AN ABSTRACT OF THE THESIS OF

Punnchalee Laothumthut for the degree of Master of Science in Chemical Engineering

presented on March 20, 1996. Title: Chemistry and Chemical Process Studies of Fluoride

Removal in a Silicon Wafer Manufacturing Wastewater Treatment Plant.

Abstract approved:

William J. Frederick

Fluoride is discharged as the effluent in many chemical industries including silicon

wafer manufacturing. The treatment of fluoride with lime is the most commonly usedmethod in fluoride removal. Industrial waste treatment can have difficulties to removefluoride to the required discharge limit. The chemistry and chemical processes of fluorideremoval in a silicon wafer manufacturing waste treatment plant was studied. The studydivided into mass balance, equilibrium effects and kinetic effects to determine the cause ofdifficulty in fluoride removal. In mass balance, the time variation of flowrate and fluoride

concentration data were measured. The mass balance analysis for fluoride ions was done

in the treatment plant. The amount of lime used in the plant indicated that underlimingwas not the cause of high fluoride concentration in the effluent.

In equilibrium study, Chem Sage 3.0 software was used to study the ions effects inthe waste to fluoride chemical activity. The components that caused the non -CaF2formation were Al3+, Na2CO3, and H3PO4. The Al 3+ caused the AIF3 and A1F2+ formation

at acidic conditions. The Na2CO3 compound competed with F to react with availableCa2+ and formed CaCO3 compound in the system. If there was a significant enoughquantities of Na2CO3 in the system, it could reduce the amount of free calcium and resultthe high amount of free fluoride ion in the system. Lastly, the H3PO4 component couldresult the formation of Cas(PO4)3F in the system. Significant amount Cas(PO4)3F

compound could exceed the discharge limit if it were to remain as fine suspended particles

in the discharge from treatment plant.

Redacted for Privacy

In kinetic study, the experiment of HF and different ratios of lime reaction were

conducted. From the rate data analysis, the rate of reaction and rate constants were

obtained. The reaction of HF at high enough ratios of lime was very rapid compared to

the average residence time in the reaction tank of treatment plant. The prediction of amixed flow conversion in the CSTR was calculated at different residence time. The results

of the prediction determined the minimum residence time required to achieve the high F

conversion in the reaction tank in order to meet the effluent discharge limit regulations.

Chemistry and Chemical Process Studies of Fluoride Removal in a

Silicon Wafer Manufacturing Wastewater Treatment Plant

by

Punnchalee Laothumthut

A THESIS

submitted to

Oregon State University

in partial fulfillment of

the requirement for the

degree of

Master of Science

Presented March 20, 1996

Commencement June 1996

Master of Science thesis of Punnchalee Laothumthut presented on March 20, 1996

APPROVED:

Major ofessor, representing Chemical Engineering

Ch. 'of Department of Chemical Engineering

Dean of Gra to School

I understand that my thesis will become part of the permanent collection of Oregon StateUniversity libraries. My signature below authorizes release of my thesis to any readerupon request.

Punnchalee Laothumthut, Author

Redacted for Privacy

Redacted for Privacy

Redacted for Privacy

Redacted for Privacy

ACKNOWLEDGMENTS

No undertaking of this magnitude can be accomplished alone, I would like to

express appreciation for the contribution of the following:

Wacker Siltronic Corp., Portland, OR for supporting this research

Tom McCue, Chip Bloomer, Tom Rothchild, Diana, and every operators in

wastewater treatment plant for cooperation during research at Wacker Siltronic Corp.

Harris Group & Barry Kelly

Dr. Jim Frederick, Dr. Shoichi Kimura, Dr. Kristina Lisa, Dr. Goran Jovanovich,

Dr. Milo Koretsky, Dr. Greg Rorrer, Dr. Skip Rochefort, and all the professors for

providing valuable education and help during this research.

Mike Littau for his contribution in this research

Yan Lu, Varut Pimolmas, Ron Armstrong, Udom Techakijkachorn, and all the fellow

colleagues in the Chemical Engineering Dept.

Special appreciation to Dr. Jim Frederick, my advisor, for his valuable suggestions,

his kindness, and help during the research and the thesis writing.

Lastly, I would like to thank my family in Thailand, especially my parents, and theBe llis family for the great support and encouragement during this research.

TABLE OF CONTENTS

Page

1. INTRODUCTION 1

2. OBJECTIVE 4

3. LITERATURE REVIEW 6

3.1 MASS BALANCE 6

3.1.1 Chemistry of hydrofluoric acid 63.1.2 Chemistry of precipitation with Lime 73.1.3 Lime precipitation with aluminum sulfate 103.1.4 Aluminum effects in acidic waters 13

3.2 REACTION KINETICS 14

4. ANALYTICAL METHODS 18

4.1 pH AND ION-SELECTIVE ELECTRODE MEASUREMENTS 18

4.2 CAPILLARY ELECTROPHORESIS ANALYSIS OF CARBONATE 19

4.3 SPECTROPHOTOMETRIC ANALYSIS OF ANION SPECIES 20

5. EXPERIMENTAL PROCEDURE5.1 EXPERIMENTAL PLAN FOR THE KINETIC STUDY 22

5.2 EQUIPMENTS 23

5.3 EXPERIMENTAL PROCEDURE 23

5.3.1 Preparing the solutions and running the experiment 245.3.2 Measuring fluoride 25

6. RESULTS AND DISCUSSION 26

6.1 MASS BALANCE ANALYSIS 26

TABLE OF CONTENTS (Continued)

Page

6.1.1 Wastewater treatment process 266.1.2 Evaluation of mass balance 296.1.3 AlF2+ formation analysis 346.1.4 Conclusion 34

6.2 EQUILIBRIUM EFFECTS 35

6.2.1 Defining the CAD reaction tank in the ChemSage environment file 366.2.2 Comparison of Chem Sage 3.0 solubility graphs to experimental data 426.2.3 Possible reaction mechanisms for increased fluoride discharge 466.2.4 Effect of input components on Harris mass balance

of CAD reaction tank 536.2.5 Conclusion 66

6.3 KINETIC STUDY 69

6.3.1 Reaction mechanisms for rate limiting step 696.3.2 Qualitative analysis of kinetic data 696.3.3 Evaluation of rate of reaction 726.3.4 Distribution of the residence time in the CAD reaction tank 746.3.5 Conclusion 78

7. SUMMARY AND FUTURE WORK 80

7.1 SUMMARY 80

7.2 FUTURE WORK 81

BIBLIOGRAPHY 82

APPENDICES 86

APPENDIX A Pitzer ion coefficients for Pitzdata.txt data file 87

APPENDIX B Comparison of solubilities between results by Chem Sage andexperimental data 95

APPENDIX C Experimental data for equilibrium study 116

APPENDIX D The rate data for kinetic study 117

LIST OF FIGURES

Figure Pate

3.1 Gas/liquid equilibrium of fluoride in water4 log Fr(mg/1) vs. pH 7

3.2 Experimental results for residual soluble fluoride at initialconcentration 87 mg/liter after lime precipitations 9

3.3 Experimental results: residual fluoride after precipitation with lime' 10

3.4 Effect of solution pH on alum addition required to reachtarget fluoride concentration. Initial sodium fluoride solution 15 mg/1 F 7 11

3.5 Alum requirement as a function of fluoride concentrationResidual fluoride 4.8 mg/1 F, at pH 7 12

3.6 Alum requirement to decrease fluoride in concentration for asodium fluoride solution containing initially 78 mg/1, pH 7.07 13

5.1 Schematic of experimental procedure 24

6.1 Schematic flow diagram of CAD wastewater treatment plant 27

6.2 Schematic mass balance in CAD reaction tank 28

6.3 Wacker Siltronic plant's record of reaction tank lime valve(% open),flow from forwarding sump, flow out of reaction tank, pH measuredin the CAD reaction tank and in lamella settlers 32

6.4 Fluoride concentration variation over 8 hours(measurement of samples from forwarding sump on January 15, 1995) 33

6.5 1Mol/L Al, 1.5 Mol/L Ca, 3 Mol/L F in 25 °CSystem Calcium and Fluoride Species as Function of pH 47

6.6 1.5 Mol/L Al, 1.5 Mol/L Ca, 3 Mol/L F in 25 °CSystem Calcium and Fluoride Species as Function of pH 47

6.7 3 Mol/L Al, 3 Mol/L Ca, 3 Mol/L F 25 °C SystemCalcium and Fluoride Species as a Functionof pH 48

6.8 4 Mol/L Ca, 1 Mol/L Al, 3 Mol/L F in 25 °C System,Calcium and Fluoride species as Function of pH 49

6.9 Simulation of Lawler and Williams System pH Varied by Ca(OH)2 51

6.10 Simulation of Lawler and Williams System nEl Varied with NaOH 52

6.11 Effect of Al2(SO4)3 on System(Mass Balance Concentration = 0.000175 Mol/L) 59

LIST OF FIGURES (Continued)

Figure Page

6.12 Effect of Al2(SO4)3 on System 59

6.13 Effect of Added Amounts of Al2(SO4)3 and Ca(OH)2 inpH Constant System 60

6.14 Effect of Added Amounts of Al2(SO4)3 and Ca(OH)2 in pH ConstantSystem 60

6.15 Effect of H3PO4 on System 62

6.16 Effect of H3PO4 on SystemNon CaF2-Fluoride Species in Dilute Region 62

6.17 Effect of Added H3PO4 and Ca(OH)2 in pH Constant System 63

6.18 Effect of Added H3PO4 and Ca(OH)2 in pH Constant System(Fluoride Limit = 0.0009 Mol/L) Non CaF2-Fluoride Speciesin Dilute Region 64

6.19 Effect of Na2CO3 on System 65

6.20 Effect of Na2CO3 on System 65

6.21 Experimental results of HF and Na2CO3 reacts with lime at room temperature,average initial concentration of HF = 190 ppm 66

6.22 The experimental data of HF react with differentratio of lime in a batch reactor 70

6.23 The experimental data of HF reacted with lime at different mixing rate,the initial concentration of HF = 200 ppm 71

6.24 Initial rate as a function of initial HF concentration(The ratio of HF : lime = 1:1.3) 73

6.25 The value of the integral in equation (6.10) versus time for mixedflow in CSTR with Tau = 100 min and k1 = 0.95 min-1 (HF:lime = 1:1) 76

6.26 The results of conversion prediction of a mixed flowover the range of the residence time in a CSTR, 0 < Tau < 400 min 77

6.27 The conversion of a mixedflow in a CSTR at different ratio of HF:lime, 0 < Tau < 50 min 77

LIST OF TABLES

Table Page

4.1 The summary of CES optimum analysis conditions for carbonate anions 205.1 The plan of experiments 226.1 Cation analysis by ICP (performed at the OSU Crop Science Laboratory) 29

6.2 Anion analysis performed by UV-visible spectrometry,Fluoride was measured by ISE 30

6.3 Material balance in the CAD system(Stream number refers to Figure 6.1) 31

6.4 Material balance in the CAD reaction tank(Stream name refers to Figure 6.2) 31

6.5 Specification in CAD reaction tank 33

6.6 Mass balance of components entering CAD reaction tame 36

6.7 Concentration of ions in CAD reaction tank 1/10/95 38

6.8 Concentration of ions in CAD reaction tank 1/29/9331 396.9 Component list in Chem Sage environment file 41

6.10 Summary of Harris mass balance analysis 54

6.11 The apparent kinetic initial rate constants for HF and lime reactionat different ratio of lime at 25° C 73

6.12 The plant's data for residence time in the CAD reaction tank 74

LIST OF APPENDIX FIGURES

Figure Page

1 Solubility Profile for NaC1-H2O-NaOH System at 25 °C 95

2 Solubility Profile for NaCI- CaSO4 -H20 System 95

3 Solubility Profile for H3PO4-CaSO4-H20 System 96

4 Solubility Profile for Na2SO4-CaSO4-H20 System 96

5 Solubility Profile for HNO3- CaSO4 -H20 System 97

6 Solubility Profile for Ca(OH)2- NaNO3 -H20 System 97

7 Solubility Profile for NaNO2- Ca(OH)2 -H20 System 98

8 Solubility Profile for NaF-HF-H20 System at 20 °C 98

9 Solubility Profile for NaOH- Na2SO4 -H20 System at 25 °C 99

10 Solubility Profile for NaCI- CaCO3 -H20 System at 25 °C 99

11 Solubility Profile for Na2SO4- CaCO3 -H20 System at 25 °C 100

12 Solubility Profile for HC1- CaF2 -H20 System at 20 °C 100

13 Effect of NaCIO2 on System (Mass Balance Concentration = 0.00 Mol/L) 101

14 Effect of NaC1O2 on System (Fluoride Limit = 0.0009 Mol/L) 101

15 Effect of Na2SO4 on System (Mass Balance Concentration = 0 Mol/L) 102

16 Effect of Na2SO4 on System (Fluoride Limit = 0.0009 Mol/L) 102

17 Effect of NaNO3 on System (Mass Balance Concentration = 0 Mol/L) 103

18 Effect of NaNO3 on System (Fluoride Limit = 0.0009 Mol/L) 103

19 Effect of H202 on System (Mass Balance Concentration = 0 MOIL) 104

20 Effect of H202 on System (Fluoride Limit = 0.0009 Mol/L 104

21 Effect of Ca(OH)2 on System 105

22 Effect of Ca(OH)2 on System (Fluoride Limit = 0.0009 Mol/L) 105

23 Effect of HNO3 on System (Mass Balance Concentration = 0.1630 Mol/L) 106

24 Effect of HNO3 on System 106

25 Effect of Added Amounts of HNO3 and Ca(OH)2 in pH Constant System 107

26 Effect of Added Amounts of HNO3 and Ca(OH)2 in pHConstant System 107

LIST OF APPENDIX FIGURES (Continued)

Figure Pane

27 Effect of HF on System(Mass Balance Concentration = 0.03573 Mol/L) 108

28 Effect of HF on System 108

29 Effect of Added Amounts of HF and Ca(OH)2 in pH Constant System 109

30 Effect of Added Amounts of HF and Ca(OH)2 in pH Constant System 109

31 Effect of H2CO3 on System (Mass Balance Concentration = 0.00 Mol/L) 110

32 Effect of H2CO3 on System 110

33 Effect of Added Amounts of H2CO3 and Ca(OH)2 in pH Constant System 111

34 Effect of H2CO3 and Ca(OH)2 in pH Constant System 111

35 Effect of HC1 on System (Mass Balance Concentration = 0 Mol/L) 112

36 Effect of HC1 on System 112

37 Effect of Added Amounts of HC1 and Ca(OH)2 in pH Constant System 113

38 Effect of Added Amounts of HC1 and Ca(OH)2 in pH Constant System 113

39 Effect of H2SO4 on System (Mass Balance Concentration = 0 Mol/L) 114

40 Effect of H2504 on System 114

41 Effect of Added Amounts of H2SO4 and Ca(OH)2 in pH Constant System 115

42 Effect of Added Amounts of H2SO4 and Ca(OH)2 in pH Constant System 115

LIST OF APPENDIX TABLES

Table Page

A.1 Pitzer ion coefficients from Haung 88

A.2 Pitzer ion coefficients used for Pitzcomp.exe datafile in Chemsage software 88

C.1 The result of experiments between HF, lime, and different concentrationof added Na2CO3 in a 1 L batch reactor at 25° C 116

D.1 Experimental data between HF and lime at different ratios in thewell-mixed batch reactor at 25°C 117

Chemistry and Chemical Process Studies of Fluoride Removalin a Silicon Wafer Manufacturing Wastewater Treatment Plant

1. INTRODUCTION

Silicon wafer manufacturing facilities use hydrofluoric acid (HF) in processing of

silicon ingots and wafers. The HF is discharged to a wastewater treatment facility where

it is neutralized and precipitated with lime. Fluoride is a priority pollutant regulated under

the Clean Water Act administered by Environmental Protection Agency (EPA). The effect

of excessive concentrations in water involves dental health and bone disease. The controlof fluoride in wastewater is strictly maintained. EPA has established an average dailyfluoride effluent limit for discharges resulting from the manufacture of semiconductors to

be 17.4 milligrams per liter. (48 Fed. Reg. 15394, April 1, 1983). Industrial waste

treatment facilities of wafer manufacturing can have difficulties in meeting this discharge

limit in the treated wastewater. Because of these problems, an investigation of the basic

chemistry and chemical processes of HF removed from wastewater is required.

The waste treatment facilities of wafer manufacturing involve essentially thefollowing steps: neutralization of wastewater with hydrated lime at a pH 11, which results

in the formation of relatively insoluble calcium fluoride; separation of the insoluble

product of the reaction in a classifier; and filtration of the sludge. The waste treatmentprocess is illustrated in more detail in a later section. The system is composed of areaction tank, lamella clarifiers, filters, and a filter press.

The waste acid stream from wafer manufacturing, contains ions including Pat ,

S042-, CO3-, C104, a other than their reaction products from processing operations, F.

2

The acid waste stream is pumped to the reaction tank in wastewater treatment plant. The

waste stream is neutralized and treated for fluoride removal by adding a lime slurry. The

amount of lime slurry to be added is controlled by maintaining a specific alkalinity of pH in

the reaction tank. The pH is measured by the pH analyzer which sends a signal to the pH

controller. The pH controller maintains the required pH by regulating the lime slurry flow

control valve. The addition of lime neutralizes the acid component on the waste stream

and supplies the required calcium ion necessary to cause the precipitation of fluoride as

CaF2(s). The neutralized stream with CaF2 precipitate, is blended with the alum in the

flocculation tank to increase the CaF2 particulate sizes. The blended stream is fed to alamella clarifier. The clarifier separates a CaF2 solid from the neutralized stream. The

CaF2 solids are pumped to the filter press. The filter press separates the CaF2 solids from

the liquid. The solids are disposed of into a dumpster and the filtered liquid is recycled

back to the reaction tank. The neutralized stream from the clarifier is fed to the filters.

The filters remove any solids that are not removed in the clarifier. The clear effluent from

filters is neutralized to a pH suitable for discharge and released.

From a process perspective, there are several possible reasons why the fluoride

concentration may not being reduced to an acceptable level by lime treatment. These

include first, mass balance: the fluctuating input of fluoride concentration to the waste

treatment system was suspected to affect the exiting fluoride concentration. The

insufficient amount of lime addition in reaction tank could also cause the high fluoride

concentration. Second, equilibrium effects: the interference of other ions could change the

chemical activity of the fluoride ions, or by pH or ionic strength effects. Lastly, kinetic

effects: there are two possibilities that could result in slow reactions of fluoride with lime.

They include an inherently slow reaction rate, and formation of a non-porous productlayer that block access of fluoride ion to the remaining lime.

The study reported here was conducted to determine which of these would result

in unacceptable fluoride concentrations in the water leaving the waste treatment plant, and

under what process conditions. The tasks include the study of mass balance, equilibrium,

3

and kinetic effects. This study was done in cooperation with the Wacker SiltronicCorporation plant in Portland, OR.

4

2. OBJECTIVE

The objective of this thesis was to provide the investigation of the chemical and

process limitations to fluoride removal from the plant's waste water treatment, and to

propose ways to improve fluoride removal. The investigation divided in to three tasks.

Task 1. Evaluation of Solubility Data and Mass Balances

a. Evaluate solubility data and fluoride removal processes from both the literature and

Wacker Siltronic studies.

b. Complete a mass balance for the waste treatment facility, including the range of

variation of total flows and species flows, including a breakdown by input source.

Obtain data by measurement of process streams as necessary.

c. Determine whether underlining could account for unacceptable fluoride ion removal.

Task 2. Equilibrium Investigation

a. Create a fluoride ion solubility "map" in terms of all of the other chemical species

present, using the CHEMSAGE equilibrium package.

b. Determine which other species can influence the solubility of fluoride and to what

extent they can increase or decrease its solubility.

c. Verify the results with published fluoride solubility data and data from Wacker

Siltronic.

d. Determine whether equilibrium constraints could be responsible for unacceptable

fluoride ion removal.

S

Task 3. Kinetic Investigation

a. Evaluation of the rate of reaction in batch reactor for typical reaction tank conditions.

b. Measurement of the residence time distribution in the CAD reaction tank to either

confirm that it acts as a mixed flow reactor or to obtain a residence time distribution

under typical flow conditions.

c. Calculate the expected conversion of fluoride ion to CaF2 based on the measured batch

kinetics and residence time distribution for typical entering fluoride concentrations,

fluoride/lime ratios, and residence times.

6

3. LITERATURE REVIEW

3.1 MASS BALANCE

3.1.1 Chemistry of hydrofluoric acid

Fluorides are discharged in significant quantities in effluents from glass

manufacturers, electroplating operations, aluminum and steel producers and processors,

fertilizer manufacturers and wafer manufacturing. Average fluoride concentration for

aluminum reduction plants are reported as 107-145 mg/1 in wastewater streams. For glass

manufacturing, the fluoride concentration was reported ranging from 1,000 to 3,000 mg/l.

Wafer manufacturing uses hydrofluoric acid (HF) in the etching process, the effluentconcentration varied from 100-1000 mg F/1.

