Hetero-functionalization of polyitaconates for developing improved polymer dielectrics: Merging sulfones with bulky/rigid cycles

Reactive and Functional Polymers 196 (2024) 105842 Available online 1 February 2024 1381-5148/© 2024 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Hetero-functionalization of polyitaconates for developing improved polymer dielectrics: Merging sulfones with bulky/rigid cycles Sebastian Bonardd a , b , * , Jon Maiz a , c , Angel Alegría a , b , Jos´ e A. Pomposo a , b , c , Ester Verde Sesto a , c , Galder Kortaberria d , David Díaz Díaz e , f a Centro de Física de Materiales (CSIC, UPV/EHU) and Materials Physics Center MPC, Paseo Manuel de Lardiz´ abal 5, E-20018 Donostia, Spain b Departamento de Polímeros y Materiales Avanzados: Física, Química y Tecnología, University of the Basque Country (UPV/EHU), Paseo Manuel de Lardiz´ abal 3, E20018 Donostia, Spain c IKERBASQUE – Basque Foundation for Science, Plaza de Euskadi 5, 48009 Bilbao, Spain d Materials +Technologies" Group, Chemical and Environmental Engineering Department, UPV/EHU, Plaza Europa 1, 20018 Donostia, Spain e Departamento de Química Org´ anica, Universidad de La Laguna, Avda. Astrofísico Francisco S´ anchez 3, La Laguna 38206, Tenerife, Spain f Instituto Universitario de Bio-Org´ anica Antonio Gonz´ alez, Universidad de La Laguna, Avda. Astrofísico Francisco S´ anchez 2, La Laguna 38206, Tenerife, Spain ARTICLE INFO Keywords: Polymer dielectrics Dipolar glass polymers Polyitaconates Energy storage ABSTRACT This work addresses the preparation of new polymer dielectrics working under the dipolar glass polymer (DGP) concept. Herein, we report the synthesis and characterization of four hetero-functionalized polyitaconates bearing sulfones as high dipole moment entities and norbornane or adamantane structures responsible for increasing the glass transition temperature (T g ) and thus the range of temperatures where they can work without degrading or showing high energy dissipation. As a result, all obtained polymers exhibited dielectric constants ( ε r ’) between 5.1 and 6.2 while presenting low loss factors (Tan(δ) <0.01), reaching the status of high-dielectric polymers with a low dissipative behavior. In addition, it was demonstrated that including bulky structures into their polymer backbone allows for an increase of up to 80 ◦C in their working temperature ranges, expanding the temperature range where they behave as DGPs in an outstanding manner. A complete structural, thermal and dielectric characterization was carried out in terms of infrared spectroscopy (FTIR), nuclear magnetic resonance (NMR), thermogravimetry (TGA), differential scanning calorimetry (DSC) and broad dielectric spectroscopy (BDS). Overall, these materials seem to fulfill the basic requirements to be considered good candidates for dielectric applications such as energy storage, confirming the versatility of polyitaconate-based materials. 1. Introduction In the last decade, all-polymer dielectrics have garnered increasing attention from the scientific community, resulting in an abrupt increase research interest [1–8]. As a consequence, the amount and diversity of new polymer materials with enhanced dielectric properties have shown a remarkable increase. This allows not only the development of new materials but also enriches our understanding of the topic from a theoretical perspective, which, in any case, reinforces the former [9–11]. Arguably, the sudden interest in the field can be attributed to the energy situation our current society is facing. Without changing our consumption habits and the sources from which we obtain energy, an upcoming energy shortage is inevitable [12]. Therefore, it is imperative to grasp that the development of efficient and sustainable energy acquisition processes from clean and renewable sources would be in vain if, once obtained, we are unable to utilize it effectively. To ensure this, one approach involves creating and/or reinforcing technologies that focus on efficient storage, conversion and distribution of obtained energy. This is where polymer dielectrics emerge as attractive candidates to achieve and promote a more efficient and effective energy storage process, particularly through the fabrication of improved capacitor devices [6,13–15]. A capacitor allows energy storage in the form of electrostatic energy that, subsequently, can be released as electrical discharge. These devices find application in a wide spectrum of areas ranging from trivial electrical circuits (such as household appliances and defibrillators, among * Corresponding author at: Centro de Física de Materiales (CSIC, UPV/EHU) and Materials Physics Center MPC, Paseo Manuel de Lardiz´ abal 5, E-20018 Donostia, Spain. E-mail address: [email protected] (S. Bonardd). Contents lists available at ScienceDirect Reactive and Functional Polymers journal homepage: www.elsevier.com/locate/react https://doi.org/10.1016/j.reactfunctpolym.2024.105842 Received 20 November 2023; Received in revised form 25 January 2024; Accepted 26 January 2024 Reactive and Functional Polymers 196 (2024) 105842 2 others) to other more complex and advanced ones such as power pulse applications, hybrid/electric vehicles, aerospace technology, robotics, (bio)sensors, rectifiers and