Design and Investigation of a Side-Chain Liquid Crystalline Polysiloxane with a N_tb-Phase-Forming Side Chain

Abstract

A new mesogenic non-symmetric dimeric monomer with a terminal olefin function, forming a twist bend nematic (N_tb) as well as a nematic (N) phase, was synthesized, using an enhanced synthetic methodology, which avoids isomerization of the terminal double bond in the preparation of the dimer. This monomer was attached to a pentamethyldisiloxane group, resulting in the SmA LC phase behavior of the ensuing material. Linking the monomer to a siloxane main chain resulted in nematic phase behavior. Detailed studies with the N_tb phase forming dimer DTC5C7 show full miscibility of the dimer and the new LC polymer in the LC state, suggesting that the side-chain LC polymer forms a N_tb phase as the low-temperature nematic phase. Copolymerizing the monomer with a cyanobiphenyl-based monomer allows us to tune the glass transition and phase behavior further.

1. Introduction

Since the first side-chain liquid crystal polymer was reported in 1978 [1,2,3,4,5,6], the interest in the synthesis and applications of side-chain liquid crystal polymers (SCLCPs) has evolved. Side-chain liquid crystal polymers are polymers where mesogenic units are attached to backbones via flexible spacers. Thus, three elements are crucial in determining the properties of liquid crystal side-chain polymers: the polymer backbones, the spacers and the mesogenic units. Among all the backbones reported in the literature, polymethylsiloxanes are arguably the most flexible; this is combined with their excellent thermal stability and low glass transition temperatures. For a given backbone, the molecular structure of the mesogenic units, together with the length and parity of spacers, can have a profound effect on liquid crystal properties and the transition temperature ranges. However, as crystallization might be observed in some of the SCLCPs, a considerable amount of effort has been carried out to prevent polymers from unwanted crystallization and to reduce glass transition temperatures. Examples of such efforts are the use of different types of spacers [7], the variation of the length of the spacers [8] and the use of mesogens that are either terminally or laterally attached to the side chains [9]. Other tools are the use of copolymers with different mesogenic units [10] or even simply the use of non-mesogenic groups [11,12].

So far, there are no reports, to the best of our knowledge on polysiloxanes coupled with side-chain dimers to form side-chain liquid crystal polymers that exhibit N_tb phase behavior. Though it is noted that a number of polymers have been reported that carry di-mesogenic side chains, mainly based on malonate branching groups [13,14,15]. The N_tb phase for dimers is known to show often non-typical optical defect textures; also, it is not easy to distinguish the N_tb from the nematic phase by X-ray diffraction. So, in this study, a monomer was designed to consist of hydrocarbon units as spacers and terphenyl and biphenyl groups as mesogenic units, together with terminal double bonds, all designed for controlling the transition temperature and phase sequence. The overall approach is outlined schematically in Figure 1.

The reason why this architecture was chosen is that: (a) it has been reported that dimers with similar structures such as DTC5C_n show the N_tb phase, even for related linear oligomers [16]. (b) The dimers structurally related to the system discussed here exhibit the N_tb phase over a large temperature range, a feature that makes it possible to tune the transition temperature of the N_tb phase by fine tuning the architecture [17]. Also, (c) a non-symmetric mesogenic system containing altogether five aromatic rings may result in a lower melting point when compared with a six-membered ring system. This is critical, as the transition temperatures of side-chain polymers are often higher than those of the monomers. Since pentamethydisiloxane derivatives can exhibit properties associated with polymers and maintain the low viscosity of monomers, it is interesting to investigate the liquid crystal functionalized N_tb system too. Additional results of a copolymerization that was carried out to eliminate crystallization are also reported. The overall approach is shown in Figure 1.

There were two challenges in this project. Firstly, there was a challenge in the preparation of the mesogen. The challenge was that isomers were formed. They are a byproduct of the Suzuki–Miyaura coupling reaction in the presence of hydrocarbon chains that carry olefinic functions. To eliminate the isomers, a co-catalyst was used together with the conventional Pd catalyst. The second challenge was that it was difficult to identify the N_tb phase of polymers; therefore, binary mixtures of the homopolymer with the N_tb-forming material were used to observe the liquid crystal defect textures using optical polarizing microscopy at controlled temperatures. The textures of the liquid crystal phases of the monomer and homopolymer 1 (HP1) were analyzed by polarized optical microscopy (POM) and differential scanning calorimetry (DSC) and through the construction of phase diagrams.

2. Experiment

2.1. Materials

Poly(methylhydrosiloxane) (average Mn 1700–3200) and a solution of platinum divinyl tetramethyl disiloxane complexes, with a platinum concentration of 2.5–5 wt%, were purchased from Sigma-Aldrich and used as received.