Hydrofluoric acid is very soluble in water. The dissociation of hydrofluoric acid in

water is expressed by equation (3.1); the value of the equilibrium constant indicates that at

pH values higher than 3.17, most of the total fluoride is present as fluoride ion, F. The

combination of these two equilibria is shown conceptually in Figure 3.1 in which the total

soluble fluoride concentration depends strongly on pH. The total soluble fluorideconcentration is defined in equation (3.2). The curve of the plot shows that the solubility

of fluoride rises with pH.4

HF + H20 = + H30+ HF(25 °C) = 10-317 (3.1)

FT = [HF] + [F] (3.2)

7

-2

-4

O

PH

4 6 8 10 12 14

Figure 3.1 Gas/liquid equilibrium of fluoride in water 4 log FT (mg/1) vs. pH.

3.1.2 Chemistry of precipitation with lime

A variety of treatment methods are available for fluoride-bearing waste streams.

Current treatment methods can be divided into two categories; precipitation methods and

adsorption methods. Precipitation methods involve addition of treatment chemicals and

formation of fluoride precipitates or coprecipitation of fluoride on a resulting precipitate.

Removal is accomplished by solids separation of the precipitate. Chemicals employed

include calcium hydroxide (lime), magnesium compounds, and aluminum sulfate (alum).

Adsorption methods involve the passage of the wastewater through a contact bed, with

fluoride being removed by general or specific ion exchange or chemical reaction with the

solid bed matrix. Since the basic mechanism of a contact process is one of ion exchange

or surface reaction, these methods are usually appropriate only for low level fluoride

wastes or as a final polishing step.

8

Using hydrated lime to precipitate calcium fluoride is a widely accepted method for

defluoridationf equation (3.3)] The addition of hydrated lime, Ca(OH)2, leads to a rise in

the pH due to dissolution and dissociation described in equation (3.4). The resulting

calcium ions, Ca24", are then available to precipitate with fluoride [equation (3.5)] and with

other ions in solution. The possibility of precipitation of fluoride ions in the presence of

sulfate [equation (3.6)] and carbonate [equation (3.7)] was investigated by Lawler andWilliams.4

2HF(I) + Ca(OH)2(s) CaF2(s) + 2H20 (I) (3.3)

Ca(OH)2 = + 20H- Keg = [Ca2+] [01-1-] = 10-53 (3.4)

CaF20) = Ca2+ + 2F Keg = [Ca2+] [F]2 = 10-'115 (3.5)

Ca Saks) = Ca2+ + S042- Keq = [Ca2+] [S042-] = 10-4'6 (3.6)

CaCO3(s) = Ca2+ + CO3 2+ j [Call [CO3 = 10-836 (3.7)

According to Standard Methods by APHA1, the results indicated that nosignificant precipitation of calcium sulfate should occur. Sulfate is added to the system in

the make-up water which is put to replenish water lost by evaporation and removal of

sludge. Only a negligible amount of sulfate precipitation occurs at any time in comparison

to the fluoride precipitation.

The possibility of calcium carbonate precipitation was investigated in detail byParker and Fong5 and Lawler and Williams4. The experimental results of Parker and

Fong5 are shown in Figure 3.2. The plot reflected a competition for the available calcium

between fluoride and carbonate. Parker and Fong's results were cited by Paulson6 in

recommending that a pH near 12 be used for fluoride precipitation with lime in order to

avoid the poor removal at lower pH values.

9

4s

.4 --

C

23

1.1J ZS

z8171G 20 -

13 --

S-

06 7 9

TREATMENT pH12

Figure 3.2 Experimental results for residual soluble fluoride at initial concentration87 mg/liter after lime precipitations.

However, Lawler and Williams' hypothesized that Parker and Fong's experimental

results were influenced by contamination of distilled water with carbonate, which could

easily occur by dissolution of CO2 from air. Lawler and William performed experiments

similar to those Parker and Fong. Their results, shown in Figure 3.3, confirmed the

hypothesis that the hump in the residual fluoride curve was caused by competition with

carbonate with maxima near pH 7 and pH 11.

10

Jc

li

60

50

40

10

04

initioiFr

(mq C11 Pte=O 106 Armeso 106 0A 240 0

1

6 8 10 12 14

pH

Figure 3.3 Experimental results: residual fluoride after precipitation with lime4.

Most wastewater treatment systems would more closely approximate closed

systems, i.e. negligible absorption of gaseous CO2, small amount of surface area in contact

with the atmosphere per unit volume per solution in deep basin, at some fixed totalcarbonate concentration. If that fixed level of carbonate is nearly zero, the residualfluoride achieved should be essentially independent of pH in the range 7.5<pH<10.5. At

measurable levels of total carbonate, the residual fluoride is highly dependent on pH, with

maxima near pH 7 and pH 11.

3.1.3 Lime precipitation with aluminum sulfate

Aluminum sulfate (alum) dissociates and hydrolyzes in water to form a hydrated

aluminum species, Al (OH)3. Previous studies have found that alum can further reduce

fluoride concentrations in classifier effluent from lime precipitation processes bycoagulating fine suspended particles. Culp and Stoltenberg3 concluded that further

11

fluoride removal depends on the efficiency of alum flocculation, which is related to pH.They discussed that alum has the optimum pH range between 6.5 and 7.5. Zabban andHelwick8 investigated the alum requirements at different pH and verified this optimumrange as shown in figure 3.4. The plot illustrates the result of alum addition indefluoridation of sodium fluoride containing 15 mg/1 at pH 7,8, and 9. The residual ofdissolved fluoride concentration were measured. Treatment of wastewater at a pH of 8requires approximately twice the quantity of alum as a pH of 7, and a pH of 9 requiresabout 3.5 times the amount of alum. Although aluminum ion is not responsible for the

precipitation of fluoride, some hydroxylated aluminum complex formed in the precipitation

of alum combines with and precipitates the fluoride ion or removes it by adsorption.'

IS ALUM/ lb FLUORIDE

Figure 3.4 Effect of solution pH on alum addition required to reach target fluorideconcentration. Initial sodium fluoride solution 15 mg/1 F7.

Besides being pH-dependent, the alum requirement also is a function of initial

fluoride concentration, desired residual fluoride content, settleability of the precipitate,

12

sludge volume and sludge dewatering characteristics. Even though greater fluoridequantities require larger amounts of alum, the alum requirements to reach 4.8 mg/1 residual

F remain constant for solutions containing initially 10-85 mg/1 F as shown in figure 3.5.

SO

SO

40

30

20

IO

- 0

O0 IS 30 GS GO

INITIAL FLUORIDE CONCENTRATION. nt1/1 F

$0

Figure 3.5 Alum requirement as a function of fluoride concentrationResidual fluoride 4.8 mg/1 F, at pH 7.

However, as the fluoride concentration is decreased below 4 mg/1, the alumrequirement to further decrease F concentration drastically increases as shown in Figure3.6.7

110

40

30

20

10 20 30 40

1$ ALUM/11, FLUORIDE

13

Figure 3.6 Alum requirement to decrease fluoride in concentration for a sodium fluoridesolution containing initially 78 mg/1, pH 7.0 7.

3.1.4 Aluminum effects in acidic waters

Further studies of alum effects in acidic conditions was done. Acidic deposition

has an important effect on the transport and speciation of soluble aluminum. The report

by Plankey and Patterson2 has indicated that the aqueous aluminum seems to be strongly

dependent on aluminum speciation and the presence of complexing ligands such asfluoride. They reported the study of the complex formation kinetics of A1F2+ in the

environmentally significant pH range 2.9-4.9. The important paths are as follows.

Al3+ + F E->AT2+

Al0H2+ + F+ H+ E->2+

(3.8)

(3.9)

Al3+ + HF AT2+ H+

+ HF H A1F2+ + H2O

3.2 REACTION KINETICS

The rate of a reaction may be expressed in terms of the concentration of any

reactant or of any product of the reaction. It may be expressed as the rate of decrease of

the concentration of a reactant, or as the rate of increase of a product of the reaction.According to equation (3.3), the reaction of HF and lime yields:

d[HF] d[Ca(OH)2] cl[CaF2]rA dt dt dt (3.12)

The reaction rate is determined by experimental observation. Although thefunctional dependence may be postulated from theory, experiments are necessary toconfirm the proposed form. One of the most common general forms of this dependence is

the product of concentrations of the individual reacting species, each of which is raised toa power, e.g.,

r A = kCACe (3.13)

The exponents of the concentrations in Equation (3.13) is the order of reaction.

The order of the reaction refers to the powers to which the concentrations are raised in the

kinetic rate law. The overall order of the reaction, n, is

15

n = a + p (3.14)

No kinetic data for the reaction of hydrofluoric acid with lime was found in the the

literature. However, batch reactor data can be used to determine rate law parameter for

homogeneous reactions. This determination is usually achieved by measuringconcentration as a function of time and then using either a differential or integral method

of data analysis to determine the reaction order, n and specific reaction rate, k.

The differential method is applicable when the rate is essentially a function of the

concentration of only one reactant: e.g., the decomposition reaction. However, by

utilizing the method of excess, the differential method can determine the rate law.

In order to find the rate order by the integral method, first we assume the reaction

order and integrate the differential equation used to model the batch system. If thereaction order is correct, plot of the integrated concentration as a function of time will belinear with slope k. In the following section a number of rate expressions are integrated

and discussed by which they are used to obtain the rate constant of HF and lime reaction.

For the second order reaction, first order with respect to each reactant A and B

with stoichiometric equation 2A + B = C + 21)13,

dA= kAB

dt

where: k = reaction rate constantA = concentration of HF [mo111]

B = concentration of Ca(OH)2 [mo111]

integration yields:

(3.15)

14 A° j+(2B0

2

k ) A (2B01

k ) ln

( Ao A \Bo

2 2Bo

16

= kt (3.16)

where Ao = initial concentration of HF [mol]

Bo = initial concentration of Ca(OH)2 [mol]

x = Ao A = numbers of moles of reacted HF [mol]

t = time (min)

For the third-order reaction, first order with respect to reactant B and second

order with respect to reactant A with stoichiometric equation 2A + B = C + 2D,

dA= kA

2Bdt

Integration yields:

where

2 7 1

A

1 2In

Bo A= kt(2B0 A0) Ao )l (2B0 AO- A0B

B = BoAo

+ A2 2

(3.17)

(3.18)

(3.19)

Another third-order reaction, first order with respect to A and second order with

respect to reactant B with the same stoichiometric equation as the earlier case 2A + B = C

+ 2D,

dA= kAB 2

dt (3.20)

integration yields:

2 1 1 kl3= kt

(2A0 Bo)

2(B Bo (2A0 Bo)

1

BoA

17

(3.21)

18

4. ANALYTICAL METHODS

4.1 pH AND ION-SELECTIVE ELECTRODE MEASUREMENTS

In this study, a large number of samples were collected to obtain the fluoride

concentrations. The ORION benchtop 720A pH/Ion-Selective Electrode meter was used

to measure the pH and fluoride concentration in this study.

The pH meter combines with a glass electrode and a reference cell. The glasselectrode consists of a tube with a bulbous bottom which is filled with a solution having a

known hydrogen ion concentration. The bulb is immersed in to a solution containing

hydrogen ions, the concentration gradient developed across the glass membrane of thebulb. The hydrogen-measuring glass electrode is connected to a reference cell with astandard potentials. The difference between the potentials of the unknown solution and

the electrode solution can be measured and the hydrogen concentration can be calculated.

The pH meter is calibrated against standards and adjusted for temperature effect each use.

The electrode slope is calculated and then meter will then proceed the measurement.33

An Ion-Selective Electrode Systems (ISE) meter is used to measure the fluoride

concentration in the sample solutions. The concept of measuring fluoride ion with the ISE

meter is the same as the concept used in the pH meter. Two electrodes connected through

a meter, a difference in potential between a sample and an electrode solutions is known,

then the concentration of the sample solution will be calculated.

19

4.2 CAPILLARY ELECTROPHORESIS ANALYSIS OF CARBONATE

A Dionex model CES-1 Capillary Electrophoresis System (CES) was used to

determine the amount of carbonate ion in the CAD samples. Capillary Ion Analysis (CIA)

is a type of capillary electrophoresis (CE) optimized for the rapid analysis of lowmolecular weight anions and cations. It separates ions according to their mobility in

electrolytic solutions. A sample is injected into the capillary and ions in the sample are

forced to move under high voltage which causes an electroosmotic flow. The separation

process takes place along the capillary by the fact that each ion has a different ionic charge

and molecular weight, resulting in different mobilities under the electroosmotic flow.32

The mobility of the UV absorbing co-ion is important. The rule states that the co-

ion must have a mobility similar to that of the analytes. The closer the match the better

the peak symmetry. Carbonate ion is a low mobility compound, therefore, a low-mobility

co-ion is needed for good peak symmetry. The peak detected in the carbonate analysis in

this study was in poor shape because a higher-mobility buffer was used in this analysis.

The voltage range used in analysis for anion species is around 15 kV to 30 kV.

A 30 kV voltage was used for carbonate analysis in this study. The length of capillary and

diameter affect the separation efficiency for each ion species. A longer capillary yields a

more efficient separation, but results in a longer analysis time. In this study, the 50 cm

length and 50 pm diameter of capillary was the most practical and used for carbonate

analysis.

The wavelength of UV light is important in the analysis. Each ion species absorb

UV light at different wavelengths. The wavelength of 210 nanometer (nm) was used even

though the manual recommended the wavelength at 250 nm for anion analysis. The peaks

detection at 210 nm yielded greater signal than at 250 nm. The summary of CESoperating condition is shown in Table 4.1.

20

Table 4.1 The summary of CES optimum analysis conditions for carbonate anions

Capillary ID, 1.tm 50

Capillary length to thedetector, cm

50

Capillary length, cm 60

Separation Voltage, kV 30

Detection wavelength, nm 210

Injection method Gravity(100mm / 30sec)

Buffer IonPhorAnion PMA electrolyte

Dionex Part 44138

4.3 SPECTROPHOTOMETRIC ANALYSIS OF ANION SPECIES

In mass balance study, the concentrations of all the components possibly found in

the samples were measured. The concentrations of PO4, S042-, NO3, C104 , and

S2032-, in the CAD waste water samples were measured by using HACH model DR/3000

UV- spectrophotometer.

The DR/3000 UV-spectrophotometer is a microprocessor-operated laboratoryinstrument. Preprogrammed with calibrations for most water management testrequirements, it provides direct, digital readouts in absorbance, percent transmittance or

concentration. When the desired program code is entered, prompting lights direct the

operator through the test by giving the appropriate wavelength and indicating control key

sequences. The DR/3000, a single-beam spectrophotometer, uses a double-pass grating

monochromator capable of wavelengths from 325 to 1000 nm. The assemblies are made

up of the pour-thru cell, the base assembly, the funnel with funnel holder and

21

interconnecting tubing. The ambient temperature must be within the range of the 10° to40°C (50° to 104°F) for proper operation of the instrument.34

22

5. EXPERIMENTAL PROCEDURE

5.1 EXPERIMENTAL PLAN FOR THE KINETIC STUDY

Two types of experiments were required in this study. One was to measure theequilibrium between HF, lime, and Na2CO3 to confirm the results of equilibrium study.

The other was to measure the rate of reaction between HF and lime. Table 5.1 shows the

summary of the experimental plan. Experiments la-le were conducted to find the rate of

reaction and experiments 2a-2d were for equilibrium study.

Table 5.1 The plan of experiments

Experiment Mol Ratio

HF Lime Na2CO3

la 1 0.5

lb 1 1

lc 1 1.3

id 1 2

le 1 10

2a 1 2 1

2b 1 2 0.1

2c 1 2 0.2

2d 1 2 0.5

23

5.2 EQUIPMENTS

All the equipments used in this experiment are listed as followed,

pH/ISE meter The ORION pH/ISE meter was used to measure fluoride

concentrations in the experiments.

Syringe with millipores membrane filters (45 p.) The filters were used to

remove the lime and CaF2 precipitate from the collected samples. They helped to stop the

further reaction between the lime and the HF after the sample was collected.

Beaker equipped with stirrer The HF and lime reacted in a well-mixed 1-L

beaker. A motor-driven stirrer was used to maintain the effective mixing within the

beaker.

Chronometer The concentration of fluoride was measured at a one-minute

interval of time. The reaction was very rapid, therefore, measurements at the first ten

minutes were the most critical data. In order to collect the sample at the correct time, the

buzzing timer was used and recommended in the kinetic study.

5.3 EXPERIMENTAL PROCEDURE

Numerous data was necessary in order to obtain accurate rate constants.

Reproducible experiments were proceeded after each experimentalrun to confirm the

24

reasonable results in each run. The schematic of the experimental procedure is shown in

Figure 5.1.

Lime

t = 0

t = ti

Collect thesample

filter lime &solid out

Measure F-

Figure 5.1 Schematic of experimental procedure

5.3.1 Preparing the solutions and running the experiment

1) Weigh the lime (CaO) to the ratio amount required to react with HF. Dissolve the

lime with DI water in a beaker. Dilute the concentrated HF solution to the desired

initial concentration. Take an extra precautions during HF solution preparation as the

concentrated HF is extremely hazardous.

25

2) Pour the HF solution into the baffled-beaker. Measure the initial concentration of HF

solution before each experimental run. Turn on the motor to begin the mixing in the

solution, set the timer to a one-minute interval, then pour the dissolved lime solution

into the beaker.

3) At one minute, set the timer to another minute and take a sample solution

from the beaker. Filter the sample solution through a millipore membrane filter to

remove the CaF2 precipitate. After filtering the sample solution, dilute it with DI

water and take a F measurement in the diluted solution with the ORION model 720A

ISE meter.

4) After making measurement, be ready to take another sample once the timer buzzes.

Continue to collect the samples until the reaction between lime and HF stops, which is

the F concentration remains constant.

5.3.2 Measuring fluoride

1) Add TSAB buffer into the sample solution. Perform a two point calibration before

measuring concentration.

2) Select the concentration mode on the display of meter.

3) Press cal key to begin calibration. When "READY" is displayed, place the electrode

into the first standard solution. Adjust the concentration, and press enter. Rinseelectrode with DI water, and continue with the other standard solution.

4) After the calibration, the electrode slope is calculated and the meter will proceed to the

measure mode. Place electrode into a sample solution, record the reading.

26

6 RESULTS AND DISCUSSION

6.1 MASS BALANCE ANALYSIS

6.1.1 Wastewater treatment process

The wafer etching process at Wacker Siltronic Corp. discharges the acid waste

stream containing HF and other ions to a forwarding sump. The Concentrated Acid Drain

(CAD) system collects the wastes streams in a forwarding sump and transfers to the CAD

waste water treatment plant. The CAD system is composed of a reaction tank, aflocculation tank, two parallel lamella clarifiers (settlers), three parallel filters and a filter

press. The fluoride waste stream is pumped in to the CAD reaction tank for fluorideremoval by adding a lime slurry. The lime neutralizes the acid component in the waste

stream and supplies the required calcium to react with fluoride. The CaF2 and neutralized

stream are blended with the alum in the flocculation tank to increase the CaF2 solid sizes.

The stream then is fed into lamella settlers. The solid settles and is removed from the

neutralized stream as sludge. The sludge is sent to a filter press and dewatered. The

solids are removed by a filter press and stored in a container before disposal in the landfill.

The neutralized stream from the settlers is fed to the sand filter. The sand filters removefurther solids from the neutralized stream. The clear effluent from filters then is feddirectly to the Weak Acid Drain (WAD) system for further treatment. The overview of

the waste water treatment is illustrated in Figure 6.1.

ForwardingI Sump Flow

2 Lime Slurry

To LandfillSolid Capturl

7

Recycle stream 11

CADReaction Tank

Alum

FlocculationTank

---4P1406

CaF2 --,/solids

Rejected water to 9WAD system

FilterPress

aSolid CaF2

6

/illlLamellaSettler

8NeutraltzedStream

Sand Filter t"

Figure 6.1 Schematic flow diagram of CAD wastewater treatment plant

10

27

28

Figure 6.2 shows the material balance in the CAD reaction tank. Forwarding sump

collects the acid waste streams from etching process. The effluent from the plant'sscrubber is fed to the CAD reaction tank and neutralizes the acid components in the

reaction tank. The stream from the scrubber contains PO4 , C104 , a, S042-, and NO3-.The effect of anions in scrubber effluent to the fluoride will be discussed in a later section.

Alum and coagulant are added in to the reaction tank to increase the CaF2 particulate sizes

and allow it to settle.

Forwarding SumD Effluent

Srubber Effluent

Lime Slur&

Alum

Polvmer(Coagulant)

CaF2Solids

.....

ReactiOn Tank

Figure 6.2 Schematic mass balance in CAD reaction tank.

29

6.1.2 Evaluation of mass balance

Material balances for the waste treatment system were calculated based on

measurement made in this study. The concentrations of all the ions possibly found and

measured in the samples, including fluoride, are shown in Tables 6.1 and 6.2. The samples

were taken from forwarding sump, scrubber, and CAD reaction tank.

Table 6.1 Cation analysis by ICP (performed at the OSU Crop Science Laboratory).

Cations and

Elements

Forwarding Sump

(ppm)

Reaction Tank

(ppm)

Scrubber

(ppm)

Al 0.24 54 8

Ca2+ LD1 403 <25

1(4 LD 9 26mg2+ 0.2 2.5 0.52

Na+ 166 119 4847

Mn LD <0.1 <0.1

Fe LD <0.1 <0.1

Ni LD <0.1 <0.1

Cu LD <0.1 <0.1

Zn <0.1 <0.01 <0.01

P LD <10 <10

S 17 32 460

Si 6 0 0

B LD <0.1 0.1

1 LD = Below Detection Limit

30

Table 6.2 Anion analysis performed by UV-visible spectrometry,Fluoride was measured by ISE.

