Effect of surface activation of alumina particles on the performances of thermosetting-based composite materials

Abstract

Two types of alumina particles, commercial (c-Al₂O₃) and iron doped (Fe-Al₂O₃), were functionalized with 3-(aminopropyl)trimethoxysilane (one-step) and two-step consecutive process, i.e. firstly using 3-(aminopropyl)trimethoxysilane followed by methyl ester of linseed oil (biodiesel) to produce Al₂O₃ATPMS-BD reinforcement, respectively. The effect of modifier type and variable amount of alumina particles on the dynamical and mechanical properties of unsaturated polyester resin–based composites was studied. The highest improvement of the tensile strength and micro Vickers hardness, 78.1 and 163%, respectively, was obtained at 1.0 wt% of Fe-Al₂O₃APTMS-BD addition. The obtained multifunctional composites can be potentially applied in construction and mining industries.

Keywords

Thermosetting resin particle-reinforcement polymer-matrix composites chemical properties mechanical properties

Introduction

The challenge of polymer composite technologies development to meet the increasing market needs for such materials has become an area of concern in recent years. Great efforts in designing suitable composite formulations using different types of organic/inorganic reinforcements resulted in obtaining materials with unique properties and good quality/price ratio.^1–3

Nano-sized alumina powders which possess excellent structural properties are ideal candidates for preparing polymer composites.^4,5 However, there are some limitations related to the hydrophilic nature of alumina particles, which cause high water/moisture absorption and incompatibility with hydrophobic segments of polymeric matrices.⁶ These drawbacks can be overcome by the surface modification/functionalization through reaction between modifying agent and large number of hydroxyl groups within the alumina structure.⁷ Alumina particles have been modified with various types of modifying agents in order to tailor/design reactive surface.^8,9 The obtained surface architecture, most frequently created by silane coupling agents, is suitable for establishment of a great linkage with polymeric matrices, such as unsaturated polyester resins (UPRs).

UPRs represent the most frequently used thermosetting matrices for obtaining the advanced composite materials. The UPR production can be carried out by re-use of the waste poly(ethylene terephthalate) (PET) with respect to the circular economy concept instead of linear one based on “take, make, use, and dispose.”^10,11 Furthermore, it is not only the challenge of resource scarcity in terms of closed material loop, but also the challenge of environmental pollution that has become acute.¹⁰

Considering that UPR is a brittle material, the addition of reinforcements is necessary for obtaining the high-performance composites. The most commonly used modifiers are silanes with methacrylic/vinyl double bonds which participate in cross-linking reactions during the curing of thermosetting resins.¹² Besides, an innovative way to prepare green alumina-based fillers is the esterification of surface alumina hydroxyl groups by bio-renewable fatty acids from vegetable oils.¹³ The presence of long fatty acid chains gives the hydrophobic alumina surface which enables better adhesion on filler/polymeric matrix interface as well as the plasticity, which plays a significant role in determining/designing the composites dynamical–mechanical properties.¹⁴ There are many studies related to the influence of modified inorganic/organic fillers on the mechanical properties of the UPR-based composites, but the effect of modified alumina particles is not much described in the literature.

Two different types of alumina particles are used in this study: commercial alumina (c-Al₂O₃) and alumina doped with ferrous oxide (Fe-Al₂O₃).¹⁵ Afterward, modification of both alumina particles is performed by 3-(aminopropyl)trimethoxysilane (APTMS) (one-step modification) and successive modification with APTMS in the first step followed by methyl ester of linseed oil fatty acids (biodiesel, BD) in the second step. Although it is known that amino functionalized reinforcements are more suitable for application in epoxy-based matrices,¹⁶ the successful introduction of BD onto amino modified alumina, by formation of the amide linkage between APTMS and BD, creates possibility of its use in the UPR-based composite as well. The influence of modification type/amount of alumina particles on dynamical–mechanical and thermal properties of UPR-based composites is investigated in this study.

Experimental

Materials

Waste PET, previously mechanically pre-treated, was procured from RKS Kompoziti Ltd. Obsolete PET bottles were crushed into small pieces (approximately 0.5 × 0.5 cm) and washed with ethanol and dichloromethane to remove impurities and residual adhesives. Propylene glycol (PG), tetrabutyltitanate styrene, sodium hydroxide, toluene, pyridine, methyl ethyl ketone peroxide, cobalt octoate, maleic anhydride (MA), hydroquinone, methanol, absolute ethanol, tetrahydrofuran (THF), commercial alumina, c-Al₂O₃, and iron(III)-chloride (FeCl₃·6H₂O) were obtained from Sigma-Aldrich. Locron L (Al₂(OH)₅Cl·2.5H₂O), used for alumina synthesis, was supplied by Clariant. The APTMS was supplied by Dynasylan, Evonik Industries. All chemicals were of analytical grade and used as-received without further purification.

Synthesis of UPR and nano-alumina reinforcement

UPR synthesis was performed through two consecutive steps: (i) catalytic depolymerization of waste PET (glycolysis) using PG (molar ratio PET:PG was 1:0.65) and tetrabutyltitanate catalyst and (ii) polycondensation of previously obtained hydroxyl terminated oligomeric product (polyol) with MA. Pilot-scale PET glycolysis and polycondensation reaction with MA (molar ratio PG:MA was 1:0.95) were performed in accordance with previous publications^2,17 in a custom-made pressure reactor (8 l of reactor volume) equipped with mechanical stirrer, nitrogen inlet system, reflux condenser, Dean–Stark azeotropic separator, temperature and pressure controllers, and indicators. Slow addition of 4 kg PET flakes into 1.54 l of preheated PG at 190℃ for 3 h provides slow dissolution/depolymerization. Glycolysis reaction was carried out for 6 h at 220℃ without ethylene glycol azeotropic removal (classical glycolysis method). The mixture was then cooled down to 90℃ when 0.015 wt% of hydroquinone was added. During the polycondensation of synthesized polyol with MA, the temperature was increased from 90 to 210℃ at heating rate 15℃/h. The obtained UPR resin possessed the following characteristics: acid value 12 and hydroxyl value 41 (synthesis path presented by blue arrows, Figure 1).

Figure 1.

Schematic overview of the procedure for the production the UPR/alumina-based composites.

Alumina particles doped with iron(III)-oxide were prepared analogously to previously published procedure using sol–gel technique.¹⁸ The sol consisting of Al₂Cl(OH)₅·2.5H₂O and demineralized water, as well as 1.5 wt% FeCl₃·6H₂O was placed into the reactor and left to gel. Further on, the obtained product was thermally treated at 900℃ for the next 2 h.