accumulators in wind and solar farms [3,6,16]. The heart of a capacitor is, precisely, the dielectric material, which is intimately related to the amount of charge (energy) that the device is able to store. Therefore, physics has taught us that the maximum energy stored that a capacitor could sustain follows the relation U e =½( ε r ’E b 2 ), where ε r ’ and E b represent the dielectric constant and dielectric breakdown, respectively [3]. Both parameters are directly related to the dielectric material placed in the capacitor, evidencing the need to employ materials with the highest possible ε r ’ and E b values. Regarding the above, using polymers as dielectrics presents advantages and disadvantages when compared, for example, to inorganic systems. Among the advantages, probably the most relevant, is that polymers offer notably higher E b values as dielectric media, in addition to their better mechanical properties, light weight, easy processability and manufacture, synthetic versatility and low-cost production. However, they lack high ε r ’ values, usually falling within the range of 2–4 [3,17]. Unfortunately, E b – which is the one contributing more from a mathematical point of view – has shown a tighter relationship with the inherent nature of the material, making it difficult to modify while on the contrary, ε r ’ has been shown to be more tunable by manipulating the chemistry of the material. Taking into account the above, it has been demonstrated that by endowing a material with a greater polarization ability, its ε r ’ can be augmented outstandingly [18,19]. Therefore, searching for new strategies to afford the preparation of polymers with ε r ’ above conventional values but also presenting a low dissipative behavior has become a demanded and fruitful research field. Regarding the latter, to date, one of the most effective and promising strategies for developing new polymer dielectrics with high ε r ’ and low dissipation behavior is the manufacture of dipolar glass polymers (DGPs). This concept, introduced by Lei Zhu [17] almost ten years ago, is referred to as polymeric materials fulfilling two main requirements: 1) to contain small-sized high dipole moment functional groups covalently attached to its structure and 2) to be an amorphous material with a glass transition temperature (T g ) as high as possible [20]. Therefore, the polarization mechanism through which DGPs base their dielectric properties is mainly governed by the orientational motions that these dipolar entities can perform within the amorphous matrix when exposed to an external electric field. In this context, the need for an amorphous structure is driven by the inherent free volume present in these materials. This, combined with the requirement for small-sized dipole moment functional groups, creates an environment conducive to unhindered orientational motions. Equally crucial is the demand for a high T g , ensuring the material’s resilience under the typically heated conditions experienced by dielectric materials. A high T g value serves a dual purpose. First, it prevents the material from entering a rubbery state that could lead to leakage and subsequent failure. Second, it delays the onset of dissipative phenomena on the temperature scale, directly influencing the energy storage capacity of the system. Among various dissipative phenomena, the rise in ionic conductivity near T g is particularly detrimental, as it impacts the insulation property of the dielectric material [18,20,21]. This is a consequence of diffusion processes facilitated by ionic impurities, which are inherently present in these polymeric materials and become activated as the polymer chains undergo long segmental motions, typically occurring when the material approaches its rubbery state. When a DGP surpasses its T g , it transitions into a paraelectric state characterized by higher polarization but also a highly dissipative behavior. Ideally, in a DGP, one seeks a material with dipolar structures capable of orientational motions from the lowest possible temperature while also possessing the highest attainable T g value. This would result in a material exhibiting a wide temperature range with a highly polarized state, devoid of significant dissipative processes. This temperature range, often referred to as the DGP’s `sweet spot´, represents the regime where the paraelectric state dominates [22]. The large number of reports within the last five years demonstrates the success of DGPs as a strategy for the preparation of new all-polymer dielectrics with high ε r ’ values (>4) [8,23–26]. In this sense, diverse polymer structures ranging from aromatic to aliphatic and from synthetic to bio-based systems have been designed and prepared to contain highly dipolar functionalities and achieve superior dielectric properties [4,26–32]. Among these functionalities, nitriles and sulfones have been the ones that have attracted the most attention due to their high dipole moments (3.9 and 4.25 Debye, respectively) and easy incorporation [17]. However, in 2015, Wei et al. set an unprecedented result in the field of DGPs, where a new sulfonyl-containing polymethacrylate achieved an ε r ’ value of 11.2 and a loss factor – Tan(δ) – of 0.02 at room temperature [20]. Since then, sulfones