are promising for developing temperature- and field-responsive luminescent devices and sensors, as well as advanced smart optical materials. In particular, such materials could enable new functionalities in wavelength regions that are difficult to access using visible light.

Tolane (diphenylacetylene), consisting of two aromatic rings connected by a carbon–carbon triple bond, is a representative LC mesogen with a simple, rigid, and rod-like π-conjugated structure. Owing to its linear and rigid geometry, it has been widely employed in the design of numerous LC compounds [8,9,10]. In contrast, its application as a luminescent material has been limited, since tolane is generally non-emissive in both solution and solid states. This non-emissive behavior arises from rapid internal conversion from the excited state to non-radiative decay pathways [11,12,13]. Recently, increasing attention has been directed toward the development of photophysical functional materials based on tolane skeleton with compact π-conjugated structures. Reported examples include: (i) tethered tolanes that efficiently emit phosphorescence [14,15]; (ii) solid-state blue-fluorophores generated through crystallization-induced emission enhancement [16] or an aggregation-induced emission [17]; and (iii) solid-state yellow-to-orange PL utilizing an intermolecular C–H···F interactions [18,19,20].

Recently, our group developed mesogenic dimers consisting of two hexyloxy-substituted, fluorinated tolane-based mesogens linked by an alkylene-1,n-dioxy linker, in which the mesogenic unit exhibited near-UV PL in the solid state [7]. In specific, we designed and synthesized a series of compounds denoted as 6-O-n (with n representing the number of carbons in the linker; n = 5–10), in which sterically and electronically distinctive fluorine atoms were introduced on the terminal aromatic rings (outer aromatic rings). After intensive investigations on the phase-transition behaviors and photophysical properties of the materials (Figure 1a), compounds that provide a PL quantum yield (Φ_PL) of up to 0.19 in the solid state were identified. However, only a limited number of derivatives exhibited LC phases (n = 8–10), and the temperature ranges of the LC phases were narrow. These shortcomings indicate the need for further studies to stabilize the LC phases.

In general, two main approaches are used to stabilize LC phases using rod-shaped compounds: (1) extending the flexible chains to increase the entropy and thus stabilize the LC phase while promoting aggregation and controlling molecular arrangement [21,22], and (2) introducing fluorine, which is bulkier than hydrogen and highly electronegative, into the aromatic rings on the inner part of the mesogenic dimer’s linking chain to enhance molecular linearity.

In this study, to further preserve the linearity of the mesogenic dimers, even number of carbon atoms were used in the linking chain. Further, following the aforementioned design criteria, two types of compounds were prepared: 10-O-n with fluorine atoms on the outer aromatic rings (Figure 1b), and 10-I-n with fluorine atoms on the inner aromatic rings (Figure 1c). The synthesis, phase-transition behavior, and photophysical properties of 10-O-n and 10-I-n were evaluated in detail.

2. Materials and Methods

2.1. General

¹H, ¹³C, and ¹⁹F NMR spectroscopic studies were performed with an AVANCE III 400 NMR spectrometer (Bruker, Billerica, MA, USA) using CDCl₃ as the solvent (¹H: 400 MHz; ¹³C: 100 MHz; and ¹⁹F: 376 MHz). Chemical shifts are reported in parts per million (ppm) based on the residual protons of the NMR solvent for ¹H, carbon signal of the NMR solvent for ¹³C, and the signal of C₆F₆ used as the internal standard (δ_F = −163 ppm) for ¹⁹F. Infrared (IR) spectroscopy was performed with an FT/IR-4100 type A spectrometer (JASCO, Tokyo, Japan) using the KBr method. High-resolution mass spectroscopy (HRMS) was performed with a JMS-700MS spectrometer (JEOL, Tokyo, Japan) using the fast-atom bombardment (FAB) method.

All reactions were performed under an argon atmosphere in dried glassware using magnetic stir bars. Thin-layer chromatography (TLC) was conducted using silica gel TLC plates (60F₂₅₄, Merck, Darmstadt, Germany).

The target compounds bearing fluorine atoms on the outer aromatic rings, viz., 10-O-n (n = 6, 8, and 10) were synthesized via a cross-coupling reactions (Scheme 1).

Specifically, 4-[2-(4-decyloxy-2,3,5,6-tetrafluorophenyl)ethyn-1-yl]phenol (1), which was readily obtained in two steps from 1-bromo-4-decyloxy-2,3,5,6-tetrafluorobenzene, was reacted with the corresponding 1,n-dibromoalkane. Similarly, the target compounds, bearing fluorine atoms on the inner aromatic rings, viz., 10-I-n (n = 6, 8, and 10) were prepared from readily available 4-[2-(4-decyloxyphenyl)ethyn-1-yl]-2,3,5,6-tetrafluorophenol (2) and the corresponding 1,n-dibromoalkane using similar reaction protocols. The detailed synthetic procedures for precursors 1 and 2, along with the ¹H, ¹³C, and ¹⁹F NMR spectra of all new compounds are provided in the Supplementary Material (Figures S1–S17).

