Greenhouse gas fluxes of microbial‐induced calcite precipitation at varying urea‐to‐calcium concentrations

Microbial‐induced calcite precipitation (MICP) is regarded as environmentally friendly, partly due to the storage of carbon as carbonates. Although CO2 emissions during MICP have been reported, quantification of its environmental impact regarding total greenhouse gas fluxes has not yet been thoroughly investigated. In particular, N2O fluxes could occur in addition to CO2 since MICP involves the microbially mediated nitrogen cycle. This study investigated the greenhouse gas fluxes during biostimulation of MICP in quartz sand in incubation experiments. Soil samples were treated with MICP cementation solution containing calcium concentrations of 0, 20, 100 and 200 mM at a fixed urea concentration of 100 mM to offer a range of carbonation potential and/or mitigation of CO2 emissions. Greenhouse gas (CO2, CH4 and N2O) measurements were determined by gas chromatography during incubations. Soil total inorganic carbon and the isotopic composition of precipitated and emitted CO2 were determined by isotope ratio mass spectrometry. CO2 emissions (0.52 to 4.08 μg of CO2–C h−1 g−1 soil) resulted from MICP, while N2O and CH4 fluxes were not detected. Increasing Ca2+ with respect to urea resulted in lower CO2 emissions, lower solution pH, similar carbonate precipitation and urea hydrolysis inhibition. The highest urea‐to‐calcium ratio (1:0.2) emitted roughly two times the amount of CO2 (112 μg of CO2–C g−1 soil) compared to the 1:1 and 1:2 ratios (47 to 58 μg of CO2–C g−1 soil) and five to six times more than samples that did not receive Ca2+ (1:0) (~18 μg of CO2–C g−1 soil). Precipitated CaCO3–C was tenfold higher than cumulative emitted CO2–C, and isotopic analysis indicated both emitted and precipitated carbon were of urea origin. Both emitted and precipitated carbon accounted for a very low percentage of total carbon applied in the system (<0.35 and <4.5%, respectively), presumably due to limited urea hydrolysis which was negatively affected by increasing the Ca2+ concentration.

emissions, lower solution pH, similar carbonate precipitation and urea hydrolysis inhibition.The highest urea-to-calcium ratio (1:0.2) emitted roughly two times the amount of CO 2 (112 μg of CO 2 -C g À1 soil) compared to the 1:1 and 1:2 ratios (47 to 58 μg of CO 2 -C g À1 soil) and five to six times more than samples that did not receive Ca 2+ (1:0) ($18 μg of CO 2 -C g À1 soil).Precipitated CaCO 3 -C was tenfold higher than cumulative emitted CO 2 -C, and isotopic analysis indicated both emitted and precipitated carbon were of urea origin.
Both emitted and precipitated carbon accounted for a very low percentage of total carbon applied in the system (<0.35 and <4.5%, respectively), presumably due to limited urea hydrolysis which was negatively affected by increasing the Ca 2+ concentration.

K E Y W O R D S
carbon dioxide, carbon stable isotopes, carbonate, CO 2 emissions, greenhouse gas fluxes, MICP In the current context of human-induced global warming (IPCC, 2021), environmental pollution, and biodiversity decline (IPBES, 2019), the environmental impact of current and future anthropogenic activities requires careful consideration.Over the past 25 years, microbial-induced calcite precipitation (MICP) via urea hydrolysis has been developing as an alternative to traditional approaches in soil and environmental engineering, mainly in soil stabilisation, manufacturing (e.g., bricks) and repair (concrete fissures) of construction materials, and bioremediation of contaminants in soil (Al Qabany et al., 2012;Castro-Alonso et al., 2019;DeJong et al., 2006;Fang et al., 2021;Fujita et al., 2004;Montoya et al., 2014;Stocks-Fischer et al., 1999;Warren et al., 2001;Whiffin et al., 2007;Zamani & Montoya, 2018).
The construction sector is one of the largest energy consuming (36%) and carbon dioxide (CO 2 ) emitting (39%) sectors globally (IEA, 2020).Manufacturing of materials (e.g., cement) accounts for 11% of total CO 2 emissions of the construction sector (IEA, 2020).In particular for soil engineering, cement and chemical binders are widely used for soil stabilisation (Chang et al., 2019), with ground improvement processes (mixing and grouting) contributing to 0.2% of global CO 2 emissions (Chang et al., 2016).Strategies to minimise CO 2 emissions associated with soil engineering practices are therefore necessary.MICP is also relevant in the context of soil science since the biogeochemical process of urea hydrolysis is the same as that occurring in agricultural settings with the application of nitrogen (N) fertilisers such as urea.The agricultural sector is estimated to be responsible for up to 8.5% of global greenhouse gas (GHG) emissions (IPCC, 2019).Synthetic nitrogen fertiliser is estimated to account for 1129.1 ± 171.1 Mt CO 2e emissions, $38% from industrial manufacturing and 58% from direct (CO 2 ) and indirect (N 2 O) emissions associated to urea application to agricultural soils worldwide (Menegat et al., 2022).
In the literature, MICP is widely regarded as a 'sustainable' and 'environmentally friendly' technique compared to traditional approaches relying on carbon intensive materials (e.g., involving cement), although there is a lack of data on the environmental impact and sustainability of the technique.Several aspects favour this perception: MICP is a low energy technique since low viscosity of treatment fluid (water) greatly reduces the requirement for high injection pressures; the chemicals used are not associated with carbon intensive industries such as the cement industry, with global contribution to CO 2 emissions of 5 to 7% (Benhelal et al., 2013); and the technique relies on naturally occurring biogeochemical processes.Nonetheless, production of calcium chloride and urea, the main components of MICP via urea hydrolysis, is associated with energy intensive processes fuelled by combustion of fossil fuels.Calcium chloride is produced by combustion of calcium carbonate rocks (e.g., limestone) which requires temperatures above 700 C, releasing 440 g of CO 2 per kg of CaCO 3 breakdown excluding emissions derived from energy input.Urea is produced in two steps where, first, natural gas largely containing methane is partially oxidised with O 2 to produce CO 2 (900-1200 C, 40-100 bar), and H 2 is combined with N 2 to produce NH 3 through the Haber-Bosch process.The second step combines NH 3 and CO 2 to produce urea (Pagani & Zardi, 1995).A recent life cycle assessment (LCA) of MICP assumed that all carbon introduced in the system is precipitated as calcium carbonate (Deng et al., 2021).However, it is likely that MICP in field case scenarios does not achieve complete conversion of urea-C into precipitated calcium carbonate.This constitutes a gap in the literature as data on soil-atmosphere GHG fluxes of MICP is scarce.
MICP via urea hydrolysis is a heterotrophic microbial pathway which produces CO 2 and NH 3 gases as by-products of microbial activity.Okyay et al. (2016) and Okyay and Rodrigues (2015) investigated biotic and abiotic CO 2 fluxes during MICP in aqueous samples obtained from travertine cave environments in incubation experiments at a concentration of 10% CO 2 in the vial headspace.They posed the relevant question of whether the amount of CO 2 produced by bacterial metabolism would exceed the bacterial capability to sequester CO 2 through MICP.Results of biotic experiments highlighted CO 2 sequestration (0.9% to up to 8.6% of CO 2 in vial headspace), but also CO 2 emissions, the net balance being directly dependant on microbial composition.Abiotic experiments elucidated concomitant CO 2 sequestration due to the alkalinity of solution pH, composition and nutrient concentration in treatment media.In a former study, we investigated soil-atmosphere CO 2 gas fluxes over 2 months during and following biostimulation of MICP on a quartz sand using calcium chloride or dolerite fines (basaltic rock composed of calcium-rich minerals by-product of the quarrying sector) as sources of calcium for MICP.Our results highlighted net CO 2 emissions during and following MICP treatment, which were dependant on reaction time and soil water saturation conditions (i.e., submerged vs freely drained) (Casas et al., 2020).Our experiment was conducted at a fixed urea-to-calcium molar ratio (100 and 20 mM, respectively).Because the molar content of calcium was smaller than urea-C, calcium could have limited the extent of soil carbonation and the capacity of the system to balance CO 2 emissions.An unexplored aspect of MICP is that release of NH 3 from urea degradation into the soil solution presents the potential occurrence of nitrite, nitrate, and nitrous oxide (N 2 O) emissions through microbial nitrification and denitrification processes.Excess of nitrate is widely recognised as environmentally detrimental to underground water quality and life due to eutrophication, while the estimated global warming potential of N 2 O is 265 times that of CO 2 over a 100-year period (IPCC, 2019).Studies on potential nitrification following MICP have received little attention, despite Gat et al. (2016) observing ammonia oxidation (production of nitrate) nearly a month following MICP treatment, demonstrating the potential for occurrence of N 2 O emissions.
To expand the work conducted by Okyay et al. (2016), Okyay and Rodrigues (2015) and Casas et al. (2020), in this study we present an investigation on the GHG (CO 2 , N 2 O and CH 4 ) fluxes during biostimulation of MICP in a quartz sand in incubation experiments.The aims were, on the one hand, to determine whether N 2 O gas emissions could result from MICP and, on the other hand, to study the dynamics of CO 2 fluxes during MICP under different conditions to determine whether CO 2 emissions could be reduced, balanced, or whether MICP in soil could act as a net sink of CO 2 .To do so, we varied the urea-to-calcium molar ratio by varying calcium and fixing urea concentration in treatment solution to provide a range of soil carbonation and CO 2 sequestration potential, under the hypotheses that with increasing calcium concentration more CaCO 3 would precipitate, and less CO 2 would be emitted.In addition, we studied the abiotic effect of alkaline pH, induced by urea hydrolysis, on CO 2 sequestration versus microbial-derived CO 2 emissions.Furthermore, the isotopic composition of precipitated and emitted CO 2 were investigated to elucidate the origin of precipitated and emitted CO 2 .