have taken the lead as dipolar entities for excellence in developing new polymer materials with improved dielectric properties, as can be verified in the following references to name a few [22,33–41]. Of course, it is evident that one of the paths to follow in seeking new and improved DGPs is the design and testing of new polymer structures decorated with these molecular dipoles. In this context, and fortunately, some of us were able to contribute to this field by introducing, for the first time, polyitaconates as dielectric media [34,42,43]. One of the main attractiveness of these polymers is the double functionality of their monomeric units – derived from itaconic acid – that permits playing not only with the type of substituent but also with the number, in this case, of dipoles. Thus, in a series of consecutive works, we have successfully proven that these materials can act as valuable dielectric media by exhibiting high ε r ’ values (from 4 to 16) while presenting low Tan(δ) values (<0.02) after the chemical incorporation of nitriles and sulfonyl groups. More importantly, we have also demonstrated that the incorporation of bulky pendant groups into dipole-containing polyitaconates turned out to be a suitable strategy to dramatically increase the T g of these systems and, consequently, the temperature range where these behave as DGPs. Thus, with the aim of continuing to explore the performance of polyitaconates as dielectric media, this work presents a novel approach to achieve an improved version of our previous report dealing with the use of dipolar polyitaconates bearing bulky substituents [43]. In that opportunity, we successfully prepared of three new polyitaconates hetero-functionalized that had, on the one hand, norbornane or adamantane pendant groups and, on the other hand, nitriles as dipolar entities. Even when these materials exhibited notably high T g s, broad DGP temperature ranges and, therefore, low dissipative behavior (Tan <0.01), their ε r ’ values – between 4.0 and 4.3 – barely surpassed the upper limit established for conventional polymers. Therefore, to reach higher ε r ’ values and inspired by the superior performance shown by sulfonyl-containing polymers, in this work we carried out for the first time the replacement of nitriles by sulfone units in four new polyitaconates having norbornane and adamantane derivatives as secondary pendant groups. These materials were successfully prepared through the polymerization of their respective monomeric units and characterized in terms of their structural, thermal, and dielectric properties. 2. Experimental section 2.1. Materials Itaconic acid (≥99%, CAS 97–65-4), acetyl chloride (≥99%, CAS 75–36-5), Exo-norborneol (98%, CAS 497–37-0), 2-Norbornanemethanol (mixture of endo and exo, 97%, CAS 5240-72-2), 1-Adamantanemethanol (99%, CAS 770–71-8), 1-Adamantanol (99%, CAS 768–95-6), N,N ′ -dicyclohexylcarbodiimide (DCC, 98%, CAS 538–75-0), and 4-(dimethylamino)pyridine (DMAP, ≥99%, CAS 1122-58-3) were all purchased from Sigma–Aldrich. 2-(Methylsulfonyl)ethanol (98%, CAS 15205–66-0) was acquired from Fluorochem. 2,2 ′ -Azobis(2-methylpropionitrile) (AIBN, 98%, CAS 78–67-1) was purchased from Fluka. All solvents used in this work were purchased from Merck and used without further purification unless stated. S. Bonardd et al. Reactive and Functional Polymers 196 (2024) 105842 3 2.2. Characterization techniques 1 H NMR and 13 C NMR spectra were recorded on a Bruker Avance II500 system using CDCl 3 and DMSO‑d 6 as solvents for monomers and polymers, respectively. FTIR spectra of specimens were recorded in a Cary 630 FTIR spectrophotometer (Agilent) equipped with an ATR accessory between 650 and 4000 cm −1 after 34 scans in transmission mode and with a resolution of 2 cm −1 . The molecular weight of polymers was determined by size-exclusion chromatography (SEC) using a Shimadzu Nexera 40 HPLC system, equipped with a polar gel-M guard (50 ×7.5 mm) and a polar gel-M column (300 ×7.5 mm), both from Agilent. A differential refractive index (dRI) detector (RID-20 A, Shimadzu) was used as a detection system. A solution of LiBr in N,N-dimethylformamide (DMF) (0.1 wt%) was used as the eluent (1 mL/min flow rate), and polystyrene (PS) standards were employed for conventional calibration. The density values of the polymer films were calculated using a toploading electronic METTLER balance (ME-33360) and methanol (density 0.792 g/cm 3 at room temperature). The thermal characterization of samples was carried out using a simultaneous thermal analyzer (TG/DSC) Discovery SDT 650 (TA Instruments) and a differential scanning calorimeter Discovery DSC 025 (TA Instruments). Thermogravimetric analysis (TGA) was performed between 25 ◦C and 800 ◦C at 10 ◦C/min under a nitrogen atmosphere. From the obtained thermograms, the degradation onset temperature (T i ), maximum weight loss rate temperature (T MD ) and residue percentage (R) were calculated. T i is defined as the temperature at which the material loses 5% of its initial weight. On the other hand, differential scanning calorimetry (DSC) analysis involved the following five consecutive steps: i) dynamic heating from 25 ◦C to 200 ◦C at 20◦/min, ii) isothermal process at 200 ◦C for 3 min, iii) dynamic cooling process going from 