According to the Intergovernmental Panel on Climate Change, global greenhouse gas emissions continue to rise and have reached a net anthropogenic value of 59 ± 6.6 Gt CO2-eq in 20191. The energy sector accounts for over 60% of such emissions, leaving nearly 40% of global emissions to other industry sectors2,3. The cement and concrete industries play a significant role in this context, with ordinary Portland cement production alone being the largest emitter and the source of ~8% of the global anthropogenic CO2 emissions4. Meanwhile, ordinary Portland cement has a high potential for CO2 sequestration owing to its inherent alkalinity, promising to turn the cement and concrete industries into carbon sinks via the conversion of CO2 into solid crystals mainly consisting of calcium carbonate (CaCO3). Stable carbonate crystals have a longer lifetime than structures made of cement and concrete, thus representing an optimal means to store CO2 in such structures5,6.

Over the past decades, various efforts have focused on the reduction of the CO2 emissions deriving from the cement and concrete industries by working on the whole life cycle of these materials, from the extraction of the involved raw materials to their processing7,8, manufacturing9,10, use11,12, and end of life13,14,15,16. In this context, two main categories of approaches have been developed to store CO2 in the cementitious materials that constitute concrete during the manufacturing of such a mixture: hardened and fresh concrete carbonation approaches. Hardened concrete carbonation approaches were introduced in the 1970s17,18 and gained limited attention until recently but remain confined to laboratory studies due to technological limitations. Fresh concrete carbonation approaches were introduced more recently19,20,21 but already benefit from commercial pilots due to technological advantages19,22.

Hardened carbonation approaches involve concrete preparation and mixing, concrete casting, concrete exposure to CO2 gas, and concrete hydration23,24,25. The pressure, temperature, duration, and general environmental conditions at which CO2 is placed in contact with hardened concrete distinguish the different types of hardened concrete carbonation approaches. Besides these differences, all hardened concrete carbonation approaches involve chemical reactions that engage the active cementitious constituents of hardened concrete with the conversion of CO2 gas into solid CaCO3 crystal forms. Nevertheless, as these approaches are driven by diffusion-controlled kinetics, they are limited by the minimal extent of CO2 diffusion in hardened concrete, which enables carbonation reactions from the surface of hardened concrete to a depth of a few millimeters26. Pressurized vessels, usually between 1 and 5 atm27, can improve the carbonation degree and penetration depth24,28,29. However, the resulting cost-effectiveness diminishes as the applied pressure increases30, with CO2 uptake rates that remain within 5–20%24 at the expense of significant energy and infrastructure needs. Furthermore, the nature of these approaches limits their applicability to precast concrete, as they require CO2-rich or high-pressure environments31.

Fresh concrete carbonation approaches involve concrete preparation and mixing with the simultaneous injection of CO2 gas, concrete casting, and concrete hydration20,32,33. With these approaches, chemical reactions engage fresh concrete with the conversion of CO2 gas into solid CaCO3 crystal forms. Therefore, two main advantages characterize fresh compared to hardened carbonation approaches: (1) the possibility to carbonate a greater volume of concrete via CO2 injections during mixing; (2) the possibility to incorporate relatively easily CO2 injection nozzles in industrial processes.

Considering the advantages of fresh compared to hardened concrete carbonation approaches, increasing experimental investigations have been performed in recent years to maximize their performance19,20,21,32,33,34,35,36. However, the understanding of the reactions, kinetics, and mechanisms involved in fresh concrete carbonation approaches remains elusive, leading to limited CO2 capture efficiencies20. This evidence, together with the fact that energy consumption induced by CO2 pumping, purification, transport, and high-pressure vessels can involve greater emissions of CO2 gas compared to those that are sequestered5, involves that the carbon capture potential of fresh concrete carbonation approaches remains largely untapped.

Motivated by the opportunity to improve the fundamental understanding and efficacy of fresh concrete carbonation approaches, this work presents an experimental laboratory study in support of an approach to carbonate fresh concrete that harnesses a key methodological difference compared to the state-of-the-art. The difference between the proposed approach and state-of-the-art approaches is that CO2 gas is injected into a diluted cement suspension instead of fresh concrete, which completely changes the carbonation reaction kinetics. Specifically, the proposed approach resorts to a multi-stage concrete preparation method whereby CO2 gas is first injected into a cement suspension, the resulting material is then used to achieve a cement paste of desired properties, and concrete is finally obtained by adding and mixing aggregates. The results of laboratory experiments (see the “Methods” section) indicate that injecting CO2 gas into a cement suspension, rather than the conventional method of injecting CO2 into fresh concrete, significantly accelerates constituent reactions and CaCO3 precipitation kinetics. Therefore, under specific treatment conditions that are precisely identified via the developed experiments, not only the properties of the carbonated concrete can remain unchanged, but the carbon footprint associated with the production of such material can be reduced compared to current standards. This discovery supports more efficient concrete carbonation approaches and advances the understanding of mineralization processes, which are widespread in materials science and engineering.


Figure 1 presents scanning electron microscopy (SEM) images to illustrate the impacts of CO2 gas injections with different volumes (2000, 4000, and 8000 standard cubic centimeters, i.e., scc; corresponding to 3.9, 7.9, and 15.7 g of CO2) on the morphology of cement particles constituting the concrete mix design selected for this work (see the “Methods” section for details). The results illustrate the features of cement particles deriving from (a) a cement suspension that did not undergo any CO2 injection (i.e., an uncarbonated cement suspension); (b–d) a cement suspension that underwent a CO2 injection (i.e., a carbonated cement suspension); (e) a cement suspension that did not undergo any CO2 injection and was subsequently mixed with additional cement (i.e., an uncarbonated cement paste); and (f–h) a cement suspension that underwent the injection of CO2 and was subsequently mixed with additional cement (i.e., a carbonated cement paste).

Fig. 1: SEM images of cement particles from suspension or paste, uncarbonated and carbonated.
figure 1

The cement suspension is made of 50 g of cement and 100 mL of deionized water (DI-H2O), involving a water-to-cement ratio of w/c = 2. The concrete is prepared from the suspension by first adding and mixing 50 g of virgin cement, achieving a cement paste with w/c = 1, and then adding another 100 g virgin cement (w/c = 0.5) and mixing of aggregates. See the “Methods” section for details. a Uncarbonated suspension. b Carbonated suspension with 2000 scc of CO2. c Carbonated suspension with 4000 scc of CO2. d Carbonated suspension with 8000 scc of CO2. e Uncarbonated paste. f Paste deriving from a carbonated suspension with 2000 scc of CO2. g Paste deriving from a carbonated suspension with 4000 scc of CO2. h Paste deriving from a carbonated suspension with 8000 scc of CO2. Red arrows: crystal structures.

By comparing Fig. 1a with Fig. 1b–d, it is possible to assess the influence of CO2 injections on the morphology of the particles that constitute a cement suspension. By comparing Fig. 1e with Fig. 1f–h, it is possible to assess the influence of CO2 injections on the morphology of the particles that constitute a cement paste deriving from a carbonated cement suspension. By comparing Fig. 1a–d with Fig. 1e–h, it is possible to assess the influence of adding cement to a given suspension, which may or may not be carbonated with CO2 gas, to achieve a cement paste for use in concrete.

