Optimal Mixing Ratios of Silica and Hybrid Resin with Epoxy Resin for Concrete Floor Surface Covering

Unlike the alkoxy group, which forms a direct bond with metals, an alkoxide cannot participate in hydrolysis and condensation reactions. As a result, the composition acts as a modifying agent for the mesh structure of inorganic materials. The organic and inorganic composite materials are divided into the following three types based on the method of adding the organic material to the mesh structure of the inorganic materials:

The first type can be easily obtained through a hydrolysis and condensation reaction of an organic metal alkoxide, which constitutes the organic composition. However, strictly speaking, it is not classified as an organic–inorganic composite material because the organic material does not form a network (Cornelius and Marand 2002). In the third type, a three-dimensional mesh structure of inorganic material is formed through hydrolysis and a condensation reaction of a metal alkoxide, independently of the three-dimensional mesh structure of organic material through the polymerization of the added organic monomers. Therefore, this material cannot be classified as an organic–inorganic composite, either, but is rather classified as an organic and inorganic composite.

The second type can be classified as an organic–inorganic composite material. If an $R$ (organic) group containing an epoxy group, a vinyl group, or other such reactors is not hydrolyzed, it is polymerized via the reactions of the organic constituents under heat or ultraviolet (UV) irradiation. Additionally, hydrolysis and condensation reactions occur at the same time. The question is whether it is possible to fabricate materials wherein the mesh structures of inorganic and organic materials are chemically combined in a molecular unit, because the organic composition of an organic metal alkoxide and the added organic monomer can be polymerized by adding a third organic monomer. Therefore, these types of materials are organic–inorganic composite materials in the truest sense. Because the mesh structures of inorganic and organic materials are connected with each other through a chemical bond, phase separation does not occur, and a new function can be implemented as each component is uniformly distributed in the molecular unit (Seo and Choi 2007; Song et al. 2011).

The bond strength test method followed KS F 4937 (KSA 2009a). The attachment, $40\times 40\mathrm{mm}$ iron, was attached on the surface of a specimen using an epoxy bond, and a hand grinder was then used to cut around it. Finally, it was pulled up at a tension speed of $5\mathrm{mm}/\min$ from the surface by universal testing machine (UTM). The results of the S.R. bonding strength test are shown in Figs. 3 and 4.

The bond strength test results showed that both S.R._A and S.R._B, when coated on the surface of dried concrete (ADC), possessed an overall higher bond strength compared to wet concrete (WCC), regardless of curing time. Based on a curing time of 28 days, ADC was the reference value—meaning 100%—and thus WCC shows decrease of the bond strength of 49.5 (S.R._A) and 21.5% (S.R._B) of the reference value. In the case of WCC, this large difference was likely caused by the high-strength condition of the concrete and by the fact that the penetration of the epoxy and hybrid resins into the capillaries of the concrete surface was not affected by water (water membrane). This trend was also seen in the mortar specimens [air-dried mortar (ADM) and wet mortar (WCM)].

Based on a curing time of 28 days, it was found that in the case of S.R._B, the bond strength increased by 14.5% in wet mortar (WCM) and 34.2% in wet concrete (WCC) compared to the corresponding values for S.R._A. Regarding the failure mode after the bond test, S.R._B showed a breakage of the floor material, together with the floor concrete (interfacial decohesion), as shown in Fig. 5(a). This was because the silane coupling agent reacted even when the amount of moisture present on the adherent’s surface was very small because of its characteristic structure, thereby forming a siloxane group on the surface. As a result, a strong chemical bond with the inorganic materials was created. This was also because the other side of the end group had a functional group that could react with organics, which plays a role in combining organic and inorganic materials.

In contrast, although S.R._A showed a breakage of the floor material along with the floor concrete in the final dried floor, both the mortar and concrete specimens were simply separated from the floor concrete (interfacial deadhesion) in the wet floor, as shown in Fig. 2(b). This separation from the floor concrete occurred because moisture in the mortar or concrete surface was present during the curing process, which created obstacles to an active reaction between the base and the curing agent. It is also possible that the water membrane formed at the interface of the hardened body disturbed the capillary penetration required for the formation of the strong root of the epoxy resin.

An impact resistance test was conducted with epoxy-coated specimens using a base of dried or wet concrete after 14 days of curing in accordance with KS F 2622. The test used 500 g iron weights falling from heights of 1.0 and 1.5 m. Fig. 6 shows the results of the impact resistance test.

For S.R._A, as shown in Figs. 6(a and b), the results of the impact resistance test showed holes and cracks at both ends, regardless of the fall height. The hardened body, coated under wet conditions, clearly had cracks at the surrounding edges. This phenomenon occurred because of brittle fracture, which is characteristic of epoxy resin. The brittle epoxy was easily destroyed by the instantaneous impact because of the high cross-linking density, which caused a low absorption capability against external impact, and thus the epoxy-hardened body could not absorb or distribute the impact. On a wet surface, resistance to impact was lower because of a degradation in the bond strength; as a result, cracks and fine ruptures around the holes were larger.

