Quantification-of-chemical-and-biological-calcium-carbonate-precipitation-Performance-of-self

Quanti?cation of chemical and biological calcium carbonate precipitation:Performance of self-healing in reinforced mortar containing chemical

admixtures

Claudia Stuckrath a ,Ricardo Serpell b ,Loreto M.Valenzuela c ,d ,Mauricio Lopez b ,d ,?

Department of Structural and Geotechnical Engineering,School of Engineering,Ponti?cia Universidad Catolica de Chile,Santiago,Chile b

Department of Construction Engineering and Management,School of Engineering,Ponti?cia Universidad Catolica de Chile,Santiago,Chile c

Department of Chemical and Bioprocess Engineering,School of Engineering,Ponti?cia Universidad Catolica de Chile,Santiago,Chile d

Research Center for Nanotecnology and Advanced Materials ‘‘CIEN-UC’’,Ponti?cia Universidad Catolica de Chile,Santiago,Chile

a r t i c l e i n f o Article history:

Received 7May 2013

Received in revised form 31January 2014Accepted 23February 2014

Available online 12March 2014Keywords:Bacteria

Calcium lactate Image analysis Crack sealing

Calcium carbonate precipitation Self-healing

Chemical admixture Lightweight aggregate Carbonation

a b s t r a c t

Cracks increase permeability affecting the durability of concrete.As they develop gradually,it is dif?cult to determine when to repair them.Self-healing materials can repair themselves gradually as cracks form.In this study,the isolated and combined effect of two self-healing agents for concrete,both based on calcium carbonate precipitation,was studied.Lightweight aggregates were impregnated with chemical and biological solution to be added as healing agents in concrete mixtures.The in?uence of two common chemical admixtures on the performance of the self-healing agents was also studied.All self-healing agents were able to seal cracks between 0.08and 0.22mm in width.The estimated effect of chemical agents on the mean healing was higher than that of biological agents.In addition,thermogravimetric analysis suggests the precipitates are different.Admixtures had no signi?cant in?uence on the performance of self-healing agents.

1.Introduction

Concrete is a ceramic material susceptible to developing cracks when it is under tensile stress.Ceramic materials are generally char-acterized by ionic-covalent like bonds.The limited atomic mobility of this kind of bond imparts intrinsic brittleness to ceramics [1].Moreover,these bonds make the material very strong to compres-sive stresses,but weak to tensile stresses.To compensate for this weakness,steel reinforcement is used in concrete due to its tensile strength and ductility.It should be noted that the role of steel in reinforced concrete is not to prevent concrete to deform or crack,but to take tensile stresses and control crack width.

Cracks below 0.05mm in width are not usually a problem since concrete has the ability to seal them in a process called autogenous healing,and fully recovering it’s mechanical and transport proper-ties [2].Autogenous healing is believed to be produced by swelling of the cement paste,hydration of the remaining unhydrated cement,precipitation of calcium carbonate (CaCO 3)crystals,and crack ?lling by impurities in water or by debris from the crack sur-face [3].Of all these mechanisms,calcium carbonate precipitation (CCP)is the only one that can be intentioned and engineered to improve the self-healing capacity of concrete and is the primary fo-cus of this research.In this research the effects of two different self-healing agents (chemical and biological)and two polar organic molecules (plasticizer and air-entrainer)were studied.

CCP has 4key factors that could be managed to increase or de-crease its effectiveness:(1)calcium ion (Ca 2+)concentration,(2)dissolved inorganic carbon (DIC)concentration,(3)pH (pK2(HCO 3

-à/CO 3

2à)=10.3at 25°C),and (4)the availability of nucleation sites [4].In addition,some organic molecules and bacteria can modify these factors increasing or decreasing CCP and can create different CaCO 3polymorphs,such as calcite,vaterite,and aragonite [5].Organic molecules can affect CCP by inhibition through Ca 2+sequestration,by acting as nucleation sites,or by modifying CaCO 3precipitate to form other crystalline phases or amorphous phases.Negatively charged acidic functional groups that can sequester Ca 2+are carboxylic acids (R–COOH),hydroxyl groups (R–OH),amino groups (R–NH 2),sulfate (R–O–SO 3H),sulfonate (R–SO 3H),

?Corresponding author.Address:Vicu?a Mackenna 4860,7820436Macul,Chile.Tel.:+56223544244.