The uncarbonated cement suspension (Fig. 1a) and paste (Fig. 1e) show the typical morphology of cement hydration products, with irregular shapes, rough surfaces, and no visible crystals. The injection of CO2 into a cement suspension and the addition of cement to such suspension influence the morphology and size of cement particles, respectively. As a result of the carbonation, crystals deposited on the surface of most cement particles become visible (Fig. 1b–d and f–h, respectively). Such crystals are uniformly found on the surface of cement particles derived from a cement suspension (Fig. 1b–d), whereas they appear both on the surface and interlocked within the bulk of cement particles deriving from a cement suspension mixed with additional cement after carbonation (Fig. 1f–h). Larger amounts of injected CO2 gas yield the formation of a larger number of crystals (Fig. 1b–d and f–h). The crystals found on cement particles that derive from cement suspensions always resemble needles for increasing volumes of injected CO2 (i.e., Fig. 1b–d). In contrast, the crystals found on cement particles that derive from cement pastes resulting from carbonated cement suspensions resemble needles for relatively limited volumes of injected CO2 (i.e., for 2000 and 4000 scc, Fig. 1f, g) but become flakes for larger volumes of CO2 (i.e., for 8000 scc, Fig. 1h). Larger and more polydisperse cement particles are obtained when cement is added to a suspension after its preparation, irrespective of whether such suspension does not undergo carbonation (Fig. 1a vs. Fig. 1e) or does undergo carbonation (Fig. 1b–d vs. Fig. 1f–h). Cement particles that derive from carbonated cement suspensions have a size <10 µm (Fig. 1b–d), whereas cement particles that derive from virtually equivalent carbonated cement suspensions and subsequently mixed with more cement have a size that can exceed 20 µm (Fig. 1f–h). Energy dispersive spectroscopy (EDS) analyses (Supplementary Fig. 1) show that the observed crystals have the major elemental compositions of O (45.1–45.6%), Ca (33.0–34.3%), C (13.1–13.8%), Si (3.0–3.5%), and Fe (1.5–1.7%). Other detected elements including S, Al, and K have elemental weights <1%.

Figure 2 illustrates the results of (a) thermogravimetric analyses (TGA), (b) differential thermal analyses (DTA), and (c) X-ray diffraction (XRD) spectra obtained for both carbonated and uncarbonated cement suspensions. The TGA curves reveal two distinct temperature ranges characterized by significant weight losses (Fig. 2a). The first range is approximately located between 380 and 440 °C, whereas the second is between 520 and 740 °C. The first range corresponds to a dehydration range associated with the loss of absorbed water in the range of 105 to 400 °C and the CH dehydration in the range of 400–500 °C37. The second range corresponds to the decarbonation process typically taking place between 530 and 900 °C38. The peaks in the DTA curves at the corresponding temperatures (Fig. 2b) corroborate these results. Uncarbonated cement suspensions exhibit the highest dehydration peak and the lowest decarbonation peak. As the amount of injected CO2 increases, the dehydration peak decreases while the decarbonation peak intensifies, with no visible dehydration peak being observed for 8000 scc of injected CO2. The decarbonation peaks shift to higher temperatures with an increasing amount of injected CO2. Meanwhile, the decarbonation peak temperatures for uncarbonated cement and carbonated cement with 2000, 4000, and 8000 scc of CO2 are 656, 668, 678, and 702 °C, respectively. The XRD spectra (Fig. 2c) substantiate the results discussed so far by indicating the presence of calcium carbonate in the form of calcite and aragonite in the carbonated cement, along with the typical constituents of cement represented by C3S, C2S, ettringite, C–S–H, and portlandite; additionally, these results enable to remark the lack of visible peaks of calcite and aragonite in the uncarbonated cement. Therefore, the observed crystals on the surfaces of carbonated cement particles (Fig. 1) are CaCO3 and not cement hydration products. As the amount of injected CO2 increases, the relative intensity of most of the peaks also increases, indicating an increase in the overall crystallization degree.

Fig. 2: Characterization of carbonated and uncarbonated cement suspensions.
figure 2

a TGA curves of cement suspensions. b DTA curves of cement suspensions. c XRD spectra of carbonated and uncarbonated cement suspensions.

Figure 3 illustrates key variables related to the efficacy of the proposed approach: (a) pH and temperature variations of the suspensions during their preparation and carbonation; carbon capture performance expressed as (b) CO2 capture efficiency, (c) CO2 capture mass, and (d) carbonation degree of the resulting concrete as a function of the CO2 treatment duration; and influences on (e) absolute unconfined compressive strength of uncarbonated and carbonated concrete and (f) normalized strength of carbonated relative to uncarbonated concrete.

Fig. 3: Physico-chemical variations, carbon capture performance, and concrete strength involved with the developed approach.
figure 3

a Suspension pH (n = 3 independent experiments) and temperature at different stages of the preparation process (raw data reported in Supplementary Tables 1 and 2). b CO2 capture efficiency with different CO2 injection durations (n = 3 independent experiments; raw data reported in Supplementary Table 3). c Carbonation degree of carbonated cement with different CO2 injection durations (n = 3 independent experiments; raw data reported in Supplementary Table 4). d CO2 capture mass per unit volume of concrete with different CO2 injection durations (n = 3 independent experiments; raw data reported in Supplementary Table 5). e Unconfined compressive strength of the concrete specimens as a function of the curing time and the volume of injected CO2 (n = 10 independent experiments; raw data reported in Supplementary Table 6). f Normalized strength relative to the uncarbonated material (raw data reported in Supplementary Table 7). Error bars show standard deviations.

The preparation of the cement suspension starts with the use of DI-H2O only, which has a pH of 7 at an ambient room temperature of 21.2 °C (Fig. 3a). Next, DI-H2O is saturated with CO2 gas, yielding a pH of ~5.3 without any significant temperature change. Afterwards, cement is added to the CO2-saturated DI-H2O (with no concurrent CO2 injection), leading to an increase in pH up to the value of 12.60 ± 0.05 and a surge in the solution temperature from 21.2 to 23.1 °C under stabilized conditions. Finally, CO2 is injected in the cement suspension for a time of up to 40 min, leading to an instantaneous decrease in pH from 12.60 ± 0.05 to 12.18 ± 0.08 during the considered time, which increases up to the value of around 12.6 under stable conditions (see the “Methods” section). After 40 min of carbonation, the stable and instant pH values are 12.18 and 12.63, corresponding to OH concentrations of 0.015 and 0.043 M, respectively. Temperature drastically increases at the beginning of the CO2 injection, but such an increase becomes less pronounced as carbonation progresses, stabilizing around 39.5 °C after 30 min.

The CO2 capture efficiency decreases for longer injection durations, from 45.4% for a 10-min injection (2000 scc) to 26.2% for a 40-min injection (8000 scc) (Fig. 3b). Despite this decrease, the carbonation degree increases with the amount of injected CO2 (Fig. 3c), starting from 3.3% when cement suspensions are uncarbonated and placed in contact with CO2 from ambient air and achieving 17.2% for a CO2 injection lasting 40 min. The absolute mass of captured CO2 also increases with a higher amount of injected CO2 (Fig. 3d). Specifically, 5.2, 7.7, and 12.0 kg of CO2 per unit volume of concrete are captured during 10, 20, and 40 min of injection, respectively.

