Geochemical kinetics of hydrated carbonates

Laura Bastianini, Stefan Baltruschat, Jens Hartmann, James Campbell, Spyros Foteinis, Rachel Millar, Aidong Yang, Xuesong Lu, Pranav Thoutam, Sean Higgins, Georgina Rosair, Jim Buckman & Phil Renforth .

​

2

CRYSTAL OCEAN

Amorphous Calcium Carbonate (ACC) & Ikaite

Formation to dissolution�

1. Technoeconomic assessment: Precipitation kinetics.

​

• Storage and transport to ocean: Stability in air and seawater.

​

• Addition to ocean:

Dissolution kinetics in seawater.

​

3

SYNTHESIS OF AMORPHOUS CALCIUM CARBONATE (ACC) CONSISTENLY – Lennie et al., 2004 Recipe

XRD patterns showing synthesis of ACC consistently with measurement of the precipitate straightaway after collection.

4

STABILITY TEST – HYDRATED CARBONATES

​

​

B1 -18°C - AFTER 9 MONTHS=

AMORPHOUS SAMPLE BUT PEAKS ARE 100% VATERITE

B2 +3°C - AFTER 9 MONTHS=

100% IKAITE

B3 +24°C - AFTER 9 MONTHS=

100% VATERITE

5

0

Lennie et al.(2004) method– HYDRATED CARBONATES

SEM images of ikaite – Sample B2 after 16 weeks – 100% ikaite

SEM images of ACC – Sample analysed straightaway after collection

6

Protocol:

1. TA measurement of 200mL SW filtered with 0.45µm PTFE filters at 2°C = 2007µmol/kg.
2. Dissolution of 0.3g ACC in 500mL SW filtered with 0.45µm PTFE filters.
3. Taking 200mL of this solution and placing it into the reactor vessel at 2°C.
4. TA measurement of the latter at 2C= 2572µmol/kg.

​

Could ACC be a good candidate for OAE?

DISSOLUTION OF ACC IN SW – INCREASE OF TA?

YES!

Solubility of ACC is at least over 10 times higher than that of any crystalline calcium carbonate (Meiron et al., 2011).

7

TA MAX OF ACC DISSOLVED IN SW EXPECTED

COMPARISON WITH PAPADIMITRIOU ET AL., 2014

ICP-OES Results:

Protocol:

Synthesis of 10 samples which are a 2% HNO3-DI water matrix (10mL) in which a small amount of ACC (50mg) has been dissolved.

Objective:

To compare the amount of ACC (0.12g ACC in 200mL SW) with the amount of ikaite added to SW (Papadimitriou et al., 2014).

Results:

81% water (molecular bound + free water) in sample of ACC.

∆TA(carbonate alkalinity)=2200µEq/kg.

TA Max expected = ∆TA+TA₀ (=TA_{SW last exp 10C})=4906 ±44µEq/kg; Papadimitriou Paper: 0.3 g of synthetic ikaite was introduced into the cold reactor (500mL), with an overall average of 0.303 ± 0.002 g (n = 31), equivalent to 2915 ± 21 µmol carbonate alkalinity.

TA Max last exp - TA₀ =1906µEq/kg.

1906/ ∆TA=86% of solid dissolved. Why not 100%? Carbonate precipitation and/or ACC not completely dissolved.

8

GEOMETRIC SURFACE AREA OF ACC

Approach:

1. Determination of average area using ImageJ by drawing the spheres of ACC.
2. Particle weight average volume of ACC from Liu et al., 2010.
3. Average density of ACC from Liu et al., 2010.
4. Mass=Density*Volume
5. Geometric surface area = Area/Mass

Results:

Average area=1.2*10^-13m²

Particle weight average volume of ACC=7.96*10^-16cm³

Average density of ACC=1.62g.cm^-3

Mass=1.29*10^-15g

Geometric surface area=92.4m^2.g^-1 – Result in the rough estimate for ACC (ACC with a surface area between 100 and 200 m₂.g₋₁ have been independently reported by Cai et al. (2010) and Gebauer et al. (2013).

​

ACC is made from the aggregation of individual ACC particles on a nm scale. The presence of nanoparticles means that ACC has a certain level of porosity. In fact, ACC has the highest specific surface area of the different polymorphs of

CaCO3 (Sun et al., 2018).