Anion Forwarding sump

(PPm)

Reaction Tank

(PM)

Scrubber

(PM)S2032- 0.075 0 3

S042- 59 38 2000

NO3- 726 1584 1848

P043- 34 127 26

F 125 10 175

0102 NT2 1500 40

Cl- NT 8 0.5

Several assumptions were made for the mass balance.

Steady state flow;

90% of solid removal in sand filters;

Approximately 2% of solid concentration in backwash filter;

Approximately 95% of solid capture in the filter press;

90% of solid concentration in the filter press;

The ratio of lime and makeup water add in the mixing tank = 50 lb/100 gal;

Conversion of HF approximately 96% (based on the fluoride concentration of the

effluent measured on 1/13/95)

Average initial HF concentration approximately = 242 ppm.

Table 6.3 shows the mass balance in CAD waste water treatment plant (Figure

6.1). The calculation in the waste treatment only based on HF, lime, CaF2, and water.

2 NT = Not Tested

31

Table 6.3 Material balance in the CAD system.(Stream number refers to Figure 6.1).

Streani# 1 2 3 4 5 6 7 8 9 10 11

lbmol/h

HF 0.008 0 0.003 0.003 2.74e-4 0 0.00026 0.00026 0.00001 0.0003Ca(OH)2 0 0.391 0.387 - 0.387 0 0 0.383 0.383 0 0CaF2 0 0 0.0037 - 0.0037 0.0036 0.0035 0.00011 1.05e-5 9.45e-5 0.00021120 33.5 27.8 61.3 - 61.3 13.5 0.02 47.8 47.8 0.0003 13.5

TSS - - - 350.5 - - 9.64 - -

F (pprn) 242 - 10 - - - - 9.8 10.8 -

Table 6.4 Material balance in the CAD reaction tank.(Stream name refers to Figure 6.2)

Components

(lbmol/hr)

Forwarding

Sump

Lime

Slurry

Alum Polymer Scrubber RXN Tank

Influent

RXN Tank

Effluent

HF 0.0077 0 0 0 0 0.0077 0.003HNO3 0.4611 0 0 0 0 0.4611 0

Surfactant 0 0 0 0 0 0 0

CaF2 0 0 0 0 0 0 0.0037

Ca(NO3)2 0 0 0 0 0 0 0

Al2(SO4)3 0 0 0.03 0 0 0.03 0

Al(OH)3(s) 0 0 0.01 0 0 0.01 0

CaS040) 0 0 0 0 0 0 0

Ca(OH)2 0 0.391 0 0 0 0.391 0.387Polymer 0 0 0 0.0035 0 0.0035 0.0035

H2O 33.5 27.8 6.486 18.53 37.1 123.416 123.424

H2SO4 0 0 0 0 0.817 0.817 0Na(C102) 0 0 0 0 0.0177 0.0177 0.0177

Flowrate (GPM) 72.4 60 14 40 80 274 274

pH 3.68 11.5 11.5

32

Table 6.4 shows the mass balance in CAD reaction tank (Figure 6.2) based on the

measurement of cation and anions in the reaction tank.

The fluctuating input of fluoride to the waste treatment system affected the

fluctuating fluoride concentration exiting the lime treatment process. The effect of flow

variation in fluoride removal was considered in this study. Figure 6.3 illustrates data for

CAD reaction tank pH, forwarding sump and reaction tank flow rate (hourly basis) over

12 hour on May 13, 1995. The fluctuation of the waste flowrate and fluorideconcentrations depends on the schedule of discharge of the waste from the processing

plant. Samples were collected and were tested for fluoride concentrations at different

time shown in Figure 6.4.

0< 160

140ioC

EI- _1cc 0 1200

Eas 0. a0)

C (130 "C 3 CC' O 0 600. =O mc.)

< aocu

ea>

E 20>

100

80

0E

:-.Or-mil -

AA, - At'lla :1'. I I=.r.

lir,.

....4.Ai

A A

tAA_A_AA__A A A A___,A_AA_A___ A_ _4

12 00 1:00 2:00 3:00 4:00 5:00 6:00 7:00 8:00 9:00 10:00 11:00 12 000`7-

AM AM AM AM AM AM AM AM AM AM AM AM PM

Time (hr) A Lime Vlv % openCAD Flowin

FlowoutLamella pl-r10

CAD Tank 01'10

Figure 6.3 Wacker Siltronic plant's record of reaction tanklime valve (% open), flow from forwarding sump, flow out of reaction tank,

pH measured in the CAD reaction tank and in lamella settlers

33

400.0350.0300.0250.0200.0150.0100.050.0

0.012:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM

Time (hr)

Figure 6.4 Fluoride concentration variation over 8 hours.(measurement of samples from forwarding sump on January 15, 1995)

The lime was added to the reactor with the ratio of 50 lb/100 gal of makeup water.

The average lime feed rate to the reactor was approximately 33 lb/min. The average input

flowrate from forwarding sump was 73 gal/min, the average fluoride concentration is 242

ppm. The summary of the plant's average flowrate and pH are shown in Table 6.5.According to the plant's record, 500 lb of lime was added to lime slurry mixing tank every

15 minutes. The amount of lime added was about 50 times of the stoichiometric amount

required for removal of HF.

Table 6.5 Specification in CAD reaction tank

Range of Fluoride Concentration (ppm) 200-1000

Lime Feed Rate (lb/min) 30

Lime/Make up Water (lb/gal) 50 /100

Average Forwarding sump Flowrate (gpm) 70-1000

pH in CAD reaction tank 9-11

34

6.1.3 AlF2' formation analysis

According to the Plankey and Patterson2 study, A1F2+ formation depends strongly

on a aluminum speciation (A13+) and the presence of F at the pH range 2.9-4.9.Therefore, A1F2+ complex formation does not occur at alkaline pH's. Data from Wacker

Siltronic's waste treatment plant shown in Figure 6.3 indicates that the pH in reaction tank

is between 8-11 and the pH in the lamella settlers (effluent of reaction tank) fluctuates but

stays within ± 0.5 pH units of the setpoint value of 11.5. This indicates that the complex

A1F2+ will not form in the CAD reaction tank and lamella settlers.

6.1.4 Conclusion

The material balance shows that the pH is high enough and sufficient calcium is

present to remove fluoride to acceptable levels in the effluent from the waste treatment

plant if equilibrium is reached or approached closely and if the precipitated CaF2 isremoved efficiently from the waste water.

35

6.2 EQUILIBRIUM EFFECTS

After the material balance study showed that underliming was not the cause of ahigh fluoride concentrations in the treated waste water, the next step in the study was an

investigation of the equilibrium effects of various chemical species on fluoride ions.

Chem Sage 3.0, a phase equilibrium computational software package, was used to

determine what components, if any, could cause the discharge problem by equilibrium

effects. Chemsage is the free energy minimization by using Pitzer's model for ion activity

coefficients. The assumption of a database of Pitzer parameter included in Appendix A.

A preliminary study was made in which calculated solubilities were compared with

published solubility data for a variety of chemical systems. The results are concluded in

section 6.2.2. They showed that the calculated solubilities were in reasonable agreement

with the available experimental data, within a factor of three for most systems, as well as a

general match of each system's behavioral trend. The poorest agreement was between

experimental and predicted solubilities in binary and ternary solutions when fluoride was

one of the ions. This may be a limiting factor in the analysis presented here.

A study was performed on the effect of increased concentrations of variouscompounds on the system found in the Wacker Siltronic process streams.th These were

based on mass balance calculated for a new waste treatment plant which has since been

built at the Wacker Siltronic Corp. plant in Portland, OR. The mass balance data areshown in Table 6.6. Most of the compounds found in ionic form in the CAD reaction tank

samples were tested in this study and some data provided by IDC.31 The mass balance

data in Table 6.6 yields a full conversion (99.9%) of HF to CaF2 at equilibrium.

The equilibrium study was broken into four sections: defining the contents of the

CAD reaction tank in Chem Sage environment file, comparison of calculated solubility

36

results to experimental data, analysis of two possible mechanisms as causes of the fluoride

discharge problems, and analysis of the effect of various components on the CAD reactiontank system.

6.2.1 Defining the CAD reaction tank in the Chem Sage environment file

Numerous samples were taken from the system, but the analysis of these samples

yielded only concentrations of dissolved species in solution. They did not account for any

suspended solids which may have been in the samples. The best mass balance found was

from a study by the Harris Group, dated May 1, 199510 [Table 6.6]. The mass balance,

dated May 1, 1995, was intended for the new waste treatment facility to be installed in

1996, but it is assumed that the relative concentrations are similar to those in the current

reaction tank. This mass balance did not take into account any species entering from thescrubber.

Table 6.6 Mass balance of components entering CAD reaction tare

Component Incoming (mg/L) Harris Groupi° OutputValue (mg/L)

Chem Sage Output

Value (mg/L)

HF 715 22.4 0.87 FionHNO3 10269 0 0

CaF2 (s) 577 1928 1968

Ca(NO3)2 545 13916 0.173

Al2(SO4)3 60 12 0

Al(OH)3 (s) 9.3 31.1 0.0267

37

Table 6.6 (continued)

Component Incoming (mg/L) Harris Group" OutputValue (mg/L)

Chem Sage Output

Value (mg/L)

CaSO4 (s) 24 81 33.8

Ca(OH)2 9464 2114 1390

H2O 1000000 1003600 1003600

NO3- ---- ---- 10516

SO4 ---- ---- 12

Ca" ---- 3761

01.1" ---- ---- 308

Total 1021663 1021665 1021592

Also included in the this mass balance were inerts, a surfactant, and a polymeric

coagulant. These are unable to be accurately simulated using ChemSage and were notincluded in the equilibrium calculations. They are not expected to have a significant effect

on the ions in solution. In addition, a specific stream containing inerts and water was

eliminated from the mass balance. None of the compounds from specific stream were

found in aqueous form in this mass balance. There was a discrepancy found in the mass

balance above in the ChemSage output. This was due to the fact that many components

had hydrated forms, other miscellaneous components were found in the output which did

not fit any of the compounds above (as with the case of the aluminum components), and

that there was a margin of error caused by rounding. Overall, ChemSage predicted a99.9% conversion of HF at equilibrium, and a pH of 11.83 based on the inputs in theTable 6.6. The Harris Group mass balance out puts were based on a 97% conversion ofHF.

To obtain an estimate of the composition of components in the CAD reaction tank,

two sets of analyses from process stream samples were examined. One was taken on

38

January 10, 1995 in this study [Table 6.7] and the other was taken on January 29, 1993 by

IDC.31[Table 6.8]

Table 6.7 Concentration of ions in CAD reaction tank 1/10/95

Anion/Cation Reaction Tank

Concentration (mg/L)

Anion/Cation Reaction Tank

Concentration (mg/L)

S042 38 Mg2+ 2.5

NO3 1584 Na+ 119

PO4 127 Si 7

Off 0.001 Mn < 0.1

F 10 Fe < 0.1

S2032 0 Ni < 0.1

C102 1500 Cu < 0.1

Cl- (total) 9 Zn < 0.01

Al(OH)4 54 P < 10

Ca2+ 803 S 32

K+ 9 B < 0.1

C032- 25

The analysis of both samples were also used to compare the amount of ions before

and after the installation of scrubber. The IDC sample was taken before the scrubber

installation, while the sample in this study was taken after scrubber installation. The

concentration of Ca2+ was doubled after the scrubber installation. The concentration of

Ca2+ represents the amount of lime used in the CAD reaction tank. This indicates that

amount of lime used was also increased after installation. The significant difference in

concentration between these two samples were the amount of P043- and C102 in the

samples. This agrees with the actual system of the scrubber, where the large amount of

39

P043- and C102 were presented in the process. The amount of C032-, NO3-, and S042-

were also detected in both samples. The results of this analysis were used in estimating

the amount of incoming components in the Chemsage environmental file.

Table 6.8 Concentration of ions in CAD reaction tank 1/29/9331

Anion/Cation Reaction Tank

Concentration (mg/L)

Anion/Cation Reaction Tank

Concentration (mg/L)

Cal' 420 HCO3- 107mgz+ 1.8 C032- 130

Na+ 71 NO3- 180

K+ ND3 P043- ND

Al(OH)4 29 F 24

OH- ND S042- 110

The DC sample had a charge misbalance with the cations and anions in a 2:1 ratio.

However, in this study, cations and anions had a charge balance. Measuements of the

anions in this study was done by using the DR/3000 Spectrophotometer. In this study,

there were high concentrations of both NO3- and C102 present in the sample. Therefore,

these two ions effects to CaF2 formation were studied carefully in the equilibriumcalculation. The primary utilization of the samples was to help determine what type of

compounds may be coming from the scrubber, so the charge misbalance was not alimitation.

The environment file, from which Chem Sage draws the possible components in the

CAD reaction tank at equilibrium, was created using the HSC Chemistry Software. The

3 ND = This ion is not detected in the sample.

40

list of compounds used in the environment file is found in Table 6.9. Since HF is dilute

enough to not have a significant vapor pressure and no other gas phase components were

needed, the gas phase was not considered in the simulation. HSC Chemistry generated a

list of all possible components containing any of the following elements: Al, F, Ca, P, 0,

N, H, Na, Cl, S, and C.

According to Table 6.7, the sample contained only small amounts of Si, Mg, Mn,

Fe, Ni, Cu, K, Zn, and B. Due to computational constraints, these elements were

considered negligible and were not considered in the simulation. In addition, the only

nitrogen compounds considered had nitrite and nitrate groups, the only sulfur compounds

considered had sulfate and thiosulfate groups, and the only carbon compounds fullyconsidered contained carbonate groups. Pitzer ion interaction parameters were used by

Chem Sage to model ion activity coefficients in the aqueous phase, were taken fromFrederick et al.19, Harvie et al.21, Haung22, Kim24, Millero25, Pitzer26, Pytkowicz27, and

Zemaitis et a130.(Appendix A)

The Pitzer model for ion activity coefficients treats all components in the aqueous

phase as dissociated. Undissociated and neutral species were assumed not to exist. All

non-aqueous components were assumed to be in the solid phase. When necessary,

compounds only found in the aqueous phase were entered into the program as ions (i.e. 1

mole of HF was entered as 1 mole of 1-1+ (aq) and 1 mole F (aq)). All simulations were

performed on a concentration basis. One or more species inputs were varied in order to

obtain equilibrium concentrations of the system over a large range of values.

41

Table 6.9 Component list in Chem Sage environment file

Compound Phase Compound Phase Compound Phase

Al(+3a) (aq) (NH4)20 (s) H2SO4*2H20 (s)Al(OH)4(-a) (aq) (NH4)2SO4 (s) H2SO4*3H20 (s)A1F(+2a) (aq) Al (s) H2SO4*4H20 (s)A1F2(+a) (aq) Al(OH)3 (s) H2SO4*6.5H20 (s)A102(-a) (aq) Al2(SO4)3 (s) H2SO4*H20 (s)A10H(+2a) (aq) Al2(SO4)3 *6H20 (s) H3PO4 (s)Ca( +2ai (aq) Al2Ca (s) H3PO4*0.5H20 (s)CaOH(+a) (aq) A1203 (s) HNO3 (s)Cl( -a) (aq) A1203*3H20 (s) HNO3*3H20 (s)C10(-a) (aq) A1203*H20 (s) HNO3*H20 (s)

C102(-a) (aq) A14Ca (s) Na (s)C103(-a) (aq) A1C13 (s) Na2CO3 (s)

C104(-a) (aq) AlC13*6H20 (s) Na20 (s)CO3(-2a) (aq) A1F3 (s) Na202 (s)F(-a) (aq) A1H3 (s) Na2SO4 (s)

H(+a) (aq) Al0 *OH (s) Na2SO4*10H20 (s)H2O (aq) Al0C1 (s) Na2SO4 *7H20 (s)

H2P207(-2a) (aq) AlP (s) Na3A1C16 (s)

H2PO4(-a) (aq) AlPO4 (s) Na3A1F6 (s)H3P207(-a) (aq) Ca (s) Na3PO4 (s)HCO2(-a) (aq) Ca(NO3)2 (s) Na4P207 (s)

HCO3(-a) (aq) Ca(NO3)2*2H20 (s) Na5A13F14 (s)

H02(-a) (aq) Ca(NO3)2*3H20 (s) NaA1C14 (s)HP207(-3a) (aq) Ca(NO3)2*4H20 (s) NaA102 (s)HP03F(-a) (aq) Ca(OH)2 (s) NaC1 (s)

HPO4(-2a) (aq) Ca2A1205 (s) NaC103 (s)HS(-a) (aq) Ca2P207 (s) NaC104 (s)HS204(-a) (aq) Ca3(PO4)2 (s) NaF (s)HS03(-a) (aq) Ca3P2 (s) NaH (s)HSO4(-a) (aq) Ca5(PO4)3F (s) NaHCO3 (s)N2H5(+a) (aq) Ca5(PO4)30H (s) NaNO2 (s)Na(+a) (aq) CaAl2 (s) NaNO3 (s)Na2P207(-2a) (aq) CaA14 (s) Na02 (s)

NaCO3(-a) (aq) CaC12 (s) NaOH (s)NaHP207(-2a) (aq) CaCO3 (s) NaP03 (s)NaS203(-a) (aq) CaF2 (s) NH4C1 (s)NaSO4( -a) (aq) CaH2 (s) NH4C104 (s)NH4(+a) (aq) CaHPO4 (s) NH4F (s)NO2(-a) (aq) CaHPO4*2H20 (s) NH4HF2 (s)NO3(-a) (aq) Ca0 (s) NH4NO3 (s)OH(-a) (aq) Ca02 (s) NH4OH (s)P207(-4a) (aq) Ca0C12 (s) P (s)PH4(+a) (aq) CaSO4 (s) P203 (s)

PO3F(-2a) (aq) CaSO4*1/2H20 (s) P205 (s)PO4(-3a) (aq) CaSO4 *2H2O (s) P4010 (s)S203(-2a) (aq) H202 (s) P406 (s)SO4(-2a) (aq) H2SO4 (s) P02 (s)

42

6.2.2 Comparison of Chem Sage 3.0 solubility graphs to experimental data

Figures 1-12 (Appendix B) show graphs comparing experimental solubility data

for a variety of chemical systems to results generated by Chem Sage. The objective of

performing these calculations was to assess the accuracy of Chemsage results. All

compounds tested were in the environment file used by Chem Sage to simulate the CAD

reaction tank. Experimental data was taken from Zemaitis et al.30, Silcock28, and Perry's

Handbook31. Overall, most of the solubility curves generated by Chem Sage agree towithin a factor of two of the experimental data and give as a match of the generalbehavioral trend as well. The results from Chem Sage generally were able to closely match

the behavior of the system at dilute concentrations (where most compound concentrations

were found in the CAD reaction tank). As concentrations became larger, the deviation

between the two became greater. All computations were performed at

25 °C and 1 bar unless otherwise specified. The molality of each substance represents the

number of moles of substance dissolved per liter of H2O.

Figure 1 shows an NaCl- NaOH -H20 system at 25 °C. The general trend is of

decreasing concentration of NaCI with increasing concentration of NaOH. While no NaC1

and only small amount of NaOH were found in mass balances for the CAD reaction tank,

it was the first system plotted to test the validity of Chem Sage's results. This is a fairly

simple system to model, and Chem Sage is able to closely match the solubility curve.

The solubility profile of an NaCl- CaSO4 -H20 system at 25 °C is found in figure 2.

The general trend is of increasing concentration of dissolved CaSO4 with increasing NaC1

concentration. The largest amount of CaSO4 found dissolved in the mass balance

analysis[Table 6.6] was 0.05 mol/L. Chem Sage is able to accurately predict the solubility

of CaSO4 in pure water as well as the general behavior of the system. At around 0.25mol/L of NaCl, Chem Sage begins to overpredict the amount of dissolved CaSO4 in the

system. This is probably due to the fact that the Pitzer interaction parameters for CaSO4

43

were originally evaluated to a maximum molality of 0.02 mol/L.24 At 0.25 mol/L of NaC1,

the dissolved concentration of CaSO4 is 0.03 mol/L.

Figure 3 shows a CaSO4-H3PO4-H20 system at 25 °C. The general trend is of

increasing CaSO4 concentration with increasing H3PO4 concentration. H3PO4 is apolyprotic acid, and at equilibrium there are several types of phosphorous compounds

present. The experimental results found in Silcock28 were not explicit as to the type of ion

phosphorous compound being described, so it was assumed that all of the phosphoric

compounds were grouped together as one (so that the x-axis represents the amount of

dissolved phosphorous atom in the system). Chem Sage overpredicted the amount of

dissolved CaSO4 in the system at higher phosphorous concentrations. However, laterfigures will show that only a very small amount of H3PO4 (0.1 mol/L) was needed to cause

reactive effects, and at this low of concentration ChemSage predicted the solubilitybehavior fairly well. At a H3PO4 concentration of 0.1 mol/L, Chem Sage predicted a

CaSO4 concentration of 0.04 mol/L, compared to a 0.025 mol/L concentration fromexperimental data.

