Effect of Cl Removal in MSWI Bottom Ash via Carbonation with CO2 andDecomposition Kinetics of Friedel’s Salt

Namil Um+

Environmental Resources Research, National Institute of Environmental Research,Hwangyeong-ro 42, Seo-gu, Incheon 404-708, Republic of Korea

In this study, the effect of Cl removal in bottom ash via a carbonation treatment with CO2 was investigated by comparing it with a waterwashing treatment. First, this was also focused on examining the existence of Cl contained in the bottom ash. The overall (soluble and insoluble)Cl content was close to that of bottom ash with fine particle. Next, the washing with water was confirmed and it was not effective in decreasingthe Cl content because of the existence of insoluble Cl. Whereas, the removal effect of Cl via carbonation with CO2 was very high compared tothe washing treatment because of the decomposition of Friedel’s salt (main insoluble Cl).

In addition, the kinetics data pertaining to the decomposed Friedel’s salt as the carbonation process proceeds was confirmed. Thetheoretical was well fitted to the kinetics data. The variation of the rate is constant upon decomposition with the reaction temperature followedthe Arrhenius equation (19.676 kJ/mol of activation energy) and the orders with respect to water-to-solution and particle size were also obtained.The decomposition rate of Friedel’s salt based on diffusion through the product layer of shrinking core model could be expressed by theequation. [doi:10.2320/matertrans.M-M2019811]

(Received June 30, 2018; Accepted February 12, 2019; Published April 5, 2019)

Keywords: MSWI bottom ash, insoluble Cl, carbonation, Friedel’s salt, decomposition kinetics

1. Introduction

Although the incineration method, which can reduce thevolume of the waste by 85­90%, is significant for treatingmunicipal solid waste, the ash remained after incineration hasbeen generated in an amount of nearly 400 thousand tons, inKorea. Approximately 90% of the ash is bottom ash with theremaining 10% fly ash. In the case of fly ash, it is unstableenvironmentally because of a high concentration of heavymetals. Whereas municipal solid waste incineration (MSWI)bottom ash, which can be described as heterogeneousparticles consisting of ferrous and non-ferrous metals,synthetic and natural ceramics, glass, minerals, etc., can bea potential substitute resource, especially as an aggregatematerial in the construction industry.1) However, the bottomash is subject to rigorous requirements with the limit valuesof pH, Cl, hazardous elements, amongst others, designated bythe law before recycling.

Among these limit values, the existence of a highconcentration of Cl may become the priority of strictrequirement because of the living habits of Koreans, wholike to eat salty food. In fact, the food with high Cl contentis well collected after being discarded and reused as therecycled-product contributing to the agriculture, such asfertilizers and feeds. Nonetheless, about 20­30% of a foodwaste is treated in incinerator intentionally because of theenergy- and the cost-effectiveness regarding the disposal offood waste. After incineration, the remained main Cl inbottom ash is composed of soluble salts such as NaCl andKCl, which can easily be dissolved by water. For the purposeof dechlorination, water washing is considered as a cheap buteffective method. However, it is difficult to reduce the Clcontent to the desired condition for recycling only by meansof washing due to the existence of insoluble Cl.

If the bottom ash is not subject to rigorous requirementswith the limit values of Cl, it can cause many environmentalproblems. For example, when the bottom ash is recycled asthe application with construction fill, sub-base material inroad construction, raw material of cement, etc., the remaininginsoluble Cl has a negative effect on the properties ofconstruction materials; potential Cl destroys the passivationof metal and results in many material and equipmentcorrosion. According to Kikuchi,2) the bottom ash containinginsoluble Cl is not applicable to ordinary Portland cement,because it can be dissolved after a certain period of timedue to pH change and carbonation with CO2. In addition,an insoluble Cl can lead to surrounding environmentalproblems; the Cl released to the environment has a negativeinfluence on the growth of plants and aquatic ecosystems.According to Gryndler et al.,3) increased concentrations ofsoil chloride can increase the ability of the soil to degradeand produce chlorinated organic compounds. Moreover, theformation and behavior of toxic heavy metals are mainlyaffected by the presence of Cl.4,5)

For this problem-solving with dechlorination effect ofinsoluble Cl, a number of related studies are currentlyunderway. According to Ito et al.,6) the acid leaching processhas been employed by using sulfuric or other acid solutionsfor removal of insoluble Cl. However, it has some demerits.Even if insoluble Cl is ionized by acid solutions, unwantedelements are dissolved along with the target Cl and additionalmethod is thus needed to treat the remaining waste-acid-solution. In another method, the thermal treatments such assintering, roasting and calcinations are often used prior to thewashing step because the insoluble Cl can be decomposed,some chlorides are vaporized, and the dechlorination effectis enhanced.7,8) However, a demerit of thermal treatment istheir high energy requirements and potential environmentalcontamination (generation of dust, gas, and etc.). Instead, tofind the effective method leading to the dechlorination, itis good to understand Friedel’s salt (Ca2Al(OH)6Cl·2H2O)