Surface modification of alumina reinforcements

Alumina particles (c-Al₂O₃ and Fe-Al₂O₃) were modified with: (i) APTMS (one-step modification) and (ii) APTMS modified Al₂O₃ subsequently reacted with methyl ester of linseed oil fatty acids (biodiesel of linseed oil—BD) synthesized according to the previously published method (two-step modification).³ Modification procedure which includes usage of BD was performed by the two-step procedure in the following way:

(a) First step—modification with APTMS. Alumina particles, c-Al₂O₃ or Fe-Al₂O₃, (0.1 kg) were dispersed in 2 l of toluene using ultrasonic probe for 1 h (synthesis path presented by red arrows, Figure 1). Obtained suspension was placed in glass reactor equipped with condenser, dropping funnel, heating system, and thermometer. Afterward, 20 ml of APTMS was added in suspension containing alumina particles and the reaction was carried out in the next 48 h at 25℃ under nitrogen atmosphere. Modified alumina, c-Al₂O₃APTMS or Fe-Al₂O₃ APTMS, was separated from toluene by vacuum filtration, washed with toluene, dried, and further used in composite production or in second modification step.

(b) Second step—modification with BD. The modified alumina from the first step which contained terminal amino groups was dispersed in 2 l of THF (synthesis path presented by green arrows, Figure 1). Further on, 160 g of BD was charged into four-necked glass reactor (5 l), equipped with a mechanical stirrer, temperature control system, condenser with calcium chloride protection tube, and nitrogen inlet tube, and the reaction took place for 12 h at 25℃. Then, the reaction mixture was heated up to 60℃ and maintained for 2 h. The obtained product, c-Al₂O₃APTMS-BD or Fe-Al₂O₃APTMS-BD, was filtered under the vacuum, washed two times with THF, again filtrated and washed with absolute ethanol, and dried at 40℃ for 12 h.

Preparation of composites based on UPR and modified alumina reinforcement

Composites, in general designated as UPR/a-Al₂O₃ N(n), were prepared via the solution blending method by mixing the determined amount of binder, UPR resin, and reinforcement, pristine, or functionalized alumina particles. Index a denotes the type of alumina particles (c-Al₂O₃ or Fe-Al₂O₃ particles), N is related to alumina modification type (c-Al₂O₃APTMS, Fe-Al₂O₃ APTMS, c-Al₂O₃APTMS-BD, or Fe-Al₂O₃APTMS-BD), while n ascribes the amount of added filler into UPR matrix: (a) 0.1, (b) 0.25, (c) 0.5, (d) 1.0, and (e) 2.5 wt%. Both, the UPR resin and alumina particles as well as curing system (cobalt octoate/methyl ethyl ketone peroxide) were placed into a vacuum injection molding device where the added ingredients were vigorously mixed and injected into the mold. Afterward, the cured samples were taken out from the mold and subjected to post-curing at 50℃ for an additional 2 h.

Experimental techniques

Qualitative structural analysis of the modified alumina and curing rate of prepared composites were performed using Fourier transform infrared (FTIR) spectroscopy (Bomem MB-102), in the range of 400–4000 cm⁻¹, at a resolution of 4 cm⁻¹. Nuclear magnetic resonance (¹H NMR) structural analysis of UPR resin was performed in CDCl₃ solvent using Bruker, USA, Ascend 400 device at 400 MHz. The gel time of the composites was determined from the cure exotherm which was measured according to ASTM D2471-99. Microstructural characterization of the modified alumina particles and the corresponding composites was obtained using a Field Emission Gun Scanning Electron Microscopy device with field emission gun TESCAN MIRA3 electron microscope at an accelerating voltage of 20 kV. Thermogravimetry (TG), derivative thermogravimetry (DTG) and differential scanning calorimetry (DSC) analyses of the cured samples were performed using SDT Q600 instrument. Samples were heated at 10℃ min⁻¹ to 800℃ in a flow of nitrogen (100 cm³ min⁻¹). Uniaxial tensile measurements of standard cured samples were performed using an AG–X plus Universal testing machine, Shimadzu Corporation according to ASTM D882-12 standard method. All the tests were performed on three replicates for each sample at room temperature and crosshead speed of 0.5 mm/min. The hardness of the composites was measured using micro Vickers hardness (VH) tester Leitz, Kleinharteprufer Durimeti with a load of 4.9 N according to ASTM E384-16 standard method. Dynamical–mechanical properties of composites were obtained using dynamic mechanical analysis (DMA), which was performed on a Discovery Hybrid Rheometer HR2. The DMA was conducted in a torsion rectangular mode (sample dimensions: 6 × 1×0.2 mm³) from 25 to 250℃ at a fixed strain amplitude of 0.1% and angular frequency of 1 Hz. The glass transition temperature (T_g), determined by the dynamic mechanical measurements, was estimated as the maximum value at the temperature dependence curve of the loss factor, tanδ.¹⁹

Results and discussion

FTIR analysis of the modified alumina particles

The FTIR spectra of the modified c-Al₂O₃ and Fe-Al₂O₃, as well as the spectra of pristine alumina, used for comparison purpose, are presented in Figure 2. The influence of applied modification processes in designing of filler surface properties is examined.

Figure 2.

FTIR spectra of pristine/modified alumina particles.

The characteristic peaks of c-Al₂O₃ and Fe-Al₂O₃ particles, observed at about 3450 and 1638 cm⁻¹, originate from hydroxyl group stretching and bending vibrations, respectively.^4,20 This indicates that most of the OH groups are on the particles surface, which points to the surface affinity for adsorption of ambient moisture exhibited by alumina.²¹ Moreover, pristine alumina particles show specific broad bands in the range 550–750 cm⁻¹ attributed to Al–O–Al and Al–O vibrations.²² The band around 553 cm⁻¹ in the FTIR spectrum of c-Al₂O₃ is shifted to the higher value (568 cm⁻¹) in the FTIR spectrum of Fe-Al₂O₃ particles due to contribution of the Fe–O bending vibration.²³

In the FTIR spectra of the modified alumina particles, the peaks observed at 2927, 2867, and 2851 cm⁻¹ originate from CH₃ and CH₂ groups stretching vibrations, respectively. The stretching ν(N–H) vibrations, overlapped with OH vibrations, are observed at ∼3400 cm⁻¹. The peaks detected at 1638 (amide I) at 1562 cm⁻¹ (amide II) (Figure 2(a)) are mainly associated with the C=O stretching vibration and N–H bending vibration coupled with the C–N stretching vibration of amide group, respectively.²⁴ In the range 1650–1600 cm⁻¹, double/unsaturated bonds, which originate from BD, contribute to the change in the intensities of the peaks. Moreover, these groups are overlapped with peaks associated to OH bending vibrations. Symmetrical C–H deformation vibrations of CH₃ groups are observed at 1493 and 1486 cm⁻¹ (Figure 2(a) and (b), respectively). Stretching vibrations of N–H, from amino and primary amide bond present in c-Al₂O₃APTMS-BD mixed with δN–H, are noticed at 1129 cm⁻¹ (Figure 2(a)). Two overlapped asymmetric bands at 1114 and 1068 cm⁻¹ are assigned to Si–O stretching vibrations.^25,26

NMR characterization of UPR

¹H NMR analysis is performed to confirm the structure of the obtained UPR, and the corresponding spectra indicate that UPR contains mostly more reactive fumaric moiety necessary for achieving high cross-linking reactivity during sample curing. The results of the analysis of ¹H NMR spectrum of UPR (Figure 3) are: ¹H NMR (CDCl₃): 1.15–1.42 (m, 6H, CH₃), 3.99–4.54 (m, 6H, CH₂CH–), 4.33–4.67 (m, 4H, –O–CH₂CH₂–O–), 5.24–5.47 and 6.69–6.77 (m, 2H, fumaric moiety), and 8.05 (s, 4H,H_Ph-terephthaloyl moiety).

