Keywords

1 Introduction

Concrete, the most utilized construction material and the second most used substance on earth after water, creates significant amount of CO2 emissions during production [1]. The CO2 emissions primarily originate from the calcination of limestone (CaCO3) to produce cement (main binder material in concrete), releasing CO2 in the process. As CO2 contributes to global warming, there has been a strong focus on the decarbonation of concrete, and reducing the amount of cement per cubic meter of concrete is a proven approach to deliver on the target CO2 reduction. Reducing the cement content can be achieved through partial cement substitution with supplementary cementitious materials (SCMs) such as slag or fly ash. Slag, being hydraulic in nature like cement, can be used at higher substitution rates than other SCMs (usually ≥ 50%), translating to better CO2 reduction. However, although slag notably improves the later-age strength as well as most concrete durability properties, including chloride ingress [2], the alkali–silica reaction [3,4,5], sulfate resistance [6, 7], and delayed ettringite formation [8], its use results in increased susceptibility of the concrete to carbonation [9,10,11].

Carbonation refers to the ingress of CO2 into the binder system, which results in the formation of carbonic acid (H2CO3) that further dissociates into H+ and CO32− and reacts with calcium ions in the pore solution, resulting in the precipitation of calcium carbonate (CaCO3) and a decrease in the pH of the pore solution. Low concrete pH (≤9.5) resulting from carbonation is detrimental to steel-reinforced concrete because steel begins to lose its passivation layer at low pH, making it susceptible to corrosion [10]. Thus, carbonation is a serious durability concern, particularly for steel-reinforced concrete [12].

Phenolphthalein is an indicator used to assess the depth of carbonation. It is colorless at lower pH values (≤9), whereas at pH > 10.5, it presents a characteristic purple or magenta. Because mortar or concrete that has been carbonated has pH ≤ 9, phenolphthalein is used to visually confirm the drop in pH due to carbonation [13]. The effect of carbonation is, however, not confined to the change in pH of the pore solution. Although it has been reported that carbonation in general is beneficial and results in an increase in compressive strength due to the conversion of Ca(OH)2 to CaCO3, which increases the volume of the binder and reduces porosity [10], it appears that this may not be true for all binder systems. It has been reported that although moderate carbonation can improve the mechanical properties of the concrete, excessive carbonation impairs mechanical strength due to the decalcification of the C-S-H [14]. High-slag concrete is also particularly susceptible to carbonation shrinkage [9]. Therefore, due to variability in the reported effect of carbonation and considering the increasing levels of slag being used in concrete production, a better understanding of the effect of carbonation on microstructure and phase development is required.

We investigated the effect of carbonation on the microstructure and phase development of high-slag mortars (50% and 70% slag replacement). The mortars were characterized after being subjected to accelerated carbonation conditions (2%CO2, 23 °C, and 50% relative humidity (RH) for 112 days.

2 Methods

2.1 Raw Materials

We used General Purpose cement and slag that complied with AS3972 and AS3582.2 respectively. Normen sand (CEN Standard Normsand according to EN196-1) was used as fine aggregate. The maximum moisture content of sand was 0.2%.

2.2 Carbonation Test

The 40 × 40 × 160 mm mortars were prepared using Normen sand in combination with ordinary Portland cement (OPC) and OPC + slag binders (slag at 50% and 70% replacement) at 0.45 water to cement ratio. Table 1 shows the mortar mixes investigated in this study.

Table 1 Mortar mixes
Full size table

The mortars were demolded after 1 day, cured for 28 days inside a sealed moisture bag and then transferred into the shrinkage room (50%RH, 23 °C) to air cure for 7 days in preparation for the accelerated carbonation test. At age 35 days (28 + 7 days), the mortars were transferred into the carbonation chamber running at 2%CO2, 50%RH and 23 °C for the accelerated carbonation test. Carbonation depth measurements using phenolphthalein (1% solution) were carried out after 1 week (7 days), 4 weeks (28 days), 9 weeks (63 days), and 16 weeks (112 days) exposure of the mortars in the carbonation chamber.

2.3 Characterization of the Mortars

After 112 days carbonation, the “colorless” and “pink regions” of the mortars were subjected to thermogravimetric analysis (TG; SDT-Q600 Simultaneous TGA/DSC equipment, TA Instruments) and scanning electron microscopy (SEM). The mortar specimens were ground and 50-mg samples of the ground material were transferred to a platinum crucible, which was placed inside the TG instrument. The thermal analysis was performed in a nitrogen gas atmosphere, within a temperature range of 23–900 °C and at a heating rate of 10 °C/min.

All imaging and elemental analyses of the fractured mortars were performed using a Zeiss Supra 55VP SEM fitted with a Bruker SDD EDS Quantax 400 system and FEI Quanta 200 with Bruker XFlash 4030 EDS detector. The microscopes were operated at 15 kV accelerating voltage and 12.5 mm working distance.

3 Results and Discussion

Figure 1 shows the carbonation depths at 16 weeks (112 days). As expected, the plain OPC mortar showed the best resistance to carbonation (largest pink region) while mortars with slag exhibited poorer resistance. The higher the slag replacement, the poorer the carbonation performance.

Fig. 1
Three photographs of plain O P C mortar, O P C plus 50 percent slag mortar, and O P C plus 70 percent slag mortar display the carbonation depths at 16 weeks.

Photos of the mortars (OPC, OPC + 50% slag and OPC + 70% slag) at 112 days exposure in the carbonation chamber showing the carbonation depth determined using phenolphthalein

Full size image

Figure 2 shows the increase in carbonation depth over time, with the mortar with 70% slag consistently having the highest carbonation depth and fully carbonated at 63 days.

