Mix Design-Driven Control of Carbonation and Hydration in CO₂-Mixed Cement Pastes: Effects of Water, Slag, and Surfactant

Abstract

This study systematically investigates the influence of mix proportion on and the early-age properties and CO₂ uptake of CO₂-mixed cement paste, focusing on variations in the water-to-binder (w/b) ratio, slag content, and air-entraining agent (AEA) dosage. Mineralogical characteristics were analyzed using X-ray diffraction (XRD) and thermogravimetric analysis (TGA), while pore structures were assessed via nitrogen adsorption. CO₂ uptake was quantified immediately after mixing. Results indicate that a low w/b ratio limits CO₂ dissolution and transport, favors hydration over carbonation, and leads to a coarser pore structure. At moderate w/b ratios, excess free water facilitates concurrent carbonation and hydration; however, thinner water films ultimately hinder CaCO₃ precipitation and C-S-H nucleation. Slag contents up to 30% slightly suppress early carbonation and hydration, while higher dosages significantly delay both reactions and increase capillary porosity. An increasing AEA dosage stabilizes CO₂ bubbles, suppressing immediate CO₂ dissolution and reducing the early formation of carbonation and hydration products; excessive AEAs promotes bubble coalescence and results in an interconnected pore network. An optimized mix design, moderate water content, slag below 30%, and limited AEA dosage enhance the synergy between carbonation and hydration, improving early pore refinement and reaction kinetics.

1. Introduction

The rapid rise in anthropogenic greenhouse gas emissions has intensified the need for carbon mitigation strategies. Cement manufacturing, as one of the most prevalent industrial processes, contributes approximately 9% of global CO₂ emissions, positioning it as a critical sector for decarbonization [1]. Among emerging strategies, carbon capture, utilization, and storage (CCUS) integrated into cement and concrete production has attracted substantial interest for its scalability and efficiency [2,3].

Within this context, CO₂ curing and CO₂ mixing have emerged as promising technologies due to their capacity to induce stable mineralization and reduce the carbon footprint of concrete products. In particular, direct CO₂ injection during the mixing of fresh cementitious materials has shown potential to lower emissions by approximately 10% through the accelerated carbonation of clinker and hydration products [4,5,6]. Furthermore, the in situ formation of nano-CaCO₃ can serve as nucleation sites for C-S-H growth, promoting early-age strength development [7,8].

Previous studies have explored the influence of process parameters such as CO₂ dosage [7,9,10], injection timing [11], and mixing duration [12], as well as alternative carbon sources like dry ice or carbonated water [13,14,15]. However, while the mixing setup has been extensively studied, the effects of mix proportion remain underexplored in the context of CO₂ mixing.

To address this knowledge gap, this study investigates how varying mix design parameters affects the hydration, pore structure, and CO₂ uptake of mixtures with various water-to-binder (w/b) ratios, slag contents, and air entraining agent (AEA) dosages considered. AEAs are used to stabilize the microbubbles within fresh cement paste that are introduced through CO₂ mixing, potentially enhancing the gas–liquid interfacial area and improving CO₂ dispersion during mixing. Different AEAs vary in their surfactant chemistry and can influence CO₂ solubility and bubble stability differently. In this study, sodium dodecyl sulfate (SDS), an anionic surfactant with well-known foaming properties, was selected due to its compatibility with cementitious systems and its effectiveness in stabilizing CO₂ bubbles, as supported by prior studies [16,17]. Mineralogical analyses were conducted using X-ray diffraction (XRD) and thermogravimetric analysis (TGA), and pore characteristics were evaluated via nitrogen adsorption. The results provide new insights into how the synergy between carbonation and hydration can be optimized through mix design for the improved sustainability and performance of cementitious materials.

2. Experiment

2.1. Materials

P•I Portland cement and grade S95 ground granulated blast furnace slag conforming to Chinese standards GB 175 [18] and GB/T 18046 [19] which are manufactured by zhongjiaojianyi Co., Ltd. (Beijing, China), respectively, were used as binders. Deionized water served as the mixing water. The chemical compositions of cement and slag, determined by X-ray fluorescence (XRF), are listed in

Table 1. Chemical compositions of cement and slag (wt.%).