Based on these results, it can be concluded that brittle polymer resins, such as epoxy, may be strengthened by adding fillers to increase their toughness, but they are still vulnerable to impact. This is because chemical absorption rarely occurs, and physical absorption occurs through weak bond energies between the filler and the matrix segment.

For S.R._B, as shown in Figs. 7(a and b), the results of the impact resistance test, performed after 14 days of curing following coating of the hybrid on the base of the dried and wet concrete, showed hollow holes, regardless of the fall height. However, the surface crack at both ends observed for S.R._A was not seen. This result was achieved because the brittle fracture feature of the epoxy resin was reduced by the buffering action of the silane coupling agent contained in the silane coupling composite resin. The agent connected the silica and the epoxy composite to facilitate the absorption and distribution of external impact through a chemical bond.

An abrasion resistance test was conducted in accordance with KS F 4041 and KS F 2813. The revolution and total revolution of the taper are 50 revolutions per minute (rpm) and 500 times at prior testing, and it was checked on the mass change and surface state every 200.

The S.R._A abrasion resistance test results, shown in Fig. 8, indicated an average abrasion of $0.14\mathrm{mg}/{\mathrm{mm}}^2$, which is close to $0.15\mathrm{mg}/{\mathrm{mm}}^2$, the quality criterion of KS F 4041, which is a self-leveling mortar. S.R._B had an average abrasion of $0.09\mathrm{mg}/{\mathrm{mm}}^2$, which satisfied the quality criterion. Compared to an average abrasion of $0.14\mathrm{mg}/{\mathrm{mm}}^2$ for S.R._A, S.R._B had 35.7% better abrasion resistance because of the strong interfacial force formation from the bonding of the GPTMS with the curing agent, which promoted coupling between the epoxy bond body and the silica interface.

The mixing ratio of the epoxy resin and silica (Nos. 7 and 10) for the flow values and interfacial deadhesion test was set; the basic mixing ratio was fixed at $4:1$ (base:curing agent), and the mixing ratio of silica was increased from 5.0 to 7.5 g in steps of 0.5 g, giving six values of resin/silica ratios ($1:1.0$ to $1:1.5$). Figs. 9 and 10 show the results of the flow values and the occurrence of interfacial deadhesion per mixing ratio of C.R._A and C.R._B.

For the mixing ratio of C.R._A, silica Nos. 7 and 10 had flow values of more than that observed for the KS F 4041 quality criterion of 190 mm in all the mixes. For silica No. 7, interfacial deadhesion did not occur at the $1:1.4$ and $1:1.5$ ratios. For silica No. 10, interfacial deadhesion did not occur at the $1:1.3$ and $1:1.5$ ratios.

For the mixing ratio of C.R._B, silica No. 7 showed that interfacial deadhesion occurred in all the mixing ratios, but silica No. 10 had no interfacial deadhesion.

Therefore, in a 1:1.3 mixed sample, shown in part No. 7 part in Fig. 9, interfacial deadhesion did not occur. However, a somewhat good flow value was seen. Further, interfacial deadhesion did not occur in the $1:1.3$ mixed sample (part No. 10 in Fig. 9), which also had a good flow value. As a result, an additional test was conducted to select the appropriate mixing ratio for Nos. 7 and 10 to suppress shrinkage after curing.

For C.R._A, a mixing mass ratio of $\mathrm{epoxy}:\mathrm{silica}$ (Nos. 7 and 10) was fixed at $1:1.3$, while the ratio of silica Nos. 7 and 10 was adjusted to determine the flow values and interfacial deadhesion. Because the interfacial deadhesion results were the same with both types of silica for C.R._B, the best flow value, observed for a S.R._B:silica ratio of $1:1.0$, was used and the mixing ratio (mass ratio) of silica Nos. 7 and 10 was adjusted to determine the flow values and interfacial deadhesion presence.

The configuration of the mixing ratio of C.R. and silica Nos. 7 and 10 is shown in Table 6; Fig. 11 shows the flow values and interfacial deadhesion test results at each mixing ratio.

For the C.R._A n2 mix, it was found that while the flow value had a high flexibility of 257 mm, no interfacial deadhesion was found. For the C.R._B n1 mix, the flow value showed 249 mm of flexibility and no interfacial deadhesion. Therefore, in the C.R._A mix, to obtain an appropriate ratio of Nos. 7 and 10, the ratio of silica added to the epoxy resin was $1:1$ (mass ratio). In C.R._B, to obtain an appropriate ratio of Nos. 7 and 10, the ratio between the silica added to the epoxy and the silane coupling agent synthesis resin was $2:3$ (mass ratio).

