Title: Effect of particle size on the destabilization of colloidal suspensions in water
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00098224/00001
 Material Information
Title: Effect of particle size on the destabilization of colloidal suspensions in water
Physical Description: xxiii, 216 l. : illus. ; 28 cm.
Language: English
Creator: Vilaret, Manuel R., 1919-
Publisher: s.n.
Place of Publication: Gainesville
Publication Date: 1965
Copyright Date: 1965
 Subjects
Subject: Colloids   ( lcsh )
Polymers and polymerization   ( lcsh )
Civil Engineering thesis Ph. D
Dissertations, Academic -- Civil Engineering -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis - University of Florida.
Bibliography: Bibliography: l. 212-215.
General Note: Manuscript copy.
General Note: Vita.
 Record Information
Bibliographic ID: UF00098224
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000423956
oclc - 11041289
notis - ACH2361

Downloads

This item has the following downloads:

effectofparticle00vila ( PDF )


Full Text










EFFECT OF PARTICLE SIZE ON THE
DESTABILIZATION OF COLLOIDAL
SUSPENSIONS IN WATER
















By
MANUEL R. VILARET









A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY









UNIVERSITY OF FLORIDA
August, 1965


























66-2052

VILARET, Manuel R., 1919-
EFFECT OF PARTICLE SIZE ON THE DE-
STABILIZATION OF COLLOIDAL SUSPENSIONS
IN WATER.

The University of Florida, Ph.D., 1965
Engineering, civil



University Microfilms, Inc., Ann Arbor, Michigan
















ACKNOWLEDGMENTS


The writer is indebted, first and above all, to his Committee

Chairman, Dr. A. P. Black, not only for his guidance and dedication

but for the opportunity to complete his work for the degree. He is

also deeply indebted to Dr. J. J. Morgan for the many helpful sugges-

tions and continued orientation during the course of the research work.

Special recognition is also due to Professor John E. Kiker, Jr. for

his friendship and advice during the entire period of graduate study.

A very special note of thanks goes to Dr. Frank B. Birkner for

his understanding, his suggestions, and for the time spent in fruitful

discussions.

The writer wishes to thank Mrs. Annie L. Smith for instructing

him in microelectrophoresis techniques and for her assistance in the

laboratory. Thanks are also due to Mrs. Janice 0. Larson and to Mrs.

Frances M. Kost for their devoted efforts in typing the manuscript.

The author wants to express his recognition to his wife and

children for their patience over the past several years.

He also wishes to extend his appreciation to the Dow Chemical

Company for supplying the latex samples used in this investigation,

and to the United States Public Health Service, whose financial support

made this study possible.
















CONTENTS

Page

ACKNOWLEDGMENTS . . . . . . . . . . ii

LIST OF TABLES . . . . . . . . . . ....

LIST OF FIGURES . . . . .. . . . . .... xiii

ABSTRACT. ................. . . . xx

CHAPTER

I. INTRODUCTION . . ... . . . . . . . 1

II. THEORETICAL CONSIDERATIONS . . . . . 4

Stability of Colloids Destabilization . . 4
Coagulation in a Static Dispersing Medium . . 4
Gradient Coagulation . . . . . . . 5
Coagulation in Turbulent Flow . . . ... 7
Flocculation by Polymers . . . ...... 8
General Predictions . . . . . .... 10

III-. EXPERIMENTAL MATERIALS AND PROCEDURES . . .. 12

Suspending Medium . . . . . . . .. 12
Coagulants and Flocculants .. . . . 12
Preparation of Suspensions . . . . .. 13
Jar Test Procedure . . . . . . .. 14
Turbidity Measurements . . . . . . 15
Electrophoretic Mobilities . . . . . 16
pH Adjustment . . . . . . . . . 17
Sample Numbering . . . . . . . .. 17

IV. EXPERIMENTS WITH LATEX SUSPENSIONS . . . . 19

General ........... . . .... 19
Variation of Turbidity with Particle Size . .. 21
Destabilization with Cationic Polymer . . 23
Polydisperse Suspensions . . . . . 27
Observations on Particle Growth . . . . 32
Destabilization with Calcium Chloride . . . 33
Destabilization with Nonionic Polymer . . 35
Destabilization of Latexes with Alum at pH 7.5 37
Effect of Particle Size and Mixing Times and
Rates . . .. ... . 37
Observations on Times for First Visible Floe . 41












Page
V. EXPERIMENTS WITH SILICA SUSPENSIONS . . 45

General ................... 45
Experimental Materials ............. 45
Destabilization with Cationic Polymer . . 47
Destabilization with Alum at pH 4.0 . . . 49
Destabilization with Alum in pH Range 4.6-5.4 50
Experimental . . . . . . 51

