THE ROLE OF THE SURFACE STRUCTURE OF NANOPARTICLES IN THEIR HARDENING OF EPOXY RESIN
A comparative study of the hardening of the cured epoxy resin with SiO2 and Al2O3 nanoparticles obtained by various methods has been carried out. The relationship between the strength of the centres on the surface of the particles, the value of their fractal dimension, and the thickness of the interfacial layer they create is shown.
Polymers are now used in many industries, including biomedicine, batteries, ceramics, composites, magnetism, electronics packaging, solid fuels and adhesives. The inclusion of various fillers [1, 2] can significantly modify such properties as mechanical strength [3, 4], thermal  and electrical conductivity , thermal stability , magnetic characteristics [8, 9], fire resistance [10, 11] and other properties. The properties of such composites depend on the size, shape, nature of the particles, interaction between their constituents, and distribution of the particles in the matrix.
The specificity of nanoparticles is that their surface has a significant number of active centres, which depend on the conditions and method of particle synthesis. Strengthening effects can be observed using the same nanoparticles in combinations with different matrices or different nanoparticles with the same matrix. Different synthesis methods have been shown to form nanoparticle surfaces with different properties while maintaining the phase composition of the material [12, 13]. The relation between the types of surface centres, their strength and interaction of nanoparticles with the dispersion medium was also shown. In particular, there was a significant difference in the strength of active centres for the samples studied, which was reflected in the rheology of epoxy and water based nanofluids.
The process of composite modification by nanoparticles can be divided into chemical and mechanical aspects. Mechanical interaction refers to the effect of particles on the structure of the rigid bonding matrix, its morphology and local density, and chemical (structural) interaction refers to the increase in the number of chemical bonds caused by the introduction of particles.
A prime example of this effect are two types of surface due to the respective active groups: hydrophilic and hydrophobic. Hydrophilic OH-groups are polar and can form hydrogen bonds in the presence of the same groups in the mixture. It is known that epoxy resin does not have such groups and is hydrophobic, but during curing the epoxy groups are broken to form cross-links and free OH-groups. In addition, a number of hardeners can work directly through the OH-groups by incorporating them in the bonding process. The hydrophobic CH groups belong to the same class of compounds as the resin molecules. On the one hand, this should contribute to better mixing of nanoparticles and resin. On the other hand, such groups are not active in the curing process, i.e. the presence of hydrophobic nanopowders does not initiate additional chemical bonds and should not affect the curing process. Thus, hydrophobic particles affect the epoxy composite as solid particles introduced into the resin matrix without explicit chemical interaction with it. For hydrophilic nanopowders there is both chemical and mechanical interaction with the resin.
To take into account the altered morphology of the composite material in the vicinity of the particle-polymer boundary, a material model with spherical and cylindrical inclusions with interfacial layer, scale and adhesion effects (modified Eshelby method) . The parameters of this interfacial layer (its size, elastic modulus and adhesion force to the solid phase) are determined by the effect of active centres present on the nanoparticle surface, e.g., changes in the reaction stoichiometry due to changes in the reactant densities, which may vary greatly depending on the way of its production [12, 13] and particle size distribution. Calculations have shown  that by excluding the influence of the surface layer (filler-matrix interaction), the Young’s modulus is independent of the filler particle size while maintaining the same bulk concentration. When the surface layer is taken into account, a decrease in the filler size caused an increase in the Young’s modulus. This behavior is probably due to an increase in the volume fraction of the surface phase if the size of the inclusion decreases while maintaining its concentration.
The aim of this work is to compare the strengthening effect of SiO2 and Al2O3 nanoparticles of different synthesis methods on cured epoxy resin.
Table 1 shows properties of the nanoparticles used.
ED20 epoxy resin (PolyMax, Russia) was used in the experiment. After doping it with nanoparticles, the suspension was subjected to ultrasound for 30 minutes in a Sapphire ultrasonic bath (Russia) as a particle deagglomeration measure. The PEPA hardener was then added in a ratio of 1 : 10 to the resin mass. Curing took place for 24 hours at room temperature.
