CO2 Sequestration through Mineral Carbonation of Iron Oxyhydroxides

ARTICLE pubs.acs.org/est

CO2 Sequestration through Mineral Carbonation of Iron Oxyhydroxides Kristin Lammers,† Riley Murphy,† Amber Riendeau,† Alexander Smirnov,‡,§ Martin A. A. Schoonen,‡ and Daniel R. Strongin*,† †

Department of Chemistry, Temple University, Philadelphia, Pennsylvania 19122, United States Department of Geosciences, Stony Brook University, Stony Brook, New York 11794, United States § Department of Earth and Marine Sciences, Dowling College, Oakdale, New York 11769, United States ‡

bS Supporting Information ABSTRACT: Carbon dioxide sequestration via the use of sulfide reductants and mineral carbonation of the iron oxyhydroxide polymorphs lepidocrocite, goethite, and akaganeite with supercritical CO2 (scCO2) was investigated using in situ attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR), X-ray diffraction (XRD), and transmission electron microscopy (TEM). The exposure of the different iron oxyhydroxides to aqueous sulfide in contact with scCO2 at ∼70
100 °C resulted in the partial transformation of the minerals to siderite (FeCO3) and sulfide phases such as pyrite (FeS2). The relative yield of siderite to iron sulfide bearing mineral product was a strong function of the initial sulfide concentration. The order of mineral reactivity with regard to the amount of siderite formation in the scCO2/sulfide environment for a specific reaction time was goethite < lepidocrocite e akaganeite. Given the presence of goethite in sedimentary formations, this conversion reaction may have relevance to the subsurface sequestration and geologic storage of carbon dioxide.

1. INTRODUCTION An increase in both global energy consumption and in the combustion of fossil fuels to meet this energy demand has led to a continuing upward trend of CO2 emissions.1 Subsurface sequestration of carbon dioxide has been proposed as a method to mitigate climate change and global warming by limiting the emission of CO2 to the atmosphere.1 With 85% of the world’s energy consumption based on fossil fuel, one of the focal points of the efforts to curb emissions is to sequester CO2 emitted from large coal- or gas-fired power plants in subsurface geological formations.1 Potentially attractive geological formations for the storage of CO2 include oil and gas reservoirs or deep saline formations.1,2 Geological sequestration involves the subsurface storage of CO2 as a supercritical fluid (hydrodynamic trapping), via solubility and mineral trapping mechanisms.2,3 The mineral trapping of CO2 through subsurface carbonation reactions to form stable solid mineral phases is particularly attractive for the long-term storage of CO2. Consequently, there have been studies on the potential of mineral trapping in CO2 sequestration, with many of these studies focused on understanding whether sulfur-bearing waste gas in the supercritical CO2 (scCO2) injectate can facilitate the formation of iron carbonates in the subsurface environment.3
5 However, mineral trapping processes have been largely thought to proceed slowly (on the scale of thousands of years) and therefore have not typically been seen as important in the context of the amount of carbon sequestered.1,4,6 r 2011 American Chemical Society

Prior estimates of the time of mineralization of CO2 during sequestration operations, however, have largely neglected the possibility that species such as H2S and SO2 can be present in the scCO2 injectate. Previous research has suggested that sulfide coinjectants can potentially increase the rate of the long-term storage of CO2 by the reduction of ferric iron in specific mineral phases and the carbonation of the ferrous iron product to form siderite.3,4 Research by Palandri et al.4 demonstrated, for example, that the reaction between the ferric iron bearing hematite mineral and a SO2
CO2 mixture resulted in a ∼10% conversion to siderite product, along with the formation of pyrite and sulfur. Subsequent research in our laboratory showed that the iron oxyhydroxide, ferrihydrite,7 and hematite8 could also be converted to siderite in the presence of aqueous sulfide and scCO2 at a reaction temperature of 70 °C. The current study extends our research to the investigation of the reactivity of different FeO(OH) polymorphs—lepidocrocite (γ-FeOOH), goethite (α-FeOOH), and akaganeite (β-FeOOH)— in the presence of aqueous sulfide and scCO2 with the aim to resolve the fundamental reaction steps leading to the formation of Received: July 25, 2011 Accepted: November 8, 2011 Revised: October 28, 2011 Published: November 08, 2011 10422