Figure 3e, f shows the unconfined compressive strength of concrete specimens obtained from uncarbonated and carbonated cement suspensions after 3, 7, 28, and 56 days of curing. The strength shows a typical increase with the curing time due to ageing, irrespective of whether the concrete derives from cement suspensions that underwent carbonation or not (Fig. 3e). Meanwhile, reference to the average strength of the prepared concrete specimens allows us to identify several noteworthy trends. Although the average strength of carbonated concrete is always higher than the strength of uncarbonated concrete at early stages of curing (i.e., at 3 days), such strength can be higher or lower than the average strength of uncarbonated concrete at successive stages of curing (i.e., at 7, 28, and 56 days) depending on the amount of injected CO2 (Fig. 3f). Concrete deriving from cement suspensions free of carbonation displays average strengths of 22.9 ± 4.0, 31.0 ± 2.5, 35.4 ± 2.7, and 36.5 ± 4.8 MPa after 3, 7, 28, and 56 days, respectively. After 3 days of curing, the average strength of concrete deriving from cement suspensions that underwent carbonation is at least 10% higher than the average strength of uncarbonated concrete. After 7, 28, and 56 days, the average strength of concrete deriving from cement suspensions that underwent carbonation is generally smaller (down to 20.7%) than the average strength of uncarbonated concrete for 2000 scc and 4000 scc of injected CO2. In contrast, after 7, 28, and 56 days, the average strength of carbonated concrete is generally higher (up to 23.1%) than the strength of uncarbonated concrete for 8000 scc of injected CO2. Besides these considerations, an analysis of the standard deviation of these latter data suggests that the actual average strength of carbonated concrete may not be statistically higher than that of uncarbonated concrete. Meanwhile, it does indicate that the strength of carbonated concrete is at least not compromised compared to the strength of uncarbonated concrete for 8000 scc of injected CO2.


CaCO3 precipitation and its impact on concrete strength

The strength of concrete strongly depends on the constituents and the chemical reactions characterizing such material. The proposed carbonation approach induces changes in the constituents and reactions that characterize concrete, which influence the structure and properties of such material.

In general, CaCO3 precipitates in three main polymorphs: calcite, aragonite, and vaterite. Calcite is the stable polymorph and exhibits thermostability39,40. Aragonite is generally less stable than calcite and transitions to such polymorph at a relatively slow rate under ambient conditions39. Vaterite is unstable and hence uncommon in natural environments, transforming into aragonite and calcite at temperatures ranging from 0 to 30 °C and 60 to 80 °C, respectively39,40. CaCO3 is recognized as calcite when it decomposes in TGA within the range 650–900 °C, whereas is considered as aragonite when it decomposes within the range 530–650 °C38. Meanwhile, weight losses in TGA between 500 and 700 °C are attributed to poorly crystalline CaCO3, whereas weight losses between 700 and 900 °C are attributed to highly crystalline CaCO337,41.

The results of this work indicate that carbonated cement suspensions subsequently used to manufacture concrete undergo the precipitation of CaCO3 crystals that vary in quantity and morphology (Fig. 1). The quantity of CaCO3 precipitations essentially depends on the amount of injected CO2 and increases for larger volumes of injected CO2. The morphology of CaCO3 precipitations depends on the crystal polymorph type and the crystallization degree (i.e., the ratio of crystalline to amorphous precipitated minerals). As TGA results support that decarbonation occurs between 530 and 900 °C (Fig. 2a), the achieved CaCO3 precipitations are a mixture of calcite and aragonite, as also substantiated by the XRD results (Fig. 2c). The results also indicate a shift toward higher temperatures of the decomposition and decarbonation peaks for different CO2 treatments (Fig. 2b). Thus, changes in the polymorph type and crystallization degree take place.

The formation of a mixture of different CaCO3 polymorphs can be attributed to two formation pathways of such compounds due to the hydration and carbonation of concrete. During the hydration and carbonation of concrete, which occur nearly simultaneously (although carbonation is faster than hydration23), two main compounds contribute to the precipitation of CaCO3: calcium silicate hydrate (C–S–H) and calcium hydroxide (Ca(OH)2). Another two compounds, represented by calcium aluminate (Ca3Al2O6) and calcium aluminoferrite (Ca4Al2Fe2O10), can also form as a result of CO2 injections but typically in a significantly smaller fraction than calcium silicates. The reason is that calcium aluminate has a fast hydration reaction rate but remains rather inactive with CO242, and calcium aluminoferrite also shows limited reactivity with CO243. Since C3S and C2S are practically insoluble in water, the direct carbonation of them is rather slow.

The hydration reaction of calcium silicates (tricalcium silicate, C3S; dicalcium silicate, C2S) yields C–S–H and Ca(OH)2 as follows:

$$2\left(3{{{{\rm{CaO}}}}}\cdot {{{{\rm{SiO}}}}}_{2}\right)+6{{{\rm{H}}}}_{2}{{{\rm{O}}}}\to \,3{{{{\rm{CaO}}}}}{{\cdot }}2{{{{\rm{SiO}}}}}_{2}{{\cdot }}3{{{\rm{H}}}}_{2}{{{\rm{O}}}}+3{{{{\rm{Ca}}}}}{\left({{{{\rm{OH}}}}}\right)}_{2}$$
$$2\left(2{{{{\rm{CaO}}}}}\cdot {{{{\rm{SiO}}}}}_{2}\right)+4{{{\rm{H}}}}_{2}{{{\rm{O}}}}\to \,3{{{{\rm{CaO}}}}}{{\cdot }}2{{{{\rm{SiO}}}}}_{2}{{\cdot }}3{{{\rm{H}}}}_{2}{{{\rm{O}}}}+{{{{\rm{Ca}}}}}{\left({{{{\rm{OH}}}}}\right)}_{2}$$

In turn, Ca(OH)2 reacts with CO2 and forms CaCO3 as follows:

$${{\rm {Ca}}}{\left({{\rm {OH}}}\right)}_{2}+{\rm {C{O}}}_{2}\,\to {{\rm {CaC}{O}}}_{3}$$

In addition, C–S–H can also react with CO2 and form CaCO3 as follows:

$$3{{{{\rm{CaO}}}}}{{\cdot }}2{{{{\rm{SiO}}}}}_{2}{{\cdot }}3{{{\rm{H}}}}_{2}{{{\rm{O}}}}+{{{\rm{CO}}}}_{2}\,\to {{{{\rm{CaCO}}}}}_{3}+2{{{{\rm{CaO}}}}}{{\cdot }}2{{{\rm{Si}}}}{{{\rm{O}}}}_{2}{{\cdot }}3{{{\rm{H}}}}_{2}{{{\rm{O}}}}$$
$$3{{{{\rm{CaO}}}}}{{\cdot }}2{{{{\rm{SiO}}}}}_{2}{{\cdot }}3{{{\rm{H}}}}_{2}{{{\rm{O}}}}+2{{{\rm{CO}}}}_{2}\to 2{{{{\rm{CaCO}}}}}_{3}+{{{{\rm{CaO}}}}}{{\cdot }}2{{{{\rm{SiO}}}}}_{2}{{\cdot }}3{{{\rm{H}}}}_{2}{{{\rm{O}}}}$$
$$3{{{{\rm{CaO}}}}}{{\cdot }}2{{{{\rm{SiO}}}}}_{2}{{\cdot }}3{{{\rm{H}}}}_{2}{{{\rm{O}}}}+3{{{\rm{CO}}}}_{2}\to 3{{{{\rm{CaCO}}}}}_{3}+2{{{{\rm{SiO}}}}}_{2}{{\cdot }}3{{{\rm{H}}}}_{2}{{{\rm{O}}}}$$