2.2. Typical Synthetic Procedure of Outer Aromatic Ring-Fluorinated Mesogen Dimer 10-O-6 with Decyloxy Chains at Both Ends

In a shield tube equipped with a Teflon^®-coated stir bar, 4-[2-(4-decyloxy-2,3,5,6-tetrafluorophenyl)ethyn-1-yl]phenol (1, 0.422 g, 1.0 mmol), 1,6-dibromohexane (0.100 g, 0.42 mmol), potassium carbonate (0.350 g, 2.5 mmol), potassium iodide (0.173 g, 1.0 mmol), and 18-crown-6 (0.058 g, 0.2 mmol) were added to dimethyl formamide (DMF; 3 mL). The resulting reaction mixture was stirred at 80 °C for 36 h. Thereafter, the mixture was poured into an aqueous NH₄Cl solution, and the product was extracted three times with Et₂O. The combined organic layers were washed with brine, dried over anhydrous Na₂SO₄, and filtered. The filtrate was concentrated under reduced pressure, and the resulting residue was purified by recrystallization from hot acetone and then a MeOH/CH₂Cl₂ (1:1) mixture to obtain the desired product, 10-O-6, as a white solid in 16% yield (0.060 g, 0.067 mmol).

2.2.1. 1,6-Bis(4-[2-(2,3,5,6-tetrafluoro-4-decyloxyphenyl)ethyn-1-yl]phenoxy)hexane (10-O-6)

Yield: 16%, white solid; M.p.: 147 °C; ¹H NMR (CDCl₃): δ 0.88 (t, J = 6.8 Hz, 6H), 1.23–1.40 (m, 26H), 1.46 (quin, J = 7.2 Hz, 4H), 1.72–1.89 (m, 10H), 4.00 (t, J = 6.8 Hz, 4H), 4.25 (t, J = 6.8 Hz, 4H), 6.88 (d, J = 8.8 Hz, 2H), 7.49 (d, J = 8.8 Hz, 4H); ¹³C NMR (CDCl₃): δ 14.1, 22.7, 25.5, 25.8, 29.1, 29.2, 29.3, 29.5 (for two carbons), 29.9, 31.9, 67.9, 72.9 (t, J = 3.7 Hz), 75.5 (t, J = 2.9 Hz), 98.4 (t, J = 17.6 Hz), 100.4 (t, J = 3.0 Hz), 113.9, 114.6, 133.4, 137.8 (t, J = 12.5 Hz), 139.6–142.5 (dm, J = 246.5 Hz), 145.6–148.6 (dm, J = 250.9 Hz), 159.9; ¹⁹F NMR(CDCl₃, C₆F₆): δ −139.72 to −139.83 (m, 4F), −158.82 to −158.94 (m, 4F); IR (KBr): ν 2924, 2854, 1606, 1521, 1491, 1252, 1175, 1031 cm⁻¹; HRMS (FAB) Calcd for (M + H) C₅₄H₆₃F₄O₄: 927.4599, Found: 927.4590.

2.2.2. 1,8-Bis(4-[2-(2,3,5,6-tetrafluoro-4-decyloxyphenyl)ethyn-1-yl]phenoxy)octane (10-O-8)

Yield: 65%, white solid; M.p.: 121 °C; ¹H NMR (CDCl₃): δ 0.88 (t, J = 6.4 Hz, 6H), 1.24–1.37 (m, 26H), 1.40–1.52 (m, 8H), 1.73–1.85 (m, 10H), 3.98 (t, J = 6.0 Hz, 4H), 4.25 (t, J = 6.4 Hz, 4H), 6.88 (d, J = 8.8 Hz, 4H), 7.49 (d, J = 8.8 Hz, 4H); ¹³C NMR (CDCl₃): δ 14.1, 22.7, 25.5, 25.9, 29.1, 29.2, 29.26, 29.29, 29.5 (for two carbons), 29.9, 31.9, 68.0, 72.9 (t, J = 3.6 Hz), 75.5 (t, J = 2.9 Hz), 98.4 (t, J = 17.6 Hz), 100.5 (t, J = 3.6 Hz), 113.8, 114.6, 133.4, 137.8 (tt, J = 12.5, 2.9 Hz), 139.7–142.5 (dm, J = 247.2 Hz), 145.7–148.6 (dm, J = 250.8 Hz), 159.9; ¹⁹F NMR(CDCl₃, C₆F₆): δ −139.72 to −139.84 (m, 4F), −158.82 to −158.93 (m, 4F); IR (KBr): ν 2918, 2854, 1604, 1517, 1491, 1290, 1248, 1172, 1037 cm⁻¹; HRMS (FAB) Calcd for (M + H) C₅₆H₆₇F₄O₄: 955.4912, Found: 955.4902.

2.2.3. 1,10-Bis(4-[2-(2,3,5,6-tetrafluoro-4-decyloxyphenyl)ethyn-1-yl]phenoxy)decane (10-O-10)

Yield: 5%, white solid; M.p.: 110 °C; ¹H NMR (CDCl₃): δ 0.88 (t, J = 6.8 Hz, 6H), 1.23–1.40 (m, 30H), 1.45 (quin, J = 7.2 Hz, 8H), 1.72–1.83 (m, 10H), 3.98 (t, J = 6.4 Hz, 4H), 4.24 (t, J = 6.8 Hz, 4H), 6.88 (d, J = 8.8 Hz, 4H), 7.49 (d, J = 8.8 Hz, 4H); ¹³C NMR (CDCl₃): δ 14.1, 22.7, 25.5, 26.0, 29.1, 29.2, 29.29, 29.33, 29.4, 29.5 (for two carbons), 29.9, 31.9, 68.1, 72.9 (t, J = 3.6 Hz), 75.5 (t, J = 3.6 Hz), 98.4 (t, J = 16.2 Hz), 100.5 (t, J = 3.6 Hz), 113.8, 114.6, 133.3, 137.8 (tt, J = 13.2, 2.9 Hz), 139.6–142.6 (dm, J = 247.2 Hz), 145.7–148.8 (dm, J = 236.1 Hz), 160.0; ¹⁹F NMR (CDCl₃, C₆F₆): δ −139.74 to −139.89 (m, 4F), −158.82 to −158.95 (m, 4F); IR (KBr): ν 2922, 2853, 1605, 1517, 1491, 1291, 1248, 1021 cm⁻¹; HRMS (FAB) Calcd for (M⁺) C₅₈H₇₀F₄O₄: 982.5146, Found: 982.5151.

