New SmAP_F Mesogens Designed for Analog Electrooptics Applications

Abstract

We have previously reported the first realization of an orthogonal ferroelectric bent-core SmAP_F phase by directed design in mesogens with a single tricarbosilane-terminated alkoxy tail. Given the potentially useful electrooptic properties of this phase, including analog phase-only electrooptic index modulation with optical latching, we have been exploring its “structure space”, searching for novel SmAP_F mesogens. Here, we report two classes of these—the first designed to optimize the dynamic range of the index modulation in parallel-aligned cells by lowering the bend angle of the rigid core, and the second expanding the structure space of the phase by replacing the tricarbosilane-terminated alkyl tail with a polyfluorinated polyethylene glycol oligomer.

1. Introduction

Potential applications of the lamellar (smectic), tilted, polar liquid crystal phase (SmCP), the lamellar, orthogonal, polar phase SmAP, and the lamellar, orthogonal nonpolar phase (SmA) exhibited by bent-core liquid crystals (LCs) motivate continuing interest in their materials design and synthesis. For example, Jákli et al. have described a bistable “scattering” display mode of the SmC_SP_A phase requiring no polarizers [1,2], O’Callaghan et al. reported fast analog electrooptic phase modulators with a SmC_AP_F mesogen [3], and Takezoe et al. have reported an “ideal display mode” [4,5] obtained with certain bent-core SmA mesogens in what has become known as the SmAP_R phase.

The SmAP_F phase, an achiral, orthogonal phase with spontaneous ferroelectric polarization parallel to the smectic layers, was described in a theoretical exercise in the early 1990s [6], and first realized experimentally in an interesting (but very slow-switching) polymeric LC system composed of calamitic monomer units in 2004 [7]. A low molar mass SmAP_F bent-core mesogen, codenamed W586 (Figure 1), obtained by directed design, was reported in 2011 [8,9], and a method for high-quality alignment of the SmAP_F phase has recently been described [10].

The approach for the design of structures exhibiting the SmAP_F phase involved structural modifications of a previously known antiferroelectric SmAP_A mesogen with a structure similar to W586, but lacking the tricarbosilane [11,12]. Our basic concept for obtaining the ferroelectric SmAP_F phase involved removing the generally strong preference for synclinic layer interfaces in smectics, which leads to antiferroelectric ordering of bent-core mesogens, an outcome thought to be due to out-of-layer fluctuations typically seen with mesogens possessing alkoxy tails [13]. The desired anticlinic layer interfaces were obtained by introducing functionality expected to suppress out-of-layer fluctuations—in the case of W586, the tricarbosilane terminating group (earlier applications of carbosilanes in smectic LCs can be found in [14,15]). This approach to suppression of the out-of-layer fluctuations removes an hypothesized entropic preference for synclinicity, but seemingly does not in itself favor anticlinic interfaces [16]. However, with the W586 structure, incorporation of the tricarbosilane-terminated tail in fact produced the desired anticlinic inter-layer structure and ferroelectric order.

As previously reported [17], the electrooptic behaviour of W586 is interesting. With rubbed polyimide-coated glass plates, the layers orient perpendicular to the substrates, with the ferroelectric polarization P parallel to the plates. Though there is effectively random planar alignment (a smectic focal conic texture is seen by polarized optical microscopy), application of fields normal to the plates induces no brush rotation, but causes a smooth, analog change in birefringence with increasing field until saturation, where the polarization is oriented normal to the plates, directly along the field (Figure 2, Left). This is an example of electrostatic V-shaped switching, as first reported for calamitic de Vries SmC* materials with high polarization, [18,19]. For the SmAP_F phase in uniformly aligned planar cells (see below), V-shaped switching gives uniform, phase-only modulation of incident polarized light, with no rotation of the optic axis in the cell, a potentially useful effect. This electrooptic mode gains added attractiveness in applications, since no power is required to maintain an optical state (i.e., the system exhibits optical latching) (Figure 2, Right) [20].

Potential applications of SmAP_F have led to further investigation. For example, as mentioned above, a method for obtaining high quality bookshelf alignment of W586 in the SmAP_F phase with excellent azimuthal orientational control has recently been reported [10]. Here, we describe work aimed at exploring the structure space of the SmAP_F, both to improve fundamental understanding of the chemical structural features leading to the phase, and with an eye to improving the electrooptic properties exhibited in this novel manifestation of electrostatic V-shaped switching.

