- Email: [email protected]
Understanding the importance of iron speciation in oil-field brine pH for CO2 mineral sequestration
Journal of CO2 Utilization 16 (2016) 78–85
Contents lists available at ScienceDirect
Journal of CO2 Utilization journal homepage: www.elsevier.com/locate/jcou
Review article
Understanding the importance of iron speciation in oil-field brine pH for CO2 mineral sequestration Patricia Córdobaa,* , Qi Liua , Susana Garciaa , Mercedes Maroto-Valera,b a Centre for Innovation in Carbon Capture and Storage (CICCS), Institute of Mechanical, Process and Energy Engineering (IMPEE), Heriot-Watt University, EH14 4AS, United Kingdom b The Institute of Petroleum Engineering (IPE), Heriot-Watt University, United Kingdom
A R T I C L E I N F O
Article history: Received 9 October 2015 Received in revised form 22 May 2016 Accepted 11 June 2016 Available online 12 July 2016 Keywords: CO2 Mineralogical sequestration Oil-field brines Above-ground brines Modelling tools
A B S T R A C T
Carbon dioxide (CO2) mineralogical sequestration using oil-field brines (above-ground field brines) is gaining attention for the storage of CO2 into stable mineral carbonates. This involves complex processes that are affected by brine composition, pressure, temperature and pH where the latter is the key parameter, as carbonation conversion rates and the specific carbonates species formed are strongly dependent on pH. This study investigates the effect of iron (Fe) in the pH of three synthetic brines prepared as analogue to oil-field brine by combining a number of experimental and modelling tools to elucidate whether or not Fe speciation may affect the pH of brines. X-Ray powder Diffraction (XRD) analyses were also carried out to characterise the mineralogy of the synthetic brines. Results indicate that the Fe speciation affects the pH of oil-field brines; brines containing Fe3+ experiment a rapid increase of the pH as a consequence of Fe2O3 precipitation by aging of Fe(OH)3 under alkaline conditions. Brines containing Fe2+ show a sharp drop of the pH as a result of air oxidation of highly alkaline Fe(OH)2 suspension induced by KOH addition by means of dissolution-oxidation-crystallisation mechanism where particles of Fe(OH)2 provide the substrate for the g-FeO(OH) growth. The master variable governing the rates at which these compounds form is pH. ã 2016 Elsevier Ltd. All rights reserved.
Contents 1. 2.
3.
4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Brine preparation . . . . . . . . . . . . . . . . . . . . . . . . 2.1. pH stability studies . . . . . . . . . . . . . . . . . . . . . . . 2.2. Mineralogical composition of solid phases . . . . 2.3. Geochemical modelling . . . . . . . . . . . . . . . . . . . 2.4. Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . pH studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Characterisation of brines . . . . . . . . . . . . . . . . . 3.2. Identification of solid phases in brines 3.2.1. Aqueous speciation of brines . . . . . . . . 3.2.2. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
78 79 79 79 80 80 80 80 80 80 80 85 85 85
1. Introduction
* Corresponding author. E-mail address: [email protected] (P. Córdoba). http://dx.doi.org/10.1016/j.jcou.2016.06.004 2212-9820/ã 2016 Elsevier Ltd. All rights reserved.
Carbon dioxide (CO2) mineral sequestration using brines is gaining attention for the storage of CO2 into stable mineral
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
carbonates [1]. Brines are commonly found either underground, namely underground brines in saline aquifers, or above-ground, most notably as a by-product of oil and natural gas extraction and known as oil-field brines, and its fate is mainly disposal [2]. However, oil-field brines (above-ground brines) could also be a promising approach to sequester CO2 as they are produced in large volumes [3,4]. Brines in addition to the dominant Na and Cl ions [5], also have significant concentrations of Ca, Mg and Fe, which react with CO2 to produce CaCO3(s), MgCO3(s), Fe2CO3(s) and other products under favourable conditions. However, most both under and aboveground brines are generally acidic in nature with a typical pH ranging from 3 to 5 [3], so that carbonates will not form in this pH range. The suitable pH range for formation of carbonate is 7.8 or higher [6], where CO32 dominates. Therefore, to boost the precipitation of mineral carbonates by reaction between brine and CO2, the pH of the brine must be modified e.g. by caustic addition before it can be used for the carbonation reaction [3]. In this regard, a factor that should be considered as the first step to boost the precipitation of mineral carbonates by the increase of the pH in above-ground brines is Fe speciation. The hydrolysis of ferrous (Fe2+) and ferric (Fe3+) in aqueous solutions involves pH changes because of the formation of a variety of partially oxidized meta-stable Fe2+-Fe3+ aqueous complexes, which could prevent oil-field brines from using for CO2 mineral carbonation. Fe2+ show high solubility in aqueous solutions, whereas Fe3+ shows a stepwise solubility at pH >1.0 [7]. The Fe2+-hexaquo-complex [Fe (H2O) 6]2+, which is present in acid solutions, hydrolyses by the formation of FeOH+ and precipitates as the hydroxides Fe(OH)2 upon increasing the pH value (Bard et al., 1985). Fe3+ exist only in acidic solutions (pH <2.5) as the hexaquo-complex Fe(H2O) 63+, which increasing pH is subsequently transformed to mono- and the dyhidroxo-complexes. In the pH range 5 at least four different Fe3+ ions coexist in aqueous solution: Fe3+; Fe (OH)2+; Fe(OH)+2 and the dimer Fe2(OH)42+ [8]. The spontaneous chemical oxidation of Fe2+ to Fe3+ by O2 is also a complex process involving pH changes with a variety of partially oxidized meta-stable Fe2+-Fe3+ intermediate species, which ultimately transform into a variety of stable Fe-oxide end-products such as hematite (Fe2O3), magnetite (Fe3O4), goethite (a-FeOOH), and lepidocrocite (g-FeOOH) [9]. These crystalline products form by competing mechanisms and the proportion of each in the final product depends on the relative rates of formation. The master variable governing the rates at which these compounds form is pH. Other important factors are temperature and the presence of additives. Although the conditions governing Fe chemistry under different environments have been identified by the above mentioned researchers among others, the role of Fe speciation in aboveground brines aimed at mineral sequestration of CO2 and how it can affect brine composition and pH has not been elucidated yet. Druckenmiller and Maroto-Valer [4] reported that brines with low Fe concentration (9 ppm) showed a pH relatively constant after addition of a strong base. By contrast, brines with high Fe concentration (121–476 ppm) had a rapid decline in pH during the first hours, after which the decline seemed to level off. However, it was not ascertained whether Fe2+ or Fe3+ caused the pH drop in their study. With the advent of making CO2 sequestration through mineral carbonate formation a viable approach by using oil-field brines, the role of Fe speciation in oil-field brines pH needs to be elucidated to reach suitable pH range for further formation of carbonate. Towards this goal, this work i) studies the evolution of the pH in three synthetic brines with different composition prepared as analogues to an oil-field brine; ii) investigates the effect of Fe in the pH of the three above-ground brines when adjusting the pH; and iii) elucidates the role of Fe speciation in brines when adjusting the
79
pH by combining a number of experimental tools and a geochemical model based on acquired experimental data. To reproduce experimental conditions when attempting to use oilfield brine for carbon sequestration, ambient temperature and ambient pressure conditions were chosen for pH studies and modelling calculations. 2. Methodology 2.1. Brine preparation Three different synthetic brines, namely B1, B2 and B3, were prepared as an analogue to the natural brine (OH-2) which comes from a natural gas well in Youngstown, Ohio [4]. Owing to the complex composition of natural brines, only major ions were considered to prepare the three synthetic brines, including Na+, K+, Mg2+, Ca2+, Fe3+/Fe2+, Sr2+, Ba2+and Cl. Brines with their corresponding duplicates were synthesised by dissolving NaCl, KCl, MgCl26H2O, CaCl22H2O, FeCl3/FeCl2, SrCl2 and BaCl22H2O (Sigma-Aldrich) in Milli-Q water. Solutions were mixed in a beaker thoroughly with magnetic stirrers until a complete dissolution was attained. After this procedure, 20 mL of each B1, B2, and B3 was filtered by glass-fibre filter of 0.2 mm pore size and acidified with HNO3 to prevent oxidation of ions and stabilize the metal concentration. Brine aqueous phases were then analysed by Inductively Coupled Plasma Mass Spectrometry (ICP-MS) by the X-SERIES II device from Thermo Fisher SCIENTIFIC to determine the concentration of the aforementioned major ions in order to compare them with target values. The quantitative analyses were performed with an external calibration, using an external standard of similar matrix to the samples, which covered concentrations range expected forming the calibration lines. The internal correction was carried out by means of an internal standard (In 10 ppb). Brine compositions only differ in the type of Fe present in the aqueous phase of brines, with Brine B1, B2, and B3, containing Fe3+, Fe2+, and non-Fe, respectively. 2.2. pH stability studies The pH evolution was studied at ambient temperature and pressure in open atmospheres using B1, B2, and B3 over a period of 8-14 days. According to the previous study [4], it was preliminarily determined to raise the pH from the initial values to an upper limit below the point at which Fe hydroxide precipitates. According to a Fe-H2O system, in a solution containing Fe2+, iron hydroxide has a much lower solubility product than the hydroxyl forms of the other abundant cationic species [7]. Because the analogue brine selected in this study showed a high Fe composition from which a pH of 6.8 would promote precipitation, a pH value of approximately 6.3 was selected for the pH stability of B1 and B2, and thus providing an upper limit well below 6.8. For B3 a pH value of 9.0 was selected as the target as the initial pH of B3 was higher than 6.3. The precipitation of mineral carbonates is mostly dependent on brine pH and it is favoured above a basic pH of 9.0. In each pH stability experiment, 100 mL of fresh synthetic brine was poured into a 250 mL conical flask. The brine solution was mixed effectively and continuously throughout the experiment by using a magnetic stirrer. The pH variation with time was measured by a Thermo Orion 420A+ bench top pH meter. The brine pH was raised to above 6.3 or 9.0 initially by adding 1.0 M KOH solution. After the brine pHs were raised to the desired values (6.3 or 9.0), addition of KOH was stopped and time set as zero. pH measurements were then made every 5–10 min for the first 2 h, then every 1 h. After the first 12 h, pH was measured every 12 h. When the
