Features of the microstructure and phase composition of hardened cement paste prepared by mechano-chemical activation of the binder

1 Introduction

Binder activation is one of the most promising methods of developing the physical and mechanical properties of cement composites [1-6]. The best-known methods include: turbulent [7-9], cavitational [10, 11], mechano-chemical [12-21], ultrasonic [22, 23] and vibrational [24, 25] activation. All the methods of activation mentioned above are aimed at increasing the dispersion and raising the surface area of new-growth hydration products.

However, identical dispersion of new-growth hydration products can be achieved by different methods of activation so the mechanical and chemical properties of the cement composites differ [26]. Fine grinding of Portland cement using different equipment has a special effect on its components; there are changes in the nature of lattice destruction and the level of particle aggregation. The energy input for binder dispersion differs when different equipment is used to achieve the same effect. It has been proved that the energy consumption for the fine grinding of cement in an aqueous environment is lower than in a dry one, assuming that a similar dispersion is achieved [27-28].

In this regard it is desirable to use equipment with a high power density and degree of hydrodynamic drag to disperse the water-cement suspension, especially when effective modifying additives are involved.

1.1 Mechanical activation of cement suspensions

Different technologies for dispersing and activating cement binder compounds in an aqueous environment are currently being developed. With the introduction of the rotor-pulsation apparatus (RPA) it is possible to activate water-cement suspensions directly in the RPA [29]. However, the technology involving water-cement activation is not often used in practice due to the lack of study of the effects of mechano-chemical binder activation on the rheology and structure of the cement system.

Authors [30, 31] point out that mechanical ­activation of the cement suspension during the ­initial period of hydration and the formation of the structure of the hardened cement paste contributes to increasing the area of the chemically active coagulation environment and its densification, which increases the strength by up to 30 %. The article [32] emphasizes that the one-day strength of ­sand-cement mortar is increased by 70 % by ­cement suspension activation in an RPA. It is also mentioned [33] that it is better to use binder activation when preparing concrete mixtures because with the traditional method of preparing concrete mixtures the cement grains in the 40-60 μm size range and above remain unhydrated.

It is emphasized that it takes 120 min for a binder dispersion to reach 650 m²/kg when it is treated in a ball mill, while this takes only 20-25 min in a planetary mill [34]. Increasing the binder surface area up to 500-600 m²/kg strengthens the cement composites but any further increase in cement particle surface area causes loss of activity [12, 13]. Binder activation in an aqueous environment is highly desirable to eliminate any loss of cement activity during storage.

Increasing the quantity of gauging/mixing water for binder activation reduces the strength of the cement composites and raises the void volume [17-19]. To eliminate these defects it is necessary to use effective superplasticizers and activate the binder with modifying additives [30-34].

1.2 Mechano-chemical activation

of modified binders

Binder dispersion in an aqueous environment can be intensified by introducing extra surface-active agents (SAA). This dispersion process can be considered as mechano-chemical binder activation (MCA).

Laboratory experiments have shown the significant intensifying effect of SAA on cement grinding. The introduction of some additives during fine grinding not only intensifies the grinding but also improves the physical and mechanical properties of the cement composites obtained.

The number of modifying additives is now growing steadily but it is still difficult to select the most effective SAA for binder dispersion when ­using different equipment.

There are no references in the literature on how mechano-chemical binder activation of the cement suspension affects specific features of the hydration process depending on the activation ­parameters. There has been little investigation of the influence of MCA of the binder on the kinetics of cement paste heat emission, the cement particle size composition after activation or the structure and morphology of the new-growth hydration products. There has also been no study of the importance of highly active superplasticizers for the above-mentioned properties.

The aim of this article is therefore to analyze the results obtained from the mechano-chemical activation of cement suspensions with respect to the hardening kinetics and the physical and mechanical properties of the mass concrete, as well as to examine the specifics and degree of Portland cement hydration by X-ray phase analysis, differential thermal analysis and electron microscopy.

2 Experimental

2.1 Materials and equipment

The following raw materials were used to prepare the concrete paste:

Portland cement ЦЕМ III/A 32,5 H, manufactured at the Ulyanovsk factory and complying with GOST 31108-2003 (National Standards), was used as the binder. Portland cement consists of the following basic minerals: C₃ S – 54 %, C₂ S – 20 %, C₃ A – 11 %, C₄AF – 12 %, pozzolanic mineral admixtures– 9.2 %, SO₃ admixtures – 2.8 %.

Sand from Kamskoe-Ustye with a fineness modulus of 2.7, complying with GOST 8736-2014, was used as the fine aggregate and the 5-20 mm fraction of crushed stone from the Urals, complying GOST 8267-93, was used as the coarse aggregate.

