Synthesis of novel cyclosiloxane monomers containing push–pull moieties and their anionic ring opening polymerization

The synthesis of three novel tetracyclosiloxane monomers modified either with a nitroaniline (NA) or with a Disperse Red 1 (DR1) push–pull group and their ring opening polymerization reaction in the presence of tetramethylammonium hydroxide are presented. The prepared monomers and polymers were characterized by different spectral methods and gel permeation chromatography. For the crystalline monomers, the structures were further proven by single crystal X-ray diffraction. Dynamic scanning calorimetry shows that the polymers that carry NA groups have a glass transition temperature (Tg) well below room temperature (RT), while the one that carries DR1 groups melts at 55 °C. The transition temperatures have a strong effect on permittivity as indicated by broadband impedance spectroscopy measurements conducted at different temperatures and frequencies. The polymers modified with NA groups have a high permittivity (maximum value of 17.3) at RT, suggesting the polar groups to be mobile and orientation polarization to be effective. However, the polar groups of the polymer modified with DR1 are frozen and thus cannot contribute to the permittivity via orientation polarization. Consequently, the permittivity is only 8.8 at RT, but increases to 22 above the melting temperature, where the dipoles are mobile. Because of the high dielectric permittivity and rather low Tg, the polymers modified with NA are attractive as active dielectric materials in actuators, capacitors, and stretchable electronics, whereas the polymer modified with DR1 may be of interest in nonlinear optical devices.


Introduction
Push-pull molecules such as nitroaniline (NA) and disperse red 1 (DR1) have a large dipole moment and are oen used as active components in nonlinear optic (NLO) polymer devices to increase the NLO coefficient 1-4 as well as in electromechanical transducers to increase the polarizability and thus the dielectric permittivity (3 0 ). 5 In nonlinear optical devices, the active polymers are non-centrosymmetric and have the push-pull moieties aligned in the same direction and in a frozen-in state. Such orientation is achieved via the poling in which the initially randomly oriented dipoles of a polymer are poled in a strong electric eld above the glass transition temperature (T g ). 6 By cooling the sample below T g while maintaining the electric eld, the oriented structure is frozen and the poled polymer shows NLO properties. For this application, polymers with a high T g and high density of push-pull moieties are required.
For the second application, however, these moieties have to be mobile to be able to contribute to the dielectric permittivity via orientation polarization. Orientation polarization is effective for low T g polymers which have permanent dipoles incorporated. When a polar, low T g polymer is exposed to an electric eld, the dipoles align in a certain time along the electric eld direction. Depending on the size of the dipole used, they may contribute to the permittivity beyond a certain frequency. [7][8][9][10][11][12][13][14][15] Push-pull groups can be introduced on polysiloxane chain using a post-polymerization modication reaction. This has been achieved starting from either azide or hydrosilyl containing polysiloxanes. 16 The azides were subsequently used to introduce nitrobenzene groups via aza-click chemistry, 17 while the hydrosilyl groups were used in hydrosilylation reactions with push-pull compounds that carry a reactive double bond. Using this strategy, Yang and Wnek introduced silyl ketene acetal pendant groups which were subsequently reacted with 4-nitrobenzenesulphenyl chloride. 18,19 Kussmaul et al. graed Nallyl-N-methylaniline to the Si-H groups of a poly(methylhydroco-dimethyl)siloxane in an in situ process optimized to produce elastomers, but only a rather small amount of push-pull dipoles was incorporated as reected by the rather low permittivity of about 3 0 ¼ 6 achieved for their materials. [20][21][22] Racles et al. introduced DR1 groups on a polysiloxane chain using a condensation reaction. 23 Unfortunately, in all cases the amount of polar groups incorporated was rather low and therefore the increase in permittivity was only moderate. To the best of our knowledge, no cyclosiloxane monomers exist that carry push-pull dipoles. Such monomers would provide access to the corresponding polysiloxanes combining high content of polarizable groups with main chain exibility. The current work therefore describes how some of these monomers can be synthesized and converted into polymers. We further present the glass transition temperatures of these polymers and their dielectric permittivity at different temperatures and frequencies. Depending on their particular transition temperatures, these novel polymers may nd application as dielectric in mechanical actuators or as active component in non-linear optical devices.