​

9

PROTOCOL:

​

1. TA measurement of 200mL SW filtered with 0.45µm PTFE filters at 2°C.
2. Placing directly 0.12g ACC in 200mL SW into the reactor vessel at 2°C .
3. Using stirrer during experiment and bubbling solution with CO2 stripped lab air.
4. Measure the pH continuously during experiment.
5. Wait for 1 hour and then filter solution with 0.45µm PTFE filters.
6. Measure TA of the solution.
7. Repeat experiment after waiting 0.5 to 170 hours (n=40).
8. Calculate saturation state (Ω) of ACC using PHREEQC model.

​

SW collected from St Abbs (Scotland, UK) in November 2019

DISSOLUTION KINETICS IN SW at -5°C to 10 °C – RESULTS

Results:

​

1. ACC dissolves within hours, whereas ikaite dissolve within days. The decline of TA is certainly related to the precipitation of carbonates in the reactor vessel. Indeed, some particles analysed by Raman spectroscopy after filtration of the reacting solution show the persistent presence of undissolved hydrated carbonates, which can act as seeds for secondary precipitation.

​

• TA is clearly increasing with temperature for both hydrated carbonates, consequently solubility of hydrated carbonates is enhanced with temperature.

​

• The values of Ω versus time for ACC are significantly lower (0.01-0.03) than the ones of ikaite (0.2-1), which corroborate the higher surface-area-normalized dissolution rates of ACC in comparison to the ones of ikaite.

​

​

10

Example of dissolution kinetics experiment at 2°C

RATE OF DISSOLUTION OF ACC IN SW

Approach:

​

1. Rate is expressed in mol.m^-2.s^-1

​

• R=K . (1-Ω)ⁿ where K is a constant depending on temperature and salinity which are fixed in this experiment, and Ω is the saturation of ACC.

​

​

• Rate could be obtained by calculating the slope of first 2 values of TA.

​

​

Method: Rate (R) has been obtained by the slope of the 2 first values of the curve of TA versus time.

KINETICS OF DISSOLUTION IN SW

​

Logarithm of the surface-area-normalized ikaite-solution reaction rate as a function of the logarithm of the deviation of the saturation state of the solution from equilibrium with respect to ikaite (Papadimitriou et al., 2014).

Results: R clearly increases with temperature for both minerals, but rates of ACC are higher than ikaite. The graph of log (1-Ω) versus log(RSA) for both minerals dissolution experiments correlates well with Figure 3b of Papadimitriou et al. (2014), since our results are in the same order of magnitude, but within a much more limited window (i.e. as for ACC: log (1-Ω)=-0.015 to 0.01; log (R_SA)=1.5 to 2; and ikaite: log (1-Ω)=-1 to -0.3; log (RSA)=1.3-1.6).

Ikaite

ACC

Method: Decomposition of hydrous ACC (7-9 mg) was studied using the thermogravimetric method. Five tests were performed in a TGA 5500 (TA Instruments) at heating rates of 2.5, 5, 10, 15, 20 and 25°C min-1 in flowing air (25 ml. min^-1). The loss of water from hydrous ACC occurred from room temperature to 460°C (176 mins) at 2.5°C.min^-1, and 514°C (20 mins) at 25°C.min^-1.

TGA RESULTS - ACC

​

Results: The mean chemical formula of hydrous ACC was determined to be CaCO3.1.7H2O, which has never been established previously. The dehydration portion (below 550°C) of the TGA data closely resembles that of Ihli et al. (2014) - CaCO3.1.4H2O and Konrad et al. (2016) - CaCO3.0.4H2O.

The TGA data suggests a rapid loss of about 0.9 moles of water below 25-100°C likely due to loss of surface-bound water, and about 0.8 mole of water lost above 100°C, likely due to structurally trapped water.

STABILITY TEST - ACC AT 4°C

​

Method: A stability test of ACC at 4°C (seawater temperature of potential operational site: Bedford Basin, in Halifax, Canada) was conducted. ACC was created following the Lennie et al. (2004) and rinsed with ethanol. A Raman measurement was acquired every day over 1 month on the same sample of ACC.

