Studies of Formation Mechanism, Structure, and Properties of Network Copolymers Obtained by Cocuring of Rolivsan Thermosetting Resins with Aromatic Diamines.
Rolivsan thermosetting resins (ROLs) are a new class of unsaturated thermosetting resins featuring high glass-transition temperatures ([T.sub.g]), excellent processability, low relative permittivity and mechanical losses, low moisture absorption, good mechanical properties, and satisfactory compatibility with various substrates [1-3]. They are obtained via acid-catalyzed reactions between disecondary aliphatic-aromatic diol (bis-(4-([alpha]-hydroxyethyl)phenyl) ether) (BHEPE) and methacrylic acid (MAA) . It was established that during thermal curing of ROLs, thermostable network copolymers are formed; these copolymers are capable of changing their chemical structure, microstructure, crosslink density and, correspondingly, thermal, mechanical, and other properties .
In the first low-temperature stage of curing (140-170[degrees]C), bulk three-dimensional free radical polymerization (TFRP) of unsaturated components of ROLs occurs and the initial highly crosslinked copolymers are formed. When the temperature exceeds 170[degrees]-180[degrees]C, thermosensitive methacrylate bridges start to cleave gradually and form MAA and styrene-like ([CH.sub.2] = CH-Ar-) pending units. Upon further increase in curing temperature (up to 200-300[degrees]C), the adjacent MAA units and the fragments containing styrene-like aromatic rings become involved in solid-phase intramolecular reactions. These reactions include dehydration, intramolecular cyclization (high-temperature acylation of neighboring aromatic units by MAA carboxyl groups and anhydrization of carboxyl groups of neighboring MAA units within the chain). Thus, the processes result in specific self-functionalization of polymer grains and interlayers between these grains [1,2].
Formation of microheterogeneous-crosslinked polymers is accompanied by microphase segregation (microsyneresis) of the initially homogeneous unsaturated system (e.g., the one consisting of oligoetheracrylates or other unsaturated monomers and/or oligomers). It is preceded by a sharp increase in viscosity of the system caused by formation and association of polymer chains and by nucleation and growth of crosslinked microgel particles (grains). As microgel particles (grains, microdomains) continue to grow, the surrounding liquid reaction medium is transformed into interlayers between the grains. In these interlayers, nonreactive impurities (such as solvents and/or various functional additives) are accumulated. The size of grains (dispersed phase) increases during formation of polymer chains in their surface (boundary) layers by "extraction" of unsaturated monomers and oligomers from the reaction medium into the continuous dispersion medium. While three-dimensional free radical polymerization (TFRP) proceeds in bulk, the grains fuse together, form a monolith, and are transformed into a continuous phase. At the same time, the reaction medium enriched in nonreactive impurities and functional additives forms solidifying interlayers between the grains (IBGs) with the [T.sub.g] lower than those of the grains. These IBGs drastically deteriorate useful properties of the network polymer (mechanical strength, heat resistance, etc.) and can be considered as fundamental structural defects of the supramolecular level .
Thus, accumulation of inert impurities and functional additives (that are usually present in the initial liquid thermosetting resin) in IBGs is an important consequence of the formation of the morphological structure of the network polymer. Earlier, we have proposed to use this feature of the microheterogeneous mechanism for delivering potentially reactive (polyfunctional) compounds (that do not participate in the main TFRP process) into IBGs. However, these latent compounds, which are present in IBGs as a peculiar kind of "microreactors," can be activated under the appropriate conditions (e.g., at elevated temperatures and/or after changing acidity of a medium). After activation, they take part in additional reactions of structurization (crosslinking) of supramolecular formations (grains and interlayers) in microheterogeneous network polymer [2, 5-7]. Thus, we propose to use certain implications of mechanism of the network formation as a new structure-forming tool for chemical modification of thermosets. The method has turned out to be effective in improving physicomechanical, thermal, and other useful properties of thermosets, as was illustrated by the example of thermochemical transformations of ROLs in binary systems containing epoxy resins [2, 3, 5] and aromatic di- and tetraamines [3, 5-7].
In the present work, dynamic mechanical analysis (DMA), FTIR, and [sup.13]C NMR spectroscopy were used in investigation of conditions and mechanism of cocuring ROLs with ADA and in the studies of the chemical structure and properties of the obtained network copolymers.
2.1. Materials. m-Phenylenediamine (m-PDA, 99%), p-phenylenediamine (p-PDA, 99%), 3, 4,-oxydianiline (3,4;-ODA, 97%), 4,4;-oxydianiline (4,4;-ODA, 97%), 4,4'-methylene dianiline (4,4;-MDA, 97%), 1,3-bis(4-aminophenoxy)benzene (1,3-BAPB, 98%), toluene (99.8%), methacrylic acid (MAA, 99%), p-toluenesulfonic acid monohydrate (p-TSA, 98.5%), and hydroquinone (HQ, 99%) were purchased from Sigma-Aldrich. 1,4-Bis-(4-aminophenoxy)benzene (1,4-BAPB, 98%) was purchased from TCI America and used as received.
2.2. Preparation of ROLs and the Cured ROL-ADA Blends. ROL samples were synthesized by condensation of bis(4-([alpha]-hydroxyethyl)phenyl) ester (BHEPE)  with methacrylic acid (MAA) that was catalyzed by p-TSA, in accordance with the following procedure. The mixture of 140.0 g (0.542 mol) of BHEPE, 3.5 g of CuCl, 1.4 g of HQ, 1300 mL of toluene, and 70.0 g (0.813 mol) of freshly distilled MAA was heated at stirring at the temperature of 70[degrees]C for ~15 min. After that, 0.420 g of p-TSA was added to the solution; the mixture was stirred, and boiling temperature (102-111[degrees]C) was maintained until the dehydration process ceased (i.e., for ~1.5h). The water released during the process was distilled off as a toluene-water azeotrope. When water release stopped, the resin solution was cooled down to room temperature, precipitate was filtered off, and the solution was washed successively with distilled water (several times, ~50 mL at a time), with 5% solution of NaHC[O.sub.3] (three times, 50 mL at a time), and with water (several times, 100 mL at a time), until pH = 7 was reached. The solution was dried over Ca[Cl.sub.2] and filtered again, and toluene was removed by distillation under reduced pressure. The yield of the resin was 145 g. Several characteristics of noncured neat ROLs are given in Table 1.