In subsequent tests conducted to investigate the physical properties (e.g., bond strength, impact resistance, flexural strength, compressive strength, length change rate), the mixing ratio of silica Nos. 7 and 10 used in all the C.R._A mixtures was that of C.R._A n2, shown in Table 6, and for C.R._B the mixing ratio of C.R_B n1 was used. The resin/silica mixing ratio was increased between $1:1$ and $1:2$ (mass ratio) by adding 0.25 wt% silica to obtain five different specimens.

The bond strength test method followed KS F 4937. The test results of the bond strength of C.R._A are shown in Figs. 12 and 13.

The broken floor material in C.R._A, together with the floor concrete, was shown in the dried mortar (ADM) and dried concrete (ADC), regardless of changes in the amount of silica added, once it exceeded $1.2\mathrm{N}/{\mathrm{mm}}^2$, the quality criterion of KS F 4937 was above 179.87% and 265.29%, on average, respectively. On the other hand, the wet mortar (WCM) and wet concrete (WCC) showed overall interfacial falling off, regardless of the changes in the amount of silica added, and all the mixing ratios, except for $1:1$ and $1:1.75$, were below the KS standard. Therefore, for the wet mortar and concrete base, the epoxy resin and silica additions to the epoxy resin did not improve the bond strength. At a $1:1.75$ mixing ratio, the bond strength improved in the wet concrete, but gradually decreased as the curing continued, as shown in the dried concrete. This phenomenon is disadvantageous in terms of the long-term durability of the floor. At mixing ratios of $1:1.25$ and $1:1.2$, a similar trend was seen in the wet concrete.

As described previously, the water membrane formed on the surface of the specimen is the likely cause of the interfacial falling off in the mortar and concrete base under wet conditions. The membrane suppresses the formation of strong roots because of capillary penetration of the epoxy resin, along with the inhibition of the bonding reaction between the epoxy resin base and the curing agent by the moisture. This phenomenon was the same as the interfacial falling off in the single epoxy resin without silica.

In C.R._B, the overall strength increased as compared to that of C.R._A. The bond strength in the dried mortar and dried concrete destroyed the base, exceeding the KS quality standard by approximately 181.67 and 283.83%, respectively. Further, in the $\mathrm{C}.\mathrm{R}.\_\mathrm{B}:\mathrm{silica}=1:1–1:1.75$ (mass ratio) range in the wet mortar and concrete base, the bond strength result exceeded the standard by approximately 145.88 and 188.04%; all the samples showed the falling apart of the floor material together with the floor concrete.

In the case of the wet concrete (WCC), mixing ratios of $1:1$, $1.:1.5$, and $1:1.75$, except for strength at curing ages of 14 days and 28 days at a $1:2$ mixing ratio, showed higher bond strength distribution compared to that of the dried mortar (ADM). This result occurred because the hybrid resin reacted with fine moisture in the surface of the adherent, thereby creating a cross link between the resin-based coating film, which did not form strong roots, and the concrete base to improve the bond strength in C.R._B in a wet-base condition compared to that in C.R._A.

Fig. 14 shows an additional observation of the curing condition after coating the mortar base based on the silica ratio prior to the bond strength test. Fig. 14 shows that epoxy resin:silica mass ratios of $1:1.75$ and $1:2$ had somewhat different curing conditions after coating on the wet mortar base; however, both of them showed foam generation. This foam generation could prevent a good finish and degrade the long-term durability.

A C.R. impact resistance test was conducted at a curing age of 14 days after coating on the wet concrete and wet concrete bases in accordance with KS F 2622. The test result was summarized as three types: dent, micro crack, and normal, as shown in Table 7 (see also Figs. 6 and 7).

For C.R._A. the samples in which silica was added to the epoxy resin showed cracks around the edge of the hollow hole, regardless of the fall height in the concretes under dry and wet conditions, irrespective of the change in the amount of added silica. This was due to the fact that dispersion was not easy owing to the interfacial decohesion in the epoxy matrix when silica was simply added to the epoxy resin, so that the cross-linking density at that region degraded, thereby creating cracks in the area of impact. Therefore, the addition of silica alone cannot solve the problem of brittle fracture, the greatest drawback of epoxy resins.

For C.R._B, the impact of a spherical weight that has fallen from a height of 1 m showed no damage to the concrete base under dry and wet conditions, regardless of the amount of silica added to the hybrid resin. However, near a hybrid resin:mixing silica ratio of $1:2$ (mass ratio), cracks were found around the hollow hole owing to the impact of a spherical weight falling from a height of 1.5 m.