| Soil sampling and characterisation
Sand was obtained from a quarry located in the Lower Rhine Basin, Germany (50 55 0 018.8184 00, 06 46 0 45.6528 00 , WGS84; operated by Quarzwerke GmbH, Frechen).The sampling location was selected close to vegetation (Figure S1) to ensure the soil was biologically active.The first 3 cm of the surface soil were removed, and the underlying soil was collected to a depth of 20 cm and sealed to prevent moisture loss.The sample consisted of a loose, homogeneous, white-greyish fine sand, with no presence of gravels, boulders, or fines.The soil water content was determined gravimetrically upon arrival at the laboratory.The grain size distribution was determined by wet sieving on two analytical replicates (ISO 11277: 1998), and the soil pH was determined in 0.01 M CaCl 2 in distilled water solution at a soil-to-solution ratio of 2.5:1 (Houba et al., 2000) on three analytical replicates (WTW SenTix 41 PLUS probe with WTW multi 340i pH meter, WTW, Weilheim, Germany; calibrated to pH = 4 and 7).

| MICP treatment solutions
MICP was induced through biostimulation of soil indigenous bacteria.The growth solution contained 1 g L À1 of cane molasses (MLS) (Rapunzel Naturkost GmbH, Germany), 0.1 g L À1 yeast extract (Vitasan Bio-Hefeextrakt; VITAM Hefe-Produkt GmbH, Germany), 100 mM anhydrous sodium acetate (CH 3 COONa) (ACS, Merck KGaA, Germany) and 250 mM urea (≥99.5%,Carl Roth GmbH C Co., KG, Germany) in distilled water.The cementation solution contained 0.1 g L À1 of MLS, 100 mM sodium acetate, 100 mM urea and 0, 20, 100 or 200 mM calcium chloride dihydrate (CaCl 2 Á 2H 2 O) (ACS, Merck KGaA, Germany) in distilled water.Ammonium was excluded from treatment solution as it has been reported not to be necessary for growth of ureolytic bacteria Sporosarcina pasteurii (Lapierre et al., 2020), and was shown not to prevent urea hydrolysis by indigenous bacteria by a previous study on this soil (Casas et al., 2020).

| MICP experiments
Greenhouse gas fluxes during MICP at varying ureato-calcium ratios were investigated in an incubation experiment by gas chromatography.Treatments included distilled water (0:0) (control), MICP treatment solutions with varying concentrations of urea-C to calcium (Ca 2+ ) molar ratios (1:0, 1:0.2, 1:1 and 1:2), and a sodium hydroxide (NaOH) solution of adjusted pH of 9.5 used to study the abiotic GHG fluxes at the pH levels induced by the urea hydrolysis reaction (NaOH).Three replicate soil samples were prepared for each treatment by adding a constant mass of 1.5 g wet soil (m dry = 1.1637 ± 0.0004 g; θ = 22.4%; n = 18) in 22-mL gas chromatography vials.Solution was applied at a soil-solution ratio of 1:1.
The treatment sequence comprised a growth and a cementation phase.Initially, treatments that included urea (i.e., urea-to-calcium molar ratios 1:0, 1:0.2, 1:1, 1:2) received the growth solution, while control (0:0) and the abiotic (NaOH) treatments received distilled water.Aliquots were decanted after a reaction time (t r ) of 96 h.The cementation phase started 2 h after decantation and comprised five cementation solution (i.e., 0:0, 1:0, 1:0.2, 1:1, 1:2 and NaOH) applications.The first four applications were performed every 24 h on consecutive days.After each application, vials were sealed with butyl rubber septa and aluminium caps and incubated for a reaction time (t r ) of 10 h.After each incubation, vials were opened, aliquots decanted, and vials were stored at room temperature covered with aluminium foil until the next cementation solution was applied.The fifth cementation solution was applied 2 h after finalising the fourth incubation and was allowed to react for 24 h, where the incubation for the GC measurement lasted 10 h from t r = 14-24 h.At the end of the incubation experiment, vials containing soil samples were oven-dried to a constant mass at 105 C for more than 24 h.Soil samples were subsequently obtained for quantification of soil total inorganic carbon and stable isotopes of soil carbonates (see Section 2.5).
The experiment was replicated in 22-mL vials using 5 g of soil (m wet ) to determine the pH of the soil aqueous phase simultaneously with gas measurements (see Section 2.4).An additional replicate experiment was run in 12-mL vials with screw-top exetainers sealed with septa using 0.1 g soil (m wet ) to determine the isotope composition of evolved CO 2 -C in the vial headspace.A single gas measurement was carried out at the end of each 10 h incubation.The isotope analysis of evolved CO 2 was conducted as specified in Section 2.5.