Relamix T-2 naphthalene-formaldehyde superplasticizer, produced in accordance with Technical Requirements 5870-002-14153664-04, was used as the modifying additive in a ratio of 1 % to the mass of cement.

Class B25 mass concrete (cement:sand:crushed stone = 490:555:1315) was chosen for the experiment.

The water/cement ratios of the mixes being studied were chosen to achieve similar consistencies (slump of 6-8 cm as specified in GOST 10181-2014).

Mechano-chemical activation of the cement suspension was carried out in a rotor-pulsation apparatus (RPA) with a driving element speed of 5000 rpm in accordance with Technical Requirements 5132-001-70447062.

The surface area was specified by the PSKh-9 air permeability method and the particles size was measured with a Horiba La-950V2 laser particle analyzer.

The structure of the hardened cement paste was studied with an electron microscope using an AZtec X-MAX energy dispersion spectrometer. The spectrometer has a resolution of 127 eV. The survey of the hardened cement paste surface was carried out using an accelerating voltage of 5 keV. The ­elementary analysis was carried out using an accelerating voltage of 20 keV and a flange focal length of 9 mm, so the penetration depth was less than 1 micron.

The X-ray phase analysis was carried out using a D2 Phaser diffractometer (Bruker, Germany) ­applying Bragg-Brentano geometry using monochromatic CuKa-radiation (λ = 1,54178 Ǻ) in a stepped scan mode.

Measurement and log mode: X-ray tube tension - 30 kV, flow - 30 mA. Scan step - 0.02°. Speed - 1 degree per minute. Scanning angle range in the Bragg-Brentano geometry - 3–60°.

The differential thermal analysis was carried out using an STA 443 F3 Jupiter simultaneous thermal analyzer (Netzsch, Germany) with Netzsch Proteus Thermal Analysis software.

2.2 Methods

The cement suspension was evaporated using a suction funnel (Buchner funnel) connected to an ejector pump in order to obtain dry cement powder and determine its particle size composition obtained after activation in an aqueous environment in the RPA. Immediately after the end of the liquid phase the filter sample was immersed in absolute alcohol and was stored then in acetone, the amount of acetone being used was not less than 5 times the sample volume. The material was then placed in a dryer at a temperature of 105 °C.

Fragments of the hardened cement paste were sprayed with fused Au/Pd in a proportion of 80/20 in a Quorum T150 ES high-vacuum unit in order to study the structure of the paste.

The differential thermal analysis was carried out on a test sample weight of 30-50 mg. The rate of temperature increase was 10 °C/min with a temperature range from 30 °C to 1000 °C.

The following procedure was used to obtain the concrete mixture: approximately 50 % of the cement was mixed with the mixing water and modified Relamix T-2 additive and then mechano-chemically activated in the RPA for 2 minutes. The rest of the cement and the fine and coarse aggregates were introduced into the suspension and mixed in the concrete mixer for 5 minutes [33]. Sample cubes (10 x 10 x 10 cm) were made and tested mechanically after 1, 3 and 28 days of normal curing. The sample strength was defined in accordance with GOST 18105-2010.

The frost resistance of mass concrete was defined as specified in GOST 10060-2012 and harmonized with EN 12390-9:2006. The pore structure was defined as specified in GOST 12730.4-78.

The sulfate resistance of the cement composites was measured using test beams (4 x 4 x 16 cm) made of sand-cement mortar in the proportion cement : sand = 1 : 3. The coefficient of sulfate resistance was defined by comparing mixtures that had hardened in an aqueous environment with mixtures that had hardened in 5 % Na₂SO₄ solution for 180 days, followed by a compressive test.

The kinetics of cement paste heat emission was defined by thermal calorimetry using the Termokhron DS1921G measuring system.

3 Results and discussion

The results of the influence of mechano-chemical activation of the cement suspension binder on the kinetics of the hardening of mass concrete are shown in Table 1. In this table composition 1 is the control; 2 is the composition modified with the Relamix T-2 additive; 3 is the composition that was subjected to mechanical activation without the Relamix T-2 additive, and 4 is the composition obtained by mechano-chemical activation of the binder.

From the results shown in Table 1 it can be seen that the greatest increase in mass concrete compressive strength for all hardening ages is found with composition No. 4 (60 – 249 % when compared with the control composition), especially in the first day of hardening. Mechanical activation of the binder leads to a more significant increase in compressive concrete strength than with the composition modified only with superplasticizer. The authors [31] indicate that with mechano-chemical activation of the binder the increase in strength of the cement composites ranges from 30 to 100 % on the first day of hardening and from 0 to 70 % at 28 dys.

