Epoxidized Block and Statistical Copolymers Reinforced by Organophosphorus-Titanium-Silicon Hybrid Nanoparticles: Morphology and Thermal and Mechanical Properties.

Glycidyl methacrylate (GMA) and a mixture of alkyl methacrylates (average chain length of 13 carbons; termed C13MA; derived from vegetable oils) were copolymerized by nitroxide-mediated polymerization to form epoxidized statistical and block copolymers with similar compositions (F GMA ∼0.8), which were further cross-linked by a bio-based diamine. Hybrid plate-like nanoparticles containing organophosphorus-titanium-silicon (PTS) with an average size of ∼130 nm and high decomposition temperature (485 °C) were synthesized via a hydrothermal reaction to serve as additives to simultaneously enhance the thermal and mechanical properties of the blend. Nanocomposites filled with PTS were prepared at different filler-loading levels (0.5, 2, 4 wt %). Transmission electron microscopy (TEM) of the cured block copolymer displayed reaction-induced macrophase-separated domains. TEM also showed an effective dispersion of PTS hybrids in the matrix without intense agglomeration. Thermogravimetric analysis at different heating rates revealed the activation energy of poly (GMA-stat-C13MA) at maximum decomposition increased from 143.5 to 327.2 kJ mol-1 with 4 wt % PTS. Decomposition temperature and char residue improved 12 °C and ∼7 wt %, respectively, and T g increased 12 °C by adding 4 wt % PTS. Targeting various PTS concentrations enabled tuning of the tensile modulus (up to 75%), tensile strength (up to 46%), and storage modulus in both glassy state (up to 59%) and rubbery plateau regions (up to 88%). Oscillatory frequency sweeps indicated that PTS makes the storage modulus frequency dependent, suggesting that the inclusion of the nanoparticles alters the relaxation of the surrounding matrix polymer.

Introduction

Polymers impart flexibility and desirable mechanical, electrical, and thermal properties to a myriad of products, but they also require flammability reduction and thermal stability improvement to be adopted for those applications. Various inherently thermally stable polymers like fluoropolymers and poly(vinylchloride) are often substituted by more flammable polymers because of recycling challenges, costs, or environmental concerns for the elimination of certain compounds such as heavy metals or halogens in materials waste.[1,2] In this context, the fabrication of polymer nanocomposites filled with organic–inorganic hybrid nanoadditives compatible with the polymer matrix for thermostability enhancement with fewer environmental repercussions has warranted considerable attention.[3]

To date, a large number of nanofillers, including carbon materials,[4,5] phosphoric compounds,[6−9] siliceous compounds,[10−13] and minerals,[14−16] have been used to improve both thermal and mechanical properties of polymers. Nanoadditives can be applied at much lower concentrations (<10 wt %) compared to traditional fillers with markedly observed enhancement in mechanical properties, conductivity, and solvent resistance.[17,18] This is due to the “nano-effect” causing changes in the local properties of the matrix and the small distances between nanofillers particles even at low mass loadings along with an extremely high surface area of nanofillers.[18] For instance, the addition of 1 wt % rod-shaped carbon nanofibers (∼200 nm in diameter) in polyurethane foams simultaneously improved thermal degradation onset by 18 °C and enhanced flexural modulus from 143 to 207 MPa.[19] Incorporating polyhedral oligomeric silsesquioxanes (POSS) at low loadings into epoxy resins not only increased the limiting oxygen index (LOI) by ∼7% but also improved the oxidation resistance of the resin.[20,21] In a recent study, we showed that POSS nanoparticles improved both tensile modulus (from 96 to 179 MPa) and thermal decomposition temperature (225–255 °C) in recyclable thermosets based on vinylogous urethane cross-linking networks.[22] Addition of clay to polymers contributes to long-term thermo-oxidative stability and thermal stability since clay can act as both a heat insulator and a mass transport barrier to oxygen and the volatiles generated during the degradation.[23,24]

During the past several years, hybridization of inorganic–organic materials has brought a new perspective to the development of high-performance materials.[25,26] Incorporation of inorganic–organic nanoadditives has been widely used for improving the thermal stability of polymers because of the synergistic effect between their inorganic and organic components.[27] In addition, the presence of organic moieties plays an important role in the morphology, dispersity, and thermal stability of such hybrid nanofillers.[28−30] Organophosphorus compounds, such as ammonium polyphosphate[31] and diphenylphosphinic acid[32] have been used as inorganic additives for improving the thermal stability and flame resistance of polymeric materials. There have been several studies concerning hybrid nanomaterials such as aluminum–organophosphorus,[30] phosphorus-doped silica,[33,34] silica–graphene oxide,[35,36] and silica–carbon[37] that have exhibited marked thermostability. For instance, only 2 wt % of Cu2O-TiO2-GO nanosheets reduced the heat release rate of unsaturated polyester by ∼30%.[38] Epoxy resin reinforced with 5 wt % Ti-Si exhibited 13 °C increase in glass transition temperature (Tg), 16 °C increase in decomposition temperature, and enhanced storage modulus compared to pure epoxy.[39] Recently, Wei et al.[40,41] reported hybrid nanofillers that contained organophosphorus–titanium–silicon (PTS) to fabricate high performance flame-retardant polycarbonate. Only 0.1 wt % of the nanoparticles endowed polycarbonate with a LOI of 29.7%, 8% improvement in tensile strength, and 8.3% improvement in elongation at break. Such an improvement in both mechanical and thermal stability was ascribed to the uniform dispersion of PTS and suitable interfacial interactions with the polymeric matrix.