80
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
brine pH was stable around 0.2 continuously for 3 days, the stability study was then completed. The brine solutions were then filtrated by using cellulose nitrate membrane filters (0.45 mm pore size) and the solid residue was saved for further XRD analysis. The error for pH measurement was 0.02 owing to the calibration of pH probe (every 2 h for the first 12 h and thereafter every 24 h) and the variation of laboratory room temperature. 2.3. Mineralogical composition of solid phases X-Ray powder Diffraction (XRD) analyses were carried out to determine the mineralogical composition and solid species of brines after KOH addition. Analyses were carried out by using a HILTONBROOKS diffractometer with monochromatic Cu Ka1,2 radiation operated at 40KV and 20 mA, from 460 2u, at a step size of 0.05 , and scan rate of 3s/step. The classification of ground soil particles was carried out by a Coulter Laser Scattering analyser employing Fraunhofer diffraction (LS 200, Variable-Speed Fluid Module Plus (VSM + )). 2.4. Geochemical modelling Theoretical calculations with the PHREEQC code (version 2.0) and the coupled thermodynamic database Lawrence Livermore National Laboratory (LLNL) are used herein to calculate the 1) aqueous speciation of the synthetic brines and saturation index (SI) of Fe minerals with the progressive addition of KOH; and 2) pH evolution of synthetic brines with the progressive addition of KOH. As a result of the different speciation of Fe in the synthetic brines, the kinetic oxidation of Fe in B2 with the progressive addition of KOH on the pH stabilisation over 8 days was also modelled. The modelling calculations were conducted by using experimental pressure and temperature conditions as well as the chemical composition and pH values of the three brines under which the process takes place as input data. Detailed description of the model is provided as supporting information. The theoretical results from the model were then compared to experimentally derived results on the same system. 3. Results and discussion 3.1. pH studies The pH values of brines and the experimental parameters in the pH stability studies are given in Table 1. The results of pH studies for B1 over a period of 192 h (8 days) are shown in Fig. 1a. B1 duplicates, namely B1A and B1B, containing Fe3+ show an initial pH of 1.68 and 1.70, respectively (Table 1). The B1A and B1 B pHs were therefore adjusted to 6.3, an upper pH limit but below the point at which Fe(OH)2 precipitates (7.0–9.0.), by the addition of KOH. The evolution of the pH of the B1 against time shows that the pH fluctuates over the first 48 h, after which it remains (0.2 pH) stable till the end of the experiment. Over a run period of 8 days, B1A and B1 B reach pH values of 6.41 and 6.36, respectively (Table 1). Table 1 Summary of pH stability studies. Brine type
Initial pH
Added KOH (mmoles)
Adjusted pH
Final pH
B1A B1B B2A B2B B3A B3B
1.70 1.68 3.43 3.45 6.58 6.59
3.27 3.28 0.26 0.26 0.08 0.08
6.33 6.32 6.35 6.36 9.08 9.10
6.41 6.36 2.86 2.86 7.57 7.52
The evolution of pH for B2 over a period of 192 h (8 days) is shown in Fig. 1c. B2 duplicates, namely B2A and B2B, containing Fe2+ shows an initial pH of 3.44 and 3.45, respectively (Table 1). The B2A and B2 B pHs were also adjusted to an upper pH limit but below the point at which Fe(OH)2 precipitates(7.0–9.0.) Accordingly, B2A and B2 B pHs were adjusted to 6.3 by KOH addition. Results show that B2 experiences a sharp drop of the pH over the first 24 h with a gradual decrease until it stabilises (0.2 pH) after 192 h. Over a run period of 8 days, B2A and B2 B reach pH values of 2.86. The evolution of pH for B3 over a period of 336 h (14 days) is shown in Fig. 1e. B3 and a duplicate, namely B3A and B3B, containing no Fe shows an initial pH of 6.58-6.59, respectively (Table 1). Owing to the absence of Fe in this brine, B3A and B3 B pHs were adjusted to 9.0. Results show that B3 experiences a gradual drop of the pH along the experiment until it stabilises after 336 h. Over a run period of 14 days, B3A and B3 B reach pH values of 7.57 and 7.52, respectively (Table 1). The different pH patterns between B1 and B2 indicate that the pH of brines is highly affected by the Fe speciation. However, it should also be noted that B3 reaches pH stability (0.2 pH) after a longer time period when compared with B1 and B2, which is also indicative that the absence of Fe affects the pH stability of brines. 3.2. Characterisation of brines The composition of the natural brine (OH-2) and prepared synthetic brines is shown in Table 2. 3.2.1. Identification of solid phases in brines The XRD analysis revealed that NaCl and Fe2O3 are the crystalline phases detected in B1 (Fig. 2a). It should be noted that B1 showed a low crystallinity and small crystalline sizes, which caused most diffraction patterns to have poor signal to noise ratios and very broad peaks. Crystalline phases detected in B2 (Fig. 2b) are g-FeO(OH) and NaCl whereas NaCl is the only crystalline phase detected in B3 (Fig. 2c). The identification of g-FeO(OH) in the B2 solid phase suggests an oxidation of Fe2+ to Fe3+ in the B2 aqueous phase, where Fe2+ would be oxidised and hydrolysed to g-FeO(OH) when adjusting the B2 aqueous phase pH via KOH addition. Therefore, it can be stated that the different mineralogical composition of brines following pH adjustment is caused by the different Fe speciation in their corresponding aqueous phases. 3.2.2. Aqueous speciation of brines Geochemical modelling reveals, with the exception of Fe, no significant differences in the B1 (pH 1.68-1.70), B2 (pH 3.43-3.45), and B3 (pH 6.58-6.59) aqueous phases despite the KOH addition (Table 3). Ca and Na as free cations (Xn,n+1,+2) are the predominant aqueous complexes in all the brines followed by K, Mg, and Sr. 3.2.2.1. Fe3+ effect on aqueous speciation. The geochemical modelling indicates that the addition of KOH to B1 produces a slight increase of pH from 1.68 to 1.74 (Fig. 1b). However, experimentally, B1 pH initially fluctuates, but it remains stable (0.2 pH) at the end of the experiment at 6.41 pH (Fig. 1a). The differences between the B1 experimental and modelling pH pattern could be related to either geochemical modelling accounting for equilibrium reactions involving Fe, and/or the pH stability of Fe species non-detected by XRD. According to the geochemical modelling, prior to the KOH addition, the B1 aqueous phase is oversaturated (SI = IAP/K >0) with respect to Fe2O3 (SI value of 3.28) and a-FeOOH (SI 1.13) and Fe3+ as free cation followed by FeOH2+, Fe(OH)2+, Fe3(OH)24+, Fe(OH)45+, and Fe(OH)3 are the Fe-aqueous complexes with highest
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
81
Fig. 1. a) B1 pH over time; b) B1 pH modelled vs KOH; c) B2 pH over time; d) B2 pH modelled vs KOH; e) B3 pH studies over time; f) B3 pH modelled vs KOH.
Table 2 Target concentrations of major ions in OH-2 and B1, B2, and B3.
mg/L Ca2+ K+ Na+ Mg2+ Sr2+ Fe3+ Fe2+
Brine (OH-2)
B1
Variation (%)
B2
Variation (%)
B3
Variation (%)
19350 1180 48500 2055 1800 476 476
20010 1243 50215 1887 1790 451 –
3.4 5.3 3.5 8.2 0.6 5. –
19490 1143 48185 1838 1765 – 452
0.7 3.1 0.6 10.5 1.9 – 5.1
19235 1204 50015 1871 1737 – –
0.6 2.0 3.1 8.9 3.5 – –
activities at the initial pH of 1.69. The occurrence of the Fe aqueous complexes in the B1 aqueous phase promotes the acidification of the aqueous phase according to the reactions: Fe3+
2+ + (aq) + H2O $ FeOH (aq) + H
Fe3+
(aq) + 2H2O $ Fe
(OH)2+
(aq) + 2H
(R1)
+
(R2)
2Fe3+
+ 4+ (aq) + 2H2O $ Fe2(OH)2 (aq) + 2H
(R3)
3Fe3+
+ 5+ (aq) + 4H2O $ Fe3(OH)4 (aq) + 4H
(R4)
Fe3+
(aq) + 3H2O $ Fe
(OH)3(aq) + 3H+
(R5)
where Fe-aqueous complexes from (R1) to (R5) remain in the B1 aqueous phase under acidic conditions. When modelling the KOH addition, the geochemical modelling predicts, in addition to the occurrence of the above mentioned Feaqueous complexes, the occurrence of Fe2+, FeOH+, Fe(OH)2, and Fe (OH)3 and a pH (Fig. 1b) showing a different pattern to that of the experimental pH (Fig. 1a). The geochemical modelling also predicts the oversaturation of the B1 aqueous phase with respect to Fe2O3 (SI of 7.77) and a-FeOOH (SI of 2.87). The modelled pH pattern and the occurrence of Fe2+, FeOH+, Fe(OH)2, and Fe (OH)3 complexes would be consistent with the reduction of Fe3+ and Fe3+-aqueous complexes when adding KOH. Therefore, it is postulated that Fe3+ as a free cation, whose activity decreases after KOH addition, would be reduced to Fe2+ whose activity is the highest among
82
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
the release of the H+ by the formation of Fe3+ aqueous complexes ((R1)–(R5)); 2) the reduction of Fe3+ to Fe2+ by OH from KOH (R6); and 3) the formation of FeOH+ and Fe(OH)2 complexes by Fe2+solvation, which acidifies the B1 aqueous phase ((R7) and (R8)). By contrast, the experimental pH pattern of B1 (Fig. 1a) would be consistent with the precipitation of Fe3+ in the B1 aqueous phase as Fe(OH)3 via KOH addition according to:
B1
a) 800
1- Fe2O3
700
1
2- NaCl
600 Counts
500 1
400
1 2
300
Fe3+
100
Because Fe(OH)3 is not detected by XRD, it is postulated that Fe (OH)3(s) could have turned into a-Fe2O3(s) by aging, which would be in agreement with detection of Fe2O3(s) in the B1 solid phase by XRD (Fig. 2b) and with the alkaline experimental conditions under which the experiments were carried out. However, it should be noted that this does not rule out the occurrence of other crystalline or solid phases which concentration is not high enough (1%) to be detected by XRD and/or a formation of amorphous Fe solid phase, which cannot be detected by XRD. Several investigations about the constitution and aging of precipitated hydrous Fe3+oxide gels [10–14] obtained amorphous non-stoichiometry hydroxides by addition of bases to Fe3+ solutions which were not detected by XRD. This finding is in line with previous works on Fe speciation and precipitation processes. Feitknecht and Wytlenbach [15] reported the occurrence of FeOOH and Fe2O3 as a the result of the Fe (OH)3nH2O aging by agitation with KOH whereas Kasuaki [16] formed a-FeOOH particles under an alkaline environment (NaOH) at 200 C for 5 h. Cornell et al. [17] reported that alkaline solutions promote the change of Fe2O3.1/2 H2O and 5Fe2O3 9H2O into FeOOH and Fe2O3(s). Therefore, the experimental pH pattern in the B1 aqueous phase might be due to the formation of Fe2O3(s) by aging of Fe(OH)3(s). Hence, the Fe3+ effect on B1 aqueous speciation and pH lies in the role that KOH could play in R6 (modelling) and R9 (experimental), respectively. The former would demonstrate the feasibility of KOH to reduce Fe3+ to Fe2+ since the estimation of the DG of formation according to R6 or chemical potential translated in terms of redox potential infers the fact that the reduction of Fe3+ in the B1 aqueous phase is thermodynamically possible. The latter demonstrates the feasibility of KOH to react with Fe3+ to form Fe (OH)3 from which the Fe2O3(s) by aging can be formed, as observed experimentally.