For the present study, we have selected ADA which are widely used in industry, e.g., for curing epoxy resins, in the synthesis of bismaleimides and aromatic polyimides. The bulk polymer samples were prepared by casting neat ROLs and ROL-ADA blends into preheated open brass molds followed by heating the resins (blends) according to the following stepwise temperature-time regime until the final curing temperature [T.sub.fin.cur.] was achieved ([degrees]C/h): 150/2, 160/2, 160-180/0.5, 180/1, 180-220/0.5, 220/1, 220-260/0.5, 260/1, 260-300/0.5, 300-320/0.5, 320/1, and 350/3 (Table 2).
2.3. Characterization. Fourier transform infrared spectroscopy (FTIR) was performed using a Bruker Vertex 70 spectrometer (Bruker, Germany; 4000-600 [cm.sup.-1]) equipped with an ATR (attenuated total reflection) attachment (PIKE Technologies, USA). The samples were studied in a thin layer and in the form of finely ground powder. Solid-state [sup.13]C cross-polarization (CP)/magic angle spinning (MAS) NMR spectra of powdered samples of 1,3-BAPB and the cured ROL-1,3-BAPB blends were taken with a Bruker AVANCE 400 spectrometer at an operating frequency of 125.77 MHz in the CP MAS mode (10 kHz). The [sup.13]C chemical shifts were externally referenced to tetramethylsilane (TMS). Temperature dependences of the storage (dynamic) Young's (E') or shear (G') moduli and mechanical loss tangent (tan [delta]) were obtained by dynamic mechanical analysis (DMA) using a DMA 242C/1/F dynamic mechanical analyzer (NETZSCH, Germany) at a frequency (v) of ~1 Hz, torsion pendulum at v = 0.30, and a setup operating at v = 660 Hz in a tension mode . DMA studies were performed using rectangular bars of cured samples at a heating rate of 5[degrees]C/min. Accelerated thermo-oxidative aging of the cured neat ROLs and ROL-ADA blends (with masses of 0.50 [+ or -] 0.10 g and average dimensions of 32 * 5 * 4 mm) was conducted in air thermostat at constant temperature (350[degrees]C); weight loss (WL) was determined every 1-3 h. Dynamic thermogravimetric analysis (TGA) of grinded samples was performed using a Paulik-Paulik-Erdei "Derivatograph C" (MOM, Hungary); the heating rate was 10[degrees]C/min. For determining gelation time ([[tau].sup.t[degrees]]), a liquid resin sample was placed in the Pyrex test tube immersed into oil bath with a constant temperature and timing was started. In these experiments, we chose the following temperatures: 140[degrees]C, 155[degrees]C, and 170[degrees]C. In the beginning (at 140[degrees]C), the resin system demonstrated very low viscosity (similar to that of water). As the copolymerization proceeded, gelation occurred; beginning of the gelation process was taken as a start of crosslinking.
3. Results and Discussion
3.1. FTIR Studies of Neat ROL-II Samples in the Process of Curing. Three-dimensional free radical polymerization (TFRP) of ROLs in bulk at 140-170[degrees]C yielded glassy highly crosslinked copolymers with heat-sensitive methacrylate crosslinks. In the course of subsequent chemical transformations occurring at higher temperatures (180-320[degrees]C), these initial copolymers change their chemical, topological, and supramolecular structures and properties . Figure 1 presents FTIR spectra of crosslinked ROL copolymers formed after curing neat ROL-II resin. The process involved stepwise heating from 140 to 160[degrees]C up to different final temperatures [T.sub.fin.cur.] (180, 220, 260, and 320[degrees]C) for several hours (Table 2). All spectra include two broad high-frequency absorption regions. The first one is a region of weak absorption (3600-3300 [cm.sup.-1], with the maximum at ~3400[cm.sup.-1], associated with carboxyl and phenol groups). The second one is the 3100-2850 [cm.sup.-1] area attributed to the following C-H stretching vibrations: 3030 ([C.sub.Ar-H]), 2959 (C-H in the methyl group), 2928 (C-H in the methylene group), and 2872 [cm.sup.-1] (C-H in the methyl group).
Restructurization of the initial network copolymer that is formed during heating of the ROL-II sample in the thermal regime up to [T.sub.fin.cur.] = 180[degrees]C (spectrum 4) is very clearly manifested in the FTIR spectra. The process can be indicated using position, shape, and intensity of the stretching vibration band of the carbonyl (>C=O) group (Figure 1, spectra 1-3). There is a strong absorption band at 1717 [cm.sup.-1] attributed to monomeric and oligomeric methacrylate components that are present in all neat (uncured) ROLs. In these components, the [CH.sub.2] = C([CH.sub.3)]- and >C = O groups are conjugated with each other. As TFRP proceeds (when ROL-II is heated at 150[degrees]-180[degrees]C), the carbonyl group appears to be in the nonconjugated position. As the intensity of this band gradually decreases, it becomes broader and shifts toward higher frequencies (spectrum 4, 1732 [cm.sup.-1]). During further heating of ROLs in the temperature range from 180[degrees] to 260 (320)[degrees]C, heat-sensitive methacrylate crosslinks of the initial network ROL copolymer begin to cleave with the formation of MAA units. The absorption band attributed to the carboxyl groups of these MAA units lies in the 1697-1699 [cm.sup.-1] area (Figure 1, spectra 2 and 3). Therefore, transformation of ester groups into carboxyl groups leads first to broadening of the carbonyl band and then to shifting of this peak toward lower frequencies. In the course of heating of the neat ROL-II resin up to [T.sub.fin.cur.] = 260[degrees]-320[degrees]C (spectra 1 and 2), new bands appear. They are associated with carbonyl fragments formed after intramolecular cyclization of neighboring methacrylic units in the ROL copolymer chains. The cyclization leads to the formation of methacrylic anhydride (MA) units (doublet at 1799-1801 and 1755-1757 [cm.sup.-1]).
3.2. FTIR Studies of the ROL-ADA Blends in the Process of Curing
3.2.1. ROL-II- (75wt.%) m-PDA (25 wt.%) Blends. In contrast to the copolymers prepared by thermal cure of neat ROL-II resin (Figure 2, spectrum 1), no absorption bands of MA units (1801 and 1756 [cm.sup.-1]) were observed in the IR spectra of cured ROL-II-1,3-PDA (II-IV) blends (Figure 2, spectra 2-4). The observed bands attributed to carbonyl groups at 1724-1726 and 1680 [cm.sup.-1], as well as weak bands at 1375, 1103, and 691 [cm.sup.-1], represent the characteristic absorptions of amide and, possibly, imide fragments [8-14]. However, it should be noted that the bands at 1780 and 727 [cm.sup.-1] (typical of imide absorbance) were not observed in the spectra of the studied blend samples.