Results and discussion
The synthetic strategy to the cyclosiloxanes monomers modied either with NA or DR1 push-pull moieties is illustrated in Scheme 1. They were prepared via a hydrosilylation reaction of heptamethylcyclotetrasiloxane (D 4 H) with NA or DR1 modied with allyl or methacryl groups. For this, N-methyl-p-nitroaniline 1 was reacted with allyl alcohol to form 2 which carries a reactive allyl group. This group was subsequently used in a hydrosilylation reaction with D 4 H to give the tetracyclosiloxane monomer 3 that carries a nitroaniline group. Fig. S1 † shows the 1 H NMR spectra of the two starting compounds D 4 H and 2 and of the product 3. The disappearance of the hydrosilyl group at 4.7 ppm of D 4 H and of the allyl group between 5 to 6 ppm and the appearance of new signals in the aliphatic part of 3 are clear indicators of the successful hydrosilylation. The other two monomers, 6 and 9, were prepared via a hydrosilylation reaction of D 4 H either with NA or DR1 moieties 5 or 8 that were modied with a methacryl group. While monomers 3 and 6 formed in rather high yields, monomer 9 modied with DR1 was obtained only in a moderate yield of 30%. The monomers were puried by column chromatography over silica gel and they were isolated as viscous oils. They were pure compounds as proven by their elemental analysis and 1 H NMR spectra (Fig. 1). Compound 3 and 9 slowly solidied at room temperature (RT).
Single crystals of 3 and 9 were grown from ethanol and their solid state structure was established by X-ray diffraction. Both monomers have a molecular crystal structure with P 1 space group of triclinic system. The asymmetric part of the unit cell comprises one discrete molecule as shown in Fig. 2 and 3. The bond lengths and angles are summarized in Table 1S (ESI †). Analysis of the packing diagrams indicates a close similarity of compounds 3 and 9. In the crystal, the asymmetric units are linked through weak C-H/O hydrogen bonds and p-p stacking interactions. The main crystal structure motif can be characterized as a parallel packing of innite supramolecular ribbons along a axis, as shown in Fig. 2 and 3.
The ring opening polymerization of 3, 6, and 9 as well as the copolymerization of 3 with octamethylcyclosiloxane (D 4 ) was performed. Polymerization of cyclosiloxanes can be conducted in cationic as well as in anionic conditions, however due to the presence of the amino groups in the monomers, which can interfere with the cationic initiator, it was decided to conduct anionic ring-opening polymerization. 24 As initiator tetramethylammonium hydroxide (TMAH) was used, since it allows for a facile purication of the polymers. This also enabled to introduce silanol end-groups which we intended to use further for cross-linking in thin lms to elastomers. 25 All reactions were conducted at 100 C and monomers 3, 6, and 9 polymerized to highly viscous liquids P-NA, P-MNA, and P-MDR1, respectively. For the copolymerization of 3 different  ratios of monomers D 4 : 3 (x : y) were used: 2 : 1, 1 : 1, and 1 : 2 (see Table 1) to furnish copolymers co-P-NA(x:y) with different contents of NA. With increasing the amount of 3, an increase in the signals of the aromatic part, characteristic for the NA moiety, was observed. The 1 H NMR spectra of the polymers were broader as compared to the starting monomer. The molecular weight and dispersity were determined by GPC using tetrahydrofuran solvent and polydimethylsiloxane standards (Table 1). For the copolymers a molar mass increase was noted with increasing content of D 4 in the copolymerization mixture. The polymerization of 3 afforded polymers of a higher M w as compared to the one obtained from 6 and 9. All polymers showed a rather broad PDI, characteristic for polysiloxanes obtained under thermodynamic equilibrium. Slight deviations of the theoretical content of the push-pull moieties and the one found by 1 H NMR was observed (Table 1).