0 0
10
20
30
40
50
60
2θ (degrees)
B2
b) 6000
1- FeO(OH) 2- NaCl
2
5000
Counts
4000
1 3000 2000 1000
2
1
1
1
1 1
1
0
0
10
20
30
40
50
60
2θ (degrees)
B3
c) 25000
1- NaCl 20000
Counts
200
1
15000 10000 5000
1 1
0 0
10
20
30
40
50
60
2θ (degrees) Fig. 2. a) Mineral and solid phases in B1; b) Mineral and solid phases in B2; c) Mineral and solid phases in B3.
Fe2+aqueous complexes in the B1 aqueous phase after KOH addition (Table 3) according to: 4Fe3+ + 4OH $ 4Fe2+
(aq) + 2H2O + O2(aq)
2+
(R6) +
Fe from (R6) would be solvated forming FeOH and Fe(OH)2 aqueous complexes, respectively, acidifying once again the B1 aqueous phase according to: Fe2+
+ + (aq) + H2O $ FeOH + H
Fe2+
(aq) + H2O $ Fe(OH)2 + 2H
(R7)
+
(R8)
The viability of (R7) and (R8) depends on 1) the thermodynamic feasibility of R6, where KOH would reduce Fe3+ to Fe2+ since the estimation of the Gibbs standard free energy of formation (DG ) or chemical potential translated in terms of redox potential infers the fact that the reduction of Fe3+ in the B1 aqueous phase is thermodynamically possible, and on 2) the kinetics of (R6). Therefore, the slight raise in the modelled B1 pH despite the KOH addition would be the result of a number of successive reactions: 1)
(aq) + 3OH
$ Fe(OH)3(s)
(R9)
3.2.2.2. Fe2+ effect on aqueous speciation. The geochemical modelling also indicates a different pH pattern between the initial modelled and the experimental pH for B2. In the initial model, the pH slightly increases at the beginning of the experiment (Fig. 1d) as opposed to a decreasing pH trend in the experiment after the KOH addition as time progresses (Fig. 1c). Because g-FeO(OH) (g-Fe3+O(OH)) is the only Fe crystalline phase detected in the B2 solid phase (Fig. 2b), the oxidation of Fe2+ to Fe3+ was identified as the cause of the experimental B2 pH value and the different pattern between the experimental and modelled pH values in the B2 aqueous phase (Fig. 1c and d). The experimental pH would be consistent with a rapid kinetic oxidation of Fe2+ to Fe3+ that would give rise to the formation of hydroxy-complexes of Fe3+ that progressively lower the B2 pH (2.86). By contrast, the modelled rise in pH would be consistent with a slowly oxidation of Fe2+ to Fe3+ by forming hydroxycomplexes of Fe3+ resulting in a slight increase of the B2 pH. Based on Fe-hydroxide compounds occurring under reducing and weakly acid to weakly alkaline conditions as intermediate phases in the formation of Fe oxides such as a-FeO(OH), g-FeO (OH), Fe3O4 [17–19], the oxidation of Fe2+ to Fe3+ and its
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
83
Table 3 Aqueous speciation of B1, B2, and B3. Before KOH addition
Ca Fe (2)
Fe (3)
Na K Mg Sr
Initial Eh (mV) Activity (mol/L) Ca2+ CaOH+ Fe2+ FeOH+ Fe(OH)2 Fe(OH)3 Fe (OH)42 Fe(OH)4 Fe3+ Fe(OH)3 Fe2(OH)24+ FeOH2+ Fe3(OH)45+ Fe(OH)2+ Na+ NaOH K+ KOH Mg2+ Mg4(OH)44+ Sr2+ SrOH+
After KOH addition
B1
B2
B3
B1
B2
B3
4.0
1.58
2.03
18.868
17.525
11.337
1.119e-01 7.351e-13 – – – – – 4.880e-19 4.159e-04 4.179e-11 4.195e-07 1.248e-04 1.685e-10 1.922e-06 1.575 1.217e-13 2.100e-02 3.385e-15 2.505e-02 3.273e-40 3.528e-03 8.412e-15
1.089e-01 4.025e-11 1.812e-03 1.500e-09 3.117e-17 3.248e-24 – – – – – – – – 1.557 6.773e-12 1.923e-02 1.745e-13 2.431e-02 2.913e-33 3.481e-03 4.672e-13
1.076e-01 5.483e-08 – – – –
1.119e-01 8.45313 5.53510 9.357e-18 3.973e-27 8.457e-36 – 8.050e-19 3.923e-04 5.994e-11 4.936e-07 1.354e-04 2.472e-10 2.093e-04 1.636 2.535e-10 2.300e-02 4.264e-15 2.505e-02 5.722e-40 3.528e-03 9.673e-15
1.089e-01 1.816e-11 5.276e-15 2.988e-10 4.249e-07 3.029e-03 5.424e-04 6.421e-06 – 1.428e-10 – 2.875e-29 – 1.705e-18 1.558 3.058e-12 1.942e-02 7.949e-14 2.433e-02 1.210e-34 3.480e-03 2.107e-13
1.081e-01 2.775e-05 – – – –
subsequent transformation to g-FeO(OH) through oxygen dissolution in a highly alkaline B2 aqueous phase is proposed. Therefore, the 1) aqueous speciation and 2) pH pattern for B2 considering the kinetic oxidation of Fe2+ to Fe3+ with the progressive addition of KOH over 8 days were modelled. The oxidation of Fe2+ to Fe3+ through O2 dissolution in a highly alkaline B2 aqueous phase, modelled using B2 aqueous phase chemical composition as input data, was based on the rate of oxidation of Fe2+ by O2 in water given by Singer and Stumm [20]:
– – – – – – – 1.620 9.713e-09 1.989e-02 2.489e-10 2.479e-02 1.138e-20 3.468e-03 6.417e-10
– – – – – – 1.567 1.991e-02 1.255e-07 2.718e-02 1.060e-09 3.550e-03 3.309e-07
where t is time in seconds, aOH is the activity of the hydroxyl ion, aFe2þ is the total molality of Fe2+ iron in the B2 aqueous phase, and Po2 is the oxygen partial pressure (atm). Detailed description of the model here is provided as supporting information. The kinetic modelling consisting of the B2 aqueous speciation over the 8 days of simulation (Fig. 3) shows that the molality of Fe3+ against time increases whereas the molality of Fe2+ against time is initially constant, but lessens as the reaction progresses, which is consistent with Eq 1. The kinetic model consisting of the B2 pH over the 8 days of simulation shows, as expected, a decrease of the pH
value (Fig. 4). As can be observed in the pH model, the initial pH of the B2 (6.35) experiences a sharp drop over the first day of simulation and a soft decrease between the 3rd and 8th day of simulation (Fig. 4). The modelled pH reaches the stability with the B2 solution at 3.10, which is nearly in line with the experimental B2 pH (2.86). The time for complete oxidation of Fe2+ is a matter of minutes in an aerated solution when pH is 7.0. However, Fe3+ forms aqueous complexes with OH precipitating as Fe oxy-hydroxides, so that pH decreases during oxidation. Because the rate has quadratic dependence on the activity of OH, the oxidation rate in the B2 aqueous phase rapidly increases. The modelling results presented above along with the experimental analyses allow us to propose three pathways for g-FeO(OH) formation via air oxidation of a highly alkaline Fe(OH)2 suspension in the B2 aqueous phase by means of dissolutionoxidation-crystallisation mechanism, in which the initial particles of Fe(OH)2 would provide the substrate for the g-FeO(OH) growth. These formation pathways are based on the occurrence of Fe3+-complexes showing the highest activities in the B2 aqueous phase when modelling KOH addition (Table 3). Accordingly, it is proposed that in line with the B2 composition, when Fe2+ is immersed in the B2 aqueous phase Fe(OH)2 is formed as the
Fig. 3. Kinetic oxidation of Fe aqueous complexes over 8 days of simulation.