2. Design Rationale for the New SmAP_F Mesogens

Prototypes of two new classes of SmAP_F mesogens are reported. In one of these, a 4,4′-thiobisphenol diester unit is incorporated into the mesogen core, while maintaining other key features of the W586 structure, giving W653 (Figure 3, top). The thioether core present in W653 was chosen based upon the expectation that the bend angle for this bent-core material should be smaller than that of W586 (~120°). Thus, in 4-4′-thiobisphenol (bis(4-hydroxyphenyl)sulfide) in the crystalline solid state, the OPh-S-PhO angle is reported to be 104.21° by single crystal X-ray crystallography [21], and the calculated angle in the gas phase (Spartan 14 at the MP2 6-31G* level) is 102°. The smaller bend angle at the center of the core was expected to decrease the “overall average bend angle” of the core, leading to an increase in the modulation depth of the index change on switching relative to W586, other things being equal.

The second new structural class, exemplified by W788 and W789, (Figure 3, bottom) was chosen to explore the efficacy of polyfluorinated polyethyleneglycol (PF-PEG) tails as an alternative to the tricarbosilane-terminated alkoxy tail for suppression of out-of-layer fluctuations, thereby removing the normal preference for synclinicity, and allowing the desired anticlinic layer structure in the SmAP phase. This approach is motivated by a considerable body of research on calamitic ferroelectric liquid crystals (FLCs) driven by researchers at 3M Company in the 1990s, showing that the PF-PEG tails incorporated into calamitic FLC structures produce de Vries smectics [22], a phase also thought to be favored by suppression of out-of-layer fluctuations [13,23].

3. Synthesis of the New Mesogens

Diphenylthioether W653 was readily prepared as indicated in Scheme 1. Synthesis of the diphenylthioether bent-core W653. Monoesterification of 4,4′-thiobisphenol (1) with 4-cyanobenzoyl chloride gave intermediate thioether 2. Esterification of the second phenolic hydroxyl group with carboxylic acid 3 [24] using DCC/DMAP provided the target thioether W653 in good yield.

Incorporation of a PF-PEG tail with commercially available fluorinated starting materials was accomplished using an ester linkage in place of a simple aryl ether, leading to structures W788 and W789, possessing a cyano group, and trifluoromethoxy group, respectively, opposite the PF-PEG tails, as indicated in Scheme 2.

Both fluoro-PEG mesogens W788 and W789 were prepared from the key tetracyclic phenol intermediate (8). Synthesis of tetracyclic 8 relies upon prior work of Finkelmann, et al., who reported the synthesis of the PF-PEG phenol 6. Thus, esterification of benzyl ether protected p-hydroxybenzoic acid (4) with the now commercially available hydroxymethylene-terminated PF-PEG 5, using carbonyldiimidazole (CDI) and 1,8-Diazabicycloundec-7-ene (DBU), followed by debenzylation, gave Finkelmann’s phenol 6. A second esterification with benzoic acid 4, then debenzylation, gave phenol (7). Esterification of 7 with m-benzyloxybenzoic acid, debenzylation, then another esterification with p-benzyloxybenzoic acid (4) followed by debenzylation gave the key intermediate tetracyclic phenol 8. Esterification of phenol 8 using either p-cyanobenzoyl chloride 9, or p-trifluoromethoxybenzoyl chloride 10, gave mesogens W788 and W789, respectively.

4. Properties of Diphenylthioether W653

The phase sequences (obtained by polarized light microscopy) and transition temperatures and enthalpies (obtained by DSC) for the new mesogens are given in Figure 3. Gratifyingly, the diphenylthioether W653 possessed the desired enantiotropic SmAP_F phase between 140 °C and 154 °C. Small-angle X-ray scattering shows a layer spacing of 57.0 Å at 140 °C, expanding to about 57.3 Å at 154 °C.

For SmAP mesogens such as these, possessing only one tail, the layer spacing was more difficult to interpret at a molecular level than for nominal calamitic smectics. A simple molecular mechanics fully extended conformation for W653 was about 49 Å in length, suggesting a “partial bilayer” layer structure. This is also seen for the prototype SmAP_F material W586. Lacking results of high-end computational dynamic simulations (currently in progress for W586), it was not possible to understand the layer structure of W653 in more detail.

As shown in Figure 4A,B, in planar-aligned electrooptic cells (rubbed polyimide on ITO-coated glass plates), the observed focal conic texture of W653 exhibits the SmAP_F signature change in birefringence, without brush rotation, upon application of a field. In addition, electrostatic V-shaped switching behavior is clearly evidenced by the electrooptics data given in Figure 4C. The ferroelectric polarization derived from the observed polarization reversal current for W653 ranges from 660 nC/cm² at 154 °C to 715 nC/cm² at 140 °C. This large observed polarization is similar to the maximum ferroelectric polarization observed for the SmAP_F prototoype W586 of 587 nC/cm² [8].