An Na2SO4-CaSO4 -H20 system at 25 °C is shown in figure 4. The general trend is

of decreasing CaSO4 concentration with increasing Na2SO4 concentration until a

concentration of 0.1 mol/L Na2SO4 is reached. The experimental data then displays an

increasing amount of CaSO4 with increasing Na2SO4 concentration. Chem Sage was able

to accurately match the behavior at dilute concentrations and also matched the "bend"found at 0.1 mol/L Na2SO4. The maximum amount of Na2SO4 tested in the massbalance[Table 6.6] was 0.5 mol/L. At this concentration both the experimental data and

Chem Sage yielded a CaSO4 concentration of 0.012 mol/L.

Figure 5 shows an HNO3-CaSO4 -H20 system at 25 °C. The general trend is of

increasing CaSO4 concentration with increasing concentration of HNO3. Similar to figures

2 and 3, Chem Sage is able to predict the behavior at dilute concentrations, butoverpredicts the amount of dissolved CaSO4 at higher concentrations. The maximum

concentration of HNO3 tested in the Harris mass balance study was 0.26 mol/L [Table6.6]. At this concentration, experimental data predicted a CaSO4 value of 0.05 mol/L,

while ChemSage predicted a value of 0.07 mol/L.

Figure 6 shows the solubility profile of an NaNO3-Ca(OH)2-H20 System at 25 °C.

The experimental data shows a trend of increasing Ca(OH)2 concentration up with

increasing NaNO3 concentration to an NaNO3 concentration of 2.25 mol/L. From there,

the concentration of dissolved Ca(OH)2 slightly declines. Chem Sage demonstrates a

similar behavioral trend, although the maximum occurs at an NaNO3 concentration of 1.17

mol/L and the overall curve underpredicts the experimental data. The averageconcentration of NO3 ion in Table 6.6 was less than 0.18 mol/L. At this concentration,

the experimental data yielded a dissolved Ca(OH)2 concentration of 0.025 mol/L while

Chem Sage predicted a value slightly under 0.024 mol/L.

An NaNO2-Ca(OH)2-H20 System at 25 °C is found in figure 7. The experimental

data shows an initial rise of Ca(OH)2 concentration with increasing NaNO2 concentration.

Eventually the concentration of Ca(OH)2 levels off and remains constant. Chem Sage was

able to match the behavioral trend at dilute concentrations fairly well. At higher

concentrations of NaNO2, the dissolved Ca(OH)2 levels found in the results fromChem Sage continued to rise while the experimental concentrations became level.

The solubility profile for an NaF-HF-H20 system at 20 °C is found in figure 8.

Since HF was considered by Chem Sage to always be dissociated in solution, only HF's

influence on the solubility of other compounds could be modeled. The experimental data

shows a general trend of decreasing NaF solubility with increasing HF concentration.

Chem Sage predicted a similar trend, but the values were generally lower than the

experimental data. It should be noted that the entering HF stream to the CAD tank on the

Harris mass balance was only 0.0357 mol/L, and that the total possible fluorideconcentration in the system tested was 0.12 mol/L. Both concentrations are below the

45

saturation limits of NaF in pure water. In addition, for most cases in the Harris massbalance study, all of the HF reacted with Ca(OH)2 or other compounds.

An NaOH-Na2SO4 -H20 system at 25 °C is shown in figure 9. The experimental

data indicates a general trend of decreasing Na2SO4 solubility with increasing NaOH

concentration, with a slight hump in the data at an NaOH concentration of 3 mol/L.

Chem Sage is able to predict fairly accurately the amount of Na2SO4 dissolved in pure

water, as well as match the general downward solubility trend with increasing NaOH

concentration. However, no hump is found in the results from Chem Sage.

Figure 10 shows an NaCl-CaCO3 -H20 system at 25 °C. The general trend is of

increasing CaCO3 concentration as NaC1 concentration is increased. Chem Sage is able to

predict the general behavioral trend at fairly high NaC1 concentrations (1-3 mol/L). The

predicted concentration of CaCO3 in pure water also matches the value found in Perry's

handbook (1.4 x 104 mol/L in Perry's compared to 1.41 x 104 mol/L from Chem Sage).

The solubility profile of an Na2SO4-CaCO3-H20 system at 25 °C is found in figure

11. The general trend is of increasing CaCO3 concentration with increasing Na2SO4

concentration. Chem Sage gave a similar behavioral trend as the experimental data, but the

results consistently underpredicted the amount of CaCO3 dissolved in the system. The

solubility of CaCO3 in pure water is the same as in the NaC1-CaCO3-H20 profile in Figure

10.

Figure 12 shows an HC1-CaF2-H20 system at 20 °C. The general behavioral trend

of the system is of increasing CaF2 concentration with increasing HC1 concentration.

Chem Sage well-underpredicted the amount of CaF2 dissolved at high HCl concentrations.

At high concentrations the difference was almost two orders of magnitude. Attempts

were made to better the match by varying the Pitzer ion interaction parameters, but no

suitable combination could be found to improve conditions. No other suitableexperimental CaF2-solubility profiles at 25 °C containing other CAD reaction tank

46

components could be found. Despite this, Chem Sage is able to accurately predict the

concentration of CaF2 in pure water with the value found in Perry's Handbook (2.2 x 104mol/L for Perry's compared with 1.5 x 104 mol/L for Chem Sage).

6.2.3 Possible reaction mechanisms for increased fluoride discharge

In analyzing the mass balance (discussed in detail later), it became clear that there

were three possible ways to obtain non-CaF2 fluoride species in significant quantities:

Formation of aluminum fluoride(A1F3) compounds in acidic conditions, formation of free

fluoride ion due to carbonate competition for calcium, and formation of Cas(PO4)3F (s)

from H3PO4. The first two are discussed in detail in the following section. The formation

of Cas(PO4)3F will be discussed later in the compounds affecting CaF2 formation in basic

conditions section.

When the system becomes acidic, equilibrium shifts away from the formation of

CaF2 and towards the formation of aluminum compounds such as AIF3 (s) and A1F2+ (aq).

To determine the relationship between aluminum and calcium in their competition for

fluoride, a series of equilibrium calculations was performed. The system investigated

contained varying amounts of aluminum, calcium, and fluoride ions. Nitrate groups werealso added in order to keep the initial system charge neutral. Hydroxide ions were thenadded to the system in order to obtain a wide pH range. In the calculated equilibria,fluoride fully combined with either calcium or aluminum. There were negligible amountsof free fluoride ion.

Figure 6.5 shows a system with 1 mol/L aluminum ion, 1.5 mol/L calcium ion, and

3 mol/L fluoride ion. At basic conditions, virtually all of the fluoride is found as CaF2. Asthe pH drops, more aluminum fluoride compounds are formed. At a pH of 3.5, only 13%

47

of the fluoride species exist as CaF2, and all aluminum present in the system is combined

with fluoride.

1.6

1.4

1.2

1

0.8

0.6

0.4

0.2

0

2.00

Figure 6.5 1 Mol/L Al, 1.5 Mol/L Ca, 3 Mol/L F in 25 °CSystem Calcium and Fluoride Species as Function of pH

- - ......... - ..A%iik:

0 N2?4II

(s)A Al

1

) 11 )...

et alc1-4A

\ A

'I g 8 ::,4... -----...-a

1.6

1.4

1.2

1

0.8

0.6

0.4

0.2

0

2.00

4.00 6.00 8.00

pH

10.00

Figure 6.6 1.5 Mol/L Al, 1.5 Mol/L Ca, 3 Mol/L F in 25°CSystem Calcium and Fluoride Species as Function of pH

12.00

- - - ---. I'/ A1F3 (s)

. ., CaF2 (s)`-', A1203*H20 (s).

A , " AlO*OH (s)AlF(+2a) (aq)-

AA Ca(+2a) (aa)

' AII

oI

0 o033101MIEN

.'2A II!` - I

A A AAAlli

4.00 6.00

pH

8.00 10.00 12.00

48

In acidic conditions, the distribution of fluoride between AIF3 (s) and A1F2+ (aq) is

dependent on the amount of aluminum present in the system. At equilibrium, the system

will bind as much aluminum as possible with as much fluoride as possible, which is why

AIF3 (s) is the dominant aluminum fluoride species in figure 6.5. By increasing the amount

of aluminum available in the system (figure 6.6), the amount of AlF2+ (aq) increases in

order to accommodate the additional aluminum while still binding all of the fluoride. At

acidic conditions in a system with equal molar amounts of aluminum and fluoride, the

fluoride will almost entirely exist as A1F2+ (aq) (figure 6.7).

3

c 2.5

2

40 1.50

1

:42

S 0.5Lzl

02.00

Figure 6.7 3 Mol/L Al, 3 Mol/L Ca, 3 Mol/L F 25 °C System,Calcium and Fluoride Species as a Function of pH

AIF3

A1203*H20

+ Al(+3a)

Ca(+2a)

(s)

(aq)

(aq)

CaF2

(s) A10413H

(s)

(s)

(aq)

Ei"

11311 AlF(+2a)

---kr

111

m

ki I%

3.00 4.00 5.00 6.00 7.00

pH8.00 9.00 10 00 11 00 12 00

There are two ways to avoid the formation of large amounts of aluminum fluoride

products at equilibrium. The first way is to keep the pH basic. The second way is tomaintain a large calcium to aluminum ratio. Figures 6.6 and 6.7 show that at equivalent

stoichiometric levels of aluminum and calcium, aluminum fluoride products will be

49

strongly favored. However, if a ratio of 4:1 calcium ion to aluminum ion is used (figure

6.8), then CaF2 will be the dominant fluoride species at equilibrium throughout the pH

range.

2.5

2

c1.5

0.5

0

Figure 6.8 4 Mol/L Ca, 1 Mol/L Al, 3 Mol/L F in 25 °CSystem, Calcium and Fluoride Species as Function of pH

- '" CaF2 (s)0 A1203411 2 0 (s)A10011 (s)0 Ca(NO3)2*4H20 (s)Ca(OH)2 (s)

1(+3a) (aa)

ma, s - - - - - - - - - NI

k

...... ...2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10 00 11.00 12 00

pH

At high pH conditions it is possible for carbonate ion to combine with calcium to

form calcium carbonate. Competition with calcium for fluoride ion occurs by the

following equilibrium reaction:

CaCO3 (s) + 2 F (aq) CaF2 (s) + C032" (aq) (6.1)

50

Experimental data was gathered by Lawler and Williams4 in which CaF2, HF, and

H2CO3 (formed from dissolved CO2) concentrations were varied and the results weredisplayed as a function of pH. It was found that there was a large increase in the fluoride

ion concentration at a pH near 11. When carbonate was excluded, no such hump could be

found. Lawler and Williams also performed equilibrium calculations to simulate theexperimental data. Their results are found in figures 6.9 and 6.10. Their equilibrium

calculations confirmed increased fluoride ion concentrations at high pH. A concentration

of 0.001 mol/L H2CO3 was enough to push the free fluoride ion levels very close to the 17

mg/L fluoride ion discharge limit. In an "open system" all of the fluoride ion became free,

since the large levels of carbonate in the system had reacted with all of the availablecalcium.

Chem Sage was used to simulate the equilibrium system of Lawler and Williams.

The fluoride ion concentration used in both simulations was 87 mg fluoride ion/L, which is

approximately one order of magnitude lower than the incoming hydrofluoric acidconcentration in the Harris mass balance. Figure 6.9 shows the results using Chem Sage

when pH was varied by adding Ca(OH)2 to the system. No levels of carbonic acid were

found to cause any rise in fluoride ion concentration at high pH's, which did not appear toagree with the results of Lawler and Williams. However, by varying the pH usingCa(OH)2, more and more calcium is introduced in the system, which combines withH2CO3 by the following reaction:

H2CO3 Ca(OH)2 CaCO3 2H20 (6.2)

Figure 6.9 Simulation of Lawler and Williams System pHVaried by Ca(OH)2

1.00E+00

1.00E-01

1.00E-02

1.00E-03

1.00E-04

1.00E-05

A 0 H2CO3" 0.001 H2CO3

0 0.5 H2CO3Na2CO3

L&W 0.001 H2CO3L&W Open System

1.00E-066.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00

pH

51

As larger concentrations of H2CO3 were introduced in the system, largerconcentrations of Ca(OH)2 were needed to react with the H2CO3's hydrogen ion groups in

order to achieve the wide pH range. When Na2CO3 was used instead of H2CO3, the pH

quickly jumped into the basic region, as HF was the only acid in the system. The

formation of HCO3" (aq) ions also quickly took up the limited hydrogen ions that were in

the system. For an Na2CO3 concentration of 0.5 mol/L, a Ca(OH)2 concentration of 0.012

mol/L was needed to obtain a pH of 11.8. To meet this same pH for 0.5 mol/L of H2CO3,

a Ca(OH)2 concentration of 0.52 mol/L was required. More than 43 times the amount of

calcium is present in the H2CO3 system than the Na2CO3 system at this pH. The calcium

in the Na2CO3 system is quickly absorbed by carbonate ion. For the Na2CO3 system, a pH

greater than 13 is needed in order to introduce enough calcium in the system to reduce the

free fluoride ion concentration below the discharge limit.

52

Figure 6.10 represents the same system simulated by Lawler and Williams, only for

this simulation NaOH was used to vary the pH instead of Ca(OH)2. This allowed the level

of calcium in the system to remain constant and in low concentrations. The results from

ChemSage with H2CO3 levels of 0.001 mol/L and of 0.5 mol/L (open system) give results

that closely approximate the results found by Lawler and Williams at a pH of around 10.

At higher pH's, the 0.001 mol/L run of Lawler and Williams decreases, while theChemSage result slightly increases. The Lawler and Williams result for an open system

matches that of ChemSage at higher pH's.

Figure 6.10 Simulation of Lawler and Williams System pHVaried with NaOH

1.00E+00

1.00E-01

w 1.00E-02cci4J

c1.00E-03

1.00E-04

1.00E-056.00 7.00 8.00 9.00 10.00 11.00 12.00

pH

D 0.5 H2CO34-0.001 H2CO3- - L&W 0.001 H2CO34L&W Open System

0 0 DODO

0-4 0000111111.,...... . . . .

I

The CAD reaction tank varies the system pH by adding Ca(OH)2. Figure 6.9suggests that any incoming H2CO3 would be neutralized by Ca(OH)2, since an equivalent

amount of calcium needed to neutralize the carbonate group is added while the pH is held

constant by the added hydroxyl groups. However, Na2CO3 does not drive down the pH as

H2CO3 does, so additional Ca(OH)2 is not added as Na2CO3 concentration is increased.

The carbonate groups are able to combine with calcium without the system adding more

calcium to compensate.

6.2.4 Effect of input components on Harris mass balance of CAD reaction tank

The May 1, 1995, Harris group mass balance was used as a base system when

determining the effect of various compounds on the equilibrium state. Each inputcompound was varied to determine its effect on the system, and to find out if any non-

CaF2 fluoride species were produced. The following compounds were tested: Al2(SO4)3,

Ca(OH)2, HC1, HF, HNO3, H2CO3, H202, H2SO4, H3PO4, NaC1O2, NaNO3, Na2CO3, and

Na2SO4. The results are classified into three sections: Compounds having no effect onCaF2 formation, compounds hindering CaF2 formation at acidic conditions, and

compounds hindering CaF2 formation at basic conditions. For each component, twographs are shown. The first graph is of the system as a whole, listing only the compounds

which were found in large enough quantities to show on the graph. The pH (divided by

100 to fit on the graph) is also shown. The second graph is of all non-CaF2 fluoride

compounds. Simulations were performed at 25 °C and 1 bar. A summary of the analysis

is shown in table 6.10.

54

Table 6.10 Summary of Harris mass balance analysis

Species Fig.

(App A)

pH Ca+ Cone CaF2 Conc Non-CaF2 Compounds

Summing to > 0.0009 mol/L F

NaC102 13/14 ---- ---- ---- Under Limit

Na2SO4 15/16 ---- U Under Limit

NaNO3 17/18 ---- ---- ---- Under Limit

F1202 19/20 4 U Under Limit

Ca(OH)2 21/22 11 11 AIF3 (s), Alr (aq), F(aq) (acidic

region)

HNO3 23/24 U ft 4 AIF3 (s), AT (aq), F(aq) (acidic

region)

HNO3 w/pH 25/26 ---- 11 ---- Under Limit

HF 27/28 4 4 ft then U AlF3 (s), AlF" (aq), F(aq) (acidic

region)

HF w/pH 29/30 ---- ---- it Under Limit

H2CO3 31/32 U ft U AIF3 (s), AlF+ (aq), F(aq) (acidic

region)

H2CO3 w/pH 33/34 ---- ---- ---- Under Limit

HC1 35/36 U 11 U AlF3 (s), AT+ (aq), F(aq) (acidic

region)

HC1 w/pH 37/38 11 Under Limit

H2SO4 39/40 4 4 U AlF3 (s), Alr" (aq), F(aq) (acidic

region)

H2SO4 w/pH 41/42 ---- ---- ---- Under Limit

Al2(SO4)3 6.11/6.12 4 4 4 AlF3 (s), AlF" (aq), F(aq) (acidic

region)

Al2(SO4)3

w/pH

6.13/6.14 ---- ---- Under Limit

H3PO4 6.15/6.16 4 4 then ft 4 then ft Cas(PO4)3F (s), AlF3 (s), AlF* (aq), F

(aq)

H3PO4 w/pH 6.17/6.18 ---- ---- 4 Cas(PO4)3F (s)

Na2CO3 6.19/6.20 ---- 4 U F(aq), NaF (s)

55

6.2.4.1 Compounds having no effect on CaF2 formation

The following section lists compounds which do not have a direct effect on the

conversion of HF to CaF2. However, there are several components which createsituations that would make it easier for other components to have an effect on thereaction. Figures 13-20 are shown in appendix B.

NaCI02: Figures 13 and 14 show the effect of NaC1O2 on the system. As figure

14 indicates, the compound dissociates without having any effect on the system. The

maximum incoming concentration of NaC1O2 tested was 0.50 mol/L.

Na2SO4: The effect of Na2SO4 on the system is shown in figures 15 and 16. As

Na2SO4 is added into the system, CaSO4 is created and the level of calcium decreases.

However, a point is reached when CaSO4 is no longer created, and the calcium level

remains constant. The pH is unaffected, and the level of free fluoride ions is well below

the discharge limit. While Na2SO4 cannot be the cause of the fluoride discharge problem

alone, it could be an accomplice to other compounds which may cause increased fluoride

levels by calcium competition (such as Na2CO3). The maximum incoming concentration

of Na2SO4 tested was 0.50 mol/L.

NaNO3: Figures 17 and 18 show the effect of NaNO3 on the system. NaNO3

behaves similar to NaC1O2 in that it dissociates without having any effect on the system.

The maximum incoming concentration of NaNO3 tested was 0.14 mol/L.

H202: The effect of increasing H202 concentration on the system is found infigures 19 and 20. The amount of free fluoride in the system remains well below the

discharge limit. Added H202 concentrations lower the pH to a value of 7, which indicates

a pH neutral system. H202 does not cause the fluoride discharge problem alone, although

since it does drive down the pH it could be an accomplice to other compounds which may

56

cause increased non-CaF2 fluoride levels by creating aluminum fluoride products in acidic

conditions. The maximum incoming concentration of H202 tested was 0.40 mol/L.

6.2.4.2 Compounds affecting CaF2 formation in acidic conditions

The following section lists compounds which were able to form aluminum fluoride

compounds in acidic conditions. While CaF2 is still the dominant fluoride species in most

cases, the aluminum fluoride species formed are in large enough quantities to bypass the

fluoride discharge limit if they were to be released. In each case the dominant aluminum

fluoride compound is AIF3 (s). When the pH is held constant at an operating pH of 12, no

aluminum fluoride products form. The free fluoride ion concentration at this pH is well

below the discharge limit. (Figure 21-42 in App. B)

Ca(OH)2: Figures 21 and 22 show the effects of varying Ca(OH)2 concentration

on the system. Acidic conditions (and the resulting aluminum fluoride products) form due

to a lack of Ca(OH)2, rather than due to the presence of it. Acidic conditions began to

form when the concentration of Ca(OH)2 was approximately 80% of the amount indicated

in the Harris mass balance. The maximum amount of Ca(OH)2 tested was 0.20 mol/L.

HNO3: HNO3's effect on the system is found in figures 23 and 24. Acidic

conditions began to form when the concentration of HNO3 was approximately 40% more

than the concentration indicated in the Harris mass balance. As HNO3 concentration was

increased, more and more of the solid Ca(OH)2 remaining became soluble. The maximum

incoming concentration of HNO3 tested was 0.24 mol/L. When pH was held constant in

the basic region by adding Ca(OH)2 along with HNO3, HF proceeded to its normalconversion to CaF2 (figures 25 and 26). Calcium ion levels also rose with increasing

HNO3 concentration. With pH control, the maximum incoming concentration of HNO3

tested was 0.26 mol/L.

57

HF: As HF was added to the system, calcium levels dropped and the level of CaF2

rose (figure 27). When enough HF was added to drop the pH into acidic conditions(requiring approximate 280% of the Harris mass balance HF concentration), aluminum

fluoride products were found (figure 28). The maximum incoming concentration tested

was 0.10 mol/L. When pH was held constant (figure 29), all incoming HF was converted

to CaF2 and fluoride levels were well below the discharge limit (figure 30). The maximum

incoming concentration of HF tested with pH control was 0.12 mol/L.