2.3. Typical Synthetic Procedure of Inner Aromatic Ring-Fluorinated Mesogen Dimer 10-I-6 with Decyloxy Chains at Both Ends

In a shield tube equipped with a Teflon^®-coated stir bar, 4-[2-(4-decyloxyphenyl)ethyn-1-yl]-2,3,5,6-tetrafluorophenol (2, 0.422 g, 1.0 mmol), 1,6-dibromohexane (0.100 g, 0.42 mmol), potassium carbonate (0.331 g, 2.5 mmol), potassium iodide (0.173 g, 1.0 mmol), and 18-crown-6 (0.058 g, 0.2 mmol) were added to acetonitrile (MeCN; 3 mL), and the resulting mixture was stirred at 80 °C for 36 h (the completion of the reaction was verified using TLC). Thereafter, the organic layer was poured into an aqueous NH₄Cl solution. The crude product was extracted three times with Et₂O and washed once with brine. The organic layer was collected, dried over anhydrous Na₂SO₄, and separated by filtration. The filtrate was evaporated in vacuo and recrystallized from hot acetone and then MeOH/CH₂Cl₂ (1:1) to obtain the desired product, 10-I-6 in 41% yield (0.242 g, 0.242 mmol).

2.3.1. 1,6-Bis(4-[2-(4-decyloxyphenyl)ethyn-1-yl]2,3,5,6-tetrafluorophenoxy)hexane (10-I-6)

Yield: 41%, white solid; M.p.: 98 °C; ¹H NMR (CDCl₃): δ 0.88 (t, J = 6.8 Hz, 6H), 1.22–1.40 (m, 26H), 1.46 (quin, J = 7.2 Hz, 4H), 1.72–1.88 (m, 10H), 3.98 (t, J = 6.4 Hz, 4H), 4.27 (t, J = 6.4 Hz, 4H), 6.88 (d, J = 8.8 Hz, 4H), 7.49 (d, J = 8.8 Hz, 4H); ¹³C NMR (CDCl₃): δ 14.1, 22.7, 25.2, 26.0, 29.1, 29.3, 29.4, 29.6 (for two carbons), 29.8, 31.9, 68.1, 72.8 (t, J = 3.6 Hz), 75.2 (t, J = 2.1 Hz), 98.6 (t, J = 18.3 Hz), 100.6 (t, J = 3.6 Hz), 113.8, 114.6, 133.3, 137.7 (tt, J = 12.5, 2.9 Hz), 139.6–142.5 (ddt, J = 246.4, 13.9, 5.1 Hz), 145.6–148.6 (dm, J = 250.1 Hz), 160.0; ¹⁹F NMR(CDCl₃, C₆F₆): δ −139.62 to −139.73 (m, 4F), −158.86 to −158.98 (m, 4F); IR (KBr): ν 2922, 2852, 1604, 1517, 1492, 1291, 1248, 1173, 1021 cm⁻¹; HRMS (FAB) Calcd for (M+) C₅₄H₆₂F₈O₄: 926.4520, Found: 926.4524.

2.3.2. 1,8-Bis(4-[2-(4-decyloxyphenyl)ethyn-1-yl]2,3,5,6-tetrafluorophenoxy)octane (10-I-8)

Yield: 55%, white solid; M.p.: 100 °C; ¹H NMR (CDCl₃): δ 0.89 (t, J = 6.4 Hz, 6H), 1.23–1.55 (m, 34H), 1.79 (quin, J = 6.8 Hz, 10H), 3.97 (t, J = 6.8 Hz, 4H), 4.25 (t, J = 6.4 Hz, 4H), 6.88 (d, J = 8.8 Hz, 4H), 7.49 (d, J = 8.8 Hz, 4H); ¹³C NMR (CDCl₃): δ 14.1, 22.7, 25.4, 26.0, 29.1, 29.2, 29.3, 29.4, 29.5 (for two carbons), 29.8, 31.9, 68.1, 72.9 (t, J = 3.6 Hz), 75.4 (t, J = 3.0 Hz), 98.5 (t, J = 18.3 Hz), 100.6 (t, J = 2.9 Hz), 113.8, 114.6, 133.3, 137.8 (tt, J = 12.5, 2.9 Hz), 139.7–142.6 (dm, J = 247.2 Hz), 145.6–148.6 (dm, J = 250.9 Hz), 160.0; ¹⁹F NMR(CDCl₃, C₆F₆): δ −139.67 to −139.82 (m, 4F), −158.83 to −158.97 (m, 4F); IR (KBr): ν 2922, 1604, 1492, 1291, 1248, 1142, 1014, 996 cm⁻¹; HRMS (FAB) Calcd for (M⁺) C₅₆H₆₆F₈O₄: 954.4833, Found: 954.4823.