Defining the three refractive indices n_Z, n_P, and n_O, in a planar-aligned SmAP_F cell as indicated in the inset, the index seen for the field off state (P parallel to the plates) is (n_Z-n_P), while that seen for the field on state (at saturation) is (n_Z-n_O). The index modulation for the field off/on states is greater for W653 than for W586, as indicated in Figure 4D, consistent with the expectation that a tighter bend angle would increase the maximum modulation depth in the V-shaped switching by increasing n_P. The higher δ(∆n) of W653 relative to the prototype W586 at the same reduced temperatures, is illustrated in Figure 4E.

5. Properties of a New SmAP_F Mesogen Possessing a PF-PEG Tail (W789)

As discussed above, the first known SmAP_F mesogens were designed to test the hypothesis that suppression of out-of-layer fluctuations allows the formation of anticlinic layer interfaces (ferroelectric order for bent-cores) in smectic LCs. In the case of the classic SmAP_F mesogen W586, and the more strongly bent diphenylthioether W653, suppression of out-of-layer fluctuations was achieved by incorporating a tricarbosilane moiety at the end of the tail, expected to enhance nano-phase segregation at the layer interfaces, thereby allowing the formation of anti-clinic layer interfaces.

Here we provide an additional positive test of this hypothesis with characterization of a SmAP_F mesogen possessing a PF-PEG tail. The first material synthesized for this effort, W788 (Figure 3), possesses a cyano group “opposite” the PF-PEG tail, similar to W586 and W653. Somewhat surprisingly, W788 did not provide a SmAP phase, but rather a monotropic B1 phase, as indicated in Figure 3. The B1 phase is a variety of columnar phase well known among bent-core FLCs. However, when the cyano group opposite the PF-PEG tail was replaced by another well-known “short” polar group, trifluoromethoxy, a new SmAP_F mesogen, W789, resulted. The phase sequence, transition temperatures, and transition enthalpies on heating obtained from DSC are given in Figure 3. As indicated, W789 exhibits an enantiotropic SmAP_F phase between 123 °C and 196 °C. The phase assignment is based upon observation of the diagnostic electrooptics expected for the SmAP_F phase, analog index modulation with no brush rotation, as indicated in the photomicrographs given in Figure 5A,B. In this case the index modulation depth is slightly less than half that seen with W586, as illustrated in Figure 5C. For W789 the maximum observed index modulation depth δ(∆n) (measured about 40 °C below the Iso–SmAP_F transition, just before crystallization) is about 0.025, while that for W586 is 0.0425 at the same reduced temperature. It is interesting to note that we observe no change in the polarized light microscope texture upon crossing the Crystal–SmAP_F phase boundary. This led initially to characterization of the material at low temperature as a glass. However, the DSC (Figure 5D) shows a sharp first-order transition from the SmAP_F to the crystal, both on heating and on cooling, with a large transition enthalpy.

Additionally, the small-angle X-ray scattering profile of this crystal phase is essentially identical to that of the SmAP_F, as shown in Figure 5E. Also notable is the prominent second harmonic of the layer peak, consistent with suppression of out-of-layer fluctuations. The observed layer peak at q = 0.143 Å⁻¹ suggests a layer spacing in the SmAP_F phase of W789 of d = 43.6 Å. The X-ray diffraction (XRD) results suggest that the structure of the crystal phase maintains the “smectic layer” metric, with apparent crystallization of the molecules in these layers.

Finally, as expected for a SmAP_F mesogen, W789 exhibits a high ferroelectric polarization P = 786 nC/cm² at 150 °C, and P = 837 nC/cm² at the lower end of the SmAP_F phase temperature range. This value is to our knowledge the highest polarization seen to date in a SmAP_F mesogen, slightly higher than the previous record measured for W623, an isomer of W586, of 830 nC/cm² [17].

6. Speculation Regarding the Layer Structure of the SmAP_F Phase of W789

As indicated in Figure 5, the layer spacing in the SmAP_F phase of W789 obtained from X-ray diffraction is 43.6 Å. A “fully extended” conformation of this molecule imposing an “all anti” PF-PEG tail (not actually expected to be a well-populated conformation) as indicated in Figure 6A, is about 46.8 Å (including two-times Spartan’s fluorine atom Van der Waals radius)—seemingly a fairly close match. However, the picture is complicated by the fact that the mesogen has only one tail, and our “first wild guess” is that the cores should prefer to overlap giving the molecule pair indicated in Figure 6B. The “molecular length” for this pair is about 61.5 Å, which is much larger than the experimental layer spacing. This model forms the basis of the often-proposed “partial bilayer” structure, where the tails overlap considerably. While the body of work on PF-PEG conformational analysis is small, computations by Glaser et al. suggest that the gauche and anti conformations of perfluoroethyleneglycol are similar in energy [25], leading to a higher incidence of gauche conformers of the tail than would be expected for a nominal straight-chain hydrocarbon tail. This would tend to produce more “twisted” tails with a shorter length on average.