All other starting materials came from Sigma-Aldrich, Alfa Aesar, Manchester Organics, Fluorochem, TCI, ACROS organics and Apollo Scientific. All the company information is listed in the following: Apollo Scientific—Whitefield Rd, Bredbury, Stockport, SK6 2QR, UK; Sigma-Aldrich-Co., 3050 Spruce Street, St. Louis, MO 63103, USA; Tokyo Chemical Industry Co., Ltd. 6-15-9 Toshima, Kita-ku, Tokyo, Japan; ACROS organics-New Jersey, ISA 1-800-ACROS-01, Ceel, Belgium; Fluo-rochem-Fluorochem Ltd., Unit,14, Graphite Way, Hadfield, SK13, 1QH, UK; Alfa Ae-sar-Shore Road, Heysham, LA32XY, England; Manchester Organics-The Heath Business &Technical Park, Runcorn, Cheshire, UK, WA7 4QK. All compounds purchased were used without further purification. Silica gel 60 F254 Merck KGaA was used for thin-layer chro-matography.

The degree of polymerization written as $\overline{\mathrm{D}\mathrm{P}}$ is calculated by $\overline{\mathrm{M}\mathrm{w}}$/${\mathrm{M}}_0$, using the values obtained by gel permeation chromatography (GPC), where Mw is the weight average molecular weight of the polymer, and M₀ is the molecular weight of the repeating unit of the monomer. The molecular weight distribution, also called polydispersity, is calculated by gel permeation chromatography (GPC) as the value of $\overline{\mathrm{M}\mathrm{w}}$/$\overline{\mathrm{M}\mathrm{n}}$.

2.2. Methods

Column chromatography: All purification by column chromatography utilized Fluorochem Ltd. Silica gel 60A 35–70-micron silica gel and Fisher Scientific Ltd. UK general purpose low iron grade sand. Company information in detail: Fisher scientific UK, Bishop Meadow Road, Loughbor-ough, Leics, LE11 5RG, UK.

¹H-NMR 400MHz, ¹³C-NMR 101 MHz and ¹⁹F-NMR 377 MHz were recorded on a Jeol JNM ECP400 spectrometer.

The thermal transitions were investigated by both differential scanning calorimetry (DSC) and polarized optical microscopy (POM). POM was performed using an Olympus BX51 Polarizing Microscope equipped with a Linkam LTS350 heating stage. The DSC data were from a Perkin Elmer Differential Scanning Calorimeter DSC 4000, an aluminum reference pan was used to load the reference and sample, and the calibration materials were standard indium. DSC results were normally quoted by the average values for the onset of the second heating and cooling curve. The heating rate, if not mentioned specifically, was at 10 °C/min.

Gel permeation chromatography (GPC) was run on a 350A HT-GPC system, equipped with two VE-1122 solvent delivery pumps, a VE-7510 solvent degasser, a Vortex 430 autosampler and the 350A HT detector module. The GPC is from Malvern Instruments, company detail as following: Malvern Instruments Ltd. Enigma Business Park, Grovewood Road, Malvern, Worcestershire WR14 1XZ, United Kingdom. Samples were passed through microfilters before measurement to remove the gel. Disposable microfilters were from Whatman Puradisc 25 TF. 0.2 μm PTFE Membrane with Polypropylene Housing Diameter 25 mm.

Elemental analysis was performed using FISONS Instruments EA 1108 CHN and using software Eager Xperience OS Version: 6.01 for processing the data.

X-ray powder diffraction data were collected from samples mounted on zero-background silicon discs. A PANAlytical Empyrean diffractometer operating in Bragg–Brentano geometry with copper Kα1 (λ = 1.541 Å) and a PIXEL detector was used for the data collection. Data were collected in the range 3° ≤ 2θ ≤ 30°.

Monomers were characterized by high-resolution mass spectrometers. The samples were run on the Bruker maXis, a Hybrid Quadrupole/Atmospheric Pressure Ionization orthogonal accelerated Time-Of-Flight mass spectrometer.

The side chain liquid crystal polymers were characterized by NMR, gel permeation chromatography (GPC), differential scanning calorimetry and optical polarizing microscopy.

2.3. Purification of Polymers

As the amount of the monomers is always in excess in the polymer analogous reaction with the poly(methylhydro)siloxane backbone, and as the traces of monomers can strongly influence the phase transition of related polymers, it was necessary to remove the monomers and purify the polymers. Purification was achieved by repeated precipitation with methanol, after dissolving the polymer with minimum volumes of toluene. The removal of monomers was monitored by DOSY-NMR and thin-layer chromatography on silica with an appropriate solvent combination.

2.4. Synthesis of Monomers

The synthesis route used to prepare the vinylic monomer is shown in Scheme 1. The monomer was synthesized in a twelve-step synthesis route, which was followed by the synthesis of the pentamethyldisiloxane derivative and the polymers.