dx.doi.org/10.1021/es202571k | Environ. Sci. Technol. 2011, 45, 10422–10428

Environmental Science & Technology

ARTICLE

siderite under conditions relevant to carbon capture and sequestration. Of these polymorphs, goethite is the most common to sedimentary phases and soils.9 The presence of lepidocrocite is more likely to be found in iron ore deposits and as an oxidation product of other iron-bearing minerals, whereas akaganeite is generally found in hot brines.9 Whereas goethite is perhaps most relevant to subsurface carbon sequestration, an investigation of all three oxyhydroxides phases allows us to shed light on the microscopic controls that underlie the carbonation process in the presence of sulfide. Prior studies have investigated the reductive dissolution of these iron oxyhydroxides phases in the presence of aqueous sulfide, 10
15 and have shown that the rate of reduction of the minerals by sulfide varies in the following order: lepidocrocite > akaganeite > goethite. A scientific goal of the present study is then to evaluate how the reduction rate of the polymorphs is affected under conditions relevant to CO2 sequestration. In this contribution, we present and interpret results obtained during and after the exposure of the different FeO(OH) polymorphs to aqueous sulfide and scCO2 at temperatures ranging from 70 to 100 °C. Such temperatures are common to those present at potential technologically relevant scCO2 injection sites.1,3,6 Attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) is used in this study as an in situ probe to determine the chemical transformations of the polymorphs during exposure to scCO2/sulfide. This in situ technique also gives insight into the relative amount of carbonate formation that occurred in the three different mineral systems for a given reaction time. X-ray diffraction (XRD) and transmission electron microscopy (TEM) were also used to characterize the structure and morphology of the solid product.

samples were characterized with TEM and XRD. The morphology determined by TEM and structures determined by XRD were consistent with prior published research for the suite of iron oxyhydroxides used in this study. Both in situ and ex situ experiments were carried out to simulate CO2 sequestration with a waste stream that contains supercritical CO2 and H2S. In these experiments aqueous sodium sulfide nonahydrate (Na2S) (Alfa Aesar) was used as the source of the sulfide-containing reductant. CO2 exists as a supercritical fluid above its critical temperature (31 °C) and pressure (71.8 bar). Elevated pressure was achieved using a CO2 tank fitted with a siphon tube (Airgas) and a high-pressure generator from High Pressure equipment (HiP). In situ infrared spectroscopy of the polymorphs during their exposure to scCO2 in the presence of Na2S(aq) was conducted using a Nicolet Nexus 670 Fourier transform infrared spectrometer (MCT-B detector, KRS-5 optics) equipped with a Specac Goldengate ATR accessory (diamond element). The FeO(OH) samples were individually prepared for the in situ experiments by first drying a suspension of the mineral of interest on the diamond element of the ATR accessory with nitrogen. The dried film adhered well to the ATR element, concentrating the analyte in the region probed by the evanescent IR wave emanating from the element. An aqueous deoxygenated solution of Na2S was prepared and ∼0.5 mL was deposited on the ATR element coated with the dried film. All the aqueous sulfide concentrations quoted in this paper refer to the initial Na2S concentration. Calculations performed using PHREEQC17 software estimated the equilibrium H2S partial pressure that would coexist with the aqueous sulfide solution at the desired temperature and under 83 bar scCO2 (see Table 1). A stainless steel microreactor cell fitted on the ATR accessory with a total volume of ∼0.5 mL was fastened over the reaction solution. This microcell was then pressurized with CO2 (83 bar) and the temperature was increased using the Goldengate heating assembly (70
100 °C) prior to acquisition and analysis of spectra (using Omnic 7.3 software). In all cases, 500 interferograms were coadded and averaged for each spectrum with a resolution of 4 cm
1. All spectra were ATR corrected and referenced against water at each corresponding temperature (70
100 °C) at 83 bar of scCO2. TEM images of the polymorphs and products prepared in an aqueous solution were obtained using a FEI Tecnai 12 microscope at 120 keV. Samples were characterized with XRD after exposure to the aqueous sulfide/scCO2 with a Bruker APEX DUO, with monochromatized MoKα radiation (λ = 0.71073 Å) and a Cu anode

2. EXPERIMENTAL SECTION Samples of lepidocrocite and goethite were synthesized by the method of Cornell and Schwertmann.9 In brief, lepidocrocite was synthesized via the oxidation of aqueous ferrous iron (prepared from FeCl2) as the pH of this solution was raised to pH 7 with NaOH addition. Goethite was also synthesized by the oxidation/ hydrolysis of aqueous ferrous iron (from FeCl2). In this circumstance oxidation of the ferrous iron was carried out in a solution buffered at pH 7 by NaHCO3. The as-formed lepidocrocite and goethite were washed with DI water several times and air-dried at room temperature. Akaganeite was prepared and characterized by another laboratory and the synthetic protocol is described elsewhere.16 The BET surface areas of synthesized lepidocrocite, goethite, and akaganeite were 115 ( 7, 53 ( 10, and 94 m2/g, respectively. The structure and morphology of the submicrometer