+Corresponding author, E-mail: inhaking96@hotmail.com, namil-um@

korea.kr

Materials Transactions, Vol. 60, No. 5 (2019) pp. 837 to 844©2019 The Mining and Materials Processing Institute of Japan

because Friedel’s salt was identified in the bottom ash asa major insoluble Cl. It also has important phenomenonwith the decomposition by CO2 gas.9,10) For this reason,the accelerated carbonation using a gas with a higher CO2

percentage can be perceived as one of the important treatmentprocess for the removal of insoluble Cl.11,12)

Therefore, in this study, the effect of Cl removal in bottomash via accelerated carbonation with CO2 was investigatedby comparison with washing using water. Carbonation wascarried out by using batch-type reactor kept at desiredtemperature and 30% CO2 was injected into it. Forconfirming the behavior of dechlorination after the desiredcarbonation time, the quantitative analysis of Friedel’s salt,which existed as a major insoluble Cl in bottom ash, wasperformed using the peak intensity of XRD pattern. Inaddition, the decomposition kinetics of Friedel’s salt duringcarbonation reaction was analyzed according to the shrinkingcore model. The equation was formulated with the rateconstant and decomposition fraction vs. carbonation-timeand fitted to the data of decomposition kinetics. Diffusionthrough the product layer was found suitable to explain thedecomposition kinetics, considered as the function of reactiontemperature, water-to-solid ratio, and particle size.

2. Experimental Methods

2.1 MaterialBottom ash sample was taken from a MSWI incineration

facility, which has the quenching process with a water-cooling system (cooling down the incinerated bottom ashwith high-temperature), located in a metropolitan area inKorea.

The bottom ash was sampled twice while on a conveyorbelt after being quenched using the water-cooling system; thequantity sampled each time was 500 kg and a total of 1000 kgwas obtained. After sampling, the coarse particles with thesize larger than 20mm were sieved out because most of thesecoarse particles are metals, glass, and ceramics and are non-representative in this study (more contaminants including Clare likely to be present in the fine particle). Next, the bottomash was dried at 100°C for 24 hrs. After drying, a magneticseparation was used to remove iron scrap. Then a sizeclassification was carried out according to a sieving methodwith under 0.15, 0.15­0.3, 0.3­0.6, 0.6­1.18, 1.18­2.36,2.36­4.75, and over 4.75mm.

2.2 CharacterizationThe initial amount of Cl in each fresh (untreated) bottom

ash sieved as required particle size fraction (CL0; mg/kg) wasobtained after dissolving it in acid solution. The concen-tration of Cl dissolved after acid digestion was measuredusing an Ion Chromatography (ICS-3000, Dionex).

To determine the initial amount of soluble Cl (CLSC;mg/kg), 100 g of the bottom ash sieved as required particlesize fraction was mixed with 1000mL of distilled water.The mixture was shaken on a vibration table at 300 rpm and20°C for 120min (according to the conditions obtained fromthe result of Cl removal percentage after desired washingtime, mentioned in section 2.3). Then, the leachate wasseparated using a filter paper with a 0.5 um pore-size. The

concentration of Cl of the leachate was measured using anIon Chromatograph.

For the quantitative analysis of the initial amount ofinsoluble Cl that exists as a Friedel’s salt (CLISC­F; mg/kg) inbottom ash, the author used the measurement of the peakintensity in the XRD pattern.13) First, Friedel’s salt wasprepared by using pure reagents with 3CaO·Al2O3 andCaCl2.14) A suspension of 0.02mol 3CaO·Al2O3 and0.02mol CaCl2 in 300ml of distilled water was stirred by amagnetic bar at 300 rpm in a 1000mL batch-type reactor keptat 50°C. Then, the synthesized sample was dried. Thecontents of Cl, Ca, and Al were measured using an IonChromatograph and an inductively coupled plasma atomicemission spectrometry (OPTIMA 5300DV, Perkin Elmer),respectively. It was found that the sample contained 7.21mass% Al, 7.81mass% Cl, and 24.84mass% Ca, indicatingthat molar ratio is Al:Cl:Ca = 1.2:1:2.8. Al and Ca in thesample were higher than that of the theoretical pure Friedel’ssalt, which is Al:Cl:Ca = 1:1:2. If all Cl exists as theFriedel’s salt, the synthesized Friedel’s salt was 61.7mass%as follows:

Calculated assay ðmass%Þ¼ 7:81 ðmass%Þ � ½280:59 ðFriedel’s salt molar mass;

g=molÞ=35:5 ðchloride molar mass; g=molÞ� ð1ÞHere, four standard products with different concentrations

of Friedel’s salt (2.1mass%, 10.4mass%, 30.2mass%, and61.7mass%) were prepared by mixing the synthesizedsample (61.7mass% of Friedel’s salt) and the reagent ofSiO2 (99.9% chemical grade; Sigma Aldrich, Ltd.). Then,10mass%MgO was added to each standard product for usingthe ratio of strongest XRD intensity between 2ª = 11.2°(Friedel’s salt) and 2ª = 43.0° (MgO) and these productswere measured using an X-ray diffractometer (Cu K¡

radiation, 45 kV, 100mA, PW3040/00, Philips). Figure 1,showing the standard line obtained by measuring the ratioof peak intensity (Friedel’s salt/MgO), helped to bring theinitial amount of Friedel’s salt. For example, if the XRD peakof the mixture of MgO and desired bottom ash shows theratio of Friedel’s salt/MgO of 4.66, 17800mg/kg of CLISC­Fcan be calculated by using the graph in Fig. 1.