The overall synthesis is based on a convergent approach. It starts with the conversion of heptanedioic acid to the related acid chloride 1 by using oxalyl chloride. The acid chloride reacts with bromobenzene through a Friedel–Crafts acylation reaction to obtain compound 2. The two keto groups were then reduced by a reaction with triethyl silane and trifluoroacetic acid to reach compound 3. Compound 3 was reacted with compound 8, which was obtained in a separate reaction pathway published earlier [17]. To make compound 8, the synthesis was based on another convergent synthesis route. After the reaction of 1,2-difluorobenzene with n-butyllithium, followed by a reaction with trimethyl borate, the boronic acid intermediate 4 was obtained. This was reacted in a coupling reaction with compound 6. Compound 6 was obtained by reacting bromobenzene with oxalyl chloride and n-pentanoic acid in a Friedel–Crafts acylation, followed by the reduction of the ketone group with triethylsilane and trifluoroacetic acid. The coupling of compound 4 and compound 6 resulted in compound 7, which was converted into a boronic acid by reacting first with n-butyllithium. This was followed by the reaction with trimethyl borate to form compound 8. Compound 3 and compound 8 were coupled to form compound 9. (shown in Scheme 1). The overall reaction pathway procedure toward this type of dimer has already been reported earlier, and hence the chemical characterization will not be reported here again [17].

Monomer M1 was synthesized through a Suzuki–Miyaura coupling reaction of compound 9 and compound 11. Compound 11 was prepared by a Williamson etherification, followed by lithiation and a Miyaura coupling reaction.

In this synthesis, for the preparation of the monomer, a cocatalyst was used in the Suzuki–Miyaura coupling to eliminate the formation of isomers in the alkene chain of the hydrocarbon groups of the monomers. The occurrence of isomerization reactions has been discussed early on, in 1993 [18]. An alternative to produce isomer-free products is to use P(Ph₃)₄ without a cocatalyst. It was found that the main catalyst concentration cannot be higher than 0.5% mol to avoid isomerization of the terminal double bond (see Figure 2b). However, in this procedure, the yield was only 15%, which is very low. Hence, a [1,1′-Bis(diphenylphosphino)ferrocene] dichloropalladium (II) complex with a dichloromethane/triphenyl arsine catalyst system was employed. The latter catalyst system was not only used to avoid formation of the isomers during the coupling reaction but also to help to maintain the yield of the isolated product after column chromatography at higher than 50%. This is due to the bulky structure of the main catalyst formed together with the cocatalyst, suppressing the beta-hydride elimination reaction which produces isomers. ¹HNMR results for the pure monomer are shown in Figure 2a, and an example of an isomer formed during the initial exploratory work is shown in Figure 2b.

The ¹HNMR (CDCl₃) spectra for HP1 are virtually identical to the monomer M1, with the addition of the signals for the silyl protons at δ = 0.07 (48, s, Si-CH₃) and δ= 0.86 (14,s, Si-CH₂) and the loss of signals for the vinylic protons at δ = 5.99–5.75 (1H, m, -CH = CH₂) and 5.15–4.95 (2H, m, -CH = CH₂). See below in Figure 2a and Figure 3 as the ¹HNMR of monomer M1 and polymer HP1 separately. As typical for polymers, the signals are somewhat broader when compared to the monomer. The synthesis of monomer 4′-(pent-4-en-1-yloxy)-[1,1′-biphenyl]-4-carbonitrile discussed later in this work and corresponding homopolymer has been reported in detail much earlier [19,20].

Synthesis of the yet unreported steps for the preparation of M1.

2′3′-difluoro-4-(7-(4′-(pent-4-en-1-yloxy)-[1,1′-biphenyl]-4-yl)heptyl)-4″-pentyl-1,1′:4′,1″-terphenyl M1.

(2,3-difluoro-4-(pent-4-en-1-yloxy) phenyl) boronic acid 11 (0.11 g, 0.52 mmol) and 4-(7-(4-bromophenyl) heptyl)-2′,3′-difluoro-4″-pentyl-1,1′:4′,1″-terphenyl 9 (0.32 g, 0.52 mmol). Caesium carbonate was put in THF (50 mL) and water, and the solution was degassed 3 times. After that, the [1,1′-Bis(diphenylphosphino)ferrocene] dichloropalladium(II) complex with dichloromethane (2.12 mg, 0.5 mol%, 2.6 × 10⁻⁶ mmol) was added. This was followed by the addition of triphenyl arsine (23 mg, 1.5 mol% and 7.8 × 10⁻⁵ mmol), and the mixture was stirred at 60 °C overnight. The cooled mixture was poured into brine, and the product was extracted into chloroform (3 × 30 mL). The combined organic layer was dried by anhydrous MgSO_4, and the solvent was removed under vacuum. The crude compound M1 was purified by column chromatography (DCM:Hexane = 1:4) and recrystallized from iso-propanol.

Yield: 0.18 g, 51.6%.

High-resolution mass spectroscopy: m/z 693.3883, err −0.4 ppm.

Elemental analysis: theoretical: C 84.14%; H 7.81%; experimental: C 84.20%; H 7.83%

¹HNMR (400 MHz, CDCl₃): δ 7.56 – 7.50 (6H, m), 7.48 (2H, d, J = 8.1 Hz), 7.33 – 7.27 (4H,m), 7.24 (4H, dd, J = 6.2, 4.5 Hz), 6.99 – 6.94 (2H, m), 5.95 – 5.81 (1H,m), 5.15 – 4.98 (2H,m), 4.02 (2H, t, J = 6.5 Hz), 2.67 (6H, dd, J = 17.4, 9.8 Hz), 2.33 – 2.19 (2H, m), 1.98 – 1.84 (2H, m), 1.67 (6H, dd, J = 13.0, 6.8 Hz), 1.46 – 1.32 (10H, m), 0.94 (3H, dd, J = 9.0, 4.7 Hz).