H2CO3: Figures 31 and 32 show the effect of H2CO3 on the system. As H2CO3 is

added to the system, CaCO3 is formed. However, as the pH drops with added H2CO3, the

amount of CaCO3 declines and HCO3 ion becomes the dominant carbonate form. The

characteristic aluminum fluoride product formation is found at acidic conditions. Acidic

conditions were achieved at an incoming H2CO3 concentration of 0.06 mol/L. As the

CaCO3 concentration decreases, the level of calcium in the system rises, to a point where

at acidic conditions, the level of calcium is actually higher than if there was no H2CO3 in

the system at all. The maximum incoming concentration of H2CO3 tested was 0.10 mol/L.

When pH was held constant (figures 33 and 34) everything in the system remained

constant, except for the continuously increasing amounts of CaCO3. Enough calcium is

added when pH is held constant to neutralize all incoming carbonate from H2CO3. The

maximum incoming concentration of H2CO3 tested with the pH held constant was 1.00mol/L.

HCI: The effect of HC1 on the system is found in figures 35 and 36. The behavior

of the acid is fairly typical of all of the other compounds in this section, with aluminum

fluoride products forming in the acidic region. Acidic conditions are achieved at anincoming HC1 concentration of 0.06 mol/L. The largest concentration tested for HC1 was

0.20 mol/L. Keeping the pH constant (figures 37 and 38) indicates that there would be no

problem. Calcium levels would increase from all of the added Ca(OH)2. With pH control,

the maximum incoming concentration of HC1 tested was 0.20 mol/L.

58

H2SO4: The effect of H2SO4 on the system is shown in figures 39 and 40. The

aluminum fluoride product formation in acidic regions is found here as well, as well as the

formation of CaSO4. CaSO4 formation decreases the amount of available calcium in the

system. Acidic conditions were achieved at an incoming H2SO4 concentration of 0.03

mol/L, and the maximum incoming concentration tested was 0.08 mol/L. When pH was

held constant (figures 41 and 42), the non-CaF2 fluoride levels were well below the limit,

and there was enough incoming calcium to combine with the sulfate groups to keep the

free calcium levels constant. The maximum incoming H2SO4 concentration tested at aconstant pH was 0.10 mol/L.

Al2(SO4)3: Figures 6.11 and 6.12 show the effect of increasing concentration of

Al2(SO4)3 on the system. Increasing the concentration of Al2(SO4)3 has two major effects:

lowering of pH and increase in aluminum concentration. With this combination, the effect

of Al2(SO4)3 on the system is much greater than any other compound in this section.

Comparable concentrations of Al2(SO4)3 to the other compounds would result in non-

CaF2 species formation several orders of magnitude higher than the others. AlF2+ (aq) also

takes up a larger and larger fraction of the acidic aluminum compounds with increasing

Al2(SO4)3 concentration. Acidic conditions were achieved at an incoming Al2(SO4)3

concentration of 0.01 mol/L, and the maximum incoming concentration of Al2(SO4)3

tested was 0.012 mol/L.

When pH is held constant (figures 6.13 and 6.14), aluminum fluoride products are

found in negligible quantities and the total non-CaF2 fluoride component levels are well

below the discharge limit. The maximum incoming concentration of Al2(SO4)3 tested at a

constant pH was 0.1 mol/L.

0.18

0.16

0.14

0.12

0.10

C-) 0.08

0.06

1:7' 0.04

0.02

0.00

Figure 6.11 Effect of Al2(SO4)3 on System(Mass Balance Concentration = 0.000175 Mol/L)

= n : = .... = :

pH/100- - CaF2 (s)+ Ca(OH)2 (s)A CaSO4*2H20 (s)

X A1F3 (s)0 AIO*OH (s)0 CaSO4 (s)0 AIF( +2a) (aq)

NO3(-a) (aq)13 Ca(+2a) (aq)

IN = = -IS

0.000

0.0030

0.0025

0.0020

c.;

C..) 0.0015Es.e" 0.0010

0.0005

0.002 0.004 0.006 0.008

Incoming Conc. Al2(SO4)3 (Mol/L)0.010

Figure 6.12 Effect of Al2(SO4)3 on System

0.012

IIi

(aq)

'AIF3 (s)

- AlF(+2a)

F(-a) (aq)13 Total F

1/I

IF0

0.00000.000 0.002 0.004 0.006 0.008 0.010

Incoming Conc. Al2(SO4)3 (Mol/L)0.012

59

0.18

0.16

0.14

0.12

0.10

0.08

0.06

0.04

0.02

0.00

Figure 6.13 Effect of Added Amounts of Al2(SO4)3 andCa(OH)2 in pH Constant System

111 Ill III lh Ill 1pH/100 ml I. CaF2 (s)A1203*H20 (s) AlO*OH (s)

--1:1-0 Ca(OH)2 (s) + CaSO4 (s) 0

CaSO4*1/2H20 (s) 0 CaSO4 *2H20 (s) 0Ca( +2a) (aq) NO3(-a) (aq) 0

A OH(-a) (aq) 0 0II0 _

00 000pg 0 + + 2 1a ''' + 2++ t I 2 2 2

* %,Nt XX XA** Xft3 A A A A 4

,

,

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

Incoming Conc. Al2(SO4)3 (Mo1/L)

Figure 6.14 Effect of Added Amounts of Al2(504)3 andCa(OH)2 in pH Constant System

0.0010

0.0009

0.0008c

g 0.00074; 0.0006=I0

C...) 0.0005EM 0.0004

.r.".

0.0003*GP 0.0002;4

0.0001

0.00000

INI NI a a M NI NI IN IN NI IN NI NI IN M MI a

' F(-a) (aq)

M MI IN 111. IN a IN IN IN a IN 1

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 01

Incoming Conc. Al2(SO4)3 (Mol/L)

60

61

6.2.4.3 Compounds affecting CaF2 formation in basic conditions

The following section describes the two compounds, H3PO4 and Na2CO3, which

cause significant amounts of non-CaF2 fluoride species to form while operating at a high

pH. Since the CAD reaction tank is operated at a fairly high pH value (8<pH<12), these

two compounds would be the most likely suspects if equilibrium effects are the reasonbehind the fluoride discharge problem.

H3PO4: H3PO4 displays similar behavior to the other acids concerning the

formation of aluminum fluoride products at acidic conditions (figures 6.15 and 6.16).

However, the compound Cas(PO4)3F (s) also forms. The trend for the formation of this

compound was a sharp rise, a maximum, and then a slow decrease in concentration. The

maximum and the decrease occur at a concentration nearly an order of magnitude higher

than that of the fluoride discharge limit. In order to meet the discharge limit, the amount

of H3PO4 allowed in the CAD system is at 0.003 mol/L.

0.18

0.16

0.14

0.12

0.10

0.08

0.06

0.04

0.02

0.000.00

0.0100.009

t 0.0080.007

oci 0060(5 0.005E0 0.0042 0.0035cr 0.002

rza0.0010.000

0.00

Figure 6.15 Effect of H3PO4 on System

- CaF2 (s)0 Ca(OH)2 (s)A H(+a) (aq)

NO3(-a) (aq)

pH/100Ca5(PO4)3F (s)

--0 --- Ca(+2a) (aq)D H2PO4(-a) (aq)X OH(-a) (aq)

0A

II

0

is al ,X

111 ÷ -1-- 3. . .0.02 0.04 0.06 0.08

Incoming Conc. H3PO4 (Mol/L)

Figure 6.16 Effect of H3PO4 on System Non CaF2-Fluoride Species in Dilute Region

....,..nnAnnft............."."(

0.10

AlF3 (s)Ca5(PO4)3F (s)

- AIF( +2a) (aq)0 F(-a) (aq)--D Total F

0.02 0.04 0.06 0.08 0.10

Incoming Conc. H3PO4 (Mol/L)

62

63

When pH is held constant (figures 6.17 and 6.18), the aluminum fluoride products

are no longer found, but the amount of Cas(PO4)3F (s) continues to rise. An incoming

concentration of 0.003 mol/L H3PO4 was needed to exceed the fluoride discharge limit

should Ca5(PO4)3F (s) be discharged as fine, suspended particles.

0.18

40 0.16

0.14

. 4 0.12o

0.10

c.;

0.08

El 0.06t.= 0.04

0.02

0.000.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07

Figure 6.17 Effect of Added H3PO4 and Ca(OH)2in pH Constant System

Incoming Conc. H3PO4 (Mo0.08 0.09 0.10

0.0010

0.0009

0 0008

0.0007e.;O 0.00060

C.) 0.0005

- 0.00041-.

0.0003Ear 0.0002

0.0001

0.0000

0.0000

Figure 6.18 Effect of Added H3PO4 and Ca(OH)2in pH Constant System (Fluoride Limit = 0.0009 Mol/L)

Non CaF2-Fluoride Species in Dilute Region

0.0005 0.0010 0.0015 0.0020 0.0025

Incoming Conc. H3PO4 (Mol/L)0.0030

64

Na2CO3: The effect of Na2CO3 on the system is found in figures 6.19 and 6.20.

As Na2CO3 was added, the pH level remained fairly constant. However, the levels of free

calcium drop as CaCO3 (s) is formed. This causes a rise in the concentration level of free

fluoride ion. While CaF2 is still the dominant fluoride species, an incoming concentration

of 0.12 mol/L Na2CO3 causes free fluoride ion formation in significant enough quantities

to exceed the fluoride discharge limit.

CL

o

IO

0.18

0.16

0.14

0.12

0.10

0.08

0.06

0.04

0.02

0.00

Figure 6.19 Effect of Na2CO3 on SystempH/100 1.1 CaF2 (s)CaCO3 (s) _Ca(OH)2 (s)21 CO3(-2a) (aq)

1F(-a) (aq)Ca( +2a) (aq)NO3(-a) (aq)

A Na(+a) (aq) NaCO3( -a) (aq)0 OH(-a) (aq):ss

A0

++

+o*++

lk0

0++

o0000013000009th000DOODOODOO+

-I-+

+

+

0a 15...i.+.....-I-

i. . - - .+ + ' ...

0.00 0.10 0.20 0.30 0.40

Incoming Conc. Na2CO3 (Mol/L)

0.009

0.008

a 0.007

0.006

0.005

0.004

0.003

0.002

0.001

0

0.00

0.50

Figure 6.20 Effect of Na2CO3 on System

0.60

a

se

to s r0

''01

00

0/ NaF (s)- - F(-a) (aq)

1:1 Total F- I

4/

0.10 0.20 0.30 0.40

Incoming Conc. Na2CO3 (Mol/L)0.50 0.60

65

66

The effect of Na2CO3 was tested experimentally. HF, lime, and various amount of

Na2CO3 reacted in the batch reactor at room temperature. Figure 6.21 shows that at low

amount of Na2CO3, the fluoride concentration was lower than discharge limit. However,

increasing the amount of Na2CO3 to 5 times the amount of HF, causes a rise in fluoride

concentration to exceed the discharge limit. The experimental results agree with thehypothesis that Na2CO3 could be the cause of the high fluoride concentration.

Figure 6.21 Experimental results of HF and Na2CO3 reacts with lime at roomtemperature, average initial concentration of HF = 190 ppm

E

210

180

HF /lime/Na2CO31:1:0

60

30

0

0

6.2.5 Conclusion

5 10 15

Time (min)

20 25 30

Chem Sage 3.0 was used to simulate the Wacker Siltronic CAD reaction tank. An

analysis was performed to determine what incoming components could hinder the

67

formation of CaF2 and cause fluoride to be discharged from the waste treatment process.

The only instances when non-CaF2 fluoride species form in significant enough quantities to

exceed the fluoride discharge limit are when aluminum fluoride products form in acidic

conditions, the formation of Ca5(PO4)3F (s) when H3PO4 is added, or when free fluoride

ion forms when Na2CO3 is added. The aluminum concentrations found in the massbalance were low and the pH level in CAD reaction tank is between 9-11. Therefore, at

present conditions, the discharge limit would only be bypassed at acidic conditions if AIF3

(s) is able to bypass the solids press. No H3PO4 was found in the Harris mass balance,

although phosphate ions were detected in the sample in Table 6.7. The compounds AlF3

(s) and Cas(PO4)3F (s) would only cause a problem if they were to remain in the system as

fine, suspended particles in the discharge from the waste treatment plant. Na2CO3 causes

a large amount of free fluoride ion to form, which would definitely be able to bedischarged.

Na2CO3 is the only compound to produce a knowingly soluble fluoride species

(free fluoride ion) at the operating pH of the CAD tank and in large enough quantities to

exceed the fluoride discharge limit. If equilibrium effects are the reason behind thefluoride discharge limit problem, then Na2CO3 is the most likely cause.

Acidic compounds, when added in such quantities to cause the pH to drop into the

acidic region, caused the formation of AlF3 (s) and Alr(aq). At acidic conditions, the

sum of fluoride contained in non-CaF2 fluoride compounds in the system was large enough

to exceed the discharge limit if it was to remain as fine, suspended particles in thedischarge from the waste treatment plant. HF was fully converted to CaF2 when pH was

held constant by adding additional amounts of lime with the incoming acid.

When H3PO4 was added, the aluminum fluoride compounds mentioned above

formed when conditions were acidic. The compound Cas(PO4)3F (s) also formed in

significant enough quantities to exceed the discharge limit if it were to remain as fine,

suspended particles in the discharge from the waste treatment plant. Even when the pH

68

was kept in the basic region, the compound formed in large enough quantities to exceed

the discharge limit if not removed as particulate.

When Na2CO3 was added, the amount of free calcium in the system decreased to

negligible levels due to formation of CaCO3. As a result, the amount of free fluoride ion

in the system increased to an amount exceeding the discharge limit. Laboratory

experiments were conducted to confirm the hypothesis that Na2CO3 could cause the high

fluoride concentrations. The result of the experiments showed that Na2CO3 compound

caused high fluoride concentration.

In summary, the only instances when non-CaF2 fluoride species would form were

when conditions were acidic (aluminum fluoride products), or when H3PO4 or Na2CO3

were added at basic conditions (pH=11.8). Na2CO3 was the only component to produce

free fluoride ions in significant enough quantities to violate the discharge limit. If

equilibrium effects are the reason behind the fluoride ion discharge problem, then Na2CO3

is the most likely cause.

69

6.3 KINETIC STUDY

6.3.1 Reaction mechanisms for rate limiting step

The CaF2 formation involves two reaction mechanisms as shown in equation (6.3)and (6.4). Equation (6.3) is known as slaking reaction. The CaO reacts with water toform slaked lime, Ca(OH)2. The consecutive reaction of the slaked lime with HF to form

CaF2 is shown in equation (6.4). The slaking reaction is discussed in this section. The

CaF2 reaction [equation (6.4)] was studied and discussed in the following section.

CaO(), + H2O = Ca(OH)20) <=> Ca2+ + 20H- (6.3)

Ca2+ + 2F = CaF2(s) (6.4)

The slaking reaction is important whenever lime is added to aqueous solutions; for

example, in the treatment of aqueous wastes or in the causticizing process in a haft pulpmills. The slaking reaction proceeds rapidly and is exothermic with a heat of reaction of65.2 KJ/mol at 373.16 K.34 The lime used in wastewater treatment plants is very reactive.

According to the experiments, the slaking reaction time for lime in wastewater treatmentplant is about 60-80 seconds.

6.3.2 Qualitative analysis of kinetic data

Laboratory experiments between HF and lime were conducted in a 1L batchreactor at 25 ° C.(Figure 6.22) The lime usage was based on stoichiometric requirements

70

to convert the HF to CaF2. HF was also reacted with lime at ratios other thanstoichiometric.

HF:Lime = 1:1.3, X = 0.94

HF:lime =1:1, X = 0.65

HF:Lime = 1:2, X = 0.99

Figure 6.22 The experimental data of HF react with different ratioof lime in a batch reactor

The effective of mixing was tested in this experiment. The samples at the same

concentrations of HF reacted with lime at different stirring rate 60, 120, and 150 rpm.

Figure 6.23 shows the experimental results of HF and lime reaction at different stirring

rate. At 60 rpm, the final concentration of the sample was at 90 mg/1 which was higher

compared with the the samples reacted at 120 and 150 rpm. From the experimental

results, the optimum stirring rate at 150 rpm was used in the experiment for the kinetic

study.

71

200

180

160

EL 140

c 1202 100

80

60LL 40

20

0

f -60 rpm

120 rpm

G 150 rpm

2 4 6

Time (min)

8 10

Figure 6.23 The experimental data of HF reacted with lime at different mixing rate,the initial concentration of HF = 200 ppm

The kinetic data shows that the reaction is fast, at a one minute or less theconversion reached its final conversion. With 1:1 ratio of HF:lime, the reaction stopped at

approximately 63-65 % conversion. This was probably due to diffusion limitations created

by the CaF2 product layer. The molar volume of CaF2 has a higher specific molar volume

(24.5 cm3/mole) than CaO (18.3 cm3 /mole). This could result in blocking of micropores

through which the fluoride and/or calcium ions diffuse, thereby slowing the rate of CO-,formation as discussed in the following paragraph.

The reaction mechanism of CaF2 is discussed in this section. The Ca(OH)2dissolves in the water to form the Ca2+ and OH-.[Equation (6.3)] The Ca2+ reacts with Fto form the CaF2 products. The CaF2 is less soluble than the Ca(OH)2 so it precipitates on

Ca(OH)2 particle and forms a product layer. For further reaction, the Ca2+ must diffuse

through the product layer to get into the solution. The higher the CaO/HF ratio, the more

surface area there is available for CaF2 product to deposit on, and the thinner the CaF2product layer will be for a given conversion of HF to CaF2. Therefore, low ratios of

72

CaO/HF can lead to low conversion of HF to CaF2 even when there is more than thestoichiometric quantity of lime available.

6.3.3 Evaluation of rate of reaction

The reaction of HF and lime is very rapid and it always has a high conversion, only

the worst case conditions were estimated in this study. The average rate over the firstminute rate was used to determine the reaction order. The initial concentrations, CAO and

average rate of reaction, -rAo, were determined for each run of experiments. The average

rate, -rAo, was found by differentiating the data and extrapolating to zero time. The

experimental data was plotted on Figure 6.24. The slope of the plot gives a reaction order

of 1. The equation obtained rate is given in equation (6.5).

= k CHF (6.5)

- 'where k [=1 s

Because of the fast rate of the reaction, the Ca2+ concentration was maintained

constant except when limited by diffusion for runs with low CaO/HF ratio.

73

3

2.5

2

1.5

1

0.5

0

0 1 2In[HF]o (molim3)

3

Figure 6.24 Initial rate as a function of initial HF concentration(The ratio of HF : lime = 1:1.3)

The first order reaction rate of HF was used to approximate the rate constant by

using the first experimental data point to estimate the apparent rate constant. This results

in the slowest possible estimate of the apparent rate constant. The integral analysis

method was used to obtain the rate constants. Table 6.11 shows the results of rateconstant calculations of HF at different ratio of lime.

Table 6.11 The apparent kinetic initial rate constants for HF and lime reactionat different ratio of lime at 25 °C.

k (M' 2s1) HF:Lime

(mol Ratio)kinitial (min-1)

ki 1:1 0.95

k2 1:1.3 2.5

k3 1:2 3.83

74

6.3.4 Distribution of the residence time in the CAD reaction tank

According to data from the plant's operating records, the average residence time in

CAD reaction tank was obtained as shown in Table 6.12. The reaction rate data indicates

that the HF and lime reaction is very rapid compared with residence time in the CADreaction tank.

Table 6.12 The plant's data for residence time in the CAD reaction tank

Level of tank full 30-90%

Volume in CAD reaction tank 11,400-38,000 gal

Influent flowrate 60-140 gpm

Average influent flowrate 100 gpm

Residence time 114-380 min

Average operating pH 11.2

The reaction rate constants [Table 6.11] were used with the residence timedistribution to determine the conversion of F to CaF2 in the effluent from the CADreaction tank. The reaction tank was assumed to be a mixed flow reactor under the typicalflow conditions. The prediction of conversion for mixed flow is the integral of theproduct of the conversion function X(t) and the exit age distribution function E(t) overtime as shown in equation (6.6)."

(X) = f o X (t)E(t)dt X ( t)E(t)dt (6.6)

where (X) = average conversion exiting the reactor

X(t) = conversion at time t calculated from the rate equation

E(t) = exit age distribution function evaluated at time t

t = time

For the CSTR, the exit age distribution function is

,E(t) = 1exp( t/T )

where i = average residence time

(6.7)

75

Recalling the rate equation in equation (6.5), HF (A) reacts with lime (B) in CAD

reaction tank, yielding the concentration of HF(A)

C A(t) = CAOCkt

Solving for X(t), we have

X (t) = 1 ekt

(6.8)

(6.9)

Combining equation (6.7) and (6.9), the mean conversion is calculated numerically as

L r 1 (t< X >. L exp 1(1 e't )At0

(6.10)

Figure 6.25 shows the value of the integral in equation (6.6) with lime for mixed

flow in a CSTR obtained from equation (6.10). The ratio of HF:lime was at 1:1 with the

average residence time of 100 minutes. The average conversion exiting the reactor is

76

obtained from the value that the integral approaches asymptotically at long time. For the

condition in Figure 6.25, the conversion of HF the CaF2 was predicted to be at 84%.

