Biotransformation Processes Relevant to Geologic Carbon Sequestration: Potential Implications for Environmental Fate

Abstract

The nature and extent of microbial reactions in formations targeted for geologic carbon sequestration (GCS), as well as in nontarget formations that may be impacted by potential CO₂ migration, are key to understanding the fate of injected CO₂. The dissolution of CO₂ into formation waters drives predictable geochemical changes, including pH reduction, shifts in redox conditions, and increased mineral solubility. These changes can alter microbial community composition (e.g., favoring acid-tolerant taxa) and stimulate microbes capable of using CO₂ as a carbon source. Resulting biotransformation processes can transfer CO₂ into the mineral phase (e.g., microbially facilitated carbonate precipitation), gas phase (e.g., methanogenesis), or organic phase (e.g., biomass formation). However, the extent, rate, and significance of these processes in both target and nontarget environments are not well understood. This paper reviews current knowledge of CO₂ biotransformation relevant to GCS, including reactions occurring in deep storage formations and those arising from potential CO₂ migration into shallow groundwater aquifers, the vadose zone, and marine environments. Additionally, factors that influence these transformations are summarized, methods for monitoring microbial processes are discussed, and key research gaps that could facilitate improved prediction of the long-term fate of CO₂ under varying environmental conditions are identified.

1. Introduction

There are ongoing efforts to stabilize and reduce greenhouse gas (e.g., carbon dioxide (CO₂)) emissions and concentrations in the atmosphere to support a net-zero future. − Carbon capture and storage (CCS) utilizing geologic carbon sequestration (GCS), also referred to as geological carbon storage, is the long-term storage of CO₂ in underground geologic formations, typically saline aquifers or depleted oil and gas reservoirs. The capture and storage of CO₂ in deep geologic formations has the potential to limit increases in atmospheric CO₂ concentrations while the world economy continues to decarbonize. − Understanding the long-term fate of CO₂ is important to operationalize and successfully implement GCS. This paper provides a review of the current knowledge on CO₂ biotransformation processes that influence its fate both within the target storage formations (such as depleted hydrocarbon reservoirs) and in the event of migration into nontarget environments, including shallow groundwater aquifers, unsaturated soils, and marine waters and sediments (Figure ). In addition, factors potentially controlling the overall magnitude of such transformation, methods for monitoring relevant biotransformation processes, key knowledge gaps, and research opportunities are discussed.

1.

Conceptual model for GCS including transport, injection, trapping within target formation, and hypothetical migration pathways to be evaluated during GCS design: (a) onshore; (b) offshore. CO₂(sc) = supercritical CO₂, CO₂(l) = liquid CO₂; CO₂(aq) = aqueous (dissolved) CO₂; CO₂(g) = gaseous CO₂.

Microbes can facilitate various carbon cycling processes that interact in either competing or synergistic ways. The relative dominance of these microbially mediated processes is shaped by site-specific environmental conditions, which may be influenced by elevated levels of CO₂(aq). CO₂-consuming processes include hydrogenotrophic methanogenesis, chemolithoautotrophy, and homoacetogenesis, while CO₂-producing processes include methanotrophy, heterotrophy, acetoclastic methanogenesis, heterofermentation, and hydrocarbon degradation (Figure and Table , detailed discussion provided in Supporting Information (SI) section S1).

2.

Microbially mediated biotransformation reactions relevant to the cycling of carbon in the subsurface. ED = electron donor (e.g., H_2, NH₄ ⁺, H₂S, Fe­(II)). EA = Electron acceptor (e.g., O₂, NO₃ ^–, Fe­(III), SO₄ ^2–). EA* = Hydrocarbon reduction can occur using an exogenous terminal electron acceptor or without an exogenous terminal electron acceptor (i.e., fermentation). This figure was adapted from Figure in ref . Copyright 2023 American Chemical Society.

1. Microbial Processes Responsible for Carbon Cycling in Carbon Sequestration Storage Sites.

reaction	effect on CO2	effect on methane	effect on pH	other (e.g., increase in dissolved iron)	example of key reactions	reactions reference	possible implications
Reactions That Generate CO2
Methanotrophy: Microbial oxidation of CH4 to CO2 using oxygen, nitrate, nitrite, manganese, iron, or sulfate as electron acceptors.	Generates CO2	Consumes CH4	Oxygen, nitrate, and sulfate reduction: Decrease in pH (associated with production of CO2).	Increase in reaction byproducts like Fe(II), H2S.	Redox reactions of major components in natural waters with methane oxidation :		Carbonate mineral formation; release of reaction byproducts (e.g., iron, H2S) or liberation of metals under reducing/low pH conditions.
					Oxygen reduction: CH4 + 2O2 → CO2 + 2H2O
					Nitrate reduction: 5CH4 + 8NO3 – + 8H+ → 5CO2 + 4N2 + 14H2O
					Manganese reduction: CH4 + 4MnO2 + 4HCO3 – + 4H+ → CO2 + 4MnCO3 + 6H2O
					Nitrate reduction: CH4 + NO3 – + 2H+ → CO2 + NH4 + + H2O
			Manganese and iron reduction: Increase in pH (associated with consumption of H+).		Iron reduction: CH4 + 8FeOOH + 8HCO3 – + 8H+ → CO2 + 8FeCO3 + 14H2O
					Sulfate reduction: CH4 + SO4 – + H+ → CO2 + HS– + 2H2O
Heterotrophy: Microbial oxidation of reduced carbon compounds (e.g., acetate) to produce CO2 and H2O. The catabolization of carbon compounds can proceed by fermentation and/or respiration.	Generates CO2	None	Depends on specific reaction: (balance of H+ consumed vs H+ liberated by CO2 production, hydration, and subsequent H2CO3 dissociation).	Increase in reaction byproducts like Fe(II), H2S.	Respiration: CH2O + O2 → CO2 + H2O		Carbonate mineral formation; release of reaction by products (e.g., iron, H2S) or liberation of metals under reducing/low pH conditions.
					Denitrification: 5CH2O + 4NO3 – + 4H+ → 5CO2 + 2N2 + 7H2O
					Iron reduction: CH2O + 4Fe(OH)3 + 4H+ → CO2 + 4Fe2+ + 11H2O
					Sulfate reduction: 2CH2O + SO4 2– + H+ → 2CO2 + HS– + 2H2O
Acetoclastic methanogenesis: Microbial reduction of acetate to CH4 and CO2.	Generates CO2	Generates CH4	Decrease in pH (associated with production of CO2).	–	Acetoclastic methanogenesis: CH3COOH → CO2 + CH4		Methane production drives decreased CO2 solubility; carbonate mineral formation; liberation of metals under reducing/low pH conditions
Hydrocarbon degradation: Microbial oxidation of hydrocarbons, yielding CO2 as a byproduct.	Generates CO2	None	Depends on specific reaction: (balance of H+ consumed vs H+ liberated by CO2 production, hydration, and subsequent H2CO3 dissociation).	Increase in reaction byproducts (e.g., Fe(II), H2S; see “methanotrophy”).	Benzene degradation: C6H6 + 4.5H2O → 2.25CO2 + 3.75CH4 (also see “methanotrophy”).		Carbonate mineral formation; release of reaction byproducts (e.g., iron, H2S) or liberation of metals under reducing/low pH conditions.
Reactions That Consume CO2
Hydrogenotrophic methanogenesis: Microbial reduction of CO2 to CH4 and H2O by using H2 as an electron donor.	Consumes CO2	Generates CH4	Increase in pH (associated with consumption of CO2)	–	Hydrogenotrophic methanogenesis: CO2 + 4H2 → CH4 + 2H2O		Methane production drives decreased CO2 solubility.
Chemolithoautotrophy: Microbial use of inorganic carbon (CO2) as a carbon source, coupled with the oxidation of electron donors including H2, H2S, Fe(II) and NH3, to produce organic molecules.	Consumes CO2	None	Increase in pH (associated with consumption of CO2)	–	H 2 S Oxidation (e.g., at hydrothermal vents): CO2 + 4H2S + O2 → CH2O + 4S + 3H2O		–
Homoacetogenesis: Microbially mediated production of acetate from CO2.	Consumes CO2	None	Increase in pH (associated with consumption of CO2)	–	Acetogenesis: 2CO2 + 4H2 → CH3COOH + 2H2O		–
Other Reactions
Heterofermentation: Microbial conversion of organic molecules into acids, gases, or alcohols.	None	None	Increase in pH (associated with consumption of H+)	–	Fermentation of acetate to form ethanol: CH3COOH + H+ + 2H2 → C2H6O + H2O		–

In addition to directly producing or consuming CO₂, microbial processes can influence carbon cycling through several secondary effects: (i) microbially facilitated carbonate mineral formation, (ii) methane (CH₄) generation, which reduces CO₂ solubility, and (iii) microbial growth, which stores carbon in the form of biomass (Figure ). Microbially facilitated mineral formation can promote carbonate mineral precipitation and includes both controlled processes, where organisms actively regulate the nucleation and growth of intracellular biominerals, and induced or influenced processes, where mineral formation results from microbial metabolic activity or simply the presence of biomass. − CH₄ production via hydrogenotrophic, acetoclastic, or methylotrophic methanogenesis decreases the effective solubility of CO₂ in formation water. Thus, production of CH₄ is likely to reduce the dissolution of separate-phase CO₂ and reduce the CO₂ storage capacity in formation water. Biomass formation occurs through chemolithoautotrophy, in which CO₂(aq) is converted to organic matter using energy derived from the oxidation of reduced inorganic compounds like Fe­(II), Mn­(II), elemental sulfur, or sulfur compounds. − These chemolithoautotrophs may therefore play a key role in subsurface carbon cycling by generating organic matter that supports other microbial processes. For more detailed discussion of these secondary processes, see SI sections S2–S4.