CaCO3 crystals formed via these two pathways yield different polymorphs: CaCO3 formed from Ca(OH)2 tends to be calcite, whereas CaCO3 formed from C–S–H tends to be aragonite38,44. Available carbonation studies either take for granted that CaCO3 precipitations are calcite or do not discuss the polymorph type. However, an analysis of these studies20 suggests that CaCO3 precipitations consist of a mixture of both calcite and aragonite even in these cases.

The changes in the polymorph type and crystallization degree can be attributed to temperature variations because these variables are strongly sensitive to temperature40. Following the formation of a calcite–aragonite mixture through pathways (3) and (4)–(6), a transition of aragonite to calcite characterizes any cement-based material exposed to CO2 (e.g., naturally present in the air or artificially injected). In the proposed carbonation approach, the injection of CO2 into a suspension involves intense exothermic reactions that enhance the temperature rise that would naturally occur due to cement hydration (Fig. 3). This enhanced temperature rise has two main effects. First, it accelerates the transition of aragonite into calcite, with a larger proportion of aragonite turned into calcite for a larger amount of injected CO2 (at a given flow rate) because of more significant temperature variations (Fig. 3). Second, it increases the crystallization degree of precipitated CaCO3 crystals. Both outcomes can be inferred from the shift of the decarbonation peak towards higher temperatures for a larger amount of injected CO2 (Fig. 2b).

CaCO3 precipitations in freshly cast concrete are generally recognized to enhance the early-stage strength of such material through so-called accelerated curing21,45. This result partly derives from the formation of dense calcium carbonate layers in some of the material pores46 and partly from the significant degree of cement reaction, either through carbonation or hydration, which is much higher than that achieved through typical steam curing regimes. The extra nucleation seeds provided by CaCO3 precipitations are the main factor of such acceleration21. Meanwhile, having more cement constituents reacted, whether through carbonation or hydration, is also beneficial for early-stage strength development47. This evidence has been observed via fresh concrete carbonation20,32,48 and hardened concrete carbonation44,49,50.

In contrast, CaCO3 precipitations and/or crystal transitions at later stages are typically detrimental to the strength of concrete. This evidence results from the fact that aragonite typically has a higher density (2.94 g/cm3) than calcite (2.71 g/cm3)51, which implies that polymorph transitions inevitably induce changes in the volume and shape of these crystals in the concrete structure, leading to deleterious cracking44,52,53. For this reason, carbonation approaches are considered successful when they do not affect the strength of concrete due to the addition of CaCO3 crystals. Accordingly, typical carbonation approaches show, at best, unchanged values of strength, especially after 28 days23.

By injecting CO2 into a cement suspension during mixing, the carbonation approach proposed in this work provides three key opportunities:

  1. 1.

    It employs less energy compared to that required to carbonate fresh and hardened concrete, as CO2 gas is injected into a liquid that has a viscosity comparable to water (i.e., significantly inferior to fresh concrete).

  2. 2.

    It allows harnessing the large surface area of cement particles floating in the water and engaging only a portion of fresh cement with reactions with CO2, as cement is added to a suspension after its carbonation to achieve a paste for use in concrete. While achieving CaCO3 precipitations around some cement particles, this approach avoids the enclosure of most cement particles. This ability is favorable because cement particles encapsulated by CaCO3 crystals cannot yield the formation of C–S–H and involve an aggregation of CaCO3-coated particles into clusters with interconnected and unwanted pores, ultimately leading to deteriorations in the strength of concrete20.

  3. 3.

    It fosters temperature variations during the CO2 treatment of a cement suspension, resulting in the precipitation of more stable and crystalline CaCO3 for larger temperature variations associated with larger amounts of injected CO2 (at a given flow rate). This ability is favorable because stable CaCO3 crystals remain unaffected by polymorph transitions during the subsequent concrete preparation with additional cement and aggregates, curing, and lifetime.

The obtained results specifically allow us to postulate that the variable strength achieved for variable amounts of injected CO2 is due to a variable conversion of CO2 into more stable and crystalline forms of CaCO3. In particular, when 2000 and 4000 scc of CO2 are injected in a cement suspension, the proportion of CO2 that is turned into stable and crystalline forms of CaCO3 is limited and subsequently results in unwanted polymorph transitions and volume changes. In contrast, when 8000 scc of CO2 is injected in a cement suspension, the proportion of CO2 that is turned into stable and crystalline forms of CaCO3 is more significant and subsequently involves minimal polymorph transitions. As a result, the strength of concrete remains uncompromised through the presence of durable and strong calcite deriving from the conversion of CO2 into solid crystals (Fig. 1).

In order to substantiate the proposed mechanism behind the performance of this carbonation approach, experiments involving the injection of the same amount of CO2 (i.e. 8000 scc) with different flow rates over different timeframes (i.e., the reference flow rate of 200 sccm over 40 min and the two additional flow rates of 100 and 400 sccm during 80 and 10 min, respectively) were designed and conducted (see the “Methods” section). The results are shown in Fig. 4. The injection of higher CO2 flow rates in cement suspensions results in faster temperature increases in such suspensions per unit time (Fig. 4a). Maximum temperatures of 16.5, 20.4, and 22.2 °C are measured for flow rates of 100, 200, and 400 sccm after 70, 52, and 45 min, respectively. However, at the time the same volume of 8000 scc of CO2 is injected in cement suspensions, temperatures of 16.2, 19.1, and 13.8 °C are recorded for 100, 200, and 400 sccm, respectively. Therefore, the highest temperature is achieved for the reference flow rate of 200 sccm. All carbonated specimens with a volume of 8000 scc of CO2 exhibit a higher strength compared to the uncarbonated concrete, regardless of the applied flow rates or the curing time (Fig. 4b). This evidence supports that there is a threshold value of injected CO2 and interconnected amount of CaCO3 crystals in concrete that overcomes the negative influence of possible polymorph transitions and yields a strength increase as opposed to a strength decrease compared to uncarbonated concrete. Meanwhile, the greatest strength characterizes the concrete deriving from a cement suspension carbonated with a flow rate of 200 sccm compared to 100 or 400 sccm at any curing time, supporting that there is a threshold temperature value for which CaCO3 crystals mostly transition to more stable calcite instead of remaining aragonite compounds, or for which an increase in the crystallization degree is triggered. Accordingly, the concrete deriving from a cement suspension carbonated with a flow rate of 200 sccm is the only one that presents both CaCO3 needles and flakes on its constituting cement particles, whereas the concrete deriving from a cement suspension carbonated with a flow rate of 100 and 400 sccm only presents CaCO3 needles (Fig. 4c–h).