¹³CNMR (101 MHz, CDCl₃): δ 158.53, 143.23, 143.15, 141.52, 138.34, 137.98, 133.75, 132.04, 129.61, 128.91,128.82, 128.07, 126.69, 124.73, 124.70,115.34, 114.86, 67.37, 35.84, 35.68, 31.71, 31.61, 31.49, 31.25, 30.28, 29.48, 29.36, 28.61, 22.71, 14.20.

¹⁹F NMR (377 MHz, CDCl₃): δ −143.21.

2.5. Synthesis of the Pentamethyldisiloxane Derivative (PMDS1)

The pentamethydisiloxane derivative was prepared by a platinum-catalyzed hydrosilylation, as shown in Scheme 2.

After the hydrosilylation reaction with 1,1,1,3,3-pentamethyl-1,3-disiloxane and M1, using Karstedt’s catalyst, the pentamethydisiloxane derivative PMDS1 was obtained (see Scheme 2). The synthesis is described in detail below:

1-(5-((4′-(7-(2′,3′-difluoro-4″-pentyl-[1,1′:4′,1″-terphenyl]-4-yl)heptyl)-[1,1′-biphenyl]-4-yl) oxy) pentyl)-1,1,3,3,3-pentamethyldisiloxane PMDS 1.

Monomer 2′,3′-difluoro-4-(7-(4′-(pent-4-en-1-yloxy)-[1,1′-biphenyl]-4-yl)heptyl)-4″-pentyl-1,1′:4′,1″-terphenyl M1 (0.17 g, 0.254 mmol) in anhydrous toluene (50 mL) was stirred vigorously under nitrogen with 1,1,1,3,3-pentamethyldisiloxane (0.5 mL, 2.56 mmol) at 50 °C for 3 days, using a Karstedt’s catalyst solution in xylene (Pt ~2%) (0.5 mL, 0.0224 mmol) as the catalyst. The solvent was removed by vacuum, and the catalyst was removed by passing the dry crude through silica gel and washing it with DCM. Silica gel column chromatography was used to further purify the crude (DCM/Petrol = 1/2). After recrystallization by isopropanol, it was obtained as a white solid, yielding 67 mg, 32.4%.

¹H NMR (400 MHz, CDCl₃): δ 7.54–7.43 (m, 8H), 7.32–7.26 (m, 4H), 7.23 (dd, J = 7.8, 5.0 Hz, 4H), 6.98–6.91 (m, 2H), 3.99 (t, J = 6.6 Hz, 2H), 2.65 (dt, J = 12.6, 8.0 Hz, 6H), 1.87–1.75 (m, 1H), 1.66 (dd, J = 13.0, 7.2 Hz, 6H), 1.45–1.19 (m, 17H), 0.99–0.81 (m, 6H), 0.10–0.03 (m,12H).

(Carbon-fluorine-proton decoupled) ¹³CNMR (101 MHZ, CDCl₃): δ 158.63, 148.63, 143.20, 143.12, 141.39, 138.36, 133.60, 132.11, 129.60, 128.89, 128.85, 128.81, 128.03, 126.66, 124.68, 114.83, 68.16, 35.83, 35.68, 31.71, 31.61, 31.49, 31.25, 29.88, 29.49, 29.21, 28.37, 23.29, 22.71, 22.64, 18.46, 14.20, 2.14, 1.18, 0.51.

¹⁹FNMR (376 MHz, CDCl₃): δ −143.12.

2.6. Synthesis of the Homopolymers (HP1 and PCN)

After a hydrosilylation reaction with poly(methylhydrosiloxane) and M1, using Karstedt’s catalyst, HP 1 and PCN were obtained. The synthesis is described in detail below (see Scheme 3).

2.6.1. PCN

PCN has been reported earlier [21,22]. The phase transition was found to be I-Sm_A 166.6 °C in this work, a value that is very close to the transition temperature reported previously of I-Sm_A 169.5 °C [21,22].

The polymer was synthesized by a hydrosilylation reaction of the monomer 4′-(pent-4-en-1-yloxy)-[1,1′-biphenyl]-4-carbonitrile (1.2 g, 4.56 mmol) and polymethylhydrosilane (Mn ≈ 1700~3200, 0.228 g, 3.8 mmol) using dry toluene as solvent with 2 drops of Karsted’s catalyst (Pt ~2%), and the solution was stirred at 60 °C for 3 days whilst under an atmosphere of nitrogen. After the reaction was completed, further purification was carried out by repeated precipitation into methanol several times, and the product was dried under vacuum.

Isolated yield: 0.25 g, 20.4%.