200 ◦C to 0 ◦C at 20 ◦C/min, iv) isothermal process at 0 ◦C for 3 min, and finally, v) a dynamic heating process from 0 ◦C to 200 ◦C at 10◦/min. The informed glass transition temperatures (T g ) were calculated from the inflection point detected in thermograms acquired from the last heating step. Broadband dielectric spectroscopy was used to perform the dielectric characterization of the materials. Measurements were carried out using a Novocontrol Alpha high-resolution analyzer with an applied AC voltage of 1.0 V, from which frequency scans (isothermal spectra) and temperature scans (isochronal spectra) were obtained. Frequency scans were measured in the range between 10 −1 and 10 6 Hz. These measurements were taken at different temperatures ranging from – 150 ◦C to 25 ◦C. On the other hand, isochronal spectra were recorded from – 150 ◦C to 150 ◦C at frequencies between 1 and 10 6 Hz. The sample preparation for dielectric measurements is outlined below. First, a certain amount of polymer powder is hot-pressed between two metal discs at a temperature T g +30 ◦C and 2 tons of pressure, achieving the formation of a bubble-free film. Later, this film was hot-pressed again between two gold-plated electrodes of different diameters (20 mm and 30 mm), using the electrode with the greater diameter as the bottom part in the sandwich configuration. Initially, the thickness of the film was determined as the difference in thickness between both electrodes before and after hot pressing the polymer film. However, the above value was corrected by measuring the thickness of the peeled-off film from the electrodes after the dielectric measurements. It is worth mentioning that samples placed between electrodes were dried overnight under vacuum at 130 ◦C and then stored in a desiccator before analysis. Relative permittivity values (ԑ r ’) were calculated using eq. 1: ԑ ′ r=C’ Pd ԑO π r2(1) where ԑ r ’ is the calculated dielectric constant, C p ’ is the real part of the measured capacitance in the parallel configuration, d is the thickness of the polymer film, ԑ o is the vacuum permittivity (8.85 ×10 −12 F m −1 ) and r is the radius of the top electrode. Once calculated ԑ r ’, the dielectric loss (ԑ r ”) parameter was calculated using the relation ԑ r ” =ԑ r ’ Tan(δ), where Tan(δ) is the loss factor obtained directly from the experimental measurements. A Metricon model 2010 prism coupler was used to determine the refraction index (n) of the polymer films after dielectric measurements. Values of 1.5285, 1.5292, 1.5298, and 1.5304 were obtained for PSO 2 mNORI, PSO 2 NORI, PSO 2 mADAI and PSO 2 ADAI, respectively. 2.3. Monomer synthesis The spectroscopic data (FTIR, 1 H and 13 C NMR) belonging to itaconic anhydride, the obtained monomers and their respective intermediates can be consulted in the supporting information file. 2.3.1. Itaconic anhydride In a two-neck round bottom glass, 10 g (7.7 mmol) of itaconic acid and 19 mL (26.3 mmol) of acetyl chloride were added and stirred at 55 ◦C for 2.5 h. Then, 10 mL of dry toluene was added to the reaction and refluxed for 3 h using a Dean-Stark apparatus. Subsequently, 25 mL of diethyl ether was added, and the solution was stored at −20 ◦C overnight. The resulting white crystals were washed with cold diethyl ether and dried in a vacuum oven until constant weight. Yield: 81%. Melting point: 65–68 ◦C. 2.3.2. Preparation of mono methylnorbornyl itaconic acid (1a), mono norbornyl itaconic acid (1b), mono methyladamantyl itaconic acid (1c) and mono adamantyl itaconic acid (1d) The synthesis of mono-ester itaconic acids 1a, 1b and 1c was achieved following a slightly modified protocol previously reported by our group, which was also useful for the preparation, for the first time, of specimen 1d [43]. Briefly, in a two-neck round bottom flask coupled to a reflux system, 2.00 g (18 mmol) of itaconic anhydride was mixed with 1.2 equivalents (21 mmol) of the corresponding alcohol and solubilized in 15 mL of dry toluene. The reaction was heated at 120 ◦C for 24 h, after which the solvent was removed in a rotary evaporator at reduced pressure, obtaining a transparent oil in all cases. 1a and 1b were successfully purified by crystallization, pouring the transparent oil into 10 mL of hexane and storing them for 24 h at −20 ◦C, allowing the formation of white crystals. In the case of 1c and 1d, the purification process was carried out by column chromatography using a mixture of hexane:acetone (7:3) as the eluent, to give a transparent oil that solidifies in the form of white crystals upon cooling at −20 ◦C. Samples 1a, 1b, 1c and 1d were obtained with yields of 68, 81, 55 and 57%, respectively. 