3.

Possible cycling of CO₂ in deep formations targeted for carbon sequestration. Abiotic fate and transport processes are shown such as CO₂ phase separation (i.e., separation of CO₂(sc) and CO₂(aq) phases) and convective mixing driven by density gradients. Further, potentially relevant abiotic and biotransformation processes are depicted, including mineral dissolution and formation, carbon fixation (i.e., biomass generation) and catabolism, and methanogenesis and methanotrophy. L = Limestone, B = Basalt, S = Sandstone, D = Depleted hydrocarbon reservoir. A depleted hydrocarbon reservoir may be a sandstone, limestone, or other sedimentary formation.

The occurrence and extent of these primary and secondary carbon cycling processes will depend on (i) physical factors affecting the phase and residence time of CO₂ within that environment, (ii) the local chemical and geochemical conditions, and (iii) the composition of the resident microbial community. These factors are summarized in more detail below.

At the most basic level, understanding the long-term fate of subsurface CO₂ begins with examining its physical state during GCS and how that state may evolve over time. CO₂ is typically injected into deep formations as either supercritical CO₂ (CO₂(sc)) or liquid CO₂ (CO₂(l)), depending on the surrounding temperature and pressure conditions (refer to a CO₂ pressure–temperature phase diagram, such as that produced by Finney and Jacobs, 2010). Once injected underground, a portion of supercritical CO₂ (CO₂(sc)) dissolves into formation water within the geologic reservoir. The resulting CO₂-rich water (CO₂(aq)) is slightly denser than CO₂-poor formation water, which can lead to gravitational instability when CO₂-saturated water overlies less dense brine. This density contrast may drive downward-moving convective “fingers.” In contrast, undissolved CO₂(sc), which is less dense than formation water, can migrate upward if conduits like natural fractures or artificial penetrations exist. During this upward migration, the drop in pressure will cause the CO₂(sc) to transition into aqueous CO₂ (CO₂(aq)) and, depending on solubility and depth, potentially into gaseous CO₂ (CO₂(g)).

Microorganisms that consume, produce, or otherwise interact with CO₂ in the subsurface are mainly located in water-filled pores or in thin water films that coat mineral surfaces. As a result, microbial activity related to CO₂ is generally expected to target its dissolved form (i.e., CO₂ in water). This holds true whether CO₂ is present in deep storage reservoirs, overlying aquifers, or the vadose zone, where microbial life depends on thin water films as a habitat. Therefore, as an initial consideration, the potential for microbially mediated CO₂ biotransformation is governed by physical processes that determine its phase and the duration of its residence in a particular environment. Beyond that, the potential for biotransformation of CO₂ depends on the resident microbiome, and chemical and geochemical conditions.

The deep subsurface (>700 m) hosts a large and diverse range of microorganisms and, as a whole, is one of the largest reservoirs of biomass on the planet. Despite the challenges involved in accessing these environments, prior efforts have demonstrated that deep subsurface microbial communities are metabolically diverse and capable of generating energy by utilizing the surrounding geochemical conditions. , The composition of microbial communities that thrive in a given location depends largely on how favorable the local conditions are for specific biotransformation processes. These conditions are shaped by a range of physical and chemical factors, including but not limited to temperature, pressure, pH, salinity, concentrations of key reactants (e.g., electron donors and acceptors), and the availability of nutrients.

A significant increase in CO₂(aq), whether in the target formation or overlying formations, can trigger various physical and chemical changes, primarily due to the formation of carbonic acid when CO₂ dissolves (Figure ). This results in a pH decrease (acidification), and changes in redox equilibria, driven by changes in the availability of electron donors and acceptors. Additionally, acidic conditions can promote mineral dissolution, leading to higher total dissolved solids (TDS), and increased solubility of metals and organic compounds. Over time, secondary minerals may precipitate as divalent cations released during mineral dissolution precipitate with carbonate ions dissolved in the water. These direct and indirect effects of increased CO₂(aq) concentrations can affect the Gibbs free energy changes associated with the various microbially mediated carbon-cycling reactions, which are largely dependent on the nature and abundance of electron donors and acceptors. As such, these geochemical shifts can create conditions that favor certain microorganisms, such as acidophiles, while inhibiting others.

4.

Geochemical changes following CO₂ dissolution into water that can drive mineral dissolution and precipitation.

Constraints in dissolved organic carbon (DOC) and essential nutrients such as nitrogen and phosphorus can also limit the microbial community in deep subsurface aquifers, many of which are considered oligotrophic. Nevertheless, while these limitations can suppress microbial growth (i.e., biomass production), microbes may still generate enough energy for basic survival. , DOC levels in oligotrophic groundwater typically range from 0.2 to 2 mg/L, though only a portion is bioavailable. In depleted hydrocarbon reservoirs, hydrocarbons can serve as electron donors, but microbial activity may be limited by a lack of electron acceptors, which have been depleted over millions of years. Currently, there are few studies documenting the levels of bioavailable nutrients, organic carbon, electron donors, and electron acceptors in deep aquifers, limiting the ability to predict microbial activity under nutrient- or electron donor-limited conditions in target formations and marine settings. Although shallow groundwater aquifers and the unsaturated zone periodically receive inputs of organic carbon and nutrients through infiltration and interactions with surface water, the precise factors that limit microbial activity and growth in these environments remain difficult to predict. Given these knowledge gaps, this review focuses on the potential for CO₂ biotransformation in relevant environments, without addressing the magnitude or rate of these processes. Factors influencing microbial activity and carbon biotransformation during geologic sequestration of CO₂ in target formations, and the potential unintended migration of CO₂ into nontarget environments (i.e., shallow groundwater aquifers, unsaturated soils, and marine waters and sediments), are discussed in detail in sections –. Additional information on microbial processes relevant to CO₂ transport infrastructure (e.g., pipelines) can be found in the SI section S5

2. Target Formation for Geologic Carbon Sequestration

To promote long-term sequestration, target geological formations should be overlain by one or more impermeable formations (i.e., caprock). CO₂ may be sequestered within geologic formations through four primary trapping mechanisms , (Figure ):

• 1. Physical trapping whereby CO₂(sc) is confined within the target geological formation by the overlying impermeable formation.
• 2 Residual trapping whereby CO₂(sc) is trapped in the formation pore spaces by capillary forces.
• 3. Dissolution or solubility trapping where CO₂ is dissolved into the formation waters (i.e., trapped as CO₂(aq)).
• 4. Mineral trapping where the CO₂ reacts to form a stable mineral phase, typically carbonate minerals.

5.

Trapping mechanisms for carbon sequestration within target geologic formations.

The relative importance of these four distinct trapping mechanisms can change over time and depend on the dominant CO₂ phase. At typical injection depths (i.e., > 800 m), the temperature and pressure will be above the critical point for CO₂ (73.8 bar and 31 °C). As a result, the CO₂ will be in a supercritical state with a density similar to CO₂(l) and a very low viscosity similar to CO₂(g). Due to its buoyancy, injected CO₂(sc) will tend to rise through the target formation during flow away from the injection point; thus physical trapping at the base of the caprock is expected to be the most important short-term trapping mechanism.

Over a time scale of decades or longer, dissolution and other transformation processes are expected to become more important. Dissolution of CO₂ will be driven by diffusion, dispersion, and advection. For example, the dissolution of CO₂ into brine increases the density of the brine potentially leading to convective mixing that speeds the dissolution of CO₂(sc) trapped in the upper layers of the target formation (Figure ). Once dissolved, abiotic or biotic (i.e., microbially facilitated) reactions may lead to mineral trapping or other transformations of the CO₂. Understanding the role of microorganisms in transforming injected CO₂ could help predict the medium- and long-term fate of CO₂, the permanency of sequestration, and the potential consequences of migration from the target formation. Similarly, understanding the impact of CO₂ on the native microbial community could be important for predicting the impacts of possible CO₂ migration from target formations and may inform methods to monitor for such migration.