Table 1. Product Yields Associated with the Exposure of Lepidocrocite and Goethite to Aqueous Sulfide at 70 or 100 °C and in the Presence of 83 bar scCO2 final mineralogy a

b

c

c

% goethitec

% lepidocrocitec

-

-

55.33

-

12.0

-

-

-

Na2S (mM)

% H2S

% siderite

lepidocrocite

14.5

0.470

44.67

-

lepidocrocite

50.0

1.60

26.0

62.0

3.10

23.0

52.0

11.0

14.0

-

-

-

-

99.38

-

-

52.0

13.0

17.0

-

18.0

goethite goethite

100. 3.70 100.

0.205

0.620

4.80

-

% sulfur

% marcasitec

FeO(OH)

lepidocrocite

% pyrite

c

Lepdicrocite and goethite reacted at 70 °C and 100 °C, respectively. b %H2S is based on equilibrium pressure of gaseous H2S above the aqueous solution (relative to 83 bar scCO2) resulting from the indicated sodium sulfide concentration using PHREEQC. c Quantitative analysis based on Rietveld refinement of the powder diffractogram using Topas Academic. a

10423

dx.doi.org/10.1021/es202571k |Environ. Sci. Technol. 2011, 45, 10422–10428

Environmental Science & Technology

ARTICLE

a refinement was satisfactory. Lattice parameters and atomic positions for the minerals relevant to our studies were taken from Majzlan et al.22 In addition to conducting experiments in the microcell associated with the ATR-FTIR assembly, larger-scale reactions were carried out in a Parr high-pressure reactor cell (100 mL) with a Teflon liner. The total volume of the polymorph/sulfide suspension used in these experiments was ∼50 mL. The cell was heated (70
100 °C) and pressurized with CO2 (∼83 bar) for 24
48 h. Following the batch reaction, the cooled solution was repeatedly centrifuged and washed to remove any soluble product. These large-scale ex situ experiments allowed for further analysis by XRD. The diffractograms were collected on a Scintag PAD X powder diffractometer with monochromatized Cu Kα radiation (λ = 1.5405 Å), 40 kV, 25 mA, 5
90° 2θ, 0.02° 2θ step, and variable scan rates (10 s/step or 1 degree 2θ per minute, respectively). Analysis and identification of the phases was done using Match! software and quantified by Rietveld refinement using Topas Academic. TEM images of the polymorphs and products were obtained using a FEI Tecnai 12 microscope at 120 keV

3. RESULTS

Figure 1. ATR-FTIR spectra obtained in situ for the transformation of lepidocrocite and akaganeite to siderite at 70 °C under 83 bar CO2 in the presence of aqueous sulfide and spectra of the transformation of goethite to siderite at 100 °C under 83 bar CO2 in the presence of aqueous sulfide. Spectra of lepidocrocite and akaganeite are referenced against water at the same temperature and were obtained in ∼10 min intervals. Spectra for goethite are referenced against water at the same conditions, obtained in ∼40 min intervals and are offset for clarity with spectrum a representing time 0. Spectrum e (goethite) was obtained after ∼3.3 h of reaction time illustrating that siderite as well as goethite are still present. The inset in the lepidocrocite and goethite plots show changes in absorbance between successive spectra. All spectra are offset for clarity.

(a fluorescence filter was not present to block Fe fluorescence). XRD patterns were recorded in the range 2
50° 2θ with a step size of 0.02° at variable scan rates of 180
300 s
1 at ambient temperature. Phases were identified and quantified by Rietveld refinement18
20 using DIFFRACplus EVA and Topas Academic,21 respectively. For any particular analysis, the qualifications for optimum Rietveld refinement were judged from the goodness of fit, χ2 = Rwp/Rexp, where Rwp is the weighted sum of the residuals of the least-squares fit and Rexp is the statistically expected value. Low R-values and hence, χ2, were used to judge whether