0 5 10 15 20 250

10

20

30

40

50

60

70

8

6

4

2

CL IS

C-F

X10

4 (mg

/ kg)

Con

tent

of F

riede

l's s

alt (

wt%

)

Ratio of XRD peak intensity (Friedel's salt / MgO)

Y = 3.033X (R2 = 0.999)

0

Fig. 1 Standard line for quantitative analysis of Friedel’s salt by XRD peakintensity. The abbreviation represents: (CLISC­F) initial amount ofinsoluble Cl that exists as a Friedel’s salt.

N. Um838

The difference between CL0 and (CLSC + CLISC­F)determined the initial amount of insoluble Cl that exists asother chlorides except Friedel’s salt (CLISC­EF; mg/kg), i.e.sodalite (Na8Si6Al6O24Cl2), mentioned in section 3.1.

In addition, the samples were taken from each sievedbottom ash and were measured using an X-ray diffractometer,to confirm the mineralogical phases.

2.3 Washing with waterTo confirm the effect of Cl removal as a function of

washing time, the washing experiments with the bottomash sample being less than 0.15mm were performed in a2000mL batch-type reactor kept at 20°C of temperature. Thewater-to-solid ratio with mixture of 1000mL of distilledwater and 100 g of bottom ash was 10mL/g. The reactor wasshaken on a vibration table at 300 rpm during variouswashing times ranging from 0 to 120min. Then, the leachatewas separated using a filter paper and the concentration of Clwas measured using an Ion Chromatography. The exper-imental data taken at different washing times were made to fitthe remaining amount of Cl in bottom ash after the desiredwashing time (CLAW; mg/kg). Using CLAW, the removalpercentage of Cl after desired washing time (RCL­AW) can becalculated from the following equation:

RCL­AW ¼ ð1� CLAW=CL0Þ � 100 ð2ÞTo confirm the particle size distributions of the bottom ash

before and after washing, the measurements were performedusing a particle size analyzer (Mastersizer 2000, MalvernInstruments Ltd.).

In addition, the samples were taken from before and afterwashed-bottom ash and were measured using an X-raydiffractometer, to confirm the mineralogical phases.

2.4 Carbonation with CO2

The bottom ash sample taken for carbonation with CO2

had a particle size of less than 0.6mm owing to the high rateof Friedel’s salt content and specific surface area suggestingfurther evidence of the removal effect of Cl and thedecomposition kinetics in this study.

All the experiments were performed by putting desiredamount (67, 100, and 200 g) of bottom ash with differentparticle sizes (under 0.15, 0.15­0.3, and 0.3­0.6mm) into2000ml distilled water (10, 20, and 30mL/g water-to-solidratios) in a 3000ml batch-type reactor kept at desiredtemperature (20, 30, 40, and 50°C) and CO2 (30%). Themixture was stirred by a magnetic bar at 300 rpm duringcarbonation times ranging from 0 to 120min. Then, theleachate was separated and pH and RCL­AC (removalpercentage of Cl in bottom ash after desired carbonationtime) were measured. The concentration of Cl of the leachatewas measured using an Ion Chromatography. The RCL­AC canbe calculated from the following equation:

RCL­AC ¼ ð1� CLAC=CL0Þ � 100 ð3ÞHere, CLAC represents the remaining amount of Cl in

bottom ash after desired carbonation time (mg/kg).The samples were measured using an X-ray diffractometer,

to confirm the changes of the mineralogical phases of thecarbonated bottom ash.

3. Results and Discussions

3.1 Characterization of bottom ash containing ClTable 1 shows CL0, CLSC, CLISC­F, and CLISC­EF for each

particle size fraction of the bottom ash; the sum of CLSC,CLISC­F, and CLISC­EF is CL0 in each particle size. It is noticed

Table 1 CL0, CLSC, CLISC­F, and CLISC­EF for each particle size fraction of untreated bottom ash. The abbreviations represent:(CL0 = CLSC + CLISC­F + CLISC­EF) initial amount of Cl; (CLSC) initial amount of soluble Cl; (CLISC­F) initial amount of insoluble Clthat exists as a Friedel’s salt; (CLISC­EF) initial amount of insoluble Cl that exists as other chlorides except Friedel’s salt.

1)Each data point was determined in triplicate.2)Error ranges corresponding to the standard deviation of the data were estimated for each case.

Effect of Cl Removal in MSWI Bottom Ash via Carbonation with CO2 and Decomposition Kinetics of Friedel’s Salt 839

that the CL0, CLSC, CLISC­F, and CLISC­EF are correlated withthe particle size. They increased with decreasing particlessizes and the fraction with a particle size under 0.15mm hasthe highest value. According to Chen et al.,15) a negativecorrelation, showing 0.886 of the correlation coefficient (R2),exists between Cl content and the particle size, indicating thatCl accumulates easily on fine particles. As shown in Fig. 2, itpresents that the overall Cl content (sum of CL0s) is close tothat of bottom ash with fine particles. The fraction with thesize smaller than 0.15mm only accounts for 7.7% of the totalweight of bottom ash, but accumulates more than 34.7% ofthe overall Cl content. Almost half of overall Cl content iscontained in the fraction of the size smaller than 0.3mm. Inaddition, CLSC and CLISC­F also increased with a decrease inthe particle size. In case of CLSC, almost half of the amount iscontained in the size smaller than 0.6mm, whereas more thanhalf of the overall CLISC­F (about 55.4%) is contained in thesize smaller than 0.15mm. However, the values of CLISC­EFwere relatively low in all particle sizes.