2.3.3. 1,10-Bis(4-[2-(4-decyloxyphenyl)ethyn-1-yl]2,3,5,6-tetrafluorophenoxy)decane (10-I-10)

Yield: 24%, white solid; M.p.: 97 °C; ¹H NMR (CDCl₃): δ 0.88 (t, J = 6.4 Hz, 6H), 1.23–1.40 (m, 30H), 1.40–1.50 (m, 8H), 1.73–1.83 (m, 10H), 3.97 (t, J = 6.8 Hz, 4H), 4.25 (t, J = 6.4 Hz, 4H), 6.88 (d, J = 8.8 Hz, 4H), 7.49 (d, J = 8.8 Hz, 4H); ¹³C NMR (CDCl₃): δ 14.1, 22.7, 25.5, 26.0, 29.1 (for two carbons), 29.3, 29.4 (for two carbons), 29.5 (for two carbons), 29.9, 31.9, 68.1, 72.9 (t, J = 3.6 Hz), 75.4 (t, J = 2.2 Hz), 98.4 (t, J = 18.3 Hz), 100.5 (t, J = 3.0 Hz), 113.8, 114.6, 133.3, 137.8 (tt, J = 11.7, 2.9 Hz), 139.7–142.5 (ddt, J = 247.2, 14.0, 4.4 Hz), 145.7–148.6 (dm, J = 249.8 Hz), 160.0; ¹⁹F NMR(CDCl₃, C₆F₆): δ −139.72 to −139.83 (m, 4F), −158.82 to −158.93 (m, 4F); IR (KBr): ν 2922, 1605, 1491, 1249, 1142, 1012, 980 cm⁻¹; HRMS (FAB) Calcd for (M⁺) C₅₈H₇₀F₈O₄: 982.5146, Found: 982.5151.

2.4. Density Functional Theory (DFT) Calculation

DFT calculations were performed using the Gaussian 16 software package (Rev. B.01; Gaussian, Wallingford, CT, USA) [23]. Geometrical optimization was conducted at the M06-2X/6-31G(d) level of theory [24] using the CPCM implicit solvation model [25] to simulate CH₂Cl₂. Vertical electronic transitions were computed via time-dependent DFT (TD-DFT) at the same level of theory [26].

2.5. Phase-Transition Behavior

The phase-transition behaviors of the specimens were examined by POM using a BX-53 microscope (Olympus, Tokyo, Japan) equipped with a heating–cooling stage (10.002 L, Linkam Scientific Instruments, Redhill, UK). The phase sequences and transition enthalpies were determined by DSC (Shimadzu DSC-60 Plus) at heating and cooling rates of 5.0 °C min⁻¹ under a N₂ atmosphere. Powder X-ray diffraction (PXRD) patterns were acquired using an X-ray diffractometer (Rigaku MiniFlex600, Tokyo, Japan) equipped with an X-ray tube (Cu Kα, λ = 1.54 Å) and semiconductor detector (D/teX Ultra2). The sample powder was mounted on a silicon non-reflecting plate set on a benchtop heating stage (Anton Paar, BTS-500, Graz, Austria). The temperature, heating/cooling rate, and time of X-ray exposure were precisely controlled.

2.6. Photophysical Behavior

UV–visible absorption spectra were recorded using a V-750 absorption spectrometer (JASCO, Tokyo, Japan). PL spectra in the solution and crystalline (Cr) states were acquired using an RF-6000 spectrofluorophotometer (Shimadzu, Kyoto, Japan). The absolute quantum yields of the materials in solution and Cr phases were measured using a Quantaurus-QY C11347-01 absolute PL quantum yield spectrometer (Hamamatsu Photonics, Hamamatsu, Japan). The PL lifetime (τ_PL) was measured using a Quantaurus-Tau C11367-34 fluorescence lifetime spectrometer (Hamamatsu Photonics, Hamamatsu, Japan).

3. Results and Discussion

3.1. Molecular Design

To address the issue of the narrow and limited temperature range for LC phase formation observed in previously developed 6-O-10, a mesogenic dimer consisting of fluorinated tolane-type mesogens with hexyloxy terminal chains connected by a decylene-1,10-dioxy linker, the molecular design was modulated to achieve a more stable LC phase by (1) elongating the terminal flexible chains and (2) fluorinating the inner aromatic rings. To evaluate the design strategy, quantum chemical calculations were performed using the Gaussian 16 program with the M06-2X hybrid functional, 6-31G(d) basis functions, and the conductor-like polarizable continuum solvation model (CPCM; CH₂Cl₂). Figure 2 shows the optimized structures of 6-O-10 developed in our previous study (Figure 2a,b), along with those of the newly designed 10-O-10 with decyloxy terminal chains (Figure 2c,d) and 10-I-10 with fluorinated inner aromatic rings (Figure 2e,f).

The optimized structures of compounds 6-O-10 (from the previous study) and 10-O-10 (from the present study), both of which have fluorine atoms on the outer aromatic rings, adopted an N-shaped conformation. In this structure, the plane defined by the connecting carbon chain is nearly coplanar with the π-conjugated plane of the mesogen, while the two terminal chains are bent out of this plane (Figure 2a–d). The slight difference observed between the two structures, based on DFT calculations, is attributed to the extended terminal flexible chains of 10-O-10, resulting in a marginally greater deviation from planarity (154° for 10-O-10 vs. 153° for 6-O-10). This non-planarity is expected to enhance the stabilization of the LC phase by increasing the entropy of the system. In contrast, the optimized structure of 10-I-10, which features fluorine atoms on the inner aromatic units of the mesogenic dimer, exhibited a different geometry. Owing to the steric and electronic effects of the fluorine atoms, the π-conjugated plane of the mesogen is oriented nearly perpendicular to the alkylene linker (Figure 2e,f). Moreover, the angle between the mesogenic core and the terminal flexible chain is 159°, indicating a significantly more linear structure than that of 10-O-10. This increased linearity is expected to promote intermolecular interactions and contribute to the formation of a more stable LC phase.