This result can be explained based on the fact that the silane coupling agent contained in the hybrid resin connected the epoxy matrix and silica to form an interfacial force by means of a chemical bond between them, so that the stability of the silica particles increased and a uniform dispersion phase formed, facilitating the buffering action owing to absorption and dispersion against the external impact. As observed at a mixing ratio of hybrid resin:mixing silica of $1:2$ (mass ratio), this phenomenon can also be explained by the fact that an increase in the silica content may reduce the cross-linking density because of the aggregation of silica owing to the high specific surface area of silica. Therefore, it was found that the hybrid resin could complement the weakness of the epoxy resin, easy-demonstrating the brittle failure by the external impact.

The compressive strength and flexural strength test method followed KS L 5105 and KS F 2408. Figs. 15–18 show the results of the compressive strength and flexural strength tests of the C.R.

The compressive strength of the specimens of C.R._A and C.R._B showed somewhat increasing trends as the curing ages increased. In C.R._A, all the mixing ratios except for $1:1$ exceeded $100\mathrm{N}/{\mathrm{mm}}^2$ at 28 days of curing age, whereas in C.R._B, all the mixing ratios had values below $100\mathrm{N}/{\mathrm{mm}}^2$. These results showed that the mesh structure of C.R._B, which had a diversified tissue structure because of the addition of silica, was somewhat disadvantageous in terms of compressive strength compared to C.R._A, which has a fine tissue structure because it consists of only a polymer resin.

The difference in the mixing ratio showed that in C.R_A, a relatively high compressive strength was observed for mixing ratios of $1:1.25$ and $1:1.75$, whereas in C.R._B, a mixing ratio of $1:1.25$ showed the highest compressive strength. In particular, both C.R._A and C.R._B showed an increased compressive strength up to a mixing ratio of $1:1.25$ based on the amount of added silica. However, for ratios greater than $1:1.25$, no compressive strength increase was observed. Even in C.R._B, the compressive strength decreased up to a mixing ratio of $1:1.75$. Based on this result, increasing the amount of added silica minimally affects the compressive strength, and $1:1.25$ should be considered the appropriate level.

The flexural strength of C.R._A showed that samples with added silica had somewhat higher results in a range of $29.45–48.30\mathrm{N}/{\mathrm{mm}}^2$ at 3 days of curing age. It also showed an increasing distribution up to 7 days of curing, followed by a decrease till the 14th day of curing. Then, at 28 days of curing, the flexural strength varied in the range $22.24–34.04\mathrm{N}/{\mathrm{mm}}^2$. Therefore, the overall flexural strength showed less distribution than that observed at 3 days of curing. At 28 days of curing, the strength increased because the amount of added silica was changed; however, the degree of change was minimal.

On the other hand, in the case of C.R._B, most of the mixing ratios showed values above $30\mathrm{N}/{\mathrm{mm}}^2$ of the flexural strength, thus demonstrating overall improvement in the flexural strength compared to that observed in C.R._A. Mixing ratios of $1:1.25$ and $1:1.75$ indicated better flexural strength compared to other mixing ratios. In particular, the initial strength at $1:1.75$ was approximately 34% higher than that at $1:1.25$. At all ages, a deviation of less than around 4% was observed, which suggested stable flexural strength values. This flexural strength characteristic can reduce the crack occurrence probability with regard to the initial deformation (behavior and deflection) after the completion of construction and can accelerate the overall project schedule, including subsequent operations, which is considered advantageous. In the case of a $1:1.25$ mixing ratio, the initial strength was lower than that observed for a mixing ratio of $1:1.75$; however, it increased gradually as the material aged by approximately 6% at 28 days. Although this result was within the error range, a mixing ratio of $1:1.25$ can maintain stable performance better than a mixing ratio of $1:1.75$, which demonstrated a gradual decrease as the material aged following the initial strength.

It can be conclusively stated that flexural strength was greatly affected by the mesh structure formed as a result of adding mixed silica to the hybrid resin. This result was consistent with that of a previous study (Ban et al. 1990), which reported more than a 20% increase due to silica addition.

This result indicated that while the compressive strength of the epoxy resin increased as the material age increased, it became vulnerable to brittle fracture, as confirmed by the impact resistance test. In the case of C.R._B, flexural strength improved owing to the buffer action of the silane coupling composite resin.

As shown in Fig. 19, C.R._A showed that the length change rate decreased as the amount of added silica increased. This result matched with the result in a previous study (Park et al. 2006), where the length change rate of epoxy resin decreased as the amount of added silica increased. In C.R._B, the effect of a change in the amount of added silica on the length change rate was relatively low. Both C.R_A and C.R_B showed that a mixing ratio of $1:1.75$ produced the lowest length change rate compared to other mixing ratios, for example, $1:1.25$, which showed an approximately 52.5% difference in the length change rate. When the length change rate in C.R._B is small, based on the characteristics of floor-finishing materials of parking lots constructed in a broad area, this is an excellent strength for reducing cracks in floor materials owing to the difference in the length change rate and dried shrinkage.