	CaO	SiO2	Al2O3	SO3	Fe2O3	MgO	K2O	LOI
Cement	64.6	18.3	4.6	3.6	3.5	3.2	0.8	1.4
Slag	51.3	21.4	14.1	1.7	1.3	6.3	0.5	1.3

. Their particle size distributions are shown in Figure 1.

Sodium dodecyl sulfate (SDS), supplied by Sinopharm Chemical Reagent Co., Ltd., was used as the AEA. High-purity CO₂ gas (99.99%) was supplied by Wen Dong (Shanghai) chemicals Co., Ltd. (Shanghai, China).

2.2. Mix Proportions and Sample Preparation

Cement pastes were prepared under various mixing conditions, as summarized in

Table 2. Mixing proportions of sample.

Sample	Mixing Atmosphere	Water-to-Binder Ratio	Slag Content	AEA Dosage
A	Air	0.50	/	/
C	CO2	0.50	/	/
C-0.45	CO2	0.45	/	/
C-0.40	CO2	0.40	/	/
C-0.35	CO2	0.35	/	/
C-S1	CO2	0.50	10%	/
C-S3	CO2	0.50	30%	/
C-S5	CO2	0.50	50%	/
C-SDS0.2	CO2	0.50	/	0.2%
C-SDS0.4	CO2	0.50	/	0.4%
C-SDS0.6	CO2	0.50	/	0.6%

. The mixing protocol followed the procedure described in [20]. Cement, slag, and the AEA were pre-blended in sealed containers prior to mixing. For CO₂ mixing, CO₂ gas was introduced continuously during the first 90 s of mixing. To directly evaluate the impact of CO₂ mixing on the performance of cement pastes with various mixing proportions, samples were submerged in ethanal to halt further hydration immediately after mixing, after which samples were vacuum dried for 48 h.

2.3. Testing Method

2.3.1. XRD

Powder XRD analysis was conducted using a Bruker D6 PHASER diffractometer equipped with Cu Kα radiation. Samples were vacuum-dried, ground to below 75 μm, and mixed with 20 wt% corundum as an internal standard. Scanning was performed from 5° to 70° (2θ) at a 0.02° step size and 0.5 s per step. Quantitative phase analysis was carried out via Rietveld refinement.

2.3.2. TGA

TGA was performed using a NETZSCH STA 2500 instrument. Approximately 30 mg of ground sample was heated from 30 °C to 1000 °C at 10 °C/min under a flowing nitrogen atmosphere.

2.3.3. Nitrogen Adsorption

The pore structure of hardened pastes at 72 h was characterized using a BEISHIDE 3H-2000PS2 BET analyzer. Nitrogen adsorption measurements were taken at 77.3 K over a relative pressure range of 0.05–0.30 using a six-point isotherm method.

2.3.4. CO₂ Uptake

The apparent CO₂ uptake (ACU) of samples was calculated from TGA data using the following equation [21]:

$$ACU={\displaystyle \frac{m_{550°\mathrm{C}}-m_{1000°\mathrm{C}}-{0.44\times m}_{initial}}{m_{30°\mathrm{C}}}}\times 100\%$$

(1)

where $m_{30°\mathrm{C}}$, $m_{550°\mathrm{C}}$, and $m_{1000°\mathrm{C}}$ are the sample masses at corresponding temperatures. $m_{initial}$ is the initial mass of CaCO₃ in cement.

For slag-containing samples, the ACU was corrected as

$$ACU={\displaystyle \frac{m_{550°\mathrm{C}}-m_{1000°\mathrm{C}}-{0.44\times (m}_{initial}+m_{initial,slag})}{m_{30°\mathrm{C}}}}\times 100\%$$

(2)

where $m_{initial,slag}$ represents the pre-existing CaCO₃ content in slag.

3. Results

3.1. Mineral Composition

3.1.1. XRD Analysis

Figure 2 presents the XRD patterns and corresponding quantitative phase analysis of cement pastes immediately after mixing. Minor phases such as quartz and periclase are grouped under “others” for clarity. The amorphous content primarily includes C-S-H gel, monosulfoaluminate, and unreacted slag [22].