Figure 6.25 The value of the integral in equation (6.10) versus time for mixed flow inCSTR with Tau = 100 min and k1 = 0.95 min' (HF:lime = 1:1)

Equation (6.10) was applied to predict the average conversion of a mixed flow

exiting the CSTR at the range of residence time. The results of prediction are shown in

Figure 6.26. The results were used to determine the minimum residence time required to

achieve the desired conversions. The longer the residence time in the reactor, the higher

the mean conversions were obtained. In the treatment plant, the average residence time is

at 100 minutes. The average F concentration of the CAD influent is about 200 ppm. At

this concentration, a conversion of 92.3% was needed in the CAD reaction tank in order

the effluent to meet the discharge limit at 17 mg/l. The required residence time in the

reaction tank was determined from the prediction in Figure 6.26. With the 1/1 ratio of

HF/lime (k1 = 0.95 min'), the minimum residence time required to meet the discharge limit

was at 300 minutes. At the 1/1.3 ratio (k2 = 2.5 min'), the required residence time in the

77

CAD reaction tank was at 40 minutes. Lastly, the residence time required for the 1/2 ratio

to achieve the high conversion was at 7 minutes.

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0 M0

CI HF/lime = 1:1HF/lime = 1:1.3

HF/lime = 1:2

I I i 1 i I

50 100 150 200 250 300 350

Residence time, Tau (min)400

Figure 6.26 The results of conversion prediction of a mixedflow over the range of the residence time in a CSTR, 0 <Tau< 400 min

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

1:1 HF/lime = 1:1HF/lime = 1:1.3

HF/lime = 1:2

0 10 20 30 40 50

Residence time, Tau (min)

Figure 6.27 The conversion of a mixed flow in a CSTR at differentratio of HF:lime, 0 <Tau< 50 minutes

78

Figure 6.27 shows the conversion of the CSTR mixed flow at the lower range ofthe residence time (0 < Tau< 50 min).

According to the data from treatment plant, the 50 lb of lime/100 gal of water ratio

was added to treat with HF in the CAD reaction tank. The HF/lime mol ratio wasapproximated to be at 1/50. The ratio of HF/lime in the CAD reaction tank was muchhigher than the ratios used in the prediction. Therefore, the conversion in the CADreaction tank was expected to be higher than the predicted conversions in Figure 6.26.

However, the flow condition in the CAD reaction tank may not be perfectly well mixed

CSTR. Thus, the average residence time in the CAD reaction tank was recommended tobe at least at the residence time determined from Figure 6.26 in order to achieve the highconversion.

6.3.5 Conclusion

The rate data obtained from the experiments in this study was used to determine

the rate of reaction and rate constants. The initial rate data determined the HF and lime

reaction to be the first order reaction with excess of lime. An integral analysis method was

used to determine the rate constants. The rate constants indicated, the reaction betweenHF and lime at high enough CaO/F ratios, to be very rapid compared to the averageresidence time in the reaction tank.

The prediction of a mixed flow conversion in a CSTR at different ratios of HF:lime

was done. It was used to determine the minimum residence time required to achieve the

high enough conversion in the CAD reaction tank. the ratio of HF and lime in the CAD

tank was much higher than the ratios used in the calculations of the prediction. Therefore,

the higher conversions were expected in the CAD reaction tank. The results of theprediction also indicated that the average residence time in CAD reaction tank was long

79

enough to convert most of the F in the reaction tank and able to meet the discharge limit

without any difficulty.

7. SUMMARY AND FUTURE WORK

7.1 SUMMARY

80

The material balance study shows that the insufficient amount of lime addition was

not the cause of the high fluoride concentration in waste treatment plant. The pH data

from the treatment plant also confirmed that the pH level was high enough that no AIF3 or

AlF2+ formation should occur in the CAD reaction tank and lamella clarifiers.

In equilibrium study, there are three components that could interfere the chemical

activity of the fluoride ions. First component was the A13+ ion in the reaction tank at

acidic conditions. The A13-' ion could cause the formation of A1F3(,) and Alr(a,l)

compounds in the system if it is present in a large quantities. The AIF3 could cause a

problem if they were to remain in the system as fine particles in the discharge from the

treatment plant. The H3PO4, and NaCO3. could result in the formation of compound other

than CaF2 if they are present in large enough concentrations in the waste treatment plant.

The H3PO4 could react with F ion to form the Ca5(PO4)3F solid in the reaction tank.

However, the Ca5(PO4)3F particles could be removed in the settling tank as the suspended

particles. If the Na2CO3 is present in the system, the CO3- ion will compete with F to

react with Cat -'. Significant quantities of Na2CO3 could decrease the amount of available

Ca2+ to react with F and could cause a large amount of free F ion to form in the reaction

tank. This would result the free F to be discharged from treatment plant. Therefore, in

order to achieve the high removal of F in the system, maintenance of a high pH is required

to eliminate the AIF3 formation. Also, the CO3- and PO4 concentrations must be

controlled to low enough concentrations so as not to affect the reaction of F with lime.

81

Kinetic study shows that the reaction of HF and lime is very rapid compared to the

average residence time in the reaction tank. Too slow reaction rate and too short aresidence time could not be the cause of high fluoride concentration in the plant.

7.2 FUTURE WORK

Another possibility that could result in slow reactions of fluoride and lime is the

formation of a non-porous product layer that blocks access of fluoride ion to the

remaining lime. A study of the lime particle used in the wastewater treatment plant is

recommended.

A faster technique to measure the conversion of F to CaF2 is needed to obtain better

kinetic data.

A study of H3PO4 effects to the HF reaction with lime is recommended.

A new reactor design is required in kinetic study. The recommended laboratory-scale

CSTR is composed with flow control valves to control the flowrate of HF and lime

slurry in to the reactor. Also, the solution mixer is needed to prepare the HF solution,

where the amount of HF added in the mixer could be controlled and the Fconcentration could be measured.

A study of the flow conditions in the reaction tank is recommended. Experiments to

determine the amount of the lime effectively bypassed and the volume of the dead zone

in the reaction tank are required. Running me tracer tests in the CAD reaction tank is

highly suggested.

82

BIBLIOGRAPHY

1 APHA(1980) Standard Methods for the Examination of Water and Wastewater, 15thEdition. American Public Health Association, Washington DC.

2 Plankey, B.J., and Patterson, H.H., Kinetics of Aluminum Fluoride Complexation inAcidic Waters, Environmental Science Technology. Vol 20, No. 2, 160-165, (1986)

3 Culp, R.L., and Stoltenberg, H.A., 'Fluoride Reduction at LaCrosse, Kansas,' JournalAWWA. 50, 423-31, (1958)

4 Lawler, D.F., and Williams, D.H., Equalization/Neutralization modeling , Water Res.Vol. 18, No 11, 1411-1419, (1984)

5 Parker, C. L., and Fong, C. C., (1975) Fluoride removal: technology and cost estimate.Ind. Wastes 21, No. 6, 23

6 Paulson, E. G., (1977) Reducing fluoride in industrial wastewater. Chem. EngngDeskbook Iss. 89, 17 October

Zabban, W. and Helwick, R., (1975) Defluoridation of wastewater. Proc. 30th PurdueInd. Waste Conf. Purdue Univ, 479

8 Zabban, W., and Jewett, H.W., (1967) The Treatment of Fluoride Wastes. Proceedingsof the 22nd Industrial Waste Conference, Purdue University, Engineering Bulletin No.129,22, 706-16

9 Varuntanya, C.P., and Shafer, D.R., Lime Precipitation Insufficient for RemovingFluoride, Industrial Wastewater, January/February 1994, 32-35 (1994)

10 Kelly, B., Harris Group Inc., (1995) Fab 2 CAD Reaction Tank Mass Balance, 10 May

83

11 Fog ler, H.S., (1992) Elements of Chemical Reaction Engineering, 2' Edition, PTRPrentice Hall New Jersey, 61-95

1: Laid ler, K.J., (1965) Chemical Kinetics, 2" Edition, McGraw-Hill Increased., 1-49

13 Moore, W.J., and Pearson, R.G., (1981) Kinetics and Mechanism, 3rd Edition, JohnWiley & Sons, Inc., 1-82

14 Laidler, K.J., Reaction Kinetics, Vol 2, Pergamon Press LTD., p. 84-86, (1963)

15 Hanson, C., Theliander, H., and Wimby, M., Tappi Jl., 78(9):105-110 September,(1995)

16 Hanson, C., Theliander, H., and Wimby, M.,Tappi Jl., 76(11 ):181-188, November,(1993)

17 Hanson, C., "Lime mud reburning- Properties and quality of the lime produced,",Ph.D.thesis, Chalmers University of Technology, Goteborg, Sweden, (1993)

18 Dorris, G. M., and Allen, L. H., J. Pulp Paper Sci. 11(4):J89, (1985)

19 Frederick, W.J., R.C. Streisel and H.A. Gasteiger. The Solubility of Aluminosilicates inAlkaline Pulping Liquors. Amer. Inst. of Chem. Engrs. Forest Products Division:Applications of Chemical Engineering Principles in the Forest Products and RelatedIndustries, 2, 67, (1988)

20 Green, D.W., ed. Perry's Chemical Engineers' Handbook, Sixth Edition. McGraw-Hill, Inc., (1984)

21 Harvie, C. E., N. Moller, and J.H. Weare. The Prediction of Mineral Solubilities inNatural Waters: The Na-K-Mg-Ca-H-C1-SO4-0H-HCO3-0O3-0O2-H20 System to HighIonic Strengths at 25 °C. Geochimica et Cosmochimica Acta, 48, 723, (1984)

84

22 Haung, H.-H. Estimation of Pitzer's Ion Interaction Parameters for ElectrolytesInvolved in Complex Formation Using a Chemical Equilibrium Model. Journal ofSolution Chemistry, 18, 1069, (1989)

23 Kelly, B. and W.J. Frederick. An Equilibrium Model for Trace Element Solubility inAqueous Inorganic Solutions. Amer. Inst. of Chem. Engrs. Forest Products Division:Applications of Chemical Engineering Principles in the Forest Products and RelatedIndustries, 1, 1, (1986)

24 Kim, H.-T., (1988) Thesis for degree of Ph.D. : Prediction of ThermodynamicProperties of Aqueous Electrolyte Solutions., Oregon State University.

25 Millero, F.J. The Estimation of the pICHA of Acids in Seawater Using the PitzerEquations. Geochimica et Cosmochimica Acta, 47, 2121, (1983)

26 Pitzer, K.S. Activity Coefficients in Electrolyte Solutions, 2nd Edition. CRC Press,Boca Raton, Florida, (1991)

27 Pytkowicz, R.M. Activity Coefficients in Electrolyte Solutions. CRC Press, WestPalm Beach, Florida, (1979)

28 Silcock, H.L. Solubilities of Inorganic and Organic Compounds, Volume 3, Part 3,Ternary and Multicomponent Systems of Inorganic Substances. Pergamon Press, NewYork, (1979)

29 Tchobanoglous, G., and Schroeder, D.E., Water Quality, Addioson-Wesley PublishingComapany, 465-554, (1987)

3° Zemaitis, J.F., Clark, D.M., Rafal M., and Scrivner N.C., Handbook of AqueousElectrolyte Solutions, Am. Ins. Chem. Engrs., (1986)

31 Young, G. IDC Memo to Wacker Siltronic: Wacker Wastewater Treatment Facility,March 5, (1993)

85

32 Sricharoenchaikul, V., (1995), Thesis for degree of Master of Science : Sulfur speciestransformation in black liquor pyrolysis, Oregon State University

33 Lee, W.L., (1974) Elementary Principles of Laboratory instruments, 3rd Edition, TheC.V. Mosby company, St. Louis

34 Holman, K.L., Warrick, R.P., and Carlson, K.R., Recausticizing Kinetics with millLiquor and lime, Kraft recovery operations short course, (1992)

35 Injander, P., and Theliander, H., Disintegration of lime mud particles: the influence ofmixing conditions, Dept. of Chemical Engineering Design, Goteburg, Sweden

86

APPENDICES

87

APPENDIX A Pitzer ion coefficients for Pitzdata.txt data file

Table A.1 and A.2 are the lists of the pitzer ion coefficients for Pitzcomp.exedatafile. Column #1 is the binary/ternary interaction(2 or 3). "REF" indicates thereference number which the interaction was taken from. The references are listed in the

following section. *'S indicate fields which should not be written in. ION #1, #2 and #3

are the names of the ions. The next four columns are the values. The first column is

either BO, THETA, or PSI depending on if it is a binary interaction, interaction of the

same charge, or interaction between three ion. The next two fields are B1 and B2 forbinary interactions. The last field is C(PHI) for binary interactions.

The following are the lists of the references correspond to the REF column

01. Pytkowicz, R.M. Activity Coefficients in Electrolyte Solutions, 2nd Ed. CRPress, Boca Raton, Florida, 1991.

02. Kim, H.-T., Prediction of Thermodynamic Properties of Aqueous ElectrolyteSolutions. 1988D PhD thesis, Oregon State University

03. Harvie, C.E., N. Moller, and J.H. Weare. The Prediction of Mineral Solubilities inNatural Waters: The Na-K-Mg-Ca-H-C1-SO4-0H-HCO3-0O3-0O2-H20 Systemto High Ionic Strengths at 25 C. Geochimica et Cosmochimica Acta, 48, 723, 1984.

04. Millero, F.J. The Estimation of the pK*(HA) of Acids in Seawater Using the PitzerEquations. Geochimica et Cosmochimica Acta, 47, 2121, 1983.

05. Pitzer, K.S. Activity Coefficients in Electrolyte Solutions, 2nd Edition. CRCPress, Boca Raton, Florida, 1991.

06. Haung, H.-H. Estimation of Pitzer's Ion Interation Parameters for ElectrolytesInvolved in Complex Formation Using a Chemical Equilibrium Model. Journal ofSolution Chemistry, 18, 1069, 1989.

88

The following 4 parameters are from Haung. It is not put in because it assumes that HF

behaves completely different that a "typical" value which Kim found for HF.

Table A.1 Pitzer ion coefficients from Haung

I REF ION#1 ION #2 ION#3 BO,THT,PSI B1 B2 C(PSI)

2 06 H F (NO COMPLEX) -0.03985 -3.859980 .00000 -0.0062012 06 H F (COMPLEX) 0.269271 0.000506 .00000 -0.0072372 06 H HF2 (COMPLEX) -.673283 -.758059 .00000 .3852162 06 F HF2 (COMPLEX) -.662393 .000000 .00000 .000000

Table A.2 Pitzer ion coefficients used for Pitzcomp.exe datafile in Chemsage software

I REF ION#1 ION#2 ION#3 BO,THT,PSI B1 B2 C (PSI)

2 01 H CL NONE .17750 .29450 .00000 .000802 01 H BR NONE .19600 .35640 .00000 .008272 01 H I NONE .23620 .39200 .00000 .001102 02 NA F NONE .03183 .18697 .00000 -.008402 01 NA CL NONE .07650 .26640 .00000 .001262 01 NA BR NONE .09730 .27910 .00000 .001162 01 NA I NONE .11950 .34390 .00000 .001802 01 NA OH NONE .08600 .25300 .00000 .004402 01 NA NO2 NONE .06410 .10150 .00000 -.004902 01 NA NO3 NONE .00680 .17830 .00000 -.000722 01 NA H2PO4 NONE -.05330 .03960 .00000 .007952 01 NA B(OH)4 NONE -.05260 .11040 .00000 .015402 01 K F NONE .08089 .20210 .00000 .000932 01 K CL NONE .04835 .21220 .00000 -.000842 01 K BR NONE .05690 .22120 .00000 -.001802 01 K I NONE .07460 .25170 .00000 -.004142 01 K OH NONE .12980 .32000 .00000 .004102 01 K NO2 NONE .01510 .01500 .00000 .000702 01 K NO3 NONE -.08160 .04940 .00000 .006602 01 K H2PO4 NONE -.06780 -.10420 .00000 .000002 01 NH4 CL NONE .05220 .19180 .00000 -.003012 01 NH4 BR NONE .06240 .19470 .00000 -.004362 01 NH4 NO3 NONE -.01540 .11200 .00000 -.000032 01 MG CL NONE .35240 1.68150 .00000 .005192 01 MG BR NONE .43270 1.75280 .00000 .003122 01 MG I NONE .49020 1.80410 .00000 .007932 01 MG NO3 NONE .36710 1.58480 .00000 -.020632 01 CA CL NONE .31590 1.61400 .00000 -.000342 01 CA BR NONE .38160 1.61330 .00000 -.002572 01 CA I NONE .43790 1.80680 .00000 -.000842 01 CA NO3 NONE .21080 1.40930 .00000 -.020142 01 SR CL NONE .28580 1.66730 .00000 -.001312 01 SR BR NONE .33110 1.71150 .00000 .001232 01 SR I NONE .40130 1.86000 .00000 .00266

eniX,00,-INENNi-nd'We\ICOMMOMW.700t-,WWr-100WM000000000000000000000000000

0101M'cVCDMWNWWWC-.0101MNWWM(DpenOMI-IMWOOMMNOOMM000000000000000000000000000

0101 MNOOC,DIDOdPNM0r1

MOW,Ar-IWMMWN0001PI--000000000000000000000000000

r1 r-1 rAt-IONOrA0000.7N00 O,ArAelMindliN0c-IOMMNdli1 enr-1000c-1000000000000000000000000000

000000000000000,1 000000000000000000000000000000000000C)00000000000

1111

111111111111111

I111

I1

0000000000000000000000000000000000000000000000000000000000000000

0000000000000000000000000000000000000000000000000000000000000000

00000000000000000000000000000MM01 MNW01 M000000000000000000000000000

0000000000000000000000000000M01 en0 NMWMM000000000000000000000000000

0000000000000000000000000000NMONdiNenW01000000000000000000000000d00

."*"""".............

enen.7NdiNIM01

11111111

0000000000000000000000000000000000000000000000000000000000000000

CD en 00 00 In M e001C0 Lf) Ln CD enenLnenLnenOen0 0 ln en en LnenC) CD 01 0 0 .7 01 N O CO

C) 0 0 0 0 0 0 CD 0 CD 0 0 CD CD CD 0 0 CD 0 0 Ci 0 C) (Z) 0

CD (Zi

OWMWNONMMNO,AWONd,Mr-idiMMWMM.7W,ANenWMI-INOWd110000000000000,000000000000000

MCIIWWOMMNME---MOldr---NMME---.4.7iNNMN0000000000001000000000000000

endilillaMMM.71COWlC)Mend,WWMWWW,AMd,NNWWMMONOWMNOd,000000,000000000000000000000

rA rA rA rA

rA rA ri rA rA rA rA ,A rA ri rA rA rA rA rA rA

,A

ri

M ro 01 ro cv ro (NI cv Pc) 0') CV

000011 000000000000000M0 M

MM

CM

00 N0000 0000000000 0000000000 0 0 0 0

WW,DMNNMPIOPIMMOWNOWNMNWI-IMONMO.700WMOM0000000000000000000000000

diNd,r-iNNLII.-ANWWOWNWNOMMWM.70WNt-id,d,d,MOOM,AMWM.71q300NNLON0100W0000MODOWONG,

MW,ANMNMWNNr-ldr0,-IWWMd,NW,-IWMd,Nd,NNNLi10WG00,Aen,AMOdi,ACOT-1NMrAenONMOr--00000M

,ANM.700101M.7MenCOMMNPOr-10000,AMNrAN,ANNNMM0000000000000000C)000000C700,-,

1I

11

1111

11

I1111

0000000000000000000000000000000000000000000000000000000000000000

zzzzzzZZZZZZZZZZZZZZZZZzzzzzzzzzzZzzzzzzzzZzzzzzZZZZ:ZZZZZZZ'ZZzzz

.41,-10

Men

rn

rn

0,-)

co ,t1 01 0 di .7 .7 .7 .7 di .7 di di di .7 di .7 0'

di U U

Ini',

04c4

C),144c4

04400c4 04000(1,000000000000001-4

mXZZKC

mwmc4<Z,,,40,,UOUrOM

Z 0 Pg FA Z CJ C) c) (A FA Z () Ej z c) m FA Z () Z M C) Z M (34

cn a4 la, cn m m En m m u) m Cl) 4 Z 4 ti Z ,t1 ,J4 Z 4 C)

4 C) C) m PQ Z PI Z CJ al

m()

Xrl

NN

.41

NE4<<W00001-1ZZZZOO<F,4KC

ol-fznorwo

HFAH<

MH46,(4,(44

cn f)=1 cO oci pg

PL,

Z C) CJ N m m m D

D Z Z Z 4

4 4, Z 4 Z Z (i N Ci C) 0 PO

D X X X X X X X ,--1 .1 ,1 Z Z 4 X X X ,-1 Z Z Z Z Z Z Z

4 4C)

