Influence of Various Soluble Carbonates on the Hydration of Portland Cement studied by X-ray Diffraction

The effect of limestone on the hydration of Portland cement has been studied by many researchers. However, a possible influence of adding more soluble carbonates was not explained. Therefore we executed a qualitative and quantitative research on the influence of slightly soluble (CaCO3, MgCO3, dolomite), medium soluble (Li2CO3) and highly soluble (K2CO3 and KHCO3) carbonates on the hydration. Blending of Portland cement with differently soluble carbonates was found to influence the hydrate assemblage of the hydrated cement. With the help of the Rietveld analysis, the study indicated that the amount of reacted carbonate in cement hydration at a 15% addition of slightly or medium soluble carbonates does not exceed 5% and is not affected by their solubility; at a 15% addition of the highly soluble carbonate K2CO3 the amount of reacted carbonate was around 6%. An increase in temperature (25 to 40 °C) gradually affects the rate of hydration and the quantity of stable phase assemblage.


Introduction
Over the last 20 years, the effects of using limestone in Portland cement (PC) have been well studied. The benefits of limestone as a partial replacement for PC are well established. Its economic and environmental advantages of reducing CO 2 emissions are well known.
However, throughout the years, the general provisions which determine the amount of limestone permitted in cement, have remained unchanged. Most Portland cement specifications allow the use of limestone up to 5%. Beyond that, Portland limestone cements (PLC) are categorized on the basis of the percentage of limestone (6-35%) added to the cement. The European Standard (EN 197-1-2000) permits up to 5% limestone regarding it as a minor additional constituent. It also identifies four types of PLC containing 6-20% limestone (types II/A-L and II/A-LL) and 21-35% limestone (types II/B-L and II/B-LL), respectively. 1 For a long time, the limestone has been considered as an inert filler. Recently it has been concluded that lime-stone serves both as an inert filler and also reacts to a limited extend. The reactivity depends on its fineness (specific surface) 2 and content. 3,4 Matschei et al found that at a low concentration, limestone (a molar ratio of CO 2 /Al 2 O 3 ∼ 0.66 with a fixed sulfate ratio SO 3 /Al 2 O 3 = 1 and an excess of portlandite, T = 25 °C) reacts completely to various forms of carboaluminate phases. The extension of limestone's reactivity is controlled by the amount of sulphate in the system. As the sulphate content increases, the likelihood of unreacted calcite increases.
It is agreed that limestone reacts primarily with the tricalcium aluminate (C 3 A) to form carboaluminates (hemicarboaluminate, C 4 AC -0.5 H 12 and monocarboaluminate, C 4 AC -H 11 are the most common) at the expense of hydrates. 5,6 On the other hand, the formation of ettringite (C 6 A 3 S -H 32 ) in PLC, the hydration reaction of C 3 A in the presence of gypsum, is still questionable. Some researchers found its formation delayed, while others found that ettringite (AFt) formation proceeded normally. 5

1. Materials
Ordinary Portland cement, CEM I 52.2 R (Salonit Anhovo, Building Materials, Joint-Stock Co., Anhovo, Slovenia) was used. The cement consisted of clinker and chemically precipitated gypsum, both ground in the laboratory mill. Chemical and mineralogical compositions of cement are shown in Table 1. The quantitative phase analyses of the phases with known structures were performed using the Rietveld method by using Topas Ver. 2.1 program (Bruker AXS, Karlsruhe, Germany). The method of internal standard addition was used to determine the absolute content of identified crystalline phases of the hydrated samples. For this purpose, 40% of α-Al 2 O 3 , MicroPolish II, 1-μm deagglomerated alumina (Buehler Ltd., Lake Bluff, IL) was added to hydrated samples and thoroughly mixed in an agate mortar.
After the specified hydration time, samples were powdered in an agate mortar and analyzed. The XRD phase analyses of the data considered three replications of each treatment, adding up to 384 samples.