In addition, a preliminary computational conformational analysis of PhCOOCH₂CF₂OCF₂CF₂OCF₃ using a “small basis set” density function model (ωB97X-D 6-31G*) in a “nonpolar solvent” (an attempt to negate strong intramolecular interactions), using the computational package Spartan’16 [26], gives an average value for the (carbonyl carbon-O-CH₂-CF₂-) dihedral angle for the 14 lowest energy conformers of 104°. Removing one outlier with energy in the middle of the range and a calculated dihedral that is close to 180°, gives an average dihedral of 98° for this molecular fragment. While these calculations should be considered a primitive starting point in the interesting conformational analysis of the Ph-PF-PEG system, if the indicated dihedral angle is indeed averaging about 100° in the LC phase, then “bent” conformations such as that shown in Figure 7 would be preferred. It is easy to imagine that a combination of “partial bilayer” tail overlap and a large majority of twisted and bent conformations could lead to the observed layer spacing.

7. Conclusions

Two new chemical structural classes exhibiting the interesting and relatively rare SmAP_F liquid crystal phase are reported. One of these, the thio-bisphenol derivative W653, shows enhanced electrooptic index modulation depth relative to the prototype SmAP_F mesogen W586, by a design accomplished by lowering the bend angle of the mesogen with a diphenylthioether core. In addition, the occurrence of this phase in a mesogen possessing a PF-PEG tail significantly expands the SmAP_F phase chemical structural space, and is consistent with the hypothesis that suppression of out-of-layer fluctuations can allow the formation of anticlinic layer interfaces, leading to a ferroelectric layer structure for a bent-core mesogen. The discovery of new SmAP_F structural classes suggests it may be possible to obtain such materials with lowered SmAP_F phase temperature ranges, allowing realistic exploration of potential applications.

8. Synthesis Experimental Workflow

Reagents and starting materials were used as purchased from qualified suppliers without additional purification. Tetrahydrofuran (THF) was freshly distilled under argon from sodium benzophenone ketyl, dichloromethane was purchased from Sigma-Aldrich^® (Now Millipore Sigma, Darmstadt, Germany) dry, sure seal grade solvent, and other solvents were used as purchased. Non-aqueous reactions were performed in oven-dried glassware under an atmosphere of dry argon. Purification by flash chromatography was performed with silica gel (40–63 microns) purchased from Zeochem AG^®. (Uetikon am See, Switzerland) Analytical thin-layer chromatography (TLC) was performed on Silica gel 60 F₂₅₄ TLC plates from Millipore Sigma (Darmstadt, Germany) Compounds were visualized with shortwave ultra-violet (UV), or by staining with I₂. Nuclear magnetic resonance (NMR) spectra were obtained with a Bruker Avance-III 300 spectrometer (Brooker Daltonics Inc., Bellerica, MA, USA) or with a Varian INOVA 500 spectrometer (no longer commercially available). NMR chemical shifts were referenced to CHCl₃ (7.24 ppm for ¹H, 77.16 ppm for ¹³C). Exact mass was determined using electrospray ionization time-of-flight (ESI-TOF) mass spectrometry. Liquid crystal phase sequences and phase transition temperatures were determined using polarized light microscopy with a Nikon Optiphot 2 POL microscope equipped with an (Instec temperature-controlled hot stage. DSC was performed using a Mettler DSC823^e differential scanning calorimeter (Mettler Toledo, Columbus, OH, USA). Electrooptics and birefringence measurements made using a Zeiss microscope (Carl Zeiss Microscopy, LLC, Thornwood, NY, USA) fitted with an Ehringhaus rotary compensator with quartz combination plates and 656.3 nm light.

4-[(4-hydroxyphenyl)sulfanyl]phenyl 4-cyanobenzoate (2)

4-[(4-hydroxyphenyl)sulfanyl]phenol (1) (6.60 g, 30 mmol) was dissolved in THF (200 mL), and cooled to 0 °C. Triethylamine (3.64 g, 36 mmol, 5 mL) was then added, followed by dropwise addition of 4-cyanobenzoyl chloride (5.00 g, 30 mmol) in THF (20 mL). The reaction mixture was then allowed to stir overnight at room temperature, then poured to a saturated aqueous solution of NH₄Cl (200 mL) and extracted with 3 aliquots of EtOAc. The combined organic layers were washed with water, brine, dried over MgSO₄, filtered and concentrated at reduced pressure. The crude product was purified by flash chromatography (silica gel, CH₂Cl₂/1% EtOH), to afford ester 2 as a white solid (3.60 g. 35%).