0000000000000000000000000000000000000000000000000000000000000000

NNNNNC\INNNNCINCANNNNNNNNNNNNNNNNNC\INNNNNNNNNCNINNNNNNNNNNNNNNNNNNNN(NINN

90

2 01 MG CA NONE .01000 .00000 .00000 .000002 01 CA CO NONE -.05500 .00000 .00000 .000002 01 CL BR NONE .00000 .00000 .00000 .000002 01 CL OH NONE -.05000 .00000 .00000 .000002 01 BR OH NONE -.06500 .00000 .00000 .000002 01 CL SO4 NONE -.03500 .00000 .00000 .000002 01 CL NO3 NONE .01600 .00000 .00000 .000002 01 CL H2PO4 NONE .10040 .00000 .00000 .000003 04 MG CA SO4 .00700 .00000 .00000 .000003 01 H LI CL .00000 .00000 .00000 .000003 01 H LI BR .00000 .00000 .00000 .000003 01 H LI cL04 -.00170 .00000 .00000 .000003 01 H NA CL -.00400 .00000 .00000 .000003 01 H NA BR -.01200 .0000 .00000 .000003 01 H NA CLO4 -.01600 .00000 .00000 .000003 01 H K CL -.00700 .00000 .00000 .000003 01 H K BR -.02100 .00000 .00000 .000003 01 H CS CL -.01900 .00000 .00000 .000003 01 H NH4 CL .00000 .00000 .00000 .000003 01 LI NA CL -.00300 .00000 .00000 .000003 01 LI NA NO3 -.00720 .00000 .00000 .000003 01 LI NA CLO4 -.00800 .00000 .00000 .000003 01 LI NA OAC -.00430 .00000 .00000 .000003 01 LI K CL -.01000 .00000 .00000 .000003 01 LI CS CL -.00940 .00000 .00000 .000003 01 NA K CL -.00180 .00000 .00000 .000003 01 NA K BR -.00220 .00000 .00000 .000003 01 NA K NO3 -.00120 .00000 .00000 .000003 01 NA K SO4 -.01000 .00000 .00000 .000003 01 NA CS CL -.00300 .00000 .00000 .000003 01 K CS CL -.00130 .00000 .00000 .000003 01 H SR CL .01800 .00000 .00000 .000003 01 H BA CL .02400 .00000 .00000 .000003 01 H MN CL .00000 .00000 .00000 .000003 01 LI BA CL .01900 .00000 .00000 .000003 01 NA MG CL .00000 .00000 .00000 .000003 01 NA MG SO4 .00000 .00000 .00000 .000003 01 NA CA CL .00000 .00000 .00000 .000003 01 NA BA CL .00000 .00000 .00000 .000003 01 NA MN CL -.00300 .00000 .00000 .000003 01 NA CO CL -.00100 .00000 .00000 .000003 01 NA CU CL -.01400 .00000 .00000 .000003 01 NA CU SO4 -.01100 .00000 .00000 .000003 01 K CA CL -.01500 .00000 .00000 .000003 01 K BA CL .00000 .00000 .00000 .000003 01 CS BA CL .00000 .00000 .00000 .000003 01 MG CA CL .00000 .00000 .00000 .000003 01 CA co CL .01300 .00000 .00000 .000003 01 NA CL BR .00000 .00000 .00000 .000003 01 K CL BR .00000 .00000 .00000 .000003 01 NA CL OH -.00600 .00000 .00000 .000003 01 K CL OH -.00800 .00000 .00000 .000003 01 NA BR OH -.01800 .00000 .00000 .000003 01 K BR OH -.01400 .00000 .00000 .000003 01 NA CL SO4 .00740 .00000 .00000 .000003 01 K CL SO4 .00000 .00000 .00000 .000003 01 MG CL SO4 .00000 .00000 .00000 .000003 01 CU CL SO4 .04300 .00000 .00000 .000003 01 LI CL NO3 -.00300 .00000 .00000 .000003 01 NA CL NO3 -.00600 .00000 .00000 .000003 01 K CL NO3 -.00600 .00000 .00000 .000003 01 MG CL NO3 .00000 .00000 .00000 .000003 01 CA CL NO3 -.01700 .00000 .00000 .000003 01 K CL H2PO4 .06000 .00000 .00000 .00000

WNP/COCOlOr--WN01401,1Mr-Nr-rA4CDONNON4cOMM.4.41,14WC\Ir--WVDWMNOCDON01COT-ALOMCDMCDWr-MOW4M Mt-.1,-immoomN01,--1c0WN=1W01WWcr---.(--CDCDOmprol---WNMCVt-IVDNMCTICDt-INCOMNOCDOWMCDV--Mk.00144.4V--014.401WV--NM4CD rA CA 4 N ri 01 ri 01 01 Ul 111 r VD CI 01 cr H 1I 0 0 CD ri CD .11 CO t--1 H CD VD C) Ln VD CD CD 01 CD CD N r- e0 0 CD 0 V-- rA 01 CD CD Ul CD 01 CI 0/ ri N VD CV ,11 co d1 C3 CO ChCD CD CD CD CD ci CD C3 CD CD CD CD CD CD CV CD ri CD CD CD CD di CD CD CD CD CD CD CD CD CD CD CD CD CD CD 01 CD CD 0 0 0 CD CD r- C) CD C3 CD CD CD C) CD ri CD CD CD 01 CD CD CD 0 4 riCD CD CD CD CD C) CD CD CD CD CD C3 CD CD CD CD CD CD CD CD CD CD CD CD C) CD CD CD CD CD C3 C) CD CD CD CD CD CD CD CD CD CD CD CD 01 CD CD C3 CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD

I 1 1 1 I I I I I I I I I I 1 1 1 1 1 1 1 1 1 1 I 1111 1111 0 1 1

CD CD CD C, CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD C) C) CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD C) CD CD 0 CD C) CD CD CD CD CD C) CD CD CD CD CD CD CDCD CD CD CD CD C) CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD C) CD CD CD CD CD CD CD CD CD CD C) CD CD CD CD CD CD CD CD CD CD CD 0 CD CD CD CD CD CD CD CD CD CD CD CD CD 0 C) CD 0CD CD CD CD CD CD C) CD CD CD CD CD CD CD CD CD CD CD CD C) CD CD CD C) CD CD C) CD CD CD CD CD CD 0 CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD C) 0 CD CD CD CD CD CD CD 0CD CD CD CD C) CD CD CD 0 CD CD CD CD CD CD CD CD CD CD CD CD CD CD C) CD CD C) CD CD CD CD CD CD C) CD CD CD 0 CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD C) CD CD CDC) CD CD CD C) CD CD CD CD CD CD CD CD CD C, CD C) C) CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD C) CD CD CD CD 0 CD CD CD CD CD CD 0 CD CD CD CD CD CD CD CD CD 0 CD CD 0 CD CD CD CD CD

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

V) CO CO CD di cr N ri t-- .41 =44 01 N r- CD In CO .41 Ul 01 di N r.-1 Ul 01 .41 01 co di in CD CO VD CO r- r- CO N 01 VD 01 rA 0/ Ul CO 141 CO CO C) CO co co ,--I CO r- on ,-I N C) 1-0 ul ,A ,-1 VDIn 1.1) 01 CO di CO -41 in In 01 .41 di 01 CD LO CO CO CO CO cil rA In VD CO rA C, 00 CV N r- 01 ri VD UD CO CD liD In CO CO CO N Ul .M co C) CO ri CO CD 01 N CO CV CO .M CD CO C) Ul ri VD LO riH N 01 01 CV 01 CI CI CO N CO N Cn CI CO CO CO CD rA 01 .4 .4 01 N ul co CO N (NI ,-1 ,A Cn H ,-1 Ul 01 r10 UD CO r- CO Ul ,-1 CD ri Ul l00 ri CO Ul ri r101 CO 0 cr Ul C4 ,A 0) .cli VDCD N CO 4 4 ul r- cv CD in cv CD r- CO In CO CD CD dl 01 CD Ul un cv In CO CV VD Ul 01 141 CV 4 ri di r- 00 CV CV CV VD 01 N CD ,A .41 ri In .Zi, N N CD V- N CD VD 01 N 00 01 CV CO ri C---.4 CI 01 di r- cD 01 di di N N N ri el CD ,-I ,-1 cl N cv cv N -11 01 01 ri ri H CD CD 01 ri CD CD CD CD 01 01 01 ri dr CD CO rA CD 01 CD 0/ CO 01 CD ri ri r.-1 Ul N Ch rA VD 01 ri 01 CO

. . . . . . . . . . . . . . .

I 1 1 1 I 1 1 I I 1 1 (N 1 1 1 I I I I 1111rA rA ri ,A ri ,A ri

N N 0 N ,Ii r-I Ul CD CO r- CO In co .cti N 0,) CO 01 01 01 QD .cti r-1 CD 01 CV CD CO N di 00 VD .1i 01 r-1 H .4 VD CO .41 N VD ri CO ri CO ri N CO N 4 H in r- u) CO in in uP r- w N COri H 01 r- In '.0 00 C) CD cr di H CD Ul CO N CO CD C\1 CO N H 01 ri 00 N VD VD CD N c) CO .11 4 ri CV C) N 01 4 CD CD rA rA .4 N CD CO CV N N r- CO di in CO H N H N .41 Cq CO riNWCOMMW0401-10C101,1 01MNWV--1--1.4WWNNO0W000OrACONWV0Mr40014.--1,--14rAM000L---L0 M0N0101.400010,00MCOMNCOCl ri CO CD di di In CD co CV r- H ,--1 cv r- N In CV 00 CO ,A CV el 01 Ul CD di 01 01 CO CD UD 01 01 N cv 01 CD VD N N QD ri CV Ul CD In C) CD 4 rl 01 ,711 H N '4' -;14 CO 00 VD LO CO CO CDC) N CD N CV ,-I CD CI H ri ri ri C) CD C) ri CD CD rA C) rA ri C3 ri 1.--1 ri CD CD CD CD CD ri ri CD CD CD ri CD ri ri CD rA ri ri VD CD CD CD CD CD CD CD CD ri CD rA ri C3 CD CD CD CD N N

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ,t1 . . . . . .

1 1 I 1 I 1 I I I I I1 01 1 1 1 I I I I I I

fil

Li 5' N 2 2 2 2 2 20000000000000000000000000000000000000000000000000000zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz w L." 0 0di .41 01.100.41 0100,14 N 01k.oN,:p .4 01 el el 0 0 di mcntOOmON OmN0 00(OWN 0101M MNU MN M Ococv OZOZ4U04.zt0ONN.1.4140,1c4 X,1001-100144Z0r4U004mvp 00K4 1.44C4 00(J0.1400U 1-10ZZU U00cOcOMOr4UZUMI-IOUZZUUMNUMPP44NZP.,0 .-11-1ZZ UMI-IZZO4ZUUZZ4HUMZ CV MM

UU

UN00100-14r1.4UM

(0 VD

clicr441.11.414,NMHHH1-11--IHHHI-ig4444604 MMOIMMMPOMMCOMMO)MCO01-14-1,-11-1XZE1WWWW i-iHg6444X4,-1,1,-141-1,14.-IZZZZZZZI:4c41c4c4UUVUOUU4E-1E-IPE--IFIZZZWXX(.1,-1,IZZZZZZN N N N CV N N N N CV N CV N N CI CI N N CV N N N Cl N CV N N N cv N CV CV N N N N N CV CV CV CV N CV CV CV N N Cl N N N N N N N CV CV CV N CV N N N NCD CD CD CD CD CD C) CD CD CD CD CD CD CD CD CD CD CD CD CD CD C) CD (D) CD CD CD CD CD CD CD CD CD 0 CD CD C) CD CD CD CD C3 CD CD CD CD CD CD CD CD 0 CD CD CD CD CD CD CD CD CD CD CD C) CD

N CV CV C4 CV CV CV CV C\I (NI CV C4 N N Cs] CA N N N N N CI N CV CV CV N N N CI N CV N N N CV CV N CI CV CV CV N CV N N N CV CV C4 CV CV N N CV CV cv cv N N cv cv N N

92

2 02 NA HASO4 NONE .13607 1.70125 .00000 .012022 02 NA C4H204 NONE .23639 .82784 .00000 -.022182 02 K H2P207 NONE -.00585 1.25198 .00000 .005242 02 K HASO4 NONE .10466 1.80042 .00000 .000002 02 K CR04 NONE .07712 1.18413 .00000 -.001072 02 K CR207 NONE -.01111 2.33306 .00000 .000002 02 K PT(CN)4 NONE .05955 2.25539 .00000 .000002 02 RB SO4 NONE .09232 .75746 .00000 -.013392 02 RB S208 NONE .20464 -.26340 .00000 .000002 02 CS SO4 NONE .14294 .66711 .00000 -.027462 02 CS 8208 NONE .13283 -.76429 .00000 .000002 02 NH4 HPO4 NONE -.04259 -.69871 .00000 .005272 02 NH4 B10H10 NONE .15824 1.46202 .00000 -.017102 02 CN3H6 CO3 NONE -.07420 .22809 .00000 .013802 02 MG CLO4 NONE .50083 1.94817 .00000 .008482 02 MG AC NONE .20801 1.05448 .00000 -.012862 02 CA CLO4 NONE .47924 2.16287 .00000 -.008372 02 SR CLO4 NONE .44138 2.00236 .00000 -.014542 02 BA CLO4 NONE .32673 2.53859 .00000 -.015762 02 BA AC NONE .22541 1.50143 .00000 -.033062 02 MN CL2 NONE .29486 2.01251 .00000 -.015282 02 MN BR NONE .44655 1.34477 .00000 -.022692 02 MN CLO4 NONE .50957 2.16209 .00000 .011442 02 NI BR NONE .44305 1.48323 .00000 -.005902 02 NI CLO4 NONE .49285 1.98517 .00000 .016792 02 NI NO3 NONE .30978 2.10644 .00000 -.003942 02 CO CLO4 NONE .50409 1.96664 .00000 .013492 02 CU BR NONE .41247 1.66270 .00000 -.042622 02 CU CLO4 NONE .48984 1.90361 .00000 .008392 02 CU C7H703S NONE .08473 1.79523 .00000 .000002 02 ZN F NONE .00144 -.08746 .00000 .000002 02 ZN CLO4 NONE .52365 1.46269 .00000 .007482 02 ZN C7H703S NONE .11840 1.67138 .00000 .000002 02 CD CL NONE .01624 .43945 .00000 .001092 02 CD BR NONE .02087 -.86302 .00000 .002842 02 CD I NONE .14916 .55935 .00000 -.011172 02 CD CLO4 NONE .38986 1.99610 .00000 .020752 02 CD NO2 NONE .00265 -2.15854 .00000 .003022 02 CD NO3 NONE .28764 1.68468 .00000 -.025872 02 CD C7H703S NONE .07161 1.75817 .00000 .000002 02 PB CL NONE .08010 -2.57126 .00000 .000002 02 PB CLO4 NONE .33500 1.61813 .00000 -.009042 02 PB NO3 NONE .01506 -.27095 .00000 -.013302 02 UO2 CLO4 NONE .66563 1.42853 .00000 .006992 02 C8H22N2 CL NONE .10390 -.10568 .00000 .001652 02 C8H22N2 I NONE -.07160 -.85778 .00000 .011562 02 LA CL NONE .59625 5.60000 .00000 -.024642 02 LA CLO4 NONE .83815 6.53330 .00000 -.012882 02 LA NO3 NONE .30507 5.13330 .00000 -.017502 02 LA C2H5SO4 NONE .80506 5.23150 .00000 -.103892 02 PR CL NONE .58804 5.60000 .00000 -.020602 02 PR CLO4 NONE .82454 6.53330 .00000 -.009142 02 PR NO3 NONE .32615 5.13330 .00000 -.018512 02 PR C2H5SO4 NONE .80996 5.31110 .00000 -.099722 02 ND CL NONE .58674 5.60000 .00000 -.018822 02 ND CLO4 NONE .81468 6.53330 .00000 -.006772 02 ND NO3 NONE .33927 5.13330 .00000 -.019452 02 ND C2H5504 NONE .79101 5.49280 .00000 -.091352 02 SM CL NONE .59361 5.60000 .00000 -.019142 02 SM CLO4 NONE .82673 6.53330 .00000 -.004872 02 SM NO3 NONE .35802 5.13330 .00000 -.018842 02 SM C2H5SO4 NONE 84486 5.80160 .00000 -.100392 02 EU CL NONE .60135 5.60000 .00000 -.019262 02 EU C2H5SO4 NONE .80148 5.67230 .00000 -.08613

93

2 02 GA CLO4 NONE .78535 5.20550 .00000 .042022 02 GD CL NONE .61142 5.60000 .00000 -.019242 02 GD CLO4 NONE .84832 6.53330 .00000 -.007922 02 GD NO3 NONE .37841 5.13330 .00000 -.019602 02 GD C2H5SO4 NONE .85152 5.46190 .00000 -.102242 02 TB CL NONE .62231 5.60000 .00000 -.019232 02 TB CLO4 NONE .88329 6.53330 .00000 -.011122 02 TB NO3 NONE .36850 5.13330 .00000 -.017942 02 TB C2H5SO4 NONE .84999 5.66880 .00000 -.096762 02 DY CL NONE .62826 5.60000 .00000 -.018952 02 DY CLO4 NONE .88021 6.53330 .00000 -.009472 02 DY C2H5SO4 NONE .85138 5.90230 .00000 -.092482 02 HO CL NONE .62346 5.60000 .00000 -.016752 02 HO CLO4 NONE .87129 6.53330 .00000 -.006992 02 HO C2H5SO4 NONE .84317 5.49720 .00000 -.093962 02 ER CL NONE .62158 5.60000 .00000 -.015242 02 ER CLO4 NONE .87506 6.53330 .00000 -.006712 02 ER NO3 NONE .43114 5.13330 .00000 -.025872 02 ER C2H5SO4 NONE .85345 5.62910 .00000 -.093712 02 TM CL NONE .62640 5.60000 .00000 -.015132 02 TM CLO4 NONE .87513 6.53300 .00000 -.006172 02 TM NO3 NONE .45394 5.13330 .00000 -.027762 02 TM C2H5SO4 NONE .84589 5.61670 .00000 -.092792 02 YB CL NONE .62580 5.60000 .00000 -.014532 02 YB CLO4 NONE .88116 6.53330 .00000 -.006442 02 YB NO3 NONE .46744 5.13330 .00000 -.028122 02 YB C2H5SO4 NONE .85915 5.66400 .00000 -.090782 02 LU CL NONE .62106 5.60000 .00000 -.013562 02 LU CLO4 NONE .86883 6.53330 .00000 -.001882 02 LU C2H5SO4 NONE .86256 5.72100 .00000 -.091672 02 AL CL NONE .68627 6.02030 .00000 .008102 02 SC CL NONE .72087 6.53170 .00000 .033672 02 CR CL NONE .69081 2.78490 .00000 -.043902 02 CR NO3 NONE .72490 6.31690 .00000 -.059932 02 Y CL NONE .62570 5.60000 .00000 -.015712 02 Y C2H5SO4 NONE .85187 5.65770 .00000 -.093222 02 CE CL NONE .63509 7.49910 .00000 -.030012 02 NA ASO4 NONE .20193 5.53660 .00000 .000002 02 K ASO4 NONE .42291 9.98090 .00000 .000002 02 K FE(CN)6 NONE .34915 5.58490 .00000 -.045082 02 K CO(CN)6 NONE .36592 1.61900 .00000 -.069462 02 CO(EN)3 CL NONE .18592 3.80000 .00000 -.027832 02 CO(EN)3 NO3 NONE .10340 3.55130 .00000 .000002 02 CO(PN)3 CLO4 NONE .14854 2.95040 .00000 .000002 02 K MO(CN)8 NONE .00575 -7.47440 .00000 .010152 02 K P207 NONE .05939 -9.29390 .00000 .015912 02 K W(CN)8 NONE .38299 6.16240 .00000 -.058102 02 K ATP NONE .08619 -4.80450 .00000 .014942 02 NA ATP NONE -.04154 -6.06310 .00000 .030442 02 NA P207 NONE .06250 -11.13640 .00000 .000002 02 TH CL NONE .47146 -9.48430 .00000 -.000782 02 TH NO3 NONE .35392 -7.64530 .00000 -.018692 02 PT(PN)3 CL4 NONE .28756 10.71310 .00000 .000002 02 N(ME)4 MO(CN)8 NONE .53495 9.66070 .00000 .086202 02 MN SO4 NONE .20563 2.93620 -38.93100 .016502 02 CL F NONE -.00280 .00000 .00000 .000002 02 CL HCO3 NONE .07350 .00000 .00000 .000002 02 H CO NONE .08290 .00000 .00000 .000002 02 H NI NONE .08950 .00000 .00000 .000002 02 H CA NONE .06820 .00000 .00000 .000002 02 H MG NONE .08910 .00000 .00000 .000002 02 H UO2 NONE .13770 .00000 .00000 .000002 02 K SR NONE .01490 .00000 .00000 .000002 02 NA UO2 NONE .02310 .00000 .00000 .00000

CD NJ NJ NJ NJ NJ N NJ NJ NJ NJ NJ N NJ NJ NJ NJ NJ N NJ NJ NJ N NJ NJ NJ NJ NJ NJ NJ NJ NJ NJ NJ

CD CJ CD CD 0 CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD 0 CD 0 CD CD CD CD 0 C, CD CD 0 CDCD ui Vl Ul Cfl Ln cn ul cn Ln un CO CO CO LO CJ W CO W C0 UJ UJ CO GO CO CO Lo CO CO CO iP W N) NJCD wwoxxoxxxXnXxxnNzmwmoononnnzo----,0OW 0 >r, 0 rn

o o o r,ooxrx>>>>wii..a.woxx----ooip.p.xxw mnxxxxx zXX xnxxoxmnriorn---xxxxoonnonnooDoonownowoo xcpp.4.0mcomw.p.w00000x00000woow0w..w WX WWWWW xxxxw Ip.W p.w000Dorrrip,,P