E-mail address:mlopez@ing.puc.cl (M.Lopez).

and sulfhydryl groups(–SH)[6].These functional groups can also act as nucleation sites if the Ca2+concentration is high[7].Surfac-tants can also control the formation of the crystal phase by in?u-encing the nucleation,crystal growth,and aggregation[8]. Likewise,amorphous phases are formed when the concentration of organic molecules is high and it can be stabilized against compaction and recrystallization[7].

The fact that some chemical admixtures for concrete have func-tional groups similar to those mentioned above might indicate that they can affect CCP and the healing process itself.For this reason, the effect of two chemical admixtures on CCP was studied:a common air-entrainer(surfactant)based on sodium naphthalene sulfonate formaldehyde condensate,and a common plasticizer based on calcium lignosulfonate.

Dupraz et al.[6]found that one of the fundamental controls of CCP is the‘‘Alkalinity Engine’’that can rely on an intrinsic(bacte-rial metabolism)and/or extrinsic(the environment)component. Since concrete has a pH above9,due to the presence of calcium hydroxide,an intrinsic approach to increase pH with bacteria metabolism is not necessary.However,bacteria were also included in this study because they have been proven to increase the amount of precipitate[9].

CCP requires free calcium(Ca2+),which can be externally pro-vided in the form of calcium salts like calcium chloride or calcium lactate.Nevertheless,an internal source of calcium might be avail-able in the hydrated cement paste in the form of calcium hydrox-ide,calcium silicate hydrates,or calcium aluminate hydrates, among others.In fact,calcium hydroxide naturally becomes cal-cium carbonate in a process known as carbonation.Carbonation is a form of concrete deterioration since it can reduce pH below the threshold needed to maintain the passivation of steel thus potentially speeding up the onset of corrosion.Even,bacteria’s spores have calcium in its protective membrane[10]and could be a source for CCP.

Jonkers and Schlangen[9]proposed a two-component self-healing system based on the addition of bacteria and calcium lactate.The mechanism is based on the metabolic conversion of suitable organic compounds to calcite, e.g.represented by the bio-conversion of calcium-formate with portlandite present in the paste matrix.

The relative size of bacteria with respect to cement paste pores suggests that it cannot be directly embedded in the concrete ma-trix because there would not be enough space for bacteria to live. An attempt to overcome this problem was proposed through the immobilization of bacteria in porous glass beads(Siran TM)[11]. Another attempt to protect bacteria was proposed by Wang et al.

[12]encapsulating silica gel or polyurethane with bacteria in glass tubes.Yet another approach used diatomaceous earth to host bac-teria[12]and observed that cracks in the range from0.15to 0.17mm in width were completely healed.Recently,porous light-weight aggregates were used to immobilize bacteria and calcium lactate[13].After100days of immersion in water,it was obtained full crack healing for cracks up to0.46mm in width,which com-pared favorably to the0.18mm in width sealed in control specimens.

This research focuses on the performance of alternative self-healing agents that improve CCP and how such performance is in?uenced by common components present in concrete.The self-healing agents were based on calcium lactate,representing chemical CCP,and bacteria,representing biological CCP.The con-crete’s admixtures were calcium lignosulfonate(plasticizer)and naphthalene sulfonate formaldehyde condensate(air-entrainer).

2.Materials and methods

Reinforced mortar?exural specimens,each with different self-healing agents and chemical admixtures,were prepared,cured, cracked,and left to self-heal immersed in water at room tempera-ture for100days.The mortar specimens were prepared using Ordinary Portland Cement,water,and siliceous sand.Three types of self-healing agents were prepared(Table1):chemical agent(C) calcium lactate solution,biological agent(B)bacteria and yeast-extract solution,and both combined(CB).The agents were embed-ded in lightweight aggregate(LWA)according to Wiktor and Jonkers [13].

2.1.Bacteria selection and grow

Bacillus pseudo?rmus LMG17944(Belgian coordinated collections of microorganisms,Ghent),a spore-forming facultative alkaliphilic bacteria,was used for this study.These bacteria grow in a pH range from7.5to11.4and can withstand large sudden increases in external pH[14].