Next, the mineralogical phase of bottom ash wasconfirmed by the XRD patterns in Fig. 3. As shown inpatterns, the peaks of Friedel’s salt, ettringite, portlandite,etc. existed and the intensities of them increased with adecrease in the particle size. Particularly, the XRD patternrevealed that the intensity of Friedel’s salt was the largest.In addition, it can be seen that the small peak of sodalite(Na8Si6Al6O24Cl2), which is one of other insoluble Cl exceptFriedel’s salt, was detected. The facts about insoluble Clsupport the aforementioned results in Table 1. To explain thereason for the existence of Friedel’s salt, it is necessary tounderstand the water quenching process. Water quenching isthe process with water cooling system for cooling down thebottom ash with high-temperature after incineration and ithas been determined to have a high impact on bottom ashcharacteristics. Inkaew et al.16) examined that water quench-ing affected the change of bottom ash’s morphology byreducing temperature, the alteration of the chemicalcomposition, and the enhancement of the quench products

formation (i.e., portlandite, ettringite, and Friedel’s salt). Thechemical composition of the bottom ash shown in manyresearcher’s data indicated that the quenching process playsa crucial role to provide the hydrate precipitation of Cl suchas Friedel’s salt on the bottom ash.8,17,18) In addition, theexistence of Ca is related to controlling the formation ofportlandite (Ca(OH)2) during the quenching process, as wellas calcite (CaCO3) formed by the reaction between Ca(OH)2and CO2 in the air. Ettringite (Ca6Al2(SO4)3(OH)12·26H2O)can also be controlled by the quenching process.

3.2 Washing treatmentThe waste washing has its disadvantages in terms of

two respects. First, the insoluble Cl is not expected to bedecomposed leading to the decrease of the Cl content.According to the washing method in section 2.3, the resultsin Fig. 4, showing the behavior of RCL­AW as a function of

0

5

10

15

20

25

30

35

40

45

50

0

5

10

15

20

25

30

35

40

45

50

under0.15

0.15-0.3

0.3-0.6

0.6-1.18

1.18-2.36

2.36-4.75

bottom ash in each size / total bottom

ash (wt.%

)

CL S

C,CL I

SC-F

, and

CL I

SC-E

F / ov

eral

l Cl c

onte

nt (w

t.%)

over4.75

CLISC-F

CLSC

CLISC-EF

Fig. 2 Weight percentages of CLSC, CLISC­F, and CLISC­EF per overall Clcontent (sum of CL0s) and bottom ash in each particle size pertotal bottom ash. The abbreviations represent: (CL0 = CLSC +

CLISC­F + CLISC­EF) initial amount of Cl; (CLSC) initial amount of solubleCl; (CLISC­F) initial amount of insoluble Cl that exists as a Friedel’s salt;(CLISC­EF) initial amount of insoluble Cl that exists as other chloridesexcept Friedel’s salt.

6 8 10 12 14 16 18 20 22 24 26 28 30

6 8 10 12 14 16 18 20 22 24 26 28 30

31

6

5

2

-0.15mm

0.3-0.6mm

0.15-0.3mm

0.6-1.18mm

1.18-2.36mm

2.36-4.75mm

+4.75mm

Inte

nsity

(arb

itrar

y un

it)

2θ / degree

4

1. Ettringite (Ca6Al2(SO4)3(OH)12.26H2O)2. Friedel's salt (Ca2Al(OH)6Cl2.H2O)3. Portlandite (Ca(OH)2)4. Sodalite (Na8Si6Al6O24Cl2)5. Quartz (SiO2)6. Calcite (CaCO3)

Fig. 3 XRD patterns of untreated bottom ash as a function of the particlesize.

0 20 40 60 80 100 120

Washing or carbonation time / min

RCL-AC

RCL-AW

pH8

RC

L

0

60

40

20

80

100

11

pH

9

13

12

10

Untreated (RCL)

Untreated (pH)

Fig. 4 RCL­AW of washed bottom ash and RCL­AC, and pH of carbonatedbottom ash as a function of the washing or carbonation time (washing orcarbonation conditions: 30% CO2 concentration (carbonation only),10mL/g water-to-solid ratio, 20°C reaction temperature and particle size<0.15mm). The abbreviations represent: (RCL­AW) removal percentage ofCl after desired washing time; (RCL­AC) removal percentage of Cl afterdesired carbonation time.

N. Um840

washing time, indicate that the washing process is effectivefor the removal of the only soluble Cl in a short time; the onlyhalf of CL0, indicating 48.3% of RCL­AW, was removed andthe solution reached almost its maximum at less than 20minbecause of the insoluble Cl content. Second, the unexpectedettringite (mentioned as quench product) is formed fromthe water washing. According to the literatures,19,20) the pHrange of stability for ettringite is between 10.5 and 13.According to the following equation, the formation is favoredover mono-sulfate at below 50°C.