¹H-NMR (400 MHz, CDCl₃): δ 7.47 (s, 14H), 7.25–7.00 (m, 8H), 6.84 (s, 4H), 3.86 (s, 4H), 2.63 (s, 7H), 1.61 (t, 26H), 0.90 (s, 4H), 0.59 (s, 4H), 0.09 (d, 19H).

The calculation of the yield is listed in Figure S1.

2.6.2. HP1

2′,3′-difluoro-4-(7-(4′-(pent-4-en-1-yloxy)-[1,1′-biphenyl]-4-yl)heptyl)-4″-pentyl-1,1′:4′,1″-terphenyl M1 (0.15g, 0.224 mmol) in dry toluene (30 mL) was stirred with polymethylhydrosilane (Mn ≈ 1700~3200, 0.01g, 0.166 mmol) at 60 °C under nitrogen, and then 1 drop of Karsted’s catalyst (Pt ~2%) was added. The reaction mixture was stirred for 3 days. After the reaction finished, purification was carried out by repeated precipitations into methanol, and the product was dried under vacuum.

Yield: 0.031g, 25.6%.

¹HNMR (400 MHz, CDCl₃): δ 7.46 (t, 9H), 7.22 (s, 7H), 6.80 (s, 2H), δ 3.84 (s,2H), 2.62 (s, 8H), 1.33 (s, 17H), 1.27 (d, 18H), 0.97–0.76 (m, 13H), 0.23–0.02 (m, 48H) (as shown in Figure 3).

The calculation of the yield is listed in Figure S1.

2.7. Synthesis of the Copolymer (CoP1)

After the hydrosilylation reaction with poly(methylhydrosiloxane) and M1 using Karstedt’s catalyst, CoP1 was obtained. The synthesis is described in detail below (see Scheme 4).

CoP1 

The polymers were synthesized by the hydrosilylation of the monomer 4′-(pent-4-en-1-yloxy)-[1,1′-biphenyl]-4-carbonitrile (1.2 g, 4.56 mmol) and polymethylhydrosilane (Mn ≈ 1700~3200, 0.051 g, 0.85 mmol) using dry toluene as solvent with 2 drops of Karsted’s catalyst (Pt ~2%) and stirred at 60 °C for 3 days while kept under an atmosphere of nitrogen. After the reaction was finished, further purification was carried out by repeated precipitation into methanol several times, and the product was dried under vacuum.

Isolated yield: 0.102 g, 21.2%.

¹HNMR (400 MHz, CDCl₃) solvent: δ 7.71–7.30 (m,6H), 6.84 (s, 2H), 3.86 (s, 2H), 1.75 (s, 2H), 1.45 (s, 4H), 0.57 (s, 2H), 0.07 (dd, J = 13.9, 8.0 Hz, 4H).

The calculation of the yield is listed in the Figure S1.

The GPC results indicate that the polydispersity is relatively narrow and broadly like that of the polymer backbone before the reaction, for the homopolymer and copolymer (see

Table 1. GPC results of polymers.

Compound No	GPC Results			Weight Average DP Obtained by GPC	Average Molecular Weight of Repeating Units
	$\overline{\mathbf{M}\mathbf{w}}$	$\overline{\mathbf{M}\mathbf{n}}$	$\overline{\mathbf{M}\mathbf{w}}$$/\overline{\mathbf{M}\mathbf{n}}$	$\overline{\mathbf{M}\mathbf{w}}$$/{\mathbf{M}}_0$
Polysiloxane backbone	830	663	1.3	13.8	60.1
HP1	21,494	15,441	1.4	29.4	730.4
PCN	28,848	14,212	2.0	89.3	323.1
CoP1 (M1: MCN = 1.2: 0.8)	22,720	12,778	1.8	40	567.5

).

3. Results and Discussion

3.1. DSC Studies

To investigate the phase behavior of materials, DSC measurements were used, and the data for the second cooling and heating curves are listed below (

Table 2. Phase transition and phase transition enthalpies for monomer, pentamethyldisiloxane derivative and polymers.

Compound No	Spacer Length	Number of F Groups Attached	Transition Temperature °C (Using Onset and Offset of the Peaks), Measured by DSC at 10 °C/min, Enthalpies Unit (J/g)	Highest LC-I Temperature Peak Width
M1	5	2	Cr 87.5 Ntb 101.5 (10.1) N 113.0 (0.5) I	3.8
			I 114.2 (−0.5) N 102.5 (−10.2) Ntb 90.6 (−3.6) Smx 59 Cr	3.9
MCN	-	-	Cr 86.4 (93.7) I	-
			I 66.9 (−1.5) [N 49.3] Cr	1.5
PMDS 1	5	2	Cr 68.7 Smx 92.5 (1.8) SmA 110.5 (17.5) I	2.4
			I 108.6 (−17.7) SmA 90.4 (−1) Smx 45.3 Cr	2.5
HP1	5	2	Cr 107.1 Nx 182.9 (10.6) I	32.8
			I 174.1 (−10.4) Nx 106.8 Cr	29.6
PCN	5	0	SA 164.9 (5.2) I	9.5
			I 166.6 (−5.3) SA	4.9
CoP1 (M1: MCN = 1.2: 0.8)	5	2	N 141.39 (7.0) I	34.5
			I 148.3 (−5.9) N	38.6

Melting point and isotropisationtemperature are both measured as the onset temperature of the 2nd heating and cooling curves as measured by DSC. The melting points for polymer PCN and CoP1 cannot be determined as below the room temperature.