Figure 3.

NMR spectra of UPR resin.

¹H NMR spectra of the synthesized UPR show that the dominant products of glycolysis are glycol esters of terephthalic acid: bis(2-hydroxylpropyl) terephthalate, (2-hydroxyethyl)(2-hydroxypropyl) terephthalate and glycols,¹⁷ which are incorporated into UPR in the course of polycondensation reactions. The results of NMR analysis confirm successful synthesis of UPR.

Curing kinetics

Curing kinetics of UPR and the corresponding composites with 1 wt% of filler loading, traced out through styrene conversion into cured network, were measured using FTIR technique. The impact of the modified alumina particles compared to pristine fillers and pure UPR on curing degree was determined. The monitoring of styrene consumption was done by change of the peak area at 909 cm⁻¹ and the obtained data were processed according to Beer's law.²⁷

Figure 4 shows the styrene conversion as a function of reaction time, for pure UPR and corresponding composites. It is noticed that for 20 min of system curing (inset S1) styrene conversion is in the range 84–87% and 77–82% for pure UPR/composites filled with pristine alumina and composites reinforced with modified alumina, respectively. Further on, the enhancement of styrene consumption for 8–11% is achieved by applying the additional heating for 2 h at 50℃. It can be observed that curing at elevated temperature results in obtaining the styrene conversion degree more than 92%, except for the composites reinforced with alumina modified with APTMS. This occurs due to the inhibitory effect of amine functionalities present in APTMS which could participate in radical transfer reactions with peroxy radical (initiator) or alkyl radicals (chain transfer radical-propagation). In that way, aminoxyl radicals of lower chain transfer capability are created, and they could participate either in propagation or could return to their amino form. Another factor which controls the system reactivity originates from the extent of creation of hydrogen bonds between amino groups from APTMS and carbonyl groups from UPR resin. Decreased flexibility of macromolecule segments, i.e. poor diffusion of reactive centers, causes lower effectiveness of radical transfer between ethylenic segments from UPR resin and styrene causing the cross-linking reactions²⁸ to slow down.

Figure 4.

Styrene conversion during curing determined by FTIR method.

Conversely, APTMS-BD modified alumina contains approximately six double bonds possibly involved in curing reaction with unsaturated segments from UPR resin and styrene. Thus, styrene consumption is higher, compared to UPR/(c, Fe)-Al₂O₃APTMS composites, regardless of steric hindrance on the reactivity caused by BD structure. The potential of reactive unsaturation is sufficient to overcome the influence of flexible/coiled BD structure, having higher end-value of styrene conversion as a consequence (Figure 5). The incorporation of pristine alumina particles ((c, Fe)-Al₂O₃) leads to auto-acceleration of curing reaction²⁹ and thus higher values for styrene conversion. A lower value was obtained for Fe-Al₂O₃ which causes slightly increased viscosity, i.e. restriction of the movements of the macromolecule segments.³⁰ These results suggest that system reactivity significantly depends on filler surface properties.

Figure 5.

Modification paths of alumina particles together with interactions between modified fillers and UPR resin.

The gel time and maximum curing temperature of UPR resin and the corresponding composites

The gel times together with maximum curing temperatures, determined from curing exotherms, of UPR resin and composites with 1 wt% of filler particles are shown in Table 1. The obtained results indicate that addition of unmodified alumina particles decreases the gel time compared to pure UPR resin, while the maximum curing temperature is slightly increased. The observed phenomenon is potentially related to a catalytic effect of nano/submicron-scaled alumina particles when the radical polymerization reaction is performed.³¹ The incorporation of larger modified alumina particles (Figure 5) due to physical/size effect prevents the approaches of vinyl moieties from UPR resin and segments introduced via modification processes causing the decrease of T_max.³ The gel time/maximum curing temperature of composites containing BD-modified alumina particles shows somewhat higher values, compared to samples with APTMS functionalized filler, due to the number of double bonds present in BD suitable to react with vinyl moieties from UPR resin/styrene. Again, the amino group showed low retardation effect to system reactivity which is reflected in lower T_max. These results are in full accordance with the trend obtained by FTIR curing kinetics measurements.

Table 1.

The gel time and maximum curing temperature of UPR resin and analyzed composites.

Sample	Gel time (min)	T_max (℃)
UPR	11.75	95.6
UPR/c-Al₂O₃(d)	11.50	97.8
UPR/c-Al₂O₃APTMS(d)	12.25	92.1
UPR/c-Al₂O₃APTMS-BD(d)	11.75	95.3
UPR/Fe-Al₂O₃(d)	11.25	97.1
UPR/Fe-Al₂O₃APTMS(d)	12.75	91.3
UPR/Fe-Al₂O₃APTMS-BD(d)	12.00	93.8

UPR: unsaturated polyester resin; APTMS: 3-(aminopropyl)trimethoxysilane; BD: biodiesel.

Thermal properties of pristine/modified alumina

The thermal properties of unmodified/modified alumina particles and the amount of grafted modifiers are determined using TG, DTG, and DSC analysis (Figure 6). The amount of grafted modifiers is calculated from the difference between residual content of modified and pristine filler particles (higher weight loss of modified particles above 125℃ is attributed to loss of organic modifier), see Table 2. The weight loss for all the analyzed samples between 25 and 125℃ originates from the loss of water/volatile compounds adsorbed at the surface of alumina particles.³² TG curves of the unmodified alumina (Figure 6(a)) show that they are more thermally stable than modified ones, with the weight loss of 7.4 and 5.5% for c-Al₂O₃ and Fe-Al₂O₃, respectively (Table 2).

Figure 6.

Thermal properties of the unmodified/modified alumina particles determined by (a) TG, (b) DTG, and (c) DSC analysis.

Table 2.

The T_0.05, residue content, grafted modifiers, and DTG peaks of the unmodified/modified alumina particles.

Sample	T_0.05 (℃)	Residual content (wt%)	Grafted modifier (wt%)	DTG peaks (℃)
c-Al₂O₃	230.9	92.6	–	94.2
Fe-Al₂O₃	516.7	94.5	–	48.8, 382.7
c-Al₂O₃APTMS	225.9	84.1	8.5	58.2, 247.2, 407.4
c-Al₂O₃APTMS-BD	108.1	75.3	19.2	102.7, 243.9, 434.5
Fe-Al₂O₃APTMS	371.3	92.0	2.5	54.2, 158.9, 406.9, 565.1
Fe-Al₂O₃APTMS-BD	188.0	86.0	8.5	102.7, 243.9, 434.6

APTMS: 3-(aminopropyl)trimethoxysilane; BD: biodiesel.