Fig. 4: Effects of different CO2 flow rates.
figure 4

a Trends of temperature variations in the carbonated cement suspension for different CO2 flow rates (n = 3 independent experiments; raw data reported in Supplementary Table 8; error bars show standard deviations). b Compressive strength of concrete deriving from a paste placed in contact with different CO2 flow rates relative to the reference uncarbonated concrete (reference values are dashed; n = 5 independent experiments; raw data reported in Supplementary Table 9; error bars show standard deviations); and SEM images of the CaCO3 crystals formed on the cement particles deriving from a suspension subjected to c 100 sccm, d 200 sccm, and e 400 sccm, or a paste resulting from a suspension subjected to f 100 sccm, g 200 sccm, h 400 sccm.

Figure 5 summarizes the key stages involved in the strength development of concrete resulting from the proposed carbonation approach. The proposed approach allows to engineer the sequence of hydration and carbonation reactions and tailor them to specific quantities of cement.

Fig. 5: Schematic of the key steps involved in the proposed approach to carbonate fresh concrete.
figure 5

a Injection of CO2 into a cement suspension. b Growth of CaCO3 crystals in such a suspension (inlets b1and b2 show the growth and transition of CaCO3 crystals in the cement suspension). c Addition of cement and aggregates to suspension (inlet c1 shows the transition of CaCO3 crystals in the cement paste). d Dominant hydration of cement in concrete. e Development of interparticle bonding within concrete. f Optional carbonation of hardened concrete.

The proposed approach starts with the mixing of a small portion of cement powder with water while CO2 is injected in the system (Fig. 5a). Carbonation and hydration reactions commence concurrently, but carbonation is faster than hydration at this stage.

As CO2 injection continues, CaCO3 precipitates in the cement suspension while temperature rises (Fig. 5b). Depending on the local environmental conditions, CaCO3 crystals undergo polymorph transitions and changes in polymerization degree, covering a wider number and surface of C3S/C2S particles as time progresses (Fig. 5b1, b2). The precipitation of CaCO3 in a suspension involves that most polymorph transitions and volume changes of such crystals occur early, limiting the potential for cracking and deleterious effects on the structure, properties, and behavior of concrete during its curing and lifetime.

Following the CO2 injection, the second portion of cement powder and aggregates are introduced (Fig. 5c), previously precipitated CaCO3 crystals can be characterized by morphology changes, and new CaCO3 crystals can be formed because of the contact with atmospheric CO2 (Fig. 5c1). At this stage, mixing is performed as in any classical approach to manufacturing concrete. Thereafter, the CaCO3 crystals can act as nucleation seeds for subsequent hydration of C3S and C2S, accelerating the hydration process (Fig. 5d) and contributing to the enhanced early-age strength of concrete.

As the curing process continues, the degree of hydration increases and C–S–H fills the concrete pores as in typical concrete curing (Fig. 5e). This process is crucial to achieving interparticle bonding within the concrete matrix, ensuring strength development.

Finally, the concrete can optionally undergo an additional step of carbonation in its hardened state (Fig. 5f), as in classical carbonation approaches of hardened concrete. In this context, carbonation can still characterize some quantity of cement and form a carbonated layer on the external surface of the concrete.

Carbon capture capacity, kinetics, and efficiency

Carbonation approaches should not only be evaluated in terms of their impacts on the structure and properties of carbonated materials, but also in terms of their ability to effectively store CO2 in durable, solid crystal forms as a result of their CO2 capture capacity, kinetics, and efficiency. The carbon capture capacity reflects the theoretical number of alkaline compounds that can react with CO2 and form stable carbonates, representing the theoretical CO2 sequestration potential of any carbonation approach. The CO2 reaction kinetics determines the amount of CO2 that can be captured for a given CO2 injection duration. The CO2 capture efficiency results from variable CO2 sequestration capacities and reaction kinetics.

The main reactive alkaline compound in cementitious materials is Ca(OH)2, which comes from the hydration of C3S and C2S (Eqs. (1) and (2)). This compound typically gives such materials a pH value between 12 and 13. In some instances, the pH can nonetheless exceed a value of 13 due to the presence of KOH or NaOH54. In this work, the pH of a well-stabilized cement suspension is 12.60 ± 0.05 (Fig. 3a). This value closely aligns with the value of 12.33 that characterizes the pH of a saturated Ca(OH)2 solution:

$${{\rm {Ca}}}{\left({{\rm {OH}}}\right)}_{2}\rightleftharpoons \,{\rm {C{a}}}^{2+}+2{\rm {O{H}}}^{-}{{K}}_{{{{{{\rm{sp}}}}}}}(298\,{{{{{\rm{K}}}}}})=5.02\times {10}^{-6}$$

This result suggests that Ca(OH)2 is the primary source of OH- in the considered suspension system. According to Eq. (3), Ca(OH)2 reacts with CO2 and forms insoluble CaCO3. However, according to Eq. (4)–(6), when Ca(OH)2 is formed through Eqs. (1) and (2), C–S–H is concurrently formed in these reactions and can also react with CO2 in conjunction with Ca(OH)2. Therefore, through a combination of Eqs. (1)–(6), the CO2-active component in cement is equivalent to CaO from a stoichiometric standpoint, allowing for the simplification of the previous equations to the following:

$${{\rm {CaO}}}+{\rm {C{O}}}_{2}\to {{\rm {CaC}{O}}}_{3}$$

Thus, the carbon capture capacity can be calculated based on the CaO content in the materials.

To determine a more universal carbon capture capacity value for the proposed approach, a typical Type I cement composition is used as reference (Table 2). CaO constitutes 61.23% of the cement, resulting in ~30.6 g of CaO in the experimental setup employing 50 g of cement. In principle, according to Eq. (8), ~12,200 scc (24 g) of CO2 can be captured and precipitated as CaCO3 in the considered cement suspension if every molecule of CaO were to react with CO2. In practice, 2992 scc of CO2 are captured when 8000 scc of CO2 are injected into the cement suspension (Fig. 3). Therefore, the achieved CO2 sequestration capacity is markedly smaller than the theoretical sequestration capacity of the carbonated cement suspension due to two reasons: (1) as CO2 is injected, the surfaces of cement particles progressively become partly or entirely covered by CaCO3 crystals (Fig. 1), and hence less reactive to CO2; (2) as CO2 is injected for a limited time, not all the CO2 can react with the available cement particles.

The obtained results also indicate a decrease in the CO2 reaction rate, which is attributed to a decrease in the OH- concentration in the material and a corresponding decrease in pH. An analysis of the reaction kinetics corroborates this evidence. Since a series of chemical and physical reactions develop during the proposed carbonation process, it is essential to know which one is the rate-control step. CaCO3 precipitation relies on the concentrations of the two reactants, Ca2+ and CO32−, as indicated by the following equation:

$${\rm {C{a}}}^{2+}+{\rm {C{O}}}_{3}^{2-}\to {{\rm {CaC}{O}}}_{3}\downarrow$$

The presence and concentration Ca2+ and CO32− in the aqueous suspension rely on their reaction routes.