):

(a)

Cr = crystalline; I = isotropic; N_x = unknown nematic phase; Sm_x = unknown smectic phase.

(b)

Monotropic phases are shown in square brackets; enthalpies are shown in parentheses, unit (J/g).

(c)

Peak width was measured as the difference between onset and offset temperatures ($°\mathrm{C}$).

To facilitate the comparison, the thermal properties of monomer M1, the pentamethydisiloxane PMDS1 and polymers are given in Table 2. After polymerization, both melting points and isotropization temperature for the polymer increase when compared with the related monomer and the pentamethydisiloxane PMDS1. For the pentamethydisiloxane, the transition temperature and the liquid crystal phase behaviors can often provide very useful information for the prediction of the related homopolymers [22,23,24]. All the phase transitions are reduced for the pentamethydisiloxane PMDS1 when compared with the related monomer. Homopolymer 1 (HP1) shows crystallization without a glass transition. A route to eliminate the crystallization is through the introduction of another mesogenic monomer or even non-mesogenic units, to separate the liquid crystal side chains. For copolymer (CoP1), both crystallization tendencies and melting points are eliminated on the second heating and cooling curve of the DSC, and the liquid crystal phase is stable at room temperature. The isotropization temperature decreases for the copolymer, compared with homopolymer. It is noteworthy that not only can copolymerization help to eliminate the crystallization, but it can also help to stabilize the liquid crystal phase at room temperature. Based on the transition temperatures of HP1 and CoP1 (Table 2), the crystalline tendencies are eliminated in the copolymer because the incorporation of the short spacer length of the monomer 4′-(pent-4-en-1-yloxy)-[1,1′-biphenyl]-4-carbonitrile has broken the π-π interactions of the phenyl rings of the long side chains. In addition, due to the flexibility of the polymer backbones, the glass transition temperature is lower than room temperature.

An interesting and important feature is the width of the N-Iso transition on linking M1 either to a pentamethyldisiloxane end group or to the polysiloxane main chain. M1 is characterized by a transition width of 3.8–3.9 °C, somewhat wider than the monomeric material MCN known in the literature but in line with results reported for other dimesogenic dimers with lateral fluoro groups. Linking MCN to a polysiloxane main chain results in PCN, as well as a widening of the transition interval and an increase in the enthalpy, in line with the original reports [21,22]. Typically, the widening of transitions on attachment to polymers is linked to kinetic effects; molecular mobility is decreased due to increased viscosity, but even more so, it is linked to the influence of the polydispersity of the main chain. Polymer chains of differing lengths exhibit different transition temperatures. A polydisperse sample is thus an ensemble of a series of systems, transitioning at different temperatures, and, together with the ensuing biphasic regions associated with first-order phase transitions, is the cause of widening the transition intervals. Attachment of M1 to a pentamethyldisiloxane end group narrows the transition somewhat to 2.4–2.5 °C when compared to M1 with the vinylic end group of (3.8–3.9 °C), but the transition enthalpy increases dramatically to 17.5–17.7 J/g, very high for a liquid crystal to isotropic transition. This is indicative of a major order–disorder transformation. It suggests also that a different LC phase may have formed as the highest stable LC state. The attachment to the polysiloxane main chain has an equally dramatic effect. The transition to the monotropic low-temperature smectic phase of M1 is lost, and the transition enthalpy is reduced to 10.4–10.6 J/g when compared to the pentamethyldisiloxane derivative PMDS1, indicative of a lower ordering. However, the transition interval is enormous; it reaches 29.6–32.8 °C, depending on whether the values are collected on heating or cooling. This suggests that multiple processes take place. For M1, the nematic phase range is ~12 °C, and the N_tb phase has a range of about 30 °C. The extreme widening of the transition in HP1 would thus be explainable by a combination of the biphasic behavior observed already in mixtures of dimers. Additional phase transition peaks seem to appear in DSC experiments but are related to the effects of the end points of the biphasic regions of the phase diagram [25]. For the polymer, additionally, an overlap of the phase transitions of polymers of different chain lengths and transitions between different phases such as the N and the N_tb phase play a role. In other words, when on cooling, some of the polymer chains in this ensemble still undergo an Iso to N transition, others are already on the way to the N to N_tb transition. This broadens the transition range to the extreme. This view is supported by the results for CoP1. Even more molecular complexity is introduced; the range of the transition interval increases marginally to 34.6–38.6 °C, but the transition enthalpy associated with the loss of order is broadly similar to that for PCN and much smaller than observed for HP1. This confirms the argument that for HP1, a broad and flat peak masks multiple transitions in the polymer. However, as only one transition was detectable by DSC, the term N_x is used for the LC phase in Table 2.