¹H NMR (300 MHz, Chloroform-d) ∂ ppm: 8.35–8.22 (m, 2H), 7.88–7.75 (m, 2H), 7.47–7.33 (m, 2H), 7.30–7.15 (m, 2H), 7.15–7.04 (m, 2H), 6.90–6.77 (m, 2H), 5.06 (s, 1H). ¹³C NMR (75 MHz, CDCl₃) ∂ ppm: 163.77, 156.28, 148.73, 136.88, 135.88, 133.36, 132.56, 130.79, 129.43, 124.27, 122.12, 117.95, 117.21, 116.75.

4-[(4-{4-[4-({11-[({dimethyl[(trimethylsilyl)methyl]silyl}methyl)dimethylsilyl]undecyl}oxy)-benzoyloxy]benzoyloxy}phenyl)sulfanyl]phenyl 4-cyanobenzoate (W653)

To a solution of 4-[4-({11-[({dimethyl[(trimethylsilyl)methyl]silyl}methyl)dimethylsilyl]-undecyl}oxy)benzoyloxy]benzoic acid (3) [7] (0.905 g, 0.14 mmol) and ester (2) in CH₂Cl₂ (100 mL), DCC was added (0.50 g, 0.24 mmol) as well as a trace of DMAP. The reaction mixture was allowed to stir at room temperature for 2 days, then filtered, washed sequentially with water, 5% CH₃COOH, water, and brine, dried over MgSO₄, filtered, and concentrated at reduced pressure. The resulting crude product was purified by flash chromatography (silica gel, CH₂Cl₂). The solvent was removed at reduced pressure, filtered, recrystallized from CH₃CN/EtOAc (40:10), and dried in vacuo overnight, giving 0.712 g, (53%) of W653.

DSC 5 (deg/min ramp):

Heating: X − 136.3 °C→SmAP_F (ΔH = −16.30 KJ mol⁻¹) − 150.1 °C→I (ΔH = −15.57 KJ mol⁻¹).

Cooling: I − 147.7 °C→SmAP_F (ΔH = 16.10 KJ mol⁻¹) − 133.2 °C→X (ΔH = 15.46 KJ mol⁻¹).

¹H NMR (300 MHz, Chloroform-d) ∂ ppm: 8.36–8.21 (m, 4H), 8.21–8.10 (m, 2H), 7.88–7.77 (m, 2H), 7.52–7.32 (m, 6H), 7.27–7.13 (m, 4H), 7.05–6.93 (m, 2H), 4.05 (t, J = 6.5 Hz, 2H), 1.83 (p, J = 6.7 Hz, 2H), 1.55–1.28 (m, 16H), 0.49 (d, J = 8.5 Hz, 2H), 0.11–0.02 (m, 21H), 0.27 (d, J = 7.0 Hz, 4H). ¹³C NMR (75 MHz, CDCl₃) ∂ ppm: 164.46, 164.44, 164.01, 163.55, 155.67, 150.52, 149.74, 134.21, 133.33, 132.89, 132.68, 132.58, 132.05, 131.99, 130.82, 126.73, 122.90, 122.48, 122.33, 121.03, 117.95, 117.30, 114.58, 68.56, 33.87, 29.79, 29.76, 29.72, 29.55, 29.53, 29.24, 26.14, 24.14, 18.22, 5.94, 4.18, 2.62, 1.63, −0.28.

HRMS (ESI-TOF) m/z for C₅₄H₆₇NO₇Si₃SLi⁺ [M + Li]⁺; calc. 964.4101, found 964.4089.

2,2-difluoro-2-{1,1,2,2-tetrafluoro-2-[1,1,2,2-tetrafluoro-2-(trifluoromethoxy)ethoxy]ethoxy}ethyl 4-(benzyloxy)benzoate

Starting material 2,2-difluoro-2-{1,1,2,2-tetrafluoro-2-[1,1,2,2-tetrafluoro-2-(trifluoromethoxy)-ethoxy]ethoxy}ethan-1-ol (5) was as purchased from Exfluor^® (Round Rock, TX, USA) Compound (6) was prepared by a modification of the procedure described in [27], using CDI in the presence of DBU as coupling reagent, yield 57% (lit. 86%).