ATR-FTIR Experiments. Figure 1 exhibits in situ ATR-FTIR spectra associated with the reaction of lepidocrocite (top) and akaganeite (middle) at 70 °C, and goethite (bottom) at 100 °C in the presence of 83 bar scCO2 and aqueous sulfide. The data presented below will show that FeCO3 forms during these reactions. It is important to mention that we also carried out control experiments where the three polymorphs were individually exposed to 83 bar scCO2 at the same respective temperatures, but in the absence of sulfide. Carbonate formation was never observed to form in these particular experiments (Supporting Information), where sulfide was not present, consistent with our results from prior studies.7,8 In the lepidocrocite circumstance, the sulfide concentration used to obtain the ATR-FTIR data presented in Figure 1 was 14.5 mM, that based on calculation produces a H2S partial pressure above the solution that is ∼0.5% of the scCO2 pressure (see Table 1). The initial pH of the aqueous solution in contact with the mineral before scCO2 was introduced was 11.3. After 10 min of reaction time, vibrational modes became apparent at ∼1400, 861, and 738 cm
1, which we attribute to the asymmetric stretch (ν3), out-of-plane (ν2) bending, and in-plane bending (ν4) modes, respectively, of siderite (spectra b
c).23 Further inspection of the data shows that the ν2 and ν4 vibrational modes associated with this phase show no significant change in intensity between 50 and 60 min of reaction, suggesting that the growth of this phase is largely complete after 50 min. To better illustrate this contention, spectra are displayed in the inset to Figure 1 (lepidocrocite) that show the change in vibrational mode absorbances between successive spectra. The spectra show that the vibrational mode at 1020 cm
1 that is associated with lepidocrocite decreases continuously in absorbance for times up to 40 min. This decrease occurs concomitantly with an increase in the 861 cm
1 mode that is associated with siderite. The spectrum associated with the difference between the 50 and 60 min spectrum shows insignificant changes in the 1020 and 861 cm
1 vibrational modes, suggesting that there is no further loss in lepidocrocite and no further increase in siderite product yield after 50 min of reaction. 10424

dx.doi.org/10.1021/es202571k |Environ. Sci. Technol. 2011, 45, 10422–10428

Environmental Science & Technology

ARTICLE

Figure 2. TEM of (a) lepidocrocite, (b) lepidocrocite exposed to 83 bar scCO2 at 70 °C, and (c) lepidocrocite exposed to 83 bar scCO2 and aqueous sulfide at 70 °C. Arrows indicate distortions around edges of lepidocrocite exposed to scCO2. Complete loss of the needle-like morphology suggests that dissolution and recrystallization reaction occurred.

Further inspection of lepidocrocite infrared data also shows that there is a shift in spectral weight between 1350 and 1450 cm
1 in spectra d
f. This is likely the result of the evolution of aqueous bicarbonate and/or adsorbed bicarbonate. An aqueous bicarbonate mode at ∼1360 cm
1 is perhaps most clearly isolated in spectrum a of the figure (denoted by an asterisk). This species is likely an intermediate species for the formation of siderite, and its contribution to the vibrational spectra, largely at 1360 cm
1, increases with reaction time. It is also of note that the 861 cm
1 mode, which has contributions from both siderite and bicarbonate, shows a marked growth up to 30 min. (see inset). The growth in the absorbance of this mode exceeds the absorbance increase of the ν4 (738 cm
1) mode that is a siderite-specific mode, consistent with the growth of the 861 cm
1 being due in part to an increase in the concentration of bicarbonate. ATR-FTIR spectra for akaganeite presented in Figure 1 show a similar reaction profile as lepidocrocite in the presence of scCO2 and aqueous sulfide (14.5 mM) at 70 °C. However, the conversion to siderite was relatively fast and started prior to heating to 70 °C and as soon as the akaganeite was exposed to the aqueous sulfide and scCO2. This conversion of akaganeite to siderite at such a low temperature was not experimentally observed to occur for the lepidocrocite or goethite (shown below) systems. Within 60 min of reaction the formation of siderite was largely complete based on the relatively constant absorbances of the siderite derived modes after this reaction time. Unfortunately, the experimental data do not allow the rate of siderite formation from lepidocrocite and akaganeite to be compared. Figure 1 also exhibits in situ ATR-FTIR spectra of goethite before and after exposure to scCO2 and sulfide at a reaction temperature of 100 °C. Spectrum a (i.e., time-zero) exhibits modes at 790 and 890 cm
1 that arise from the out-of-plane and in-plane stretching of goethite’s hydroxyl group, respectively. The experimental conditions used to obtain the goethite’s spectral data involved a sulfide concentration of 3.7 mM (partial pressure of H2S is 0.21% of scCO2 pressure). It is important to mention that we also carried out experiments that exposed goethite to scCO2/sulfide (e.g., 14.5 mM) at 70 °C, but in this circumstance, siderite product was not formed (data not shown) after 12 h of reaction time. Also, additional in situ experiments carried out on goethite suggested that conversion is more sensitive to sulfide concentration in the context of siderite formation than the other polymorphs. The transformation to siderite in the presence of scCO2 and sulfide was only experimentally observed in the presence of a relatively low concentration of sulfide (3.7 mM)