3.2.2. ROL-II- (85wt.%) 1,3-BAPB (15wt.%) Blends. Let us consider FTIR spectra of ROL-II-1,3-BAPB (V-VIII) blends cured in the thermal regime up to final temperatures of 180 (V), 220 (VI), 260 (VII), and 320[degrees]C (VIII). In the high-frequency regions of their spectra (Figure 3), a wide absorption band is observed (3440-3250 [cm.sup.-1]) with a peak at 3368 [cm.sup.-1]. This absorption is typical of O-H and amide N-H stretching vibrations in polyamic and polymethacrylic acids [8-14]. In contrast to thermally cured neat ROL-II heated at temperatures not exceeding 180[degrees]C (Figure 1, spectrum 4), no peaks attributed to anhydride (1801 and 1757 [cm.sup.-1]) and carboxylic groups (1701 [cm.sup.-1]) were observed in this case. The band of the carbonyl group (>C = O) at ~1715 [cm.sup.-1] (Figure 3, spectra 1 and 2) that is attributed to monomeric methacrylate components of ROL becomes broader. This is caused by formation of methacrylate units in the initial ROL copolymer. Upon further heating, this band splits (see Figure 3, spectra 3 and 4 of the ROL-II-1,3-BAPB-VII and VIII samples, cured up to final temperatures ([T.sub.fin.cur.]) of 260[degrees] and 320[degrees]C, respectively).
At the same time, the peaks at 1728-1732 [cm.sup.-1] and 1690-1695 [cm.sup.-1] appear. They indicate formation of the nonconjugated C=O group of the methacrylate unit in ROL copolymer and, apparently, formation of MAA and methacrylamide units (Figure 5, spectra 3 and 4). Weak absorption at 1801 [cm.sup.-1] is related to the presence of traces of MA units in the studied samples (Figure 3, spectra 3 and 4). Note that the spectrum of the diamine sample subjected to similar heat treatment (1.3 - BAPB) does not include absorption bands in the 1850 - 1650 [cm.sup.-1] region . To summarize, IR spectra of the cured ROL-II-1,3-BAPB (V-VIII) blends with relatively low concentration of amino groups (Figure 3) contain the peaks typical of amide rather than of imide units. It is seen that in the spectrum of the ROL-II-1,3-BAPB (VIII) sample cured in the regime with the final temperature [T.sub.fin.cur.] = 32 0[degrees]C, virtually no absorption at 1780 [cm.sup.-1] is present.
3.2.3. ROL-II- (50wt.%) 1,3-BAPB (50wt.%) Blends. Analysis of FTIR spectra of the ROL copolymers prepared from ROL-II-1,3-BAPB blends with relatively low amounts of amino groups has not clarified the point of efficiency of the thermochemical interaction between the carboxylic (anhydride) and amino groups. Therefore, we carried out the experiments involving the blends with higher concentration of this diamine and those cured at the highest temperatures ([T.sub.fin.cur.] = 320[degrees] - 350[degrees]C).
As seen from Figure 4, in the high-frequency area of FTIR spectra of cured blends, two broadbands at (i) 3500-3400 [cm.sup.-1] 1 (with a maximum at 3445 [cm.sup.-1]) and (ii) 3400-3300 [cm.sup.-1] (with a maximum at 3370 [cm.sup.-1]) are observed. As [T.sub.fin.cur.] is increased from 320[degrees] to 350[degrees]C (ROL-II-1,3-BAPB (XIII and XIV), spectra 2 and 4), these bands become more intense. Intensity of the bands related to the neat 1,3-BAPB (XV) heat-treated (oligomerized) in air (spectrum 1) depends on the content of this diamine in the initial ROL-II-1,3-BAPB blend. Intensity of the bands attributed to stretching vibrations of hydrogen atoms in methyl and methylene groups (3100-2800 [cm.sup.-1]) is apparently determined by the ROL-II content in the blends. The bands associated with possible formation of (poly)amic acid units in the cured ROL-II-1,3-BAPB blends are overlapped with the abovementioned broad bands of carboxyl and phenol groups of the restructured ROL-II copolymers and amino groups of 1,3-BAPB (XV). Therefore, this range is hardly suitable for evaluating the efficiency of the reactions under consideration.
We proceed now to discussing the 1900-550 [cm.sup.-1] range of the FTIR spectra (Figure 4). First of all, it should be noted that carbonyl absorption bands in the 1850-1600 [cm.sup.-1] interval are naturally not observed for 1,3-BAPB (XV) (spectrum 1). In contrast to the restructured neat ROL-II copolymer (ROL-II (IV), spectrum 5), spectra of ROL-II-1,3-BAPB (VIII) and ROL-II-1,3-BAPB (XIII) blends (spectra 2 and 4) exhibit no band characteristic of MA (1801, 1757 [cm.sup.-1]), methacrylate (1736 [cm.sup.-1]), and MAA (1701 [cm.sup.-1]) units. The broad band of medium intensity appears around 1674 [cm.sup.-1]. This band is characteristic of amide units . It should be noted that the absorption band at 1674 [cm.sup.-1] is absent in the spectrum of neat ROL-II (IV) (spectrum 5). Spectra 2-5 contain a weak absorption band at 1377 [cm.sup.-1], which could confirm the formation of an imide ring [12-14] if it were absent in the spectrum of cured neat ROL-II (IV).
Comparison of spectra 2 and 4 of the cured blends with that of neat ROL-II (IV) (Figure 4, spectrum 5) reveals appearance of the broad band at 3445 [cm.sup.-1] that can be assigned to N-H bonds . The presence of [C.sub.Ar-N] bands at 1476 and 1308 [cm.sup.-1]  and of the band at 1250 [cm.sup.-1] in spectra 1-3 (while they are absent in spectrum 5 (the cured neat ROL-II)) confirms that the cured blends contain significant amounts of heat-treated 1,3-BAPB fragments.