1. Portland Cement (PC)
The following crystalline or partly crystalline phases were identified in the clear Portland cement and were included into the model for Rietveld refinement: CH (ICSD code 34241), Aft (ICSD code 16045), Mc (ICSD code 59327), partly crystalline C-S-H (ICSD code 87689), CaCO 3 (CC -) (ICSD code 73446), after 28 days of hydration also in vaterite modificiation (ICSD code 15879), calcium aluminium hydrate C 3 AH 6 (ICSD code 202316), unreacted clinker phases C 3 S (ICSD code 81100), C 2 S (ICSD code 39006), C 3 A (ICSD code 6287), C 4 AF (ICSD code 9197), and substances in traces (below 2%): SiO 2 (S) (ICSD code 67117), gismondine CAS 2 H 4 (ICSD code 15838) and carbosilicate tilleyite C 3 S 2 (CC -) 2 (ICSD code 14256). In Figure 1 a representative Rietveld plot is presented. At a certain temperature (25 or 40 °C) and after the hydration process is finished (3, 7, 28, 90 days), parts of each individual phase are visible -presented in Figure 2. Each value of a given phase content presents an average of 3 measurements taken. There is also a minimum and a maximum value, so one can picture the discrepancy between the three values included in the calculated average.
In the PC samples most of CH (12-14%) is produced. Its amount is decreased after 90 days of the hydration process, which means that the hydration process is finished. This is also confirmed by the amount of crystalline C-S-H, which does not significantly change after 90 days of hydration and the minimal presence (2% and less) of C 3 S and C 2 S.
After 90 days, even AFt is no longer being produced; its amount is app. 3.7% at both temperatures. (Less AFt occurs at 40 °C after any days of hydration.) The occurrence of Ms is obviously hampered in the starting stages of hydration, as the amounts of Mc (app. 4 %) and CaCO 3 (app. 1%) are too high and there is not enough Al 2 O 3 . The amount of C 3 A is minimal after only 3 days of hydration (app. 0.15%), and after 28 days the substance is no longer present. AFt is stabilized in the presence of Mc and CC -.
Hemicarboaluminate (Hc) was not identified even after the 3-day hydration, in spite of the fact that the samples were kept in airtight containers and were not dried with solvents, e.g. acetone or diethyl ether. 6 The reason might be the air carbonization, which can occur in the airtight containers. We presume that all of the Hc converted into the more stable Mc in the 3-day hydration. The in- creased temperature (from 25 to 40 °C) does not influence hydration significantly -it accelerates it a bit. This can be seen in the results of the comparison of the unreacted phases of the cement clinker (C 3 S, C 2 S, C 3 A) after a certain number of hydration days and the increased amount of CH.
The crystallinity (mass fraction of all crystalline phases in a sample) of PC samples is app. 42% after 3 days of hydration and is decreased to app. 31% after 90 days of hydration at both temperatures. The exact values are shown in Table 2   11.2% at 40 °C). The amount of the reacted CaCO 3 is around 4%, while the rest serves as a filling. This value of the reactive carbonate matches with articles published in the past (up to 5%), or even a bit higher than suggested by T. Matschei  The increased temperature (from 25 to 40 °C) does not considerably affect hydration. It is just a bit accelerated, which is indicated by the comparison of the reactivity of the cement clinker phases (C 3 S, C 2 S) and thus the formation of an increased amount of CH and C-S-H. The   It would be difficult to conclude that the addition of carbonate has any significant influence on the formation of AFt (either impeding the formation or accelerating it). 4 For a clearer demonstration and possible comparison all contents of the main hydration products (CH, AFt, Mc and CaCO 3 ) are collected in Table 3 in relation to hydration time (3, 7, 28, 90 days) and temperature (25 and 40 °C). The minimum and maximum determined contents of the main hydration products are collected bellow, in Table 4 and Table 5.
C-S-H can include CO 3 2anions in its structure, and consequently calcium silicocarbonate hydrates can be formed. We have not identified any with an XRD analysis, but we assume their formation. If the hydration of clinker phases in samples with added calcite or CC is accelerated, slightly more SCH should form in these samples than in clear PC samples. Actually we have identified less C-S-H during the hydration in samples with calcite compared to PC samples probably just due to formation of calcium silic-   The crystallinity of samples with added carbonate does not depend on temperature, after 3 days of hydration it is around 44% and after 90 days of hydration it is around 41%. Exact values are collected in Table 2. After 3 days of hydration the crystallinity in samples with added carbonate and samples with PC is comparable, whereas after 90 days of hydration the crystallinity of samples with added carbonate is higher than that of PC samples (around 10%).