2.3.3. Preparation of 2-(methylsulfonyl)ethyl methylnorbornyl itaconate (2a), 2-methylsulfonylethyl norbornyl itaconate (2b), 2-methylsulfonylethyl methyladamantyl itaconate (2c) and 2-methylsulfonylethyl adamantyl itaconate (2d) The synthesis of diitaconates 2a, 2b, 2c and 2d was performed starting from their corresponding mono itaconate esters 1a, 1b, 1c and 1d, following the same experimental protocol in each case. The protocol is described below, using the synthesis of 2a as an example: A total of 3.0 g of 1a (13.0 mmol) and 2.4 g of 2-(methylsulfonyl) ethanol (19.5 mmol, 1.5 equiv) were solubilized in 30 mL of dry dichloromethane (DCM) and cooled in an ice-water bath. Simultaneously, a second solution was prepared in a round-bottom glass by dissolving 3.2 g of DCC (15.6 mmol, 1.2 equiv) and 0.3 g of DMAP (2.6 mmol, 0.2 equiv) in 70 mL of dry DCM, which was left under constant stirring in an ice-water bath. Afterwards, when both solutions reached 0 ◦C, the first solution was added dropwise over the second solution under constant stirring, and after concluding the addition, the reaction mixture was allowed to reach room temperature and react for 24 h, maintaining the stirring conditions. Then, the mixture was stored overnight at −20 ◦C to favor the precipitation of the dicyclohexyl urea (DCU) subproduct. The cooled mixture was filtered and subjected to S. Bonardd et al. Reactive and Functional Polymers 196 (2024) 105842 4 consecutive washing steps with brine (5 ×50 mL), after which the organic phase was dried over anhydrous MgSO 4 and filtered. The solvents were removed under reduced pressure, and the whitish oil obtained (with clear presence of DCU traces) was successfully purified by column chromatography (7:3, ethyl acetate:hexane), affording a transparent oil that slowly formed a white solid over time (2a). Yield =71%. Similarly, after column chromatography, 2b, 2c and 2d turned out to be transparent oils that, after a certain period of time, became white solids. The calculated yields for 2b, 2c and 2d were 68, 61 and 78%, respectively. 2.4. Synthesis of polyitaconates The synthesis of polymers was carried out in bulk using AIBN as the radical initiator. Typically, 2.0 g of monomer was placed into a Schlenk tube equipped with a magnetic stirrer, followed by the addition of the respective amount of AIBN to afford an initiator concentration of 0.5 mol%. The system was subjected to three consecutive freeze–thaw cycles employing liquid nitrogen and finally purged with dry N 2 . The entire system was immersed in an oil bath thermostated at 70 ◦C and stirred for 48 h. After completing the reaction time, the mixture was exposed to air, dissolved in 10 mL of N,N ′ -dimethylformamide (DMF), precipitated into 150 mL of methanol and filtered. The above process was repeated three times. Finally, the obtained white powders were dried at 120 ◦C in a vacuum oven for 72 h prior to characterization. The nomenclature chosen for the obtained polymers was PSO 2 mNORI, PSO 2 NORI, PSO 2 - mADAI and PSO 2 ADAI, which were obtained directly from the polymerization of monomers 2a, 2b, 2c and 2d, respectively. Conversion values of 72%, 64%, 58% and 50% were calculated for PSO 2 mNORI, PSO 2 NORI, PSO 2 mADAI and PSO 2 ADAI, respectively. 3. Results and discussion The experimental part of the present project began with the design and preparation of the corresponding monomers that would allow the obtainment of this new family of polyitaconates. Scheme 1 illustrates the synthetic route employed to achieve the preparation of the devised monomers 2a, 2b, 2c and 2d. All monomers were derived from itaconic anhydride (1), which was successfully synthesized from itaconic acid using well-established methods [43]. Subsequently, this anhydride was reacted under reflux conditions with the corresponding alcohols derived from norbornane and adamantane structures. This induced the opening of the anhydride structure, leading to the formation of mono itaconate esters 1a, 1b, 1c and 1d in yields ranging from regular to good. This was confirmed by 1 H and 13 C NMR analysis (Figs. S1 – S5). Subsequently, each monoester underwent a second esterification process using the Steglich protocol. [44] This led to the incorporation of sulfone entities into the final monomeric structures, resulting in the obtainment of diitaconates 2a, 2b, 2c and 2d. This was also confirmed by 1 H and 13 C NMR analysis (Figs. S6 – S9). Importantly, all diitaconates were successfully polymerized using conventional radical polymerization under bulk conditions, as depicted in Scheme 2. These polyitaconates, all containing sulfonyl groups as high dipole moment entities, can be categorized into two families based on the structure of the attached aliphatic cycle. In this sense, one family would be represented by PSO 2 mNORI and PSO 2 NORI, having norbornane-like structures, with the only structural difference between them being the Scheme 1. Synthetic route for obtaining hetero – functionalized itaconate monomers. S. Bonardd et al. Reactive and Functional Polymers 196 (2024) 105842 5 presence of a methylene spacer unit in PSO 2 mNORI. Likewise, the second family includes PSO 2 mADAI and PSO 2 ADAI, both featuring adamantane-based moieties. Once more, the primary structural distinction between them lies in the presence of the methylene unit in PSO 2 mADAI. The conversion values for all obtained polymers were in the range of 50–72%. Interestingly, when comparing within the same family, systems with the spacer unit exhibited higher conversion values. Conversely, comparing specimens with or without the spacer unit revealed that those bearing norbornane groups also exhibited higher conversions than those with adamantyl pendant groups. This can be attributed to the higher steric hindrance generated: 1) the absence of the methylene spacer unit, bringing the aliphatic cycle closer to the polymerizable