Thermogravimetric analysis shows that all polymers are stable up to 300 C, where they start to decompose (see ESI †). Differential scanning calorimetry was used to nd the phase transition of monomers and polymers. Monomer 3 has a T m ¼ 112 C, while the polymerization of 3 allowed formation of P-NA with a T g ¼ À33 C. As expected with increasing the content of dimethylsiloxy units in the copolymers co-P-NA, a decrease in the T g was observed ( Table 2). Monomer 6 has a larger exible chain between the NA moiety and the tetracyclosiloxane ring as compared to 3. Monomer 6 is therefore liquid at RT (T m ¼ À40 C). The T g of P-MNA was however the same as the one of P-NA. Monomer 9 was solid at RT (T m ¼ 44.9 C). Interestingly, while the polymerization of 3, which is solid at RT, allowed formation of a polymer with a rather low T g , the ring opening polymerization of 9, which is also solid, gave a polymer with a T m ¼ 55.4 C, likely due to stronger dipolar interactions. While this T m may be too low for some NLO properties applications, by increasing the density of DR1 dipoles in the polymer it should be possible to further increase the T m .
To nd how the dielectric properties of the prepared polymers are affected by the presence of push-pull groups, dielectric spectroscopy measurements were performed at different temperatures and frequencies. The measured dielectric response contains in all cases a dipolar and a conductivity contribution from mobile ions. 26 Therefore, we modelled the data using the following expression for the complex permittivity:  The background permittivity (3 s ), the dipolar enhancement (D3), the conductivity (s), the characteristic dipolar activation frequency (u 0 ), and the characteristic exponents (a and b) of the dipolar loss peak were extracted from the data t at every temperature and their temperature dependence is shown in Fig. 4 and ESI. † To ensure the correctness of the t, we also extracted the parameters from ts to the electric modulus and veried that both results are consistent. Here the parameters in the temperature range where the ts were conclusive are presented. For instance, the data at temperatures below 30 C could not be analyzed as the signal is dominated by the nearly constant loss from caged ions. At the highest temperatures, the loss peak falls out of the frequency range and therefore neither the characteristic frequency nor the characteristic exponents can be precisely determined.
The dielectric enhancement shows a consistent increase with the concentration of dipoles and a nearly (1/T) dependence on the temperature, with the exception of the P-NA and P-MDR1 which have the highest dipoles concentration and/or melting point above RT. The ratios of the enhancements among co-P-NA(1:2), co-P-NA(1:1), and co-P-NA(2:1) follow closely the ratios of the corresponding concentrations. Thus, with increasing the content of NA in (co)polymers an increase in the permittivity was observed from 7.5 for co-P-NA(2:1), to 12 for co-P-NA(1:2), and reached a maximum value of 17.3 for P-NA which had the maximum content of NA (Fig. 5, Table 3).
Polymers P-NA and P-MNA contain the same NA push-pull group, but their push-pull groups are slightly differently connected to the polymer chain. Because of this, the concentration of push-pull moieties in P-MNA is slightly lower as compared to P-NA, which is reected by a slightly lower value for the dielectric permittivity at high frequencies. For example P-MNA has a 3 0 ¼ 15.8, whereas P-NA has a 3 0 ¼ 17.3 (Fig. 5). Although the dipole moment of DR1 is higher as compared to NA and one would expect that the polymer modied with DR1 to have the highest permittivity, this was not the case. The reason behind this is the T m ¼ 55.4 C of this polymer. Therefore, at RT, the push-pull dipoles are frozen and the orientation of the dipoles in an alternative electric eld is not possible at this temperature. However, when heated to temperatures above T m , where the dipoles are mobile and can contribute to orientation polarization, an increase in permittivity was observed to 3 0 ¼ 21. Thus, the behavior of P-MDR1 is anomalous due to the strong link among the dipoles in the solid state. P-NA shows the inuence of the glass transition as the temperature approaches À30 C. Consistent with this, the characteristic exponents of the relaxation peaks are systematically smaller for P-MDR1 and P-NA than for the other samples. For an ideal system of noninteracting dipoles a ¼ b ¼ 1. As the correlations between dipoles become stronger, the exponent 0 < b < 1 takes smaller values. Values of 0 < a < 1 indicate strongly correlated (glassy) dynamics with power-law (slow) relaxation in time.