3.3. [sup.13C] NMR Spectra of the Cured Neat ROL Samples. The nonrestructurized (neat) ROL-II (I) copolymer (obtained by radical copolymerization of styrene-like and (di)methacrylate ROL-components according to the thermal regime with [T.sub.fin.cur.] = 180[degrees]C (Figure 5) is similar to amorphous alternating styrene/methyl methacrylate copolymers (P(S-MMA)) in the chemical structure and ratio between comonomer units. Solid-state CP/MAS [sup.13]C NMR spectra of the neat ROL-II-copolymers (Figure 6) and the cured ROL-II-1,3-BAPB blends (Figure 7) were obtained. The majority of the observed resonance lines were very broad, since they are attributed to [sup.13]C carbon atoms surrounded by kindred units of amorphous polymers (which differ in composition). Many of the peaks produced by different types of carbon atoms are superimposed on each other. The spectra of neat ROLs (Figure 6) consist of numerous resonance lines, which originate from carbon sites in aromatic rings (160-110 ppm). The resonances at 158,154,140, and 136 ppm are assigned to substituted (nonprotonated) aromatic carbons ([sup.13][C.sub.Ar]-), and the resonances at 128 and 122 ppm are assigned to protonated atoms ([sup.13][C.sub.Ar].-H]). The signals in the ranges of 185-165 ppm and 60-15 ppm are related to carbonyl, methyl, methylene, and methine groups.
Assignment of chemical shifts in the spectra of neat ROLs and ROL-ADA blends was consistent with the previously reported data [17-24]. These data are related to [sup.13]C atoms included in kindred structures (groups, units, or segments).
For the purpose of comparison, we correlated the signals of [sup.13]C atoms in NMR spectra of the cured neat ROLs and the known kindred (co)polymers (i.e., polystyrene (PS), poly(-methyl methacrylate) (PMMA), poly(methacrylic acid) (PMAA), and poly(methacrylic anhydride) (PMAN)). The corresponding resonances were assigned to specified carbons; the data are summarized in Table 3. The sample of neat ROL-II (I) copolymer with [T.sub.fin.cur.] = 180[degrees]C (Figure 6, spectrum 4) shows wide [sup.13]C signals at 176.6 and 175.4 ppm (>C=O ester group (h)). These signals almost completely coincide with the signals in NMR spectra of an alternating S-MMA copolymer  and PMMA . With increasing [T.sub.fin.cur.] of the cured ROL-II samples (II, III, and IV) up to 220, 260, and 320[degrees]C, the intensity of [sup.13]C signals (>C=O, heat-sensitive methacrylate units) of the neat ROL copolymer (Figure 6, spectra 1-3) decreases. This change is related to thermocatalytic cleavage of these units (Figure 5) . At the same time, the spectra of samples cured in the regime with [T.sub.fin.cur.] 260[degrees] and 320[degrees]C (spectra 1 and 2) contain more intensive [sup.13]C (>C=O) signals at 173-172 and 185-183 ppm. These peaks are attributed to MA units (m)  and MAA units (j) , respectively. Note that there is a significant difference between positions of [sup.13]C (>C=O) signals of cyclic aliphatic (e.g., 1,2,3,4-cyclo-pentanetetracarboxylic dianhydride, 169.29 and 173.34 ppm ) and aromatic anhydrides (164.5 and 165.5 ppm ).
Analysis of FTIR and [sup.13]C NMR spectra of the cured neat ROL-II (I-IV) samples (Figures 1 and 6) demonstrates a good agreement between the data obtained by these two methods. The presence of carbonyl groups in the samples is confirmed by (i) the absorption bands related to >C=O valence vibrations and (ii) the [sup.13]C NMR signals attributed to carbonyl C atom: ([cm.sup.-1]/ppm): (1801 and 1756)/(171.0-172.5; 172.8ppm ) (MA units), 1732/(176-178); 178ppm  (methacrylate units) and 1700/(181.5-184.5); and 181 ppm  (MAA units).
3.4. [sup.13]C NMR Spectra of the Cured ROL-II-1,3-BAPB Blends. Figure 7 presents [sup.13]C NMR spectra of the samples prepared by curing ROL-II-1,3-BAPB blends according to the same thermal regime as was used in curing the neat ROL-II. Composition of the blend containing comparatively low amount (15wt.%) of 1,3-BAPB attracted special interest. As was shown above, the ROL copolymer samples obtained on the basis of this blend possessed maximum high-temperature strength and thermo-oxidative stability. Naturally, the spectra of this blend (Figure 7) are almost similar to that of the neat ROL resin (Figure 6) prepared by curing in the regime with [T.sub.fin.cur.] increasing from 180 to 320[degrees]C. This similarity is caused by high content of ROL component. In both series of spectra, wide signals are observed; they originate both from carbon atoms ([sup.13]C) of aromatic rings (at 158, 154, 140, 136, 128-129, and 122 ppm) and aliphatic groups (at 57, 45, 40, 24, and 18 ppm). The differences in spectra are related to carbonyl signals > [sup.13]C=O in the 180-160 ppm area, since these signals are sensitive to the nature of groups adjacent to carbonyl C atom.
The spectra of cured blends contain a wide signal in the 174-178ppm area with a peak at 176ppm. This signal continues to broaden with increasing [T.sub.fin.cur.] from 180[degrees] to 220-320[degrees]C (Figure 7, spectra 1-3). There is also a relatively narrow signal at 166.5 ppm (spectra 3, 4), which disappears at heating the blend up to [T.sub.fin.cur.] = 320[degrees]C (spectrum 1).