Fig. 4. Kinetic variation of the B2 pH over 8 days of simulation.
dmFe2þ ¼ 2:92e9 þ 1:33e12 a2OH PO2 mFe2þ dt
ð1Þ
84
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
primary product under the alkaline conditions by KOH addition in a two-step formation process: + (aq) + 2OH (aq) $ Fe(OH) (aq) + OH (aq) $ Fe(OH)2
(R11)
2+
As the R11 proceeds, dissolved Fe is consumed and the solution becomes saturated with respect to Fe (OH) 2 involving the formation of green rust. According to previous studies [21,22], this phenomenon is characteristic of Fe(OH)2 formation. Green rust is the precursor from which the oxidation of Fe2+ ions and formation of brown-black rust characteristic of g-FeO(OH) take place. In the first g-FeO(OH) formation pathway, Fe(OH)2(s) produced in reaction (R11) would immediately be oxidised precipitating the highly insoluble g-FeO(OH) as the time progresses: Fe(OH)2(aq) + 1/4O2(aq) $ g-FeO(OH)
(s) + 1/2H2O
(R12)
An overall reaction is obtained by combining R11 and R12 resulting in the g-FeO(OH) (s) precipitation: Fe2+
(aq) + 2OH (aq) + 1/4O2(aq) $
g-FeO(OH) (s) + 1/2H2O
(R13)
The second mechanism for g-FeO(OH) formation via air oxidation of highly alkaline Fe(OH)2 suspension in the B2 aqueous phase would involve (R11) and a reaction sequence as follows: (R14) Fe(OH)2(aq) + 1/4O2(aq) + 1/2H2O $ Fe(OH)3(aq) Fe(OH)3(aq) produced in reaction R14 would immediately react with OH resulting in the Fe(OH)4(aq) formation and subsequent dissociation into g-FeO(OH): (R15) Fe(OH)3(aq) + OH(aq) $ Fe(OH)4(aq) $ g-FeO(OH) + H2O + OH An overall reaction is obtained by combining (R11), (R14), and (R15) resulting in the g-FeO(OH) (s) precipitation as the time progresses: Fe2+
(aq) + 2OH (aq) + 1/4O2(aq) $
g-FeO(OH) + 1/2H2O
(R16)
The third mechanism for g-FeO(OH) formation via air oxidation of highly alkaline Fe(OH)2 suspension in the B2 aqueous phase, would involve (R11) and a reaction sequence as follows: Fe(OH)2(aq) + OH $ Fe(OH)3(aq)
(R17)
Fe(OH)3 (aq)
produced in reaction (R17) would immediately be oxidised resulting in the Fe(OH)3(aq): Fe(OH)3(aq) + 1/4O2(aq) + 1/2H2O $ Fe(OH)3(aq) + OH which dissociates into g-FeO(OH) Fe(OH)3(aq) $ g-FeO(OH)
(R18)
(s):
(s) + H2O
(R19)
An overall reaction is obtained by combining (R11), (R17), (R18) and (R19) resulting in the g-FeO(OH) precipitation as the time progresses: Fe(OH)2(aq) + 1/4O2(aq) + OH(aq) $ g-FeO(OH) + 1/2H2O + OH
(R20)
As it can be observed above, (R13), (R16) and (R20) follow different pathways for the g-FeO(OH) formation involving the occurrence of different Fe-aqueous complexes but all of them end in the same final reaction for g-FeO(OH). The thermodynamic reliability of the proposed pathways for g-FeO(OH) formation according to (R13), (R16), and (R20) can be demonstrated by comparing the actual IAP obtained from the experimental data from the B2 aqueous phase and the equilibrium constant (K) of the g-FeO(OH) formation calculated from the Gibbs free energy (DG ) according to (R13), (R16), and (R20). The IAP is calculated as: IAPR13,R16,R20 = [a (g-FeO(OH)) x a(H2O) 1 (O2) /4]
1/2
2+
3.2.2.3. Aqueous speciation in Fe-free brines. The geochemical modelling also indicates a different pH pattern between the modelled and the experimental pH for B3. Fig. 1f shows that the modelled pH increases when adding KOH whereas the experimental pH (Fig. 1e) for B3 decreases and reaches the equilibrium with the B3 solution at 7.52. Given that the geochemical modelling predicts the undersaturation of Mg(OH)2 prior to the KOH addition and the oversaturation of this specie when adding KOH (Fig. 5), the modelled pH trend for B3 would be consistent with the raise in the B3 pH when adding KOH and as a consequence, the formation of the highly insoluble Mg(OH)2. Experimentally, when adding KOH (0.08 mmoles), the B3 pH experiences an increase and reaches the equilibrium with the B3 solution at 7.52. The experimental pH pattern is also consistent with the formation of the highly insoluble Mg(OH)2 since the precipitation of this specie would potentially transfer OH from the aqueous to the B1 solid phase, thereby decreasing the pH value to the equilibrium. The non-detection of Mg(OH)2 in the B3 solid phase by XRD can be due to the insufficient required concentration to be detected by XRD and/or an formation of amorphous Mg(OH)2 phase.
2
]/[a(Fe ) x a(OH ) x a (R21)
Mg(OH)2 SI
1 0 0 SI = log (AIP/K)
Fe2+
where a(FeO(OH)) is assumed to be the unity, a(OH) and a(Fe2+) are calculated from their solute concentrations and the extended Debye-Huckel model (PHREEQc), and a(O2) is equated with its partial atmospheric pressure (Po2 ). The IAP calculated for the B2 is 101.32 below 1054.68, which is the equilibrium constant of the g-FeO(OH) formation calculated at 25 C. The IAPR13,R16,R20
0.01
0.02
0.03
0.04
0.05
0.06
-1 -2 -3 -4 -5 -6
mmoles KOH Fig. 5. Mg(OH)2 SI variation vs KOH.
0.07
0.08
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
4. Conclusions The work reported here reveals that Fe speciation in oil-field brines would be a key factor to consider for above-ground mineral carbonation processes that make use of the aforementioned brines. The pH of brines containing Fe is affected when it is adjusted by a strong base whereas the pH of Fe-free brines experiment shows no significant variations. The pH of brines containing Fe2+ diminishes to very low pH (2.86) by air oxidation of highly alkaline Fe(OH)2 suspension induced by KOH addition by means of dissolution-oxidationcrystallisation mechanism where particles of Fe(OH)2 provide the substrate for the g-FeO(OH) growth. The pH pattern of brines containing Fe3+ is consistent with the formation of the highly insoluble Fe(OH)3 as a primary product due to the pH rise and subsequent formation of Fe2O3 by aging of Fe (OH)3nH2O under alkaline conditions. Results from this study, which determine the importance of Fe speciation in oil-field brine pH, are then of high value when evaluating the potential capacity of oil-field brines (above-ground brines) to permanently sequestering CO2. Acknowledgments The financial support of the Centre for Innovation in Carbon Capture and Storage (CICCS) through the Engineering and Physical Sciences Research Council, EPSRC (EP/F012098/1 and EP/F012098/ 2) is gratefully acknowledged. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcou.2016.06.004. References [1] United State Environmental Protection Agency. Accessed September 2015. http://water.epa.gov/type/groundwater/uic/wells_sequestration.cfm.
85
[2] R.W. Roach, R.S. Carr, C.L. Howard, B.W. Cain, An Assessment of Produced Water Impacts at Two Sites in the Galveston Bay System, U.S, Fish and Wildlife Service Houston, Texas, 1993. [3] Y. Soong, A.L. Goodman, J.R. McCarthy-Jones, J.P. Baltrus, Experimental and simulation studies on mineral trapping of CO2 with brine, Energy Convers. Manage. 45 (11–12) (2006) p. 1845–1859.10. [4] M.L. Druckenmiller, M.M. Maroto-Valer, Carbon sequestration using brine of adjusted pH to form mineral carbonates, Fuel Process. Technol. 86 (2005) 1599–1614. [5] Y.K. Kharaka, L.Y. Leong, C.G. Doran, G.N. Breit, Can produced water be reclaimed, experience with placeritra oil field, california, Proceedings of the 5th International Petrol Environment Conference, Albuquerque, NM, 1998, pp. 1–23. [6] Y. Soong, A.L. Goodman, J.R. Jones, J.R. Baltrus, Experimental and simulation studies on mineral trapping of CO2 with brine, Energy Convers. Manage. 45 (2004) 1845–1859. [7] F.A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, 5th ed., John Wiley & Sons, New York, Chichester, Brisbane, Toronto, Singapore, 1988, pp. p.71. [8] A.J. Bard, R. Parsons, J. Jordan, Standard Potential in Aqueous Solutions, Marcel Dekker, Inc, 270 Madison Avenue, New York, 2016, pp. 10016. [9] B. Morgan, O. Lahav, The effect of pH on the kinetics of spontaneous Fe(II) oxidation by O2 in aqueous solution basic principles and a simple heuristic description, Chemosphere 68 (2007) 2080–2084. [10] W. Feitknecht, W. Michaelis, Ober die hydrolyse yon eisen(III)-perchloratL6seungen, Helv. Chim. Acta 26 (1962) 212–224. [11] W. Feitknecht, R. Giovanoli, W. Michaelis, M. Muller, Die hydrolyse der LOsungen yon eisen (III)-chlorid, Helv. Chirn. Acta 56 (1973) 2847–2856. [12] M. Magini, R. Caminiti, On the structure of highly concentrated iron (III) salt solutions, J. Inorg. Nucl. Chem 39 (1977) 91–94. [13] Kobayashi, Uda, Structure of ferric hydroxide gel, J. Non-Cryst. Solids 29 (3) (1978) 419–422. [14] I.P. Saraswat, A.C. Va’pei, V.K. Garg, V.K. Sharma, N. Prakash, On the sorption studies of some metal ions with chromium ferrocyanide gel, J. Colloid Interface Scz. 73 (1980) 373. [15] A. Feitkneeht Wand Wytlenbach, Reactivity of soils, 4th Int. Symp. Ed, Elsevier, Hol., 1960, pp. p234. [16] A. Kasllaki, Osaka Kogyo Gijut. 30 (1979) 161. [17] R.M. Comell, R. Giovanoli, P.W. Sehindlcr, Glays and Clay minerals 35 (1987) 21. [18] T. Misawa, K. Hashimoto, S. Shimodaira, The mechanism of formation of iron oxide and oxyhydroxides in aqueous solutions at room temperature, Corros. Sci. 14 (1974) 131–149. [19] U. Schwertman, H. Fetcher, The formation of green rust and its transformation to Lepidocrocite, Clay Miner. 29 (1994) 87–92. [20] P.C. Singer, W. Stumm, Acidic mine drainage; rate-determining step, Science 167 (1970) 1121–1123. [21] J.D. Bernal, D.R. Dasgupta, A.L. Mackay, The oxides and oxyhydroxides of iron and their structural inter-relationships, Clay Miner. Bull. 4 (1959) 15–30. [22] O.W. Lau, The precipitation of ferrous hydroxide: a lecture demonstration, J. Chem. Educ. 56 (1979) 474.