section of the monomer, and 2) the replacement of norbornane structures by bulkier adamantane entities. [45–48] The structural characterization of these polyitaconates was first addressed through FTIR analysis, where spectra of the obtained polymers are displayed in Fig. 1. The FTIR analysis revealed very similar spectra among the four Scheme 2. Preparation of sulfonyl-containing polyitaconates through bulk radical polymerization. Fig. 1. FTIR spectra recorded for PSO 2 mNORI (A), PSO 2 NORI (B), PSO 2 mADAI (C) and PSO 2 ADAI (D). S. Bonardd et al. Reactive and Functional Polymers 196 (2024) 105842 6 specimens, which was anticipated due to the notable similarities in their chemical structures. In light of this, three colored regions were highlighted in all spectra to identify the groups of signals that would confirm the presence of the desired functionalities. The yellow zone, defined between 3000 and 2800 cm −1 , encompasses two vibrational bands arising from the C – H linkages originating from the aliphatic content in the polymer backbone. Additionally, it includes a significant contribution of −CH−and −CH 2 −units belonging to norbornane and adamantane pendant structures. Indeed, this group of signals stands out by their remarkable intensities, being the most intense in the PSO 2 mADAI and PSO 2 ADAI spectra due to the presence of adamantane units. Interestingly, the second band appearing at lower wavenumber values (⁓2860 cm −1 ) has been previously assigned to certain vibrational modes of C – H linkages present in this class of rigid cycles [49]. At lower wavenumber values, within the cyan zone, an intense and sharp signal centered around ⁓ 1735 cm −1 would accuse the presence of the typical carbonyl structures of polyitaconates. Crucially, the presence of sulfone groups was confirmed by identifying the two characteristic bands of this dipolar entity centered at ⁓ 1315 cm −1 and ⁓ 1132 cm −1 (highlighted in gray), corresponding to the symmetric and asymmetric stretching motions of these functionalities, respectively [50]. Finally, indicated by black arrows, a signal centered at approximately 1180 cm −1 can be identified in all spectra. This corresponds to the vibrational motions of C – O – C linkages that are part of the ester groups present in the monomeric units of these materials. As a complement to the FTIR analysis, the chemical structure of these polymers was also studied by 1 H and 13 C NMR, and the results are shown in Fig. 2 and Fig. S10, respectively. NMR analysis of the four obtained polymers successfully identified distinct fragments belonging to their respective monomeric structures. In this context, the presence of sulfonyl-containing structures (blue fragments in Fig. 2) was discerned in all 1 H NMR spectra. This was primarily accomplished by identifying the methylene units from the 2methylsulfonyl ethyl pendant groups. More notably, the characteristic signal attributed to the methyl group directly attached to the sulfonyl moiety provided crucial confirmation. Conversely, the presence of norbornane or adamantane structures (red fragments in Fig. 2) is substantiated by the multitude of densely packed signals appearing at high field. These signals exhibit significant intensity and a notable degree of overlap, making it difficult to assess the main-chain protons labeled as 1 and 2, which should appear in the range between 3.0 and 2.0 ppm as evidenced in previous polyitaconate systems. Additionally, these signals tend to exhibit a low intensity and definition due to tacticity phenomena, affecting their visualization. [34,42,43,51] Similarly, the analysis of 13 C NMR spectra – shown in Fig. S11 – corroborates the previously extracted information. Successful signal assignment for the various carbon nuclei was accomplished, with the exception of certain mainchain carbon atoms (e.g., carbon C), likely due to its overlap with the DMSO solvent signal. Thus, by complementing the FTIR and NMR analyses, the structure of the obtained polymers can be confirmed. Therefore, once demonstrated, their macromolecular nature was Fig. 2. 1 H NMR spectra registered for PSO 2 mNORI (A), PSO 2 NORI (B), PSO 2 mADAI (C) and PSO 2 ADAI (D). S. Bonardd et al. Reactive and Functional Polymers 196 (2024) 105842 7 confirmed by gel permeation chromatography (GPC), through which their average molecular weights, along with their dispersity indexes (Ð), were obtained. The results are summarized in Table 1. The molecular weights obtained for PSO 2 mNORI, PSO 2 NORI and PSO 2 mADAI correlate well with our previous reports on the preparation of counterparts bearing nitrile groups as dipolar entities [43]. An exception arises with PSO 2 ADAI, which exhibits a molecular weight value approximately one order of magnitude lower. Overall, considering the experimental conditions employed during the polymerization step, the obtained values are significantly lower than those reported in previous studies concerning the preparation of other polyitaconate materials [43,45]. We believe that the incorporation of norbornane and adamantane groups into the monomeric units significantly affects their addition/reactivity toward the terminal radicals of growing chains, and the main argument supporting this should be the increased steric hindrance