2.1. Environmental Conditions

The addition of CO₂ can impact the physical environment, the chemical conditions, and the microbial community. Information on baseline geochemistry of formation waters (i.e., water present within a target formation) can inform both the abiotic and microbially mediated reactions that may occur following CO₂(sc) injection. Most target formations are deep aquifers characterized by elevated salinity (i.e., greater than 10,000 mg/L TDS, and commonly >35,000 mg/L TDS). This salinity can originate from multiple sources, including prolonged contact time with soluble minerals in the aquifer matrix and saline parent water (e.g., seawater trapped in sedimentary basins), among others. The degree of salinity, and its composing ions, may vary significantly between strata, and laterally within formations. , There are several important formation characteristics that are likely to vary by formation types and/or reservoirs. These characteristics include:

• 1.the ability to resist changes in pH by either absorbing or desorbing H⁺ and OH^– ions (i.e., buffering capacity);
• 2.the presence and availability of divalent metal-bearing silicates, which can release metals (e.g., Fe­(II), Ca²⁺, Mg²⁺) that interact with dissolved CO₂ (as HCO₃ ^– or CO₃ ^2–) to form carbonate minerals; and
• 3.the presence of nutrients that support microbial growth, and various microbial reactions.

While site-specific characterization of target formation water chemistry will be required to understand specific geochemical conditions, the common variations across different formations targeted for carbon sequestration are summarized below:

• 1.Carbonate-bearing formations (i.e., limestones, dolomites): Carbonate-bearing formations commonly exhibit a high buffering capacity associated with the availability of carbonate (CO₃ ^2–) and bicarbonate (HCO₃ ^–) ions, which neutralize acid by reacting with H⁺. At the same time, carbonate-bearing formations tend to lack divalent metal-bearing silicates, which are largely dissolved during sedimentation.
• 2.Sandstones: Without a mineralogy dominated by carbonates, sandstones have a lower buffering capacity than limestones and dolomites. However, like other sedimentary formations, sandstones largely lack divalent metal-bearing silicates sedimentation.
• 3Basalt: Igneous basalt formations have little initial buffering capacity, although they are rich in divalent metal-bearing silicates. Basalts are also characterized by relatively fast water-rock reaction rates that can result in increased pH and alkalinity over time. In addition, serpentinization, the reaction of water with iron-rich minerals in basalt formations, can serve as a source of H₂.
• 4.Depleted hydrocarbon reservoirs: While depleted hydrocarbon reservoirs may be limestones, sandstones, or other sedimentary formations, these reservoirs are characterized by the presence of high levels of organic compounds (i.e., hydrocarbons), which can support microbial growth.

2.2. Microbial Community

In the subsurface, temperatures generally increase with depth, consistent with an average geothermal gradient of 2.5 °C/100 m. Although target formations can exhibit significantly higher temperatures and pressures compared to shallow groundwater and soils, virtually all target geologic formations for CO₂ sequestration still exhibit temperatures low enough to sustain microbial activity. Only at depths greater than 3000 m are water temperatures likely to exceed the 122 °C threshold for microbial life. The diversity and function of microbial consortia vary from site to site depending on environmental conditions. Nevertheless, while deep reservoirs are considered to be biologically and chemically isolated, they commonly contain diverse microbial populations including organisms that can utilize CO₂ as electron acceptors (e.g., methanogenic archaea and homoacetogenic bacteria), SO₄ ^2– reducing microorganisms (particularly in formations where SO₄ ^2– concentrations exceed 100 μM or 9.6 mg/L), and Fe­(III) reducing bacteria. In depleted hydrocarbon reservoirs and other organic-rich formations, organic material can serve as electron donors for microbial populations.

In the short term following CO₂ injection, the extant microbiome will be shaped by those that can survive the initial CO₂ influx and decrease in pH. Under low pH conditions, active microbial communities are typically those that are more adept at tolerating CO₂ stress. In addition, chemical reactions that consume protons are more energetically favorable at low pH. Prior work has shown that chemolithoautotrophs (e.g., hydrogenotrophic methanogens), heterotrophic Fe­(III)-reducing bacteria, and homoacetogenic bacteria are favored under low pH conditions. , For these microorganisms, a lower pH can also help the acquisition of essential nutrients for microbial growth. A number of studies have monitored changes in microbial community structure in response to high CO₂(aq) either in the laboratory or in the field. ,,− Although the nature of these changes varies depending on the site-specific geochemistry and microbial community prior to injection, common changes include enrichment of methanogenic archaea and metal-reducing bacteria.

The microbial community within natural CO₂-rich waters (e.g., natural CO₂ source fields, volcanic vents) may be representative of microbiomes that will develop within a target formation over the longer term following the injection of CO₂(sc). The McElmo Dome in Colorado is a natural CO₂ source field considered as an analogue for GCS reservoirs. The microbial communities in this field are dominated by chemolithoautotrophic microorganisms that utilize sulfur compounds, hydrogen, or Fe­(II) as electron donors, as well as fermenters and microorganisms performing anaerobic respiration (e.g., NO₃ ^–, SO₄ ^2– or As­(V) utilized as terminal electron acceptors). Interestingly, methanogenesis was not observed, potentially due to the high abundance of SO₄ ^2– and SO₄ ^2– reducing bacteria which can outcompete methanogens for available electron donors. In Crystal Geyser, Wyoming (a CO₂-venting geyser), metagenome analysis showed the potential for anaerobic respiration, nitrogen fixation, CO₂ fixation, and fermentation, ,, and in the Okinawa Trough off the coast of Japan and Taiwan (a hydrothermal system characterized by CO₂ seeps), the microbiome was dominated by methanotrophs and chemolithotrophs. , In summary, the available literature suggests that most or all target and nontarget formations will have diverse microbial communities capable of mediating a wide range of CO₂ biotransformation reactions.

2.3. Biotransformation Processes

Deep target formations, except for depleted hydrocarbon reservoirs, are typically nutrient poor (i.e., oligotrophic) and low in energetic compounds (i.e., electron donors) that support microbial growth and activity. As a result, the baseline rates of microbial processes in these environments are thought to be generally low. As noted above, once dissolved, CO₂(aq) can initiate a number of abiotic reactions that change the geochemical environment, and in turn, can impact the nature and rates of biotransformation processes. CO₂(aq) will also be bioavailable and potentially subject to biotransformation. These processes are described below.

As CO₂(sc) dissolves into the formation water, the CO₂ will react with the water to form carbonic acid, which itself will dissociate to form a hydrogen ion (H⁺) and a bicarbonate ion (HCO₃ ^–), or two hydrogen ions and a carbonate ion (CO₃ ^2–), as described by the following equations:

$${\mathrm{CO}}_2(\mathrm{aq})+{\mathrm{H}}_2\mathrm{O}\rightleftarrows {\mathrm{H}}_2{\mathrm{CO}}_3$$

$${\mathrm{H}}_2{\mathrm{CO}}_3\rightleftarrows {\mathrm{H}}^{+}+{{\mathrm{HCO}}_3}^{-}$$

$${\mathrm{HCO}}_3-\rightleftarrows {\mathrm{H}}^{+}+{{\mathrm{CO}}_3}^{2-}$$

Modeling of pH changes resulting from CO₂(sc) injection suggests that formation pH can drop 1.5 to 4 pH units below baseline conditions, depending on factors such as formation type, water chemistry, and temperature. , The shift to more acidic conditions can drive several biogeochemical changes. These include increased mineral dissolution, which results in the release of divalent metal ions as well as carbonate ions, where present. Several laboratory- and field-scale studies have observed an increase in cation concentrations over time, − which is largely attributed to increased rock weathering as a consequence of pH reduction associated with CO₂(sc) injection as described above.

The reduction in pH associated with dissolution of CO₂(sc) changes the thermodynamic favorability of various biogeochemical reactions. For example, Onstott (2005) reported that decreasing pH can substantially favor Fe­(III) reduction via oxidation of acetate or H₂. Such changes in pH may be particularly important in iron-rich formations like basalts. By contrast, decreasing pH does not significantly affect the thermodynamics of SO₄ ^2– reduction. Therefore, under lower pH conditions, Fe­(III) reducing microorganisms are likely to outcompete SO₄ ^2– reducers for the available electron donors in environments where both SO₄ ^2– and Fe­(III) are available electron acceptors.

The pH is likely to remain below baseline conditions as long as dissolution of CO₂(sc) is ongoing. Nevertheless, laboratory and computational experiments suggest that, in many reservoirs, chemical reactions that predominate at lower pH conditions ultimately drive an increase in pH over time. This increase results from both higher alkalinity associated with mineral dissolution and possible effects of microbial processes like heterotrophic Fe­(III) reduction, which consumes H⁺ and produces CO₂ (Table ). The time scale of this pH rebound is subject to significant uncertainty and likely situational- or site-specific. However, pH may recover close to or somewhat above baseline (i.e., initial) pH levels, depending on factors like salinity, formation mineralogy (i.e., buffering capacity), and formation temperature. ,

The increase in pH and the abundance of divalent cations can drive the precipitation of various mineral phases (e.g., carbonate and clay minerals including chalcedony, dolomite, siderite, kaolinite, and sulfide minerals; see Table S.1). By contrast, the precipitation of these minerals is unlikely to occur while formation water pH remains below 5 or in the absence of high concentrations of cations. Carbonate mineral formation may be particularly important in silicate-rich formations (e.g., basalts), where the initial drop in pH can result in the release of divalent metals through the dissolution of noncarbonate minerals. In sandstone formations lacking in divalent metals, carbonate mineral formation is likely to be limited. Similarly in limestone formations, carbonate mineral formation is likely to replace the carbonate minerals initially dissolved, resulting in little or no net change in solid-phase carbon.