Inspired by Wei et al.[40,41] we prepared organic–inorganic hybrid PTS nanoparticles with a lower organophosphorus content. We used PTS to simultaneously enhance the mechanical and thermal stability properties of epoxy-functionalized methacrylate thermosets. We synthesized statistical and block copolymers with a similar composition of glycidyl methacrylate (GMA) and bio-based C13 methacrylate (C13MA, from vegetable oils) using nitroxide-mediated controlled radical polymerization (NMP) (Figure ). Since the incorporation of vegetable oil-derived monomers in the polymer chains affects the strength of GMA, we decided to reinforce the copolymer by adding low amounts of PTS with potential applications in adhesive and antiflammable coatings. We targeted statistical copolymers to introduce epoxy functionality randomly throughout the chain while a block copolymer was used to introduce localized epoxy segments. Previously, many efforts have been devoted to toughen epoxy resins with various renewable polymer additives such as triscardanyl phosphate[42] and bio-based polymers derived from castor oil,[43] vanillin,[44] and soybean oil.[45] Additionally, the synthesis of copolymers containing long chain alkyl methacrylates provides low-Tg, soft, and elastic domains capable of sustaining deformation through viscous flow and chain entanglements.[46−48] Thus, in the present study, we incorporated a low concentration (∼20 mol %) of a rubbery phase (C13MA, whose polymer has a Tg ∼ −45 °C) to increase the toughness of GMA. The synthesized polymers were then converted to thermosets via an epoxy-amine cross-linking reaction using a dimeric amine. Finally, the influence of incorporation of PTS on the morphology, and thermal, mechanical and rheological properties was investigated and discussed accordingly.

Figure 1

NMP scheme of poly(GMA-stat-C13MA) and poly(GMA-b-C13MA) using Dispolreg 007.

Experimental Section

Materials

Tetrabutyl titanate (TBT, 97%), tetraethyl orthosilicate (TEOS, 99%), diphenylphosphinic acid (DPPA, 98%) and 1,4-dioxane (≥99.9%) were purchased from Millipore-Sigma. GMA (Sigma, 97%) and C13 methacrylate (C13MA, a mixture of alkyl methacrylates (∼C10–C16), Evonik) were passed through a column of basic alumina to remove the inhibitors. 3-(((2-Cyanopropan-2-yl) oxy) - (cyclohexyl) amino)-2 and 2-dimethyl-3-phenylpropanenitrile (Dispolreg 007) initiator were synthesized following Ballard et al.’s method.[49] Priamine 1075 was received from Croda. Ethanol, toluene, tetrahydrofuran (HPLC, ≥99.9%), and methanol were purchased from Fisher Scientific. Deuterated chloroform (CDCl3) was received from Cambridge Isotopes Laboratory.

Polymer Synthesis

We synthesized statistical and block copolymers of GMA and bio-based C13MA with similar compositions by NMP. For poly(GMA-stat-C13MA), GMA (54.0 g, 0.38 mol), C13MA (25.5 g, 0.09 mol), Dispolreg 007 (0.90 g, 2.65 mmol), and dioxane (80.5 g) were added to a 250 mL three-neck flask. After purging the solution with nitrogen for 30 min it was heated at 90 °C for 4.5 h. Samples taken during the reaction were analyzed by gel permeation chromatography (GPC) and proton nuclear magnetic resonance (1H NMR) spectroscopy. At the end of the reaction, the solution was precipitated into methanol and dried in vacuum at room temperature for 24 h. For the synthesized poly(GMA-stat-C13MA), the final conversion was 65%, Đ = 1.52, Mn = 21.8 kg mol–1, and FGMA = 0.81.

To prepare poly(GMA-b-C13MA), we first synthesized the poly(GMA) macroinitiator by adding GMA (40.0 g, 0.28 mol), Dispolreg 007 (0.4 g, 1.18 mmol), and dioxane (40 g) to a flask and after purging with nitrogen, the solution was heated at 90 °C for 1.5 h (X = 32%). After precipitation from methanol, the product was dried in a vacuum oven overnight. For the poly(GMA) macroinitiator, Mn= 19.1 kg mol–1 and Đ = 1.55. Next, poly(GMA) (14.1 g, 0.73 mmol), C13MA (10.8 g, 0.04 mol), and dioxane (24.9 g) were added into a 125 mL flask and the same procedure at 90 °C was followed and the reaction took 3 h. For the synthesized poly(GMA-b-C13MA), the final conversion was 32%, Mn= 25.3 kg mol–1, Đ = 1.87, and FGMA = 0.83.