4-(4,4,6,6,7,7,9,9,10,10,12,12,12-tridecafluoro-2,5,8,11-tetraoxadodecanoyl)phenyl 4-(benzyloxy)benzoate

To a solution of 2,2-difluoro-2-{1,1,2,2-tetrafluoro-2-[1,1,2,2-tetrafluoro-2-(trifluoromethoxy)-ethoxy]ethoxy}ethyl 4-hydroxybenzoate (6) (11.72 g, 23 mmol) and 4-benzyloxy benzoic acid (4) (5.16 g, 23 mmol) in CH₂Cl₂ (136 mL) DCC was added, (5.7 g, 28 mmol) as well as a trace of DMAP. The reaction mixture was allowed to stir at room temperature for 3 days, then filtered, washed with water, 5% CH₃COOH, water, brine, dried over MgSO₄, filtered, and concentrated at reduced pressure. The crude product was purified by flash chromatography (silica gel, CH₂Cl₂). White solid (11.88 g, 71%). Glass 128.7–130 °C SmA 161–163.5 °C I.

¹H NMR (300 MHz, Chloroform-d) ∂ ppm: 8.22–8.07 (m, 4H), 7.51–7.27 (m, 7H), 7.14–7.02 (m, 2H), 5.17 (s, 2H), 4.72 (t, J = 9.3 Hz, 2H).

¹³C NMR (75 MHz, CDCl₃) ∂ ppm: 164.28, 164.24, 163.48, 155.71, 136.16, 132.61, 131.73, 128.88, 128.47, 127.64, 125.81, 122.29, 121.55, 114.97, 70.38, 62.61, 62.16, 61.72.

4-(4,4,6,6,7,7,9,9,10,10,12,12,12-tridecafluoro-2,5,8,11-tetraoxadodecanoyl)phenyl 4-hydroxy-benzoate (7)

A solution of the benzyl ether-protected phenol from above (25.00 g, 34 mmol) in CH₂Cl₂ (400 mL) and EtOH (150 mL) was first briefly evacuated and purged with argon, then 10% Pd/C catalyst (4 g) was added. The argon atmosphere was replaced by hydrogen gas, and the reaction mixture was stirred at room temperature for 8 hours. Hydrogen was pumped out of the system and the flask was purged thoroughly with argon. The mixture was filtered through Celite and solvents were removed under reduced pressure. This gave product as a white crystalline powder, m.p. 117–119.5 °C (21.60 g, 99%).

¹H NMR (300 MHz, Chloroform-d) ∂ ppm: 8.19–8.03 (m, 4H), 7.39–7.27 (m, 2H), 6.99–6.85 (m, 2H), 6.13 (s, 1H), 4.72 (t, J = 9.3 Hz, 2H). ¹³C NMR (75 MHz, CDCl₃) ∂ ppm: 164.44, 164.31, 160.79, 155.65, 132.93, 131.78, 125.87, 122.31, 121.53, 115.71, 62.66, 62.21, 61.76.

4-{[4-(4,4,6,6,7,7,9,9,10,10,12,12,12-tridecafluoro-2,5,8,11-tetraoxadodecanoyl)phenoxy]carbonyl}-phenyl 3-(benzyloxy)benzoate

To a solution of 3-benzyloxy benzoic acid and phenol 7 resulting from the prior debynzylation (6.26 g, 25 mmol) (15.95 g, 25 mmol) was added was added DCC (5.70 g, 28 mmol) and trace of DMAP in CH₂Cl₂ (400 mL). The reaction mixture was allowed to stir at room temperature for 3 days, then filtered, washed with water, 5% CH₃COOH, water, brine, dried over MgSO₄, filtered, and concentrated at reduced pressure. The resulting product was purified by flash chromatography (silica gel, CH₂Cl₂), followed by crystallization from CH₃CN. White solid, I 130 °C SmA 66 °C glass (17.54 g, 83%).

¹H NMR (300 MHz, Chloroform-d) ∂ ppm: 8.37–8.23 (m, 2H), 8.23–8.08 (m, 2H), 7.93–7.78 (m, 2H), 7.55–7.24 (m, 11H), 5.16 (s, 2H), 4.74 (t, J = 9.3 Hz, 2H). ¹³C NMR (75 MHz, CDCl₃) ∂ ppm: 164.52, 164.18, 163.89, 159.08, 155.60, 155.46, 136.49, 132.14, 131.82, 130.38, 129.97, 128.85, 128.38, 127.73, 126.67, 126.10, 123.08, 122.32, 122.24, 121.36, 115.97, 70.46, 62.20.