and 100 °C (in contrast to lepidocrocite and akaganeite). At these reaction conditions, the exposure of goethite to scCO2/sulfide resulted in the evolution of siderite modes after ∼1 h of reaction (goethite, spectrum b). Within ∼3 h, the intensity of the vibrational modes associated with siderite reached a maximum, which is evident in spectrum f. Over the period of siderite formation there was a shift in the peak maxima associated with the spectral weight in the 1350
1400 cm
1 region. Similar to our interpretation of the lepidocrocite and akaganeite systems, this spectral change is attributed to the growth of solution and/or surface-bound bicarbonate. Prior research also has suggested that the ν3 mode of the carbonates can show a redshift and increased broadening due to an alteration and distortion of siderite symmetry,24 and hence changes in the siderite structure may also contribute to the observed spectral changes. Based on the retention of goethite vibrational modes throughout the reaction, it is surmised that there is significant amount of goethite remaining along with the siderite product. A closer investigation of the proceeding reaction is illustrated by the inset in goethite’s infrared data, which contains each spectrum referenced against the previous spectrum. It is only after ∼200 min that goethite starts to diminish (see regions denoted by asterisks) and siderite grows based on the concomitant appearance of modes at 861 and 737 cm
1. TEM. Figure 2 displays a TEM image of lepidocrocite before reaction, after exposure to scCO2 at 70 °C, and after exposure to aqueous sulfide (14.5 mM) and scCO2 at 70 °C. Figure 2a shows the characteristic “needles” that is a morphological characteristic of lepidocrocite. After the exposure of lepidocrocite to scCO2, TEM analysis shows that the characteristic needle-like appearance of the mineral was largely retained, although some distortions around the edges of the needles become apparent (indicated by black arrows) indicating partial dissolution (Figure 2b). The distortion of the needle shape is possibly caused by a reduction in the pH (to ∼3.5) of the aqueous phase resulting from the presence of the contacting scCO2 phase. No detectable amount of siderite was experimentally observed under these experimental conditions. In the presence of aqueous sulfide, the pH in the reaction system was determined experimentally to be 5.8 immediately after the reaction cell was opened and the scCO2 released from the cell (the calculated pH using PHREEQC was determined to be 4.6). TEM of the post-reaction mass resulting from the reaction of lepidocrocite with aqueous sulfide/scCO2 did not show evidence for the presence of lath-like crystals associated with lepidococite, and instead equant crystals dominated the product morphology (Figure 2c). Based on these 10425

dx.doi.org/10.1021/es202571k |Environ. Sci. Technol. 2011, 45, 10422–10428

Environmental Science & Technology

ARTICLE

Figure 3. TEM of (a) goethite, (b) goethite exposed to 83 bar scCO2 at 100 °C, and (c) goethite exposed to 83 bar scCO2 and aqueous sulfide at 100 °C. In some cases the goethite particle morphology changed from acicular suggesting dissolution/reprecipitation reactions. Post-reaction XRD showed that about ∼1% of the goethite converted to other phases in the presence of scCO2 and sulfide.

particular morphological changes in crystal structure and size distribution, it is speculated that a large fraction of the lepidocrocite underwent (reductive) dissolution and this was followed by the reprecipitation of solid product. This notion is consistent with the results of the XRD analysis, presented below. Although not shown (included in Supporting Information), a TEM micrograph of post-reaction akaganeite showed morphological and particle size changes also consistent with reductive dissolution. Figure 3a shows needle-like particles that are characteristic of goethite. Comparison of TEM images of post-reaction products of a control reaction of goethite particles exposed to only scCO2 (Figure 3b) at 100 °C and reacted with scCO2 and 3.5 mM aqueous sulfide (Figure 3c) at the same temperature suggested that a fraction of the goethite converted to other phases during the reaction in the presence of sulfide. In particular, the alteration in the acicular shape of the siderite product in Figure 3c is suggestive of a dissolution, reprecipitation reaction on the goethite surface. XRD. A Rietveld refinement analysis of the XRD associated with the post-reaction mass associated with the ATR-FTIR experiments detailed above was carried out. The results of these analyses are presented in Table 1. An implicit assumption here is that the postreaction products are representative of those products formed in the presence of scCO2. In the circumstance where goethite was exposed to scCO2 and 3.5 mM aqueous sulfide at 100 °C for ∼3 h, the analysis shows that siderite made up less than 1% of the post-reaction mass. In contrast, results shown in Table 1 suggest that the exposure of lepidocrocite to scCO2 and 14.5 mM aqueous sulfide at 70 °C for ∼3 h resulted in a greater amount of siderite formation (44.67%). We also carried out XRD on samples used for a series of batch experiments that exposed lepidocrocite and goethite to higher concentrations of aqueous sulfide compared to the levels used in the in situ ATR-FTIR experiments. A Rietveld refinement analysis was also carried out on these samples before and after reaction and the results are presented in Table 1. For each iron oxyhydroxide, an increase in the amount of sulfide relative to what was used in the in situ ATR-FTIR experiments led to an increase in the conversion of the iron oxyhydroxides to other phases. The relative amount of siderite in the post-reaction mass, however, decreased upon this increase in sulfide concentration. For example, an analysis of the post-reaction mass resulting from the exposure of goethite to scCO2 and an 100 mM aqueous sulfide concentration at 100 °C contained FeIII OH þ HS
¼ > FeIII S
þ H2 O