6Ca2þ þ 2Al3þ þ 3SO2�4 þ 38H2O

! 12Hþ þ Ca6Al2ðSO4Þ3ðOHÞ12�26H2O ð4ÞIn case of a washing process, as the washing time

proceeds, the washing water becomes high alkaline conditionwith a pH in excess of 12 at the temperature of 25°C, and thiscondition leads to the formation of ettringite during thewashing process. Indeed, the XRD diffractions of untreatedand washed bottom ash show the formation of ettringite, asshown in Fig. 5. The peaks of KCl and NaCl disappearedafter washing, whereas that of ettringite increased. Inaddition, the data for the change of particle size distributionsbefore and after washing, as shown in Fig. 6, may providefurther evidence related with the dissolution of soluble Clparticles and the formation of ettringite; some of them wereagglomerated with the bottom ash, as shown in scanningelectron microscopy image of the surface of bottom ash afterwashing (in Fig. 6).

Ettringite has an important phenomenon. This structureconsists of columns of {Ca6[Al(OH)6]2·24H2O}6+ with theinter-column space (channels) occupied by SO2�

4 molecules,and it can be substitutable for heavy metal’s oxyanions, suchas CrO2�

4 and AsO2�4 .21,22) However, the important phe-

nomenon23,24) is that heavy metal ions substituted intoettringite can be easily released from the particle surface tothe outside because the ettringite is easily decomposed bycarbonation with CO2. In addition, the occurrence of volume

change due to the carbonation with CO2 in the bottom ashcan be also expected. Therefore, when recycled as theapplication with construction fill, sub-base material in roadconstruction, etc. their phenomenon may not lead to suitablesubstitute for natural resource.

3.3 Accelerated carbonation treatmentFor confirmation of the carbonation effect, the behaviors

of RCL­AC and pH observed at different carbonation timeswere investigated by comparing with RCL­AW in Fig. 4. Asthe result, the pH fell into under pH 9.5 after 20min ofcarbonation. It indicates that the neutralization of pH aftercarbonation is controlled mainly by the decomposition ofquench products (alkaline compounds with high value ofpH). Next, the removal effect of Cl is very high comparedto washing treatment. The removal percentage of Cl, whichwas only 50.8% (RCL­AW) using washing, increased to 95.4%(RCL­AC) at 120min of carbonation time. However, itcouldn’t reach 100% because of the existence of otherinsoluble chlorides except Friedel’s salt, such as sodalite,which is not easily decomposed with CO2. According tothe literatures,23,24) the quench products are the main alkalineinorganic materials in the reaction with CO2. Thesecarbonation reactions of Friedel’s salt (eq. (5)), portlandite(eq. (6)), and ettringite (eq. (7)) are described below:

2½Ca2AlðOHÞ6Cl�2H2O� þ 3CO2

! 3CaCO3 þ Al2O3�xH2Oþ CaCl2 þ ð10� xÞH2O ð5ÞCaðOHÞ2 þ CO2 ! CaCO3 þ H2O ð6ÞCa6Al2ðSO4Þ3ðOHÞ12�26H2Oþ 3CO2

! 3CaCO3 þ 3½CaSO4�2H2O� þ Al2O3�xH2O

þ ð26� xÞH2O ð7ÞFigure 7, showing XRD patterns of the bottom ash at

different carbonation times, provides an evidence of theabove chemical reactions. The intensities of Friedel’s salt,portlandite, and ettringite in untreated bottom ash decreasedwith the carbonation reaction and no quench product peakwas observed after 120min of carbonation while that of

5 10 15 20 25 30 35 40 45 50

5 10 15 20 25 30 35 40 45 50

3

Before washing

After washing

3

45

5

1. Ettringite (Ca6Al2(SO4)3(OH)12.26H2O)2. Friedel's salt (Ca2Al(OH)6Cl2.H2O)3. Sodalite (Na8Si6Al6O24Cl2)4. KCl5. NaCl

2

1

Inte

nsity

(arb

itrar

y un

it)

2θ / degree

Fig. 5 XRD patterns of untreated bottom ash and bottom ash treated bywashing with distilled water (washing conditions: 120min washing time,10mL/g water-to-solid ratios, 20°C reaction temperature, and particlesize <0.15mm).

0.1 1 10 100 10000

1

2

3

4

5

6

7

Untreated Washed

Particle Size (um)

volu

me

(%)

Ettringite

Surface of bottom ash after washing

Fig. 6 Particle size distribution of untreated and washed bottom ash(washing conditions: 120min washing time, 10mL/g water-to-solidratios, 20°C reaction temperature, and particle size <0.15mm).