| Gas measurements
A gas chromatograph (GC Clarus 590, PerkinElmer, Rodgau, Germany) equipped with an Elite Plot-Q column (20 m Â 0.53 mm id Â 20 μm), a flame ionisation detector (FID) and an electronic capture detector (ECD), interfaced to an autosampler (Turbo Matrix 110, PerkinElmer) was used to determine CO 2 , CH 4 and N 2 O concentration in the vial headspace during incubations using nitrogen as a carrier gas.
Gas measurements of the first four cementation applications were taken at t r = 1, 4, 7 and 10 h and during the fifth application at t r = 15, 19, 21 and 24 h.During the 10-h incubation time, four gas measurements were made on the same vial.The sampling procedure of the GC involved injecting N 2 into the vial until an overpressure of 1 bar was achieved prior to sampling 10% of the headspace volume.Therefore, the pressure drop within the vial was constant, whereas the amount of air remaining in the vial after each measurement was 90% of the air prior to the measurement.Gas concentration measurements were corrected by calculating the gas concentration equivalent to the initial amount of air as follows: where C ½ corr is the corrected gas concentration of the i gas (ppm or ppb), C ½ meas is the measured gas concentration of the i gas (ppm or ppb), f air is the fraction of air remaining after each measurement, equal to 0.9, and m is an integer that represents the gas measurement number and takes values of 1 for the first measurement, 2 for the second, and so on (for further details see Table S3).Corrected gas concentrations were used to calculate variations in gas concentrations in the vial headspace and, subsequently, gas fluxes.To compute gas fluxes, variations in gas concentration data were fitted to linear and convex quadratic equations.The polynomial fit was selected over the linear if the following three conditions were met: the coefficient of determination of polynomial fit (R 2 pol ) was >0.8, the quadratic fit was convex (a > 0), indicating saturation of the gas in the vial headspace, and R 2 pol > R 2 lin .Cumulative emissions were calculated by multiplying the gas fluxes obtained from the linear or polynomial regressions by the incubation time (t = 10 h).

| Quantitative and isotopic analysis of carbon
Total organic carbon (TOC), total nitrogen (TN) and the isotopic signature of organic carbon (δ 13 C TOC ) in soil and carbon-containing compounds used for the preparation of growth and cementation solutions (i.e., molasses, urea, and sodium acetate) were determined in triplicate by elemental analysis-isotope ratio mass spectrometry (EA-IRMS) using an elemental analyser (EA; Flash 2000, Thermo Fisher Scientific, Bremen, Germany), interfaced with a continuous flow IRMS (CF-IRMS; Delta V Advantage, Thermo Fisher Scientific).Soil total inorganic carbon (TIC) and the C and O isotopic composition of precipitated carbonate (δ 13 C TIC and δ 18 O, respectively), and evolved CO 2 (δ 13 C CO2 ) in vial headspace during incubations were determined by gas chromatography-continuous flow-isotope ratio mass spectrometry (GC-CF-IRMS, GasBench II interfaced with a Delta V Advantage, Thermo Fisher Scientific).Soil TOC and TIC were determined in separate runs.For analysis of soil TOC, TIC was previously removed by fumigation of pre-weighed samples with concentrated hydrochloric acid vapour (Ramnarine et al., 2011).Soil TIC was determined by analysis of evolved CO 2 gas released by the reaction of carbonate and pure phosphoric acid (Debajyoti & Skrzypek, 2007).Evolved CO 2 in the vial headspace was estimated by comparing peak areas of samples against air samples.Certified isotopic standards were used for 13 C, referenced against the Vienna Pee Dee Belemnite (VPDB) scale.

| Material characterisation
The soil used in this study was the same as used in Casas et al. (2020).The soil classified as poorly graded sand (SP, Unified Soil Classification System) with 29% medium, 67% fine sand, and fines content <4% (Figure S2).Soil organic C (C TOC = 0.0211 wt%) and carbonate (C TIC = 0.0003 wt%) content were very low, the C:N ratio was 13 and the soil pH was slightly acidic (pH = 6.4).The isotopic signature of soil inorganic and organic C was δ 13 C TIC = À13.9‰and δ 13 C TOC = À25.7‰,respectively (Table S1).C-containing compounds used for the preparation of the growth and cementation solutions had a similar organic carbon content (C org = 20-32%), while the C org isotopic signature differed (δ 13 C urea = À40‰, δ 13 C Na-Ac = À31‰ and δ 13 C MLS = À13‰) (Table S2).