Contents lists available at ScienceDirect
Journal of CO2 Utilization journal homepage: www.elsevier.com/locate/jcou
Review article
Understanding the importance of iron speciation in oil-field brine pH for CO2 mineral sequestration Patricia Córdobaa,* , Qi Liua , Susana Garciaa , Mercedes Maroto-Valera,b a Centre for Innovation in Carbon Capture and Storage (CICCS), Institute of Mechanical, Process and Energy Engineering (IMPEE), Heriot-Watt University, EH14 4AS, United Kingdom b The Institute of Petroleum Engineering (IPE), Heriot-Watt University, United Kingdom
A R T I C L E I N F O
Article history: Received 9 October 2015 Received in revised form 22 May 2016 Accepted 11 June 2016 Available online 12 July 2016 Keywords: CO2 Mineralogical sequestration Oil-field brines Above-ground brines Modelling tools
A B S T R A C T
Carbon dioxide (CO2) mineralogical sequestration using oil-field brines (above-ground field brines) is gaining attention for the storage of CO2 into stable mineral carbonates. This involves complex processes that are affected by brine composition, pressure, temperature and pH where the latter is the key parameter, as carbonation conversion rates and the specific carbonates species formed are strongly dependent on pH. This study investigates the effect of iron (Fe) in the pH of three synthetic brines prepared as analogue to oil-field brine by combining a number of experimental and modelling tools to elucidate whether or not Fe speciation may affect the pH of brines. X-Ray powder Diffraction (XRD) analyses were also carried out to characterise the mineralogy of the synthetic brines. Results indicate that the Fe speciation affects the pH of oil-field brines; brines containing Fe3+ experiment a rapid increase of the pH as a consequence of Fe2O3 precipitation by aging of Fe(OH)3 under alkaline conditions. Brines containing Fe2+ show a sharp drop of the pH as a result of air oxidation of highly alkaline Fe(OH)2 suspension induced by KOH addition by means of dissolution-oxidation-crystallisation mechanism where particles of Fe(OH)2 provide the substrate for the g-FeO(OH) growth. The master variable governing the rates at which these compounds form is pH. ã 2016 Elsevier Ltd. All rights reserved.
Contents 1. 2.
3.
4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Brine preparation . . . . . . . . . . . . . . . . . . . . . . . . 2.1. pH stability studies . . . . . . . . . . . . . . . . . . . . . . . 2.2. Mineralogical composition of solid phases . . . . 2.3. Geochemical modelling . . . . . . . . . . . . . . . . . . . 2.4. Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . pH studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Characterisation of brines . . . . . . . . . . . . . . . . . 3.2. Identification of solid phases in brines 3.2.1. Aqueous speciation of brines . . . . . . . . 3.2.2. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
78 79 79 79 80 80 80 80 80 80 80 85 85 85
1. Introduction
* Corresponding author. E-mail address: [email protected] (P. Córdoba). http://dx.doi.org/10.1016/j.jcou.2016.06.004 2212-9820/ã 2016 Elsevier Ltd. All rights reserved.
Carbon dioxide (CO2) mineral sequestration using brines is gaining attention for the storage of CO2 into stable mineral
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
carbonates [1]. Brines are commonly found either underground, namely underground brines in saline aquifers, or above-ground, most notably as a by-product of oil and natural gas extraction and known as oil-field brines, and its fate is mainly disposal [2]. However, oil-field brines (above-ground brines) could also be a promising approach to sequester CO2 as they are produced in large volumes [3,4]. Brines in addition to the dominant Na and Cl ions [5], also have significant concentrations of Ca, Mg and Fe, which react with CO2 to produce CaCO3(s), MgCO3(s), Fe2CO3(s) and other products under favourable conditions. However, most both under and aboveground brines are generally acidic in nature with a typical pH ranging from 3 to 5 [3], so that carbonates will not form in this pH range. The suitable pH range for formation of carbonate is 7.8 or higher [6], where CO32 dominates. Therefore, to boost the precipitation of mineral carbonates by reaction between brine and CO2, the pH of the brine must be modified e.g. by caustic addition before it can be used for the carbonation reaction [3]. In this regard, a factor that should be considered as the first step to boost the precipitation of mineral carbonates by the increase of the pH in above-ground brines is Fe speciation. The hydrolysis of ferrous (Fe2+) and ferric (Fe3+) in aqueous solutions involves pH changes because of the formation of a variety of partially oxidized meta-stable Fe2+-Fe3+ aqueous complexes, which could prevent oil-field brines from using for CO2 mineral carbonation. Fe2+ show high solubility in aqueous solutions, whereas Fe3+ shows a stepwise solubility at pH >1.0 [7]. The Fe2+-hexaquo-complex [Fe (H2O) 6]2+, which is present in acid solutions, hydrolyses by the formation of FeOH+ and precipitates as the hydroxides Fe(OH)2 upon increasing the pH value (Bard et al., 1985). Fe3+ exist only in acidic solutions (pH <2.5) as the hexaquo-complex Fe(H2O) 63+, which increasing pH is subsequently transformed to mono- and the dyhidroxo-complexes. In the pH range 5 at least four different Fe3+ ions coexist in aqueous solution: Fe3+; Fe (OH)2+; Fe(OH)+2 and the dimer Fe2(OH)42+ [8]. The spontaneous chemical oxidation of Fe2+ to Fe3+ by O2 is also a complex process involving pH changes with a variety of partially oxidized meta-stable Fe2+-Fe3+ intermediate species, which ultimately transform into a variety of stable Fe-oxide end-products such as hematite (Fe2O3), magnetite (Fe3O4), goethite (a-FeOOH), and lepidocrocite (g-FeOOH) [9]. These crystalline products form by competing mechanisms and the proportion of each in the final product depends on the relative rates of formation. The master variable governing the rates at which these compounds form is pH. Other important factors are temperature and the presence of additives. Although the conditions governing Fe chemistry under different environments have been identified by the above mentioned researchers among others, the role of Fe speciation in aboveground brines aimed at mineral sequestration of CO2 and how it can affect brine composition and pH has not been elucidated yet. Druckenmiller and Maroto-Valer [4] reported that brines with low Fe concentration (9 ppm) showed a pH relatively constant after addition of a strong base. By contrast, brines with high Fe concentration (121–476 ppm) had a rapid decline in pH during the first hours, after which the decline seemed to level off. However, it was not ascertained whether Fe2+ or Fe3+ caused the pH drop in their study. With the advent of making CO2 sequestration through mineral carbonate formation a viable approach by using oil-field brines, the role of Fe speciation in oil-field brines pH needs to be elucidated to reach suitable pH range for further formation of carbonate. Towards this goal, this work i) studies the evolution of the pH in three synthetic brines with different composition prepared as analogues to an oil-field brine; ii) investigates the effect of Fe in the pH of the three above-ground brines when adjusting the pH; and iii) elucidates the role of Fe speciation in brines when adjusting the
79
pH by combining a number of experimental tools and a geochemical model based on acquired experimental data. To reproduce experimental conditions when attempting to use oilfield brine for carbon sequestration, ambient temperature and ambient pressure conditions were chosen for pH studies and modelling calculations. 2. Methodology 2.1. Brine preparation Three different synthetic brines, namely B1, B2 and B3, were prepared as an analogue to the natural brine (OH-2) which comes from a natural gas well in Youngstown, Ohio [4]. Owing to the complex composition of natural brines, only major ions were considered to prepare the three synthetic brines, including Na+, K+, Mg2+, Ca2+, Fe3+/Fe2+, Sr2+, Ba2+and Cl. Brines with their corresponding duplicates were synthesised by dissolving NaCl, KCl, MgCl26H2O, CaCl22H2O, FeCl3/FeCl2, SrCl2 and BaCl22H2O (Sigma-Aldrich) in Milli-Q water. Solutions were mixed in a beaker thoroughly with magnetic stirrers until a complete dissolution was attained. After this procedure, 20 mL of each B1, B2, and B3 was filtered by glass-fibre filter of 0.2 mm pore size and acidified with HNO3 to prevent oxidation of ions and stabilize the metal concentration. Brine aqueous phases were then analysed by Inductively Coupled Plasma Mass Spectrometry (ICP-MS) by the X-SERIES II device from Thermo Fisher SCIENTIFIC to determine the concentration of the aforementioned major ions in order to compare them with target values. The quantitative analyses were performed with an external calibration, using an external standard of similar matrix to the samples, which covered concentrations range expected forming the calibration lines. The internal correction was carried out by means of an internal standard (In 10 ppb). Brine compositions only differ in the type of Fe present in the aqueous phase of brines, with Brine B1, B2, and B3, containing Fe3+, Fe2+, and non-Fe, respectively. 2.2. pH stability studies The pH evolution was studied at ambient temperature and pressure in open atmospheres using B1, B2, and B3 over a period of 8-14 days. According to the previous study [4], it was preliminarily determined to raise the pH from the initial values to an upper limit below the point at which Fe hydroxide precipitates. According to a Fe-H2O system, in a solution containing Fe2+, iron hydroxide has a much lower solubility product than the hydroxyl forms of the other abundant cationic species [7]. Because the analogue brine selected in this study showed a high Fe composition from which a pH of 6.8 would promote precipitation, a pH value of approximately 6.3 was selected for the pH stability of B1 and B2, and thus providing an upper limit well below 6.8. For B3 a pH value of 9.0 was selected as the target as the initial pH of B3 was higher than 6.3. The precipitation of mineral carbonates is mostly dependent on brine pH and it is favoured above a basic pH of 9.0. In each pH stability experiment, 100 mL of fresh synthetic brine was poured into a 250 mL conical flask. The brine solution was mixed effectively and continuously throughout the experiment by using a magnetic stirrer. The pH variation with time was measured by a Thermo Orion 420A+ bench top pH meter. The brine pH was raised to above 6.3 or 9.0 initially by adding 1.0 M KOH solution. After the brine pHs were raised to the desired values (6.3 or 9.0), addition of KOH was stopped and time set as zero. pH measurements were then made every 5–10 min for the first 2 h, then every 1 h. After the first 12 h, pH was measured every 12 h. When the
80
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