Compared with the control sample (A), the CO₂-mixed sample (C) shows decreased contents of clinker phases (C₃S, C₂S) and gypsum, indicating accelerated dissolution under CO₂-rich conditions [23]. Concurrently, the reduction in portlandite (CH) and the increase in calcite and ettringite (AFt) suggest carbonation and altered ion equilibria due to the presence of carbonic acid.

Effect of Water-to-Binder Ratio (Figure 2a):

As the w/b ratio decreases from 0.50 to 0.35, the relative contents of clinker and gypsum increase, while those of calcite and AFt decrease. This indicates that CO₂ dissolution and associated carbonation reactions are suppressed at lower w/b ratios due to reduced aqueous phase volume and limited ion mobility. The lowered undersaturation for calcium and silicate species also hinders dissolution and nucleation processes during early hydration.

Effect of Slag Content (Figure 2b):

The incorporation of slag leads to a rise in amorphous content due to its glassy phase. However, increasing slag content reduces calcite formation, attributed to a dilution effect and a lower availability of reactive CaO, which together hinder early carbonation [24].

Effect of AEA Dosage (Figure 2c):

With increasing SDS dosage, calcite and AFt contents decline, while clinker contents rise. This trend suggests that SDS stabilizes CO₂ bubbles in suspension, limiting their dissolution and subsequent carbonation. Moreover, SDS modifies the gas–liquid interface, impeding CO₂ diffusion into the pore solution [16].

3.1.2. TGA/DTG Analysis

Figure 3 displays the TGA/DTG curves of samples. Key endothermic peaks appear at ~110 °C (C-S-H and AFt dehydration), ~160 °C (gypsum dehydration), ~460 °C (CH dehydration), and ~680 °C (CaCO₃ decarbonation), with the relative masses at different temperatures illustrated in

Table 3. Mass loss at various temperatures.

Sample	Relative Mass at Different Temperature/%
	350 °C	550 °C	1000 °C
A	98.76	98.27	97.01
C	97.91	97.53	95.39
C−0.45	98.16	97.76	95.84
C−0.40	98.55	98.23	96.36
C−0.35	98.37	98.09	96.30
C−S1	98.52	98.29	96.43
C−S3	98.62	98.15	96.00
C−S5	98.84	98.53	96.70
C−SDS0.2	98.41	97.85	96.38
C−SDS0.4	98.38	97.66	95.98
C−SDS0.6	98.52	98.01	96.52

[25].

Compared with the control sample, sample C shows enhanced peaks at 110 °C and 680 °C, indicating an increased formation of C-S-H/AFt and carbonates, while the CH peak is diminished, confirming the competitive consumption of Ca²⁺ by carbonation.

Effect of Water-to-Binder Ratio (Figure 3a):

Decreasing the w/b ratio reduces the intensity of the first (C-S-H/AFt) and third (carbonate) peaks, reflecting reduced carbonation. The CH peak initially increases at w/b = 0.45, then declines, indicating an optimal hydration–carbonation balance around this ratio. Excessively low water content (<0.40) slows both hydration and carbonation.

Effect of Slag Content (Figure 3b):

All hydration-related peaks decrease with higher slag content, suggesting reduced reactivity. While the main CaCO₃ peak weakens, a broader peak at >800 °C becomes more prominent, indicating the formation of more stable polymorphs such as aragonite or vaterite [26]. The reduced OH⁻ release from slag lowers alkalinity, suppressing CO₃²⁻ formation and favoring dissolved carbonic acid retention.

Effect of AEA Dosage (Figure 3c):

Samples with increasing SDS content show diminishing peaks for C-S-H/AFt and carbonate. The CH peak reduction caused by CO₂ injection is also less pronounced at higher SDS dosages. This supports the view that stabilized CO₂ bubbles reduce gas dissolution, thereby limiting both carbonation and hydration kinetics. Additionally, SDS may chelate Ca²⁺, further suppressing CH and C-S-H precipitation.