The TG profile of the unmodified alumina is similar, while the degradation pattern of the modified alumina depends on the surface molecular structure. The weight loss of the modified alumina particles, in the range 125–260℃, is attributed to dehydration/thermal transformation of the functional groups linked via silanol groups at the alumina surface. Thermal dehydration and conformational rearrangements of flexible structures of BD residues occur at lower temperatures (endothermic peak in Figure 6(c)).⁶ This phenomenon is not observed for Fe-Al₂O₃-based particles, probably due to thermal stability of more ordered/tightly bonded surface functionalities. At temperatures above 260℃ conjugated double bonds in linoleic and α-linolenic fatty acids contribute to the decrease of the thermal stability of corresponding composites.⁶ The modification with APTMS-BD shifts the temperature interval of the highest decomposition from 460 to 600℃, in comparison to APTMS-modified alumina. The highest weight loss, 24.8%, is observed for c-Al₂O₃APTMS-BD. The unmodified/modified Fe-doped alumina exhibits higher thermal stability, compared to c-Al₂O₃ particles, which is attributed to modification extent and structural/spatial arrangements of lower amount of attached organic modifiers.

The temperatures where 5 wt% of the materials weight decrease (T_0.05), residue content at 800℃, the amount of grafted modifiers as well as DTG characteristic peaks are presented in Table 1. Higher thermal stability of the unmodified/modified Fe-doped alumina particles indicates relationship between larger weight loss and increased percentage of the thermodegradable organic functionalities in APTMS-BD modified particles.

The DTG peaks of APTMS and APTMS-BD modified alumina (Figure 6(b)) are grouped and, regardless to type of particles, show similar trend within the same group. Introduction of the BD onto the alumina structure leads to releasing higher amount of low molecule weight volatile compounds, which are reflected in shifted peaks (∼103℃) compared to APTMS modified particles (∼55℃). Peaks about 240℃ originate from the partial decomposition and peak at 407/435℃ corresponds to the complete thermal degradation of modifying agents and dehydration of silanol group.

Microstructural properties of alumina particles

The microstructural properties of the unmodified/modified alumina particles are determined using Field Emission Gun Scanning Electron Microscopy technique (Figure 7). The nano-sized commercial alumina has high tendency to absorb the ambient moisture which causes forming of the micro-scaled, regularly shaped agglomerates (Figure 7(a)). Fe-doped alumina forms smaller, irregular shaped agglomerates which are related to applied modification procedure¹⁸ (Figure 7(d)). The modification of c-Al₂O₃ leads to breaking the larger to smaller particle cluster (Figure 7(b) and (c)). The agglomeration degree is higher for APTMS-BD modified c-Al₂O₃ due to dipole–dipole and hydrogen bonding intermolecular interactions.^3,17 Slightly larger agglomerates of Fe-Al₂O₃APTMS are observed, compared to the c-Al₂O₃APTMS (Figure 7(e)), while similar morphology of c- and Fe-Al₂O₃APTMS-BD fillers is found (Figure 7(f)).

Figure 7.

SEM micrographs of pristine/modified alumina particles: (a) c-Al₂O₃, (b) c-Al₂O₃APTMS, (c) c-Al₂O₃APTMS-BD, (d) Fe-Al₂O₃, (e) Fe-Al₂O₃APTMS, and (f) Fe-Al₂O₃APTMS-BD.

Mechanical properties of UPR/alumina composites

The mechanical properties, tensile strength and micro VH, which are determined via stress at break (σ_t), elongation at break (ɛ_t), Young modulus (E), and VH value, are shown in Table 3. The obtained results suggest that incorporation of alumina enhances the mechanical properties of composites, especially micro VH. These properties gradually increase up to 1.0 wt% of reinforcement addition, and with higher loading, the mechanical properties follow decreasing trends. The appearance of particle agglomerates at higher loading (>1.0 wt%) causes formation of the concentrated stress sites and deterioration of mechanical properties is a consequence.²⁴

Table 3.

Values of σ_t, ɛ_t, E, and VH of UPR/alumina composites.

Sample	σ_t (MPa)	ɛ_t (%)	E (GPa)	VH (kgf/mm²)
UPR	26.10 ± 0.28	3.06 ± 0.13	0.96 ± 0.07	31.23 ± 0.73
UPR/c-Al₂O₃(a)	31.14 ± 1.34	3.80 ± 0.14	0.84 ± 0.07	37.21 ± 0.63
UPR/c-Al₂O₃(b)	35.09 ± 1.48	4.31 ± 0.15	0.82 ± 0.08	39.16 ± 2.32
UPR/c-Al₂O₃(c)	36.59 ± 1.64	3.55 ± 0.14	1.06 ± 0.12	41.60 ± 1.04
UPR/c-Al₂O₃(d)	41.65 ± 1.13	4.81 ± 0.13	0.96 ± 0.05	49.63 ± 3.63
UPR/c-Al₂O₃(e)	30.54 ± 1.55	3.04 ± 0.13	0.96 ± 0.11	38.38 ± 2.45
UPR/c-Al₂O₃APTMS(a)	30.79 ± 1.14	2.61 ± 0.16	1.49 ± 0.16	59.89 ± 2.15
UPR/c-Al₂O₃APTMS(b)	34.09 ± 0.30	3.37 ± 0.12	1.07 ± 0.05	62.32 ± 2.91
UPR/c-Al₂O₃APTMS(c)	36.43 ± 1.22	3.59 ± 0.23	1.18 ± 0.13	65.36 ± 1.78
UPR/c-Al₂O₃APTMS(d)	37.11 ± 1.24	3.12 ± 0.17	1.21 ± 0.19	68.30 ± 1.72
UPR/c-Al₂O₃APTMS(e)	27.68 ± 1.16	2.28 ± 0.15	1.39 ± 0.11	50.31 ± 8.22
UPR/c-Al₂O₃APTMS-BD(a)	30.78 ± 1.34	3.40 ± 0.14	1.01 ± 0.08	54.58 ± 1.39
UPR/c-Al₂O₃APTMS-BD(b)	32.25 ± 1.52	3.64 ± 0.13	0.98 ± 0.09	56.81 ± 1.29
UPR/c-Al₂O₃APTMS-BD(c)	38.44 ± 1.73	4.13 ± 0.16	1.02 ± 0.10	59.67 ± 1.86
UPR/c-Al₂O₃APTMS-BD(d)	40.79 ± 1.66	4.37 ± 0.19	1.06 ± 0.11	70.55 ± 0.94
UPR/c-Al₂O₃APTMS-BD(e)	30.78 ± 1.34	3.40 ± 0.14	1.01 ± 0.08	52.90 ± 2.20
UPR/Fe-Al₂O₃(a)	30.85 ± 1.60	3.80 ± 0.13	0.87 ± 0.11	24.78 ± 0.70
UPR/Fe-Al₂O₃(b)	36.56 ± 1.42	4.05 ± 0.14	0.99 ± 0.07	25.93 ± 0.39
UPR/Fe-Al₂O₃(c)	38.58 ± 1.13	4.56 ± 0.13	1.04 ± 0.06	27.87 ± 1.68
UPR/Fe-Al₂O₃(d)	39.69 ± 1.34	4.81 ± 0.17	1.01 ± 0.08	35.68 ± 0.65
UPR/Fe-Al₂O₃(e)	23.99 ± 1.85	2.23 ± 0.14	1.08 ± 0.12	26.93 ± 0.89
UPR/Fe-Al₂O₃APTMS(a)	32.08 ± 1.21	2.98 ± 0.14	1.41 ± 0.06	45.79 ± 4.79
UPR/Fe-Al₂O₃APTMS(b)	33.20 ± 0.99	3.07 ± 0.12	1.23 ± 0.08	54.99 ± 2.63
UPR/Fe-Al₂O₃APTMS(c)	35.09 ± 1.63	3.02 ± 0.13	1.07 ± 0.12	57.59 ± 3.24
UPR/Fe-Al₂O₃APTMS(d)	37.64 ± 1.47	3.11 ± 0.19	1.51 ± 0.11	61.10 ± 5.67
UPR/Fe-Al₂O₃APTMS(e)	31.22 ± 1.26	3.01 ± 0.18	1.06 ± 0.06	41.99 ± 2.33
UPR/Fe-Al₂O₃APTMS-BD(a)	32.42 ± 1.28	2.99 ± 0.18	1.00 ± 0.12	65.44 ± 2.30
UPR/Fe-Al₂O₃APTMS-BD(b)	36.46 ± 1.28	3.73 ± 0.19	1.12 ± 0.06	67.64 ± 4.35
UPR/Fe-Al₂O₃APTMS-BD(c)	41.93 ± 1.35	4.12 ± 0.15	1.01 ± 0.11	73.38 ± 3.73
UPR/Fe-Al₂O₃APTMS-BD(d)	46.48 ± 1.47	4.79 ± 0.17	0.97 ± 0.15	82.32 ± 2.20
UPR/Fe-Al₂O₃APTMS-BD(e)	30.22 ± 1.15	2.90 ± 0.15	1.18 ± 0.13	56.11 ± 4.87