Acceleration of the cement hydration process by mechano-chemical activation of the binder is probably due to the dispersion of its particles. The following compositions were investigated to determine the specific surface area and dispersed particle composition of the cement powder samples obtained after hydration:

1: original Portland cement

2: composition without additives and without mechanical activation

3: composition subjected to mechanical activation

4: composition with Relamix T-2 additive

5: composition subjected to mechano-chemical activation with Relamix T-2 additive

The results of the experiment are shown in Table 2. This table shows that the specific surface area of the cement powder sample subjected to mechanical activation (composition No. 3) increased by 10 % when compared with the composition without mechanical activation (composition No. 2). The specific surface area of the cement powder when Relamix T-2 additive had been added to the cement suspension and it was subjected to mechano-chemical activation, (composition No. 5) increased by 29 % when compared with the composition modified by the Relamix T 2 additive but without activation (composition No. 4).

The average particle size of the original Portland cement (composition No. 1) was 1.26 times larger than for the particles of cement powder subjected to mechanical activation (composition No. 3) and 2.8 times larger than for the particles subjected to mechano-chemical activation in the presence of additives (composition No. 5).

During mechanical activation the yield of the cement fraction smaller than 20 μm increased by 1.34 times when compared with the original Portland cement, while with mechano-chemical activation of the binder the yield of the fraction smaller than 20 μm increased by 1.82 times when compared with the Portland cement. No particles larger than 60 μm were detected when determining the disperse particle composition of the cement powder after mechano-chemical activation.

In [35] the grinding with the RPA achieved an average particle size of 27.5 μm. The differences in dispersion occurred because of the use of various types of Portland cement and dispersion times.

Increasing the strength of cement compositions is of scientific interest for studying the durability of the compounds involved. The frost resistance and indicators of the pore structure of mass concrete were defined (Table 3) as well as the sulfate resistance of sand-cement mortar (SCM) (Table 4) in order to assess the impact of mechano-chemical activation on the durability of cement compositions. The numbering of the compositions in ­Table 3 and 4 corresponds to that in Table 1.

Table 3 shows that mechano-chemical activation of a binder leads to a significant increase in the concrete frost resistance (up to F600) of the mass concrete. This is caused by a decrease in total porosity by 39 %, a decrease in capillary porosity by 74.8 % and an increase in the proportion of closed pores by 53 %

The data presented here show that mechano-chemical activation of the binder helps to increase the sulfate resistance coefficient of cement mortar, and therefore increases its resistance in corrosive environments.

The increase in the strength and durability of cement compositions is due to a change in the morphology of the new-growth hydrated products from the cement paste.

Mechano-chemical activation of the binder leads to the formation of a denser and more fine-grained structure of the hydration products, which is one of the reasons for the increased strength of the mass concrete, especially in the early stages of hardening [36]. The calcium hydrosilicates (CSH) are formed as spherical globules with an average size that is 1.5 - 2 times smaller than in composition No. 3. This is also a strong indicator for high compressive strength of the mass concrete.

X-ray phase analysis and differential scanning calorimetry (DSC) of the samples under investigation were carried out at 1 and 28 days to determine the nature of the new hydration products. The research results are shown in Figures 1 to 4.

From Figure 1 it can be seen that the largest quantity of ettringite (Ca₆Al₂(SO₄)₃(OH)₁₂ x 26H₂O) is formed in composition No. 4, while the smallest is formed in composition No. 1. It is known that in a saturated solution of Ca(OH)₂ ettringite is first released in a finely dispersed colloidal state. This settles on the surface of the 3CaO-Al₂O₃ particles, retards their hydration and extends the cement setting time [37]. This is confirmed by experiments to determine the setting time of the cement pastes. The retardation of the setting time of composition No. 3 can be seen when compared with composition No. 4 (65 minutes for the initial set and 100 minutes for the final set).

The lowest content of the initial clinker minerals Ca₃SiO₅ and Ca₂SiO₄ α-Ca₂SiO₄ (alite and belite) and the largest content of hydrated calcium silicate (CSH) were observed in composition No. 4. This indicates a greater degree of hydration of the cement, which is confirmed by the higher compressive strength of the concrete.

Approximately the same content of initial clinker minerals was observed in compositions No. 1 and No. 2, indicating the absence of retardation of cement hydration in the presence of the Relamix T-2 additive.

The increase in the quantity of calcium hydroxide (Ca(OH)₂), a product of the hydrolysis of clinker minerals, in the composition of the hardening paste indicates an acceleration of the cement hydration. The high content of Ca(OH)₂ is very marked in composition No. 4. The results of quantitative analysis show that the content of Ca(OH)₂ in composition No. 4 is 12 % higher than in composition No. 1 and 2.8 times higher than in composition No. 2.