3.2. Pore Structure Analysis

Figure 4 presents nitrogen adsorption/desorption isotherms for cement pastes cured for 72 h. All samples exhibit a steep increase in adsorption within the relative pressure range P/P₀ = 0.48~1.0, suggesting the dominance of capillary and transition pores. The cumulative pore volume and pore size distribution of samples are shown in Figure 5. Compared to the control sample, sample C exhibits a higher cumulative pore volume in the 2 nm–200 nm range, primarily due to an increased pore size distribution peak between 30 nm and 70 nm. This suggests that CO₂ mixing predominantly enhances the volume of transition pores (10 nm–100 nm [27]) within the cement paste matrix.

Effect of Water-to-Binder Ratio (Figure 5a):

The sample with a w/b ratio of 0.45 shows the highest cumulative pore volume, indicating enhanced pore connectivity at this intermediate water content. As the w/b ratio decreases, the pore size distribution broadens and shifts toward larger pores (100 nm–1000 nm), reflecting increased capillary porosity. This shift results from a reduced CO₂ uptake (limiting nano-CaCO₃ nucleation) and reduced fluidity of fresh paste, leading to poor particle packing and a less dense microstructure [28].

Effect of Slag Content (Figure 5b):

Samples with slag contents of 10% and 30% exhibit minimal changes in total porosity. However, at 50% slag content (C-S5), the cumulative pore volume increases markedly, especially for pores >20 nm. This increase is attributed to lower cement content and hinders early C-S-H formation, which delays pore refinement.

Effect of AEA Dosage (Figure 5c):

The incorporation of SDS increases cumulative pore volume due to the stabilization of entrained CO₂ bubbles. At 0.4% dosage (C-SDS0.4), the total porosity is the highest, dominated by pores >20 nm. The unexpected result in Figure 5c, where C-SDS0.4 shows a higher pore volume than C-SDS0.6, may be attributed to the non-linear behavior of SDS at high concentrations. Excessive SDS may lead to micelle formation, reducing surface activity and limiting additional bubble stabilization. Additionally, the excessive AEA dosage leads to bubble coalescence and upward migration, reducing effective CO₂ utilization and creating interconnected voids. Over time, these bubbles may either escape or collapse, contributing to the formation of larger pores.

3.3. CO₂ Uptake

The ACU of samples was calculated from TGA data and is summarized in Figure 6.

Effect of Water-to-Binder Ratio:

A clear inverse correlation exists between CO₂ uptake and the w/b ratio. As the ratio decreases from 0.50 to 0.35, the ACU declines by 40.4%. This reduction reflects limited aqueous volume for CO₂ dissolution and reduced ion mobility, which hinder carbonation during mixing.

Effect of Slag Content:

Slag incorporation produces a non-linear trend. The ACU initially decreases slightly with 10–30% slag content, reaching a minimum (4.9% reduction) at 30% (C-S3), which appears to be the most favorable blend. Beyond this threshold, further slag addition significantly reduces carbonation, likely due to the dilution of reactive Ca²⁺ sources and reduced alkalinity.

Effect of AEA Dosage:

Increasing SDS dosage from 0.2% to 0.6% significantly suppresses the ACU, with reductions ranging from 52.1% to 75.3% relative to the sample C. This substantial decrease confirms that the bubble-stabilizing action of AEAs limits CO₂ dissolution and impedes early carbonate formation.

4. Discussion

4.1. Effect of Water-to-Binder Ratio

The w/b ratio plays a dual role in CO₂-mixed cement systems by simultaneously affecting hydration kinetics and carbonation efficiency. At moderate w/b ratios, sufficient free water facilitates the rapid dissolution of CO₂, allowing its transformation into HCO₃⁻ and CO₃²⁻ ions, which diffuse to cement particle surfaces and precipitate as CaCO₃. This process occurs concurrently with hydration, supporting the nucleation and growth of C-S-H and AFt [29,30].

However, as the w/b ratio decreases, the reduced volume of pore solution limits CO₂ solubility and ion mobility, which suppresses the formation of CaCO₃ and delays carbonation [31]. This shift favors hydration as the dominant reaction pathway, as evidenced by the reduced carbonate content and increased portlandite at intermediate ratios (Figure 3a).