2. 2. MgCO 3 (Solubility at 25 °C is 0.18g/100g
Water 20 ) The hydration products in samples with added MgCO 3 are the same as the hydration products in PC sam-ples, with the difference that due to the dissolution of MgCO 3 in the presence of CH, the additional formation of Mg(OH) 2 occurs after 28 days of hydration at 25 °C or after 7 days of hydration at 40°C. (The reaction: MgCO 3 + Ca(OH) 2 → CaCO 3 + Mg(OH) 2 ) After 28 days no vaterite forms at 25°C as well as at 40°C. Hydration products and their contents at 3, 7, 28 and 90 days of hydration and both temperatures are demonstrated in Figure 5. Figure 5 and the above reaction clearly explain that the amount of reacted MgCO 3 increases with temperature (after 90 days of hydration: 25 °C 7.52%, 40 °C 11.71%). Consequently, at a higher temperature Mg(OH) 2 is formed earlier (after 7 days) and in larger amount (90 days hydration: 25 °C 0.58%, 40 °C 1.95%). The same holds for CaCO 3 and more CH is being used (see Table 3). The hydration of the clinker phases (C 3 S, C 2 S, C 3 A) is accelerated. The increased temperature (from 25 to 40 °C) accelerates the hydration of C 2 S after 3 days, whereas the amount of reacted C 3 S is lerger after 90 days of hydration. In sam- The added carbonate which is converted into inert CaCO 3 is considered to be unreactive.
The SiO 2 , gismondine, C 4 AF, C 3 AH 6 and tilleyite portion are small and comparable to samples of clear PC.
The crystallinity of samples with added MgCO 3 does not depend on temperature and after 3 days of hydration it is approximately at 49% and after 90 days of hydration it is around 41%. The exact values are collected in Table 2. The crystallinty in the samples with added MgCO 3 is always higher than in PC samples and after 90 days of hydration the values are around 10% higher than in samples of clear PC.

2. Dolomite
Hydration products in samples with added dolomite are the same as hydration products in PC samples, with the difference, due to the dilution of dolomite in water, that besides crystallizing CaCO 3 on the surface of the dolomite grains MgCO 3 crystallizes as well (after 3 and 7 days of hydration at both temperatures). We predict that a small quantity of MgCO 3 then reacts with CH (as in the samples with MgCO 3 ), although after 28 as well as 90 days of hydration via the Rietveld method we did not identify the formed Mg(OH) 2 , due to insufficient quantities. Moreover, we confirmed that there was no dedolomitisation reaction, no expansion reaction between the dolomite and the base (CH): CaMg(CO 3 ) 2 + Ca(OH) 2 = Mg(OH) 2 + 2CaCO 3 even after 90 days of hydration at an increased temperature, 40 °C. 21 After 28 days vaterite is not formed both at 25 °C and 40 °C. Hydration products and their contents at 3, 7, 28, 90 days of hydration and both temperatures is clearly visible on Figure 6. Figure 6 clearly shows the amount of diluted dolomite, which after 90 days of hydration is around 5% (25 °C 4.74%, 40°C 5.68%). Thus, after 90 days of hydration in samples with added dolomite there was a lot of unreacted carbonate present (unreacted dolomite +CaCO 3 : 25 °C 11.93%; 40°C 11.79%). We anticipate that it acts like a filling and at the same time, together with Mc, stabilizes AFt, the portions of which, in samples with or without the added slightly soluble carbonate, are comparable after 90 days of hydration at both temperatures and are around 3.5% (see Table 3). The amount of reacted carbonate after 90 days of hydration is thus around 1.4% at both temperatures, which is even lower comparable to samples with added MgCO 3 or samples with added CC -/ calcite.