VI. EXPERIMENTS WITH KAOLINITE CLAYS . . . . 52

General . . . . . . 52
Previous Studies ... . . . 52
Experimental Materials . . . ....... 53
Destabilization with Cationic Polymer . . 55
Times for First Visible Floe .. . ..... 57

VII. EXPERIMENTS WITH OTHER MATERIALS . . . 58

General .................. 58
Carbon Black Suspensions . . . . .. 58
Experimental . . . . . . . 60
Ion Exchange Resin . . . . . . 62
Colored Water . . . . . . . . 63

VIII. DISCUSSION AND CONCLUSIONS . . . . 65

Discussion .................. 65
Conclusions . . . . . . . . . 69

APPENDICES

A. Figures 21 to 73 . . .. . . . . . 71

B. Tabulated Data .. . . .... . . . . 125

C. Derivation of a Formula for the Location of the
Stationary Layer in a Cylindrical Microelectro-
phoresis Cell . . . .. . . . . 209

LIST OF REFERENCES .................. . .. 211

BIOGRAPHICAL SKETCH .............. . .... 216















LIST OF TABLES


Table Page

1. Characteristics of the Monodisperse Latex Stock
Suspensions . . . . . . . . .... 20

2. Physical Characteristics of the Four Silicas Investigated 47

3. Characteristics of Carbon Black Samples .... ..... . 60

4. Optical Densities of Monodisperse Latex Suspensions at
Various Concentrations. Light Path = 75 mn. Wave Length
of Filter = 650 a . . . . . . .. . 126

5. Optical Densities of Monodisperse Latex Suspensions at
Various Concentrations. Light Path = 37.5 mm. Wave Length
of Filter = 650 mu ... ................. 127

6. Optical Densities of Monodisperse Latex Suspensions at
Various Concentrations. Light Path = 75 mm. Wave Length
of Filter = 420 m .. . . . . . . . .. 128

7. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 88 mi, Suspension Concentration
5 g/ . . . . .. . . . . . . . 129
8. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 88 mP, Suspension Concentra-
tion 10 g/1 ........ ......... .... 130

9. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 88 mp, Suspension Concentra-
tion 20 g/l ... .... .... . . . 132

10. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 88 ma, Suspension Concentra-
tion 25 mg/l . . . . . . . . .... 133

11. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 88 mu, Suspension Concentra-
tion 50 mg/ ................ ... . 134












Table rage

12. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 88 mP, Suspension Concen-
tration 100 mg/1 . . . . . . . .... .. 136

13. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 88 mp, Suspension Concentra-
tion 200 mg/1 . . . . . .. .. . . . 138

14. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 88 lm, Suspension Concentra-
tion 500 mg/l . . . . . . . .. . 139

15. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 126 mu, Suspension Concen-
tration 50 mg/ . . . . . . . . ... ... 140

16. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 126 mU, Suspension Concen-
tration 100 mg/l ..................... 141

17. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 126 am, Suspension Concentra-
tion 200 mg/ . . . . . . . . . . . 143

18. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 264 mP, Suspension Concentra-
tion 25 mg/1 .............. e .... ... 144

19. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 264 mP, Suspension Concentra-
tion 50 mg/1 ...... ................ 45

20. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sue-
pension. Particle Diameter 264 nm, Suspension Concentra-
tion 100 g/ . . . . . . . . . . 146

21. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Sus-
pension. Particle Diameter 264 mU, Suspension Concentra-
tion 200 mg/ . .. . . . . . . . . 148











Table


22. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 264 ml, Suspension Con-
centration 500 mg/1 . . .... .. . . . .. 149

23. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 365 mI, Suspension Con-
centration 100 mg/l . . . . . . . ... 150

24. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 557 mi, Suspension Con-
centration 100 mg/ . . . . . . . . . 151

25. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 557 mI, Suspension Con-
centration 200 mg/l . . . . . . . ... 152

26. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 1305 mi, Suspension Con-
centration 25 mg/l . . . . . . . . . 153

27. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse PolystyreneLatex
Suspension. Particle Diameter 1305 nL, Suspension Con-
centration 50 mg/l . . . . . . . . . 154

28. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 1305 mrp, Suspension Con-
centration 100 mg/l . . . . . . . . . 155

29. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 1305 m, Suspension Con-
centration 200 mg/l . . . . . . . ... 156

30. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polyvinyltoluene Latex
Suspension. Particle Diameter 3490 mP, Suspension Con-
centration 25 mg/1 . .. . . . . . . 157

31. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polyvinyltoluene Latex
Suspension. Particle Diameter 3490 mri, Suspension Con-
centration 50 mg/l . .... . . . . . 158


vii











Table Page

32. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polyvinyltoluene Latex
Suspension. Particle Diameter 3490 mu, Suspension Con-
centration 100 mg/ . . . . . . . . . .. 159

33. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polyvinyltoluene Latex
Suspension. Particle Diameter 3490 mI, Suspension Con-
centration 200 mg/l ... . . . . . . . 160

34. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Polydisperse Polystyrene Latex
Suspension. Particle Diameter 88 mT, Concentration
10 mg/1 and Particle Diameter 264 up, Concentration
25 mg/1 . . . . . . . . ... ....... 161

35. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Polydisperse Polystyrene Latex
Suspension. Particle Diameter 88 mu, Concentration
20 mg/l and Particle Diameter 264 mi, Concentration
50 mg/1. .......... . . . . . . 162

36. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Polydisperse Polystyrene Latex
Suspension. Particle Diameter 88 w, Concentration
10 mg/l; Particle Diameter 264 up, Concentration 25 mg/l;
and Particle Diameter 1305 mP, Concentration 100 mg/1 . 163

37. Effect of Various Dosages of Calcium Chloride on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 88 ml, Suspension Con-
centration 50 mg/ . . . . . . . . . . 164

38. Effect.of Various Dosages of Calcium Chloride on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 88 am, Suspension Con-
centration 100 mg/1 ........ . . . . 165

39. Effect of Various Dosages of Calcium Chloride on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 264 mu, Suspension Con-
centration 100 mg/l . . . . . . . . . .. 166

40. Effect of Various Dosages of Calcium Chloride on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 557 uA, Suspension Con-
centrat on 100 mg/l . ... . . . . . . . 167









Table rage

41. Effect of Various Dosages of Nonionic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 88 mo, Suspension
Concentration 50 mg/ . . . . . . . .... 168

42. Effect of Various Dosages of Nonionic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 88 mg, Suspension
Concentration 100 mg/l . . . . . . . . .. 169

43. Effect of Various Dosages of Nonionic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 264 mu, Suspension
Concentration 100 mg/ . . . . . . . . .. 170

44. Effect of Various Dosages of Nonionic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex
Suspension. Particle Diameter 557 mp, Suspension
Concentration 100 mg/l . . . . . . . . .. 171

45. Effect of Times of Mixing on the Destabilization of an
88 my Dia., 100 mg/l Latex Suspension with Cationic
Polymer (Rate of Mixing 20 rpm). ........... . 172

46. Effect of Times of Mixing on the Destabilization of an
88 mu Dia., 100 mg/1 Latex Suspension with Cationic
Polymer (Rate of Mixing 100 rpm).. . . . .. ..... 173

47. Effect of Times of Mixing on the Destabilization of a
264 mr Dia., 100 mg/1 Latex Suspension with Cationic
Polymer (Rate of Mixing 20 rpm). . . . . . ... 174

48. Effect of Times of Mixing on the Destabilization of a
264 mu Dia., 100 mg/l Latex Suspension with Cationic
Polymer (Rate of Mixing100 rpm) . . . . . . 175

49. Effect of Times of Mixing on the Destabilization of a
1305 mA Dia., 100 mg/l Latex Suspension with Cationic
Polymer (Rate of Mixing 20 rpm) . . . . .. . . 176

50. Effect of Time of Mixing on the Destabilization of a
1305 mu Dia., 100 mg/1 Latex Suspension with Cationic
Polymer (Rate of Mixing 100 rpm). . . . . .... 177

51. Physical Properties of Monodisperse Polystyrene Latexes,
and Required Unit Dosages of Cationic Polymer for Optimum
Destabilization of the Suspensions . . . . .... 178











Table


52. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Silica Suspension. Min-u-Sil "53",
Average Particle Size 1.1P, Suspension Concentration
100 mg/ . . . . . . . . ... . . . . 179

53. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Silica Suspension. Min-u-Sil "51",
Average Particle Size 1.14, Suspension Concentration
200 mg/l . . . . . . . . ..... . 180

54. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Silica Suspension. Min-u-Sil "5n",
Average Particle Size 1.1i, Suspension Concentration
300 g/1 . . . . . . . . . . . . 181

55. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Silica Suspension. UCAR-Submicron,
Average Particle Size 120 m, Suspension Concentration
100 mg/1 . . . . . . . . ... ....... .. 182

56. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Silica Suspension. UCAR-Submicron,
Average Particle Size 120 mp, Suspension Concentration
200 mg/1 . . . . . . . . . . 184

57. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Silica Suspension. UCAR-Submicron,
Average Particle Size 120 ml, Suspension Concentration
300 mg/1 . . . . . . . . . . 185

58. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Silica Suspension. Cab-0-Sil M-5,
Average Particle Size 15 mW, Suspension Concentration.
100 mg/l . . . . . . . . . . . 186

59. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Silica Suspension. Cab-0-Sil M-5,
Average Particle Size 15 mA, Suspension Concentration
200 mg/l . . . . .. . . . . 187

60. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Silica Suspension. Cab-0-Sil M-5,
Average Particle Size 15 mi, Suspension Concentration
300 ag/1 . . . . . .i . . . . 188

61. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH = 4.0 Approximately.
Min-u-Sil "51", Average Particle Size 1.1C, Suspension
Concentration 200 mg/1 . . . . . . . .... 189













62. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH = 4.0 Approximately.
Min-u-Sil "5", Average Particle Size 1.1l, Suspension
Concentration 400 mg/1 . .... .. . . . . . 191

63. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH = 4.0 Approximately.
UCAR-Submicron, Average Particle Size 0.12P, Suspension
Concentration 200 mg/1 .. . . . . . . . 192

64. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH = 4.0 Approximately.
UCAR-Submicron, Average Particle Size 0.12p, Suspension
Concentration 400 mg/1 . . . . . . . ... 193

65. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH Range 4.7 5.4.
Min-u-Sil "SM", Average Particle Size 1.12, Suspension
Concentration 200 mg/l . . . . . . . . 194

66. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH Range 4.7 5.4.
Min-u-Sil "5M", Average Particle Size 1.19, Suspension
Concentration 400 mg/l. . . . . . . . . 195

67. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH Range 4.6 5.2.. UCAR-
Submicron, Average Particle Size 0.12P, Suspension Con-
centration 200 ma/1 . . . . . ... 196

68. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH Range 4.6 5.2. UCAR-
Submicron, Average Particle Size 0.12P, Suspension Con-
centration 400 mg/ . . . . . . . . . .. 197

69. Effect of Various Dosages of Cationio Polymer on the
Destabilization of aKaolinite Clay Suspension. Particle
Information Service Sample 30-4, Suspension Concentra-
tion 25 mg/l ........................198

70. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-4, Suspension Concentra-
tion 50 mg/ . . . . . . . . . . . 199


Table


Page













71. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-4, Suspension Concentra-
tion 100 mg/1 ..................... .201

72. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-4, Suspension Concentra-
tion 150 mg/l . . . . . . . . .. 202

73. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-4, Suspension Concentra-
tion 200 mg/ . . . . . . . . ... . .. 203

74. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-5, Suspension Concentra-
tion 25 mg/ . . . . . . . . . ... 20

75. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-5, Suspension Concentra-
tion 50 mg/l . . . . . . . . . . .. 205

76. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-5, Suspension Concentra-
tion 100 g/1 . . . . . . .. . . ... 206

77. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-5, Suspension Concentra-
tion 150 g/l . . . . . . . . . . . 207

78. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-5, Suspension Concentra-
tion 200 mg/1 . . . . . . . .... ..... 208


Table


Page














LIST OF FIGURES
Figures Page

1. Projection on the XZ Plane of the Sphere of Influence
Assumed to Surround Particle at 0 .. .. .... 6

2. Effect of Particle Diameter upon the Optical Density of
Monodisperse Latex Suspensions at Various Weight Concen-
trations . . . . . . . . . 22

3. Dosages of Cationic Polymer Required for Maximum Deatabili-
zation of Three Monodisperse Polystyrene Latex Suspensions
at Various Concentrations . . .. . . . . . 25

4. Effect of Particle Diameter upon the Optimum Cationic
Polymer Dosage Required for Maximum Destabilization of
Monodisperse Latex Suspensions of Given Weight Concen-
trations Variation of System Area with Particle Size . 26

5. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Polydisperse Polystyrene Latex Sus-
pension Comprising To Discrete Particle Sizes, as Compared
to the Effect on Each of the Monodisperse Components Taken
Separately (88 np Particle Diameter at 10 mg/1. and 264 iw
Particle Diameter at 25 mg/l.) . . . . . . . 28

6. Effect of Various Dosages of Cationio Polymer on the
Destabilization of a Polydisperse Polystyrene Latex Sus-
pension Comprising To Discrete Particle Sizes, as Compared
to the Effect on Each of the Monodisperse Components Taken
Separately (88 mp Particle Diameter at 20 mg/1. and 264 ms
Particle Diameter at 50 mg/l.). . . . . . . 29

7. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Polydisperse Polystyrene Latex Sus-
pension Comprising Three Discrete Particle Sizes, as Compared
to the Effect on Each of the Moncdisperse Components Taken
Separately. . . . . .. . ....... .. 30
8. Effect of Various Dosages of Cationic Polymer on the Electro-
phoretic Mohilities of the Three Suspensions Compared for
Residual Turbidity in Figure 5 ............ * 31


xlii











Figures Page

9. Effect of Various Dosages of Cationic Polymer on the
Electrophoretic Mobilities of the Three Suspensions
Compared for Residual Turbidity in Figure 6 .. . .. 31

10. Effect of Various Dosages of Cationic Polymer on the
Electrophoretic Mobilities of the Four Suspensions
Compared for Residual Turbidity in Figure 7 ....... 31