Both a and b increase as the temperature is increased (see ESI †), as expected from the loss of correlation when interaction becomes less relevant in comparison with thermal uctuations.
The temperature dependence of the characteristic frequency of the loss peak for all samples is shown in Fig. 6. A systematic shiing to lower frequencies with the increase in the dipoledipole interactions is observed. Characteristic conductivity relaxation times s ¼ (3 s + D3)3 0 /s, and values of conductivity versus temperature as well as separate plots of s and u 0 for each sample are shown for completeness in ESI. † Looking at the general trend we conclude that although higher concentrations of dipoles may lead to stronger dielectric enhancement, it may also cause a slower dynamics and loss of dielectric response due to stronger interactions. In this respect, P-MNA was found to be an interesting candidate for applications as actuator as it has a relatively large D3, similar to P-  MDR1 and P-NA, but also retains larger values for a and b. This performance could be related to the effective dilution of the dipoles due to the longer chain used for the attachment to the backbone. Although P-MNA has the largest ion conductivity of the set investigated, at high temperatures, it also displays the strongest temperature dependence resulting in ion conductivity similar to those of the P-NA and co-P-NA(x:y) samples at room and lower temperatures. A direct comparison of the conductivity of the polymers reported here and other high permittivity silicones reported in the literature is not possible, as the conductivity is inuenced in a complex way by several factors. The currently highest dielectric permittivity values were reported for silicone based polymers with either nitrile (3 0 ¼ 18.2, T g ¼ À52.6 C) or methylsulfone (3 0 ¼ 22.7, T g ¼ 19.2 C) groups at every siloxy unit. 5 The polymers reported here carry polar groups at every fourth siloxy unit. Some of them have a T g < RT and therefore a high permittivity at RT. For example, polymers P-NA and P-MNA have 3 0 ¼ 17.3 and 3 0 ¼ 15.8, respectively. While these values are slightly lower than the reported ones, the rather low T g of P-NA and P-MNA and the possibility of introducing even more polar push-pull groups to the polysiloxane chain offer interesting options. Optimizing these factors will open access to polymers with even higher permittivity or with increased NLO coefficient. Supposed the T g of such polymers is below 37 C, applications in dielectric elastomer actuators for muscle replacement applications can be envisioned.
Thermogravimetric analysis (TGA) was conducted with a Perkin Elmer TGA7 at a heating rate of 20 C min À1 under a nitrogen gas ow. Differential scanning calorimetry (DSC) investigations were undertaken on a Pyris Diamond DSC (Perkin Elmer USA) instrument under a nitrogen ow.
The dielectric permittivity was measured using a Novocontrol Alpha-A Frequency Analyzer equipped with a temperature controller. The samples were placed between two golden plated stainless steel electrodes (diameter 20 mm) and the distance between electrodes was adjusted to 100 mm using spacers.