| Solution pH
Soil solution pH data are presented in Figure 1.At the end of the growth phase (t = 0), pH ranged from 8.9 to 9.2 for all treatments containing urea, indicating similar production of ammonium ions through urea hydrolysis by soil indigenous ureolytic microbial communities.The pH of control samples (0:0) was generally between 6.2 and 6.7, similar to the untreated quartz sand (pH = 6.4,Table S1).Abiotic samples (NaOH) showed stable pH values of about 9 throughout, a decrease of 0.5 units compared to the NaOH solution (pH = 9.5).The pH of MICP treatments (1:0, 1:0.2, 1:1 and 1:2) was alkaline, ranging from 7 to 9. Lower pH was measured with increasing Ca 2+ concentration: samples that did not receive Ca 2+ (1:0) showed a stable value at a pH of about 9 during incubations, similar to NaOH samples.Instead, samples that received Ca 2+ (1:0.2, 1:1 and 1:2), showed a decreasing pH trend, stabilising at lower pH with increasing Ca 2+ (i.e., approx.8, 7.5 and 7.3 for urea-to-calcium molar ratios 1:0.2, 1:1 and 1:2, respectively).pH dropped during reaction time t r = 1 to 4 h and flattened from t r = 4 to 10 h.pH drops were less pronounced as the number of cementation solutions applied increased.For example, during the fourth incubation period (t = 72-82 h), the pH showed flatter trends than during the first incubation period (t = 0-10 h).Flattening of pH trends was observed earlier with increasing the Ca 2+ content: for 1:0.2 samples, F I G U R E 1 Soil solution pH during biostimulation of MICP in quartz sand treated with distilled water (0:0, control), ureato-calcium molar ratios 1:0, 1:0.2, 1:1, 1:2, and solution of adjusted pH of 9.5 with sodium hydroxide (NaOH).pH measurements conducted over 10 h incubation periods at reaction time (t r ) of 1, 4, 7 and 10 h during the first four incubations (t = 0-82 h), and at t r = 15, 19, 21 and 24 h during the fifth incubation (t = 84-108 h), coinciding with greenhouse gases measurements in parallel experiment.Markers and error bars indicate the average and standard deviation of three replicate samples.
flattening was observed during the fourth incubation period (t = 72-82 h), while for 1:2 samples flattening occurred during the second incubation period (t = 24-34 h).An increasing trend in pH was recorded for urea-to-calcium molar ratio of 1:0.2 during the fifth incubation period (t r = 14 to 24 h), where pH increased over time, while treatments with higher Ca 2+ content (1:1 and 1:2) showed a stable pH.
during biostimulation of MICP in fine quartz sand after application of solutions with different urea-to-calcium molar ratios (1:0, 1:0.2, 1:1 and 1:2), distilled water (0:0), and distilled water with pH of 9.5 adjusted with sodium hydroxide (NaOH) calculated from headspace concentrations of CO 2 , N 2 O and CH 4 , measured by gas chromatography, during incubations in gas-tight vials at t = 1, 4, 7 and 10 h.Markers and error bars indicate the average and standard deviation of three replicate samples.indicate emissions.CO 2 emissions were determined for samples that received urea (i.e., molar urea-to-calcium ratio of 1:0, 1:0.2, 1:1 and 1:2), while control (0:0) and abiotic (NaOH) samples showed CO 2 fluxes near 0. Variation of CO 2 in the vial headspace during incubations was found in F I G U R E 3 Comparison of CO 2 (ppm) evolution in the vial headspace during MICP over five 24 h turnaround incubation for the treatments with solutions with different urea-to-calcium ratios: distilled water (0:0); urea-to-calcium molar ratio of 1:0; 1:0.2; 1:1; 1:2; and distilled water with adjusted pH of 9.5 with NaOH (NaOH).Incubations after solution additions 1, 2, 3, and 4 were run between 0 and 10 h, and from 14 to 24 h after solution addition 5.A total of four gas measurements were performed during every 10 h incubation period at times t = 1, 3, 7 and 10 h since vial closure.Gas concentrations are the difference between measured average concentrations and atmospheric concentration at t = 0. Markers and error bars indicate the average and standard deviation of three replicate samples.each treatment, but not for control samples (Figure 3, '0:0'), indicating soil-atmosphere CO 2 fluxes did not occur.Samples that received NaOH, instead showed negative CO 2 values at t r = 1 h, equivalent to a complete removal of CO 2 from the vial headspace (Figure 3, 'NaOH') and a negative flux of À2.8 μg of CO 2 -C h À1 g À1 soil.Thereafter, CO 2 levels remained stable, hence explaining the CO 2 fluxes close to zero in Figure 2. CH 4 and N 2 O fluxes were detected at very low levels (<0.01 μg h À1 g À1 dry soil) with insignificant variation in concentration in the vial headspace (Figures S3 and S4), indicating fluxes of these gases during MICP did not occur for any case scenario.

| Greenhouse gas fluxes
Samples that received urea and Ca 2+ at various proportions showed CO 2 emissions ranging between 0.52 and 4.08 μg of CO 2 -C h À1 g À1 soil (Figure 2).The CO 2 flux dynamics varied with the amount of Ca 2+ in solution and increasing applications of cementation solution.For samples that did not receive Ca 2+ (1:0), CO 2 emissions increased slowly but progressively from 0.05 ± 0.11 to 0.62 ± 0.26 μg of CO 2 -C h À1 g À1 soil (Figure 2).With the highest urea-to-calcium ratio (1:0.2),CO 2 fluxes increased markedly from 0.53 ± 1.19 to 4.08 ± 0.84 μg of CO 2 -C h À1 g À1 soil between the first and third incubations and plateaued during the third and fourth incubation period.With equimolar urea-to-calcium ratio (1:1), CO 2 fluxes increased between the first and second incubation period but remained stable in subsequent incubation periods at 0.94 ± 0.19 μg of CO 2 -C h À1 g À1 soil.When Ca 2+ was twice as high as urea (1:2), CO 2 fluxes were higher in the first incubation and remained lower and stable in following incubations at 0.68 ± 0.01 μg of CO 2 -C h À1 g À1 soil.
During incubations, CO 2 concentration in the vial headspace increased linearly with reaction time (Figure 3, cf.Table S3 for R 2 values).During the third and fourth incubation periods, however, 1:0.2 samples showed a rapid increase in CO 2 between t r = 1 and 7 h, followed by a plateau between t r = 7 and 10 h (Figure 3, '1:0.2'), a pattern that was not observed for any other treatment.With increasing applications of cementation solutions, higher accumulation of CO 2 in the vial headspace was recorded for 1:0 and 1:0.2 samples, while for 1:1 and 1:2 samples the variation in headspace CO 2 was similar, except for the 1:2 samples that showed significantly higher CO 2 in the vial headspace during the first compared to the subsequent incubation periods (Figure 3, '1:2').
Interestingly, CO 2 emissions were lowest when Ca 2+ was not present (1:0) and highest with the highest ureato-calcium ratio (1:0.2).The highest CO 2 emissions were recorded for 1:0.2 samples during the third and fourth treatment applications (t = 48 to 82 h), with CO 2 emissions two times higher ($4 μg of CO 2 -C g À1 soil h À1 ) than for 1:1 and 1:2 samples (<2 μg of CO 2 -C h À1 g À1 soil), and up to four times higher than 1:0 samples (<1 μg CO 2 -C h À1 g À1 soil).Figure 4 presents a comparison of CO 2 in the vial headspace across treatments.During the first treatment application, evolved CO 2 in the vial headspace was higher with higher Ca 2+ content (1:2 > 1:1 $ 1:0.2) (Figure 4, '1').Instead, from the second to the fourth treatment applications, CO 2 in the headspace was consistently higher with lower Ca 2+ content (1:0.2 > 1:1 $ 1:2) (Figure 4, '2' to '4').Remarkably, without Ca 2+ (1:0), CO 2 increase in the vial headspace remained near zero the first two treatment applications, showing a clear accumulation of CO 2 in the vial headspace only from the third application of cementation solution onwards, then reaching similar values to the 1:2 samples during the fourth treatment application.
Observations from the second half of the 24 h reaction time during the fifth treatment application indicated that CO 2 in the vial headspace continued to increase with reaction time for all treatments containing urea (Figure 3).CO 2 accumulation in the vial headspace was highest for the equimolar urea-to-calcium ratio (1:1), followed by the highest (1:0.2) and lowest (1:2) urea-to-calcium ratios, respectively (Figure 4).Equimolar urea-to-calcium ratio (1:1), double calcium compared to urea (1:2) and samples that did not receive Ca 2+ (1:0) showed higher CO 2 emissions compared to incubations conducted during the first 10 h of reaction time (first to fourth treatment applications) (2.51 ± 0.41, 1.44 ± 0.77 and 0.65 ± 0.51 μg CO 2 -C g À1 soil h À1 , respectively), while the treatment containing the highest urea-to-calcium ratio (1:0.2) showed a marked decrease in CO 2 fluxes (1.48 ± 1.11 μg CO 2 -C g À1 soil h À1 ) compared to the initial 10 h reaction time of the third and fourth treatment applications (Figure 2).Estimated cumulative CO 2 -C emissions during MICP at various urea-to-calcium ratios over 10-h incubation periods based on average fluxes (Figure 2) are presented in Figure 5.Despite the different observed dynamics, the equimolar urea-to-calcium ratio (1:1) and the double Ca 2+ (1:2) treatments showed similar cumulative CO 2 emissions (47 to 58 μg CO 2 -C g À1 soil), two to three times higher than the treatment without Ca 2+ (1:0) ($18 μg CO 2 -C g À1 soil).The treatment with the highest ureato-calcium ratio (1:0.2) emitted roughly two times the amount of CO 2 (112 μg CO 2 -C g À1 soil) compared to the 1:1 and 1:2 samples, and five to six times to samples that did not receive Ca 2+ (1:0).