Unfortunately, the sought-for signals > [sup.13]C=O (C-m and C-n) that arise upon formation of nonaromatic amide and imide derivatives of methacrylic acid (according to the scheme in Figure 7) also lie in this range [11, 23-26]. Thus, for example, the resonance lines at 177.5-177.8 and 173.9 ppm were assigned to the imide carbonyl in cycloaliphatic-aromatic polyimides . The spectrum of aromatic (poly)amic acid contains the signals attributed to carbonyl carbon atoms at 172.80, 171.81, and 166.19 ppm , and the signal at 165.8ppm  and 162.9ppm  corresponds to the (O=[sup.13]C-N) fragment in polyimides. In the solid-state spectra, the signal related to aromatic imide carbonyl at 166 ppm is observed; three signals at 146.6, 144.2, and 141.1 ppm can be clearly distinguished in the region typical of substituted aromatic rings . The peak at 162.9 ppm in the spectra presented in Figure 7 is possibly overlapped with the wide signal of [sup.13][C.sub.Ar]-O-[sup.13][C.sub.Ar] (C-d, C-e) with the maximum at 158 ppm. Apparently, these wide signals of carbonyl groups in the spectra of the cured blend (Figure 7) do not provide unambiguous confirmation of the expected reactions, when they are considered separately. However, our purpose was to prove the proceeding of thermocatalytical reactions between diamine (1,3-BAPB) and the units of MA and (or) MAA in the restructurized ROL copolymer. In order to clarify this point, we analyzed other signals that could serve as indirect confirmation of chemical interactions (appearance of C-m and C-q and/or C-n and C-p signals, see scheme in Figure 7). Indeed, all spectra contain the most intense signal at 128-129 ppm that can be attributed both to [sup.13][C.sub.Ar]-H [19, 21-23] and to C-p (i.e., to [sup.13][C.sub.Ar]-N-C=O) [23, 25]. There is also a signal appearing at 136 ppm that can be assigned both to nonprotonated ([sup.13][C.sub.Ar]-) atom  and to imide carbon (C-p ([sup.13][C.sub.Ar]-N-C=O)) .
Thus, the spectra of ROL-II-1,3-BAPB blends (Figure 7) containing relatively low amounts of ADA include wide overlapping signals related to carbon atoms >[sup.13]C=O at 170180 ppm. These atoms may be incorporated in various units, and it is not possible to ascribe the signals with confidence to the desired amide and/or imide structures.
Figure 8 presents solid-state [sup.13]C NMR spectra of ROL-II-1,3-BAPB blends containing 50% of the diamine (spectrum 1) and spectrum of the cured neat ROL resin (spectrum 2).
Comparison of these two patterns shows that, contrary to expectations, these spectra contain signals of carbonyl carbon atoms (with almost similar positions and intensities) at 178, 176, 172.4, and 170.8 ppm. These signals are absent in the spectrum of oligomerized 1,3-BAPB (spectrum 3). However, the signals at 172.4 and 170.8 ppm are more pronounced in spectrum 1.
3.5. [sup.13]C NMR Spectra of the Cured ROL-II-m-PDA Blends. The ROL-II-1,3-PDA (IV) sample heated according to the regime that provides for the highest [T.sub.fin.cur.] = 350[degrees]C was studied by [sup.13]C NMR spectroscopy. Along with the signal at 177 ppm that is observed in the spectra of blends (Figures 7 and 8), the spectrum (Figure 9, spectrum 1) contains the signal at 173.3 ppm. The latter is absent in the spectra of samples after curing at lower final temperatures. Naturally, all spectra include rather intensive signals of the substituted [sup.13][C.sub.Ar]-O-(158 and 154ppm) and [sup.13][C.sub.Ar]-N< (146, 144, 141, and 138.5) groups and protonated (128, 122, 119, and 108 ppm) aromatic carbons.
3.6. Dynamic Mechanical Analysis (DMA). Figures 10-12 show that after thermal cocuring of ROLs with comparatively low amounts of ADA (10 [+ or -] 5 wt.%), the values of dynamic (storage) Young's modulus of the modified samples increase significantly over a wide temperature range. Besides, the apparent glass transition temperature [T.sub.g] shifts toward higher temperatures by 50-70[degrees]C. Moreover, modifying ROLs in this manner causes significant increase in stability of ROL copolymers against thermo-oxidative destruction (as follows from Figure 13).
3.7. Thermal Analysis
3.7.1. Thermal Oxidative Aging of Cured ROL-1,3-BAPB Blends at 350[degrees]C. The influence of the composition of cured ROL-1,3-BAPB blends on their resistance to thermal degradation is shown in Figure 13. Samples of ROL-II compounds with small (10 [+ or -] 5 wt.%) amounts of 1,3-BAPB showed the smallest weight loss in thermal oxidative aging at 350[degrees]C (Figure 13, curve 5). However, for the heat-treated 1,3-BAPB (sample II, curve 1) and samples with high diamine content (curves 2 and 4), the weight loss considerably increases. These experiments show that monomeric 1,3-BAPB sublimes at a high temperature (350-400[degrees]C).
3.7.2. The TG and DTA Curves for Cured Neat ROL Resins and ROL-1,3-BAPB Blends. The TG and DTA curves (Figure 14) recorded for the initial monomeric 1,3-BAPB ([T.sub.m] = 115 - 118[degrees]C) show that, at fast heating in air at a rate of 10deg * [min.sup.-1] in the temperature interval 20-280[degrees]C, its weight loss is close to zero. In the DTA curve, there is an endothermic peak at T ~120[degrees]C, apparently corresponding to the sample melting; on further heating, a broad exothermic effect is observed (120-350[degrees]C), suggesting the occurrence of oxidative polymerization, with appreciable weight loss observed neither for the diamine nor for its polymerization products. This means that fast oxidative polymerization (OP) of 1,3-BAPB starts virtually immediately after melting (from 130-140[degrees]C). Hence, OP can occur simultaneously with acylation of this diamine with anhydride groups of the ROL copolymer. Active decomposition of 1,3-BAPB cured in this temperature interval is observed at temperatures exceeding 400[degrees]C. According to the temperature dependence of weight loss, the resistance of this sample to thermal oxidative degradation is appreciably lower than that of the cured ROL-II and its mixtures with 1,3-BAPB.
Standard TGA and DTA experiments are usually performed at a high heating rate (10 deg * [min.sup.-1]), at which the differences in the thermal characteristics of related polymer materials are largely leveled off. Nevertheless, it can be noted that the TG curves of ROL-II-1,3-BAPB are shifted toward higher temperatures compared to neat ROL-II. The temperature peaks in the DTA curves for the regions of initial (425 [+ or -] 10 [degrees]C) and deep (610 [+ or -] 4 [degrees]C) thermal oxidative degradation virtually coincide, suggesting similar mechanisms of the degradation of these materials.