ð1Þ

> FeIII S
¼ > FeII S 3

ð2Þ

> FeII S 3 þ H2 O ¼ > FeII OH2 þ þ S 3


ð3Þ

> FeII OH2 þ f surface site þ FeIIðaqÞ

ð4Þ

where > denotes a surface bound complex, eq 2 is related to the reduction of ferric iron to ferrous iron, eq 3 is related to the release of a sulfur radical that may proceed to sulfate, and eq 4 is related the pH dependent desorption of ferrous iron into solution. The sulfur radical that is released (eq 3) is responsible for reducing additional Fe(III) at the oxide surface. Fe(II) released into the aqueous phase can either react with sulfide to form iron sulfide or under our experimental conditions react with HCO3
to form iron(II) carbonate. Thus, in addition to eqs 1
4, our experimental observations suggest that the following reaction is operative: FeII þ HCO3
f FeCO3 þ Hþ

ð5Þ

in the presence of scCO2. Also, at the sulfide concentrations used in this study, we observe the formation of iron sulfide mineral 10426

dx.doi.org/10.1021/es202571k |Environ. Sci. Technol. 2011, 45, 10422–10428

Environmental Science & Technology

ARTICLE

phases. Pyrite formation could include the following reaction sequence: FeIIðaqÞ þ HS
¼ FeSðsÞ þ Hþ

ð6Þ

FeSðsÞ þ HS
f FeS2 þ Hþ þ 2e


ð7Þ

or pyrite can form on the FeS coated iron oxyhydroxides particle > FeS þ HS
f > FeII S2 þ Hþ þ 2e


ð8Þ

These reactions are likely coupled with the reduction of ferric iron either in the aqueous phase and/or in the iron oxyhydroxide. We suspect that pyrite formation occurs on FeS particles or on oxyhydroxide particles coated with FeS, considering that the growth of pyrite from solution-phase ferrous iron and sulfide is kinetically hindered.26 We point out that the prior research13,15 that investigated the reduction of iron oxyhydroxide by aqueous sulfide used sulfide concentrations that were significantly lower (∼10 times less) than those used in the current study. However, similar to the redox chemistry proposed in the prior studies, it is expected that the initial reaction of sulfide with the iron oxyhydroxides results in the rapid partitioning of ferrous iron into the aqueous phase. The reaction of this ferrous iron to form siderite and/or pyrite is expected to depend on the relative rates of reaction 5 that leads to carbonate formation and reactions 6 and 7 that lead to sulfide product. Analysis of our XRD-based results suggest that the most conversion occurs in the aqueous sulfide/scCO2 lepidocrocite system. This conclusion is consistent with the in situ FTIR results that show significantly more siderite product formation in the lepidocrocite (and akaganeite) system than in the goethite circumstance. It is mentioned that based on prior work,13 we suspect that the protonation of the surface that facilitates ferrous iron release into solution (eq 4) becomes increasingly kinetically hindered on the iron oxyhydroxide surfaces as they undergo a large degree of sulfide surface complexation with increasing reaction time. Our results analyzed in view of prior work suggest that differences in the amount of siderite forming from the reaction of different oxyhydroxides with aqueous sulfide and scCO2 is largely due to differences in the rate of reductive dissolution for the different iron oxyhydroxides. Prior research by Alfonso and Stumm,13 for example, showed that the rate of reductive dissolution of lepidocrocite by aqueous sulfide was significantly higher than for goethite. It is likely that the relatively fast reductive dissolution of lepidocrocite leads to more ferrous iron and siderite production than in the goethite circumstance over a similar reaction time. It was also proposed in prior research that the lower dissolution rate of goethite compared to lepidocrocite in the presence of aqueous sulfide is associated with its higher free energy for the reaction FeOðOHÞ þ 3H