brine pH was stable around 0.2 continuously for 3 days, the stability study was then completed. The brine solutions were then filtrated by using cellulose nitrate membrane filters (0.45 mm pore size) and the solid residue was saved for further XRD analysis. The error for pH measurement was 0.02 owing to the calibration of pH probe (every 2 h for the first 12 h and thereafter every 24 h) and the variation of laboratory room temperature. 2.3. Mineralogical composition of solid phases X-Ray powder Diffraction (XRD) analyses were carried out to determine the mineralogical composition and solid species of brines after KOH addition. Analyses were carried out by using a HILTONBROOKS diffractometer with monochromatic Cu Ka1,2 radiation operated at 40KV and 20 mA, from 460 2u, at a step size of 0.05 , and scan rate of 3s/step. The classification of ground soil particles was carried out by a Coulter Laser Scattering analyser employing Fraunhofer diffraction (LS 200, Variable-Speed Fluid Module Plus (VSM + )). 2.4. Geochemical modelling Theoretical calculations with the PHREEQC code (version 2.0) and the coupled thermodynamic database Lawrence Livermore National Laboratory (LLNL) are used herein to calculate the 1) aqueous speciation of the synthetic brines and saturation index (SI) of Fe minerals with the progressive addition of KOH; and 2) pH evolution of synthetic brines with the progressive addition of KOH. As a result of the different speciation of Fe in the synthetic brines, the kinetic oxidation of Fe in B2 with the progressive addition of KOH on the pH stabilisation over 8 days was also modelled. The modelling calculations were conducted by using experimental pressure and temperature conditions as well as the chemical composition and pH values of the three brines under which the process takes place as input data. Detailed description of the model is provided as supporting information. The theoretical results from the model were then compared to experimentally derived results on the same system. 3. Results and discussion 3.1. pH studies The pH values of brines and the experimental parameters in the pH stability studies are given in Table 1. The results of pH studies for B1 over a period of 192 h (8 days) are shown in Fig. 1a. B1 duplicates, namely B1A and B1B, containing Fe3+ show an initial pH of 1.68 and 1.70, respectively (Table 1). The B1A and B1 B pHs were therefore adjusted to 6.3, an upper pH limit but below the point at which Fe(OH)2 precipitates (7.0–9.0.), by the addition of KOH. The evolution of the pH of the B1 against time shows that the pH fluctuates over the first 48 h, after which it remains (0.2 pH) stable till the end of the experiment. Over a run period of 8 days, B1A and B1 B reach pH values of 6.41 and 6.36, respectively (Table 1). Table 1 Summary of pH stability studies. Brine type
Initial pH
Added KOH (mmoles)
Adjusted pH
Final pH
B1A B1B B2A B2B B3A B3B
1.70 1.68 3.43 3.45 6.58 6.59
3.27 3.28 0.26 0.26 0.08 0.08
6.33 6.32 6.35 6.36 9.08 9.10
6.41 6.36 2.86 2.86 7.57 7.52
The evolution of pH for B2 over a period of 192 h (8 days) is shown in Fig. 1c. B2 duplicates, namely B2A and B2B, containing Fe2+ shows an initial pH of 3.44 and 3.45, respectively (Table 1). The B2A and B2 B pHs were also adjusted to an upper pH limit but below the point at which Fe(OH)2 precipitates(7.0–9.0.) Accordingly, B2A and B2 B pHs were adjusted to 6.3 by KOH addition. Results show that B2 experiences a sharp drop of the pH over the first 24 h with a gradual decrease until it stabilises (0.2 pH) after 192 h. Over a run period of 8 days, B2A and B2 B reach pH values of 2.86. The evolution of pH for B3 over a period of 336 h (14 days) is shown in Fig. 1e. B3 and a duplicate, namely B3A and B3B, containing no Fe shows an initial pH of 6.58-6.59, respectively (Table 1). Owing to the absence of Fe in this brine, B3A and B3 B pHs were adjusted to 9.0. Results show that B3 experiences a gradual drop of the pH along the experiment until it stabilises after 336 h. Over a run period of 14 days, B3A and B3 B reach pH values of 7.57 and 7.52, respectively (Table 1). The different pH patterns between B1 and B2 indicate that the pH of brines is highly affected by the Fe speciation. However, it should also be noted that B3 reaches pH stability (0.2 pH) after a longer time period when compared with B1 and B2, which is also indicative that the absence of Fe affects the pH stability of brines. 3.2. Characterisation of brines The composition of the natural brine (OH-2) and prepared synthetic brines is shown in Table 2. 3.2.1. Identification of solid phases in brines The XRD analysis revealed that NaCl and Fe2O3 are the crystalline phases detected in B1 (Fig. 2a). It should be noted that B1 showed a low crystallinity and small crystalline sizes, which caused most diffraction patterns to have poor signal to noise ratios and very broad peaks. Crystalline phases detected in B2 (Fig. 2b) are g-FeO(OH) and NaCl whereas NaCl is the only crystalline phase detected in B3 (Fig. 2c). The identification of g-FeO(OH) in the B2 solid phase suggests an oxidation of Fe2+ to Fe3+ in the B2 aqueous phase, where Fe2+ would be oxidised and hydrolysed to g-FeO(OH) when adjusting the B2 aqueous phase pH via KOH addition. Therefore, it can be stated that the different mineralogical composition of brines following pH adjustment is caused by the different Fe speciation in their corresponding aqueous phases. 3.2.2. Aqueous speciation of brines Geochemical modelling reveals, with the exception of Fe, no significant differences in the B1 (pH 1.68-1.70), B2 (pH 3.43-3.45), and B3 (pH 6.58-6.59) aqueous phases despite the KOH addition (Table 3). Ca and Na as free cations (Xn,n+1,+2) are the predominant aqueous complexes in all the brines followed by K, Mg, and Sr. 3.2.2.1. Fe3+ effect on aqueous speciation. The geochemical modelling indicates that the addition of KOH to B1 produces a slight increase of pH from 1.68 to 1.74 (Fig. 1b). However, experimentally, B1 pH initially fluctuates, but it remains stable (0.2 pH) at the end of the experiment at 6.41 pH (Fig. 1a). The differences between the B1 experimental and modelling pH pattern could be related to either geochemical modelling accounting for equilibrium reactions involving Fe, and/or the pH stability of Fe species non-detected by XRD. According to the geochemical modelling, prior to the KOH addition, the B1 aqueous phase is oversaturated (SI = IAP/K >0) with respect to Fe2O3 (SI value of 3.28) and a-FeOOH (SI 1.13) and Fe3+ as free cation followed by FeOH2+, Fe(OH)2+, Fe3(OH)24+, Fe(OH)45+, and Fe(OH)3 are the Fe-aqueous complexes with highest
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
81
Fig. 1. a) B1 pH over time; b) B1 pH modelled vs KOH; c) B2 pH over time; d) B2 pH modelled vs KOH; e) B3 pH studies over time; f) B3 pH modelled vs KOH.
Table 2 Target concentrations of major ions in OH-2 and B1, B2, and B3.
mg/L Ca2+ K+ Na+ Mg2+ Sr2+ Fe3+ Fe2+
Brine (OH-2)
B1
Variation (%)
B2
Variation (%)
B3
Variation (%)
19350 1180 48500 2055 1800 476 476
20010 1243 50215 1887 1790 451 –
3.4 5.3 3.5 8.2 0.6 5. –
19490 1143 48185 1838 1765 – 452
0.7 3.1 0.6 10.5 1.9 – 5.1
19235 1204 50015 1871 1737 – –
0.6 2.0 3.1 8.9 3.5 – –
activities at the initial pH of 1.69. The occurrence of the Fe aqueous complexes in the B1 aqueous phase promotes the acidification of the aqueous phase according to the reactions: Fe3+
2+ + (aq) + H2O $ FeOH (aq) + H
Fe3+
(aq) + 2H2O $ Fe
(OH)2+
(aq) + 2H
(R1)
+
(R2)
2Fe3+
+ 4+ (aq) + 2H2O $ Fe2(OH)2 (aq) + 2H
(R3)
3Fe3+
+ 5+ (aq) + 4H2O $ Fe3(OH)4 (aq) + 4H
(R4)
Fe3+
(aq) + 3H2O $ Fe
(OH)3(aq) + 3H+
(R5)
where Fe-aqueous complexes from (R1) to (R5) remain in the B1 aqueous phase under acidic conditions. When modelling the KOH addition, the geochemical modelling predicts, in addition to the occurrence of the above mentioned Feaqueous complexes, the occurrence of Fe2+, FeOH+, Fe(OH)2, and Fe (OH)3 and a pH (Fig. 1b) showing a different pattern to that of the experimental pH (Fig. 1a). The geochemical modelling also predicts the oversaturation of the B1 aqueous phase with respect to Fe2O3 (SI of 7.77) and a-FeOOH (SI of 2.87). The modelled pH pattern and the occurrence of Fe2+, FeOH+, Fe(OH)2, and Fe (OH)3 complexes would be consistent with the reduction of Fe3+ and Fe3+-aqueous complexes when adding KOH. Therefore, it is postulated that Fe3+ as a free cation, whose activity decreases after KOH addition, would be reduced to Fe2+ whose activity is the highest among
82
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
the release of the H+ by the formation of Fe3+ aqueous complexes ((R1)–(R5)); 2) the reduction of Fe3+ to Fe2+ by OH from KOH (R6); and 3) the formation of FeOH+ and Fe(OH)2 complexes by Fe2+solvation, which acidifies the B1 aqueous phase ((R7) and (R8)). By contrast, the experimental pH pattern of B1 (Fig. 1a) would be consistent with the precipitation of Fe3+ in the B1 aqueous phase as Fe(OH)3 via KOH addition according to:
B1
a) 800
1- Fe2O3
700
1
2- NaCl
600 Counts
500 1
400
1 2
300
Fe3+
100
Because Fe(OH)3 is not detected by XRD, it is postulated that Fe (OH)3(s) could have turned into a-Fe2O3(s) by aging, which would be in agreement with detection of Fe2O3(s) in the B1 solid phase by XRD (Fig. 2b) and with the alkaline experimental conditions under which the experiments were carried out. However, it should be noted that this does not rule out the occurrence of other crystalline or solid phases which concentration is not high enough (1%) to be detected by XRD and/or a formation of amorphous Fe solid phase, which cannot be detected by XRD. Several investigations about the constitution and aging of precipitated hydrous Fe3+oxide gels [10–14] obtained amorphous non-stoichiometry hydroxides by addition of bases to Fe3+ solutions which were not detected by XRD. This finding is in line with previous works on Fe speciation and precipitation processes. Feitknecht and Wytlenbach [15] reported the occurrence of FeOOH and Fe2O3 as a the result of the Fe (OH)3nH2O aging by agitation with KOH whereas Kasuaki [16] formed a-FeOOH particles under an alkaline environment (NaOH) at 200 C for 5 h. Cornell et al. [17] reported that alkaline solutions promote the change of Fe2O3.1/2 H2O and 5Fe2O3 9H2O into FeOOH and Fe2O3(s). Therefore, the experimental pH pattern in the B1 aqueous phase might be due to the formation of Fe2O3(s) by aging of Fe(OH)3(s). Hence, the Fe3+ effect on B1 aqueous speciation and pH lies in the role that KOH could play in R6 (modelling) and R9 (experimental), respectively. The former would demonstrate the feasibility of KOH to reduce Fe3+ to Fe2+ since the estimation of the DG of formation according to R6 or chemical potential translated in terms of redox potential infers the fact that the reduction of Fe3+ in the B1 aqueous phase is thermodynamically possible. The latter demonstrates the feasibility of KOH to react with Fe3+ to form Fe (OH)3 from which the Fe2O3(s) by aging can be formed, as observed experimentally.