On the one hand, the presence of CO32− comes from CO2 dissolution in an aqueous solution, which follows a series of chemical equilibria. First, CO2 transitions from a gaseous to an aqueous environment:

$${\rm {C{O}}}_{2}({\rm {g}})\rightleftharpoons \,{\rm {C{O}}}_{2}({{\rm {aq}}})$$

As CO2 comes in contact with H2O, the formation and ionization of carbonic acid take place, which supply H+ and CO32− according to the following:

$${\rm {C{O}}}_{2}\left({{\rm {aq}}}\right)+\,{{\rm {H}_{2}O}}\rightleftharpoons \,{{\rm {H}_{2}C{O}}}_{3}$$
$${{\rm {H}_{2}C{O}}}_{3}\rightleftharpoons \,{{\rm {H}}}^{+}+{{\rm {HC}{O}}}_{3}^{-}$$
$${{\rm {HC}{O}}}_{3}^{-}\rightleftharpoons \,{{\rm {H}}}^{+}+{\rm {C{O}}}_{3}^{2-}$$

As CO2 injection occurs, the concentration of CO2 (g) can be considered infinite (Eq. (11)). The consumption rate of H+ plays a crucial role in shifting the chemical equilibrium to the right and increasing the concentration of CO32− (Eqs. (11)–(13)). The following acid-base neutralization reaction is very fast:

$${{\rm {H}}}^{+}+{\rm {O{H}}}^{-}\to \,{{\rm {H}_{2}O}}$$

Therefore, the dissolution of Ca(OH)2 is the rate-determining step, according to the following:

$$\,{{\rm {Ca}}}{\left({{\rm {OH}}}\right)}_{2}\rightleftharpoons \,{\rm {C{a}}}^{2+}+2{\rm {O{H}}}^{-}$$

As Eq. (15) is the main reaction that generates OH in the system, the pH turns out to be the indicator of the overall reaction rate. During CO2 injection, the pH decreases from 12.60 to 12.18, leading to a decrease in OH concentration from 0.040 to 0.016 M and, hence, a concentration change of 0.024 M (Fig. 3). This OH concentration change is larger than half of the initial OH concentration. This decrease in OH concentration affects the rate of CO2 dissolution, which provides CO32−. Notably, the temperature increase during the CO2 injection also impacts the Ca(OH)2 solubility (for Ca(OH)2, Ksp (22 °C) ≈ 7.95 × 10−5; Ksp (40 °C) ≈ 5.4 × 10−5). The OH concentration under saturated conditions decreases from 0.043 to 0.038 M under such temperature increase and involves a concentration change of 0.005 M. However, the observed OH concentration decrease is significantly larger than the OH concentration drop caused by temperature. Therefore, the primary limitation to OH concentration maintenance still comes from Ca(OH)2 dissolution.

On the other hand, Ca2+ comes directly from dissolution of Ca(OH)2. The reactions involved are as follows:

$${{\rm {CaO}}}+{{\rm {H}_{2}O}}\,\to {{\rm {Ca}}}{\left({{\rm {OH}}}\right)}_{2}$$
$${{\rm {Ca}}}{\left({{\rm {OH}}}\right)}_{2}\,\rightleftharpoons \,{\rm {C{a}}}^{2+}+2{\rm {O{H}}}^{-}\,({{\rm {surface}}})$$
$${\rm {C{a}}}^{2+}+2{\rm {O{H}}}^{-}\,({{\rm {surface}}})\,\rightleftharpoons \,{\rm {C{a}}}^{2+}+2{\rm {O{H}}}^{-}\,({{\rm {bulk}}})$$

Equation (16) indicates the formation of Ca(OH)2. Equation (17) indicates the dissolution of Ca(OH)2. Equation (18) indicates the diffusion of ions from the particle surface to the bulk solution. Under unagitated conditions, the rate-controlling step is the diffusion step expressed in Eq. (18)55. However, in a vigorously agitated system like the one that characterizes the proposed approach to inject CO2 into a cement suspension while it is mixed, diffusion is accelerated, and the rate-controlling step is the dissolution of Ca(OH)2 at the surface instead56, as expressed in Eq. (17). Therefore, the decreased CO2 capture rate is caused by a decrease in the OH- concentration in the material and a corresponding decrease in pH.

Most fresh concrete carbonation approaches have limited CO2 capture efficiency of up to 10%20. The proposed carbonation approach achieves a CO2 capture efficiency that surpasses 45% when injecting 2000 scc of CO2 into the system and remains high at 26.2% when injecting a volume of 8000 scc CO2 (Fig. 3). In this context, the following aspects are noteworthy: (1) the developed experiments are performed by using an uncapped beaker container, which does not prevent the dispersion of CO2 into the atmosphere; (2) the experiments do not incorporate any looping system, which could further allow to enhance the treatment efficacy by lengthening the retention time of CO2 in the suspension; (3) any upscaled application of such an approach would lengthen the retention time, enhancing CO2 capture efficiency. Therefore, significant margins for improvement characterize upscaled implementations of the proposed carbonation approach, which is already more advantageous than state-of-the-art approaches. Emerging carbonation approaches involving the carbonation of wastewater are also being proposed for applications in concrete manufacturing and may be considered as a valuable pathway57,58. Notably, the approach proposed in this work could also carbonate appropriately treated wastewater instead of fresh water for the purpose of concrete manufacturing.

Carbon footprint

Carbon footprint is another key metric that can help determine the competitiveness of any CO2 capture technology. Based on a Cradle-to-Gate Life Cycle Assessment (LCA), classical cement manufacturing has a global warming potential of 519 kg CO2-eq per functional unit, with 311.4 kg CO2-eq from fuel burning and 207.6 kg CO2-eq from limestone calcination. By assuming to carbonate the same functional unit of cement with the proposed approach for subsequent use in concrete, the global warming potential is reduced to 500.3 kg CO2-eq. In other words, the production of one ton of cement leads to a reduction of 18.7 kg CO2-eq in global warming potential. When applied to the entire global cement production, this amounts to a substantial reduction of 82.28 million tons of CO2 emissions59,60. Put into perspective, this reduction is approximately equivalent to the annual CO2 emissions of 18 million automobiles resorting to combustion61.

Promisingly, the developed LCA calculations unfavorably consider various aspects in the calculations, which would likely result in even more positive outcomes in the carbon footprint of industrial applications of the proposed technique. First, the carbon footprint calculations use a CO2 capture efficiency of 26.2%. However, the actual CO2 capture efficiency in upscaled applications of the proposed approach can be significantly higher due to the longer retention time of CO2 in the mixer CO2. Second, CO2 injection is considered a one-round procedure. However, the CO2 injection in an upscaled process can be recursively applied or implemented with a looping system, increasing the CO2 capture efficiency. Third, the proportion of carbonated cement for a functional unit is set at 1/4, which aligns with the laboratory conditions considered in this work. However, this proportion can be increased in an upscaled process, as long as the strength of the carbonated concrete is not compromised.