Bacterial stocks were stored atà80°C in glycerol.Bacteria were cultured for4h in liquid nutrient media according to the supplier’s recommendations.The medium was comprised of5g peptone,3g meat extract,10mg MnSO4?H2O,0.5%NaCl in1L MilliQ ultra-pure water.pH was adjusted to9.7with a solution of0.42g NaHCO3and0.53g Na2CO3in100mL MilliQ ultra-pure water.This culture(batch1)was used to inoculate a second culture(batch2) to enhance spore formation according to[15].Cultures were aero-bically incubated in Erlenmeyer?asks on a shaker table at250rpm and37°C.The growth of bacteria was monitored using Optical Density set at600nm.Bacterial concentration was estimated with the most probable number method cultivation–dilution technique.

Bacteria were centrifuged for20min at6000rpm and re-suspended twice in sterile MilliQ ultra-pure water.The?nal concentration of bacteria was1.5?109cells/mL and it was stored at4°C(batch3)until the preparation of the self-healing agent. 2.2.Preparation of self-healing agent

Self-healing agent consist in impregnated LWA with the chem-ical and/or biological solution.LWA used was expanded clay sieved between ASTM standards n°4and16(4.75mm and1.18mm).The Table2shows the main properties of the LWA used[16].

For the CB agent,the LWA particles were vacuum-impregnated with a solution of50g/L calcium lactate pentahydrate that corre-sponds to maximum solubility and1g/L yeast extract(according to Wiktor and Jonkers[13]),followed by a?nal impregnation step with bacteria at4°C from batch3(Section2.1).After each impreg-nation treatment,LWA was oven-dried for5days at37°C.Impreg-nated LWA containing2%by weight calcium lactate and 1.3?108cells/g particles was obtained.The preparation of agents

Table1

Nomenclature of mortar specimens used in this study.

Chemical admixture No agent Calcium lactate Bacteria Calcium lactate+bacteria

Self-healing agent

No admixture Control C B CB

Plasticizer P CP BP CBP

Air-entrainer AE CAE BAE CBAE

C.Stuckrath et al./Cement&Concrete Composites50(2014)10–1511

B and

C followed the same procedures as above,but only bacteria

and yeast-extract were impregnated in agent B,and only calcium lactate was impregnated in agent C.The self-healing agent must be stored dry,otherwise calcium lactate reacts with humidity pro-ducing calcium carbonate which renders the healing agent inactive.

2.3.Preparation of reinforced mortar specimens

Mortar specimens were prepared according to the factorial experimental design using chemical admixtures and self-healing agents as experimental factors.Two chemical admixtures were used:calcium lignosulfonate23.5%,a common plasticizer,and so-dium lauryl ether sulfate4%,a common air-entraining agent.Self-healing agents CB,B,and C were used and LWA without any impregnation as control.Twelve mixtures were prepared in tripli-cate.Table1shows the combination and nomenclature used to prepare these mortar specimens.Each specimen consisted of 4?4?16cm prisms with shear and?exural reinforcement to al-low controlled crack formation upon loading.Two types of rein-forcements were considered:shear and?exural reinforcement. The shear reinforcement is comprised of two bent wires of 1.24mm in diameter.The?exural reinforcement is compressed of0.5-in diameter thread bar with a nut at the ends(Fig.1).They were prepared with Ordinary Portland Cement,water,siliceous sand,LWA and chemical admixtures(Table3).Chemical admix-tures were dosed according to supplier’s recommendations:0.5% of cement weight for the plasticizer and120mL per100kg cement testing machine setup for3-point bending.The maximum load was such that the yield stress of the steel was exceeded,so cracks remained open after unloading.Each cracked-specimen was im-mersed separately in tap water(Total alkalinity190.5mg CaCO3/ L and pH7.6)to begin the self-healing process.Containers with the cracked-specimen were maintained open to allow free diffu-sion of gases,speci?cally CO2,with the environment.Each speci-men was removed periodically from water for crack-healing quanti?cation by image analysis as explained in Section2.4.