The amounts of formed Mc after 90 days of hydration in all samples with added slightly soluble carbonate are comparable (see Table 3). As with other samples with added slightly soluble carbonate the hydration of the clinker phases is accelerated (C 3 S, C 2 S, C 3 A) and after 90 days the amounts of unreacted di-/tri-silicates are small (C 3 S: 25 °C 0.69%, 40 °C 0.66% and C2S: 25 °C 2.05%, 40 °C 1.24%). The amount of the formed CH is around 12%, after 90 days of hydration, its contents are comparable to those in the samples with added calcium carbonate or CC -(see Table 3), because CH is not being used up like in samples with added MgCO 3 . Furthermore, in samples with added dolomite the formation of partly crystalline C-S-H is unhindered and is after 90 days of hydration at both temperatures quantitatively slightly higher than in samples of clear PC or other slightly soluble carbonates (25 °C 6.74%, 40 °C 7.16%). However we assume partial formation of calcium silicocarbonate hydrates but we could not identify them. The amount of SiO 2 , gismondine, C 4 AF, C 3 AH 6 and tilleyite are small, their contents do not exceed 1%. The increased temperature (from 25 to 40 °C) somewhat accelerates the hydration speed and increases the degree of reactivity of the clinker phases C 3 S and C 2 S. The amount of stable hydration products (CH, Mc, CC -) is changed slightly during the hydration process, the differences being minor (see Table 3). The crystallinty of the samples with added dolomite does not depend on temperature. After 3 days of hydration it is around 50%, after 90 days it is around 43%. The exact values are collected in Table 2. The crystallinity is in the samples with added dolomite always higher than in PC samples and after 90 days of hydration the values exceed the samples with clear PC by 12%.

Medium Soluble Carbonate
Li 2 CO 3 (solubility at 25 °C is 1.30g/100g water 20 ) A 15% addition of the moderately soluble carbonate Li 2 CO 3 affects the cement hydration. With the applied Rietveld method the following phases were identified: CH, AFt (3 and 7 days hydration at 25 °C) Mc, partly crystalline C-S-H, CaCO 3 , Li 2 CO 3 , calcium aluminum hydrate C 3 AH 6 , unreacted clinker phases C 3 S, C 2 S, C 4 AF and compounds in traces (under 2%): stratlingite C 2 ASH 8 , lithium sulfate hydrate Li 2 SO 4 (H 2 O), SiO 2 , gismondine CAS 2 H 4 and carbosilicate tilleyite 3CaO·SiO 2 ·2CaCO 3 . The hydration of the clinker phases of C 3 A and C 2 S 22 is accelerated, whilst the hydration of C 3 S at both temperatures is decelerated (see Figure 7). What remains after 90 days of hydration in samples with added Li 2 CO 3 is 2.44% of C 3 S at 25 °C and 1.78% of C 3 S at 40 °C. (C 3 S in PC after 90 days: 25 °C 1.33%; 40 °C 1.28%.) Consequently, the amount of formed C-S-H is increased and the differences between samples with and without the added Li 2 CO 3 increase with the level of hydration. After 90 days of hydration in samples with Li 2 CO 3 at 25 °C there is 8.58% and at 40 °C there is 8.22% C-S-H. (PC 25 °C 5.19%; 40 °C 5.99%.) After 90 days of hydration around 4% Li 2 CO 3 is reacted (25 °C 4.07%, 40 °C 3.8%). Lithium with sulfate ions forms the lithium sulfate hydrate Li 2 SO 4 (H 2 O). At both temperatures the contents during the hydration remain low, under 1%, but apparently high enough to not allow the formation of AFt or Ms, probably due to the lack of availability of sulfate ions. AFt only occurs in small amounts (under 1%) at 25 °C during the beginning stages of hydration (after 3 and 7 days), whereas at 40 °C it does not even occur. The amount of the formed Mc increased, the exact values are collected in Table 3. A part of the aluminum is used for the formation of stratlingite C 2 ASH 8 (the contents during the 90 days of hydration at both temperatures do not exceed 1%). We propose that a part of Li 2 CO 3 dilutes in CH, through which CaCO 3 precipitates (its contents increase with hydration levels and increased temperature and are collected in Table 3), as the contents of CH after 3, 7 and 28 days of hydration are too low with regard to the high amounts of formed C-S-H or the reactivity of di-/tri-silicates. After 90 days, the amount of formed CH at both temperatures (around 12%, the exact values are collected in Table 3  increased temperature (from 25 to 40 °C) somewhat accelerates the hydration speed and increases the degree of reactivity of the clinker phases. The amount of stable hydration products (CH, Mc, CC -) is changed slightly during the hydration process, the differences being minor, around 1% or less (see Table 3). The crystallinity of the samples with added Li 2 CO 3 does not depend on temperature. After 3 days of hydration it is around 53% and after 90 days of hydration it is 49%.