One of the goals of this study was to confirm usefulness of the novel method for structural and chemical modification of microheterogeneous network polymers. We investigated thermochemical curing reactions of unsaturated rolivsan thermosetting resins with aromatic diamines that lead to increase in cohesion between morphological formations (grains and interlayers) in the ROL copolymer. Combination of FTIR and [sup.13]C NMR spectroscopy was used to determine the chemical structure of the cured ROL-ADA blends. IR spectra of the blends demonstrate wide absorption bands typical of amide N-H stretching vibrations in polyamic acid and carbonyl absorption peaks similar to those of amide and imide groups. There is a good agreement between the FTIR and NMR data (the presence of carbonyl groups in the samples was confirmed both by >C=O valence vibrations and by signals of [sup.13]C carbonyl atoms). However, some of the signals either are absent in the spectra of the studied blends or overlap with wide bands (carbon signals) attributed to the adjacent groups. Thus, it is concluded that the increase in cohesion between morphological formations in microheterogeneous ROL copolymer occurs at the expense of formation of additional amide fragments (polyamic acid) rather than imide crosslinks.
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that there is no conflict of interest regarding the publication of this paper.
This research was performed as part of the employment of the authors at the Institute of Macromolecular Compounds of Russian Academy of Sciences, supported by the Ministry of Science and Higher Education, Russian Federation.
 B. A. Zaitsev and I. D. Shvabskaya, "High-temperature properties of rolivsan thermosetting resins (network copolymers of (di)vinylaromatic ethers and cyclized (di)methacrylates)," Journal of Polymer Research, vol. 22, no. 7, p. 127, 2015.
 B. A. Zaitsev, I. D. Shvabskaya, and L. G. Kleptsova, "Progress in improving high-temperature properties of thermosetting resins: development of the microheterogeneous model for the formation of crosslinked rolivsan-epoxy blends," Journal of Polymer Research, vol. 23, no. 6, p. 111, 2016.
 B. A. Zaitsev, "Combination of polymerization and polycondensation in the synthesis, chemical modification, and cure of rolivsan thermosetting resins," High Performance Polymers, vol. 30, no. 2, pp. 211-223, 2018.
 G. V. Korolev and M. M. Mogilevich, Three-Dimensional Free-Radical Polymerization, Springer-Verlag, Berlin, Heidelberg, 2009.
 B. A. Zaitsev, I. D. Shvabskaya, and L. G. Kleptsova, "Novel polycondensation method of improving high-temperature properties of microheterogeneous rolivsan copolymers modified by inserting epoxy and imide bridges between spherical microdomains," High Performance Polymers, vol. 29, no. 6, pp. 636-645, 2017.
 B. A. Zaitsev, I. D. Shvabskaya, and L. G. Kleptsova, "High-temperature transformations of aromatic diamines in the rolivsan matrix," Russian Journal of Applied Chemistry, vol. 90, no. 6, pp. 946-955, 2017.
 B. A. Zaitsev, I. D. Shvabskaya, and L. G. Kleptsova, "Heat-resistant network block copolymers based on rolivsans modified with tetracarboxylic anhydrides and aromatic tetraamines," Russian Journal of Applied Chemistry, vol. 91, no. 6, pp. 1029-1034, 2018.
 S.-H. Hsiao and C.-Y. Yang, "Preparation and properties of polymides, polyamides and poly(amide-imide)s based on a spirobichroman dietheramine," Macromolecular Chemistry and Physics, vol. 198, no. 7, pp. 2181-2195, 1997.
 E. Mallakpour Sh, A. R. Hajipour, and R. Vahabi, "Synthesis and characterization of novel poly(amide imide)s based on bis(p-amidobenzoic acid)-n-trimellitylimido-L-leucine," Journal of Applied Polymer Science, vol. 84, no. 1, pp. 35-43, 2002.
 S. Mallakpour and A. Zadehnazaria, "Synthesis of novel nanostructured chiral poly(amide-imide)s containing dopamine and natural amino acids," Journal of Chemical Sciences, vol. 125, no. 1, pp. 203-211, 2013.
 M. J. M. Abadie, High Performance Polymers--Polyimides Based. From Chemistry to Applications, InTech, Rijeka, 2012.
 C. A. Pryde, "IR studies of polyimides. I. Effects of chemical and physical changes during cure," Journal of Polymer Science Part A: Polymer Chemistry, vol. 27, no. 2, pp. 711-724, 1989.
 M. K. Ghosh and K. L. Mittal, Polyimides: Fundamentals and Applications, Dekker, New York, NY, USA, 1996.
 W. Y. Chang, S. H. Chen, C. H. Yang, C. N. Chuang, C. K. Wang, and K. H. Hsieh, "Preparation and characterization of aromatic polyimides derived from 4,4,-oxydiphthalic anhydride and 4,4,-diaminodiphenylmethane with different alkyl substituents," Journal of Polymer Research, vol. 22, no. 3, p. 38, 2015.
 S. Ni, J. Tang, F. Wang, and L. Shen, "[sup.13]C n.m.r. characterization of soluble polyaniline," Polymer, vol. 33, no. 17, pp. 3607-3610, 1992.
 M. Hatamzadeh, A. Mahyar, and M. Jaymand, "Chemical modification of polyaniline by N-grafting of polystyrenic chains synthesized via nitroxide-mediated polymerization," Journal of the Brazilian Chemical Society, vol. 23, no. 6, pp. 1008-1017, 2012.
 G. Mertz, T. Fouquet, C. Becker, F. Ziarelli, and D. Ruch, "A methacrylic anhydride difunctional precursor to produce a hydrolysis-sensitive coating by aerosol-assisted atmospheric plasma process," Plasma processes and polymers, vol. 11, no. 8, pp. 728-733, 2014.
 R. M. Souto-Maior, M. I. B. Tavares, and E. E. C. Monteiro, "Solid-state [sup.13]C NMR study of methyl methacrylate-methacrylic acid copolymers," Annals of Magnetic Resonance, vol. 4, pp. 69-72, 2005.
 A. M. Aerdts, Microstructure of Styrene Methyl Methacrylate Copolymers Grafted onto Polybutadiene Seeds, Technische Universiteit Eindhoven, Eindhoven, 1993.
 C. R. Martins, F. Hallwass, Y. M. B. De Almeida, and M. A. De Paoli, "Solid-state [sup.13]C NMR analysis of sulfonated polystyrene," Annals of Magnetic Resonance, vol. 6, pp. 46-55, 2007.
 E. M. Woo, I. C. Chou, L. L. Chang, and H. M. Kao, "Solid-state NMR characterization on the molecular-level homogeneity in lower critical solution temparature mixtures of poly([alpha]-methyl styrene) and poly(2,6-dimethyl-p-phenylene oxide)," Polymer Journal, vol. 35, no. 4, pp. 372-378, 2003.