According to preliminary DFT calculations, compounds 10-O-n, which possess extended flexible chains, and compounds 10-I-n, which feature fluorinated inner aromatic rings, can exhibit more stable LC phases than the previously studied 6-O-n series (n = 6, 8, and 10). Therefore, this study focuses on a detailed evaluation of the thermo- and photo-physical properties of the 10-O-n and 10-I-n series of compounds.

3.2. Thermophysical Properties

Our initial investigation focused on evaluating the phase-transition behaviors of the synthesized compounds, viz., 10-O-n and 10-I-n (n = 6, 8, and 10). The phase transitions of these materials were characterized through POM and DSC. Figure 3 and Figures S31–S38 present the DSC thermogram of each compound, and Figure 4 shows representative POM micrographs of the mesophases observed for some compounds. The extracted thermophysical data are summarized in

Table 1. Thermophysical properties of the 10-O-n and 10-I-n compounds (n = 6, 8, and 10) evaluated during the second heating [H] and cooling [C] cycles.

Compd.	Phase Sequence and Temperature [°C] a (Phase-Transition Enthalpy [kJ mol−1]) a	Compd.	Phase Sequence and Temperature [°C] a (Phase-Transition Enthalpy [kJ mol−1]) a
10-O-6	[H] Cr 149 (80.2) Iso	10-I-6	[H] Cr1 54 (20.1) Cr2 101 (46.4) N 119 (4.3) Iso
	[C] Cr 147 (−83.4) Iso		[C] Cr1 50 (−19.7) Cr2 100 (−49.6) N 117 (−5.0) Iso
10-O-8	[H] Cr1 103 (10.4) Cr2 123 (64.4) SmA 128 (–) Iso	10-I-8	[H] Cr 104 (44.7) N 112 (4.8) Iso
	[C] Cr1 92 (−8.7) Cr2 121 (−51.3) SmA 126 (−13.9) Iso		[C] Cr 100 (−47.8) N 112 (−5.6) Iso
10-O-10	[H] Cr1 114 (72.1) Iso	10-I-10	[H] Cr 103 (71.2) Iso
	[C] Cr1 106 (–) b Cr2 110 (−71.4) Iso		[C] Cr 97 (−56.8) N 102 (−7.1) Iso

^a Determined by DSC under a nitrogen atmosphere (scan rate: 5.0 or 10 °C min⁻¹). ^b Not determined due to the narrow temperature range between the two phases.

and Tables S1–S8.

In the 10-O-n series, which feature decyloxy-terminated flexible chains and fluorine atoms on the outer aromatic rings, 10-O-6 with a hexylene-1,6-dioxy linker and 10-O-10 with a decylene-1,10-dioxy linker, only exhibited a phase transition between a crystalline phase (Cr), exhibiting a bright-field POM image without fluidity, and an isotropic phase (Iso), exhibiting a dark-field POM image with fluidity, during both heating and cooling cycles; no mesophase formation was detected. In contrast, 10-O-8 with an octylene-1,8-dioxy linker exhibited a mesophase, characterized by a fan-shaped texture in the POM image, between the Cr and Iso phases. These observations suggest that the mesophase has a layered aggregated structure typical of a smectic phase. Additionally, the PXRD pattern recorded at 125 °C during the cooling of the Iso phase of 10-O-8 revealed a strong diffraction peak at 2θ = 2.95° (Figure 5 and Figure S40).

The molecular length of compound 10-O-8 was estimated to be approximately 6.0 nm by DFT calculations. Using this value, the diffraction angles were calculated using Bragg’s equation, 2d sin θ = nλ (where d is the interlayer distance, θ is the diffraction angle, λ is the wavelength of the Cu Kα radiation, and n is an integer), obtaining 2θ = 1.48° (n = 1), 2.95° (n = 2), and 4.42° (n = 3). The diffraction peak observed at 2θ = 2.95° in the PXRD measurement corresponds to the second-order reflection, indicating that the compound forms a smectic A (SmA) phase with the molecules aligned along the normal layer.

In the 10-I-n series, which features fluorine atoms on the inner aromatic rings and flexible decyloxy terminal chains, 10-I-6 and 10-I-8, which have hexylene-1,6-dioxy and octylene-1,8-dioxy linkers, respectively, exhibited enantiotropic LC behavior. In contrast, 10-I-10, with a decylene-1,10-dioxy linker, showed monotropic LC behavior. For all compounds in this series, POM observations revealed a four-brush Schlieren texture, which is the characteristic of nematic (N) phases with an orientational order but no positional order. PXRD measurements on the mesophase resulted only in a broad halo at 2θ = ~20°, indicating short-range order along the molecular short axis (Figure S41). These results strongly suggest that the mesophase observed for the 10-I-n series is an N phase.

Next, we investigated the effects of the position of the fluorinated aromatic rings and the length of the connecting chain in the mesogenic dimers on their melting temperature (T_m; the temperature corresponding to the transition of the Cr phase to other phases) and clearing temperature (T_c; the temperature corresponding to the transition from a certain phase to the Iso phase). Figure 6 presents the T_m and T_c measured during the second heating cycle as functions of the connecting chain length (n) for the 10-O-n and 10-I-n series.