þ




þ e ¼ Fe

2þ

þ 2H2 O

ð9Þ

and hence, the difference in reactivity can be related to the difference in the free energy gain in the reductive dissolution process.9,27 Although we are unaware of studies that have investigated the rate of akaganeite dissolution in the presence of sulfide, we suspect it should be similar to lepidocrocite (in agreement with our experimental observations), based on the Gibbs free energies of formation for the different phases ΔGof:
490.6 ( 1.5,
482.7 ( 3.1, and
481.7 ( 1.9 kJ/mol for goethite, lepidocrocite, and akaganeite, respectively.28 We

point out that it has been shown in prior research28 that the relative stability of the iron oxides is a function of particle size at the nanoscale. However, based on the surface areas of the particles used in this study we are in a particle size regime where nanoscale effects should not significantly affect the relative phase stability of the three iron oxyhydroxide phases. Overall, the results highlighted in this contribution suggest that the carbonation of lepidocrocite and akaganeite with a CO2 waste stream containing H2S would sequester both the carbon and sulfide efficiently. Hence, it might be possible to develop a process that could be operated associated with large CO2 point sources in locations without suitable sedimentary strata for subsurface sequestration. The results with goethite, which is present in some deep sediments, suggest that it will react with a CO2 waste stream containing H2S resulting in the mineralization and immobilization of a fraction of the CO2 as an iron carbonate.

’ ASSOCIATED CONTENT

bS

Supporting Information. (1) ATR-FTIR spectra associated with control experiments obtained in situ during reaction of goethite, akaganeite, and lepidocrocite in the presence of scCO2 and absence of aqueous sodium sulfide; (2) TEM images of akaganeite before and after reaction at 70 °C in the presence of scCO2 and sulfide (14.5 mM); (3) XRD patterns associated with the product formed via the reaction of lepidocrocite with scCO2 and 100 mM aqueous sulfide at 70 °C and goethite with scCO2 100 mM aqueous sulfide at 100 °C; and (4) selected PHREEQC output. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT D.R.S. and M.A.A.S greatly appreciate support from the Department of Energy, Basic Energy Sciences from grants DEFG029ER14644 and DEFG0296ER14633, respectively. D.R.S. also acknowledges partial support from NSF grants CHE0714121 and CHE0923077 (MRI grant to obtain the TEM used in this research). We also thank Alan Coelho for help with the Topas analysis and Dr. Frederick C. Monson (Center for Microanalysis and Imaging Research and Training Center, West Chester University) for help with the TEM. Wei Li is also thanked for providing akaganeite samples. ’ REFERENCES (1) IPCC. Carbon Dioxide Capture and Storage; Cambridge University Press: New York, 2005. (2) Oelkers, E. H.; Schott, J. Geochemical aspects of CO2 sequestration. Chem. Geol. 2005, 217, 183–186. (3) Palandri, J. L.; Kharaka, Y. K. Ferric iron-bearing sediments as a mineral trap for CO2 sequestration: Iron reduction using sulfur-bearing waste gas. Chem. Geol. 2005, 217, 315–364. (4) Palandri, J. L.; Rosenbauer, R. J.; Kharaka, Y. K. Ferric iron in sediments as a novel CO2 mineral trap: CO2-SO2 reaction with hematite. Appl. Geochem. 2005, 20, 2038–2048. (5) Renforth, P.; Washbourne, C.-L.; Taylder, J.; Manning, D. A. C. Silicate Production and Availability for Mineral Carbonation. Environ. Sci. Technol. 2011, 45 (6), 2035–2041. 10427