 J. Roychen, W. T. Ford, S. Zhang, M. P. Tsyurupa, A. V. Pastukhov, and V. A. Davankov, "Solid-state 13C-NMR analysis of hypercrosslinked polystyrene," Journal of Polymer Sciences A: Polymer Chemistry, vol. 35, no. 4, pp. 695-701, 1997.
 G. Robertson, M. Guiver, M. Yoshikawa, and S. Brownstein, "Structural determination of Torlon[R] 4000T polyamide-imide by NMR spectroscopy," Polymer, vol. 45, no. 4, pp. 1111-1117, 2004.
 B. Deng, S. Zhang, C. Liu et al., "Synthesis and properties of soluble aromatic polyimides from novel 4,5-diazafluorene-containing dianhydride," RSC Advances, vol. 8, no. 1, pp. 194-205, 2018.
 E. Schab-Balcerzak, L. Grobelny, A. Sobolewska, and A. Miniewicz, "Cycloaliphatic-aromatic polyimides based on diamines with azobenzene unit," European Polymer Journal, vol. 42, no. 10, pp. 2859-2871, 2006.
 D. T. Padavan and W. K. Wan, "Synthesis and characterization of a novel versatile poly(amic acid) derived from ethylenediaminetetraacetic dianhydride," Materials Chemistry and Physics, vol. 124, no. 1, pp. 427-433, 2010.
 N. A. Mokri, P. C. Oh, H. Mukhtar, and T. L. Chew, "Synthesis of high molecular weight polyimide consisting hexafluoroisopropylidene moiety for gas separation," Journal of Mechanical Engineering and Sciences, vol. 11, no. 4, pp. 3144-3151, 2017.
 X. L. Meng, L. Jiang, and Y. Huang, "Characterization and thermal degradation of polyimide derived from 5-amino-2(paminophenyl) benzoxazole monomer with pyromellitic dianhydride," Advanced Materials Research, vol. 455-456, pp. 806-814, 2012.
 A. Martinez-Richa and R. Vera-Graziano, "A solid-state NMR study of aromatic polyimides based on 4,4'-diaminotriphenylmethane," Journal of Applied Polymer Science, vol. 70, no. 6, pp. 1053-1064, 1998.
B. A. Zaitsev [ID], L. G. Kleptsova, and I. D. Shvabskaya
Institute of Macromolecular Compounds, Bolshoy Prospect 31, Russian Academy of Sciences, St. Petersburg 199004, Russia
Correspondence should be addressed to B. A. Zaitsev; firstname.lastname@example.org
Received 14 November 2018; Accepted 13 January 2019; Published 9 April 2019
Academic Editor: Nobuhiro Kawatsuki
Caption: Figure 1: FTIR spectra of the cured neat ROL-II (I-IV) samples prepared by heating in the thermal regime up to the following final temperatures [T.sub.fin.cur.]: 1: 320[degrees]C; 2: 260[degrees]C; 3: 220[degrees]C; 4: 180[degrees]C.
Caption: Figure 2: FTIR spectra of the cured neat ROL-II resin (1) and the cured RII-1,3-PDA (II-IV) blends prepared by heating in the thermal regime up to [T.sub.fin.cur.] ([degrees]C): 260 (4), 320 (3), and 350 (2).
Caption: Figure 3: FTIR spectra of the cured ROL-II-1,3-BAPB (V-VIII) blends obtained by heating up to the following final temperatures: 180[degrees] (1), 220[degrees] (2), 260[degrees] (3), and 320[degrees] (4)
Caption: Figure 4: FTIR spectra of heat-treated 1,3-BAPB (XV) (1) and the following cured blends: ROL-II-1,3-BAPB (XIV) (2), ROL-II-1,3-BAPB (VIII) (3), ROL-II-1,3-BAPB (XIII) (4), and neat ROL-II (IV) (5).
Caption: Figure 5: Chemical structures of the neat and restructurized ROL copolymers.
Caption: Figure 6: [sup.13]C NMR spectra of samples of neat ROL-II (I-IV) resins cured according to the thermal regime with [T.sub.fin.cur.] increasing from 180[degrees] to 320[degrees] C.
Caption: Figure 7: Schematic representation of interaction between ADA and restructurized ROL copolymers; [sup.13]C NMR spectra of the samples of ROL-II- (85 wt.%) 1,3-BAPB (15 wt.%) cured in the thermal regime up to the following [T.sub.fin.cur.] ([degrees]C): 1: 320; 2: 260; 3: 220; 4: 180.
Caption: Figure 8: [sup.13]C NMR spectra of the samples of ROL-II- (50wt.%) 1,3-BAPB (50wt.%) (XIII) blends (1), ROL-II (100%) (2), and the 1,3-BAPB sample oligomerized in the regime with [T.sub.fin.cur.] = 320[degrees]C (3).
Caption: Figure 9: [sup.13]C NMR spectra of ROL-II- (75wt.%) m-PDA (25wt.%) blends, cured in the thermal regime up to the following [T.sub.fin.cur.] ([degrees]C): 1: 350; 2: 320; 3: 260.
Caption: Figure 10: Temperature dependencies of storage Young's modulus E' of the cured ROL-I-1,3-PDA (I) blend (1) and the cured neat ROL-I (2). The DMA experiments were performed at a frequency (v) of ~ 1.0 Hz.
Caption: Figure 11: Temperature dependencies of storage Young's modulus E' and tan [delta] of the ROL-I-3,4,-ODA blend (1) and the cured neat ROL-I (2). The DMA experiments were carried out in a tension mode at 660 Hz.
Caption: Figure 12: Temperature dependences of storage Young's modulus E' and tan [delta] of ROL-I-1,4-BAPB (1) and neat ROL-I (2). The DMA experiments were carried out at 660 Hz in a tension mode.
Caption: Figure 13: Time dependences of weight loss (100[DELTA]m/m) of ROL-II-1,3-BAPB (I-X) blends and oligomerized 1,3-BAPB (XI) prepared by heating in the regime with [T.sub.fin.cur.] =320[degrees]C. Thermal oxidative aging was performed in air at 350[degrees]C. 1,3-BAPB content (wt.%): (1) 100, (2) 50, (3) 0, (4) 35, and (5) 10.