After100days of immersion,each cracked-specimen was longi-tudinally split,and a sample from inside the cracks(‘‘healing mate-rial’’)was removed with a scalpel from inside the cracks.The longitudinal split was made to measure the calcium hydroxide con-sumption through change in pH.The pH was measured spraying the surfaces with a pH solution indicator within a range of9–13(Fluka). Images of the colored surface were obtained for image analysis.The healing material was analyzed using thermal analysis(thermo-gravimetric analysis(TGA)and differential scanning calorimetry (DSC))and scanning electron microscopy(SEM)(Section2.5).

2.4.Image analysis

Crack-healing quanti?cation was performed by image analysis. The specimens were photographed at20,40,50,70,and100days. 16-megapixel images were obtained from a?xed distance using an SLR camera and lens combination giving a resolution of55pixels per mm(18.2l m/pixel).Two photographs of the specimen face were combined to cover the complete region of interest.After trial testing,the SLR camera and lens combination was preferred over an optical microscope,because it provided a resolution high en-ough to accurately detect and measure the cracks and,at the same time,an image size large enough to cover the entire crack.

Image analysis was performed using a Matlab script specially developed for this study.All images from the same specimen face were automatically aligned to the?rst image of the time series(po-sition,rotation and scale wise).The graphical user interface of the script allowed the user to manually select multiple?xed length segments along visible cracks in the?rst image,de?ning sub-regions of interest to be analyzed.For this study,the length of the segments was held constant at1.6mm.The centerline of the crack in the?rst image was detected using a combination of pixel inten-sity classi?ers and morphological?lters.Starting from the crack centerline,crack boundaries were detected comparing the slope of the pixel intensity pro?le across the crack to the variance of pixel intensities in the perimeter of the sub-region of interest(i.e.,out-side the crack).In the following images crack boundaries were de-tected using the same classi?er,but restricting the search to the general area of the sub-region where a crack was detected in the ?rst image.For any given crack,the number of crack segments was chosen as required.The healing at time t can be calculated as follows:

HealingetT%?ew iàw tT=w i?100e1Twhere w i is the initial crack width,and w t the width at time t.

Table2

Expanded clay properties.

Properties

3-Day absorption32.5%

Dry speci?c gravity1653kg/m3 24-h SSD speci?c gravity1920kg/m3 Fineness modulus 4.16mm/mm Open porosity41.0%

1.Reinforcement used in the mortar specimens.(a)Flexural reinforcement; shear reinforcement;and(c)surface?nish.Table3

Mixing proportions of mortar specimens(in mass and volume basis).

kg/m3L/m3

Ordinary Portland Cement400129 Tap water261261 Siliceous sand968373 Impregnated LWA304217 Air entrapped20 w/c0.65

12 C.Stuckrath et al./Cement&Concrete Composites50(2014)10–15

2.5.Thermal analysis and scanning electron microscopy

TGA and DSC were used to detect the presence of CaCO 3in the healing material [18].Samples of crack healing material ranging from 2.5to 42mg were heated up to 1000°C at a rate of 10°C/min in a nitrogen atmosphere at a rate of 50mL/min.The weight loss and heat ?ow of the samples during the process of heating were recorded and shown in a weight-heat ?ow-temperature graph.SEM was used to study the morphology of the healing mate-rial.Samples were supercritically dried and gold coated.Magni?ca-tions between 300?and 5000?with an accelerating voltage of 15kV were used.

3.Results and discussion

All specimens,regardless of the self-healing agent used,showed a pH above 12on the entire surface exposed by the splitting;how-ever,the control case,without any self-healing agent,showed areas near the borders with pH of 9or less.This suggests that cal-cium hydroxide is not being depleted by CCP.Moreover,the self-healing agents could decrease gas permeability thus decreasing the natural process of carbonation.

Since the initial crack width cannot be controlled during the loading process,the distribution of crack widths varied widely among samples (Fig.2).Consequently,only cracks between 0.08and 0.22mm in width were considered in the analysis.This range of width was present in all samples.

The median of crack healing at 100days is over 58%for all self-healing agents,whereas for control samples without self-healing agents such median is under 38%(Fig.3).C,CAE,and CBP show the best healing performance with a median of over 90%healing.According to this,the best self-healing treatment for a concrete with or without air-entrainer is chemical agent (C),and for concrete with plasticizer it is chemical and biological agent com-bined (CB).