Microbially Facilitated Mineral Formation

Even when formation waters are supersaturated with respect to carbonate minerals, kinetic limitations prevent rapid abiotic precipitation of minerals like dolomite, magnesite, and siderite. , As further discussed in SI section S2, microbial activity may overcome this barrier by mediating carbonate mineral formation on a much shorter time scale (e.g., weeks to months) than would otherwise occur.

Microbial reactions that can facilitate carbonate precipitation include the microbial reduction of Fe³⁺, SO₄ ^2–, and NO₃ ^–, and the microbial oxidation of CH₄, which produces CO₂ as a reaction byproduct. Metabolic processes like urea hydrolysis (conversion of urea into ammonia and bicarbonate catalyzed by urease enzymes) and the activity of carbonic anhydrase (a family of enzymes that catalyzes the interconversion of carbonic acid and CO₂(aq) and the dissociation of carbonic acid) can also control biomineralization. ,−

Microbial extracellular polymeric substances (EPS) (polymers synthesized by various microorganisms particularly during biofilm production) can also play an important role in increasing rates of mineral weathering and leaching, as well as the precipitation of carbonate minerals. For the latter, EPS serves as a matrix for the adsorption of metal cations to which carbonate ions bind. However, there is relatively little information on the extent of EPS in deep subsurface formations or its effect on mineral formation. Similarly, little is known about the natural occurrence of urease or carbonic anhydrase activity within storage reservoirs targeted for CO₂ injection, although some researchers have suggested the engineered use of these metabolic processes to enhance biomineralization. An improved understanding of the occurrence and specific processes enabling microbially facilitated mineral formation in target formations is likely required in order to support development of methods to enhance or monitor these desirable transformations.

Reactions That Support CO₂ Fixation

There are a number of microbially mediated reactions that convert inorganic CO₂ into organic carbon (i.e., chemolithoautotrophic reactions, carbon fixation). Chemolithoautotrophic microorganisms can utilize sulfur, nitrogen (NH₃ or NO₂ ^–) or Fe­(II) as electron donors, however, these reactions typically require oxygen or NO₃ ^– as the electron acceptors. In GCS reservoirs where oxygen and NO₃ ^– are likely to be limited or absent, there are several processes utilizing other electron acceptors that may support CO₂ fixation. These include Fe­(III) reduction, SO₄ ^2– reduction, homoacetogenesis, and methanogenesis, all of which can utilize H₂ as the electron donor. For homoacetogenesis and methanogenesis, CO₂(aq) is used as the electron acceptor resulting in conversion of some fraction of CO₂(aq) into biomass. With iron and SO₄ ^2– reduction, Fe­(III) and SO₄ ^2– are used as electron acceptors and, thus, these reactions do not directly result in CO₂ fixation. However, the energy obtained from these reactions can be used for CO₂ fixation via other pathways. As a result, Fe­(III) and SO₄ ^2– reduction processes may still support microbial growth and increase biomass.

Microorganisms and genes associated with anaerobic chemolithoautotrophy have been detected in a number of deep reservoirs targeted for GCS. At the Carbfix basaltic site, bacteria and archaea associated with anaerobic chemolithoautotrophy were more abundant following injection of dissolved CO₂. In addition, several genes associated with chemolithoautotrophy were detected, specifically those associated with the Calvin-Benson-Bassham (CBB) cycle, a key biochemical pathway for converting CO₂ to organic carbon. In the same study, the authors reported the detection of genes associated with CO₂ fixation via reductive carboxylation by heterotrophic bacteria. Importantly, Trias et al. (2017) noted that it was “impossible to assess the proportion of CO₂ that was assimilated through autotrophic and heterotrophic pathways in the aquifer,” largely due to the temporal and spatial variability observed in the microbial populations. Due to the dynamics of microbial processes over spatial and temporal scales, it is difficult to predict which specific factors (e.g., electron donors, electron acceptors, or nutrients) will limit biomass formation in full scale GCS projects.

Of the reactions listed above, hydrogenotrophic methanogenesis and homoacetogenesis are of additional importance in GCS reservoirs because they either directly or indirectly result in CH₄ formation, with acetogenesis generating a substrate for subsequent acetoclastic methanogenesis. As described previously, CH₄ production can have negative impacts on carbon storage by decreasing the solubility of CO₂. These reactions are discussed more in the next two paragraphs.

Reactions That Consume H₂

H₂ can be generated in deep formations through heterofermentation of organic compounds such as petroleum hydrocarbons, serpentinization, and radiolysis. However, there is relatively little information on H₂ fluxes (i.e., the rates of generation and consumption) in different environmental settings. Nevertheless, Tyne et al. (2023) suggest that H₂ abundance and availability are likely to be highest in hydrocarbon reservoirs and coal beds where H₂ is a product of heterofermentation. This is evidenced by the regular occurrence of hydrogenotrophic methanogenesis in these reservoirs (identified via isotopic characterization of CH₄). H₂ generation by serpentinization (i.e., the reaction of water with iron-rich minerals) occurs in basalt formations. Serpentinization has been reported to produce up to 350 mmol H₂/kg of rock. In addition, H₂ can be generated by radiolysis (dissociation of water due to radiation from radioactive decay of ²³⁵U, ²³⁸U, ²³²Th, and ⁴⁰K), though rates of radiolytic hydrogen production in the continental crust are reported to be insignificant on a GCS storage time scale.

In environments where H₂ is being generated, hydrogenotrophic microorganisms that compete for available H₂ in formation waters include homoacetogens and methanogens, as well as SO₄ ^2– and Fe­(III) reducers. The predominance of certain microorganisms and associated reaction types over others depends both on the availability of electron acceptors (CO₂, SO₄ ^2– _, Fe­(III)) and the H₂ concentration. Specifically, reactions using CO₂ as the electron acceptor are only energetically favorable in the presence of higher H₂ concentrations. For example, hydrogenotrophic Fe­(III) reduction can be energetically favorable at a low H₂ concentrations (e.g., < 1 nM) and hydrogenotrophic SO₄ ^2– reduction can be energetically favorable at somewhat higher H₂ concentrations (<5 nM) that are still too low to support methanogenesis or homoacetogenesis. As long as Fe­(III) or SO₄ ^2– are available, iron reducers and SO₄ ^2– reducers are likely to keep H₂ concentrations poised too low for hydrogenotrophic methanogenesis. However, once Fe­(III) and SO₄ ^2– are depleted, hydrogen concentrations may increase to a level (e.g., > 5–10 nM) that supports hydrogenotrophic methanogenesis. , The specific H₂ threshold concentration required to support these different reactions varies with pH due to the pH dependency of reaction energetics. Lastly, it is worth noting that iron reduction is generally favored over methanogenesis across a broad range of pH.

3. Shallow Terrestrial Groundwater Aquifers

GCS targets deep aquifer formations that are isolated from shallower formations by impermeable caprock layers. However, in order to understand potential human health and environmental risks of GCS, it is important to consider the fate of CO₂ in the event of unintended migration to shallower nontarget formations through unidentified natural fractures or artificial penetrations, such as abandoned wells (Figure ). Similar to target formations, carbon cycling within nontarget formations could serve to transform carbon into less mobile forms (e.g., CO₂(aq), carbonate minerals, or biomass) limiting the potential migration of CO₂ to the atmosphere. For the purpose of this review, “shallow” groundwater aquifers are defined as aquifers where the temperature and pressure are consistent with separate phase CO₂ as CO₂(g) rather than CO₂(sc) or CO₂(l) (i.e., formation pressure less than 30 to 74 bar, depending on temperature). Therefore, separate phase CO₂ in shallow aquifers will be present as a gas that transports upward as bubbles or is trapped beneath impermeable units. Shallow groundwater aquifers are discussed below, while unsaturated terrestrial soils and marine waters and sediments are discussed in sections and , respectively.