RESULTS: Main characteristics showing that this sample remains hydrous ACC after 120 days:

​

1. A very high broad peak from 60 to 290 cm-1 (Gierlienger et al., 2013) characteristic of an amorphous phase.

​

• Moderately shifted and broader peaks at approximately 716 cm^-1 compared to crystalline calcium carbonate (Gierlienger et al., 2013).

​

• The signal in the C–H stretching region 2800–3050 cm^-1 (Gierlienger et al., 2013).

​

• The weak peak at 3400 cm^-1 diagnostic of H2O stretching (Tlili et al., 2001).

​

• The peak at 1077 cm^-1 which is characteristic of hydrous ACC (Tlili et al., 2001).

​

​

​

​

​

STABILITY TEST- ACC AT 23°C

​

Method: A stability test of ACC at 23°C was conducted since that if we can store ACC for up to 2hours at ambient temperature, this is significant in terms of life cycle assessment as we do not need a freezer to store ACC at the industrial scale. ACC was created following the Lennie et al. (2004) and rinsed with ethanol. A Raman measurement was acquired every day over 1 month on the same sample of ACC.

RESULTS: Main characteristics showing that this sample remains hydrous ACC after 113 days:

​

1. A very high broad peak from 60 to 290 cm-1 (Gierlienger et al., 2013) characteristic of an amorphous phase.

​

• Moderately shifted and broader peaks at approximately 716 cm^-1 compared to crystalline calcium carbonate (Gierlienger et al., 2013).

​

• The signal in the C–H stretching region 2800–3050 cm^-1 (Gierlienger et al., 2013).

​

• The weak peak at 3400 cm^-1 diagnostic of H2O stretching (Tlili et al., 2001).

​

• The peak at 1077 cm^-1 which is characteristic of hydrous ACC (Tlili et al., 2001).

​

​

​

​

​

15

IMPLICATIONS ON OAE

1. ACC has the potential of being a better candidate for ocean alkalinity enhancement (OAE) since the solubility of ACC is at least 10 times higher than any crystalline calcium carbonate (Meiron et al 2011) and ACC has the highest specific surface area of all the different polymorph of calcium carbonate (up to 350m2.g-1) (Sun et al., 2018).

​

​

• Since ACC is a less hydrated form of calcium carbonate than ikaite, ACC is prone to remove greater amount of CO2 with respect to mass than ikaite.

​

​

• The saturation states of ACC are significantly lower (0.01-0.03) than the ones of ikaite (0.2-1), which corroborate the higher surface-area-normalized dissolution rates of ACC in comparison to the ones of ikaite (ACC dissolves in hours, ikaite in days).

16

​

​

1. According to XRD results, ikaite remains stable after 9 months at 3C and ACC remains stable as well but peaks are 100% vaterite with Lennie et al. (2004) recipe.

​

• Good SEM images of ACC and ikaite for publication.

​

• ACC is a good candidate for OAE: dissolution of ACC in SW does increase TA.

​

• ICP-OES results for ACC dissolved in SW: TA Max expected =4906 ±44µEq/kg; 86% of solid dissolved. Why not 100%? Carbonate precipitation in the reactor vessel and/or ACC not completely dissolved.

​

• Geometric surface area of ACC: 92.4m².g^-1 - Result in the rough estimate for ACC (ACC with a surface area between 100 and 200 m2.g−1 have been independently reported by Cai et al. (2010) and Gebauer et al. (2013).

​

• Dissolution kinetics of ACC in SW: TA is increasing over time and reaches a plateau; After 24hours, TA starts to decrease (due certainly to precipitation of carbonates in the reactor vessel).

​

• Rate of dissolution of ACC in SW clearly increases with temperature and the curve of log (1-Ω) versus log(R) in our experiment correlates well with Papadimitriou et al 2014.

​

• TGA of ACC: The mean chemical formula of hydrous ACC was determined to be CaCO3.1.7H2O, which has never been established previously. The dehydration portion (below 550°C) of the TGA data closely resembles that of Ihli et al. (2014) and Chaliulinan (2019).

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

SUMMARY

17

Thank you for your attention!

​

Any questions?