Fig. 2
A multi-bar graph depicts depth in millimeters at 0, 7, 28, 63, and 112 days for O P C, O P C plus 50 percent slag, and O P C plus 70 percent slag. The depth is low on day 0 and high on day 112 for all the cases. The depths at days 63 and 112 are the same in O P C plus 70 percent slag.

Carbonation depth measurements before carbonation (Day 0) and after 7, 28, 63 and 112 days exposure to 2%CO2 at 23 °C and 50% relative humidity. OPC, ordinary Portland cement; SL, slag

Full size image

Figure 3 shows the TG curves of the non-carbonated (pink) and carbonated parts (colorless) of the plain OPC mortar, mortar with 50% slag and the mortar with 70% slag. Mass loss in the range of calcium hydroxide (CH) decomposition at ≈400–500 °C corresponds to the dehydroxylation of CH, (Ca(OH)2 → CaO + H2O) [15]. Therefore, the area under the curve at ≈400–500 °C corresponds to the amount of CH, and thus a larger area means more CH. Comparing the CH content of plain OPC mortar and the 50% slag mortar (pink regions), OPC notably has more CH, as may be expected, which explains its better carbonation resistance. During carbonation, portlandite, which is the most soluble source of calcium in the binder, serves as a buffer and maintains the pH of the pore solution by dissolving and releasing OH− and Ca2+ ions. The OH− neutralizes the H+ while Ca2+ binds CO32−, precipitating CaCO3 [11]. Therefore, due to the lower amount of portlandite, high-slag binders carbonate much faster (i.e. pH drops faster) than pure cement. Absence of portlandite in the “colorless regions” of the plain OPC and 50% slag mortars are notable indicating the full consumption of portlandite in the carbonated regions. There is also no portlandite remaining in the 70% slag mortar, consistent with it being fully carbonated (top and middle areas of the mortar were tested). Moreover, consequent to the full consumption of portlandite in the “colorless regions” of all mortars (i.e., fully carbonated regions), there was a drastic increase in the amount of CaCO3. A notable decrease in the amount of C-S-H, carboaluminates, and ettringite (TG region 0–300 °C) can also be seen, because once portlandite has been fully consumed, the other calcium-bearing phases start to react with CO2 and carbonate as well [10]. Decalcification of the C-S–H can occur, resulting in carbonation shrinkage [12].

Fig. 3
3 graphs labeled a, b, and c depict the derivative mass loss percentage per degree Celsius versus temperature in degrees Celsius for plain O P C, 50 percent slag mortar, and 70 percent slag mortar. Graphs a and b plot derivative mass loss for colorless and pink regions, and they indicate C a (O H) 2 and C a C o 3.

Derivative thermogravimetric curves of the carbonated and non-carbonated part of plain OPC, 50%SL mortar and 70%SL mortar. OPC, ordinary Portland cement; SL, slag

Full size image

Figure 4 shows the amount of CaCO3 in the different binder systems (carbonated regions) calculated from the decarbonation region (CaCO3→CaO + CO2). The higher the amount of CaCO3 formed, the higher the CO2 binding capacity. CO2 binding capacity is related to the amount of CaO in the binder and because the higher the slag replacement, the lower the CaO available, the CO2 binding capacity also decreases.

Fig. 4
A vertical bar graph plots the percentage of C a C O 3 in O P C, mortar plus 50% slag, and mortar plus 70% slag. The percentages of C a C O 3 in O P C, mortar plus 50% slag, and mortar plus 70% slag are 17, 12, and 10.

Percentage CaCO3 in the plain OPC, 50%SL mortar and 70%SL mortar determined from the thermogravimetric mass loss measurements. OPC, ordinary Portland cement; SL, slag

Full size image

SEM images of the fractured “carbonated” 50% and 70% slag mortars are shown in Figs. 5 and 6 respectively. The presence of aragonite (a CaCO3 polymorph) is very prominent in both systems. The microstructure also appears to be porous, although the change in porosity due to carbonation should be quantified.

Fig. 5
Two scanning electron microscopy images indicate the porous microstructure and the presence of C a C O 3 and C S H.

Scanning electron microscopy images of the carbonated 50%slag mortar

Full size image
Fig. 6
Two scanning electron microscopy images indicate the porous microstructure and the presence of C a C O 3 and C S H. The images have 70 percent slag mortar.

Scanning electron microscopy images of the carbonated 70%slag mortar

Full size image

4 Conclusions

In this study we investigated the effect of carbonation on the microstructure and phase development of high-slag binders. Relevant results are as follows.

  1. 1.

    High-slag binder carbonates faster than plain OPC binder. The higher the slag content, the higher the carbonation rate.

  2. 2.

    TG analysis confirmed the absence of portlandite and increased CaCO3 in the carbonated regions (“colorless” regions), which suggests that the trigger for the drop in the pH (i.e., the change in color of phenolphthalein from “pink” to “colorless”) is the absence of portlandite to buffer the pH.

  3. 3.

    Significant reduction in the amount of C-S-H was also observed, suggesting decalcification of C-S-H, and that all calcium-bearing phases are prone to carbonation.

  4. 4.

    The higher the slag replacement, the lower the CO2 binding capacity, which is consistent with the reduced CaO content of high-slag binders as well as their higher carbonation rate.

  5. 5.

    SEM images confirmed the prominent presence of aragonite (a CaCO3 polymorph) in the carbonated mortars.