3.1. Environmental Conditions

Compared to many deep target formations for GCS, shallow groundwater aquifers are more likely to exhibit greater abundances of electron acceptors (O₂, NO₃ ^–, Mn­(III/IV), Fe­(III), SO₄ ^2–) and electron donors (e.g., organic material). These differences reflect the younger geologic age of shallow groundwater, which limits the time available for biogeochemical processes to deplete these constituents, and the fact that shallow aquifers are less physically isolated and therefore more biologically and chemically accessible than deeper formations. ,

Potential changes in shallow aquifer geochemistry related to CO₂ migration have been evaluated in numerous laboratory studies, , controlled-release experiments, − and studies of shallow aquifers impacted by natural CO₂ seeps. − In general, these studies reported that, similar to target formations, CO₂ migration decreases pH and increases concentrations of alkaline earth metals and alkali metals (e.g., Ca²⁺, Mg²⁺, Sr²⁺, Na⁺, K⁺), as well as alkalinity. The magnitude of pH decrease is typically on the order of 1–3 units and is governed by both the solubility of CO₂ in the formation water (which can change with depth/pressure, salinity, and the partial pressures of other gases), and the buffering capacity of the formation. For example, CO₂ solubility is lower in shallow fresh groundwater than in deeper saline aquifers. All else being equal (e.g., no difference in buffering capacity), CO₂ solubility (and associated pH decrease) is greater in deeper aquifers due to the higher pressure. Based on longer-term monitoring and laboratory experiments, water chemistry in shallow groundwater aquifers will continue to evolve following the immediate effects of CO₂ release, with some studies showing evidence of rapid return to baseline conditions, though the time frame of this rebound is likely to be highly site-specific.

In some field studies evaluating the effects of increased CO₂(aq) on shallow aquifers, increases in trace metals (As, Pb, Cd, Ba, and U) were observed, although concentrations were often close to the detection limit and/or observed increases were minimal. In addition, changes in Fe­(II) and Mn­(II) concentrations were common, although the nature of those changes (increase or decrease) was site-specific. Collectively, these changes have been largely attributed to mineral dissolution (calcite, dolomite, iron- and manganese-oxides), ion exchange, and desorption.

3.2. Microbial Community and Biotransformation Processes

Few studies have investigated the effects of CO₂ migration on the microbial community in shallow groundwater. ,− Generally, these studies report a decrease in microbial abundance and diversity following CO₂ injection, coupled with a population shift toward acidophilic and anaerobic lithoautotrophs and chemolithotrophs. For example, in a controlled release field study in Mississippi (the freshwater aquifer at Plant Daniel), Gulliver et al. reported a marked decrease in microbial diversity in response to CO₂ migration, and genomic evidence of CO₂ fixation associated with the oxidation of reduced inorganic compounds including sulfur, hydrogen, nitrogen, and iron species, as well as methanogenesis. ,, Similarly, in a study of the microbial community at a CO₂-release analog site in South Korea, Ham et al. (2017) reported lower bacterial diversity in samples from groundwater with high concentrations of CO₂(aq), although the specific microbial composition was found to primarily depend on local geochemical conditions (e.g., lower oxygen and nondetect NO₃ ^– concentrations were associated with higher relative abundances of methanogens and metal reducers). In a field study designed to simulate CO₂ migration into a shallow drinking water aquifer in the Newark Basin (United States), O’Mullan et al. (2015) found that immediately following injection, groundwater was acidified (pH decreased from slightly basic (8.2 to 8.5) to <5) and iron- and metal-reducing bacteria increased in abundance, followed by an increase in methanogen abundance during the early and midphases of injection (in this case, coinciding with decreases in SO₄ ^2–).

Carbonate precipitation in shallow terrestrial groundwater is generally limited by undersaturation with respect to the majority of carbonate minerals (calcite being the exception). Even where waters are supersaturated, similar to deep formations, there is little understanding of whether and under what conditions microbially facilitated mineral formation may be important or feasible. Nevertheless, microbially facilitated mineral formation is likely to be driven by similar microbial reactions and metabolic pathways to those described for deep formations that alter precipitation-relevant parameters (e.g., ion concentrations, pH) in microenvironments.

4. Unsaturated Terrestrial Soils

4.1. Environmental Conditions

The unsaturated zone above the groundwater table includes surface soil, and unsaturated earth materials. The capillary fringe (i.e., the depth interval right above the water table in which pores are filled with water held by capillary forces) is the boundary layer between the unsaturated soils and the shallowest aquifer. The unsaturated zone can be less than 1 m deep (i.e., very shallow) to several hundred meters or more (very deep), depending on the local depth to groundwater table. Pores within the unsaturated zone are typically filled with air and with solid particles coated by thin water films. During rainfall infiltration, these pore spaces become filled with water, which subsequently drains, and dries out. Within this zone, atmospheric air (78% N₂, 21% O₂, 0.041% CO₂, 1.8 ppm of CH₄, in dry air) is drawn in by barometric pumping and diffusion. The concentrations of these gases can subsequently be altered by processes in the subsurface; in particular, root respiration, microbial respiration, and CH₄ oxidation, increasing concentrations of CO₂, while decreasing concentrations of O₂ relative to the atmosphere.

Root respiration, in which plant roots utilize carbohydrates and O₂ to produce energy, releasing CO₂ as a product (on a 1:1 molar basis), is commonly dominant in shallow (less than approximately 10 cm) unsaturated zone layers. Microbial respiration is dominant below the shallow root zone or in areas without a shallow root zone, where soil moisture, organic matter, and O₂ are present. In this process, microorganisms utilize organic matter as an electron donor and O₂ as an electron acceptor, again producing CO₂ on a 1:1 molar basis (CH₂O + O₂ → CO₂ + H₂O; note that CH₂O is used as a proxy for organic material). When a CH₄ source is present, microbial oxidation of CH₄ consumes O₂ and produces CO₂ on a 2:1 molar basis (CH₄ + 2O₂ → CO₂ + 2H₂O). CH₄ in the unsaturated zone may originate from underlying hydrocarbon reservoirs, degradation of petroleum contamination in groundwater, and/or methanogenesis from natural organic matter in aquifers. These processes alter soil gas composition in predictable ways based on the stoichiometry of the reactions and the associated dilution or enrichment of other gas components.

Important factors that govern rates of root and microbial respiration (as well as CH₄ oxidation) include oxygen supply, nutrient availability, soil moisture content, and temperature. Such factors can vary diurnally, seasonally, and over longer-term (i.e., years or more) time scales. Nevertheless, most unsaturated zones exhibit gradual decreases in oxygen matched by increases in CO₂(g) with depth, reflecting the ongoing effects of microbial respiration and replenishment of atmospheric gases entering the unsaturated zone from ground surface.

CO₂(g) is denser than air. An additional intrusion of CO₂(g) into the bottom of the unsaturated zone, such as that from a release from an underlying storage reservoir, is likely to accumulate at the base of the unsaturated zone displacing O₂ and other atmospheric gases (e.g., N₂). An increase in CO₂(g) can also drive acidification (lower pH) of soil water, a reduction in oxidized species (where a lack of O₂ prevents oxidation), a general shift in redox conditions, and changes in soluble salts, minerals, and nutrients. − These processes may impact the microbial community composition and potential for biotransformation within the unsaturated zone although the scale and magnitude of potential impacts warrants further research or site-specific studies, as discussed further in section .

4.2. Microbial Community and Biotransformation Processes

Microbial population density in the unsaturated zone is typically highest within the organic-rich surface soil, lowest within the low-organic interval between the surface layer and the water table and somewhat higher at the capillary fringe and water table. Although lower density, the microbial populations at depth are subject to less extreme fluctuations in temperature and moisture than those in the near-surface environment, and thus deeper microbial populations are generally more stable.

The microbial distribution in the unsaturated zone is also governed by the physical properties of the matrix, as well as the amount of organic matter within the bulk matrix below the surface layer. For example, microbial densities are lower in low-porosity rocks like basalt and materials with low organic content (e.g., sands and certain silts) than in organic-rich porous materials like buried soils. Clays can have a higher amount of organic substrates, but limited permeability can restrict nutrient diffusion. Higher CO₂(aq) abundances at depth are favorable to microaerophiles (organisms that require oxygen, but lower levels of oxygen than that present in the atmosphere).

The effects of an increase in CO₂(g) (i.e., from potential migration) into the unsaturated zone on microbial activity have been modeled, inferred from natural proxies (i.e., naturally occurring CO₂ vents), − and observed during controlled injection experiments. − Many studies report that the direct and indirect impacts of CO₂ result in a significant decrease in unsaturated zone microbial population and diversity, with the surviving population adapting to anaerobic metabolism. This shift is also driven by the consumption and physical displacement of O₂.

In particular, numerous studies have reported evidence for SO₄ ^2– reduction and methanogenesis, in response to increased CO₂, where the latter was associated with acetoclastic methanogenesis, rather than hydrogenotrophic methanogenesis. ,,,,− , For example, detected transcripts of genes for the reductive acetyl-CoA pathway, found in acetogens and nonacetogens (e.g., SO₄ ^2– reducers and methanogens), are used for acetate oxidation or CO₂ reduction. The same study also detected the activity of genes implicated in the autotrophic CBB cycle, a key biochemical pathway for converting CO₂ to organic matter. These results suggest the capacity for conversion of CO₂(aq) to biomass within the unsaturated zone. However, the magnitude of this conversion is suggested to be low, as discussed below.