UPR: unsaturated polyester resin; APTMS: 3-(aminopropyl)trimethoxysilane; BD: biodiesel.

The stress at break values show low differences between composites reinforced with pristine and modified alumina particles (Table 3), while significant differences are observed in comparison to neat UPR resin. The highest increase in σ_t is achieved for 1.0 wt% of reinforcing filler loading: 59.6, 52.1, 42.2, 56.3, 44.2, and 78.1% for incorporation of c-Al₂O₃, Fe-Al₂O₃, c-Al₂O₃ APTMS, c-Al₂O₃ATPMS-BD, Fe-Al₂O₃APTMS, and Fe-Al₂O₃APTMS-BD, respectively.

A slightly higher increase of stress at break and VH for incorporation of the c-Al₂O₃, compared to Fe-Al₂O₃, is a consequence of higher amount of surface hydroxyl sites which participate in intermolecular interaction with polymer matrix, as evidenced by amount of grafted modifiers from Table 2. Higher amount of APTMS and subsequently BD modifier found for c-Al₂O₃ particle reflect forming of thicker layers via available hydroxyl groups. On the contrary, at Fe-Al₂O₃ lower quantity of modifiers is homogeneously distributed and tightly bonded to filler surface providing better interfacial filler/matrix adhesion (Figure 6 and Table 2).

Generally, lower reinforcing effect, obtained by the incorporation of APTMS modified fillers, is the consequence of steric repulsion/interactions and increased particles geometry which disables the effective approach/diffusion of reactive centers. This phenomenon is overcome with APTMS-BD-modified alumina due to numerous ethylenic bonds in flexible structure of BD efficiently involved in cross-linking reactions.⁶ Besides, it is proved that the length/functionalities of the fatty acid chains increase the ductility of the composites through establishing a disordered conformation which enables easier deformation of materials.¹⁴ Conversely, it is also reported that ethylenic segments disable chains elongation resulting in the decrease of the maximum strain.³³ According to the values of ɛ_t and E (Table 2), both factors are into play, considering somewhat higher elongation at break of composites. Moreover, higher ɛ_t, observed for APTMS-BD modified alumina, are related to flexibility of BD segments which also exhibit appropriate plasticizing effect.³⁴

In general, better improvement of mechanical properties is achieved by incorporation of APTMS-BD modified alumina. Still, VH significantly increases for embedding both the modified alumina. The highest increase of 119, 126, 96, and 163% for the addition of c-Al₂O₃APTMS(d), c-Al₂O₃ATPMS-BD(d), Fe-Al₂O₃ APTMS(d) and Fe-Al₂O₃ATPMS-BD(d), respectively, is obtained. Such results indicate that established covalent bonds between ethylenic groups from BD and UPR resin provide means for a load transfer from polymer matrix to filler particles during stress transfer.³⁵ The addition of pristine alumina causes moderate increase of VH (59 and 14% for c-Al₂O₃(d) and Fe-Al₂O₃(d), respectively), but still significant if compared to neat UPR.

Scanning electron microscopy (SEM) analysis of composites

To support previous claims, SEM analysis of fracture surface after tensile strength tests is performed. Differences in the fracture surface properties of the selected composites samples could be observed (Figure 8).

Figure 8.

SEM micrographs of composites: (a) UPR/c-Al₂O₃APTMS, (b) UPR/c-Al₂O₃APTMS-BD, (c) UPR/Fe-Al₂O₃APTMS, and (d) UPR/Fe-Al₂O₃APTMS-BD.

The presence of reinforcement changes the crack mechanism, i.e. the filler particle behaves as a barrier for the fracture propagation. Conversely, concentrated stress at the filler/matrix boundary causes a partial segregation of filler particle from UPR matrix which is reflected in forming of defect into the material (Figure 8(a)). The brittle fracture is noticed when the Fe-Al₂O₃APTMS particles are embedded into the UPR resin which is the consequence of poor interfacial interaction (Figure 8(c)).²⁴ The good interfacial adhesion with APTMS-BD modified reinforcement is provided by numerous linkages at filler/UPR contact surface. Apparently, better adhesion at the fillers/UPR matrix interface leads to efficient load transfer from matrix to reinforcement which contributes in greater mechanical properties of the composites with APTMS-BD modified alumina (Figure 8(b) and (d)).³⁵ The polymer matrix/reinforcements interactions strengthen interfacial bonding which causes increase of the thickness of interfacial layer.^24,36 Hence, the flexibility/bonding of interfacial layer in UPR/c-Al₂O₃APTMS-BD or UPR/Fe-Al₂O₃APTMS-BD (Figure 8(b) and (d)) provides efficient stress transfer from polymer matrix to reinforcements (Table 3).³⁷

Dynamic-mechanical analysis of UPR-based composites reinforced with unmodified and modified alumina particles

To quantify the changes of dynamical–mechanical properties as influenced by adding the unmodified/modified alumina, some characteristics are tabulated in Table 3. The data include temperature dependences of storage modulus in glassy ( $G_{GS}'$ ) and rubbery ( $G_{RP}'$ ) regions at 50 and 235℃, as well as damping factor (tanδ) height, glass transition temperature determined from tanδ peak position (T_g(tanδ)), and cross-linking density (ν). The cross-linking density is calculated using equation which describes the rubbery elasticity theory (equation S1) assuming the absence of macromolecule entanglements³⁸

ν = \frac{G_{RP}'}{RT}

(1)

where ν represents the cross-linking density,

G_{RP}'

the storage modulus in rubbery plateau at temperature T = T_g + 50℃, and R the universal gas constant.