Effect of Cl Removal in MSWI Bottom Ash via Carbonation with CO2 and Decomposition Kinetics of Friedel’s Salt 841

calcite and gypsum increased. However, even though thepeak of Al2O3·xH2O is expected through the formation of anew Al-material from eqs. (5) and (7), it was not detectedbecause Al would precipitate mainly as amorphous Al-material (gel-like).25)

3.4 Decomposition kinetics of Friedel’s saltBefore studying the decomposition kinetics of Friedel’s

salt, we have to know the carbonation mechanism with CO2

dissolution into water (step 1) and the reaction between CO2

dissolved into liquid and bottom ash (step 2).24) In the case ofStep 1, CO2 dissolution into liquid (eqs. (8)­(10)) is depend-ent on injection type, CO2 pressure and concentration,temperature, etc. Carbon dioxide enters the water throughequilibrium with the atmosphere (eq. (8)) and it can reactwith the water to form carbonic acid (eq. (9)). Dissolved CO2

in the form of H2CO3 may lose protons to form bicarbonate,HCO3

3¹, and carbonate, CO33¹ (eq. (10)).

CO2ðgÞ $ CO2ðaqÞ ð8ÞCO2ðaqÞ þ H2O $ H2CO3 ð9ÞH2CO3 $ HCO�

3 þ Hþ $ CO2�3 þ 2Hþ ð10Þ

Since these consecutive chemical reactions in Step 1 aremuch faster than those of Step 2, the carbonation reactionbetween the bottom ash and CO2 in liquid (Step 2) becomes arate-determining step. In addition, in Step 2, the carbonationreaction may be generally controlled by the solid-liquidheterogeneous system.23,24) Thus, it can be expressed asthe shrinking core model with three steps such as surfacechemical reaction, diffusion through the product layer, andfluid film diffusion control. Because no fluid film covers theunreacted bottom ash particle as the reaction proceeds, therecould be only surface chemical reaction and diffusion throughthe product layer, as follows.

Surface chemical reaction: ½1� ð1�XtÞ1=3� ¼ kRt ð11ÞDiffusion through the product layer:

½1� ð2=3ÞXt � ð1�XtÞ2=3� ¼ kDt ð12Þ

Assuming that this carbonation process is simply describedby the decomposition of Friedel’s salt (eq. (5)), kR and kDare the reaction rate constants (h¹1) of Friedel’s saltdecomposition for chemical reaction and diffusion control,respectively. The calculation was then performed as eq. (13);Xt is the decomposed fraction vs. carbonation time t.

Xt ¼ XCO2=CLISC­F ð13ÞIn this equation, XCO2 is the amount of dissolved Cl, which

existed as a Friedel’s salt initially, after desired carbonationtime (mg/kg). The values of XCO2 were measured by usingthe measurement of the peak intensity in the XRD pattern andthe standard line for quantitative analysis of Friedel’s salt, asmentioned in section 2.2.

To consider the temperature effect on the decompositionkinetics of Friedel’s salt, the carbonation of the bottom ashwith less than 0.15mm particle size was performed using30% CO2 concentration at different temperatures rangingfrom 20 to 50°C. The water-to-solid ratio was 10mL/g.Figure 8(a) indicates that the decomposition rate decreasedwith a decrease in the temperature and the reaction under allconditions was reached the maximum value of Xt after120min. Figure 8(b) and Fig. 8(c) present the experimentaldata plotted on 1 ¹ (1 ¹ Xt)1/3 and 1 ¹ (2/3)Xt ¹ (1 ¹ Xt)2/3,respectively. The results show that the correlation coefficients(R2) of diffusion through of the product layer are closer to1 than those of surface chemical reaction. In addition, ifconsidered as the three functions of temperature, waste-to-solid ratio, and particle size, the decomposition rate constantof Friedel’s salt can be expressed as follows:

kD ¼ k00e�Ea=RT ðW=SÞmd0n ð14Þ

Where Ea is the activation energy (kJ/mol); T, the reactiontemperature (K); R, the ideal gas constant, 8.314 © 10¹3

(kJ/mol); W/S, water-to-solid ratio (mL/g); d0, particle size(µm); m and n, constants; and k00 is pre-exponential factor.Rearranging eq. (14), three equations are obtained as follows:

k1 ¼ k01e�Ea=RT

ðbetween rate constant and activation energyÞ ð15Þk2 ¼ k02ðW=SÞm ðbetween rate constant and

water-to-solid ratio constantÞ ð16Þk3 ¼ k03d0

n ðbetween rate constant and particle

size constantÞ ð17ÞWhere k1­3 (here kD ¼ k1 � k2 � k3) and k01­3 (here k00 ¼

k01 � k02 � k03) are the rate constants and the pre-exponentialfactors, respectively. In eqs. (15)­(17), the ln k1­3 valuescalculated from these k1­3 values were plotted against 1/T,lnW/S, and ln d0.