Nanoparticle Synthesis

Following the approach by Wei et al.,[40,41] we synthesized PTS nanoparticles by applying the same method but with a lower organic content to decrease the decomposition at high temperatures, which resulted in different nanoparticle structures. We first added TBT (6.66 mmol) and TEOS (6.66 mmol) in a mixed solution of ethanol/deionized water (0.13 mol/0.13 mol) containing a few drops of acetyl acetone to prevent the hydrolysis of TBT. After 1 h, DPPA (6.66 mmol) was added and mixed in the Ti-Si sol gel, and the mixture was transferred to an oven at 160 °C for 12 h. Finally, the product was washed sequentially with ethanol, deionized water, ether, and acetone followed by drying in a vacuum oven at 80 °C overnight.

Nanocomposite Synthesis

We prepared poly(GMA-stat-C13MA)/PTS and poly(GMA-b-C13MA)/PTS nanocomposites by a solvent casting method. The sample ID and formulation of nanocomposites are provided in Table . As an example, for S-2 which contains 2 wt % PTS, poly(GMA-stat-C13MA) (4.0 g, Mn = 21.8 kg mol–1, and FGMA = 0.81) was first dissolved in toluene (4.0 g). In a separate vial, PTS nanoparticles (80.3 mg) were dispersed in toluene (4.0 g) by sonication using a Hielscher sonicator UP200S (50% duty cycle and amplitude 70%) for 15 min and then added to the polymer. The sonicated energy employed for nanoparticle dispersion can be obtained according to the literature.[50] The polymer solution containing PTS nanoparticles was stirred for 10 min followed by adding cross-linker solution (0.8 g of Priamine 1075 in 1.0 g of toluene) and further stirring for 1–2 min. The prepared mixture was cast into the silicone molds with rectangular, dumbbell, and circular shaped molds. The samples were dried and cured at room temperature for 5 days and then post-cured in an oven at 40 °C under vacuum for 6 h to ensure full dryness and curing. Figure S1 shows the picture of the cured nanocomposites.

Table 1

Formulation of Poly(GMA-stat-C13MA)/PTS and Poly(GMA-b-C13MA)/PTS Nanocomposites

sample ID	polymer (wt %)	PTS (wt %)
S-0	poly(GMA-stat-C13MA) (100)	0
S-0.5	poly(GMA-stat-C13MA) (99.5)	0.5
S-2	poly(GMA-stat-C13MA) (98)	2
S-4	poly(GMA-stat-C13MA) (96)	4
B-0	poly(GMA-b-C13MA) (100)	0
B-4	poly(GMA-b-C13MA) (96)	4

Measurements

The characterization of the synthesized polymers in terms of molecular weight (Mn) and dispersity Đ was carried out using GPC (Waters Breeze) with tetrahydrofuran (HPLC grade) as the eluent at 40 °C at a flow rate of 0.3 mL min–1 with polymethyl methacrylate (PMMA) standards. Three Waters HR Styragel GPC columns equipped with a guard column were used. More details on the instrument and the columns were given previously.[51] Dynamic light scattering (DLS) was conducted using a Malvern Zetasizer Nano ZS equipped with a 4 mW He–Ne laser at 633 nm and an avalanche photodiode detector. The measurement was conducted at 25 °C at 173° angle on diluted samples (concentration of 0.01–1000 mg mL–1). X-Ray photoelectron spectroscopy (XPS) was conducted using a Fischer Scientific Kα spectrometer using a spot size of 200 μm, running 5 survey scans at 200 mV for 50 ms residence times, and 10 scans for 50 ms residence times for specific elements. Differential scanning calorimetry (DSC, Q2000, TA Instruments) and thermogravimetric analysis (TGA, Q500, TA Instruments) were performed at a rate of 10 °C min–1 under nitrogen using aluminum pans. 1H NMR spectra were recorded using a Varian NMR Mercury spectrometer at 32 scans and 300 MHz with deuterated chloroform solvent. Fourier transform infrared spectroscopy (FTIR) spectra were obtained using a Bruker Alpha FTIR. Dynamic mechanical analysis (DMA) and rheology measurements were performed using an Anton Paar MCR 302 in dynamic oscillatory mode (0.1–100 Hz, amplitude 1%) with a CTD 450 oven and 25 mm parallel plate geometry under nitrogen at 100 °C. DMA was carried out on rectangular-shaped samples (60 mm length, 10 mm width and 2 mm thickness) at a rate of 5 °C min–1 and a frequency of 1 Hz. Tensile testing was performed using an MTS Insight material testing system on dumbbell-shaped specimens (ASTM D638 type V, length = 60 mm, width = 10 mm) at a cross-head speed of 10 mm min–1and a 5 kN load cell. Transmission electron microscopy (TEM) of nanocomposites and the neat PTS were carried out on ultramicrotome-cut 1 μm-thick sections on conductive grids without chemical staining using Thermo Scientific Talos F200X G2 (S/TEM) at an acceleration voltage of 200 kV. This instrument allowed high-angle annular dark-field (HAADF) in scanning mode (STEM) images to be achieved with improved contrast within the macrophase-separated material.

Results and Discussion

Synthesis and Characterization of Poly(GMA-stat-C13MA) and Poly(GMA-b-C13MA)

We targeted methacrylate-based statistical and block copolymers containing epoxy functional groups, which are widely used in coatings.[52,53] We adopted NMP and synthesized poly(GMA-stat-C13MA) using Dispolreg 007 initiator at 90 °C with a high initial GMA fraction, fGMA,0 0.8 and targeting Mn,theo = 30 kg mol–1 (molecular weight at 100% conversion). As shown in Figure , the plot of monomer conversion versus time revealed pseudo first-order kinetics of both C13MA and GMA. In addition, the growth of polymer molecular weight remained consistent with monomer conversion and Đ was consistently low (1.33 < Đ <1.52), superficially indicating a controlled polymerization.