The effect of each factor on the mean and variance of healing was analyzed using regression analysis.The regression model con-sidered 4factors (C,B,P and AE).Fig.4shows a standardized Par-eto chart in which the length of each bar is proportional to the value of the t -statistic calculated for the corresponding effect.C was the only factor that has a signi?cant effect,but B also shows

that it has an important in?uence on the mean of the healing.Con-versely,none of the chemical admixtures factors has a signi?cant effect on self-healing.A negative interaction is estimated for C and B,however,the in?uences of C and B are signi?cant by them-selves.This happens because the expected combined effect of C and B would be the sum of both factors,which is not possible

C.Stuckrath et al./Cement &Concrete Composites 50(2014)10–1513

due to saturation in the response(maximum healing of100%).On the other hand,when analyzing the variance in the healing re-sponse,no factor has a signi?cant importance in explaining it.Nev-ertheless,the air-entraining seems to lower the variance in healing and the B agent seems to increase it.

Results from the TGA show an important decrease in weight be-tween500and800°C,which correspond to the decomposition of analysis shows an important endothermic peak between700and 800°C,which also corresponds to CaCO3decomposition.For sam-ples with agent C,this reaction occurs at lower temperatures con-?rming the results obtained by TGA.

The morphology of the healing material was assessed by SEM (Fig.6).In samples without chemical admixtures(Fig.6a and b), rhombohedra precipitates are identi?ed,which can be calcite,the

different healing materials removed from inside the cracks:(a),(c),and(e)with only bacteria as a self-healing

self-healing agents.(c)and(d)has plasticizer as chemical admixture and(e)and(f)has air-entrainer as chemical 14 C.Stuckrath et al./Cement&Concrete Composites50(2014)10–15

tion of rhombohedra precipitates that were bound together.In samples with bacteria and air entraining agent(Fig.6e),a needle precipitation can be observed,which might be aragonite,a poly-morphism of CaCO3,as identi?ed by Jiang et al.and Kitamura et al.[5,24].

4.Conclusions and future perspectives

All of the self-healing agents studied herein increased the autogenous healing capacity of concrete.Among them,the com-bined self-healing agents chemical and biological were the one showing the highest healing.It is noteworthy from the regression analysis that chemical agent proved to be more effective in explaining the healing of cracks than the biological agent.Since both self-healing agents promoted nearly a100%of healing,it was not possible to prove if the combination of chemical and bio-logical self-healing had the added effect of the separate agents.The ?ndings of this study should be corroborated in the future,includ-ing wider cracks to avoid the saturation of the response and to bet-ter assess the effect of the combined self-healing agent.

Thermal analysis of the healing material proved the presence of CaCO3.Precipitate promoted by chemical agent was degraded at a lower temperature than those promoted by biological agent and the combination of biological and chemical agent.This could indi-cate that some of the compounds of the biological agent promote the formation of larger crystals and/or a more stable crystal com-pared to those promoted by chemical agent only.It is not possible, however,to ensure that one self-healing agent is better than the other.Such a response can only be obtained by quantifying other properties of the precipitates such as its impact on permeability and strength of concrete.

As a?nal conclusion,the chemical admixtures studied herein have no signi?cant in?uence over the performance of the self-healing agents;therefore,they could be used without affecting self-healing of concrete.

Acknowledgments

Authors greatly appreciate Dr.ángel Leiva for the use of ther-mal analysis instruments,and https://www.360docs.net/doc/7a16276089.html,ardo Agosin for the use of laboratory facilities for bacteria cultivation.The help of Mauricio Guerra in the materials laboratory and of Jorge Torres in the bacteria cultivation are also appreciated.

References

[1]van der Zwaag S,van Dijk NH,Jonkers HM,Mookhoek SD,Sloof WG.Self-

healing behaviour in man-made engineering materials:bioinspired but taking

into account their intrinsic character.Philsos Trans A Math Phys Eng Sci 2009;367:1689–704.

[2]Li VC,Yang EH.Self-Healing in Concrete Materials.In:Zwaag S van der,editor.

Self Heal.Mater.An Altern.Approach to20Centuries Mater.Sci.,Springer;

2007,p.161–93.

[3]Edvardsen C.Water permeability and autogenous healing of cracks in concrete.

ACI Mater J1999;96:448–54.