The G′ value is almost constant below 50℃ and decreases as the temperature increases (Figure 9). A slightly different behavior and higher $G_{GS}'$ , observed for UPR/c-Al₂O₃(c), is related to the strength of intermolecular forces and the way the polymer chains packing.³⁹ As the temperature increases, G′ decreases for all samples, which is attributed to the increase of the polymer chains mobility. With increasing temperature, the composites with higher filler loading exhibit a less significant drop in G′, observed as two separate relaxation processes. Hereupon, the observed broadening of the transition region can be related to heterogeneity in the molecular weight between cross-links.⁴⁰ Furthermore, composites containing higher filler loading exhibit a higher rubbery modulus since the rigidity of the filler and the filler/matrix interface is less affected by the temperature than the rigidity of the bulk matrix. That leads to a decrease in resin mobility at the filler/matrix interface which is reflected in higher cross-linking density of the corresponding composites.²⁴

Figure 9.

Temperature dependences of storage modulus.

Generally, the maximum G′ values are found for the composites with 1 wt% filler addition for all the samples (Figure 9). The UPR/c-Al₂O₃ exhibits the highest increase of storage modulus in rubbery plateau region and cross-linking density, 93.6 and 88.3% in comparison to neat UPR resin. Besides, the presence of 1.0 wt% of c-Al₂O₃APTMS-BD filler leads to 67.9 and 61.7% enhancement of $G_{RP}'$ and ν. At filler addition higher than 1 wt%, the storage modulus tends to decrease markedly. Nevertheless, shifted rubbery plateau (T from equation (1)) of composites (T > 235℃) indicates higher cross-linking density compared to neat UPR resin (T > 220℃).³

As the storage modulus of all the composites is higher than the neat UPR, the extent of reinforcing is evaluated using the coefficient of reinforcement (C), calculated by equation (2). The obtained results are given in Table 4.⁴¹

C = \frac{{(G'_{GS} / G'_{RP})}_{composite}}{({G'_{GS} / G'_{RP})}_{resin}}

(2)

Table 4.

DMA results of UPR and the corresponding composites.

Sample	$G_{GS}'$ (MPa)	$G_{RP}'$ (MPa)	T_g (tanδ peak) (℃)	tanδ peak height	ν·10³ (mol/dm³)	Constant (C)
UPR	971	24.4	168.7	0.29	6.0	–
UPR/c-Al₂O₃(a)	901	26.2	182.1	0.25	6.2	0.87
UPR/c-Al₂O₃(b)	890	30.0	189.9	0.23	7.0	0.82
UPR/c-Al₂O₃(c)	1081	35.8	185.3	0.24	8.5	0.77
UPR/c-Al₂O₃(d)	959	48.9	197.3	0.20	11.3	0.50
UPR/c-Al₂O₃(e)	999	32.6	185.7	0.23	7.7	0.78
UPR/c-Al₂O₃APTMS(a)	1009	23.4	170.3	0.25	5.7	1.10
UPR/c-Al₂O₃APTMS(b)	1085	22.3	171.3	0.27	5.4	1.24
UPR/c-Al₂O₃APTMS(c)	1071	25.5	175.1	0.26	6.2	1.07
UPR/c-Al₂O₃APTMS(d)	1021	28.7	175.8	0.22	6.9	0.90
UPR/c-Al₂O₃APTMS(e)	953	27.9	172.7	0.25	6.8	0.87
UPR/c-Al₂O₃APTMS-BD(a)	951	21.4	169.0	0.27	5.1	1.13
UPR/c-Al₂O₃APTMS-BD(b)	1052	25.2	174.0	0.27	6.1	1.06
UPR/c-Al₂O₃APTMS-BD(c)	1148	35.1	184.4	0.24	8.3	0.83
UPR/c-Al₂O₃APTMS-BD(d)	1216	41.8	193.6	0.22	9.7	0.74
UPR/c-Al₂O₃APTMS-BD(e)	1285	31.4	167.0	0.27	7.7	1.04
UPR/Fe-Al₂O₃(a)	882	26.5	182.8	0.25	6.3	0.85
UPR/Fe-Al₂O₃(b)	956	29.8	182.5	0.25	7.1	0.82
UPR/Fe-Al₂O₃(c)	1020	32.7	184.2	0.25	7.7	0.79
UPR/Fe-Al₂O₃(d)	1173	35.2	189.8	0.26	8.3	0.85
UPR/Fe-Al₂O₃(e)	1061	30.5	180.8	0.25	7.3	0.88
UPR/Fe-Al₂O₃APTMS(a)	1036	23.8	175.0	0.25	5.7	1.11
UPR/Fe-Al₂O₃APTMS(b)	966	25.0	175.4	0.26	6.0	0.98
UPR/Fe-Al₂O₃APTMS(c)	1071	28.5	185.6	0.25	6.7	0.96
UPR/Fe-Al₂O₃APTMS(d)	1010	29.9	189.5	0.25	7.0	0.86
UPR/Fe-Al₂O₃APTMS(e)	923	28.3	172.5	0.26	6.9	0.83
UPR/Fe-Al₂O₃APTMS-BD(a)	871	31.0	175.7	0.22	7.5	0.71
UPR/Fe-Al₂O₃APTMS-BD(b)	836	33.1	176.0	0.22	8.0	0.64
UPR/Fe-Al₂O₃APTMS-BD(c)	871	35.0	177.2	0.22	8.4	0.63
UPR/Fe-Al₂O₃APTMS-BD(d)	915	36.3	181.5	0.22	8.6	0.64
UPR/Fe-Al₂O₃APTMS-BD(e)	884	35.3	175.5	0.22	8.5	0.64

UPR: unsaturated polyester resin; APTMS: 3-(aminopropyl)trimethoxysilane; BD: biodiesel.

Generally, the C value decreases as filler loadings increases which implies higher reinforcement effect exhibited in composites containing more embedded alumina particles. These results follow the ones obtained for ν value showing that higher restriction of polymer segment movements is noticed for composites with c-Al₂O₃ and Fe-Al₂O₃APTMS-BD particles. Moreover, cross-linking density of samples which contain APTMS modified particles follows trend obtained for mechanical properties of the corresponding composites, which is a result of the previously discussed phenomena (see the ‘Mechanical properties of UPR/alumina composites’ section).