Figure 2

NMP of GMA/C13MA at 90 °C in dioxane solvent: (a) semi-logarithmic kinetic plot of ln[1/(1-X)] (X = monomer conversion) versus reaction time and (b) Mn, and Đ versus conversion. For the final poly(GMA-stat-C13MA), final X = 65%, Đ = 1.52, Mn = 21.8 kg mol–1, and FGMA = 0.81.

In addition to the synthesis of poly(GMA-stat-C13MA), we also prepared a diblock copolymer by NMP. We intentionally targeted a low C13MA concentration for the second block to achieve a similar overall composition to the statistical copolymer. Using this approach, a poly(GMA) macroinitiator (Mn= 19.1 kg mol–1, Đ = 1.55) was synthesized beforehand, which then initiated the polymerization of C13MA leading to poly(GMA-b-C13MA). After precipitation into methanol, the block copolymer was characterized by GPC to indicate that the second block had successfully grown from the macroinitiator. Figure presents a moderate shift of Mn toward higher molecular weights because of the relatively high Mn of the poly(GMA) macroinitiator and the small poly(C13MA) segment. The high Đ = 1.87 is due to the low molecular weight tail indicative of some dead macroinitiator present. We carried out multiple Gaussian peak fitting for the macroinitiator to roughly estimate the fraction of dead polymer chains. According to the fitting analysis in Figure S2, 13% of the macroinitiator chains were estimated to be dead. We discussed the effect of the dead chains and high Đ on the morphology of the block copolymer in Section .

Figure 3

GPC chromatogram of the poly(GMA) macro-chain transfer agent and poly(GMA-b-C13MA) diblock copolymer (THF eluent at 0.3 mL min–1 with PMMA standards at 40 °C).

Characterization of Synthesized PTS

DLS, TGA, FTIR, and TEM were used to characterize the synthesized PTS nanoparticles. Figure a shows the average particle size of PTS ∼130 nm, obtained from DLS. TGA of PTS in Figure b indicates a very low weight loss (<10%) as heating increased from 50 to 485 °C which could be attributed to the release of bound water in nanoparticles. With increasing temperature, PTS starts to decompose through a single-step decomposition with a DTG peak appearing at about 560 °C which is mainly involved in the degradation of organic functional groups (organophosphorus moieties) and ultimately having a relatively high residue (54.6% at 800 °C).[54,55] The chemical structure of PTS was analyzed by FTIR as shown in Figure c. Peaks at 1088 and 1050 cm–1 correspond to the Ti-O-P and Si-O-P, respectively. Additionally, the vibration peaks of Si-O and Ti-O exist at 795 and 566 cm-1.[41,56] The peak around 3500 cm–1 is ascribed to H2O. To further analyze the structure of PTS, XPS was performed. Figure S3a shows the peaks located at 533.0, 458.8, 284.6, 132.7, and 103.6 eV in the XPS spectrum, corresponding to O 1 s, Ti 2p, C 1 s, P 2p, and Si 2p, respectively. Figure S3b shows the P 2p XPS spectra of the nanoparticles with three peaks at 132.1, 132.7, and 133.4 eV corresponding to P-C, P-O-Ti, and P-O-Si, respectively.

Figure 4

(a) DLS, (b) TGA, and (c) FTIR spectra of the synthesized PTS hybrid nanoparticles.

The microstructure of PTS hybrid nanoparticles was further investigated by TEM. Figure a,b shows the TEM micrographs of PTS which indicate PTS nanoparticles formed irregular amorphous aggregated structures, ranging from 70 to 170 nm in size. Such irregular microstructures of hybrid nanoparticles were previously observed and deemed to be platelet-like.[57−59] This was not as noticeable due to the aggregates caused by the concentrated solution we applied prior to TEM observation. Figure c presents a typical selected area electron diffraction (SAED) pattern for PTS without any diffraction spots or rings observed, indicating an amorphous structure. Further, the elemental composition of PTS was probed by EDX. The EDX results confirmed that P and C both exist in PTS besides the Si, Ti, and O, confirming the formation of hybrid nanoparticles (Figure d).

Figure 5

(a, b) TEM images of the hybrid PTS dispersed in toluene, (c) SAED pattern, (d) EDX result of PTS.