The method of instrumental indentation on Nanoscan-4D nanomechanical testing complex (Tisnum, Russia) was used to assess the mechanical characteristics. Indentation experiments were carried out using NanoScan-3D nanohardness meter designed to measure hardness of materials by indentation scales, modulus of elasticity and a number of parameters, including those described in GOST R 8.748-2011. Modulus of elasticity and hardness were determined using the Oliver-Farr method. The method consists in selecting the parameters of a power function describing the experimental dependence of the indentation depth and contact area on the applied force, and calculating the hardness and modulus of elasticity from these data. The microhardness of the cured resin was determined at 70 × 70 µm. A series of pinholes were made on the surface (3 × 3) with a force of 0.03 N. The depth of the indentation was ~ 2 μm.
RESULTS AND DISCUSSION
The difference in the effects of hydrophilic (Ts) and hydrophobic powder (Tsf) on epoxy is demonstrated in Fig.1. It can be seen that when the maximum of the Young’s modulus is reached, the added structural influence has a significant magnitude compared to the mechanical influence.
Fig.2 and 3 show the experimental dependencies of the Young’s modulus of cured resin on the concentration of nanoparticles. From the data obtained it is evident that the maximum of the Young’s modulus of composites is located at different nanoparticle concentrations. That is, the particle-resin interaction has different significance for the particles obtained by different synthesis methods. The intensity of interaction introduced by nanoparticles coincides with the assumptions obtained in , i.e. it is due to the strength of surface centres formed during nanoparticle synthesis. The fastest hardening effect in silicon dioxide is produced by As particles with strong Lewis acid centres. And the weakest delayed effect is shown by Ls particles which surface has only Brensted main centres. The efficiency of particle-resin interaction can be determined by the increase in particle concentration due to the attached layer (see Table 1) calculated from the increase in viscosity corresponding to nanofluid .
A similar situation occurs for samples containing aluminium oxide nanoparticles. Aa particles, having more active surface , harden the composite faster than other particles. The Lа particles having the most passive surface harden the composite at the highest particle concentration. In the same way the effectiveness of particle-resin interaction can be traced through the increase of particle concentration due to the attached layer (see Table 1) .
In [17, 18] a method of calculating polymer hardening by dispersed particles is presented. It takes into account the size of the matrix-filler interfacial layer as well as the interaction force between the particles and the polymer through the thickness of the interfacial layer and the fractal dimension of the particles. Depending on the variation of these parameters, the polymer hardening coefficient can behave as illustrated in Fig.2 and 3. However, it must be taken into account that the interfacial layers of neighbouring particles may overlap at a certain concentration, the particles may agglomerate, etc. In addition, when the method of particle synthesis is changed, same as the activity of their surface, the intensity of interaction between the particles and the medium may vary. The combination of these factors can affect the particle-medium interaction and, consequently, the strength of the composites.
Based on the obtained results, the relationship between intensity of the particle-resin interaction determined by the concentration of maximum Young’s modulus and the fractal dimension calculated in  (Fig.4) was obtained.
It is observed that the maximum of interaction between particles and medium does not correspond to the maximum of fractal dimensionality. It is likely that close to ideal value of fractal dimension corresponds to nanoparticles which are no longer capable of strong interaction. An analogy can be drawn here with the condensation of molecules onto a cluster in the synthesis of nanoparticles. When a cluster reaches the form of a metastable isomer, the probability of condensation of new molecules on it falls dramatically because vacancies disappear and new clusters start growing. Here, too, it happens that a less-than-ideal structure turns out to be more active in its interaction with the dispersed medium. The most active are particles: in SiO2 samples with D = 2.2, and in Al2O3 samples with D = 2.55.
It has been shown that concentrations of the maximums of the composite’s Young’s modulus depend on the surface activity of the nanoparticles, which is determined by their synthesis method.
The surface activity can be reflected through the thickness of the polymer layer attached to the particle, causing an increase in the volume concentration of the dispersed phase. The relationship between the fractal dimension of the nanoparticles and their concentration in the composite at the maximum hardening has been traced.
Within the framework of serial production and for quality control and development of technological processes it is necessary to take into account the method of nanoparticles synthesis and, additionally, control the stability of particle size distribution and fractal dimension value.
The work was carried out on the equipment of SC Nanotechnologies and Nanomaterials” of GSOTU named after acad. M.D. Millionshchikov, with the support of SRC “Mechanics” (Khristianovich Institute of Theoretical and Applied Mechanics SB RAS)
PEER REVIEW INFO
Editorial board thanks the anonymous reviewer(s) for their contribution to the peer review of this work. It is also grateful for their consent to publish papers on the journal’s website and SEL eLibrary eLIBRARY.RU.
Declaration of Competing Interest. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.