dx.doi.org/10.1021/es202571k |Environ. Sci. Technol. 2011, 45, 10422–10428

Environmental Science & Technology (6) Bachu, S. CO2 storage in geological media: Role, means, status and barriers to deployment. Prog. Energy Combust. Sci. 2008, 34, 254–273. (7) Murphy, R.; Lammers, K.; Smirnov, A.; Schoonen, M. A. A.; Strongin, D. R. Ferrihydrite phase transformation in the presence of aqueous sulfides and supercritical CO2. Chem. Geol. 2010, 271, 26-30. (8) Murphy, R.; Lammers, K.; Smirnov, A.; Schoonen, M. A. A.; Strongin, D. R. Hematite reactivity with supercritical CO2 and aqueous sulfide. Chem. Geol. 2011, 283, 210–217. (9) Cornell, R. M.; Schwertmann, U. The Iron Oxides: Structure, Properties, Reactions, Occurrence and Uses; Wiley-VCH Verlag GmbH & Co: Weinheim, 2003. (10) Rickard, D. T. Kinectics and mechanism of the sulfidation of goethite. Am. J. Sci. 1974, 274, 941–952. (11) Poulton, S. W. Sulfide Oxidation and Iron Dissolution Kinetics during the Reaction of Dissolved Sulfide with Ferrihydrite. Chem. Geol. 2003, 202, 79–94. (12) Pyzik, A.; Sommer, S. E. Sedimentary iron monosulfides: Kinetics and mechanism of formation. Geochim. Cosmochim. Acta 1981, 45, 687–698. (13) dos Santos Afonso, M.; Stumm, W. Reductive Dissolution of Iron(III) (Hydr)oxides by Hydrogen Sulfide. Langmuir 1992, 8, 1671–1675. (14) Peiffer, S.; dos Santos Afonso, M.; Wehrll, B.; Gachter, R. Kinetics and mechanism of the reaction of H2S with Lepidocrocite. Environ. Sci. Technol. 1992, 26, 2408–2413. (15) Poulton, S. W.; Krom, M. D.; Raiswell, R. A Revised Scheme for the Reactivity of Iron (oxyhydr)oxide minerals towards dissolved sulfide. Geochim. Cosmochim. Acta 2004, 68, 3703–3715. (16) Kim, J.; Li, W.; Phillips, B. L.; Grey, C. P. Phosphate adsorption on the iron oxyhydroxides goethite (α-FeOOH), akaganeite (b-FeOOH), and lepidocrocite (γ-FeOOH): A 31P NMR Study. Energy Environ. Sci. 2011, 4 (10), 4298–4305. (17) Parkhurst, D. L.; Appelo, C. A. J. User’s guide to PHREEQC (version 2) - a computer program for speciation, reaction-path, 1D transport, and inverse geochemical calculations; US Geol. Surv. Water Resour. Inv. Rep. 99-4259, 312p, 1999. (18) Bish, D. L.; Howard, S. A. Quantitative phase analysis using the Rietveld method. J. Appl. Crystallogr. 1988, 21, 86–91. (19) Hill, R. J.; Howard, C. J. Quantitative phase analysis from neutron powder diffraction data using the Rietveld method. J. Appl. Crystallogr. 1987, 20, 467–474. (20) Rietveld, H. M. A profile refinement method for nuclear and magnetic structures. J. Appl. Crystallogr. 1969, 2, 65–71. (21) Coelho, A. Topas Academic Users Manual Version 4.1; Coelho Software: Brisbane, Australia, 2007. (22) Majzlan, J.; Lang, B. E.; Stevens, R.; Navrotsky, A.; Woodfield, B. F.; Boerio-Goates, J. Thermodynamics of Fe oxides: Part I. Entropy at standard temperature and pressure and heat capacity of goethite (α-FeOOH), lepidocrocite (γ-FeOOH), and maghemite (γ-Fe2O3). Am. Mineral. 2003, 88, 846–854. (23) Dubrawski, J. V.; Channon, A.-L.; Warne, S. S. J. Examination of the siderite-magnesite mineral series by Fourier transform infrared spectroscopy. Am. Mineral. 1989, 74, 187–190. (24) Blengino, J. M.; Keddam, M.; Labbe, J. P.; Robbiola, L. PhysicoChemical Characterization of Corrosion Layers Formed on Iron in a Sodium Carbonate-Bicarbonate Containing Environment. Corros. Sci. 1995, 37, 621–643. (25) Liu, H.; Li, P.; Zhu, M.; Wei, Y.; Sun, Y. Fe(II)-induced transformation from ferrihydrite to lepidocrocite and goethite. J. Solid State Chem. 2007, 180, 2121–2128. (26) Schoonen, M. A. A.; Barnes, H. L. Reactions forming pyrite and marcasite from solution: I. Nucleation of FeS2 below 100 °C. Geochim. Cosmochim. Acta 1991, 55, 1495–1504. (27) Larsen, O.; Postma, D.; Jakobsen, R. The reactivity of iron oxides towards reductive dissolution with ascorbic acid in a shallow sandy aquifer (Rømø, Denmark). Geochim. Cosmochim. Acta 2006, 70, 4827–4835.

ARTICLE

(28) Navrotsky, A.; Mazeina, L.; Majzlan, J. Size-driven structural and thermodynamic complexity in iron oxides. Science 2008, 319.

10428

dx.doi.org/10.1021/es202571k |Environ. Sci. Technol. 2011, 45, 10422–10428