Influence of PTS on the Thermal Behavior of Poly(GMA-stat-C13MA) and Poly(GMA-b-C13MA)

The thermal stability of poly(GMA-stat-C13MA) and poly(GMA-b-C13MA) was evaluated by TGA (Figure a,c) under nitrogen. The TGA thermograms for all the nanocomposites exhibit a two-step degradation at 330 and 400 °C. The weight losses observed at lower temperatures are attributed to the decomposition of side chains of polymers while those at higher temperatures are attributed to main-chain degradation.[60,61] According to Table , the onset degradation temperature increased 12 °C from cured, unloaded poly(GMA-stat-C13MA) (S-0) to the 4 wt % PTS-reinforced nanocomposite (S-4). Additionally, there was an increase in the maximum decomposition temperatures from 317 and 404 °C to 337 and 421 °C, respectively, indicating an improvement in the thermal stability of the nanocomposites. These enhancements in thermal stability are due to the incorporation of thermally stable PTS into poly(GMA-stat-C13MA) where the polymer-nanoparticle interaction is strong.[62] As a comparison, the addition of 1 wt % TiO2, nanoclay, and carbon nanofibers to polyurethane improved the decomposition temperature by 7.1, 15.6, and 17.6 °C, respectively.[19] Furthermore, the residual weight percentage at the conclusion of the TGA tests increased from 2.5 to 9.2 wt %, which is the sum of the inorganic compounds plus the char products from the polymer.[54] In the case of block copolymers, the onset degradation temperature increased 6 °C and the residual weight percentage increased by 5.2 wt % from B-0 to B-4.

Figure 6

Influence of PTS on TGA and DSC of cured (a, b) poly(GMA-stat-C13MA) and (c, d) poly(GMA-b-C13MA).

Table 2

Thermal Characterization of Poly(GMA-stat-C13MA) and Poly(GMA-b-C13MA) Cured Polymers and Nanocomposites

ID	polymer	PTS (wt %)	FGMA	Tg (°C)	T10% (°C)	Tmax,1 (°C)/Tmax,2 (°C)	residue at 600 °C (%)
S-0	poly(GMA-stat-C13MA)	0	0.81	65	293	317/404	2.5
S-0.5	poly(GMA-stat-C13MA)	0.5	0.81	69	298	326/407	4.2
S-2	poly(GMA-stat-C13MA)	2	0.81	74	300	331/415	6.3
S-4	poly(GMA-stat-C13MA)	4	0.81	77	305	337/421	9.2
B-0	poly(GMA-b-C13MA)	0	0.83	–39, 84	298	328/430	4.9
B-4	poly(GMA-b-C13MA)	4	0.83	–33, 91	304	325/426	10.1

DSC curves shown in Figure b,d show that the glass transition temperature (Tg) for the poly(GMA-stat-C13MA) nanocomposite increased from 65 °C (for S-0) to 77 °C (for S-4) with increasing concentration of PTS from 0 to 4 wt %, which indicates that the interaction of polymer chains was enhanced by the incorporation of thermally stable PTS. The uniform dispersion of PTS in the polymer (Section ) restricts the mobility of polymer chains which enhances the Tg of the polymer.[63,64] The literature Tgs of the rubbery poly(C13MA) and rigid poly(GMA) are −45 °C[51,65] and 83 °C,[66] respectively. In the case of block copolymers, the epoxy groups of GMA in poly(GMA-b-C13MA) were able to react with amine end groups of Priamine, so that the block copolymer could cure within the epoxy network with the possibility of microphase separation.[67,68] The microphase separation of poly(GMA-b-C13MA) was observed by DSC. According to Figure d, two different Tg values at approximately −40 and ∼85 °C were detected for poly(GMA-b-C13MA) corresponding to the poly(C13MA) and poly(GMA) blocks, respectively. To confirm microphase separation of poly(GMA-b-C13MA), we estimated the total solubility parameters for C13MA (δ= 15.43 MPa1/2) and GMA (δ= 19.25 MPa1/2) from Hansen solubility parameters.[69] The Hansen solubility parameters of lauryl methacrylate were used to roughly estimate the total solubility parameter, δ, for C13MA as the information for C13MA was not available. The difference in estimated solubility parameters indicates the immiscibility between the poly(C13MA) and poly(GMA) blocks. After adding PTS into diblock copolymer, both Tg peaks shifted toward higher temperatures, increasing by 6 and 7 °C for the poly(GMA) and poly(C13MA) block segments, respectively, compared to the cured poly(GMA-b-C13MA), which was observed for similar block copolymer nanocomposites in the literature.[70]

The cross-linked poly(GMA-b-C13MA), B-0, was examined for phase separation behavior by HAADF-STEM (Figure ) to give contrast depending on the mass density. On the basis of previous studies demonstrating that polymers with higher density have dark contrast by HAADF,[71,72] we propose that the darker regions correspond to the cross-linked GMA segments. The observation of phase-separation based on different mass densities has been previously reported in nonreactive block copolymers[73,74] and cross-linked block copolymers.[75−77]

Figure 7

HAADF-STEM micrographs of cross-linked poly(GMA-b-C13MA) (B-0). Darker domains are due to increased density which is attributed to the cross-linked GMA segments.