[4]Hammes F,Verstraete W.Key roles of pH and calcium metabolism in microbial

carbonate precipitation.Rev Environ Sci Biotechnol2002;1:3–7.

[5]Jiang J,Chen S-F,Liu L,Yao H-B,Qiu Y-H,Gao M-R,et al.Template-free

polymorph discrimination and synthesis of calcium carbonate minerals.Chem Commun2009:5853–5.

[6]Dupraz C,Reid RP,Braissant O,Decho AW,Norman RS,Visscher PT.Processes

of carbonate precipitation in modern microbial mats.Earth-Sci Rev 2009;96:141–62.

[7]Decho AW.Overview of biopolymer-induced mineralization:what goes on in

bio?lms?Ecol Eng2010;36:137–44.

[8]Wei H,Shen Q,Zhao Y,Wang D,Xu D.Crystallization habit of calcium

carbonate in presence of sodium dodecyl sulfate and/or polypyrrolidone.J Cryst Growth2004;260:545–50.

[9]Jonkers HM,Schlangen E.Development of a bacteria-based self healing

concrete.Tailor Made Concr Struct New Solut Soc2008:425–30.

[10]Powell JF,Strange RE.Biochemical changes occurring during sporulation in

Bacillus species.Biochem J1956;63:661–8.

[11]Bang SS,Lippert JJ,Yerra U,Mulukutla S,Ramakrishnan V.Microbial calcite,a

bio-based smart nanomaterial in concrete remediation.Int J Smart Nano Mater 2010;1:28–39.

[12]Wang JY,De Belie N,Verstraete W.Diatomaceous earth as a protective vehicle

for bacteria applied for self-healing concrete.J Ind Microbiol Biotechnol 2012;39:567–77.

[13]Wiktor V,Jonkers HM.Quanti?cation of crack-healing in novel bacteria-based

self-healing concrete.Cem Concr Compos2011;33:763–70.

[14]Janto B,Ahmed A,Ito M,Liu J,Hicks DB,Pagni S,et al.Genome of alkaliphilic

Bacillus pseudo?rmus OF4reveals adaptations that support the ability to grow in an external pH range from7.5to11.4.Environ Microbiol 2011;13:3289–309.

[15]Jonkers HM,Thijssen A,Muyzer G,Copuroglu O,Schlangen E.Application of

bacteria as self-healing agent for the development of sustainable concrete.

Ecol Eng2010;36:230–5.

[16]Paulá,Lopez M.Assessing lightweight aggregate ef?ciency for maximizing

internal curing performance.ACI Mater J2011;108:385–93.

[17]Taylor HFW,Famy C,Scrivener KL.Delayed ettringite formation.Cem Concr

Res2001;31:683–93.

[18]Van Tittelboom K,De Belie N,De Muynck W,Verstraete https://www.360docs.net/doc/7a16276089.html,e of bacteria to

repair cracks in concrete.Cem Concr Res2010;40:157–66.

[19]Dick J,De Windt W,De Graef B,Saveyn H,Van der Meeren P,De Belie N,et al.

Bio-deposition of a calcium carbonate layer on degraded limestone by Bacillus species.Biodegradation2006;17:357–67.

[20]Kim HK,Park SJ,Han JI,Lee HK.Microbially mediated calcium carbonate

precipitation on normal and lightweight concrete.Constr Build Mater 2013;38:1073–82.

[21]De Muynck W,Cox K,De Belie N,Verstraete W,De Belie N.Bacterial carbonate

precipitation as an alternative surface treatment for concrete.Constr Build Mater2008;22:875–85.

[22]Cuthbert MO,Riley MS,Handley-Sidhu S,Renshaw JC,Tobler DJ,Phoenix VR,

et al.Controls on the rate of ureolysis and the morphology of carbonate precipitated by S.Pasteurii bio?lms and limits due to bacterial encapsulation.

Ecol Eng2012;41:32–40.

[23]Mann https://www.360docs.net/doc/7a16276089.html,A:Oxford University Press;2002.

[24]Kitamura M,Konno H,Yasui A,Masuoka H.Controlling factors and mechanism

of reactive crystallization of calcium carbonate polymorphs from calcium hydroxide suspensions.J Cryst Growth2002;236:323–32.

C.Stuckrath et al./Cement&Concrete Composites50(2014)10–1515