0 0
10
20
30
40
50
60
2θ (degrees)
B2
b) 6000
1- FeO(OH) 2- NaCl
2
5000
Counts
4000
1 3000 2000 1000
2
1
1
1
1 1
1
0
0
10
20
30
40
50
60
2θ (degrees)
B3
c) 25000
1- NaCl 20000
Counts
200
1
15000 10000 5000
1 1
0 0
10
20
30
40
50
60
2θ (degrees) Fig. 2. a) Mineral and solid phases in B1; b) Mineral and solid phases in B2; c) Mineral and solid phases in B3.
Fe2+aqueous complexes in the B1 aqueous phase after KOH addition (Table 3) according to: 4Fe3+ + 4OH $ 4Fe2+
(aq) + 2H2O + O2(aq)
2+
(R6) +
Fe from (R6) would be solvated forming FeOH and Fe(OH)2 aqueous complexes, respectively, acidifying once again the B1 aqueous phase according to: Fe2+
+ + (aq) + H2O $ FeOH + H
Fe2+
(aq) + H2O $ Fe(OH)2 + 2H
(R7)
+
(R8)
The viability of (R7) and (R8) depends on 1) the thermodynamic feasibility of R6, where KOH would reduce Fe3+ to Fe2+ since the estimation of the Gibbs standard free energy of formation (DG ) or chemical potential translated in terms of redox potential infers the fact that the reduction of Fe3+ in the B1 aqueous phase is thermodynamically possible, and on 2) the kinetics of (R6). Therefore, the slight raise in the modelled B1 pH despite the KOH addition would be the result of a number of successive reactions: 1)
(aq) + 3OH
$ Fe(OH)3(s)
(R9)
3.2.2.2. Fe2+ effect on aqueous speciation. The geochemical modelling also indicates a different pH pattern between the initial modelled and the experimental pH for B2. In the initial model, the pH slightly increases at the beginning of the experiment (Fig. 1d) as opposed to a decreasing pH trend in the experiment after the KOH addition as time progresses (Fig. 1c). Because g-FeO(OH) (g-Fe3+O(OH)) is the only Fe crystalline phase detected in the B2 solid phase (Fig. 2b), the oxidation of Fe2+ to Fe3+ was identified as the cause of the experimental B2 pH value and the different pattern between the experimental and modelled pH values in the B2 aqueous phase (Fig. 1c and d). The experimental pH would be consistent with a rapid kinetic oxidation of Fe2+ to Fe3+ that would give rise to the formation of hydroxy-complexes of Fe3+ that progressively lower the B2 pH (2.86). By contrast, the modelled rise in pH would be consistent with a slowly oxidation of Fe2+ to Fe3+ by forming hydroxycomplexes of Fe3+ resulting in a slight increase of the B2 pH. Based on Fe-hydroxide compounds occurring under reducing and weakly acid to weakly alkaline conditions as intermediate phases in the formation of Fe oxides such as a-FeO(OH), g-FeO (OH), Fe3O4 [17–19], the oxidation of Fe2+ to Fe3+ and its
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
83
Table 3 Aqueous speciation of B1, B2, and B3. Before KOH addition
Ca Fe (2)
Fe (3)
Na K Mg Sr
Initial Eh (mV) Activity (mol/L) Ca2+ CaOH+ Fe2+ FeOH+ Fe(OH)2 Fe(OH)3 Fe (OH)42 Fe(OH)4 Fe3+ Fe(OH)3 Fe2(OH)24+ FeOH2+ Fe3(OH)45+ Fe(OH)2+ Na+ NaOH K+ KOH Mg2+ Mg4(OH)44+ Sr2+ SrOH+
After KOH addition
B1
B2
B3
B1
B2
B3
4.0
1.58
2.03
18.868
17.525
11.337
1.119e-01 7.351e-13 – – – – – 4.880e-19 4.159e-04 4.179e-11 4.195e-07 1.248e-04 1.685e-10 1.922e-06 1.575 1.217e-13 2.100e-02 3.385e-15 2.505e-02 3.273e-40 3.528e-03 8.412e-15
1.089e-01 4.025e-11 1.812e-03 1.500e-09 3.117e-17 3.248e-24 – – – – – – – – 1.557 6.773e-12 1.923e-02 1.745e-13 2.431e-02 2.913e-33 3.481e-03 4.672e-13
1.076e-01 5.483e-08 – – – –
1.119e-01 8.45313 5.53510 9.357e-18 3.973e-27 8.457e-36 – 8.050e-19 3.923e-04 5.994e-11 4.936e-07 1.354e-04 2.472e-10 2.093e-04 1.636 2.535e-10 2.300e-02 4.264e-15 2.505e-02 5.722e-40 3.528e-03 9.673e-15
1.089e-01 1.816e-11 5.276e-15 2.988e-10 4.249e-07 3.029e-03 5.424e-04 6.421e-06 – 1.428e-10 – 2.875e-29 – 1.705e-18 1.558 3.058e-12 1.942e-02 7.949e-14 2.433e-02 1.210e-34 3.480e-03 2.107e-13
1.081e-01 2.775e-05 – – – –
subsequent transformation to g-FeO(OH) through oxygen dissolution in a highly alkaline B2 aqueous phase is proposed. Therefore, the 1) aqueous speciation and 2) pH pattern for B2 considering the kinetic oxidation of Fe2+ to Fe3+ with the progressive addition of KOH over 8 days were modelled. The oxidation of Fe2+ to Fe3+ through O2 dissolution in a highly alkaline B2 aqueous phase, modelled using B2 aqueous phase chemical composition as input data, was based on the rate of oxidation of Fe2+ by O2 in water given by Singer and Stumm [20]:
– – – – – – – 1.620 9.713e-09 1.989e-02 2.489e-10 2.479e-02 1.138e-20 3.468e-03 6.417e-10
– – – – – – 1.567 1.991e-02 1.255e-07 2.718e-02 1.060e-09 3.550e-03 3.309e-07
where t is time in seconds, aOH is the activity of the hydroxyl ion, aFe2þ is the total molality of Fe2+ iron in the B2 aqueous phase, and Po2 is the oxygen partial pressure (atm). Detailed description of the model here is provided as supporting information. The kinetic modelling consisting of the B2 aqueous speciation over the 8 days of simulation (Fig. 3) shows that the molality of Fe3+ against time increases whereas the molality of Fe2+ against time is initially constant, but lessens as the reaction progresses, which is consistent with Eq 1. The kinetic model consisting of the B2 pH over the 8 days of simulation shows, as expected, a decrease of the pH
value (Fig. 4). As can be observed in the pH model, the initial pH of the B2 (6.35) experiences a sharp drop over the first day of simulation and a soft decrease between the 3rd and 8th day of simulation (Fig. 4). The modelled pH reaches the stability with the B2 solution at 3.10, which is nearly in line with the experimental B2 pH (2.86). The time for complete oxidation of Fe2+ is a matter of minutes in an aerated solution when pH is 7.0. However, Fe3+ forms aqueous complexes with OH precipitating as Fe oxy-hydroxides, so that pH decreases during oxidation. Because the rate has quadratic dependence on the activity of OH, the oxidation rate in the B2 aqueous phase rapidly increases. The modelling results presented above along with the experimental analyses allow us to propose three pathways for g-FeO(OH) formation via air oxidation of a highly alkaline Fe(OH)2 suspension in the B2 aqueous phase by means of dissolutionoxidation-crystallisation mechanism, in which the initial particles of Fe(OH)2 would provide the substrate for the g-FeO(OH) growth. These formation pathways are based on the occurrence of Fe3+-complexes showing the highest activities in the B2 aqueous phase when modelling KOH addition (Table 3). Accordingly, it is proposed that in line with the B2 composition, when Fe2+ is immersed in the B2 aqueous phase Fe(OH)2 is formed as the
Fig. 3. Kinetic oxidation of Fe aqueous complexes over 8 days of simulation.