Caption: Figure 14: TGA (1-3) and DTA (1'-3') curves of the (1, 1') monomeric 1,3-BAPB sample heat treated in air (on a thermobalance) at a heating rate of 10 deg-min-1, (2, 2') ROL-II sample (100%), and (3, 3') sample of the ROL-II (75%)-1,3-BAPB (25%) blend, both cured using the temperature schedule with [T.sub.fin.cur] = 320[degrees]C.
Table 1: Characteristics of uncured rolivsan thermosetting resins. Sample [n.sup.20.sub.D] [M.sup.a.sub.n] [[tau].sup.t * Cb] (min) ROL-I 1.581 430 [65.sup.140] ROL-II 1.592 470 [82.sup.155], [17.3.sup.170] Sample Unsaturated groups (mol per 100 g of resin)  [CH.sub.2] = [CH.sub.2] = -CH = CH- C([CH.sub.3])- CH-(trans) ROL-I 0.274 0.239 0.120 ROL-II 0.283 0.192 0.109 (a) Number average molar mass. (b) Gelation time at a given temperature (t, [degrees]C). Table 2: Contents of ADA in blend samples and their final curing temperatures [T.sub.fin.cur.] Sample ADA (wt.%) [T.sub.fin.cur.] ([degrees]C) ROL-I 0.0 300 ROL-I-1,3-PDA (I) 12.0 300 ROL-I-3,4'-ODA 7.0 260 ROL-I-1,4-BAPB 10.0 260 ROL-II (I) 0.0 180 ROL-II (II) 0.0 220 ROL-II (III) 0.0 260 ROL-II (IV) 0.0 320 ROL-II-1,3-PDA (II) 25.0 260 ROL-II-1,3-PDA (III) 25.0 320 ROL-II-1,3-PDA (IV) 25.0 350 ROL-II-1,3-BAPB (I) 3.0 320 ROL-II-1,3-BAPB (II) 5.0 320 ROL-II-1,3-BAPB (III) 7.0 320 ROL-II-1,3-BAPB (IV) 10.0 320 ROL-II-1,3-BAPB (V) 15.0 180 ROL-II-1,3-BAPB (VI) 15.0 220 ROL-II-1,3-BAPB (VII) 15.0 260 ROL-II-1,3-BAPB (VIII) 15.0 320 ROL-II-1,3-BAPB (IX) 25.0 320 ROL-II-1,3-BAPB (XI) 30.0 320 ROL-II-1,3-BAPB (XII) 35.0 320 ROL-II-1,3-BAPB (XIII) 50.0 320 ROL-II-1,3-BAPB (XIV) 50.0 350 1,3-BAPB (XV) 100.0 320 Table 3: Assignment of CP/MAS [sup.13]CNMR chemical shifts of neat ROL copolymers and model polymers. ROL copolymer [T.sub.fin.cur.] >C=0 (h, j, m) ([degrees]C) ROL-II (I) 180 176.6; 175.4 ROL-II (II) 220 184.5; 182.9; 179.2; 177.0 175.8; 173.6 ROL-II (III) 260 185.1:184-183; 179.4 177.9;176.4; 172.4; 169.9 ROL-II (IV) 320 185.2; 183.0; 179.7; 178.0 176.0; 172.3; 170.8 PMMA  -- 178.1 Neat PMAA  -- 180.7 Neat PMA  -- 172.8 CPDA  169.29; 173.34 Neat PS  -- -- SMMA copolymer [16, 19] -- 175;176; 177 Pa-ms-PPO (f)  -- -- ROL copolymer [delta] (ppm from TMS) >C< (g) [C.sub.Ar]H (a) (c) ROL-II (I) 46.7 128; 122 ROL-II (II) 46.4 128; 122 ROL-II (III) 44.4 128; 122 ROL-II (IV) 43.9 128; 122 PMMA  45.2 -- Neat PMAA  44.2 -- Neat PMA  -- CPDA  -- -- Neat PS  -- 127 SMMA copolymer [16, 19] 44-47 128-126 Pa-ms-PPO (f)  45 127 ROL copolymer [delta] (ppm from TMS) [C.sub.Ar-] (b) (b) [C.sub.Ar]-O- (d, e) ROL-II (I) 140 154; 158 ROL-II (II) 140 154; 158 ROL-II (III) 140 154; 158 ROL-II (IV) 140 154; 158 PMMA  -- -- Neat PMAA  -- -- Neat PMA  -- -- CPDA  -- -- Neat PS  146 -- SMMA copolymer [16, 19] 143-145 -- Pa-ms-PPO (f)  145 155 ROL copolymer [delta] (ppm from TMS) >CH- (c) (a) -[CH.sub.2] (d) (f) ROL-II (I) 40 52-56 ROL-II (II) 40 53.8 ROL-II (III) 40 56.0 ROL-II (IV) 40 53-55 PMMA  52.2 Neat PMAA  53.2 Neat PMA  CPDA  39.79 -- Neat PS  40 SMMA copolymer [16, 19] 38-39 Pa-ms-PPO (f)  -- 57-63 ROL copolymer [delta] (ppm from TMS) [alpha]-[CH.sub.3] (e) (i) ROL-II (I) 18.9 ROL-II (II) 19.4 ROL-II (III) 19.8 ROL-II (IV) 19.8;15.1 PMMA  16.7 Neat PMAA  16.0 Neat PMA  CPDA  -- Neat PS  -- SMMA copolymer [16, 19] 25-15 Pa-ms-PPO (f)  -- (a) Protonated aromatic carbons. (b) Quaternary aromatic carbons. (c) Aliphatic methine carbons. (d) Aliphatic methylene carbons give a broad signal at 40-50 ppm . (e) [alpha]-Methyl carbons in polymethacrylic esters. (f) Poly([alpha]-methylstyrene) (P[alpha]-ms)-poly(2,6-dimethyl-p-phenyloxyde) (PPO).
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Research Article|
|Author:||Zaitsev, B.A.; Kleptsova, L.G.; Shvabskaya, I.D.|
|Publication:||International Journal of Polymer Science|
|Date:||Apr 30, 2019|
|Previous Article:||Evaluation of Fatty Acid Waste in the Synthesis of Oligo(Ether-Ester)s.|
|Next Article:||Chitosan-Based Nanogel Enhances Chemotherapeutic Efficacy of 10-Hydroxycamptothecin against Human Breast Cancer Cells.|