The exact values are collected in Table 2. The crystallinity in the samples with added medium soluble carbonates is always higher than in PC samples. After 90 days of hydration the values are 18% higher compared to samples with clear PC. This is attributed to the high amounts of the present carbonate (Li 2 CO 3 + CaCO 3 ) in samples with added Li 2 CO 3 .

4. 1. KHCO 3
A 15% addition of the highly soluble carbonate KHCO 3 significantly affects the cement hydration due to the high solubility and its important role in changing the pH-value (towards making it less basic). With the applied Rietveld method the following phases were identified: CH, Mc, partly crystalline C-S-H, CaCO 3 , K 2 SO 4 , calcium aluminum hydrate C 3 AH 6 , unreacted clinker phases C 3 S, C 2 S, C 4 AF and compounds in traces (under 2%): SiO 2 , gismondine CAS 2 H 4 and carbosilicate tilleyite 3CaO·SiO 2 · 2CaCO 3 . The hydration of the clinker phases (C 3 A in C 2 S) is accelerated, whereas the hydration of C 3 S is decelerated. It reaches contents comparable with those in PC only after 90 days of hydration at 25 °C and at 40 °C after 90 days of hydration there is some more reacted C 3 S than in PC. (90 days hydration: KHCO 3 25 °C 1.52%, 40°C 0.40%; PC 25 °C 1.33%, 40 °C 1.28%.) The amount of formed C-S-H is increased and the differences compared with clear PC only increase with hydration stages. The amounts of the formed C-S-H are at both temperatures comparable and are slightly higher at 40 °C (See Figure  8). After 90 days of hydration in the samples with KHCO 3 there is 7.56% C-S-H at 25 °C and 7.76% at 40 °C. (PC 25 °C 5.19%; 40 °C 5.99%.) Already after 3 days of hydration at both temperatures no KHCO 3 is visible. The hydration products make it clear that part of KHCO 3 reacts with gypsum (CaSO 4 · 2H 2 O) at both temperatures, during which K 2 SO 4 and CaCO 3 is formed. Apparently the entire available sulfate in gypsum is used for the formation of K 2 SO 4 during the reaction, thus we predict that the formation of AFt or Ms is hindered. After 90 days at 25 °C in samples with KHCO 3 there is 3.40% K 2 SO 4 present and at 40 °C there is 3.31% K 2 SO 4 present. The amount of formed Mc increased, the values are collected in Table 3. The CH contents are low at both temperatures during the hydration process (see Table 3) compared to the high amounts of formed C-S-H or the reactivity di-/tri-silicates. We anticipate the KHCO 3 dissolving in CH and furthermore the CH reacting with dissolved CO 2 during which CaCO 3 is formed (the contents are collected in Table 3). Thus, in samples with KHCO 3 after 90 days of hydration there is 25 °C 10.92% CaCO 3 present and 12.57% at 40 °C. The amount of reactive carbonate is around 3%.
The amounts of SiO 2 , gismondine, C 4 AF, C 3 AH 6 and tilleyite are small, not exceeding 2%. At a higher temperature (from 25 to 40 °C) the amount of the reacted clinker phases C 3 S and C 2 S is higher, which is consequently shown on a higher amount of hydration products (CH, C-S-H) - Figure 8. At 40 °C the amount of CaCO 3 is slightly higher during the hydration process as there was more CH available, which could then react with dissolved CO 2 .
The crystallinity of samples with added KHCO 3 does not depend on temperature. After 3 days of hydration it is 50% and after 90 days of hydration it is 44%. The exact values are collected in Table 2. The crystallinity in samples with added KHCO 3 is always higher compared to PC samples and after 90 days of hydration the values exceed those of clear PC samples by 13%.