that these monomers expose due to the presence of these bulky structures. Based on the above, it could be suggested that the lower molecular weight of PSO 2 ADAI may be attributed to its monomer unit, which, among all tested monomers, presents the highest steric hindrance. Conversely, concerning the Ð values, all of them fall within the anticipated range for this type of conventional radical polymerization. It should be noted that, in accordance with previous studies the average molecular weight and dispersity values exhibited by these types of materials should not substantially affect their polarization properties [20,52]. This is because the polarization phenomena of dipolar glass polymers primarily depend on the local movements of dipolar structures rather than long-range motions. The thermal properties of these new polyitaconates were evaluated in terms of TGA and DSC analyses, and the results are shown in Fig. 3 and summarized in Table 2. TGA analysis reveals that all samples exhibit remarkably similar degradation profiles (Figs. 3 A – D), with two discernible degradation steps characterized by temperatures T MD1 and T MD2 . It is important to note that, in all cases, >80% of the mass loss occurs during the first step. Arguably, various phenomena occurring simultaneously within the temperature range initiate the cleavage of chemical bonds present in the materials. This could be brought about by the cleavage of pendant groups or even the random fragmentation of the polymer backbone, leading to the generation of radical species that, under these high temperature conditions, can initiate depolymerization processes. In this sense, low molecular weight species would be generated and volatilized as degradation products. Similarly, albeit with a significantly lower mass contribution, the second step may be associated with the degradation of certain chemical structures that persist after the first step. However, the thermal stability of these materials is better characterized by the onset degradation temperature (T i ), defined in this case as the temperature at which the material loses 5% of its initial weight. Based on the above, it is evident that all the obtained polyitaconates commence degradation well above 250 ◦C, a condition established to deem as materials possessing adequate high-thermal resistance for application as dielectric media [16]. In line with our previous report on bulky polyitaconates [43], it is evident once again that the materials with monomeric structures lacking methylene spacer units (PSO 2 NORI and PSO 2 ADAI) exhibited lower T i and T MD1 values than their counterparts, PSO 2 mNORI and PSO 2 mADAI (Table 2). This lower thermal resistance has been previously attributed to a possibly more rigid structure that would result in a less stable system against bond vibrations induced by thermal energy. Importantly, these samples possess a markedly higher thermal stability than PSO 2 MeiT (T i = 277 ◦C), a dipolar glass polymer previously reported by us, consisting of a sulfone-containing polyitaconate obtained after the replacement of norbornane/adamantane structures by methyl units [34]. In DSC analysis, as shown in Fig. 3E, it is evident that all obtained polymers display an amorphous behavior, completely devoid of crystallization/melting phenomena. This fulfills a crucial criterion for defining them as dipolar glass polymers. Furthermore, these materials distinguish themselves by exhibiting glass transition temperatures (T g , indicated by black arrows) well above room temperature, surpassing the 100 ◦C threshold (Table 2). The elevated T g values can be attributed to the presence of norbornane/adamantane structures, which, owing to their bulkiness and rigid nature, augment the rigidity of the macromolecular structures. Experimentally, the role of these cyclic entities in elevating T g was demonstrated by comparing these results with the T g value of PSO 2 MeiT, reported at 101 ◦C, even though the latter exhibited a significantly higher molecular weight [34]. Another intriguing observation arises when comparing the T g values between PSO 2 mNORI and PSO 2 NORI, where a notable increase of approximately 32 ◦C was achieved by simply removing the methylene spacer unit between the norbornane structure and the polymer backbone. This could be attributed to a potentially favorable contribution to the free volume within the system that these spacer units could offer, where the absence of these CH 2 units would decrease the degrees of freedom of the bulky pendant groups, restricting their possibility to achieve different spatial orientations within the material’s structure. On the contrary, for PSO2mNORI, the greater orientational dispositions available for the norbornane structures could favor an increment of the free volume in the material, resulting in a lower T g value. Unexpectedly, this trend was not observed in the case of PSO 2 mADAI and PSO 2 ADAI, likely due to the markedly lower molecular weight exhibited by PSO 2 ADAI since, as is well known, there is a close dependence between T g and molecular weight in polymeric materials. [53,54] Nevertheless, this suggests that in the event of successfully preparing a PSO 2 ADAI specimen with an increased molecular weight, significantly higher T g values