11. Effect of Calcium Chloride upon the Stability of Mono-
disperse Latex Suspensions of Various Concentrations
and Particle Sizes ... .. .......... 34

12. Effect of Dosages of Nonionic Polymer on the Destabili-
zation of Four Monodisperse Polystyrene Latex Suspensions
of Different Diameters and Concentrations, in the Presence
of 2.0 g/l Calcium Chloride .... ........... 36

13. Effect of Various Times of Mixing, Expressed as Total
Number of Paddle Revolutions, on the Destabilization of
Three Monodisperse Polystyrene Latex Suspensions with
Cationic Polymer. (Mixing Rate 20 rpm). . . . . 39

14. Effect of Various Times of Mixing, Expressed as Total
Number of Paddle Revolutions, on the Destabilization of
Three Monodisperse Polystyrene Latex Suspensions with
Cationic Polymer. (Mixing Rate 100 rpm). .. . . . 40

15. Effect of Particle Concentration (Expressed as Weight
Concentration) on the Approximate Times of Mixing Required
for the Appearance of the First Visible Floe for Three Mono-
disperse Latex Suspensions, at Optimum Dosages of Cationic
Polymer (Mixing Rate 100 rpm). . . . . .. 42
16. Plot of Surface Areas of Several Latex Suspensions Versus
Approximate Times of Mixing Required for the Appearance of
the First Visible Floc (Mixing Rate 100 rpm). . .... 43

17. Dosages of Cationic Polymer Required for Maximum Destabili-
zation of Various Silica Suspensions .. ... . .. 48

18. Particle Size Distribution of the Two Kaolinite Clay
Samples Investigated . . . .. ........ 54

19. Dosages of Cationic Polymer Required for Maximum Destabili-
zation of Two Kaolinite Clay Suspensions with Different
Particle Size Distributions . .... . . . ... 56











Figures Page

20. Particle Size Distributions of the Three Carbon Samples
Investigated .... . . . . .. . * *.. 59

21. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 88 mp, Suspension Concentration
5 mg/1 . . . . . . . . . . . . . 72
22. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 88 mrp, Suspension Concentration
10 mg/1 . . . . . . . . . . . . . 73

23. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 88 mp, Suspension Concentration
20 mg/1. ...... ....... .... . .. . . . . . 74

24. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 88 ma, Suspension Concentration
25 mg/. . . . . . . . . . . . .. 75

25. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 88 mu, Suspension Concentration
50 mg/1. . . . . . . . . . . .. . ... 76
26. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 88 mg, Suspension Concentration
100 ag/1. . . . . . . . ... . ... . 77

27. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 88 mj, Suspension Concentration
200 mg/1. . . . . . .. . ... 78

28. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 88 mp, Suspension Concentration
500 mg/1. . . . . . . . . . . . 79
29. Effect of Various Dosages of Cationic Polymer on the De.
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 126 m. Suspension Concentration
50 mg/1. . . . . .. .. . . . . . 80









Figures Page

30. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 126 wp, Suspension Concentration
100 mg/1. . . . . . . . . . . . . 81

31. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 126 ml, Suspension Concentration
200 mg/l. . . . . . . . . . . 82

32. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 264 mp, Suspension Concentration
25 mg/I. . . . .. . . . . ........ . 83

33. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a bonodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 264 mP, Suspension Concentration
50 mg/1 . . . . . . . . . . . . .. 84
34. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 264 mi, Suspension Concentration
100 mg/1. . . . . . . . . . . . . 85

35. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 264 mp, Suspension Concentration
200 mg/1. . . . . . . . ... .. . . .... 86

36. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 264 mp, Suspension Concentration
500 mg/ ... . . . . . . . . 87

37. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 365 mp, Suspension Concentration
100 mg/1. . . . . . . . . . . . . 88

38. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 557 mp, Suspension Concentration
1oo mg/1. .. . . . . . . . .. . . 89

39. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 557 mp, Suspension Concentration
200 mg/1. . . . . . . . . . . . . 90











Figures Page
40. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 1305 mt, Suspension Concentration
50 mg/1. ...... . . . . . . ... 91
41. Effect of Various Dosages of Cationic Polymer on the De-
Stabilization of a Monodisperse Polystyrene Latex Suspen-
sion.' Particle Diameter 1305 mIr, Suspension Concentration
100 ag/I . . . . . . . . ... ....... 92

42. Effect of Various Dosages of Cationio Polymer on the De-
stabilization of a Monodisperse Polystyrene Latex Suspen-
sion. Particle Diameter 1305 nim, Suspension Concentration
200 mg/ . . . . .. . . . . . . 93

43. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polyvinyltoluene Latex
Suspension. Particle Diameter 3490 mw, Suspension Concen-
tration 25 mg/1 .......... . . . . . 94

44. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polyvinyltoluene Latex
Suspension. Particle Diameter 3490 mp, Suspension Concen-
tration 50 mg/l . . . . . . . . .... . 95

45. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Monodisperse Polyvinyltoluene Latex
Susplesion. Particle Diameter 3490 mgr, Suspension Concen-
tration 100 mg/1 . . . . . . . ... .. 96

46. Effect of Various Dosages of Cationio Polymer on the De-
stabilization of a Monodisperse Polyvinyltoluene Latex
Suspension. Particle Diameter 3490 m~, Suspension Concen-
tration 200 mg/ . . . . . . . . . 97

47. Effects of Various Dosages of Cationic Polymer on the De-
stabilization of a Silica Suspension. MIN-U-SIL "5i",
Average Particle Size 1.3j, Suspension Concentration
100 mg/l . . . . . . . . . . . .. 98

48. Effects of Various Dosages of Cationic Polymer on the De-
stabilization of a Silica Suspension. MNN-U-SIL "I50",
Average Particle Size 1.3k, Suspension Concentration.
200 mg/l . . . . . . . . . . 99

49. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Silica Suspension. MIN-U-SIL "5V",
Average Particle Size 1.21, Suspension Concentration.
300 mg/l . . . . .. . . . 100


xvii











Figures Page

50. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Silica Suspension. UCAR.Submicron,
Average Particle Size 120 la, Suspension Concentration
100 mg/l . . . . . . . . . . . . 101

51. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Silica Suspension. UCAR-Submicron,
Average Particle Size 120 mI, Suspension Concentration
200 mg/1 . . .....* * . . . . . . .. 102

52. Effect of Various Dosages of Cationic Polymer.on the De-
stabilization of a Silica Suspension. UCAR-Submicron,
Average Particle Size 120 mnp Suspension Concentration
300 mg/1 . . . . . . . . . . . . 103

53. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Silica Suspension. Cab-O-Sil M-5,
Particle Size 15 m., Suspension Concen:r-tion 100 mg/1 . 104

54. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Silica Suspension. Cab-0-Sil M-5,
Particle Size 15 ma, Suspension Concentration 200 ng/l . 105

55. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Silica Suspension. Cab-0-Sil M-5,
Particle Size 15 mn, Suspension Concentration 300 mg/l . 106

56. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH = 4.0 Approximately.
Min-U-Sil "5", Average Particle Size 1.1lj Suspension
Concentration 200 mg/l . . . . . . . 107

57. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH= 4.0 Approximately.
Min-U-Sil "5g", Average Particle Size l.1p, Suspension
Concentration 400 mg/l . . . . . . . . . 108

58. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH 4.0 Approximately.
UCAR-Submicron, Average Particle Size 0.12&, Suspension
Concentration 200 mg/l . . . . . . . ... 109

59. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH = 4.0 Approximately.
UCAR-Submicron, Average Particle Size 0.121, Suspension
Concentration 400 mg/1 . . . . . . . . 110


xviii













60. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH Range 4.7 5.4.
Min-U-Sil "5Ni", Average Particle Size 1.1, Suspension
Concentration 200 mg/1 . . ... . . . . . .. 111

61. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH Bange 4.7 5.4.
Min-U-Sil "51" Silica, Average Particle Size l.lt, Suspen-
sion Concentration 400 mg/1 . . . . . . . 112

62. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH Range 4.6 5.2. UCAR-
Submicron, Average Particle Size 0.12p, Suspension Concen-
tration 200 mg/l . . . . . . . . .. .. 113

63. Effect of Various Dosages of Alum on the Destabilization
of a Silica Suspension at Final pH Range 4.6 5.2. UCAR-
Submicron, Average Particle Size 0.12p, Suspension Concen-
tration 400 mg/l . . . . . ... . . . . 114

64. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-4, Suspension Concentration
25 mg/1 . . . . ... . . . . .115
65. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-4, Suspension Concentration
50 mg/1 . . . . . . . .. . . . . 116
66. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-4, Suspension Concentration
100 mg/l . . . . . . . . .. . . 117

67. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-4, Suspension Concentration
150 mg/1 . . . . . . . . . . . . .. 118
68. Effect of Various Dosages of Cationic Polymer on the De-
stabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-4, Suspension Concentration
200 ag/1 . . . . . . . . ... 119


Page


Figures











Figures Page

69. Effect of Various Dosages of Cationic Polymer on the
Destaldlization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-5, Suspension Concentration
25 mg/1 . . . . . . . . . . . .. 120

70. Effect of Various Dosages of Cationic Polymer on the
Destahtlization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-5, Suspension Concentration
50 mg/1 . ... . . . . . .... ... . 121

71. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-5, Suspension Concentration
100 mg/l . . . . ............ . 122

72. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-5, Suspension Concentration
150 mg/1 . . . . . . . . . . . . 123

73. Effect of Various Dosages of Cationic Polymer on the
Destabilization of a Kaolinite Clay Suspension. Particle
Information Service Sample 30-5, Suspension Concentration
200 mg/l . . . . . . . . . . . . 124













Abstract of Dissertation Presented to the Graduate Council in
.Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy

EFFECT OF PARTICLE SIZE ON THE DESTAE(LIZAION OF
COLLIIAL SUSPENSIONS IN WATER

By
Manuel R. Vilaret

August, 1965

Chairman Dr. A. P. Black
Major Department: Civil Engineering

The effect of particle size on the destabilization of aqueous

colloidal suspensions has been intensively investigated. The study

includes both coagulation by metal coagulants and flocculation by poly-

mers.

Most of the work has been done with latex suspensions, but some

studies employing suspensions of clays and silicas are also included.

The closely monodisperse latex spheres provided an ideal model

to study size effects. By means of a simple series of observations it

is demonstrated that suspensions having the same weight concentration

but differing in particle size may exhibit widely different optical

turbidities.

The experimental work was based on the well-established jar-

testing technique, and a synthetic water was used throughout. Working

with a cationio polymer, it was found that optimum turbidity removal

occurred over narrow dosage ranges, and that these optimum dosages

increased directly with an increase in the suspension concentration and












also with a decrease in particle size i.e., were directly proportional

to the surface areas of the systems. Experiments with polydisperse

suspensions confirmed these findings, both for turbidity removal and

for electrophoretic mobility reversal.

Studies on rates and lengths of mixing showed that the size and

number concentration of particles influence the kinetics of the process.

Times for the appearance of the first visible floor have been observed

to vary with the reciprocal of the weight concentrations and apparently

of the surface areas of the systems. Experiments with calcium chloride

indicate that particle size does not influence the required dosages

where pure coagulation is the prevailing mechanism. A nonionic polymer

in the presence of calcium chloride produced optimum turbidity removals

at dosages related to both concentration and particle size.

Experiments with silica suspensions of various average particle

sizes destabilized with cationic polymer show similar effects. Work with

alum at pH = 4.0, where a tripositive ion is believed to predominate in

the solution, indicates that dosages for coagulation are not affected

by suspension concentration or particle size. At pH = 5.2, however,

where hydrolyzed polymerized complexes of the aluminum ion are believed

to exist, much lower dosages are required for destabilization, and an

optimum occurs at values that are little affected by suspension concen-

tration and particle size, indicating an intermediate situation between

the effect of size in the case of a metal coagulant and in that of a

high polymer.

Further experiments using as a practical model suspensions of two

samples of kaolinite clay of known particle size distributions, when












destabilized with a cationic polymer, produced results which are in

strong semiquantitative agreement with the results obtained for the

latexes and silicas.

The investigation was extended to carbon black samples and

finally to a colored water as an extreme case of very small particle

size, and the results obtained agreed with those predicted from the

previous findings.

It is concluded that particle size, being related to surface

area, has a direct effect on the dosages required for flocculation or

destabilization with polymers, a decrease in size requiring increased

dosages. It is of little importance in the case of coagulation by metal

coagulants. Particle size has also an effect on the kinetics of the

destabilization process; work with polymers shows decreasing floe times

for increasing suspension concentrations and surface areas.


xxiii














I. INTRODUCTION


Very little information is to be found in the literature concern-

ing the effects of physical factors in general on the destabilization

of colloidal suspensions as applied to water treatment. Still less is

found specifically on the effect of particle size as one of these

factors.

In 1952, Langelier and his coworkers1 were probably the first to

discuss these effects. They found that the particle size distribution

of a clay suspension had a definite effect on turbidity and on coagulant

demand when alum was used in its destabilization at uncontrolled but

nearly neutral pH conditions. Linke and Booth,2 investigating the effect

of polyacrylamides on concentrated silica suspensions, found that particle

size was an important factor in determining the amount of p6lymer needed

to flocculate the system. More recently Kane et al.,3 also working with

concentrated silica suspensions, indicated that at a given solids content,

the optimum concentration of polymer increased with decreasing average

particle size.

In spite of these observations, an intensive investigation of the

effect of this parameter upon the destabilization mechanisms was still

lacking. This investigation includes studies employing several different

types of particles and suspension concentrations within the ranges

normally found in water treatment.

Close control of the size of the particles is indispensable in


- 1 -







-2-


this type of study. It is only in recent years that a series of extremely

small, remarkably uniform latex spheres has been made available. These

particles can be obtained in the form of monodisperse suspensions over a

range of sizes that covers those of most of the materials producing tur-

bidity in water. These latex suspensions have been found to be satis-

factory for studies planned to determine the effects of particle size on

destabilization. Suspensions of other materials, including clays, silicas,

activated carbon, ion exchange resins and the color colloids comprising

the "organic color" of some surface waters have also been studied.