X-ray diffraction measurements for 3 and 9 were carried out with an Oxford-Diffraction XCALIBUR E CCD diffractometer equipped with graphite-monochromated MoKa radiation. Single crystals were positioned at 40 mm from the detector and 365, and 468 frames were measured each for 15, and 50 s over 1 scan width for 3, and 9, respectively. The unit cell determination and data integration were carried out using the CrysAlis package of Oxford Diffraction. 27 The structures were solved by direct methods using Olex2 and rened by full-matrix leastsquares on F with SHELXL-97 using an anisotropic model for non-hydrogen atoms. 28,29 All H atoms were introduced in idealized positions (d CH ¼ 0.96Å) using the riding model with their isotropic displacement parameters xed at 120% of their riding atom. The positional parameters parts in 9 rened using available tools (PART, DFIX, and SADI) of SHELXL97. The molecular plots were obtained using the Olex2 program. The structure of 9 was rened to a relatively low level solely due to a poor quality of the crystals. CCDC -1811155 (3) and CCDC -1811156 (9) contain the supplementary crystallographic data for this contribution. † Synthesis of N-allyl-N-methyl-4-nitroaniline (2). The synthesis of N-allyl-N-methyl-4-nitroaniline was performed according to the literature with slight modications. 20 N-Methyl-p-nitroaniline 1 (1 g, 6.57 mmol), allyl alcohol (0.8 ml, 11.76 mmol), palladium acetate (14.7 mg, 0.066 mmol), triphenyl phosphine (69 mg, 0.263 mmol), and titanium isopropoxide (80 ml, 1.68 mmol) were dissolved in 80 ml dry toluene. The reaction mixture was evacuated, ushed with nitrogen, and heated for 5 h at 115 C. The water was removed by azeotropic distillation. The solution was ltered, the solvent was removed, and the residue was chromatographed using hexane/ethyl acetate (4 : 1) as eluent. Compound 2 was isolated in a yield of 96% as orange-yellowish oil which crystalized under cooling. 1   Synthesis of P-NA and co-P-NA(x:y). A solution of tetramethylammonium hydroxide in methanol 25% (TMAH) (1.5 ml) was dried. To the dry TMAH different molar ratios of monomers 3 : D 4 were added (1 : 0; 1 : 0.5; 1 : 1; 0.5 : 1). The ask was immersed in a preheated oil bath (104 C) and stirred for 6 h. Then, the temperature was raised to 140 C for 1 h to decompose the initiator. The reaction mixture was washed several times with methanol and precipitated twice from THF with methanol in order to remove unreacted cycles and oligomers. .09 (C Ar (g)); 126.23 (C Ar (h)); 136.70 (C Ar -NO 2 ); 153.33 (C Ar -N); Anal. calcd for C 16.5  2-(Methyl(4-nitrophenyl)amino)ethyl methacrylate (5). To an ice-cooled solution of 4 (0.03 mol) and distilled triethylamine (0.05 mol) in dry dichloromethane (27 ml) freshly distilled methacryloyl chloride (0.04 mol) was added slowly under argon atmosphere. The ice bath was removed one hour aer methacryloyl chloride was added and the solution was le to react overnight. It was then washed with brine. The organic phase was separated and the solvent was removed. The residue was puried by column chromatography on silica gel using ethyl acetate: heptane (1 : 4) as eluents, to give 5 in over 90% yield. 1 (6). A solution of D 4 H (5 g, 17.67 mmol), 5 (20 mol% excess to the D 4 H) and 1 wt% Karstedt's catalyst in dry toluene (50 ml) was heated to 110 C overnight under argon. The reaction mixture was concentrated and the residue was puried by column chromatography (heptane : ethyl acetate ¼ 4 : 1) to obtain an organic yellow oil (6) in 85% yield. 1 (8). To an ice-cooled solution of N-ethyl-N-(2hydroxyethyl)-4-(4-nitrophenylazo)aniline 7 (0.03 mol) and distilled triethylamine (0.05 mol) in dry dichloromethane (27 ml) freshly distilled methacryloyl chloride (0.04 mol) was added slowly under argon atmosphere. The ice bath was removed one hour aer methacryloyl chloride was added and the solution was le to react overnight. It was then washed with brine. The organic phase was separated and the solvent was removed. The residue was puried by column chromatography on silica gel using ethyl acetate : heptane (1 : 4) as eluents, to give 8 in over 90% yield. 1