4. 2. K 2 CO 3
A 15% addition of the highly soluble carbonate K 2 CO 3 significantly affects the cement hydration. With the applied Rietveld method the following phases were identified: CH, Mc, partly crystalline C-S-H, CaCO 3 , K 2 SO 4 , K 2 CO 3 , calcium aluminum hydrate C 3 AH 6 , unreacted clinker phases C 3 S, C 2 S, C 4 AF and compounds in traces (under 2%): SiO 2 , gismondine CAS 2 H 4 and carbosilicate tilleyite 3CaO·SiO 2 · 2CaCO 3 . The hydration of the clinker phases (C 3 A and C 2 S) is accelerated, whereas the hydration of C 3 S is decelerated in the first stages of the hydration (3 and 7  The hydration products make it clear (see Figure 9) that K 2 CO 3 at both temperatures reacts with gypsum (CaSO 4 · 2H 2 O), at which point K 2 SO 4 is formed. The entire available sulfate is apparently used for the formation of K 2 SO 4 and thus the formation of AFt or Ms is indirectly hindered. After 90 days at 25 °C in samples with K 2 CO 3 there is 5.03% of K 2 SO 4 present and 4.82% at 40 °C. A part of K 2 CO 3 remains unreacted, but the amount after 90 days of hydration at both temperatures is small, not exceeding 1.5% (25 °C 1.29%, 40 °C 1.22%). The amount of formed Mc is minimally increased after 90 days of hydration at both temperatures, the values are collected in Table 3. The CH amounts in the beginning stages of hydration (3 and 7 days at 25 °C and 3 days at 40 °C) are low compared to the high amounts of formed C-S-H or reactivity of di-/trisilicates. Thus we predict that a part of K 2 CO 3 dilutes in CH during the beginning phases of hydration, during which CaCO 3 is formed. After 90 days of hydration at 25°C and 28 days of hydration at 40 °C the CH is not being used up anymore and so the contents of CH exceed those in clear PC (see Table 3). In samples with K 2 CO 3 after 90 days of hydration there is around 7% calcium (K 2 CO 3 + CaCO 3 : 25 °C 6.91%; 40 °C 7%) present for which we predict that it serves as an inert filling during the later stages of the cement hydration process. The amount of reactive carbonate is after 90 days of hydration and the both temperatures around 6%.
The amounts of SiO 2 , gismondine, C 4 AF, C 3 AH 6 and tilleyite are small, not exceeding 2%. At an increased temperature (from 25 to 40 °C) the reactivity of the clinker phase C 2 S is slightly higher, whereas the majority of the reacted C 3 S is only visible after 28 days of hydration (see Figure 9).
The crystallinity of the samples with added K 2 CO 3 does not depend on temperature. After 3 days it is between 45 and 50% and after 90 days of hydration it is 45%. The exact values are collected in Table 2. The crystallinity in samples with added K 2 CO 3 is always higher compared to PC samples and after 90 days of hydration the values exceed those of clear PC samples by 14%.

Discussion
Results presented above show that the addition of differently soluble carbonates does have an influence on cement hydration, the higher the solubility the more obvious the differences are (see Figure 10).
Hydration process and the contents of stable hydration products do not change significantly in the temperature range 25 to 40 °C. The hydration of the clinker phases (C 3 A, C 3 S and C 2 S) is at 40 °C somewhat accelerated and the amount of stable phases (E in slightly soluble carbonates, Mc, CH, calcite and C-S-H) of cement hydration is slightly changed. Presence of stable phases in our samples confirm T. Matschei calculations. 3 Slightly soluble carbonates (CC -, calcite, MgCO 3 , dolomite) accelerate the hydration of the clinker phases C 3 A, C 3 S and C 2 S. Medium soluble (Li 2 CO 3 ) and highly soluble (KHCO 3 , K 2 CO 3 ) carbonates accelerate the hydra-tion of C 3 A and C 2 S, whereas the hydration of C 3 S is decelerated: -in samples with added Li 2 CO 3 even after 90 days of hydration; -in samples with added KHCO 3 at 25 °C even after 90 days of hydration, at 40 °C in the first stages of the hydration (3, 7 days), after 28 and 90 days the amount of unreacted C 3 S is smaller than in PC samples; -in samples with K 2 CO 3 in the first stages of the hydration (3, 7 days), after 28 and 90 days the amount of unreacted C 3 S is smaller than in PC samples. The amount of reacted carbonate in cement hydration at a 15% addition of slightly or moderately soluble carbonates does not exceed 5% and is not affected by their solubility. At a 15% addition of the highly soluble carbonate K 2 CO 3 a higher amount of carbonate is used up during the hydration process and after 90 days it is around 6%. At the addition of the highly soluble carbonate KHCO 3 the change in the pH value also plays a role in addition to the solubility and great amounts of CaCO 3 are formed in samples due to the reaction between dissolved CO 2 and CH (after 90 days they exceed 10%). The rest of the carbonate serves predominantly as an inert filling.