When eq. (15) is applied, the rate constants obtained fromthe linear relationship between ln k and 1/T in Fig. 8(c) wereused to determine the Arrhenius plot and activation energywas 19.676 kJ/mol, as shown in Fig. 8(d). For examiningthe effect of water-to-solid ratio, the rate constants for variousW/S values (10, 20, and 30mL/g) were determined by usingthe linear regressions obtained from the data in Fig. 9(a). Alinear plot between ln k and lnW/S in Fig. 9(b) indicates thatm of the water-to-solid ratio constant was calculated to be¹0.065. This value shows no effect of water-to-solid ratio on

6 8 10 12 14 16 18 20 22 24 26 28 30

6 8 10 12 14 16 18 20 22 24 26 28 30

6

3120min

Inte

nsity

(arb

itrar

y un

it)

1. Ettringite (Ca6Al2(SO4)3(OH)12.26H2O)2. Friedel's Salt (Ca2Al(OH)6Cl2.H2O)3. Gypsum (CaSO4.2H2O)4. Potlandite (Ca(OH)2)5. Sodalite (Na8Si6Al6O24Cl2)6. Calcite (CaCO3)

60min

30min

45min

20min

10min

5min

0min

2θ / degree

1

2

4 5

Fig. 7 XRD patterns of bottom ash carbonated at different reaction times(carbonation condition: 30% CO2 concentration, 20°C reaction temper-ature, 10mL/g water-to-solid ratio and a particle size <0.5mm).

N. Um842

the decomposition rate. In addition, the data in Fig. 10(a)determined the effect of particle sizes (mean diameter d0;75 (under 0.15mm), 225 (0.15­0.3mm), and 450 µm (0.3­0.6mm)) on the decomposition rate of Friedel’s salt and thedecrease in particle size increased the decomposition rate.The linear plot between ln k and ln d0 in Fig. 10(b) indicatesthat n of the particle size constant was calculated to be¹0.234.

The calculated activation energy and the constants ofwater-to-solid ratio and particle size conform to the shrinkingcore model for diffusion through the product layer. Thus thedecomposition of Friedel’s salt during the carbonationreaction with CO2 can be clearly presented as follows:

½1� ð2=3ÞXt � ð1�XtÞ2=3�¼ kDt ¼ k00ðW=SÞ�0:065d0

�0:234 expð�19:676=RTÞ t ð18Þ

0.00

0.05

0.10

0.15

0.20

0.25

0.30

(c)

20oC: Y=0.0045X, R2=0.989330oC: Y=0.0057X, R2=0.990840OC: Y=0.0079X, R2=0.995450OC: Y=0.0093X, R2=0.9913

1 - (

2/3)X

t - (1

- X

t)2/3

Time, t / min

-5.4

-5.2

-5.0

-4.8

-4.6

(d)

ln (k

/ h-1

)

(1 / T) X 103 / K-1

0.0

0.2

0.4

0.6

0.8

1.0

20oC 30oC 40oC 50oC

Xt

Time, t / min

(a)

0 5 10 15 20 25 30 35

3.05 3.10 3.15 3.20 3.25 3.30 3.35 3.40 3.45

0 20 40 60 80 100 120

0 5 10 15 20 25 30 350.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Time, t / min

(b)

20oC: Y=0.0172X, R2=0.931930oC: Y=0.0198X, R2=0.938840OC: Y=0.0242X, R2=0.958050OC: Y=0.0272X, R2=0.9557

1 - (

1 - X

t)1/3

Fig. 8 Effect of the reaction temperature on the decomposition kinetics ofFriedel’s salt (a), plots of 1 ¹ (1 ¹ Xt)1/3 versus reaction time for differentreaction temperatures (b), plots of 1 ¹ (2/3)Xt ¹ (1 ¹ Xt)2/3 (c) and effectof reaction temperature on the decomposition rate constants (d) (30% CO2

concentration, 10mL/g water-to-solid, and particle size <0.15mm).

0 20 40 60 80 100 1200.0

0.2

0.4

0.6

0.8

1.0

2.2 2.4 2.6 2.8 3.0 3.2 3.4-6.0

-5.8

-5.6

-5.4

-5.2

-5.0

(b)

ln (k

/ h-1)

ln (W/S / mL/g)

(a)W/S=10W/S=20W/S=30

Xt

Time, t / min

Fig. 9 Effect of water-to-solid ratio on the decomposition kinetics ofFriedel’s salt (a) and on the decomposition rate constants (b) (30% CO2

concentration, 20°C temperature, and particle size <0.15mm).

0.0

0.2

0.4

0.6

0.8

1.0

0 20 40 60 80 100 120

4.4 4.8 5.2 5.6 6.0 6.4-6.0

-5.9

-5.8

-5.7

-5.6

-5.5

-5.4

-5.3

(b)

ln (k

/ h-1)

ln (d0 / μm)

(a) 75 μm 225 μm 450 μm

Xt

Time, t / min

Fig. 10 Effect of particle size on the decomposition kinetics of Friedel’ssalt (a) and effect of particle size on the decomposition rate constants(b) (30% CO2 concentration, 10mL/g water-to-solid, and 20°C temper-ature). d0 indicates the mean diameter of bottom ash; 75 µm, mean ofunder 0.15mm; 225µm, mean of 0.15­0.3mm; and 450µm, mean of0.3­0.6mm.

Effect of Cl Removal in MSWI Bottom Ash via Carbonation with CO2 and Decomposition Kinetics of Friedel’s Salt 843

Plotting of 1 ¹ (2/3)Xt ¹ (1 ¹ Xt)2/3 against (W/S)¹0.065d0¹0.234 exp(¹19.676/RT) t gives a k00 amount of 47.5.