3.2. POM Studies

The liquid crystal defect textures of M1, obtained from POM are shown below (see Figure 4a–f). The material forms a nematic phase at 114 $°\mathrm{C}$, the texture of the nematic phase is shown in Figure 4a, and a further phase transition takes place at 103 $°\mathrm{C}$ (shown in Figure 4b), close to the DSC measured value of 102.5 $°\mathrm{C}$ (the DSC scan was recorded at the rate of 10 °C/min). The texture changes indicate a transition from the nematic phase (see Figure 4a) to the twist-bend nematic phase (Figure 4c). The twist-bend nematic phase texture does not change under slight shearing (Figure 4d). After being sheared at 80 $°\mathrm{C}$, the texture does not change much but flows very slowly due to high viscosity (Figure 4e). If the sample is sheared more vigorously at 95 $°\mathrm{C}$, a texture change occurs, and only a homotropic texture is observed (Figure 4f). Texture changes associated with the phase transition at 89.9 $°\mathrm{C}$ could not be identified that clearly under POM; however, a more ordered smectic phase had formed noticeably in shearing experiments where a very high viscosity was noticed. It was also found that the cooling rate strongly influences the POM textures. When cooling at a high rate of 10 $°\mathrm{C}$/min, the texture shows no focal conic type of shape defects (Figure 5a). Reducing the cooling rate to 3 $°\mathrm{C}$/min induces small focal conic features (Figure 5b), further reducing the cooling rate to 0.1 $°\mathrm{C}$/min, results in larger focal conic domains (see Figure 5c).

It is interesting to analyze the liquid crystal phase formed by the pentamethydisiloxane functionalized monomer M1 that exhibits N_tb phase behavior. As shown in Figure 6a, PMDS1 exhibits an SmA phase that emerges from the isotropic state with atypical bãtonet defect texture [26]. The bãtonets merge to form the smectic phase texture (see Figure 6b). As the temperature is reduced, the bãtonets grow into mini-focal-conic domains. As the end group is a widely reported pentamethydisolane group, this is not surprising, as this type of siloxane end group can produce nano-segregation to form smectic phase behavior [27]. The short siloxane group disturbs the formation of the heliconical structure and favors layer formation. This results in an SmA phase.

To identify the phase texture of the HP1, contact preparations of DTC5C7 (see Figure 7) with homopolymer 1 (HP1) were carried out and monitored by POM. DTC5C7 was chosen as a contact compound for HP1. One reason is that DTC5C7 has broadly similar transition temperatures when compared to HP1; the other reason is that its N_tb phase transition temperature is located within the phase transition range of HP1. This makes it a good tool for identifying the existence of a N_tb phase in HP1. The results can be found in Figure 8 and Supporting Information Figure S1. On the left-hand side, DTC5C7 is shown in the micrographs, while on the right-hand side, the defect textures of HP1 can be seen. From the Supporting Information Figure S1a,b, it can be seen that as the temperature is reduced to 120 °C, the texture changes from the Nto the N_tb phase. After annealing the polymer for 1 day, the two mesophases’ textures began to merge (see Supporting Information Figure S1c). Further annealing for 2 days made the textures clearer, the whole area of DTC5C7 and HP1 completely merged and exhibited the same texture as shown in Supporting Information Figure S1f. Figure 8a,b, show the coalescence of both compounds at 120 °C over a wider viewing field. This result needs further confirmation; the current data, however, reveal that the N_tb phase of DTC5C7 and the unknown nematic phase, provisionally identified as N_x of HP1, mix fully. As both molecules are structurally very similar, this suggests that the N_x phase of the HP1 is a N_tb phase. This would be the first example of a N_tb found in a side-chain polymer.

3.3. X-ray Powder Diffraction

The results of X-ray powder diffraction analysis for the copolymer that exhibits one liquid crystal phase, which is stable at room temperature, are discussed. The X-ray powder diffraction was collected to gain more information on the phase structure (see Figure 9 and

Table 3. Summary of X-ray powder diffraction data of copolymer 1 (CoP1) at room temperature.

2ϴ (°)	d-Value (Å)
4.25	20.78
12.28	7.20
19.77	4.49

). The X-ray powder diffraction pattern is dominated by a wide-angle intensity at 2ϴ = 19.8° (d = 4.49 Å). This is related to the lateral packing of the fluidic mesogens in the liquid crystal state. The shoulder-like peak at 2ϴ = 12.3° (d = 7.2 Å) is often found for siloxane polymers. The small angle peak at 2ϴ = 4.25° (d = 20.8 Å) is typical for N_tb and N materials and is simply associated with the electron densities of the mesogenic units which are only orientationally ordered but do not have positional ordering [17]. This confirms that the overall phase structure is that of a nematic-like phase.

3.4. Phase Diagram

As it was found to be difficult to identify the phase structure of the monomer unambiguously, an exploratory phase diagram with defined mixtures was constructed, following the results of DSC experiments. The phase diagram of binary mixtures of CBC9CB. the structure is shown in Figure 10, and monomer M1 was constructed to identify the phase transitions of the N_tb phase.