Several studies have noted that differences in soil pH values (which in turn, were related to differences in the CO₂ flux), as well as differences in soil properties, can significantly affect the response of the microbial community to CO₂ influx in the unsaturated zone. ,− Importantly, a few studies have cited evidence of increased biodegradation of soil organic matter in the presence of elevated CO₂. , The mechanism for enhanced biodegradation is unclear, though possibilities include an effect known as “priming”, whereby changes in microbial community in response to increased CO₂ drive the microbial mining of soil organic matter for nutrients. If this effect is common, then an influx of CO₂ into the unsaturated zone could accelerate the conversion of organic carbon stored in the unsaturated zone to CO₂, potentially resulting in an increased flux of CO₂ into the atmosphere.

Ultimately, the biotransformation of CO₂ in the unsaturated zone will be limited by low water saturation where microbes reside within the water films surrounding particles in the unsaturated zone, the availability of electron donors within that water, and the transit time of CO₂ through the unsaturated zone to surface. Because advective and diffusive transport of gas through the unsaturated zone is orders of magnitude faster than that through saturated aquifers, much of the CO₂ would likely escape to the atmosphere before biotransformation could occur. For the CO₂ that dissolves into water, potential transformation reactions include: (i) carbonate dissolution and/or formation (depending on the mineralogy), (ii) conversion to CH₄, and (iii) fixation into biomass. The latter will primarily occur via anaerobic processes, to the extent that there are electron donors present within the water film.

5. Marine Waters and Sediments

5.1. Environmental Conditions

The open ocean consists of three mixing layers: (i) a surface layer of approximately 100 m thickness, (ii) a stable thermocline layer extending from approximately 100 to 1000 m below the ocean surface, and (iii) a deep layer below approximately 1000 m. The surface layer is well mixed and exchanges with the atmosphere on a time scale of days. Mixing within the thermocline layer is inhibited by the vertical temperature and density gradient. Within this layer, exchange with the atmosphere occurs over a time scale of decades. The deep layer is isolated from the atmosphere by the thermocline layer and exchanges with the atmosphere on a time scale of 100s to 1000s of years through broad ocean circulation patterns.

The ocean comprises a majority of the earth’s surface, but the deeper portions of the ocean (>100 m) are poorly studied due to the difficulties accessing or sampling this environment. The deep ocean water is characterized by an absence of light, low temperatures (4 to 10 °C), and low nutrients. Hydrostatic pressure increases with depth and the organic matter supply decreases. Large portions of the deep ocean floor are covered with soft sediments with organic carbon and nutrients primarily supplied by the disposition of organic material originating in shallower ocean waters.

The short-term and long-term fate of potential CO₂ released into ocean water may depend on the depth of release related to the varying ocean layers, associated interlayer mixing as described above, and other factors (e.g., phase of released CO₂). For potential CO₂ released into the ocean as a gas (i.e., at a depth shallower than approximately 400 m), the gas bubbles would be less dense than the seawater and would rise. The rising bubbles would dissolve at a radial speed of about 0.1 cm/h. A potential release of gaseous CO₂ into the shallow ocean layer could either escape to the atmosphere as gas without dissolving, or could initially dissolve and then escape into the atmosphere through equilibration over a time scale of days. This relatively rapid release to the atmosphere (i.e., shorter residence time) would leave little time for biotransformation.

With a potential release of CO₂(l) into the thermocline (below approximately 400 m) or deep ocean layers, the CO₂(l) would most likely dissolve into the ocean water and would be retained within the ocean for years to centuries absent any biotic or abiotic transformation. Ocean pressure increases by roughly 1 bar every 10 m, exceeding the critical pressure of CO₂ at a depth of approximately 500 m. Below this depth, CO₂ will be a liquid (i.e., CO₂(l)). Because CO₂(l) is more compressible than seawater, CO₂(l) will be less dense than seawater between approximately 0–2700 m water depth and denser than seawater at water depths > ∼2700 m. Thus, CO₂(l) entering the ocean at depths shallower than 2700 m will rise due to buoyant forces and dissolve into the ocean water over some depth interval above the point of release. In a controlled experimental release of CO₂(l) droplets at a depth of 800 m, Brewer et al. (2002) found that 90% of the mass was lost to dissolution over a depth interval of 200 m above the release point, with the remaining 10% dissolving over the next 200 m.

CO₂(l) can also react with ocean water to form a solid CO₂ hydrate (i.e., CO₂·nH₂O). In theory, CO₂ hydrate has 5.75 water molecules for each CO₂ molecule (i.e., n = 5.75) but the actual hydration number varies and can be higher. CO₂ hydrates are denser than ocean water. However, the nature of hydrate nucleation and formation is yet to be fully understood and difficult to predict. As long as the surrounding water is not saturated with CO₂(aq), any hydrates that do form would subsequently dissolve into the water. Thus, in most potential release scenarios, hydrate formation is not likely to affect the ultimate fate of CO₂.

Depending on the magnitude of potential release, CO₂ dissolving into the ocean may disperse into the bulk ocean water with no discernible effect on dissolved CO₂ or may result in a localized plume of seawater with elevated CO₂(aq) concentrations. Water containing elevated concentrations of dissolved CO₂ (CO₂(aq)) is slightly denser than water with background CO₂ levels. − The resulting density contrast can drive downward migration until the CO₂(aq) plume reaches a level of neutral buoyancy within the thermocline layer or toward the ocean floor.

The potential release of CO₂ to the ocean from a deep geologic formation would increase CO₂(aq) concentrations in the local ocean sediment porewater through dissolution during transport through the sediments. In addition, dissolution of CO₂ into the overlying ocean waters would create a CO₂(aq) plume that, due to its density, would remain near the seabed and in contact with seabed surface sediments. Such a plume could create a larger area of shallow sediments with elevated CO₂(aq) due to downward diffusion from the overlying CO₂(aq)-rich plume of seawater.

5.2. Microbial Community and Potential for Biotransformation in Marine Waters

Much of the deep ocean water column is oxygenated and supports a diverse community of metabolically active bacteria and archaea. , Historically, it was believed that biotransformation of carbon within this environment was largely limited to heterotrophic oxidation of organic carbon originating from the shallow ocean. However, more recent research has demonstrated that carbon fixation through chemolithoautotrophy also contributes to carbon cycling within the water column. , In the water column, ammonia is an important electron donor for chemolithoautotrophy, with reduced sulfur compounds, H₂, and carbon monoxide (CO) also serving as electron donors and oxygen serving as the electron acceptor. A primary source of these electron donors (i.e., NH₃, sulfur compounds, H₂, and CO) is tied indirectly to photosynthesis in shallow ocean waters. Therefore, some chemolithoautotrophy in deep ocean waters cannot be classified strictly as primary production (i.e., creation of new organic matter) but rather as a mechanism for conversion of chemical energy from shallow waters into deep ocean biomass. However, the magnitude of CO₂ fixation through chemolithoautotrophy in deep ocean waters appears to be greater than can be explained by the flux of electron donors from the shallow ocean, suggesting that other energy sources support chemolithoautotrophy. ,

Studies on the effect of increased dissolved CO₂ on ocean microorganisms have primarily focused on shallow ocean waters (<200 m). In response to increased CO₂(aq), experiments have documented increased primary production and increases in dissolved organic matter (DOM) across a range of eutrophic (nutrient rich) and oligotrophic (nutrient poor) conditions indicating biotransformation of CO₂ into organic matter with a greater response under oligotrophic conditions. The applicability of these findings to deeper ocean waters is unclear. Model analysis and experimental studies indicate that increased CO₂(aq) and the corresponding decrease in pH results in decreased rates of ammonia-dependent chemolithoautotrophy. Thus, the potential release of CO₂ into deep ocean waters could decrease the conversion of CO₂(aq) to organic carbon via this pathway. However, the incomplete understanding of the energy sources that support deep ocean chemolithoautotrophy limit the ability to predict the effects of increased CO₂(aq) on overall rates of CO₂ fixation.

5.3. Microbial Community and Biotransformation Processes in Marine Sediments

Chemolithoautotrophic carbon fixation is generally widespread in marine sediments; however, the specific mechanisms and pathways vary by location (e.g., cold seeps, thermal vents, deeply buried marine sediments). Similar to the other settings discussed in this paper, elevated CO₂ in marine sediments will result in reduced pH and increased dissolution of CaCO₃ (if present in the sediments). A number of laboratory experiments have examined the effects of elevated CO₂ on microbial community structure and function in seabed sediments. In a 140-day experiment, Rastelli et al. (2015) found that 1000 to 20,000 ppm of CO₂ in seawater overlying marine sediments resulted in decreased heterotrophic organic carbon production (i.e., fixation of CO₂ by heterotrophs) and organic matter biodegradation within the sediments. In 40-day experiments, Łukawska-Matuszewska et al. (2023) found that 9150 μatm (approximately 9400 ppm) CO₂ concentrations in seawater (corresponding pH = 6.3) increased the release of phosphorus to the water column during mineral formation of organic matter, resulting in decreased sequestration of phosphorus in the marine sediments. Yu and Chen (2019) attribute the adverse effects of elevated dissolved CO₂ on the microbial community (i.e., changes in community structure and decreases in productivity) to the direct influences of CO₂ in addition to the effects of reduced pH.