In summary, the innovative suspension-based concrete carbonation approach developed here not only demonstrates a reduced footprint but also exhibits potential for further scalability, rendering it an attractive candidate for industrial applications. This approach effectively addresses two prominent concerns within the domain of carbon capture through concrete carbonation. The first concern pertains to the potential weakening of concrete strength as a result of carbonation, necessitating additional cement usage to compensate for strength loss. This study reveals that by facilitating carbonate crystal depositions in the cement suspension phase, the proposed approach can, in fact, maintain an uncompromised strength. The second concern relates to the energy consumption associated with the carbonation treatment process, which encompasses CO2 purification, pumping, transportation, and the use of high-pressure vessels. The cumulative effects of these processes could render the concrete carbonation approach carbon-positive5. However, the approach developed in this study minimizes energy consumption during CO2 introduction through simplified and less rigorous procedures.


This work introduced an approach to trap CO2 into fresh concrete by carbonating a portion of its cement in suspension. Laboratory experiments and LCA calculations support that the developed approach: (1) involves an active absorption of CO2 during concrete manufacturing as opposed to conventional concrete production procedures; (2) benefits from a CO2 capture efficiency of up to 45%; and (3) yields carbonated concrete with an uncompromised strength. The study provides a mechanistic analysis of the physico-chemical processes that characterize the proposed carbonation approach, which not only may underpin future enhancements of the proposed method but also inform the development of other approaches to carbonate fresh concrete and cementitious materials. By modifying classical concrete manufacturing processes only minimally, the proposed carbonation approach offers promise for industrialization. The possibility to enhance the treatment efficacy of the proposed approach in industrial scenarios, as described in this work, further supports its potential to decarbonize the cement and concrete industries.



The promise of the carbonation approach proposed in this work is assessed through four central sets of experiments:

  • Set 1: The first set of experiments applies CO2 injections to cement suspensions.

  • Set 2: The second set of experiments imposes the same CO2 injections on virtually equivalent cement suspensions, which are subsequently mixed with additional cement to achieve a cement paste of the desired water-to-cement ratio.

  • Set 3: The third set of experiments imposes the same CO2 injections to virtually equivalent cement suspensions, which are subsequently mixed with additional cement and aggregates to achieve a concrete mix design.

  • Set 4: The fourth set of experiments imposes different CO2 injection flow rates but reaches the same volume of injected CO2 via different injection durations in cement suspensions.

Set 1 allows determining the chemical and physical properties of carbonated cement suspensions. Set 2 allows assessing the chemical and physical properties of cement pastes deriving from carbonated cement suspensions. Set 3 allows assessing the strength of concrete deriving from carbonated cement suspensions subsequently mixed with additional cement and aggregates. Set 4 allows to expand on the mechanistic understanding of the physico-chemical processes that characterize the proposed carbonation approach with specific reference to the role of temperature and amount of injected CO2 on the results of the carbonation. Experiments without any CO2 injection are also performed for reference.

Materials preparation

The materials addressed by experimental sets 1–3 are prepared by using a 150 mL beaker and following the steps detailed hereafter (Fig. 6). Initially, 100 mL of deionized water (DI-H2O) with a resistivity of 18.2 MΩ cm is introduced into the beaker. Next, CO2 gas of high purity (99.9%) is introduced into the DI-H2O at a flow rate of 200 standard cubic centimeters per minute (sccm) for a duration of 20 min while stirring is continuously imposed. This process ensures the establishment of a CO2-saturated condition and maintains a consistent initial condition for each test. Meanwhile, experiments support that the absence of this step does not lead to different outcomes associated with the carbonation process of the cement suspensions. Subsequently, 50 g of Type 1 Portland cement (CEM I 42.5R) obtained from Vigier Cement Inc. is added to the CO2-saturated DI-H2O. Afterwards, CO2 gas is injected into the prepared cement suspension for 10, 20, and 40 min. CO2 injection flow rate is controlled by a mass flow controller (Alicat MC-300SCCM-D). The carbonated cement suspension achieved at this stage is the material addressed by experimental set 1, which, after preparation is moved to a Petri dish, dried in a CO2-free desiccator under 0% relative humidity and room temperature (~22 °C) for 24 h, and pulverization in an agate mortar for subsequent characterization.

Fig. 6: Schematic of the suspension-based fresh concrete carbonation approach from steps I–VI.
figure 6

(Step I) Initial condition; (Step II) CO2 saturation; (Step III) Creation of cement suspension; (Step IV) Carbonation of cement suspension; (Step V) Creation of cement paste; (Step VI) Creation of concrete.

After the achievement of a carbonated cement suspension, experimental set 2 involves the addition of another 50 g of cement to achieve a fresh cement paste that is stirred for another 10 min. The cement paste achieved at this stage is the material addressed by experimental set 2, which is subjected to characterization following the same procedure employed for experimental set 1.

After the achievement of a cement paste deriving from a carbonated cement suspension, experimental set 3 involves the addition and mixing of cement and natural rounded crystal quartz sands and gravel sourced from Carlo Bernasconi AG Inc. to achieve a concrete mix design of a prescribed mass of constituents (Table 1). The resulting concrete achieved at this stage is the material addressed by experimental set 3, which is promptly molded into cylindrical specimens measuring 33 mm in diameter and 67 mm in height. A general-purpose lubricant (WD-40) is applied to the inner walls of the molds to facilitate the demolding process. The specimens are demolded following casting and a curing period of 24 h in a desiccator with a relative humidity of 100% and no CO2. Subsequently, they are subjected to curing in a dedicated room with a relative humidity of 100% for various durations: 3, 7, 28, and 56 days. Finally, the resulting concrete specimens are subjected to testing. Each experimental condition is analyzed based on the results of ten replicates.

Table 1 Concrete mix design per m3

Two sets of considerations are noteworthy with respect to the approach proposed here. On the one hand, although concrete preparation standards for laboratory use, such as ASTM C19262, do not specify a particular type of water, it is indeed recognized that tap water is more commonly utilized for concrete manufacturing as opposed to the deionized water employed for this work. Nevertheless, deionized water was chosen for this study to minimize variations in the properties (e.g., mineral content, pH, CO2 saturation level) of the materials to be carbonated and hence achieve optimal control of the experimental conditions for the sake of this scientific investigation. In alignment with industrial approaches, the life cycle assessment conducted for this study indeed considered tap water, rather than deionized or distilled water, to provide representative results of real-world conditions.

On the other hand, although concrete mix designs with variable (e.g., lower) quantities of cement may be used in practice compared to those used for this work, such an aspect does not compromise the observed performance of the proposed approach to carbonate fresh concrete (see the “Results” section). In fact, only a minimal fraction of the total amount of cement that characterizes the concrete mix design chosen for this work is initially carbonated through the proposed approach. As this amount of cement is limited and lower than that considered in most concrete mix designs employed in practice, and its carbonation yields a carbon capture efficiency that remains superior compared to state-of-the-art approaches (even if not 100% efficient, as shown in the “Results” section and highlighted in the “Discussion” section), it is possible to consider the proposed approach, and especially its overarching methodological principle, as valuable alternatives compared to state-of-the-art methods.