The HAADF-STEM image of B-0 demonstrated macrophase-separated domains on scales larger than that expected for microphase separated block copolymer length scales (∼ several radii of gyration (Rg) of the polymers, Rg ∼ 10–50 nm), which could be ascribed to reaction-induced macrophase separation, where the structures of several nanometers to microns can be formed via spinodal decomposition and/or nucleation and growth mechanisms.[78−80] In reaction-induced macrophase separation, there is a competition between the curing speed and mobility (viscosity) at a given temperature, resulting in phase-separated structures with domain lengths ∼ several hundred nanometers.[78,81] In such systems, it is essential to use a good diluent (nonreactive solvent) to attain a homogeneous mixture and subsequently remove the solvent. If the solvent is not evaporated rapidly, the viscosity of the system could be low enough to allow an increase in the nanodomain coarsening rate. This may in turn affect the morphology dimensions and eventually lead to macrophase separation.[82,83] Macrophase-separated epoxidized polymers are usually characterized by opacity because of the light scattering of separated domains with dimensions comparable to the wavelength of visible light. The images of B-0 and S-0 provided in Figure S4 indicate the opacity of the cured poly(GMA-b-C13MA) compared to the transparent poly(GMA-stat-C13MA). It should be noted that according to Section , the diblock copolymer (B-0) in the present study contains a blend of a low fraction of dead poly(GMA) chains and poly(GMA-b-C13MA) chains. Solubilizing the diblock copolymer in toluene before the addition of the cross-linker provides a high mobility and the C13MA block remains free for the macrophase to separate. After the addition of Priamine, the mobility of the C13MA block is inhibited by the rigidity of the cured epoxy groups.[78] For comparison, Grubbs et al.[80] obtained opaque samples with macrophase-separated thermosets using polyisoprene-block-poly(methyl acrylate-co-glycidyl acrylate) copolymers and 4,4′-methylenedianiline hardener. They observed that the macrophase-separated domains were greater than 100 nm to millimeter-scale regions.[80] The incorporation of rubbery modifiers or block copolymers as toughening agents can lead to the formation of macro-phase separated structures in epoxy resins.[78,84] For instance, Pearson et al.[85] observed large multiphase particles (10 μm) with “salami”-type structures at 15 phr rubber content in butadiene-acrylonitrile modified epoxy resin. Additionally, very small (approximately 0.1 μm) particles populated the interstitial space between the large particles.[85]

Influence of PTS on the Thermal Degradation Kinetics of Poly(GMA-stat-C13MA)

The thermal degradation kinetics of poly(GMA-stat-C13MA) and the influence of PTS on the degradation kinetics was investigated by TGA at different heating rates to be further evaluated by Kissinger’s method.[86,87] According to Figure a,b, the heating rate has a great effect on the thermal stability, i.e., the higher heating rate, the higher the thermal stability.[88] In fact, the characteristic temperature point would move to a higher value by increasing the heating rate.[89] Kissinger’s method can be applied to determine the activation energy without a precise knowledge of the degradation mechanism.[90] The activation energy at the maximum decomposition rate was thus obtained by using the following equation:where Tmax is the absolute temperature, αmax is the conversion at the maximum weight-loss rate, A is the pre-exponential factor, β is the heating rate, E is the activation energy, and R is the universal gas constant. Figure c shows the calculation of activation energy from graphs of versus at heating rates of 5, 10, 15, and 20 °C/min according to eq . According to Figure c, the activation energy of poly(GMA-stat-C13MA) at αmax increased from 143.5 to 327.2 kJ mol–1 with the addition of 4 wt % PTS. This could be due to the interactions between the nanoparticles and the escaping moieties during degradation which provide a mass transport barrier against volatile degradation products and increases the activation energy.[91] The increase in activation energy also indicates an improvement in the formation of thermally stable char by PTS nanoparticles.[90] For comparison, the average values of activation energy for neat epoxy and epoxy/graphene oxide (GO) are 57 and 65 kJ mol–1, which means that the introduction of GO slows down the degradation reactions.[92] In another study, the presence of polythiophene increased the activation energy of epoxy resin at αmax from 165.6 to 221.6 kJ mol-1.[93] In the case of poly(methacrylates), the activation energies of PMMA degradation increased at high degradations (above 40%, α > 0.4) by the incorporation of 5 wt % silica nanoparticles.[91]

Figure 8

TGA curves of (a) cured poly(GMA-stat-C13MA) (S-0) and (b) nanocomposite (S-4) at different heating rates, (c) Kissinger plots for S-0 and S-4.

Mechanical and Rheological Behavior and Morphology of the Nanocomposites

The tensile properties for cured statistical and block copolymers (S-0 and B-0) and their composite counterparts with different filler loadings are summarized in Table . A comparison between the tensile properties of cured statistical and block copolymers, S-0 and B-0, reveals significantly lower Young’s modulus and stress at break in B-0 (14.05 and 0.76 MPa, respectively) compared to S-0 (71.13 and 10.29 MPa, respectively). The lower glass transition temperature (−45 °C) of the soft segment and difference in morphologies contribute to lower modulus and stress at break of B-0 at room temperature. This could be also due to the un-cross-linked poly(C13MA) segment attached to the cross-linked poly(GMA) segment instead of being dispersed randomly with GMA.[94] In fact, the poly(C13MA) soft segment block acts like a plasticizer as previously observed with other C13MA-containing copolymers.[51] Additionally, the tensile modulus taken from the initial linear region of the stress–strain curve depends on the total crystallinity,[95] thus, a lack of soft-segment crystallizability under strain and the possibility of phase separation could limit the tensile properties of the soft-segment-based block copolymer.[96] On the contrary, the presence of covalent cross-linking bonds throughout the polymer chains in the statistical copolymer results in the reduction in the mobility of whole polymer chains. However, the C13MA segments in the block copolymer are attached to the rigid GMA segments at only one end and therefore are not restricted as much.[97] This consequently slightly increases the elongation observed of the poly(GMA-b-C13MA) block copolymer compared to the poly(GMA-stat-C13MA) statistical copolymer.