In order to evaluate the possible long-term effects of potential CO₂ release on sediment microbial communities, a number of researchers have studied sediment effects at natural CO₂ seeps. Molari et al. (2018) found that marine sands off the Mediterranean island Panarea impacted by natural CO₂ seepage had lower faunal biomass (i.e., lower total abundance of benthic macroinvertebrates) and trophic diversity compared to reference areas. The authors also observed carbonate dissolution and an associated reduced buffering capacity in the sediments affected by CO₂ seepage. The lower biomass combined with carbonate dissolution indicated lower mass of carbon in these sediments compared to reference sediments.

CO₂ migration from a target formation into the ocean would most commonly be expressed as a seep or set of seeps on the ocean floor similar in characteristics to natural CO₂ seeps. Based on the available studies discussed above, potential CO₂ migration through marine sediments could potentially cause carbonate dissolution, decrease microbial diversity, decrease total biomass in marine sediments, and reduce or alter microbial productivity (i.e., the rate of biomass generation). Due to the limited transport time of CO₂ through sediments, the majority of CO₂ passing through such seeps is anticipated to escape untransformed into the ocean. For the fraction of CO₂ that dissolves into sediment porewater, the potential for biotransformation and specific transformation reactions are likely to be similar to those for deep or shallow aquifers discussed in sections and (e.g., possible mineral formation, CH₄ generation, or biomass generation depending on the specific environmental conditions).

6. Monitoring in Target and Nontarget Environments

Biotransformation reactions relevant to GCS can be initiated by a diversity of microorganisms and can involve a wide range of biochemical reactions and pathways. While the potential for these biotransformation reactions has been documented across the range of environmental settings reviewed here, the actual occurrence and potential magnitude of these reactions is less well understood. Thus, field-scale monitoring of CO₂ biotransformation combined with controlled laboratory-scale studies will be important for advancing the understanding of CO₂ fate associated with GCS.

Due to the complexity and interrelated nature of carbon cycling in the subsurface, when developing laboratory- and field-scale studies, a multiple lines of evidence framework is needed to accurately characterize biotic and abiotic transformation processes. Measurement of physical, chemical, and biological parameters will likely be needed to support the identification and quantification of (bio)­transformation processes. Similar frameworks have been developed and utilized at the field-scale to achieve regulatory requirements for measuring the fate of chemical contaminants in shallow subsurface settings. −

A general framework for measuring biogeochemical processes involved in GCS can be conceptualized by leveraging these prior frameworks. , Some parameters to be considered for laboratory-scale and/or field-scale GCS biogeochemical studies include:

• 1.Physical Parameters that impact the phase and/or residence time of CO₂: temperature, pressure, salinity, partial pressure of other gases (which effects CO₂ solubility), and advective or convective flow parameters (e.g., formation effective porosity, formation fluid density, injected CO₂ density).
• 2.Chemical and Geochemical Parameters: Nutrient concentrations (compounds containing nitrogen, phosphorus, and sulfur) can be evaluated to determine potential limitations for microbial growth. Measurement of O₂, NO₃ ^–, NO₂ ^–, Mv­(IV) and Fe­(III) oxides, Mn (II) and Fe­(II), SO₄ ^2–, H₂S, CH₄, short chain fatty acids (SCFAs: formate, acetate, propionate, butyrate), H₂, and oxidation reduction potential (ORP) provides an understanding of electron donor and acceptor utilization. pH, dissolved inorganic carbon (DIC), and alkalinity inform buffering capacity and CO₂ speciation.
• 3.Biological Parameters: A range of molecular biological tools (MBTs) can be used for measurement of taxonomic groups, functional genes, and specific biotransformation products. These include quantitative polymerase chain reaction (qPCR), stable isotope probing (SIP), compound-specific isotope analysis (CSIA), and omics techniques (e.g., metagenomics, transcriptomics, metabolomics, proteomics). ,, For many studies, application of multiple tools may provide the most complete understanding of the microbes and biochemical pathways involved in carbon cycling.

While monitoring data is important to better understand the fate of sequestered carbon, intrusive monitoring in the deep subsurface (i.e., target formation) is likely to be challenging. Artificial penetrations into the target formation (i.e., wells or borings) for monitoring and sample collection have the potential to create additional pathways for CO₂ transport from the target formation. The risk-benefit of intrusive monitoring programs for the deep subsurface should be carefully evaluated. Thus, the application of such monitoring frameworks may be more apt for nontarget environment monitoring (e.g., shallow groundwater).

7. Implications, Knowledge Gaps, and Research Opportunities

7.1. Implications of CO₂ Biotransformation in Target and Nontarget Environments

An important component of understanding the human health and environmental risks of GCS is understanding biotransformation of CO₂ both (i) within formations targeted for GCS and (ii) within nontarget environments in the event of potential migration from target formations. This review has identified commonalities across the environments evaluated including:

• 1.Diverse microbial communities have been documented in each setting capable of a wide range of biochemical processes that may influence the short-term and long-term fate of CO₂. Relevant potential processes include biogas formation (e.g., CH₄, H₂S), generation of biomass, and microbially facilitated mineral formation. Of these processes, the potential for microbially facilitated mineral formation is generally least well understood.
• 2.The addition of CO₂ to subsurface environmental settings results in predictable geochemical changes including: (i) increased CO₂(aq) concentrations, (ii) initially decreased pH, and (iii) increase in alkalinity, rock dissolution, and liberation of divalent metal ions. These changes, in turn, drive adaptation of microbial communities toward community members that can tolerate or benefit from the altered conditions.
• 3.Under low pH conditions, biochemical reactions that consume electrons (e.g., Fe­(III) reduction) are more energetically favorable. This favors microbial populations that can utilize these biochemical reactions such as metal-reducing chemolithoautotrophs and hydrogenotrophic methanogens. , These reactions consume CO₂(aq) and raise the pH.
• 4.The magnitude of biotransformation (i.e., the percentage of CO₂ that is transformed) will depend, at least in part, on the residence time of CO₂ within the target or nontarget environmental setting. For example, in a confined aquifer, separate phase CO₂ (i.e., CO₂(sc) or CO₂(g)) will be trapped at the top of the aquifer allowing an extended time for dissolution. In contrast, in unsaturated terrestrial soils, separate phase CO₂ has a more direct pathway to the atmosphere. Within confined aquifers, the bioavailability of CO₂ will also be important; specifically the partitioning between separate phase (i.e., CO₂(sc) or CO₂(g)) and CO₂(aq).

Despite this level of understanding, it is not yet possible to predict the occurrence or magnitude of biotransformation reactions in response to CO₂ addition, or the extent to which these biotransformation reactions will influence or drive short-term and long-term fate of CO₂ within these settings.

7.2. Knowledge Gaps and Research Opportunities

There are several important questions that remain to be addressed with respect to biotransformation of CO₂ in target GCS reservoirs and nontarget environments.

7.2.1. Factors That Underpin Biotransformation Potential and Influence Carbon Fate

• 1.Knowledge Gap 1, Magnitude of Biotransformation: While the potential for CO₂ biotransformation is well documented, the actual occurrence, rates, and magnitude in target storage formations and nontarget environments is poorly understood.

  Research Opportunity: The current state of knowledge does not support identification of the specific factors controlling the fate, direction of transformation (e.g., gas phase products such as CH₄ and H₂S, solid phase minerals, and/or biomass), and/or the overall magnitude of transformation in target and nontarget environments. Understanding the physical (temperature, pressure), chemical (nutrients, pH, electron donors, electron acceptors) and biological (absence of microbes or biochemical pathways) limitations is key to understanding how to limit unfavorable reactions and enhance favorable transformations. These data will support an improved understanding of reaction rates (i.e., fluxes), which are crucial for understanding the balances between competing reactions and systems modeling approaches.

• 2.Knowledge Gap 2, Biotic-Abiotic Interactions: There is limited understanding of the linkage between biotic and abiotic transformation reactions in target and nontarget environments.

  Research Opportunity: Many abiotic (i.e., chemical) reactions depend on reactants generated by microbial activity. Predictive understanding of abiotic transformation reactions requires detailed understanding of microbial processes that generate essential reactants. Laboratory and field studies that document the interaction of microbial activity and abiotic transformation reactions will enhance the ability to predict the overall fate of sequestered CO₂·

• 3.Knowledge Gap 3, Dissolution of Separate-Phase CO₂: Separate phase CO₂ (CO₂(g), CO₂(l), and CO₂(sc)) is not bioavailable. Within target and nontarget aquifers, the time scale for dissolution is poorly understood.