ZZZZZZZZZZZZZZZZZZZZZZZZZZZZo000000000000000000000000000000000MMMMMMMMMMMMMMMMMMMHMM

I I ONJ I

0001-"F,01-000) P.000000000001-,0000) 1-,N001--01 NJ01 WO--lMWWCDCDNONJN0 0 000i-'NJCDA.cDWi-,ocp-ZI-,0lON0100010100010000000000000WOOP.--.1W000000004=.010000001-400000000000000--14=.00000000000000000000000000000000000

CON) OD00000100010001 10 0000000000000000N0000000P.0)0)000)000000000000000000)000000000000000000000000000000000000000000000N00000000000000000 00000000000000000000000000000000000000

0000000000000000000000000000000--1000000000000000000000000000000000N00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000

00000000000000000000000000000000 00000000000100000000000000000000I=000000000001ON0000011 0000000 0000000000000000000000000000000000000000000 0 0 0 0 0 0 0 0 0 0 0 0 0000000(D0 00 0 O0 00000 0 0

APPENDIX B Comparison of solubilities between results by Chem Sageand experimental data

7-

5

4

3

2

Figure 1. Solubility Profile for NaCl-H2O-NaOH System at25 °C

f 1 1 T 1

Exper. Data (Seidell)

--Chem Sage

00

0.09

0.08

0.07

0.06

0.05

0.04

0.03

0.02

0.01

0.00

1 2 3 4 5 6 7 8 9

NaOH Molality (Mo1/1-)

Figure 2. Solubility Profile for NaCI- CaSO4 -H20 System

10

Chem Sage

Exper. Data (Seidell)

0.0 0.5 1.0 1.5 2.0 2.5 3.0

NaC1 Molality (Mo1/1-)

95

0.400.350.300.250.200.150.100.050.00

Figure 3. Solubility Profile for H3PO4-CaSO4-H20 System

ChemSage

Exper. Data(Silcock)

0

0.0180.016

-6 0.014?-- 0.012

0.01-6 0.008

0.006O) 0.004U 0.002

0

0.2 0.4 0.6

HxPxOx Molality (Mol/L)

0.8

Figure 4. Solubility Profile for Na2SO4-CaSO4-H20 System

r

itExper. Data(Seidell)ChemSage

i

0 0.2 0.4 0.6 0.8

Na2SO4 Molality (Mol/L)

1

96

Figure 5. Solubility Profile for HNO3-CaSO4-H20 System

0.35

0.3

0.25

0.2

0.15

0 0.1

CT) 0.05

0

Data

ChemSage

Exper.(Silcock)

0

0.035

q 0.030

0.025

74 0.0200.015

0.0100ct 0.005

0.0000.00

0.5 1 1.5

HNO3 Molality (Mol/L)

Figure 6. Solubility Profile for Ca(OH)2-NaNO3-H20System

2

_a

ChemSage

Exper. Data(Silcock)

1

0.50 1.00 1.50 2.00 2.50

NaNO3 Molality (Mol/L)

3.00

97

Figure 7 Solubility Profile for NaNO2-Ca(OH)2-H20System

0.050q 0.045,-(5) 0.040.-S 0.035.,..'.' 0.0306'1 0.025

0.02014 0.0150 0.010U 0.005

0.0000.00

1.2

1.0

0.8

0.6

0.4

0.2

0.0

ChemSage

Exper. Data(Silcock)

0.50 1.00 1.50 2.00 2.50

NaNO2 Molality (Mo1(1--)

Figure 8 Solubility Profile for NaF-HF-H20 Systemat 20 °C

3.00

Exper.(Silcock)

Data

ChemSage

0.0 0.5 1.0 1.5

HF Molality (Mol/L)

2.0

98

2.50

-6 2.00Nil

1.50

0Tt

1.00O`C.i 0.50z

0.000.00

Figure 9 Solubility Profile for NaOH-Na2SO4-H20System at 25 °C

ChemSage

Exper. Data(Seidell)

5.00 10.00

NaOH Molality (Mol/L)

Figure 10 Solubility Profile for NaCl-CaCO3 -H20 Systemat 25 °C

1.20E-03

1.00E -036'

8.00E-04

-c-d 6.00E -04

4.00E-0402.00E-04

U0.00E+00

0.00

ChemSage

Exper. Data(Silcock)Exper. Data(Perry's)

I I

0.50 1.00 1.50 2.00 2.50

NaC1 Molality (Mol/L)

3.00

99

0

0U

0.00E+000.00

Figure 11 Solubility Profile for Na2SO4- CaCO3 -H20System at 25 °C

4.50E-034.00E-033.50E-033.00E-032.50E-032.00E-031.50E-031.00E-035.00E-04

1.00E+00

1.00E-01

0.)

DP.00E 02

21.00E-03

1.00E-04

0.00

-ChernSageExper. Data (Silcock)Exper. Data (Perry's)

0.50 1.00 1.50

Na2SO4 Molality (Mol/L)

Figure 12 Solubility Profile for HC1-CaF2-H20 Systemat 20 °C

2.00

ChemSage 3.0 IIExper. Data (Silcock)iixper. Data (Perry's)

IIMME11111111111

0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80

HC1 Molality (Mol/L)0.90 1.00

100

0.208 0.18rt, 0.16

0.140.12

0.10L5' 0.085 0 06E:9. 0.04gi 0.02

0.000.00

Figure 13 Effect of NaC1O2 on System(Mass Balance Concentration = 0.00 Mol/L)

a

a

Aa

A.

DDDDDDDDDDDODDMODDO DDOW 401/1000 Ca(OH)20 C1( -a)

NO3(-a)+ OH(-a)

- `" CaF2 (s)6 (s) -8- Ca(+2a) (aq)

(aq) £ C104(-a) (aq)(aq) Na(+a) (aq)

(aq)aa

a.. .. .. .. .. .. .. ..

I =I.. ..

0.00100.0009

t 0.0008

dgC5

0.000700.00060.0005

g 0.0004ea 0.0003cr 0.0002

0.00010.0000

0.00

0.10 0.20 0.30 0.40

Incoming Conc. NaC1O2 (Mol/L)

Figure 14 Effect of NaC1O2 on System(Fluoride Limit = 0.0009 Mol/L)

0.50

- F(-a) (aq)

0.10 0.20 0.30 0.40 0.50

Incoming Conc. NaC1O2 (Mol/L)

101

8

O

0U

LIJ

Figure 15 Effect of Na2SO4 on System(Mass Balance Concentration = 0 Mol/L)

(s)

pHCaF2 (s )Ca(OH)2 (s)

X CAS04 (s)o CaS 04* 2H20

0.40

0.35Ca(+2a) (aq )NO3(-a) (aq)Na(+a) (aq)NaSO4( -a) (aq)OH(-a) (aq)s04-2a) (aq)

0.30

0.25

0.20 tiI

0.15 A

0.10

oit 6 oogsaW00000 00000ill It -A- it *+ + + +

0.05 0ill rit!

0.00 0.10

0.00100.0009

0 0.00080.0007

cc; 0.00065 0.0005§ 0.0004

0.00030.00020.00010.0000

0.00

0.20 0.30 0.40 0.50

Incoming Conc. Na2SO4 (Mol/L)

Figure 16 Effect of Na2SO4 on System(Fluoride Limit = 0.0009 Mol/L)

- F(-a) (aq)

- °

0.10 0.20 0.30 0.40 0.50

Incoming Conc. Na2SO4 (Mol/L)

102

0.358

0.30

0.25

0.20

o 0.15

.2 0.10

0.05

0.00

Figure 17 Effect of NaNO3 on System(Mass Balance Concentration = 0 Mol/L)

-CaF2pH/100 - - - - (s)X Ca(OH)2 (s ) Ca(+2a) (aq)- NO3(-a) (aq) 0 Na(+a) (aq)p OH(-a) (aq)

III

II 11

0o0

Eli0

0

i 60

"W"lic- -a- i" IC lc" 'it at si "i ")0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

0.00100.00090.00080.00070.00060.00050.00040.00030.00020.00010.0000

0.00

Incoming Conc. NaNO3 (Mol/L)

Figure 18 Effect of NaNO3 on System(Fluoride Limit = 0.0009 Mol/L)

F(-a) (aq)

0.05 0.10

Incoming Conc. NaNO3 (Mol/L)

103

0.408'i 0.35

0.30

1 0.25

rj 0.208

C.) 0.155

.p., 0.10:51 0.05crw

0.000.00

Figure 19 Effect of H202 on System(Mass Balance Concentration = 0 Mol/L)

1 I

A0

pH/100Ca02 (s)H202 (s)NO3(-a) (aq)

- - CaF2 (s)Ca(OH)2 (s)

AA

0 Ca(+2a) (aq)

X OW-a) (aq) AA

AA00000000000000000A

A

#*

-*.ak

* A .6 ..4.X

0.00100.00090.00080.00070.00060.00050.00040.00030.00020.00010.0000

0.00

0.10 0.20 0.30

Incoming Conc. H202 (Mol/L)

Figure 20 Effect of H202 on System(Fluoride Limit = 0.0009 Mol/L)

F(-a) (aq)

. . = . .

0.10 0.20 0.30

Incoming Conc. H202 (Mol/L)

0.40

0.40

104

8 0.18E 0.16

0.140.120.10

C.;g 0.080U= 0.06

0.040.02

ral 0.000.08

Figure 21 Effect of Ca(OH)2 on System

CatitillaiiitlAAAAAALtAALiLi4AAAAAA4AiiiituitiiiiitiAAAA4pH/100 - um CaF2 (s)

+ Ca(OH)2 (s) -El- Ca(+2a) (aq)0 H(+a)P NO3(-a)

(aq)(aq)

OH(-a) (aq)

+++

++++++++morm=====....rturtiiimr.mr=mrww NEWINIumulwwwww onowli

__++++.74.4;iiiial:+eeeeeeeeeeeeeeeeeeeeeeeeeeeec

0.0014

0.0012

0.0010

0.0008Ug 0.0006

0.0004

0.0002

0.10 0.12 0.14 0.16 0.18 0.20

Incoming Conc. of Ca(OH)2 (Mol/L)

Figure 22 Effect of Ca(OH)2 on System(Flouride Limit = 0.0009 Mol/L)

313131:1130 D

0.00000.08 0.10 0.12 0.14 0.16 0.18

Incoming Conc. Ca(OH)2 (Mol/L)

0.20

105

8 0.25

0.20

0.15

U 0.10

.., 0.05

crw 0.00

Figure 23 Effect of HNO3 on System( Mass Balance Concentration = 0.1630 Mol/L)

pH/100o Ca(OH)2 (s)X H(+a) (aq)

OH(-a) (aq)

CaF2 (s)"BOMISMACa(4.2a)

0 NO3(-a) (aq)

000000

0

0

or,---4+mt-Jr+19-41-04-41"Trig cr. 7ixxxxxxxxxxxxxxxxxxxxx001001*-54..0

0.08 0.13 0.18

0.0014

0.0012

0.0010

0.0008U§ 0.0006

a 0.00040.0002

0.00000.08

Incoming Conc. HNO3 (Mol/L)

Figure 24 Effect of HNO3 on System

0.23

AlF3 (s)

'AIF( +2a) (aq)

F(-a) (aq)

`Total F

0.13 0.18

Incoming Conc. HNO3 (Mol/L)

0.23

106

0.30

80.25

a.

0.20

j 0.150

E 0.10

0, 0.05

Figure 25 Effect of Added Amounts of HNO3 andCa(OH)2 in pH Constant System

pH/100 CaF2 (s)

0 Ca(OH)2 (s) 'D'ICa(+2a) (aq)A NO3 (-a) (aq) X OH(-a) (aq)

wW W

0.2600.200 0.220 0.240

Incoming Conc. HNO3 (Mol/L)

Figure 26 Effect of Added Amounts of HNO3 and Ca(OH)2in pH Constant System

0.00100.00090.0008

s?.., 0.0007g 0.0006

c.5 0.0005g 0.0004an 0.0003.g 0.000241 0.0001

0.00000.180

F(-a) (aq)

0.200 0.220 0.240

Incoming Conc. HNO3 (Mol/L)

0.260

107

108

Figure 27 Effect of HF on System( Mass Balance Concentration = 0.03573 Mol/L)

1113111111:1111111111111:111

MIL

0.18

8 0.16t 0.14

g 0.120.10

'co, 0.08UE 0.06

x20.04

0.02

0.000.00

0.0014

0.0012

0.0010

g 0.0008

g 0.0006

0.0004

0.0002

pH/100

o Ca(OH)2 (s)

X II(+a) (aq)NO3(-0 (aq)

CaF2 (s)0.40g4,waggsaata(+2a)

OH(-a) (aq)

MAW

OM=0.02 0.04 0.06 0.08

Incoming Conc. HF (Mol/L)

Figure 28 Effect of HF on System

0.10

0.00000.00 0.02 0.04 0.06 0.08 0.10

Incoming Conc. HF (Mol/L)

Figure 29 Effect of Added Amounts of HF and Ca(OH)2 inpH Constant System

0.180 0 0 0 0 0 0 D 0 0 0 0 0 0 0

8 0.160.14

0.12

0.10

g 0.08UE

0.06

0.04

a 0.020.00

0.04 0.06 0.08 0.10

Incoming Conc. HF (Mol/L)

0000000 D 0000000 C

pH/100 CaF2 (s) Ca(OH)2 (s)

Ca(+2a) 0 NO3(-a) (a9) A 01A-a) (acp

OMMWEM.MtaW,, mgemoo.onwouom.,,01,14WONtOg sMftgfaliBMIANUOVAISW4461,4o3oamintimmageMi4Mee

mmmmmmmmmmmmmm

mmmmm mmmmmmmmm

AAAAAAAAAAA

0.00100.00090.0008

g 0.000706 0.0006C5 0.0005§ 0.0004

0.0003.cr5 0.0002

0.00010.0000

0.04

0.12

Figure 30 Effect of Added Amounts of HF and Ca(OH)2in pH Constant System

- - - F(-a) (aq)

0.06 0.08 0.10 0.12

Incoming Conc. HF (Mol/L)

109

0.18

8 0.16sk 0.14

g 0.120.10

g 0.08§ 0.06.Zr9.. 0.04

0.02

0.000.00

Figure 31 Effect of H2CO3 on System(Mass Balance Concentration = 0.00 Mol/L)

I

pH/100 -X CaCO3 (s) 0

Ca(+2a) (aq) CIHCO3(-a) (aq)

0 OH(-a) (aq)

CaF2 (s)Ca(OH)2 (s)H(+a) (aq)NO3(-a) (aq)

===Z == ; X X ik_"p m

x 1(1.

m m .. ... ....

0 Ca X

0.0014

0.0012

0.001

g 0.0008U§ 0.0006

0.0004

14 0.0002

00.00 0.02 0.04 0.06 0.08

Incoming Conc H2CO3 (Mol/L)

0.02 0.04 0.06 0.08 0.10

Incoming Conc H2CO3 (Mol/L)

Figure 32 Effect of H2CO3 on System

AIF3 (s)AlF(+2a) (aq)

- F(-a) (aq)F

0.10

110

0.208. 0.18

0.160.140.12

c., .0 10=U 0.08E 0.062Al 0.04=0-, 0.02

ral0.00

0.00

Figure 33 Effect of Added Amounts of H2CO3 andCa(OH)2 in pH Constant System

IiI

/...

9-----4 9 9 9 9 9 9 9 13-Ef/ '" .' '" CaF2 (s)

0 Ca(OH)2 (s)NO3(-a) (aq)

pH/100CaCO3 (s)/ -X

-El- Ca(+2a) (aq)A OH(-a) (aq)

1- -cr - -o- - -* - - - - - - -o- - - -*A A

A A A A

0.0010.00090.00080.0007

(a) 0.00060(-) 0.0005= 0.0004g 0.00035cr 0.0002

L11

0.00010

0.00

0.20 0.40 0.60 0.80

Incoming Conc H2CO3 (Mol/L)

Figure 34 Effect of H2CO3 and Ca(OH)2 inpH Constant System

i

1.00

4-

F(-a) (aq)

0.20 0.40 0.60 0.80 1.00

Incoming Conc H2CO3 (Mol/L)

111

0.208 0.18

0.16

0.140.12

;.; 0.10L5 0.08

0.06

0.040.02

0.000.00

Figure 35 Effect of HC1 on System(Mass Balance Concentration = 0 Mol/L)

----- ----

#0 Ca(OH)2 (s)

CaF2 (s)

IpH/100 asimCCa(+2a)

(aq)+ Cl(-a) (aq) X H(+a) (aq)

NO3(-a) aq) 0 OH(-a) (aq)

..h.

-----++

V * iIQ 0 0 0 0 0 0I, 2

*1X

+ xx

= = MS = x= le a - ---

0.0014

0.0012

0.0010

0.00080

0.0006

0.0004L.1J

0.0002

0.00000.00

0.05 0.10 0.15

Incoming Conc. HC1 (Mol/L)

Figure 36 Effect of HC1 on System

0.20

'A1F3 (s)AIF( +2a) (aq)

F(-a) (aq)sm°°...ITotal F

0.05 0.10 0.15

Incoming Conc. HCl (Mol/L)

0.20

112

Figure 37 Effect of Added Amounts of HC1 and Ca(OH)2 inpH Constant System

0.208 0.18

0.16

0.14° 0.126 0.10eL5' 0.08

0.06r2 0.041 0.02

0.000.00 0.05 0.10 0.15

Incoming Conc. HC1 (Mol/L)

Figure 38 Effect of Added Amounts of HC1 and Ca(OH)2in pH Constant System

0.001

75 0.0008

0 0.0006UJ 0.0004

H 0.0002

00.00 0.05 0.10 0.15

Incoming Conc. HCI (Mol/L)

- - F(-a) (aq)

0.20

0.20

113

0.18

E 0.16fin. 0.14

0.12-6

0.10

g 0.08

E 0.060.04

Lg 0.02

0.000.00

Figure 39 Effect of H2SO4 on System(Mass Balance Concentration = 0 Mol/L)I

pH/I 00 - - CaF2 (s)

X Ca(OH)2 (s) CaS0.4 (s)+ CaSO4*2H20 (s) Ca(+2a) (aq)

11(+a) (aq) A FISC4( -W4IIII NO3(-a) (aq) 0 OH(-a) (aq)0 SO4(-2a) (aq)

0 0 0 0 0

MI II+A 4- 4.- "nunr10.1111WWW1

4. 4- + +mi+= =to = = Ns

'I 51Q 0

0.0014

0.0012

0.0010

0.0008

0.0006

0.0004 -

0.0002

0.00000.00

0.02 0.04 0.06

Incoming Conc. (Mol/L)

Figure 40 Effect of H2SO4 on System

0.08

AlF3(s)

AJF( +2a)(aq)

F( -a) (aq)

"m1:1..""Total F

0.02 0.04

Incoming Conc. (Mol/L)

0.06 0.08

114

0.18O-.9 0.16

sa. 0.14

0.12

0.10

g 0.08

§ 0.06

.52, 0.04

23-0.02

0.000.00

Figure 41 Effect of Added Amounts of H2SO4 andCa(OH)2 in pH Constant System

Illi

i

-.--

pH/100 - CaF2 (s), Ca(OH)2 (s) :1( CaSO4 (s)0 CaSO4*2H20 (s) fOmmca(+2a)13 OH(-a) (aq) NO3(-a) (aq) .

- -- .' . X * X

*., 13 13 0 0 0 13

0.00 100.0009

c) 0.0008S 0.0007g 0.0006

c.5 0.0005§ 0.0004E 0.0003t. 0.0002

0.00010.0000

0.00

0.02 0.04 0.06 0.08

Incoming Conc. H2504 (Mol/L)

Figure 42 Effect of Added Amounts of H2SO4 andCa(OH)2 in pH Constant System

0.10

- - F(-a) (aq)

MI 11. MI IN 11Ma

0.02 0.04 0.06 0.08

Incoming Conc. H2SO4 (Mo1/1--)

0.10

115

116

APPENDIX C Experimental data for equilibrium study

Table C.1 The result of experiments between HF , lime, and different concentrationof added Na2CO3 in a 1 L batch reactor at 25 °C

Run # 1 2 3 4 5 6 7 8

HF (mol) 0.0083 0.0092 0.0089 0.0095 0.009 0.01 0.0095 0.009

Lime 0.0083 0.018 0.018 0.019 0.018 0.02 0.019 0.018

Na2CO3 0 0 0.0089 0.0048 0.0009 0.001 0.0019 0.0018

Time(min) Fluoride Concentration(PPIn)

0 180 184 180 190 180 200 190 180

1 67 4 128 33 11 10 6 9

2

3 66 2.5 11 5

4 8

5 62 40 31 11 5

6 8 5

7

8

9 8

10 30 31 8 4 5

15 60 2.3 30 15 8 4

20 30 14

25 60 14 5

30 30 8 5

35 60 5

40 4

45 13 5

117

APPENDIX D The rate data for kinetic study

Table D.1 Experimental data between HF and lime at different ratios in the well-mixedbatch reactor at 25 °C

Run # 1 2 3 4 5

HF:Lime

(mol Ratio)

1:1.3 1:13 1:1 1:1 1:2

Initial HF 0.0104 0.0055 0.0083 0.0076 0.0089

Initial Lime 0.0135 0.007 0.0083 0.0076 0.018

Time Fluoride Concentration (ppm)0 208 110 170 160 184

1 19.2 9 50 4

2 18 66 50

3 16 50 3

4 62

5 7 50

6 16

7 14

8

9

10 6 50

12

15 3

18

20 14 6 60

22

25 6 60

30 6 50

35 60

40

45 14