Comparisons revealed that both the T_m and T_c values were consistently higher for the 10-O-n series than for the 10-I-n series, regardless of the linker chain length. This difference can be explained using the stable molecular conformations obtained through DFT calculations for the two series (see Figure 2). In the case of the 10-O-n series with outer aromatic ring fluorination, the two mesogenic units adopt nearly coplanar conformations across the linker, promoting intermolecular stacking, which stabilizes the Cr phase. In contrast, in the case of the 10-I-10 series, steric and electronic repulsions between the fluorine atoms on the inner aromatic rings and the methylene (-CH₂-) groups in the linker prevent coplanar alignment, resulting in a zig-zag molecular conformation. This geometry likely hinders intermolecular stacking, destabilizing the Cr phase. In the 10-O-n series, both the T_m and T_c values decreased significantly with increasing chain length. In the 10-I-n series, the T_c decreased with increasing chain length, but the T_m showed a slight increase. As longer linkers generally reduce the degree of ordered molecular aggregation, leading to lower T_m and T_c values, the trend observed for the 10-O-n series is reasonable. However, the observed increase in T_m with increasing chain length in the 10-I-n series deviates from this trend, suggesting that interactions between the linker and mesogenic conformation significantly affect the thermal behavior of these molecules. Moreover, in the 10-I-n series, compounds 10-I-5 and 10-I-7, which contain odd-numbered pentylene-1,5-dioxy and heptylene-1,7-dioxy linkers, respectively, exhibited significantly lower T_m and T_c values than 10-I-6 and 10-I-8, indicating a pronounced odd-even effect (Figure S39). This is likely due to the tendency of odd-numbered linkers to induce bent molecular conformations, disrupting ordered aggregation of molecules.

These results indicate that the position of the fluorinated aromatic rings in the mesogenic dimers plays a key role in controlling molecular aggregated structures. Introducing the fluorinated aromatic rings into the inner part of the dimer reduces the stability of the Cr phase, facilitating mesophase formation. Additionally, linkers with an even number of carbon atoms form more stable mesophases over a broader temperature range.

3.3. Photophysical Behavior

We next investigated the effect of the position of the fluorinated aromatic rings and the chain length of the linker in the mesogenic dimers on their photophysical properties. The measurements were performed on dichloromethane (CH₂Cl₂) solutions. A 1.0 × 10⁻⁵ mol L⁻¹ solution, prepared from the solid samples purified by recrystallization, was used for UV-visible absorption analysis, and a 1.0 × 10^–6 mol L^–1 solution was used for PL analysis. The absorption and PL spectra are shown in Figure 7 and Figure S42, and the corresponding photophysical data are summarized in

Table 2. Photophysical properties of 10-O-n and 10-I-n series in CH₂Cl₂ solutions.

Compound	λabs [nm] a (ε [103, L mol−1 cm−1])	λPL [nm] b (ΦPL) c	τave [ns]	τ1 [ns]	τ2 [ns]	kr [ns−1] d	knr [ns−1] e	knr/kr
10-O-6	300 (80.6), 316 (74.0)	358 (0.033)	1.28	0.83	2.43	0.026	0.755	29.0
10-O-8	300 (63.9), 316 (58.3)	359 (0.032)	0.77	–	–	0.041	1.26	30.7
10-O-10	299 (64.9), 316 (58.7)	358, 418sh (0.035)	0.85	–	–	0.041	1.14	27.8
10-I-6	300 (24.7), 317 (24.0)	360, 384sh (0.028)	1.06	–	–	0.026	0.917	35.3
10-I-8	300 (35.8), 316 (32.4)	358 (0.034)	0.79	–	–	0.043	1.22	28.3
10-I-10	300 (61.9), 316 (55.7)	357 (0.033)	0.86	–	–	0.038	1.12	29.5

^a Observed at a concentration of 1.0 × 10⁻⁵ mol L⁻¹. ^b Observed at a concentration of 1.0 × 10⁻⁶ mol L⁻¹. ^c Determined using an absolute quantum yield measurement system with an integrating sphere. ^d Radiative deactivation rate constant k_r = Φ_PL/τ_PL. ^e Nonradiative deactivation rate constant k_nr = (1−Φ_PL)/τ_PL. sh: shoulder peak.

.

The CH₂Cl₂ solution of compound 10-O-6, with outer aromatic ring fluorination, showed two absorption bands with maximum absorption wavelengths (λ_abs) at approximately 300 and 316 nm (Figure 7a). Similarly, 10-O-8 and 10-O-10, differing only in the linker chain length, exhibited the same absorption features. Compounds 10-I-6, 10-I-8, and 10-I-10, with inner aromatic ring fluorination, also displayed comparable absorption spectra. These results indicate that neither the position of the fluorinated aromatic rings nor the chain length of the linker significantly affects the photophysical properties of the molecules in the dispersed state, i.e., the electronic structure of the ground state.

TD-DFT calculations were conducted to clarify the electronic transitions responsible for the absorption spectra of the 10-O-n and 10-I-n series. Figure 8 presents the molecular orbital diagrams associated with the first excited-state transitions of 10-O-6 and 10-I-6.

Table 3. Theoretical orbital energy calculation data for 10-O-6, 10-I-6, 10-O-10, and 10-I-10 ¹.

Compound	HOMO−1/ HOMO [eV]	LUMO/ LUMO+1 [eV]	ΔEH-L [eV]	Theoretical Electronic Transition (Population)	Calculated λabs [nm]	Oscillator Strength (f)
10-O-6	−7.19/−7.19	−0.96/−0.96	6.23	HOMO−1 → LUMO (47.3%) HOMO → LUMO+1 (47.5%)	295	3.3737
10-I-6	−7.20/−7.19	−0.97/−0.96	6.23	HOMO−1 → LUMO (34.9%) HOMO−1 → LUMO+1 (10.7%) HOMO → LUMO (11.0%) HOMO → LUMO+1 (38.4%)	295	3.3645
10-O-10	−7.32/−7.32	−0.93/−0.93	6.39	HOMO−1 → LUMO+1 (42.5%) HOMO → LUMO (51.1%)	284	3.2167
10-I-10	−7.32/−7.31	−0.93/−0.93	6.39	HOMO−1→ LUMO+1 (47.0%) HOMO → LUMO (47.1%)	284	3.2227

¹ Computed at the M06-2X/6-31+G(d)//M06-2X/6-31G(d).

summarizes the TD-DFT calculation results for 10-O-6, 10-I-6, 10-O-10, and 10-I-10, including the energies of HOMO−1, HOMO, LUMO, and LUMO+1 levels (HOMO: highest occupied molecular orbital; LUMO: lowest unoccupied molecular orbital), the main electronic transitions and their contributions, theoretical absorption wavelengths, and oscillator strengths.