Table 3

Effects of PTS on the Tensile Properties of Poly(GMA-stat-C13MA) and Poly(GMA-b-C13MA)

ID	polymer	PTS (wt %)	Young’s modulus (MPa)	stress at break (MPa)	strain at break (%)
S-0	poly(GMA-stat-C13MA)	0	71.13 ± 7.25	10.29 ± 0.70	10.29 ± 1.91
S-0.5	poly(GMA-stat-C13MA)	0.5	87.08 ± 5.69	11.19 ± 0.44	10.15 ± 0.94
S-2	poly(GMA-stat-C13MA)	2	103.15 ± 1.22	14.04 ± 0.11	11.13 ± 1.34
S-4	poly(GMA-stat-C13MA)	4	124.01 ± 5.59	15.03 ± 1.11	10.61 ± 1.23
B-0	poly(GMA-b-C13MA)	0	14.05 ± 0.21	0.76 ± 0.12	11.74 ± 0.83
B-4	poly(GMA-b-C13MA)	4	17.02 ± 0.32	1.59 ± 0.23	11.21 ± 0.33

Figure a shows representative tensile stress versus strain curves. Among statistical copolymers and composites (samples S-0, S-0.5, S-2, and S-4), the highest increase in tensile modulus is 74% (from 71.13 to 124.01 MPa) and tensile strength is 46% (from 10.29 to 15.03 MPa) at 4 wt % PTS loading. Essentially, the strain at break is about the same and slightly increases at the loading of 2 wt % (from 10.29 to 11.13%) and decreases slightly to 10.61% with further increasing the PTS content. For the block copolymers (samples B-0 and B-4), we observed a 21% increase in tensile modulus and 109% increase in tensile strength. The elongation at break of the B-4 nanocomposite decreases slightly from 11.74 to 11.21%. For comparison, the tensile modulus and strength of epoxy/GO nanocomposites with 0.5 wt % GO increased from 3.15 ± 0.11 to 3.36 ± 0.17 GPa (6%) and 52.98 ± 5.82 to 64.79 ± 4.01 MPa (22%), respectively, compared to the neat epoxy resin.[98] Bortz et al.[99] showed that the addition of 0.1 wt % GO in epoxy resin yielded a 12% increase in tensile modulus. Melt-mixing addition of 7 wt % multiwall carbon nanotubes in styrene–butadiene–styrene tri-block copolymers improved the tensile strength of the nanocomposite by 66%.[70] 3 wt % TiO2/polyurethane nanocomposites had ∼15% improvement in tensile strength compared to the cured polymer counterpart.[100] According to Buxton et al.[101] rod and plate type particles provide better mechanical reinforcement in polymer matrices compared to spherical nanoparticles. Among the anisotropic nanoparticles, plate-like nanoparticles provide higher mechanical reinforcement than nanorods as they have a propensity to orient to the tensile direction.[101,102]

Figure 9

Effects of PTS on (a) tensile properties and (b) dynamic mechanical properties of cured poly(GMA-stat-C13MA) (S-0) and poly(GMA-b-C13MA) (B-0).

DMA probed the thermomechanical effects of PTS incorporation in poly(GMA-stat-C13MA) and poly(GMA-b-C13MA). For S-0 to S-4 specimens, targeting various PTS concentrations enabled tuning of the storage modulus in both the glassy state (up to 59%) and rubbery plateau regions (up to 88%), as shown in Figure b. The storage modulus in the glassy region mainly depends on the stiffness of the filler, good dispersion of filler, and polymer-filler interactions.[41] The damping factor (tan δ, indicative of Tg) shifted toward higher temperatures and became slightly less intense at higher PTS concentrations because of interactions between PTS and the surrounding polymer network. This is in good agreement with the evidence derived from DSC and TGA experiments. B-0 and B-4 samples exhibited very different dynamic mechanical behavior at temperatures above 25 °C because of macrophase structures (data below room temperature are not available). Generally, the DMA of phase-separated copolymers in the full temperature range covering both Tgs displays two distinct transitions; one at a lower temperature corresponding to the glass–rubber transition of the elastomeric soft phase and the other at higher temperatures related to the transition of the rigid phase.[103,104] In the present study, both B-0 and B-4 samples exhibited neither a sharp drop in G’ nor a real plateau at high temperatures. This behavior is in good agreement with previous studies on DMA behavior of block copolymers.[105,106] The damping factor is less pronounced than the statistical copolymers, corresponding to the transition of the rigid poly(GMA) phase (second Tg). Although the data for the glassy region are not available, we can generally conclude that PTS increases the modulus, as the rigid particles stiffen the network.