Fig. 4. Kinetic variation of the B2 pH over 8 days of simulation.
dmFe2þ ¼ 2:92e9 þ 1:33e12 a2OH PO2 mFe2þ dt
ð1Þ
84
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
primary product under the alkaline conditions by KOH addition in a two-step formation process: + (aq) + 2OH (aq) $ Fe(OH) (aq) + OH (aq) $ Fe(OH)2
(R11)
2+
As the R11 proceeds, dissolved Fe is consumed and the solution becomes saturated with respect to Fe (OH) 2 involving the formation of green rust. According to previous studies [21,22], this phenomenon is characteristic of Fe(OH)2 formation. Green rust is the precursor from which the oxidation of Fe2+ ions and formation of brown-black rust characteristic of g-FeO(OH) take place. In the first g-FeO(OH) formation pathway, Fe(OH)2(s) produced in reaction (R11) would immediately be oxidised precipitating the highly insoluble g-FeO(OH) as the time progresses: Fe(OH)2(aq) + 1/4O2(aq) $ g-FeO(OH)
(s) + 1/2H2O
(R12)
An overall reaction is obtained by combining R11 and R12 resulting in the g-FeO(OH) (s) precipitation: Fe2+
(aq) + 2OH (aq) + 1/4O2(aq) $
g-FeO(OH) (s) + 1/2H2O
(R13)
The second mechanism for g-FeO(OH) formation via air oxidation of highly alkaline Fe(OH)2 suspension in the B2 aqueous phase would involve (R11) and a reaction sequence as follows: (R14) Fe(OH)2(aq) + 1/4O2(aq) + 1/2H2O $ Fe(OH)3(aq) Fe(OH)3(aq) produced in reaction R14 would immediately react with OH resulting in the Fe(OH)4(aq) formation and subsequent dissociation into g-FeO(OH): (R15) Fe(OH)3(aq) + OH(aq) $ Fe(OH)4(aq) $ g-FeO(OH) + H2O + OH An overall reaction is obtained by combining (R11), (R14), and (R15) resulting in the g-FeO(OH) (s) precipitation as the time progresses: Fe2+
(aq) + 2OH (aq) + 1/4O2(aq) $
g-FeO(OH) + 1/2H2O
(R16)
The third mechanism for g-FeO(OH) formation via air oxidation of highly alkaline Fe(OH)2 suspension in the B2 aqueous phase, would involve (R11) and a reaction sequence as follows: Fe(OH)2(aq) + OH $ Fe(OH)3(aq)
(R17)
Fe(OH)3 (aq)
produced in reaction (R17) would immediately be oxidised resulting in the Fe(OH)3(aq): Fe(OH)3(aq) + 1/4O2(aq) + 1/2H2O $ Fe(OH)3(aq) + OH which dissociates into g-FeO(OH) Fe(OH)3(aq) $ g-FeO(OH)
(R18)
(s):
(s) + H2O
(R19)
An overall reaction is obtained by combining (R11), (R17), (R18) and (R19) resulting in the g-FeO(OH) precipitation as the time progresses: Fe(OH)2(aq) + 1/4O2(aq) + OH(aq) $ g-FeO(OH) + 1/2H2O + OH
(R20)
As it can be observed above, (R13), (R16) and (R20) follow different pathways for the g-FeO(OH) formation involving the occurrence of different Fe-aqueous complexes but all of them end in the same final reaction for g-FeO(OH). The thermodynamic reliability of the proposed pathways for g-FeO(OH) formation according to (R13), (R16), and (R20) can be demonstrated by comparing the actual IAP obtained from the experimental data from the B2 aqueous phase and the equilibrium constant (K) of the g-FeO(OH) formation calculated from the Gibbs free energy (DG ) according to (R13), (R16), and (R20). The IAP is calculated as: IAPR13,R16,R20 = [a (g-FeO(OH)) x a(H2O) 1 (O2) /4]
1/2
2+
3.2.2.3. Aqueous speciation in Fe-free brines. The geochemical modelling also indicates a different pH pattern between the modelled and the experimental pH for B3. Fig. 1f shows that the modelled pH increases when adding KOH whereas the experimental pH (Fig. 1e) for B3 decreases and reaches the equilibrium with the B3 solution at 7.52. Given that the geochemical modelling predicts the undersaturation of Mg(OH)2 prior to the KOH addition and the oversaturation of this specie when adding KOH (Fig. 5), the modelled pH trend for B3 would be consistent with the raise in the B3 pH when adding KOH and as a consequence, the formation of the highly insoluble Mg(OH)2. Experimentally, when adding KOH (0.08 mmoles), the B3 pH experiences an increase and reaches the equilibrium with the B3 solution at 7.52. The experimental pH pattern is also consistent with the formation of the highly insoluble Mg(OH)2 since the precipitation of this specie would potentially transfer OH from the aqueous to the B1 solid phase, thereby decreasing the pH value to the equilibrium. The non-detection of Mg(OH)2 in the B3 solid phase by XRD can be due to the insufficient required concentration to be detected by XRD and/or an formation of amorphous Mg(OH)2 phase.
2
]/[a(Fe ) x a(OH ) x a (R21)
Mg(OH)2 SI
1 0 0 SI = log (AIP/K)
Fe2+
where a(FeO(OH)) is assumed to be the unity, a(OH) and a(Fe2+) are calculated from their solute concentrations and the extended Debye-Huckel model (PHREEQc), and a(O2) is equated with its partial atmospheric pressure (Po2 ). The IAP calculated for the B2 is 101.32 below 1054.68, which is the equilibrium constant of the g-FeO(OH) formation calculated at 25 C. The IAPR13,R16,R20
0.01
0.02
0.03
0.04
0.05
0.06
-1 -2 -3 -4 -5 -6
mmoles KOH Fig. 5. Mg(OH)2 SI variation vs KOH.
0.07
0.08
P. Córdoba et al. / Journal of CO2 Utilization 16 (2016) 78–85
4. Conclusions The work reported here reveals that Fe speciation in oil-field brines would be a key factor to consider for above-ground mineral carbonation processes that make use of the aforementioned brines. The pH of brines containing Fe is affected when it is adjusted by a strong base whereas the pH of Fe-free brines experiment shows no significant variations. The pH of brines containing Fe2+ diminishes to very low pH (2.86) by air oxidation of highly alkaline Fe(OH)2 suspension induced by KOH addition by means of dissolution-oxidationcrystallisation mechanism where particles of Fe(OH)2 provide the substrate for the g-FeO(OH) growth. The pH pattern of brines containing Fe3+ is consistent with the formation of the highly insoluble Fe(OH)3 as a primary product due to the pH rise and subsequent formation of Fe2O3 by aging of Fe (OH)3nH2O under alkaline conditions. Results from this study, which determine the importance of Fe speciation in oil-field brine pH, are then of high value when evaluating the potential capacity of oil-field brines (above-ground brines) to permanently sequestering CO2. Acknowledgments The financial support of the Centre for Innovation in Carbon Capture and Storage (CICCS) through the Engineering and Physical Sciences Research Council, EPSRC (EP/F012098/1 and EP/F012098/ 2) is gratefully acknowledged. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcou.2016.06.004. References [1] United State Environmental Protection Agency. Accessed September 2015. http://water.epa.gov/type/groundwater/uic/wells_sequestration.cfm.
85
[2] R.W. Roach, R.S. Carr, C.L. Howard, B.W. Cain, An Assessment of Produced Water Impacts at Two Sites in the Galveston Bay System, U.S, Fish and Wildlife Service Houston, Texas, 1993. [3] Y. Soong, A.L. Goodman, J.R. McCarthy-Jones, J.P. Baltrus, Experimental and simulation studies on mineral trapping of CO2 with brine, Energy Convers. Manage. 45 (11–12) (2006) p. 1845–1859.10. [4] M.L. Druckenmiller, M.M. Maroto-Valer, Carbon sequestration using brine of adjusted pH to form mineral carbonates, Fuel Process. Technol. 86 (2005) 1599–1614. [5] Y.K. Kharaka, L.Y. Leong, C.G. Doran, G.N. Breit, Can produced water be reclaimed, experience with placeritra oil field, california, Proceedings of the 5th International Petrol Environment Conference, Albuquerque, NM, 1998, pp. 1–23. [6] Y. Soong, A.L. Goodman, J.R. Jones, J.R. Baltrus, Experimental and simulation studies on mineral trapping of CO2 with brine, Energy Convers. Manage. 45 (2004) 1845–1859. [7] F.A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, 5th ed., John Wiley & Sons, New York, Chichester, Brisbane, Toronto, Singapore, 1988, pp. p.71. [8] A.J. Bard, R. Parsons, J. Jordan, Standard Potential in Aqueous Solutions, Marcel Dekker, Inc, 270 Madison Avenue, New York, 2016, pp. 10016. [9] B. Morgan, O. Lahav, The effect of pH on the kinetics of spontaneous Fe(II) oxidation by O2 in aqueous solution basic principles and a simple heuristic description, Chemosphere 68 (2007) 2080–2084. [10] W. Feitknecht, W. Michaelis, Ober die hydrolyse yon eisen(III)-perchloratL6seungen, Helv. Chim. Acta 26 (1962) 212–224. [11] W. Feitknecht, R. Giovanoli, W. Michaelis, M. Muller, Die hydrolyse der LOsungen yon eisen (III)-chlorid, Helv. Chirn. Acta 56 (1973) 2847–2856. [12] M. Magini, R. Caminiti, On the structure of highly concentrated iron (III) salt solutions, J. Inorg. Nucl. Chem 39 (1977) 91–94. [13] Kobayashi, Uda, Structure of ferric hydroxide gel, J. Non-Cryst. Solids 29 (3) (1978) 419–422. [14] I.P. Saraswat, A.C. Va’pei, V.K. Garg, V.K. Sharma, N. Prakash, On the sorption studies of some metal ions with chromium ferrocyanide gel, J. Colloid Interface Scz. 73 (1980) 373. [15] A. Feitkneeht Wand Wytlenbach, Reactivity of soils, 4th Int. Symp. Ed, Elsevier, Hol., 1960, pp. p234. [16] A. Kasllaki, Osaka Kogyo Gijut. 30 (1979) 161. [17] R.M. Comell, R. Giovanoli, P.W. Sehindlcr, Glays and Clay minerals 35 (1987) 21. [18] T. Misawa, K. Hashimoto, S. Shimodaira, The mechanism of formation of iron oxide and oxyhydroxides in aqueous solutions at room temperature, Corros. Sci. 14 (1974) 131–149. [19] U. Schwertman, H. Fetcher, The formation of green rust and its transformation to Lepidocrocite, Clay Miner. 29 (1994) 87–92. [20] P.C. Singer, W. Stumm, Acidic mine drainage; rate-determining step, Science 167 (1970) 1121–1123. [21] J.D. Bernal, D.R. Dasgupta, A.L. Mackay, The oxides and oxyhydroxides of iron and their structural inter-relationships, Clay Miner. Bull. 4 (1959) 15–30. [22] O.W. Lau, The precipitation of ferrous hydroxide: a lecture demonstration, J. Chem. Educ. 56 (1979) 474.