The broadening of the G″ peak in comparison to the neat UPR resin indicates the influence of filler incorporation (Figure 10). It is attributed to the inhibition of the relaxation process within the composites as a consequence of higher number of chain segments upon filler addition. Other factor potentially originates from the heterogeneous structure of the composites caused by both phase separation during copolymerization with styrene and different intramolecular interactions.⁴² The heterogeneity is mainly attributed to the formation of tighter microgel structures composed of distinct regions of densely cross-linked network distributed among a loosely cross-linked matrix.⁴³ Great dynamic heterogeneities of a thermosetting resins established via formation of a complex structure are also reported, where a bimodal distribution of local relaxation times due to slow and fast relaxing regions is found.⁴⁴

Figure 10.

Temperature dependences of loss modulus.

The peak of tanδ, positioned between 175 and 200℃, originates from the glass transition temperature, T_g, i.e. relaxation corresponding to the neat polyester matrix (Figure 11). The addition of filler causes the appearance of two lower temperature relaxation regions. Maximum damping occurs in a region where most of the chain segments take part in a cooperative motion caused by applied deformation. The transition observed at 75℃ may be related to the motion of the phenyl groups in the styrene sequences. The transition about 120℃ originates mainly from the local modes of polyester units remote from the styrene cross-links. The main segmental relaxation of UPR resin is related to the glass transition of the whole network, while the minor relaxation at a lower temperature is ascribed to the polyester segments between cross-links.⁴⁵ The complexity of the transition region is due to multiple relaxations, arising from the heterogeneous network structure of cross-linked polyesters. The peak intensity of the main relaxation decreases as the transition becomes progressively broader, reflecting a more heterogeneous motional environment with increased filler content. The introduction of cross-links leads to a decrease in the conformational freedom of the chains and results in areas of restricted mobility in the vicinity of the cross-link junctions. The constraint imposed by the cross-links typically manifests itself by increase in T_g with increasing cross-link density.⁴⁶ The temperature location of this relaxation is shifted to higher temperatures in the composite as a result of the mobility reduction caused by filler incorporation. Temperature shifts ranging from 10 to 25℃ are common and the most prominent at high cross-link densities, where the average distance between cross-links approaches the characteristic length scale of local segmental rearrangement.

Figure 11.

Temperature dependences of damping factor. UPR: unsaturated polyester resin.

Composites reinforced with pristine alumina particles exhibit higher T_g values compared to UPR and other composites. The maximum increase of T_g is 16.9 and 12.5% for composites with c-Al₂O₃ and Fe-Al₂O₃, respectively. This suggests establishment of dipolar/hydrogen bonds between filler and polymer matrix which contributes to motions of the restricted macromolecule segments. Although modification processes tune filler surface reactivity, steric hindrance caused by the presence of modifiers plays a significant role, reflecting in higher mobility of polymer segments. Moreover, flexible BD segments in composites with APTMS-BD modified particles participate in overall segment motions which are reflected in lower T_g values (plasticizing effect).

Generally, incorporation of both the two-step modified alumina leads to further improvement of dynamical–mechanical properties of composites with APTMS-BD modified particles. Availability of the NH₂ and OH surface groups within APTMS modified particles, which participate in the creation of hydrogen/dipole–dipole bonding interactions with polymer matrix, is limited as a result of the steric hindrance of APTMS segments (Figure 3). Conversely, this phenomenon is overcome by introducing the fatty acid segments with ethylenic bonds (BD) suitable for establishing covalent bonds with UPR matrix.²⁹

Conclusion

The aim of this work was to obtain the cross-linkable alumina particles capable of establishing the strong interactions with ethylenic segments within the UPR resin structure. The effect of these interactions was traced-out by investigation of mechanical and dynamical–mechanical properties of the resulting composites. Mechanical testing showed that increasing of tensile strength was achieved with filler amount up to 1.0 wt%. The significant increase in σ_t was achieved with incorporation of pristine and modified fillers in comparison to neat UPR resin. Generally, better reinforcing effect was obtained with embedding pristine alumina particles into UPR matrix. Somewhat better mechanical improvement was achieved by the incorporation of the APTMS-BD modified alumina into UPR matrix.

Dynamical–mechanical properties of the composites were significantly improved by introducing pristine/activated alumina particles, which was mostly reflected in elevated T_g. T_g showed lower values for functionalized alumina due to steric hindrance exhibited by modifiers segments attached onto filler particles. Moreover, free BD segments in composites with APTMS-BD-modified particles additionally lowering the T_g values (plasticizing effect).

The obtained eco/multifunctional materials are potentially suitable for application in industry, mining, and construction. According to the mechanical properties, composites with embedded pristine alumina are more suitable for materials which will be subjected to dynamic loadings, while the ones containing modified fillers are better for the materials which will be exposed to high static loadings.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was financially supported by the Ministry of Education, Science and Technological development of Republic of Serbia, Project No. TR34033 and OI172057.

References

Sawpan

Pickering

Fernyhough

. Analysis of mechanical properties of hemp fibre reinforced unsaturated polyester composites. J Compos Mater 2013; 47: 1513–1525.

Kovačević

Rusmirović

Tomić

et al.

New composites based on waste PET and non-metallic fraction from waste printed circuit boards: mechanical and thermal properties. Compos Part B Eng 2017; 127: 1–14.

Rusmirović

Trifković

Bugarski

et al.

High performances unsaturated polyester based nanocomposites: effect of vinyl modified nanosilica on mechanical properties. Express Polym Lett 2016; 10: 139–159.

Soleimani

Zamani

. Surface modification of alumina nanoparticles: a dispersion study in organic media. Acta Chim Slov 2017; 64: 644–653.

Moreira

Braga

Sphaier

. Size effect on the thermal intensification of alumina-filled nanocomposites. J Compos Mater 2016; 50: 3699–3707.

Rusmirović

Ivanović

Pavlović

et al.

Novel modified nanocellulose applicable as reinforcement in high-performance nanocomposites. Carbohydr Polym 2017; 164: 64–74.

Truong

Larsen

Holme

et al.

Dispersibility of silane-functionalized alumina nanoparticles in syndiotactic polypropylene. Surf Interface Anal 2010; 42: 1046–1049.

Guo

Pereira

Choi

et al.

Surface functionalized alumina nanoparticle filled polymeric nanocomposites with enhanced mechanical properties. J Mater Chem 2006; 16: 2800–2808.

Prado

LASA

Sriyai

Ghislandi

et al.

Surface modification of alumina nanoparticles with silane coupling agents. J Braz Chem Soc 2010; 21: 2238–2245.

10.

Lieder

Rashid

. Towards circular economy implementation: a comprehensive review in context of manufacturing industry. J Clean Prod 2016; 115: 36–51.

11.

Kim

Jeong

Son

et al.

Synthesis and applications of unsaturated polyester resins based on PET waste. Korean J Chem Eng 2007; 24: 1076–1083.

12.

Tasić

Rusmirović

Nikolić

et al.

Effect of the vinyl modification of multi-walled carbon nanotubes on the performances of waste poly(ethylene terephthalate)-based nanocomposites. J Compos Mater 2016; 51: 1–15.