could be anticipated. Finally, as a noteworthy trend, when comparing these polymers with their counterparts featuring nitrile groups as dipolar entities, the former exhibited higher T g values. Considering that molecular weight values are on the same order of magnitude, the above result could be related to stronger dipolar interactions within the sulfone-containing specimens. This could also be reflected in the higher-density values that the sulfonebearing systems exhibited against those with nitrile moieties [34]. In this context, density values of 1.28, 1.24, 1.22 and 1.20 g/cm 3 were obtained for PSO 2 mNORI, PSO 2 NORI, PSO 2 mADAI and PSO 2 ADAI, respectively. After completing the thermal characterization of these materials, their dielectric properties were evaluated in terms of BDS measurements. Fig. 4 displays isochronal measurements of the dielectric constant ( ε r ’), dielectric loss ( ε r ”) and loss factor (Tan(δ)) parameters at different frequencies and temperatures for all obtained polyitaconates. Isochronal representations are particularly useful for directly observing how ε r ’, ε r ” and Tan(δ) parameters vary with changes in frequency and temperature. Importantly, they facilitate the visual assessment of any relaxation phenomena occurring in the material as temperature varies. It is worth noting that all obtained polymers exhibit a similar dielectric response characterized by closely matching profiles ε r ’, ε r ” and Tan(δ). This outcome was anticipated due to the structural similarities between them. Regarding the dielectric constant, a qualitative inspection of ε r ’ profiles reveals an increase in this parameter as the temperature rises. This corresponds well with the thermal activation needed for dipolar entities to overcome the potential energy barriers imposed by their surroundings [55,56]. This increase is marked by two distinct jumps: the first occurring at low temperatures (approximately – 100 ◦C) and the second becoming apparent from ⁓ 100 ◦C onwards. It is noteworthy that between these jumps, a slight but consistent rise in ε r ’ is Table 1 Number (M n ) and weight (M w ) average molecular weights along with dispersity values (Ð) of synthesized polymers. PSO 2 mNORI PSO 2 NORI PSO 2 mADAI PSO 2 ADAI M n (g/mol) 1.23 ×10 4 2.99 ×10 4 3.10 ×10 4 3.08 ×10 3 M w (g/mol) 2.29 ×10 4 6.05 ×10 4 6.07 ×10 4 4.43 ×10 3 Ð 1.86 2.02 1.96 1.44 S. Bonardd et al. Reactive and Functional Polymers 196 (2024) 105842 8 observed. These variations in ε r ’ can be correlated with ε r ” profiles through diverse relaxation phenomena. In this sense, it can be seen that in all cases, the abrupt increase in ε r ’ at low temperatures manifests as a clear peak in the ε r ” profiles (labeled γ), indicating the occurrence of molecular relaxation phenomena. Drawing on previous dielectric investigations on polymethacrylates and polyitaconates, especially for PSO 2 MeiT, this γ relaxation can primarily be attributed to the orientational motions of sulfone entities [20,34,52]. We use “primarily” here, as it has been demonstrated in past reports that both carbonyl species in polyitaconates, acting as dipolar entities in their respective monomeric units, contribute to polarization through different relaxation processes known as β FAST and β SLOW transitions [57]. Carbonyl species separated by a methylene unit from the polymer backbone are responsible for β FAST relaxation, which occurs within the same temperature range as γ relaxation. However, the dielectric strength of these molecular motions is notably lower than that produced by sulfones. Thus, it is reasonable to consider sulfone entities as the main drivers of these γ relaxations and, consequently, of the polarized state exhibited by these materials. Importantly, the maxima of γ relaxations at 1 Hz are centered at −96 ◦C, −101 ◦C, −98 ◦C and – 97 ◦C for PSO 2 mNORI, PSO 2 NORI, PSO 2 mADAI and PSO 2 ADAI, Fig. 3. TGA (A, B, C and D) and DSC (E) thermograms registered for polyitaconates. T g s are indicated by black arrows. Table 2 Thermal degradation onset temperature (T i ), maximum weight loss rate temperature (T MD ), residue percentage (R) and glass transition temperature for the obtained polyitaconates. Sample T i (◦C) T MD1 (◦C) T MD2 (◦C) R (%) T g (◦C) PSO 2 mNORI 299 329 472 12 111 PSO 2 NORI 297 320 453 10 143 PSO 2 mADAI 319 339 426 7 166 PSO 2 ADAI 301 324 392 9 136 S. Bonardd et al. Reactive and Functional Polymers 196 (2024) 105842 9 respectively, indicating that sulfone movements are active and thus contribute to polarization from remarkably low temperatures. Conversely, carbonyl species directly attached to the polymer main chain begin to contribute to their molecular motions (β SLOW transitions) at temperatures above – 50 ◦C [34,57]. For clarity, in Figs. 4 B, E, H, K β FAST and β SLOW transitions have been grouped together and labeled as β contributions. It is important to note that since γ, β FAST and β SLOW molecular motions occur at temperatures below the T g , these sub-T g Fig. 4. Isochronal scans of ε r ’, ε r ” and Tan(δ) for PSO 2 mNORI (A, B, C), PSO 2 NORI (D, E, F), PSO 2 mADAI (G, H, I) and PSO 2 ADAI (J, K, L). Scheme 3. Assignment of molecular motions to the observed dielectric relaxations in isochronal spectra. S. Bonardd et al.