CBC9CB was chosen as a contact compound for monomer M1. One reason is that it has a broad temperature range of the N_tb phase; the other reason is that its N_tb phase transition temperature is broadly similar to the phase transitions of M1. This makes it a good indicator of the existence of a N_tb phase in M1 [28].

The exploratory phase diagram for the mixtures of CBC9CB and monomer M1 shows the results from the second cooling DSC run (Figure 11a). The transition temperature from the nematic to the isotropic decreases somewhat linearly with an increasing amount of M1. Progressing from pure CBC9CB to pure monomer M1, the nematic phase range firstly becomes wider, then is reduced to the narrowest area at 50 wt% and finally grows slightly on the side of the monomer M1. The N_tb region for both pure compounds is wider until approaching ~50 wt%, which exhibits the narrowest range. Interestingly, for the mixtures, the N_tb phase is enantiotropic, independent of the composition change. For 25%wt and 75%wt mixtures, the N_tb phase is stable at room temperature, whereas for 50 wt%, the N_tb phase is only stable over a small temperature range and finally changes to a smectic phase whose structure is still to be determined.

Figure 11b shows the enthalpy values of the mixtures as a function of the concentration of monomer M1; the enthalpies are associated with the Iso-N and N-N_tb transition and are based on the measurements obtained from the second cooling runs of the DSC scans. They are plotted as a function of the concentration of monomer M1 in CBC9CB binary mixtures. The enthalpy values for the Iso-N transition exhibit a substantial decrease with an increasing molar fraction of the monomer M1. The values of the transition enthalpy of pure CBC9CB of 1.0 KJ/mol to 0.3 KJ/mol for pure monomer M1 are quite different. It is noted that the CBCnCB series exhibit different enthalpy values for the N-Iso transition compared to dimers containing fluoro groups in the mesogenic units [29]. The enthalpy values for the N-N_tb transition display a dramatic increase with an increasing concentration of monomer M1. With the value from the transition enthalpy of pure CBC9CB of 0.4 KJ/mol increasing to 6.8 KJ/mol for pure monomer M1, this is almost seventeen times larger in value. Figure 11c shows the DSC results of one of the mixtures 25 wt% M1–75 wt% CBC9CB as an example. In addition to the wide melting transitions, the exothermic peak related to the crystallization temperature was not observed by DSC, because the system forms a glassy liquid crystal state above 30 $°\mathrm{C}$. Similar results were observed by DSC on the second heating and cooling curves for all the other mixtures.

3.5. Polysiloxanes

For homopolymer 1 (HP1), there is only one liquid crystal phase recorded by DSC. However, it is noted that the peak is broad and unresolved, and thus a phase transition between the nematic and the N_tb phase might be obscured or simply not measurable due to a biphasic region. From Figure 8, it can be provisionally deduced that this is a N_tb phase since there is no boundary or miscibility gap between the binary mixtures of DTC5C7 and HP1 in the N_tb phase. POM pictures for HP1 depicted in Supporting Information Figure S2 show that this liquid crystal texture remains the same until crystallization at 107.9 $°\mathrm{C}$.

POM pictures for the copolymer CoP1 are also shown in Figure 12. Schlieren textures are clearer at temperatures close to isotropization temperature. Crystallization peaks were not traced by DSC, because the polymer remains in the liquid crystal phase at room temperature. This was already evidenced by X-ray powder diffraction experiments measured at room temperature (Figure 9). Recording pictures at higher magnification shows that branches between long rope-like features occur (see Figure 12e,f), and long annealing times reveal an oily streak-like texture which can often be observed for N_tb-forming materials. This might be interpreted as a variation of the rope texture that is found often in twist-bend systems.

The thermal behavior of a comparative mixture of M1 and MCN (molar ratio of 1.2:0.8, the same as in CoP1) was studied by POM. The POM data, for that mixture, show a nematic to smectic A phase transition on cooling at 105 °C (see Figure 13a–c). When these two monomers are linked to the polymer backbone, the smectic A phase is suppressed in CoP1, and the nematic phase remains. As observed for HP1, the phase transition range as measured by DSC is extremely wide, possibly masking a N-N_tb transition; this would be in line with POM observations. However, in the absence of more detailed studies, potentially with more complex behavior as a copolymer involved, the phase is provisionally identified as N_x.

4. Conclusions

A new vinylic dimesogenic monomer showing the N_tb phase was synthesized. Current synthetic methods were improved for achieving material containing an isomer-free olefin group. A pentamethyldisiloxane group was attached to this monomer, resulting in the SmA LC phase behavior of the ensuing material. Linking the monomer to a siloxane main chain results in a polymer, forming nematic phase behavior with an extremely wide transition range. Detailed studies with the N_tb-phase-forming dimer DTC5C7 show full miscibility of the dimer and the new LC polymer in all LC phases, suggesting that the side-chain LC polymer forms a N_tb phase as a low-temperature nematic phase. Copolymerizing the monomer with a cyanobiphenyl-based monomer allows us to tune the glass transition, prevent crystallization and tune the phase behavior further to a nematic phase behavior at room temperature.