TD-DFT calculations revealed that, in all compounds, the first excited state is obtained through vertical transition from degenerate HOMO−1 and HOMO to degenerate LUMO and LUMO+1. These orbitals are delocalized over the two π-conjugated mesogens, indicating a π-π transition resulting from the overlap of π orbitals. The energy levels of HOMO (HOMO−1) and LUMO (LUMO+1) are nearly identical for the compounds in the 10-O-n and 10-I-n series with the same linker length, regardless of the position of the fluorinated aromatic rings. This consistency supports the experimental absorption spectra. Quantum chemical calculations also suggested that elongating the linker chain destabilizes both HOMO (HOMO−1) and LUMO (LUMO+1), slightly increasing the HOMO–LUMO gap (ΔE_H-L) from 6.23 eV for the compounds with hexylene-1,6-dioxy linkers to 6.39 eV for those with decylene-1,10-dioxy linkers. This widening of the energy gap corresponds to a slight blue shift in the calculated λ_abs from 295 nm for the compounds with a shorter linker to 284 nm for those with the longer one. However, no clear shift in the absorption wavelength was observed experimentally with varying linker length, likely due to solvent or environmental effects. Similarly, while two distinct absorption bands appeared in the experimental spectra, only a broad single band was predicted computationally, which is also attributed to environmental influences.

We investigated the solution PL behaviors of the two series of mesogenic dimers by irradiating their solutions in CH₂Cl₂ (1.0 × 10^–6 mol L^–1) with 316 nm light (Figure 7b). Regardless of the fluorinated aromatic ring position or the linker length, all solutions exhibited a PL band with a maximum PL wavelength (λ_PL) located at 357–360 nm, indicating emission in the near-UV region. This result suggests that the deactivation of the excited state proceeds from a similar level in all cases. For certain compounds, such as 10-O-10 and 10-I-6, shoulder peaks appeared at ~418 and 384 nm, although the λ_PL remained unchanged. The origin of these shoulder peaks remains unclear. The PL quantum yield (Φ_PL) was below 0.1 for all compounds. PL lifetime (τ) measurements at the emission wavelength of ~360 nm revealed that 10-O-6 exhibits a biexponential decay with an average lifetime (τ_ave) of approximately 1.28 ns (Figure S44). Other compounds presented a single-exponential decay, with τ values ranging from 0.77 to 1.06 ns. As all lifetimes are on the nanosecond scale, the observed PL is attributed to fluorescence from the first excited singlet state (S₁). Using the measured Φ_PL and τ values, the rate constants of radiative deactivation (k_r) and non-radiative deactivation (k_nr) were calculated, revealing that the k_nr was at least 27.8 times greater than the k_r. This dominant non-radiative decay is presumed to result from rapid structural relaxation from the emissive ππ* excited state with a linear geometry to the non-emissive πσ* excited state with a trans bent geometry [11,12,13]. While such structural relaxation can be suppressed by tuning the electron density distribution using donor-π-acceptor (D-π-A) structures or through intermolecular interactions, the π-conjugated mesogens used in this study likely do not provide sufficient electron distribution to effectively prevent the structural relaxation.

Next, we evaluated the solid-state PL properties of the mesogenic dimers in the 10-O-n and 10-I-n series. For this, samples purified by recrystallization were used. Figure 9 and Figure S43 present the solid-state PL spectra of the samples, and

Table 4. Photophysical properties of the 10-O-n and 10-I-n series in the crystalline state.

Compd.	λPL [nm]	ΦPL a	τave [ns]	τ1 [ns]	τ2 [ns]	kr [ns–1] b	knr [ns–1] c	knr/kr
10-O-6	358	0.14	0.75	–	–	0.187	1.15	6.15
10-O-8	358, 377, 396sh	0.28	1.23	–	–	0.228	0.585	2.56
10-O-10	360, 376, 398sh	0.14	1.49	–	–	0.094	0.577	6.14
10-I-6	355, 372sh	0.11	1.25	0.82	2.60	0.088	0.715	8.12
10-I-8	355, 372sh	0.051	0.83	–	–	0.061	1.14	18.7
10-I-10	372	0.091	2.36	1.46	3.80	0.038	0.385	10.1

^a Determined using an absolute quantum yield measurement system equipped with an integrating sphere. ^b Radiative deactivation rate constant k_r = Φ_PL/τ_PL. ^c Non-radiative deactivation rate constant k_nr = (1 − Φ_PL)/τ_PL. sh: shoulder peak.

summarizes the corresponding photophysical data.

Compound 10-O-6, which contains fluorine atoms on the outer aromatic rings, exhibits a single PL band at approximately 358 nm in the solid state, that is, PL in the near-UV region (Figure 9a). In the cases of 10-O-8 and 10-O-10, which have octylene-1,8-dioxy and decylene-1,10-dioxy linkers, respectively, apart from the main band at 358–360 nm, additional PL bands were observed near 376 nm, along with a shoulder peak at ~396 nm. This result indicates more complex near-UV PL behavior. These λ_PL values closely agree with those observed in dilute solutions, suggesting that π-conjugated mesogens remain relatively isolated even in the solid state, resulting in monomer-like emission. However, the Φ_PL in the solid state ranged from 0.14 to 0.28, being at least 4.2 times higher than those in dilute solutions. This enhancement is attributed to the suppression of non-radiative deactivation pathways owing to restricted molecular motions caused by molecular aggregation [7,19,20,27].