In PC samples as well as samples with differently soluble carbonates the stable hydration products are Mc, AFt (at the addition of slightly soluble carbonates), CC -, C-S-H and CH. While the samples were prepared with PC which had a molar ratio SO 3 /Al 2 O 3 = 0.9 and even after 3 days of hydration in all samples Mc and Hc were not identified, we conclude, that the molar ratio CO 2 /Al 2 O 3 is equal or higher from 1. That means that all our samples the Mc was stabilized in the presence of slightly soluble carbonates, consequently causing the stabilization of AFt. The amount of Mc after 90 days of hydration at the addition of any kind of carbonates is slightly higher than in PC. In the presence of either medium or highly soluble carbonates AFt is not formed or its formation is delayed. Their amount is comparable in samples with or without slightly soluble carbonates. Carbonates accelerate the hydration of the clinker phases C 2 S and C 3 A, whereas the hydration of C 3 S is accelerated in slightly soluble carbonates and decelerated in moderately and highly soluble carbonates. The amount of reacted carbonate in cement hydration at a 15% addition of slightly or medium soluble carbonates does not exceed 5% and is not affected by their solubility; at a 15% addition of the highly soluble carbonate K 2 CO 3 it is around 6%. An increase in temperature (25 to 40 °C) gradually effects one, the rate of hydration and two, the quantity of stable phase assemblage (E in slightly soluble carbonate-blended cements, monocarboaluminate, calcite and portlandite). The crystallinity of the samples is independent on the temperatures (25 to 40 °C) and is after the 90 days of hydration always lower than after 3 days of hydration in the samples of clear PC and in the samples to which carbonate was added. Acta Chim. Slov. 2017, 64, 381-396 Medve{~ek et al.: Influence of Various Soluble Carbonates on the Hydration ...
(also include samples of pure PC) had excess of reactive carbonate after 3 days of hydration already.
The amount of Mc after 90 days of hydration at the addition of any kind of carbonates is slightly higher than in PC samples. With the presence of differently soluble carbonates Mc is stabilized and in turn causes AFt to be stabilized in slightly soluble carbonates, whereas the formation of AFt in moderately and highly soluble carbonates is hindered. The amount of AFt in samples with or without the added slightly soluble carbonates is comparable and after 90% days of hydration the content is around 4%. It is well known that the addition of carbonate affects the change of the Ca/Si ratio in C-S-H, but it cannot cause the formation of new C-S-H gel, as it has no pozzolanic qualities. Scientific writings show us that a part of the carbonate is incorporated in the C-S-H, at which point calcium silicocarbonate hydrates are formed but we could not identify any with Rietveld method. We do, however, predict their formation, as with the addition of slightly soluble carbonates (CC -, calcite, MgCO 3 ) we can identify comparable C-S-H contents only after 90 days of hydration compared to clear PC samples, even if the hydration of di-/tri-silicates is accelerated. The amount of identified C-S-H increases with the solubility of the added carbonates and after 90 days of hydration with the addition of the slightly soluble carbonate dolomite the C-S-H contents exceed those of PC samples by 1%, with the addition of the medium soluble carbonate Li 2 CO 3 and the highly soluble carbonate KHCO 3 it is even higher for about 2%, with the addition of K 2 CO 3 the C-S-H contents are higher for about 5%. Even though the quantification of C-S-H with XRD analysis is more difficult and harder to determine due to its semi amorphous nature, we can conclude based on the gathered data that the addition of carbonates affects its formation during the hydration -in the way of changing the Ca/Si relation and/or partial inclusion of CO 3 2into its structure. The crystallinity of the samples does not depend on temperature and after 90 days of hydration is always lower than after 3 days of hydration both in PC samples as well as at the addition of differently soluble carbonates. After 90 days of hydration the crystallinity in PC samples is 31%, after the addition of slightly soluble carbonates the crystallinity is higher for about 10%, after the addition of moderately soluble carbonates for about 18% and after the addition of highly soluble carbonates for about 14% (when compared to clear samples of PC).

Conclusions
We can conclude that blending of PC with differently soluble carbonates was found to influence the hydrate assemblage of the hydrated cement. With the help of the Rietveld analysis the study indicated that Mc and not Ms were stable in the presence of any carbonate. Even more,