  Research Opportunity: Improved knowledge of the specific mechanisms (e.g., convective mixing) and the time scales for dissolution of CO₂ is needed to better understand the bioavailability of CO₂ and, thus, the potential for biotransformation.

7.2.2. Biotransformation Products with Implications on Carbon Fate

• 1.Knowledge Gap 4, Microbially Facilitated Mineral Formation: While the phenomena of microbially controlled, microbially induced, and microbially influenced mineral formation are well documented, a detailed understanding of the biogeochemical processes involved is lacking. In addition, in many situations, the benefits to the microbes (if any) of mineral formation is not understood.

  Research Opportunity: A better understanding of the specific mechanisms of microbially controlled, microbially induced and microbially influenced mineral formation is important to better understand the potential for these processes to drive mineral formation in target and nontarget formations. Several specific enzymes, such as urease and carbonic anhydrase, have been identified as mediators of microbially induced mineral formation. Enumeration of biomarker genes can characterize the prevalence of these processes in the microbial communities in target and nontarget environments. As the specific role of EPS in microbially facilitated mineral formation is better understood, the presence or absence of relevant EPS in target and nontarget environments can be better characterized. Understanding the benefits to the microbial community of microbially facilitated mineral formation is critical to developing methods to enhance mineral formation in target and nontarget environments. Finally, additional experimental work is needed to evaluate the recalcitrance (or lability) of minerals formed by microbially facilitated processes compared to abiotic processes.

  Better understanding microbially facilitated mineral formation is important for predicting where within target formations mineral formation may occur. Mineral formation within the target formation during the period of CO₂ injection (especially at the near wellbore region) has the potential for reduced formation permeability thus reducing overall storage capacity.

• 2.Knowledge Gap 5, Generation of Biogas: The potential for significant generation of biogas (e.g., CH₄ and H₂S) and the short- and long-term fate of such gas is not understood.

  Research Opportunity: The formation of biogas may affect formation pressure, partitioning of other soluble gases, and other parameters related to CO₂ fate. The formation of biogas depends on the microbial community, the availability of reactants, competition with other potentially more energetic reactions, and other factors. Understanding potential scenarios and probability for biotransformation of CO₂ to generate CH₄ and/or H₂S generation within target formations or in nontarget environments (e.g., shallow terrestrial groundwater) is important to understand the reliability and potential hazards associated with GCS.

• 3.Knowledge Gap 6, Generation of Biomass: The potential for significant generation of biomass (including biofilms) and the short- and long-term fate of such biomass is not well understood.

  Research Opportunity: Understanding nutrient availability and other limiting factors discussed above will inform the potential for generation of biomass. Additional research is needed to understand the fate of new biomass. In the short term, biofilms may reduce formation permeability and, as a result, reduce or alter formation storage capacity and/or reduce injectivity into the formation. In the longer term, the potential for additional biotransformation of this additional biomass is not well understood. In deep oceans, refractory dissolved organic carbon (generated through cycling of biomass) is a significant reservoir of sequestered carbon. In target and nontarget aquifers, it is not known what fraction of new biomass would persist as active biomass, form refractory organic carbon, or serve as reduced carbon electron donor for other biotransformation reactions.

7.2.3. Application of Knowledge for Risk Assessment and Management

• 1.Knowledge Gap 7, Monitoring Tools: While there are a variety of tools for the general purpose of monitoring biotransformation reactions (see Section 6), these tools need to be validated for the purpose of monitoring relevant CO₂ and organic carbon biotransformation in target and nontarget environments.

  Research Opportunity: A robust toolbox is needed for identifying and monitoring specific biotransformation processes. This toolbox must account for the challenge of biological monitoring at depth, both in terms of costs and practicability. Relevant technologies for monitoring CO₂ biotransformation may include tools for microbial characterization that provide rapid and more complete characterization of the microbial community, genetic assays (i.e., MBTs) for specific enzymes in key biotransformation pathways, and stable isotope analysis of reactants and products. , Once relationships between microbial taxa or functional genes, and CO₂ biotransformation reactions can be more firmly established, it may be possible to develop field-portable rapid biosensors for simple microbiological signals that can be widely implemented at a variety of CCS sites.

  Once developed, a toolbox for identifying and monitoring specific biotransformation processes could be refined for applicability to specific target and nontarget environments. Based on knowledge of how microbial communities and biogeochemical processes respond to the addition of CO₂, a subset of tools may be useful for identification of releases into nontarget environments.

• 2.Knowledge Gap 8, Modeling: Current reservoir models typically simulate the transport of injected CO₂ but are unable to account for either biotic or abiotic transformation reactions. This limits the ability to understand how transformations may affect the long-term fate of injected CO₂.

  Research Opportunity: GSC reservoir models that can account for transformation of CO₂ would provide an improved understanding of likely GCS robustness. Models might draw on existing petroleum hydrocarbon models (e.g., oil spill response and reservoir souring models). These models need to account for both complex transport processes (e.g., density-driven flow) and complex interactions between biotic and abiotic reaction processes and the effect of site geochemistry (e.g., pH) on these processes.

• 3.Knowledge Gap 9, Microbially-Induced Corrosion: MIC has significant impact during current oil and gas operations and could potentially affect the integrity of CO₂ pipelines under certain circumstances, however understanding the consequence and probability of MIC for CO₂ pipelines is needed (see Section S5 of Supporting Information).

  Research Opportunity: Further evaluation is needed of MIC in CO₂ pipelines under atypical conditions, such as low or stagnant flow conditions, and on infrastructure outside the influence of the injection zone and after CO₂ is injected. Abiotic corrosion will likely be the dominant corrosion mechanism during normal operations at locations where water may drop out. However, MIC could pose a risk for pipeline integrity at locations of low or stagnant flow (i.e., dead legs) and it is currently not known how CO₂(sc) would affect already established biofilms. In addition, MIC might not be important to sites where CO₂ is injected due to the likely low pH but MIC could become a potential problem for monitoring wells and verification wells further away from injection zones where pH may be higher. Furthermore, there is uncertainty regarding the long-term impacts on steel infrastructure by microorganisms once CO₂ injection has ceased.

• 4.Knowledge Gap 10, Assimilative Capacity in Nontarget Environments: The potential and likely magnitude of biotransformation (and abiotic transformation) of CO₂ in nontarget environments that could aid or hinder assimilative capacity is not well understood.

  Research Opportunity: In the event of potential migration of CO₂ into nontarget formations, biotic and abiotic transformations could result in long-term storage in dissolved, mineral, or other phases, which may or may not aid or hinder assimilative capacity. Laboratory and field studies focused on biotransformation in nontarget environments (shallow aquifers, unsaturated zone, marine waters and sediments) would provide an improved understanding of potential for natural attenuation, resource recovery, as well as human health and environmental risks of GCS.

• 5.Knowledge Gap 11, Risks to Health and the Environment: The effect of CO₂ biotransformation processes on the risk of adverse impacts on human health and environmental receptors associated with GCS is not well understood.

  Research Opportunity: Potential release of CO₂ or potentially associated gases (e.g., CH₄ or H₂S) from the target formation or during CO₂ transport could result in direct adverse impacts to human health or environmental receptors or could have indirect adverse impacts (e.g., mobilization of metals). Research is needed to determine how biotransformation reactions may enhance or mitigate these risks.

8. Literature Curation and Search Methods

Scientific publications relevant to the review topic were identified primarily using Google Scholar search engine and database. A variety of search terms were used singly and in combination. Search terms included carbon dioxide, super critical, sequestration, geologic storage, biotransformation, biomineralization, biocorrosion, microbially influenced, abiotic, reactions, transformation, microbes, microbial, bacteria, chemolithoautrophy, oxidation, reduction, ocean, marine, aquifer, sediment, unsaturated zone. Publications identified through Google Scholar were manually curated and supplemented with publications present in the files of the authors and publications cited in publications obtained using these two approaches.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.est.5c02389.

• Additional material (PDF) including: S1: Carbon biotransformation processes – overview of microbial pathways that generate or consume CO₂; S2: Microbially facilitated mineral formation – definitions of microbially controlled, induced, and influenced mineral formation; roles of biofilms, rock dissolution, and metal mobilization in mineral formation; S3: Methanogenesis – overview of microbial methane production pathways and their effects on CO₂ solubility and storage in subsurface formations; S4: Generation of biomass – processes by which subsurface microorganisms convert CO₂ and inorganic compounds into organic biomass; S5: Microbially mediated carbon transformation during transport/injection – discussion of carbon transformation from the point of capture to the location for injection into the target formation (PDF)

The authors declare the following competing financial interest(s): S.L., K.M.M., P.G.K.v.G., L.C., and T.A.K. are employed by a company that offers carbon capture and storage as a low carbon solution. The manuscript was written as part of normal employment and is the sole responsibility of these coauthors in collaboration with other affiliated coauthors.