Further, compounds 10-I-6 and 10-I-8, with fluorine atoms on the inner aromatic rings, also showed near-UV PL in the solid state, with the λ_PL at ~355 nm and a shoulder peak near 372 nm (Figure 9b). Compounds 10-I-5 with a pentylene-1,5-dioxy linker and 10-I-7 with a heptylene-1,7-dioxy linker also exhibited similar PL behavior to the other 10-I-n series (Figure S43). Compound 10-I-10, with a longer linker, exhibited similar PL features, although with a slight red-shift in the λ_PL to ~372 nm. The compounds in this series also displayed similar PL behavior to that in dilute solutions, indicating monomer-like emission. However, the 10-I-n series showed lower Φ_PL values than the 10-O-n series. This difference is consistent with their phase-transition behavior: the 10-O-n series, with higher crystallinity, effectively suppresses molecular motion in the solid state, enhancing Φ_PL, whereas the 10-I-n series exhibits lower crystallinity, leading to less effective suppression of molecular motion, enhanced non-radiative deactivation, and thus decreased Φ_PL.

Time-resolved measurements revealed that all 10-O-n compounds exhibit single-exponential decay, with τ values of 0.75–1.49 ns, confirming PL even in the solid state. In the 10-I-n series, 10-I-8 showed a single-exponential decay, while the other two compounds exhibited biexponential decays, exhibiting τ_ave of 0.83–2.36 ns, also indicative of fluorescence (Figure S45). The calculated k_r and k_nr values revealed that the 10-O-n series has increased k_r and decreased k_nr compared with those in the dilute solution. In contrast, the Φ_PL of the 10-I-n series remained low, which is due to the increased k_nr relative to that in the dilute solution. These findings suggest that molecular aggregation in the solid state partially restricts the molecular motion, influencing the PL behavior.

The phase transition behavior of thermotropic LCs during heating–cooling cycles can be interpreted as structural changes in molecular aggregation induced by thermal cycling. To examine this phenomenon, temperature-dependent PLQY measurements were conducted on the thermotropic LC compounds, viz., 10-O-8 and 10-I-n (n = 6, 8, 10). As shown in Figures S46–S49, Φ_PL gradually decreased upon heating due to non-radiative deactivation associated with thermal molecular motions, whereas recovery of Φ_PL was observed during cooling as these molecular motions were suppressed. In the case of 10-O-8, the Cr–SmA phase transition induced a change in the PL peak shape, and a slight shift in λ_PL was detected at the SmA–Iso phase-transition (Figure S46). For 10-I-6 and 10-I-8, the Cr–N phase transition resulted in a decrease in Φ_PL accompanied by a redshift in λ_PL, attributable to the transition into the more fluid N phase (Figures S47 and S48). In contrast, for 10-I-10, λ_PL in the N phase was found to be blueshifted relative to that in the Cr phase (Figure S49). This discrepancy may be ascribed to the fact that shorter linkage chains enhance the rigidity of the dimeric mesogens, thereby favoring intermolecular electrostatic interactions. However, conclusive experimental evidence supporting this interpretation has not yet been obtained.

Overall, these results demonstrate that the position of the fluorinated aromatic rings exerts little influence on the PL properties in dilute solutions but plays a significant role in the solid states owing to differences in the crystallinity. In particular, the highly crystalline 10-O-n series exhibited a markedly enhanced Φ_PL. Moreover, the linker chain length affected the aggregated structures, thereby contributing to variations in the PL spectral profile. Importantly, thermal phase transitions further modulated the PL behavior, as structural reorganization during heating–cooling cycles led to reversible changes in both PL intensity and spectral features.

4. Conclusions

We designed π-conjugated mesogenic dimers with extended flexible terminal chains and repositioned the fluorinated aromatic rings from the outer to the inner position to improve LC and near-UV PL properties. Further, to maintain molecular linearity, we employed even-numbered alkylene-1,n-dioxy linkers, including hexylene-1,6-dioxy, octylene-1,8-dioxy, and decylene-1,10-dioxy. Phase-transition characteristics revealed that compounds with fluorine atoms on the outer aromatic rings exhibited high crystallinity and suppressed mesophase formation. In contrast, those with fluorine atoms on the inner rings formed nematic phases, regardless of the linker length. This is likely because of their zig-zag structures, which disrupt ordered crystal packing. In dilute solutions, all compounds displayed near-UV fluorescence with a low PL efficiency (<0.01). In the solid state, the PL efficiency depended on the position of the fluorinated aromatic rings and the chain length of the linker. Compounds with fluorine atoms on the outer aromatic rings exhibited high PL efficiencies (0.14–0.28) owing to stable crystallinity, whereas those with fluorine atoms on the inner aromatic rings showed no improvement, likely because increased molecular motion in less stable crystalline states enhanced non-radiative deactivation. A red-shift in the emission wavelength was also observed for the compounds with longer linkers. These findings underscore the critical role of the fluorination position and linker structure in governing the phase behavior and PL properties of mesogenic dimers, providing valuable molecular design guidelines for the development of efficient near-UV fluorescent LC materials. Moreover, the demonstrated dependence of emission color and efficiency on fluorination pattern and phase transitions indicates that these compounds represent promising prototypes for temperature-responsive luminescent sensing devices.