| Soil total inorganic carbon
An increase in soil TIC occurred only in treatments containing urea and Ca 2+ (Figure 6).The initial soil TIC content was 3.2 μg C g À1 of dry soil and increased drastically during the five cementation periods to values between 1135 and 1470 μg C g À1 of dry soil.Significant differences in precipitated TIC across treatments containing Ca 2+ were not observed.On average, a ureato-calcium ratio of 1:1 was associated with the largest TIC accumulation in soil, followed by the 1:0.2 and 1:2 treatments (1470 ± 531, 1217 ± 198 and 1136 ± 336 μg C g À1 dry soil, respectively).

| Isotopic composition of precipitated and emitted carbon
The isotopic composition of precipitated CaCO 3 in treatments containing urea and Ca 2+ was similar for all treatments, with an average δ 13 C TIC of À40.7 ± 0.8‰ and δ 18 O TIC of À14.7 ± 1.0‰ (Figure 7), indicating the origin of precipitated carbon was urea (Table S2).Soil samples F I G U R E 4 Comparison of CO 2 (ppm) evolution in vial headspace during MICP over 10 h incubation for solution treatments: distilled water (0:0); urea-to-calcium molar ratio of 1:0; 1:0.2; 1:1; 1:2; and distilled water with adjusted pH of 9.5 with NaOH (NaOH), over five incubation periods (1-5).Incubation time (1) to (4) t = 0-10 h and of 5) t = 14 h to 24 h.Four gas measurements were performed during every 10 h incubation period at t = 1, 3, 7 and 10 h since vial closure.Gas concentrations are the difference between measured average concentrations and atmospheric concentration at t = 0. Markers and error bars indicate the average and standard deviation of three replicate samples.
treated with distilled water (0:0), the alkaline pH solution (NaOH) and urea without Ca 2+ (1:0) showed a similar isotopic signature to the untreated soil (Table S1).Further, the isotopic composition of evolved CO 2 in the vials' headspace during MICP across treatments was δ 13 C CO2 = À48.0 ± 3.3‰, indicating that the source of the emitted CO 2 was mainly urea for all treatments with urea (Figure 8).

| Carbon mass balance
The carbon applied to vials in a single incubation estimated from the mass of chemical compounds added and known concentration of carbon in each compound (Table S2) was 8617 μg C g À1 dry soil, and the total carbon applied 43,068 μg C g À1 dry soil (for further detail see Tables S6-S8).Based on the cumulative CO 2 -C emitted (Figure 5) and precipitated CaCO 3 -C (Figure 6), emitted and precipitated carbon were estimated to account for <0.35% and <4.5% of the total carbon applied, while >95% of the total carbon applied remained unaccounted for (Table S9), suggesting it remained in solution and/or taken up by microbial biomass.Urea-C represented 60% of the total carbon applied (Table S8).If both emitted and precipitated carbon were solely of urea origin (based on results of Section 3.5), the data suggest that between 5.9 to 7.6% of the total urea-C introduced was emitted and/or precipitated.Based on the soil TIC data (Figure 6), the equivalent amount of precipitated Ca 2+ was calculated assuming precipitated carbonates were pure CaCO 3 .The ratio of total applied Ca 2+ in the Ca 2+ treatments to the Ca 2+ precipitated as CaCO 3 revealed that 91.5% of introduced Ca 2+ was precipitated in the 1:0.2 treatment, while the ratio decreased to 22 and 8.5% in the 1:1 and 1:2 urea-to-calcium treatments, respectively (Table S9).

| Main results
Our study on GHG fluxes during biostimulation of MICP in a quartz sand revealed CO 2 emissions, but no indication of N 2 O and CH 4 fluxes (Figure 2).The results of this study demonstrate that varying the urea-to-calcium ratio in solution influenced CO 2 flux dynamics, solution pH and the extent of carbonate precipitation.Overall, we observed that for reaction times of 10 h, increasing Ca 2+ concentration in treatment solution from 20 to 200 mM decreased cumulative CO 2 emissions (Figure 5), rendered lower solution pH (Figure 1), and produced a similar amount of precipitated carbonates in soil (Figure 6).