The pore structure is also strongly influenced by water content. Insufficient water at low w/b ratios hinders the formation of nano-CaCO₃ particles, reducing their nucleation effect for C-S-H formation and slowing early-age gel growth [32,33]. Additionally, lower paste fluidity leads to poor particle dispersion and increases voids due to entrapped air, resulting in a coarser pore network.

4.2. Effect of Slag Incorporation

Slag incorporation increases the amorphous fraction in the paste due to its glassy nature. However, high slag content (>30%) significantly reduces early-age carbonation and hydration. This effect arises from two main factors, namely (i) a dilution of clinker, which reduces reactive Ca²⁺ sources, and (ii) slower slag activation in early hydration stages, which contributes less to OH⁻ release and alkalinity, both essential for CaCO₃ formation [34,35].

Although carbonation is suppressed, slag particles with a high surface area provide nucleation sites for late-stage hydration and carbonation [36]. This may partially compensate for early reductions in reaction rates but does not prevent the broadening of pore size distribution observed in high-slag mixes (Figure 5b). Thus, an optimal slag content near 30% achieves a balance between reduced CO₂ uptake and preserved matrix integrity.

4.3. Effect of AEA Incorporation

The addition of SDS significantly alters carbonation behavior and pore development. As a surfactant, SDS adsorbs at gas–liquid interfaces, stabilizing CO₂ bubbles and preventing their rapid dissolution. This inhibits the immediate formation of carbonic acid and, consequently, carbonate precipitation.

The AEA preferentially adsorbs at the air–liquid interface, reducing surface tension and creating a stabilizing film around gas bubbles [37,38]. For the gas phase in fresh paste, this film suspends CO₂ bubbles, significantly decreasing the rate of CO₂ dissolution into the pore solution immediately after mixing. With fewer dissolved CO₂ particles available to form carbonic acid, the carbonation around the cement particle surface is delayed. Beyond the bubble stabilization, the complexation effect of the AEA in a solution partially bonds with Ca²⁺; as an excessively high dosage of AEA is added, it decreases the precipitation of hydration and carbonation products [39].

Moreover, SDS may complex with Ca²⁺ ions in a solution, reducing their availability for hydration and carbonation reactions. The result is a marked reduction in early reaction products, as seen in both XRD and TGA data (Figure 2c and Figure 3c). Excessive AEAs leads to bubble coalescence and upward migration, forming interconnected voids that increase total porosity (Figure 5c).

The pore network generated by entrained CO₂ bubbles shifts the dominant pore size to >20 nm, compromising mechanical performance and increasing permeability. Therefore, AEA dosage must be carefully optimized to avoid the degradation of structural integrity while ensuring effective gas distribution.

5. Conclusions

This study systematically investigated the influence of mix proportion, specifically the w/b ratio, slag content, and AEA dosage, on the carbonation behavior, hydration, pore structure, and CO₂ uptake of cement paste subjected to CO₂ mixing. The following key conclusions are drawn:

(1)

Lower w/b ratios limit CO₂ dissolution and mass transport, shifting the reaction pathway toward hydration. This results in reduced carbonate formation and coarser pore structures. At moderate w/b ratios, concurrent carbonation and hydration are promoted, but excessively low water content hinders both reaction kinetics and pore refinement.

(2)

Incorporating slag increases the amorphous content and influences early-age reaction rates. Up to 30% slag causes minimal reductions in carbonation and hydration. However, higher contents significantly delay both processes and lead to wider capillary pores due to the dilution of reactive phases and reduced alkalinity.

(3)

Increasing SDS dosage stabilizes CO₂ bubbles but significantly reduces immediate CO₂ dissolution and associated reaction products. Entrained gas increases porosity, particularly in the 20 nm–200 nm range, and promotes an interconnected pore network if overdosed.

(4)

The early-age performance of CO₂-mixed cement paste depends on a well-balanced mix design. A moderate w/b ratio, slag content below 30%, and restrained AEA dosage collectively enhance carbonation–hydration synergy, improve pore structure, and support effective CO₂ sequestration.

These findings provide a scientific basis for tailoring mix designs to maximize early-age performance and CO₂ sequestration efficiency in carbonated cement-based systems. The insights are particularly valuable for developing low-carbon concrete technologies aligned with carbon neutrality goals in the construction industry.