4-{[4-(4,4,6,6,7,7,9,9,10,10,12,12,12-tridecafluoro-2,5,8,11-tetraoxadodecanoyl)phenoxy]-carbonyl}phenyl 3-hydroxybenzoate

A solution of the benzyl-protected tricyclic phenol (16.60 g, 20 mmol) in CH₂Cl₂ (150 mL) and EtOH (150 mL) was first briefly evacuated and purged with argon, then 10% Pd/C catalyst (2.82 g) was added. The argon atmosphere was replaced by hydrogen gas, and the reaction mixture was stirred at room temperature for 3 hours. Hydrogen was pumped out of the system and the flask was purged thoroughly with argon. The mixture was filtered through Celite and solvents were removed under reduced pressure. This gave tetracyclic phenol 8 as white powder with phase sequence: I 163 °C SmA 110 °C X (14.75 g, 99%).

¹H NMR (300 MHz, Chloroform-d) ∂ ppm: 8.35–8.22 (m, 2H), 8.21–8.10 (m, 2H), 7.81 (ddd, J = 7.7, 1.6, 1.0 Hz, 1H), 7.71–7.63 (m, 1H), 7.48–7.30 (m, 5H), 7.15 (ddd, J = 8.1, 2.7, 1.0 Hz, 1H), 5.13 (s, 1H), 4.73 (t, J = 9.3 Hz, 2H).

¹³C NMR (75 MHz, CDCl₃) ∂ ppm: 155.98, 155.54, 132.16, 131.83, 130.22, 126.71, 122.94, 122.30, 122.24, 121.36, 116.97, 115.70, 62.23.

4-{[4-(4,4,6,6,7,7,9,9,10,10,12,12,12-tridecafluoro-2,5,8,11-tetraoxadodecanoyl)phenoxy]carbonyl}-phenyl 3-[4-(benzyloxy)benzoyloxy]benzoate

To a suspension of compound the tricyclic phenol resulting from the above debenzylation (14.08 g, 18 mmol) and 4-benzyloxy benzoic acid (4) (4.21 g, 18 mmol) in CH₂Cl₂ (800 mL), DCC was added (4.5 g, 22 mmol), as well as a trace of DMAP. The reaction mixture was allowed to stir at room temperature for 3 days, then filtered, washed with water, 5% CH₃COOH, water, brine, dried over MgSO₄, filtered, and concentrated at reduced pressure. The crude product was purified by flash chromatography (silica gel, CH₂Cl₂), followed by two crystallizations from CH₃CN and later from toluene to give a white solid, I 156 °C SmA 150 °C Sm (not assigned) 127 °C Glass (9.92 g, 57%).

¹H NMR (300 MHz, Chloroform-d) ∂ ppm: 8.36–8.03 (m, 8H), 7.67–7.30 (m, 11H), 7.14–7.03 (m, 2H), 5.18 (s, 2H), 4.73 (t, J = 9.2 Hz, 2H). ¹³C NMR (75 MHz, CDCl₃) ∂ ppm: 164.79, 164.18, 163.87, 163.85, 163.46, 155.44, 151.40, 136.16, 132.59, 132.17, 131.82, 130.60, 129.99, 128.89, 128.47, 127.79, 127.65, 126.78, 126.10, 123.92, 122.29, 122.23, 121.60, 114.98, 70.39, 62.20.

4-{[4-(4,4,6,6,7,7,9,9,10,10,12,12,12-tridecafluoro-2,5,8,11-tetraoxadodecanoyl)phenoxy]carbonyl}-phenyl 3-(4-hydroxybenzoyloxy)benzoate (8)

A solution of compound the tetracyclic benzylether-protected phenol from above (9.37 g, 9.67 mmol) in CH₂Cl₂ (250 mL) and EtOH (150 mL) was first briefly evacuated and purged with argon, then 10% Pd/C catalyst (3 g) was added. The argon atmosphere was replaced by hydrogen gas, and the reaction mixture was stirred at room temperature for 3 h. Hydrogen was pumped out of the system and the flask was purged thoroughly with argon. The mixture was filtered through Celite and solvents were removed under reduced pressure. This gave the key intermediate tetracyclic phenol 8 as a white crystalline solid mp 170–172 °C.

¹H NMR (300 MHz, Chloroform-d) ∂ ppm: 8.35–8.10 (m, 8H), 7.67–7.30 (m, 6H), 6.98–6.87 (m, 2H), 5.58 (s, 1H), 4.73 (t, J = 9.3 Hz, 2H). ¹³C NMR (75 MHz, CDCl₃) ∂ ppm: 164.83, 164.21, 163.93, 160.76, 155.43, 151.35, 135.61, 132.88, 132.19, 132.11, 131.83, 130.60, 130.02, 127.85, 126.78, 126.11, 123.92, 122.29, 122.24, 121.60, 115.71, 62.22.