| GHG fluxes
CO 2 emissions in this study are in line with CO 2 emissions derived from MICP on the same soil observed by Casas et al. (2020).In Casas et al. (2020), urea and Ca 2+ concentrations in treatment solution were 100 and 20 mM, respectively, just as the highest urea-to-calcium ratio (1:0.2) treatment used in this study.Average CO 2 emissions of eight cementation treatments of 24 h reaction time were reported to be 1.33 ± 0.44 g CO 2 -C m À2 , equivalent to 0.58 ± 0.19 μg CO 2 -C h À1 g À1 soil, and to increase linearly during reaction time.In this study, F I G U R E 5 Cumulative CO 2 -C emitted during MICP at ureato-calcium molar ratios of 1:0, 1:0.2, 1:1 and 1:2 estimated from average CO 2 -C fluxes over 10 h incubation periods in Figure 2.
F I G U R E 6 Soil total inorganic carbon at the end of MICP treatments of varying urea-to-calcium molar ratios (1:0, 1:0.2, 1:1 and 1:2), distilled water (0:0, control) and solution of adjusted pH of 9.5 with sodium hydroxide (NaOH) determined by gas chromatography-continuous flow-isotope ratio mass spectrometry.Bars and error bars indicate the average and standard deviation of three replicate samples.
average CO 2 emissions of the 1:0.2 treatment increased from 0.53 ± 1.19 to 4.08 ± 0.84 μg of CO 2 -C h À1 g À1 soil, linearly between the first and third incubations, plateauing during the third and fourth incubation, and dropping for reaction time longer than 10 h period.While results of the first incubation period are comparable with the results obtained by Casas et al. (2020), CO 2 emissions of subsequent incubations determined in this study are significantly higher.CO 2 concentration in Casas et al.
(2020) were recorded with a soil chamber system (LI-8100, LI-COR Biosciences, LICOR Inc., United States) every hour for 10 min at a frequency of 1 s À1 by closing a 3 kg soil column which otherwise was open to the atmosphere, while in this study vials remained closed throughout the entire incubation, the mass of soil was 1.5 g and the measuring frequency was four times every 10 h.While the cause of discrepancy could be partially due to the different experiment conditions and measuring devices, the main cause remains unclear.CO 2 emissions of MICP in aqueous samples obtained from travertine environments have also been reported by Okyay et al. (2016) where both CO 2 emissions and sequestration were observed in incubation experiments and attributed to MICP either acting as a source or a sink of CO 2 depending on bacterial community composition.
N 2 O fluxes are likely to occur in MICP due to the abundance of NH 4 þ .N 2 O is a minor product of both nitrification and denitrification (Stein & Nicol, 2018).Ammonia oxidation in quartz sand was observed after 27 days of being in contact with treatment solution (Gat et al., 2016), and a later study indicated an increase in activity of nitrifying bacteria and/or archaea during MICP (Tsesarsky et al., 2016).However, nitrite (NO 2 À ) and nitrate (NO 3 À ) were found at very low concentrations to below limit of detection in a previous MICP experiment on the same soil as used in this study during and 29 days following end of MICP treatment, indicating nitrification did not occur (Casas et al., 2020).Environmental factors influencing the nitrification process are F I G U R E 8 Carbon isotopic signature of CO 2 evolved in the vial headspace at the end of 10 h incubation periods of quartz sand treated with distilled water (0:0, control) and MICP solution of varying urea-to-calcium molar ratios (1:0, 1:0.2, 1:1, and 1:2) determined by gas chromatography-continuous flow-isotope ratio mass spectrometry.Markers and error bars indicate the average and standard deviation of three replicate samples.
substrate availability, soil matrix (e.g., clay content), water status, oxygen availability, temperature, salinity and pH (Norton, 2011;Sahrawat, 2008).In particular, nitrifying bacteria require molecular oxygen as electron acceptors.Oxygen availability is hindered in water saturated conditions thus reducing nitrification rates.Nitrification is observed at pH between 6 and 10 however, nitrifiers are sensitive to free NH 3 , which is exacerbated by alkaline pH and high concentration of NH 4 þ (Breuillin-Sessoms et al., 2017;Kim et al., 2006;Venterea et al., 2015).Nitrifying bacteria are chemolithotrophs, thus grow in media containing mineral salts, where magnesium and phosphate, despite at very low concentrations, were determined essential (Meiklejohn, 1952).Key enzymes involved in nitrification include ammonia monooxygenase, hydroxylamine dehydrogenase and nitrite oxidoreductase, the former containing iron/copper and the latter requiring molybdenum as cofactor (Stein & Nicol, 2018).While the reason for the observed results with respect to N 2 O emissions remains unclear, it is plausible that a lack of oxygen combined with free NH 3 could have inhibited nitrification during MICP.Additionally, it could be that elements required for enzyme synthesis were a limiting factor.In any case, despite the results of our study indicating that nitrification may not occur during MICP, additional studies are necessary to test whether these observations hold over time when excess ammonia degasses from the system, whether inhibition is related to nutritional requirements, lack of oxygen, presence of essential elements, or other factors.

| CO 2 flux dynamics
The dynamics of CO 2 fluxes during MICP at varying Ca 2+ concentrations were complex.CO 2 emissions were lowest without Ca 2+ (1:0) and highest with the lowest Ca 2+ concentration (1:0.2), while increasing Ca 2+ resulted in lower CO 2 emissions (Figure 2).Urease enzyme reaction kinetics are affected by pH, such that its affinity for urea decreases with decreasing pH, but reaction rates are higher at circumneutral pH and decrease with increasing pH (Cabrera et al., 1991;Stocks-Fischer et al., 1999).With higher urea hydrolysis reaction rates, more CO 2 is produced in the same period of time and thus can potentially be emitted.Treatments that contained Ca 2+ (1:0.2, 1:1 and 1:2) resulted in lower solution pH compared to the treatment that excluded Ca 2+ (1:0) (Figure 1), which may have resulted in increased enzyme activity (i.e., production of CO 2 ) and therefore result in higher CO 2 emissions, explaining the observed higher CO 2 emissions obtained when Ca 2+ was present.On the other hand, the high pH induced by the treatment that excluded Ca 2+ (1:0, pH = 9) may have acted as a sink of atmospheric CO 2 and partially balanced CO 2 emissions produced by urea hydrolysis, particularly during the initial treatment applications (Figure 3).In abiotic experiments with various MICP treatment media, Okyay and Rodrigues (2015) observed CO 2 sequestration by the alkaline treatment solutions where pH had been adjusted to the alkaline pH achieved by urea hydrolysis using that same media.In this study, complete CO 2 removal from the vial headspace was evident in the NaOH treatment (Figure 3), which was accompanied by a reduction in pH of 0.5 units (Figure 1).Urea hydrolysis by the treatment that contained solely urea (1:0) induced a similar pH compared to the NaOH treatment (Figure 1).This demonstrated that abiotic CO 2 sequestration was taking place due to solution alkalinity, contributing to the lower CO 2 emissions observed with the treatment without Ca 2+ .
Based on the effect of pH on both urease activity and abiotic CO 2 sequestration, the lowest CO 2 emissions in treatments containing Ca 2+ should have been observed for the treatment that contained the lowest Ca 2+ content (i.e., 1:0.2).Instead, cumulative CO 2 emissions were higher (Figure 5) for the treatment that induced the highest pH (Figure 1) (i.e., 1:0.2), indicating that pH could not solely explain the observed results.Urease activity can be negatively affected by Ca 2+ content, implying that the higher the Ca 2+ concentration, the stronger is the inhibitory effect.A recent study on urease activity under varying Ca 2+ concentrations reported no inhibition to complete inhibition of urease activity at Ca 2+ concentrations <10 mM and >200 mM, respectively (Cui et al., 2022).The carbon and Ca 2+ mass balances of our study indirectly indicated a lower amount of urea hydrolysed with increasing Ca 2+ , supporting the results of Cui et al. (2022).Thus, presumably, increasing Ca 2+ concentrations from 20 to 200 mM may have increasingly inhibited urease activity, resulting in less CO 2 produced by urea hydrolysis and lower CO 2 emissions.