Materials characterization

The characterization of the cement suspensions and cement pastes resorts to pH and temperature measurements, scanning electron microscopy (SEM), X-ray diffraction analyses (XRD), and thermogravimetric analyses (TGA). The characterization of the concrete resorts to unconfined compressive strength tests.

pH and temperature measurements resort to an Elite pocket tester by Thermo Scientific. They allow us to measure the considered variables during the preparation and possible carbonation of cement suspensions and pastes. The reported pH values during carbonation refer either to immediate measurements after a certain duration of CO2 injection (“instant pH “ in Fig. 3a) or to measurements collected for stabilized solutions after stirring them for 10 min to let excessive hydroxide donors (if present) to dissolve (“equalized pH” in Fig. 3a).

SEM analyses use JEOL JSM-7900FLV equipment facilitated with energy-dispersive X-ray spectroscopy (EDS). They allow us to characterize the morphology of dried particles of cement and CaCO3 crystals.

XRD analyses use an STOE-STADI-P powder diffractometer equipped with an asymmetrically curved Germanium monochromator (CuKα1 radiation, λ = 1.54056 Å) and 1D silicon strip detector (MYTHEN2 1K from DECTRIS). The line-focused Cu X-ray tube is operated at 40 kV and 40 mA. Intensity data from diffraction angles of 5°–70° (2θ) are collected for 45 min per sample. Phase identification is made by Rietveld refinements using the GSAS-II software for semi-quantitative analysis. Inorganic crystal structure database (ICSD) is used to obtain the standard spectra of crystals. Specifically, ICSD 15194, ICSD 15471, ICSD 18166, ICSD 79550, ICSD 87951, ICSD155395, and ICSD 162744 spectra are used, corresponding to aragonite, portlandite, calcite, belite, C–S–H, ettringite, and alite, respectively.

TGAs employ Netzsch’s simultaneous thermal analysis system and allow to quantify key features of CaCO3 crystal precipitations from weight loss data, including decarbonation peaks, CO2 capture rates, and carbonation degree. Approximately 20 mg of the powdered samples are subjected to TGA. The samples are heated from room temperature to 1000 °C at a heating rate of 10 °C/min under an inert N2 atmosphere. The weight loss occurring between 500 and 900 °C is determined as the total loss of CO2 resulting from the decomposition of carbonates. The CO2 capture efficiencies are calculated as follows:

$${{{{{\rm{{CO}}}}}_{2}}}\,{{{{{{\rm{capture}}}}}}\; {{{{{\rm{efficiency}}}}}}} \, [ \% ]=\frac{{{{{{{\rm{Overall}}}}}}}\,{{{{{\rm{{{CO}}}}}}}}_{2}\,{{{{{{\rm{loss}}}}}}}-{{{{{{{\rm{CO}}}}}}}}_{2}{{{{{{\rm{loss}}}}}}\; {{{{{\rm{in}}}}}}\; {{{{{\rm{untreated}}}}}}\; {{{{{\rm{cement}}}}}}}}{{{{{{{\rm{Overall}}}}}}}\,{{{{{{{\rm{CO}}}}}}}}_{2}\,{{{{{{\rm{injection}}}}}}}}\times 100$$

The carbonation degree is calculated as follows:

$${{\rm {Carbonation}}\; {\rm {degree}}} \, [ \% ]=\frac{{{\rm {Overall}}}\,{{{\rm {CO}}}}_{2}{{\rm {loss}}}-{{{\rm {CO}}}}_{2}{{\rm {loss}}\; {\rm {in}}\; {\rm {untreated}}\; {\rm {cement}}}}{{{\rm {Theoretical}}}\,{\rm {C{O}}}_{2}\,{{\rm {uptake}}\; {\rm {capability}}}}\times 100$$

The theoretical CO2 sequestration capacity is calculated via the composition of cement outlined in Table 2. A typical composition of type I cement is utilized to calculate a based on the consideration of calculating a universal value of such an approach to provide a concept of capacity level. While Na2CO3 and K2CO3 both provide alkalinity, they do not have a significant effect on the CO2 capture capacity calculation in this study because they are both highly soluble in water and CO32− content in the suspension is thus not affected. The alkali content does influence the reaction kinetics, but the composition in Table 2 is solely used for capacity calculation.

Table 2 Representative composition of Type 1 cement

Unconfined compressive strength tests are performed with a universal testing machine equipped with circular platens and a closed-loop hydraulic system manufactured by MTS Inc. These tests impose a displacement rate of 0.18 mm/min according to the ASTM C3963. The concrete specimens are initially capped using a sulfur-based capping compound (H-2959) obtained from Humboldt Mfg. Co. Capping is applied to achieve smooth bearing surfaces on both ends of the specimens and limit possible stress concentrations that would affect the results of the tests, thereby targeting the achievement of uniform loading conditions during the compression tests.

Carbon footprint analysis

The carbon footprint analysis developed as a part of this study is conducted based on an Environmental Product Declaration provided by CEMEX Polska Sp64. The Environmental Product Declaration adheres to standard EN 1580465 and is verified by ISO 1402566. The primary analysis is conducted using a “Cradle-to-Gate” Life Cycle Assessment (LCA). The functional unit considered in the LCA is 1 ton of cement. The LCA process system boundary is shown in Fig. 7. By following the same “Cradle-to-Gate” scenario as in the existing LCA, the carbon footprint of the developed suspension-based carbonation approach is evaluated within the scope of cement production, considering the additional CO2 injection process required by such an approach. In this manner, the net reduction of global warming potential characterizing a standard cement production serving subsequent use for concrete production can be assessed.

Fig. 7: Schematic of the system boundary of the process.
figure 7

Gray box: Cement production involved in the LCA calculation; Out of the gray box: the subsequent route of cement.

The following assumptions are made to perform the calculations:

  • 40% of the total CO2 emissions deriving from a cement plant is attributed to fuel combustion, whereas 60% is attributed to limestone calcination67.

  • Flue gas from a cement clinker calcination plant is used as the CO2 source, with a CO2 content of 21.76% (dry basis). CO2 is regarded as the sole active constituent in the gas.

  • 25% of the total employed cement in a given volume of concrete is considered to be carbonated with CO2 in suspension (i.e., the same ratio applied in the laboratory test), whereas the remaining 75% is subsequently employed after treatment for concrete manufacturing.

  • Tap water instead of deionized water is used in the calculation for consistency with the classical concrete preparation scenario. Water consumption serving concrete mixing step is excluded because the same amount of water is consumed in a classical concrete manufacturing process.

  • Commercial Air Pumps (EcoPlus Eco Air 7, each with a power of 200 W) are used for CO2 injection into the suspension, and their energy consumption is duly considered for the developed carbonation approach.

  • Each kWh of electricity consumption emits 0.39 kg CO2-eq, in alignment with the average value of the US electricity grid68.

  • Irrespective of whether concrete derives from a functional unit of cement carbonated with the proposed approach or not, its following use in practice is assumed to be the same. Therefore, the CO2 emissions associated with concrete use and recycling are excluded.

  • A CO2 injection volume of 8000 scc per 50 g cement is considered as a reference for the carbonation approach. The corresponding CO2 capture rate of 26.2% (i.e., the lower capture rate achieved in the laboratory experiments for a CO2 injection lasting 40 min) is also considered.

  • The continuous mixing process of the suspension-based fresh concrete carbonation approach is considered a part of the mixing process with subsequent concrete mixing with additional cement and aggregates; thus, it is excluded from the calculation.