13.

Almasi

Ghanbarzadeh

Dehghannya et al.

Novel nanocomposites based on fatty acid modified cellulose nanofibers/poly(lactic acid): morphological and physical properties. Food Packag Shelf Life 2015; 5: 21–31.

14.

Kulomaa

Matikainen

Karhunen

et al.

Cellulose fatty acid esters as sustainable film materials – effect of side chain structure on barrier and mechanical properties. RSC Adv 2015; 5: 80702–80708.

15.

Milanović

Vuksanović

Mitrić

et al.

Electrospun alumina fibers doped with ferric and magnesium oxides. Sci Sintering 2018; 50: 77–83.

16.

Garg

Sharma

Mehta

. Pristine and amino functionalized carbon nanotubes reinforced glass fiber epoxy composites. Compos Part A Appl Sci Manuf 2015; 76: 92–101.

17.

Rusmirović

Radoman

Džunuzović

et al.

Effect of the modified silica nanofiller on the mechanical properties of unsaturated polyester resins based on recycled polyethylene terephthalate. Polym Compos 2017; 38: 538–554.

18.

Lazouzi

Vuksanović

Tomić

et al.

Optimized preparation of alumina based fillers for tuning composite properties. Ceram Int 2018; 44: 7442–7449.

19.

Brzić

Bogosavljević

Ušćumlić

et al.

Effect of tris(2,3–epoxypropyl)isocyanurate on dynamic modulus of CTBN-based composite rocket propellant. Sci Tech Rev 2016; 66: 18–27.

20.

Wang

Eli

Zhang

et al.

A new method for surface modification of nano-CaCO₃ and nano-Al₂O₃ at room temperature. Adv Powder Technol 2010; 21: 203–205.

21.

Jin

Zhang

Sun

et al.

Fabrication and characterization of ethylenevinyl acetate copolymer Al₂O₃ nanocomposites. J Ceram Process Res 2007; 8: 336–340.

22.

Prabhu

Cheirmadurai

Raghava Rao

et al.

Glycine functionalized alumina nanoparticles stabilize collagen in ethanol medium. Bull Mater Sci 2016; 39: 223–228.

23.

Ahangaran

Hassanzadeh

Nouri

. Surface modification of Fe₃O₄@SiO₂ microsphere by silane coupling agent. Int Nano Lett 2013; 3: 23.

24.

Kovačević

Rusmirović

Tomić

et al.

Effects of oxidized/treated non-metallic fillers obtained from waste printed circuit boards on mechanical properties and shrinkage of unsaturated polyester-based composites. Polym Compos 2017. Early View.

25.

Palimi

Rostami

Mahdavian

et al.

Surface modification of Fe₂O₃ nanoparticles with 3-aminopropyltrimethoxysilane (APTMS): An attempt to investigate surface treatment on surface chemistry and mechanical properties of polyurethane/Fe₂O₃ nanocomposites. Appl Surf Sci 2014; 320: 60–72.

26.

Kurth

Bein

. Thin films of (3-aminopropyl)triethoxysilane on aluminum oxide and gold substrates. Langmuir 1995; 11: 3061–3067.

27.

Cao

Lee

. Control of shrinkage and residual styrene of unsaturated polyester resins cured at low temperatures: I. Effect of curing agents. Polymer (Guildf) 2003; 44: 1893–1902.

28.

Shim

Kim

. Reaction mechanism of an unsaturated polyester system containing thickeners. Eur Polym J 2001; 37: 1075–1078.

29.

Wypych

. Handbook of fillers, Toronto, ON, Canada: ChemTec Publishing, 2016, pp. 1035.

30.

Gogoi

Boruah

Dolui

. Jatropha curcas oil based alkyd/epoxy/graphene oxide (GO) bionanocomposites: effect of GO on curing, mechanical and thermal properties. Prog Org Coat 2015; 84: 128–135.

31.

Ramires

Gomes

Jaramillo

. Gel time and polymerization kinetics of unsaturated polyester resin/clay montmorillonite nanocomposites. Polym Compos 2015; 36: 1931–1940.

32.

Costa

TMH

Gallas

Benvenutti

et al.

H. Study of nanocrystalline γ-Al₂O₃ produced by high-pressure compaction. J Phys Chem B 1999; 103: 4278–4284.

33.

Crépy

Chaveriat

Banoub

et al.

Synthesis of cellulose fatty esters as plastics-influence of the degree of substitution and the fatty chain length on mechanical properties. ChemSusChem 2009; 2: 165–170.

34.

Rusmirović

Rančić

Pavlović

et al.

Cross-linkable modified nanocellulose/polyester based composites: effect of unsaturated fatty acid nanocellulose modification on material performances. Macromol Mater Eng 2018; 1700648: 1–12.

35.

Muthu

Dendere

. Functionalized multiwall carbon nanotubes strengthened GRP hybrid composites: improved properties with optimum fiber content. Compos Part B Eng 2014; 67: 84–94.

36.

Boluk

Schreiber

. Interfacial interactions and the properties of filled polymers: I. Dynamic-mechanical responses. Polym Compos 1986; 7: 295–301.

37.

Gunel

Basaran

. Influence of filler content and interphase properties on large deformation micromechanics of particle filled acrylics. Mech Mater 2013; 57: 134–146.

38.

Džunuzović

Pergal

Poreeba

et al.

Analysis of dynamic mechanical, thermal and surface properties of poly(urethane-ester-siloxane) networks based on hyperbranched polyester. J Non Cryst Solids 2012; 358: 3161–3169.

39.

Ornaghi

Bolner

Fiorio

et al.

Mechanical and dynamic mechanical analysis of hybrid composites molded by resin transfer molding. J Appl Polym Sci 2010; 118: 887–896.

40.

Nielsen

Landel

. Mechanical properties of polymers and composites, 2nd edn. New York: Marcel Dekker, 1994, pp. 138.

41.

Devi

Bhagawan

Thomas

. Dynamic mechanical analysis of pineapple leaf/glass hybrid fiber reinforced polyester composites. Polym Compos 2010; 31: 956–965.

42.

Hsu

Lee

. Free-radical crosslinking copolymerization of styrene/unsaturated polyester resins: 1. Phase separation and microgel formation. Polymer (Guildf) 1993; 34: 4496–4505.

43.

Qazvini

Mohammadi

. Dynamic mechanical analysis of segmental relaxation in unsaturated polyester resin networks: effect of styrene content. Polymer (Guildf) 2005; 46: 9088–9096.

44.

Wang

Ediger

. Spatially heterogeneous dynamics in thermoset resins below the glass-transition temperature: effect of temperature and composition. J Polym Sci B Polym Phys 2000; 38: 2232–2239.

45.

Yang

. Viscosities of unsaturated polyester resins: combining effects of prepolymer structure, resin composition, and temperature. J Appl Polym Sci 1996; 60: 2387–2395.

46.

Kalakkunnath S. Viscoelastic relaxation characteristics of rubbery polymer networks and engineering polyesters. PhD Thesis, University of Kentucky, US, 2007.