| Sources of CO 2 emissions
In MICP, potential sources of CO 2 emissions originate from the breakdown of urea into CO 2 and NH 3 , and additional microbial respiration through the breakdown of other organic carbon sources present both in treatment solution (e.g., molasses, sodium acetate) and soil (e.g., humus).The soil used in this study had very low carbon content (TOC = 0.02%, TIC = 0.0003%, Table S1), hence sources of CO 2 emissions could be assumed to originate predominantly from the treatment solution.The similarity in isotopic signature of evolved CO 2 (À48.0 ± 3.3‰, Figure 8) with urea-C (À41.03 ± 0.12%, Table S2) proved that the evolved CO 2 was of urea origin and the contribution of other carbon sources was small.In line with results of this study, experiments conducted on agricultural soils amended with synthetic urine containing δ 13 C labelled urea indicated urea-C was the main contributor to CO 2 emissions the first 2 days after application but declined rapidly from day three onwards (Ambus et al., 2007).

| Sources of CO 2 precipitation
CO 2 sequestration in MICP can occur due to CO 2 precipitation as Ca 2+ carbonate minerals (mineral trapping) and CO 2 storage in solution due to an increase in solution pH resulting from urea hydrolysis, which consumes protons from the soil solution due to conversion of NH 3 to NH 4 þ (solubility trapping).With an increase in solution pH, atmospheric CO 2 is drawn into the solution to neutralise excess OH À .In the absence of soil organic matter, sources of CO 2 for MICP may thus originate from atmospheric CO 2 , dissolved inorganic carbon (DIC) from minerals (e.g., carbonates) or CO 2 produced by microbial respiration (Okyay et al., 2016).In this study, MICP was indicated by the increase in soil TIC in samples that received urea and Ca 2+ , as opposed to samples that did not receive Ca 2+ , where increase in soil TIC was not observed (Figure 6).The similarity in isotopic C composition of precipitated carbonates (À40.7 ± 0.8‰, Figure 7) and urea-C (À41.03 ± 0.12%, Table S2) indicated that precipitated carbon was of urea origin, in line with previous observations by Millo et al. (2012) and Casas et al. (2020).

| Changes in pH during MICP
During urea hydrolysis, there are two main processes which alter pH, namely the protonation of NH 3 molecules released by urea hydrolysis, and the change in speciation of dissolved CO 2 .NH 3 molecules undergo protonation with hydrogen (H + ) from water to produce ammonium (NH 4 þ ) and hydroxide ions (OH À ), thus increasing pH.CO 2 molecules dissolved in water produce carbonic acid (H 2 CO 3 ), which deprotonates to bicarbonate (HCO 3 À ) and carbonates (CO 3 2À ), releasing one and two H + , respectively, decreasing solution pH.For every mole of urea hydrolysed, 2 moles of NH 3 and 1 mole of CO 2 are produced.Below a pH of 8.5, HCO 3 À is the dominant speciation of CO 2 in solution, so that only 1 mole of H + will be produced.Therefore, two OH À moles are produced for every mole of H + , resulting in a net solution pH increase of up to 9, when NH 4 þ begins to deprotonate to NH 3 and degas from the soil-solution system.The results of the treatment solution without Ca 2+ (1:0) are in agreement with the latter, where a net increase in pH to 9 occurred followed by stabilisation (Figure 1).With the addition of Ca 2+ , CaCO 3 minerals can form, with favourable conditions for precipitation above pH of 8.5.As HCO 3 À deprotonates to CO 3 2À , an additional mole of H + is released, such that the 2 moles of OH À initially produced by NH 3 protonation are balanced, resulting in a decrease in pH and stabilisation of pH around 8.25, which is the pH of pure CaCO 3 solution (Bache, 1984).In this study, when Ca 2+ was introduced to the system, the initially higher pH and decreasing trend with increasing reaction time (Figure 1) were consistent with initial rapid increase in solution pH due to urea hydrolysis and subsequent decrease due to CaCO 3 precipitation (Dupraz et al., 2009).
4.7 | Effect of increasing the Ca 2+ concentration on pH pH stabilised at lower values with increasing Ca 2+ concentration (i.e., approx.8, 7.5 and 7.3 for urea-to-calcium molar ratios 1:0.2, 1:1 and 1:2, respectively) (Figure 1) than the earlier reported pH values of MICP in solution (pH = 8-8.7 for a treatment solution containing 20 mM Ca 2+ and 333 mM urea) (Dupraz et al., 2009).CaCl 2 produces slightly acidic solutions (salt of a weak base and a strong acid), which may have partly contributed to the observed lower pH with increasing CaCl 2 .Additionally, release of protons from the adsorber matrix of the soil being exchanged with Ca 2+ likely contributed to further lowering of the pH (Bache, 1974;Houba et al., 2000).For urea-to-calcium molar ratios of 1:1 and 1:2 (containing 100 and 200 mM Ca 2+ , respectively), values were similar to pH values of CaCO 3 dominated soils (pH = 7.5) (Bache et al. 1984).

| Effect of Ca 2+ concentration on the TIC
Contrary to what was hypothesised, increasing Ca 2+ in solution from 20 to 200 mM did not result in increased CaCO 3 precipitation and resulted in similar soil TIC values (Figure 6).The fact that the total amount of TIC was not significantly different between the 1:0.2, the 1:1 and the 1:2 treatments (Figure 6) indicated that Ca 2+ was not a limiting factor for carbonate formation but instead it was the concentration of CO 3 2À in the soil solution.This was likely related to the inhibitory effect of Ca 2+ on urease activity, which induced lower CO 2 availability for carbonation, and pH ≤8, which limited the extent to which carbonation could occur.