Frequency sweep measurements were carried out to better understand how PTS interacts with the polymer network. The elastic modulus (G’) and loss modulus (G”) of cured poly(GMA-stat-C13MA) and poly(GMA-b-C13MA) and their nanocomposite counterparts are shown in Figure . As expected, G’ of PTS-filled systems is substantially higher than G’ of the neat counterparts. Figure shows small increases in G’ of the PTS-filled blends while the G’ of S-0 and B-0 do not significantly increase with frequency. In all cases, the dependence of G” on frequency is more pronounced and increases substantially (around an order of magnitude) with frequency. Since the frequency dependence of G’ has been observed for other cross-linked nanocomposites,[107,108] we concluded that the G’ of PTS nanocomposites is frequency dependent and it seems that the PTS might physically interact with the polymer chains. This observation implies that at low frequencies, the polymer chains have enough time to relax after deformation which results in lower modulus.[109] However, at high frequencies (>10 rad/s) the polymer chains do not have enough time to relax, and thus the modulus increases. The neat thermosets however do not show a substantial frequency dependence of G’ within the range measured, indicating the covalently cross-linked network structures.[107] As observed in Figure , for S-0, S-0.5, S-2, and S-4 samples, the storage modulus at low frequencies up ∼10 rad/s is roughly an order of magnitude higher than the loss modulus. In the case of B-4, this occurs at higher frequencies (above 50 rad/s).

Figure 10

G’ (storage modulus) and G” (loss modulus) versus frequency of (a) cured poly(GMA-stat-C13MA) (S-0) and (b) poly(GMA-b-C13MA) (B-0) and their nanocomposites with different filler loadings under nitrogen at 100 °C.

As the dispersion and the interfacial interaction of nanoparticles could markedly affect the mechanical properties of the polymer, the dispersion of PTS nanoparticles was investigated by HAADF-STEM. Figure shows the thin slices of the nanocomposites. For S-0.5, S-2, and S-4 samples, the dispersion of PTS particles can be clearly observed in Figure without intense agglomeration. Additionally, the shape of nanoparticles is the same as that shown in Figure and there is no obvious gap between the polymer and PTS, showing a blurred interface. Additionally, a comparison between Figure d,e and Figure shows that the incorporation of PTS did not alter the morphology of the cured block copolymer blend. Indeed, the poly(GMA-b-C13MA) nanocomposites consisted of the same macrophase-separated morphology as the network without any nanoparticles. Similarly, in a study by Esposito et al.,[110] the morphology of an epoxidized poly(styrene-b-butadiene-b-styrene) copolymer remained unaltered after the incorporation of carboxylic acid-modified MWCNTs. Based on the STEM results, we can conclude that the interfacial compatibility between PTS and the synthesized polymer was sufficient to disperse the particles, without obvious agglomeration of the nanoparticles.

Figure 11

HAADF-STEM of cross-linked poly(GMA-stat-C13MA) and poly(GMA-b-C13MA) containing PTS hybrids: (a) S-0.5, (b) S-2, (c) S-4, and (d,e) B-4 (small white arrows indicate PTS).

Overall, DMA, TGA, and tensile testing showed that the addition of PTS more effectively improved the thermal and mechanical properties of the cured poly(GMA-stat-C13MA) compared to the cured poly(GMA-b-C13MA). This could be attributed to the dispersion of the PTS in the statistical and block copolymer matrix. Previous studies have shown that the dispersion of nanoparticles in the block copolymers is complicated due to the phase separation behavior.[111,112] In fact, the competition between the macrophase separation of the block copolymer and the aggregation of the nanoparticles determines the size of aggregates and the ultimate structure of the nanocomposites. Since the mechanical properties and thermal stability did not improve in B-4 as much as it did in S-4, it can be concluded that NP dispersion is reduced in the block copolymer because of the formation of phase-separated domains.[111,113,114] However, we still chose to study nanoparticle dispersion in the block copolymers as preferential segregation of the nanoparticles in one of the domains could enhance a property such as thermal stability or barrier enhancement.

Conclusions

Epoxy-functional copolymers (statistical and segmented block) were synthesized by NMP with potential applications in adhesive and antiflammable coatings. PTS hybrids were added to enhance the thermal and mechanical properties, simultaneously. For the cured block copolymer, microphase separation was verified by the appearance of two Tgs. The TEM micrographs of the cured poly(GMA-b-C13MA) showed dispersed spherical microdomains in a range of ∼100–500 nm because of the reaction-induced macrophase separation upon curing. PTS hybrid nanoparticles were synthesized via a hydrothermal reaction with the average size of 130 nm, T10% = 485 °C, 55% residue at 800 °C, and plate-like structures, and were further incorporated in the polymers to improve the thermal stability and mechanical properties. DSC revealed that Tg of poly(GMA-stat-C13MA) nanocomposites increased from 65 to 77 °C by adding 4 wt % PTS. TGA showed that T10% and char residue was enhanced, up to 12 °C and ∼7 wt %, respectively. The activation energy of poly(GMA-stat-C13MA) at αmax increased from 143.5 to 327.2 kJ mol–1 with the addition of 4 wt % PTS. In addition to thermal stability, tensile modulus and tensile strength increased from 71.13 to 124.01 MPa and 10.29 to 15.03 MPa, respectively, upon the addition of 4 wt % PTS. Further, storage modulus improved in both the glassy region (up to 59%) and the rubbery plateau region (up to 88%). We concluded PTS sufficiently interacted with the host resin as it was effectively dispersed into the resin, based on the TEM, mechanical, and rheological measurements. Overall, these PTS hybrids indicate some promise as modifiers in polymer resins with enhanced thermal stability.