4. Conclusion

In this study, Cl removal effect, related with thedecomposition of Friedel’s salt, on the accelerated carbo-nation of MSWI bottom ash was investigated by comparingwith a water washing. In addition, the decomposition kineticsof Friedel’s salt was also examined. The obtained results wereas follows.(1) In Korea, the bottom ash generated after incineration

of municipal solid waste has a high Cl content. Itconsists of the soluble and insoluble chloride, especiallyFriedel’s salt, which is formed during quenchingprocess with water in incineration facility, wasidentified as a major insoluble Cl. The overall Clcontent was close to that of bottom ash with fineparticle.

(2) Water washing, considered a cheap and simple methodfor the purpose of dechlorination, was carried out and itwas not effective in reducing the Cl content because ofthe existence of insoluble Cl; the only half of Cl wasremoved. In addition, the result indicated that thewashing does not provide stable condition because theettringite formed during washing has a high reactivitywith CO2 to affect the volume change.

(3) Removal effect of Cl was obtained from the decom-position of Friedel’s salt according to the reaction withCO2 and was very high compared to the washing. Theremoval percentage of Cl, which was only half of Clusing washing, increased to more than 95% aftercarbonation.

(4) Theoretical model was well fitted to the kinetics datapertaining to the decomposed Friedel’s salt. Consider-ing as the function of temperature (Ea = 19.676kJ/mol), water-to-solid ratio (W/S), and particle size(d0), the decomposition rate based on diffusion throughthe product layer of shrinking core model can beexpressed by the following equation (here Xt is thedecomposed fraction vs. carbonation time t.):

½1� ð2=3ÞXt � ð1�XtÞ2=3�¼ kt ¼ 47:5ðW=SÞ�0:065d0

�0:234 expð�19:676=RTÞ t:

Acknowledgement

This work was supported by a grant from the NationalInstitute of Environmental Research (NIER), funded by theMinistry of Environment (MOE) of the Republic of Korea(NIER-2014-01-01-015).

REFERENCES

1) J.M. Chimenos, M. Segarra, M.A. Fernandez and F. Espiell: J. Hazard.Mater. A 64 (1999) 211­222.

2) R. Kikuchi: Resour. Conserv. Recycling 31 (2001) 137­147.3) M. Gryndler, J. Rohlenova, J. Kopecky and M. Matucha: Chemosphere

71 (2008) 1401­1408.4) K.Y. Chiang, K.S. Wang, F.L. Lin and W.T. Chu: Sci. Total Environ.

203 (1997) 129­140.5) K.S. Wang, K.Y. Chiang, S.M. Lin, C.C. Tsai and C.J. Sun:

Chemosphere 38 (1999) 1833­1849.6) R. Ito, T. Fujita, J. Sadaki, Y. Matsumoto and J. Ahn: Int. J. Sci. Eng.

Res. 13 (2006) 70­74.7) S. Wei, F. Chang, Y. Shen, M. Tsai and C. Ko: J. Hazard. Mater. 237

(2012) 116­120.8) S. Yang, A. Saffarzadeh, T. Shimaoka and T. Kawano: J. Hazard. Mater.

267 (2014) 214­220.9) S. Goñi and A. Guerrero: Cement Concr. Res. 33 (2003) 21­26.10) A.K. Suryavanshi and R.N. Swamy: Cement Concr. Res. 26 (1996)

729­741.11) S. Arickx, T. Van Gerven and C. Vandecasteele: J. Hazard. Mater. 137

(2006) 235­243.12) T. Van Gerven, E. Van Keer, S. Arickx, M. Jaspers, G. Wauters and C.

Vandecasteele: Waste Manag. 25 (2005) 291­300.13) R. Ito, G. Dodbiba, T. Fujita and J.W. Ahn: Waste Manag. 28 (2008)

1317­1323.14) S. Rasheeduzzafar, S.E. Hussain and S. Al-Saadoun: Cement Concr.

Res. 21 (1991) 777­794.15) C.H. Chen and I.J. Chiou: J. Hazard. Mater. 148 (2007) 346­352.16) K. Inkaew, A. Saffarzadeh and T. Shimaoka: Waste Manag. 52 (2016)

159­168.17) M.V.A. Florea and H.J.H. Brouwers: Cement Concr. Res. 42 (2012)

282­290.18) A.G. Kalinichev and R.J. Kirkpatrick: Chem. Mater. 14 (2002) 3539­

3549.19) M. Chrysochoou and D. Dermatas: J. Hazard. Mater. 136 (2006) 20­33.20) D. Damidot and F.P. Glasser: Cement Concr. Res. 23 (1993) 1195­

1204.21) S.C.B. Myneni, S.J. Traina, T.J. Logan and G.A. Waychunas: Environ.

Sci. Technol. 31 (1997) 1761­1768.22) M. Zhang and E.J. Reardon: Environ. Sci. Technol. 37 (2003) 2947­

2952.23) N. Um, S.Y. Nam and J.W. Ahn: Mater. Trans. 54 (2013) 1510­1516.24) N. Um and J.W. Ahn: Process Saf. Environ. 111 (2017) 560­568.25) J.A. Meima, R.D. van der Weijden, T.T. Eighmy and R.N.J. Comans:

Appl. Geochem. 17 (2002) 1503­1513.

N. Um844