Testing the hydration of the compositions at 28 days showed an increase in the content of calcium hydroxide (Ca(OH)₂). The greatest amount of calcium hydroxide was found in composition No. 2 and the lowest was found in composition No. 1.

The content of ettringite was increased in composition No. 2, which is due to the sodium sulfate contained in the Relamix -T2 additive, leading to the formation of calcium sulfate dihydrate (Ca₂SO₄*2H₂O). The greatest reduction in the initial clinker phases was observed in composition No. 4: the content of Ca₃SiO₅ was reduced by a factor of 1.66 and the content of Ca₂SiO₄ α-Ca₂SiO₄ by a factor of 1.47, indicating a higher level of hydration of the Portland cement.

Figures 3 and 4 show the differential scanning calorimetry curves. The first endothermic effect was observed at a temperature of 100-105 °C and is associated with the removal of moisture from the pores and capillaries. The second endothermic ­effect was observed at a temperature of 450-460 °C and is associated with the decomposition of calcium hydroxide (Ca(OH)₂). The largest peak value was observed in composition No. 4, obtained by mechano-chemical activation of the binder, which indicates a higher content (Ca(OH)₂) in the composition of the hydration products. This is confirmed by the data obtained by X-ray phase analysis. The third endothermic effect is due to the decomposition of the calcium carbonate (CaCO₃) and was observed at a temperature of 650-670 °C. The data that had been obtained indicated that the content of CaCO₃ after the 1st day of hydration varies slightly in the range of 1-3 % of the total weight. On the 28^th day of ­hydration the largest quantity of CaCO₃ was found in the control composition and the smallest was found in composition No. 4 (26 % lower). The increasing CaCO₃ content indicates carbonization of the cement paste, which is detrimental to the formation of composites that can increase the durability. The DSC shows that the degree of hydration of the cement paste of composition No. 1 increased from 12.35 % to 15.9 %, of composition No. 2 from 8.7 % to 21.9 % and of composition No. 4 from 16.27 % to 26.9 %. At all hardening times the degree of hydration of the cement composition No. 4 was higher than that of the other compositions, resulting in better physical and mechanical characteristics of the cement composites obtained.

The increase in concrete compressive strength of the compositions investigated, especially on the first day of hardening, is due to the increase in the specific surface area of the cement powder samples and the overall reduction in size of the Portland cement fraction. An experiment to determine the kinetics of heat generation in the cement suspension by thermal calorimetry (Figure 5) was carried out to determine the cement hydration characteristics of the compositions investigated.

It can be seen from Figure 5 that mechano-chemical activation of the cement suspension raised the temperature of the hydration by 20-25 °C, so there was a dramatic shift of the temperature peak to the left, indicating an intensification of the hydration processes in the cement paste. In the presence of the Relamix T-2 additive, the cement paste hydration was retarded in the early stages of hardening (composition No. 2), but with the composition No. 4 there was a sharp increase in hydration and the temperature peak increased by 25 °C.

4 Conclusions

Mechano-chemical activation of the binder in the presence of superplasticizer leads to a sharp increase in the compressive strength of mass concrete, especially on the 1^st day of hardening.

The specific surface area of the cement powder samples obtained after mechano-chemical activation of the binder was 29 % higher than that of the control composition. The average particle size was significantly reduced and the number of particles smaller than 20 μm increased by a factor of 1.34.

The frost resistance of the mass concrete obtained by mechano-chemical activation of the binder was increased by a factor of 3 (up to F600). This is due to a decrease in total porosity by 39 %, a decrease in capillary porosity by 74.8 % and an increase in the proportion of closed pores by 53 %. The coefficient of sulfate resistance SCM increased by a factor of 1.4, which indicates an increase in the durability of the compositions obtained by mechano-chemical activation of the binder.

The sharp increase in the strength of compositions obtained by mechano-chemical activation of the binder is due to the formation of a denser, fine-grained, structure of the cement paste; the hydrated calcium silicates (HCS) are mainly formed as finely dispersed spherical globules.

The phase composition of the cement paste obtained by mechano-chemical activation of the binder was characterized by an increased content of calcium hydroxide (Ca(OH)₂), an increased content of ettringite (Ca₆Al₂(SO₄)₃(OH)₁₂ x 26H₂O), and the lowest content of the initial Portland cement clinker phases. The degree of cement hydration was greatest in the composition obtained by mechano-chemical activation of the binder. It increased from 16.27 % on the 1st day of hardening to 26.9 % on the 28th day.