| Mechanisms for observed CO 2 emissions/sequestration
During MICP, the CO 2 (g) $ CO 2 (aq) $ H 2 CO 3 (aq) $ HCO 3 À (aq) $ CO 3 2À (aq) $ CO 3 2À (s) equilibrium in a system is disequilibrated by production of additional CO 2 from urea hydrolysis.The precipitation of calcium carbonate in MICP occurs extracellularly.That and the increase in NH 4 , DIC and pH in soil-solution necessarily imply that NH 3 and CO 2 from urea are to some extent excreted outside the cell.Once inorganic carbon is in the outer cell environment, its speciation is dependent on the total inorganic carbon in the system, solution pH and CO 2 partial pressure in the air-filled pore space of the soil and ultimately in the headspace of the GC vials.With increasing pH and CO 2 partial pressure, the liquid phase has more capacity to store inorganic carbon and vice versa, while the total carbon in the system determines whether there is sufficient inorganic carbon in the system for it to be quantifiable in all its possible speciation forms and phases.CO 2 gas measurements with the NaOH treatment (Figure 4), where no carbon was applied to the system, evidenced the effect of pH on CO 2 solubility trapping, where atmospheric CO 2 entered the solution due to alkaline pH inducing a decrease with respect to initial solution pH (Figure 1) and quantifiable CO 2 in the headspace was negligible due to limited total carbon in the system (Figure 4).When urea but no Ca 2+ were present (1:0 treatment), CO 2 gas measurements near zero indicated that the CO 2 produced via urea hydrolysis mostly accumulated in solution (Figure 4) due to solubility trapping, as indicated by the increase in solution pH to 9 (Figure 1).However, as microbial activity increased with repeated treatment applications (i.e., more CO 2 was produced), solubility trapping could not balance the amount of CO 2 produced from urea hydrolysis, resulting in an accumulation of CO 2 in the vial headspace/CO 2 emissions.Several interconnected factors might explain the varying CO 2 emissions observed when Ca 2+ was introduced at varying urea-to-calcium ratios.On the one hand, increasing Ca 2+ in solution and precipitation of calcium carbonate decreased solution pH (Figure 1), resulting in less capacity of the solution to store carbon and higher accumulation of CO 2 in the vial headspace.
On the other hand, slower urea hydrolysis rates with increasing Ca 2+ directly resulted in a lower amount of CO 2 in the system, which would explain the lower accumulation of CO 2 in the vial headspace observed with increasing Ca 2+ .Finally, limited urea hydrolysis and insufficient increase in solution pH to reach favourable conditions for calcium carbonate precipitation (pH >8.5) resulted in similar calcium carbonate precipitation across treatments including urea and Ca 2+ .Accumulation of CO 2 in the vial headspace observed in this study thus reflected an increase in dissolved CO 2 produced from urea hydrolysis that could not be balanced by increased capacity to store CO 2 in solution provided by increase in pH (solubility trapping) and storage of CO 2 as solid in calcium carbonate (mineral trapping).

| Chemical efficiency
Finally, it is worth highlighting that CO 2 -C emissions and precipitated CaCO 3 -C accounted for less than 0.35% and <4.5%, respectively, of the total applied carbon (Table S9).This indicated that the amount of precipitated carbon was about tenfold higher than emitted carbon.Furthermore, the Ca 2+ precipitated represented 91, 22 and 8.5% of applied Ca 2+ for treatment solutions containing 20, 100 and 200 mM, respectively.These results further indicate a small amount of urea was hydrolysed and highlight a very low efficiency of usage of applied chemicals, and that longer reaction times were necessary for full degradation of urea.Reaction times of 8 to 10 h have been reported optimum for complete urea hydrolysis in bioaugmentation experiments on quartz sand with S. pasteurii (Al Qabany et al., 2012).While different bacteria express different urea hydrolysis rates, experiments on biostimulation vs bioaugmentation showed similar overall urea hydrolysis rates (Gomez et al., 2019).This highlights that reaction time of MICP in soil via biostimulation should be adjusted to specific soil bulk urea hydrolysis rates and Ca 2+ concentration.

| CONCLUSIONS
Our study on GHG fluxes during biostimulation of MICP on a quartz sand indicated MICP was a source of CO 2 emissions, while no N 2 O nor CH 4 was produced during MICP treatment.Varying the relative proportion of Ca 2+ with respect to urea-C had an effect on CO 2 emissions, soil-solution pH, and the extent of carbonate precipitation in soil.Increasing the Ca 2+ concentration from 20 to 200 mM had an inhibitory effect on urease activity, such that less CO 2 and NH 4 þ were produced.On the one hand, the lower production of NH 4 þ and increasing exchange of protons from soil with increasing Ca 2+ concentration resulted in soil solutions of lower pH.On the other hand, the lower production of CO 2 limited availability of carbonates for precipitation and resulted in lower CO 2 emissions.When urea exceeded Ca 2+ , CaCO 3 precipitation was maximised, but CO 2 emissions were highest.When Ca 2+ exceeded urea, CO 2 emissions decreased but the extent of CaCO 3 precipitation was limited by the availability of CO 3 2À , resulting in a low chemical usage efficiency of applied urea and Ca + with respect to precipitated CaCO 3 .Based on the balance of precipitated carbon and CO 2 emissions, we recommend using treatment solutions of equimolar urea-to-calcium concentrations to reduce overall CO 2 emissions while maintaining the same degree of soil carbonation.However, results suggest that longer reaction times were necessary with increasing Ca 2+ with respect to urea, and additional studies are required to compare emitted and precipitated CO 2 at similar amounts of urea hydrolysed.During reaction time, urea was the main source of precipitated and emitted CO 2 .Results further indicate that the abiotic CO 2 sequestration mechanism by the highly alkaline solution pH induced by urea hydrolysis was relevant in balancing CO 2 emissions at pH of 9.The results of this study are expected to serve as a benchmark for future studies on CO 2 fluxes of MICP in soils (e.g., varying reaction time, urea concentration, soil type, soil organic matter content) and inform life cycle assessment of MICP to quantify the environmental impact of the technique.

Highlights•
The study determines MICP related soilatmosphere greenhouse gas fluxes.•During MICP net CO 2 emissions occur, but no N 2 O nor CH 4 fluxes were observed.• Increasing Ca 2+ -to-urea ratio resulted in lower CO 2 emissions and similar carbonate precipitation.•Urea-to-Ca 2+ ratio of 1:1 lowered CO 2 emissions and maintained soil carbonation.

Figure 2
Figure 2 presents calculated GHG fluxes of CO 2 -C, CH 4 -C, and N 2 O-N during MICP at varying urea-to-calcium molar ratios, where positive values

F
I G U R E 7 Carbon and oxygen isotopic signatures of precipitated carbonates in soil prior to treatment (Soil, untreated), following end of treatment with distilled water (0:0, control), MICP solution of varying urea-to-calcium molar ratio (1:0, 1:0.2, 1:1 and 1:2), and solution of adjusted pH of 9.5 with sodium hydroxide (NaOH) determined by gas chromatography-continuous flow-isotope ratio mass spectrometry.Additionally, the graph includes results of the isotopic composition of MICP at equimolar urea-to-calcium ratio ofCasas et al. (2020) on the same soil for comparison.Markers and error bars indicate the average and standard deviation of three replicate samples.