4-{[4-(4,4,6,6,7,7,9,9,10,10,12,12,12-tridecafluoro-2,5,8,11-tetraoxadodecanoyl)phenoxy]carbonyl}-phenyl 3-[4-(4-cyanobenzoyloxy)benzoyloxy]benzoate (W788)

Tetracyclic phenol 8 (1.01 g, 1.15 mmol) was dissolved in CH₂Cl₂ (150 mL) and cooled to 0 °C. Triethylamine (0.133 g, 1.31 mmol, 0.2 mL) was then added, followed by dropwise addition of 4-cyanobenzoyl chloride (9) (0.190 g, 1.15 mmol), followed by addition of a trace of DMAP. The reaction mixture was stirred overnight at room temperature, poured into the saturated NH₄Cl, and extracted to the CH₂Cl₂ (three times). The combined organic layers were washed with water, brine, dried over MgSO₄, filtered and concentrated at reduced pressure. The crude product was purified by flash chromatography (silica gel, CH₂Cl₂/1% EtOAc), then filtered and crystallized from CH₃CN twice. The result was a white solid (0.750 g. 68 %).

¹H NMR (300 MHz, Chloroform-d) ∂ ppm: 8.40–8.25 (m, 6H), 8.22–8.06 (m, 4H), 7.92–7.80 (m, 2H), 7.70–7.51 (m, 2H), 7.48–7.30 (m, 6H), 4.73 (t, J = 9.2 Hz, 2H). ¹³C NMR (75 MHz, CDCl₃) ∂ ppm: 164.20, 163.84, 155.43, 155.39, 155.00, 151.14, 132.93, 132.67, 132.25, 132.19, 131.82, 130.89, 130.76, 130.13, 128.09, 127.24, 126.84, 123.78, 122.27, 122.23, 122.11, 117.85, 117.60, 62.20.

HRMS (ESI-TOF) m/z for C₄₃H₂₂F₁₃NO₁₃Li⁺ [M + Li]⁺; calc. 1014.1038, found 1014.1040.

DSC 5 deg/min:

Heating: Glass 134.4 °C SmX (ΔH = −1.19 KJ mol⁻¹) 169.9 °C I (ΔH = −33.95 KJ mol⁻¹).

Cooling: I 167.3 °C Col (ΔH = 8.09 KJ mol⁻¹) 157.9 °C SmX (ΔH = 23.86 KJ mol⁻¹) 105.5 °C (ΔH = 0.70 KJ mol⁻¹).

4-{[4-(4,4,6,6,7,7,9,9,10,10,12,12,12-tridecafluoro-2,5,8,11-tetraoxadodecanoyl)phenoxy]carbonyl}-phenyl 3-{4-[4-(trifluoromethoxy)benzoyloxy]benzoyloxy}benzoate (W789)

Tetracyclic phenol 8 (1.34 g, 1.52 mmol) was dissolved in CH₂Cl₂ (200 mL), and cooled to 0 °C. Trietheylamine (0.182 g, 1.80 mmol, 0.25 mL) was then added, followed by dropwise addition of 4-trifluoromethoxy benzoyl chloride (10) (0.341 g, 1.52 mmol, 0.24 mL), followed by addition of a trace of DMAP. The reaction mixture was allowed to stir overnight at room temperature, poured to the saturated NH₄Cl, and extracted to CH₂Cl₂ (three times). The combined organic layers were washed with water, brine, dried over MgSO₄, filtered, and then concentrated at reduced pressure. The crude product was purified by flash chromatography (silica gel, CH₂Cl₂/1% EtOAc), then filtered and crystallized from CH₃CN to give W789 as a white solid (1.25 g, 82%).

¹H NMR (300 MHz, Chloroform-d) ∂ ppm: 8.39–8.24 (m, 6H), 8.20–8.07 (m. 4H) 7.67–7.57 (m, 2H), 7.45–7.33 (m, 8H), 4.73 (t, J = 9.3 Hz, 2H). ¹³C NMR (75 MHz, CDCl₃) ∂ ppm: 164.30, 164.18, 163.84, 163.79, 163.54, 155.44, 155.40, 155.30, 151.17, 132.52, 132.19, 131.83, 130.74, 130.12, 128.07, 127.71, 127.41, 126.94, 126.83, 126.11, 123.81, 122.28, 122.24, 120.66, 120.64, 62.21.

HRMS (ESI-TOF) m/z for C₄₃H₂₂F₁₆O₁₄Li⁺ [M + Li]⁺; calc. 1073.0909, found 1073.0914.

DSC 5 deg/min:

Heating: Glass 123.3 °C SmAP_F (ΔH = −29.73 KJ mol⁻¹) 196.2 °C I (ΔH = −14.32 KJ mol⁻¹).

Cooling: I 194.6 °C SmAP_F (ΔH = 15.89 KJ mol⁻¹) 98.8 °C (ΔH = 20.29 KJ mol⁻¹)^.