It is important that a distinction be made between the terms floc-

culation and coagulation. They have long been used interchangeably, but

as our understanding of the basic mechanisms involved in the destabiliza-

tion of colloids has progressed, it has become increasingly clear that

different forces are involved. By coagulation is meant a general kinetic

process obeying the simple Smoluchowski equation. It is brought about by

neutralization of the repulsive potential of the electrical double layer,

When this occurs, the forces of attraction between adjacent particles

come into play and micro flocs are formed.

On the other hand, in flocculation by polymers a bridging mecha-

nism between particles characterizes the process. The polymers, being

long chain molecules, attach themselves to a particle on a few points

or segments while the rest of the molecule extends into the solution.

Upon collision with other particles, these extended segments can then

be adsorbed on them, and this process results in flocculation.

In the case of destabilization with alum and with a few other







-3-



hydrolyzing coagulants, complicated reactions resulting in the formation

of polymeric species are believed to take place,4' 5, 6, 7, 8 and a

mechanism falling somewhere in between that for metal ions and for high

polymers appears to be involved.

An investigation of the effect of particle size on destabilization

should consider each of these three cases separately. In addition, in

the case of flocculation by high polymers, both cationic and nonionic

materials should be studied since the mechanisms involved may be dif-
9, 10
ferent.















II. THEORETICAL CONSIDERATIONS


Stability of Colloids Destabilization

It is well known that two factors are mainly responsible for

the stability of lyophobic sols. The electrostatic repulsion of

similar double layers surrounding the sol particles, and hydration or

protective fluid layers are sufficient to overcome Brownian motion and

the van der Waals' attraction forces that tend to make particles

coalesce and destabilize the suspension.

Fluid motion, as produced by mechanical mixing, may under the

proper conditions supplement Brownian motion and assist in destabilizing

the suspension.


Coagulation in a Static Dispersing Medium

When no fluid motion is present, coagulation depends on Brownian

motion alone. M. von Smoluchowskill made a classical study of the rate

of this process under "rapid" coagulation conditions, i.e., assuming

the double layer repulsion to be entirely suppressed by an addition of

electrolyte. Under these conditions all collisions result in bonding

of the particles. He studied also the case of "slow" coagulation where

repulsion is not entirely suppressed and all collisions are not effec-

tive. His analysis is based on the assumptions that all the particles

are spherical and of the same size, and that the average number of par-

ticles colliding at a given moment is small as compared to the total

number of particles per unit volume, so that the probability of








-5-


simultaneous multiple collisions is very small.

The problem is considered as one of diffusion and, as an approxi-

mation, the Van der Waalb' forces are replaced by a sphere of attraction

surrounding each particle. When the coagulation process is initiated

and the center of attraction of one primary particle moves into the

attraction sphere of another, they combine into a larger binary particle

having reduced speed. This process is repeated and also collisions occur

between primary and binary and then with tertiary particles and so on.

As a result, the dispersed material combines into individual masses

large enough to settle. As a result of his mathematical analysis,

Smoluchowski showed that the coagulation time, as defined by the reduc-

tion of the number of particles to one-half the initial number, is

inversely proportional to this initial number. This means that the

more concentrated the suspension, the faster coagulation should-proceed.

As stated above, the Smoluchowski's analysis is based on the

assumption of a monodisperse system. However, Mtller12 has shown that

polydispersity increases the probability of interparticle collision.

Also, deviations from the spherical form produce a similar effect.

Wiegnerl3 used the terms perikinetic to designate coagulation by

Browmian motion alone, and orthokinetic when it is caused mainly by

induced agitation. The latter will be discussed below

Gradient Coagulation

When the fluid is not static but subject to motion the number of

collisions per unit time is very much increased.

If laminar flow is assumed, having a velocity gradient along Z,







-6-


perpendicular to the direction of flow, an expression can be easily

found for the probability of the particles colliding. In Figure 1


Fig. 1 Projection on the YZ Plane of the Sphere of Influence Assumed
to Surround Particle at 0.


where laminar flow is assumed to exist in the X direction, perpendicular

to the plane of the paper, particles moving either slower or faster than

the one at the origin will collide with it if they enter its sphere of

attraction, shown here projected as a circle. Since the relative
du
velocity of the particles at height z is z -, the area of the circle

between z and z + dz is d2r -z dz, and if the number of particles per

unit volume is n, the collision probability will be:

C = 2 nz -du r- d
d-r z
or integrating: C = n
3 H








-7-


This expression shows a strong dependence on particle size, since the

radius of the sphere of attraction or radius of coagulation is directly

related to it.

Coagulation in Turbulent Flow

Conditions for laminar flow in water are very limited, and turbu-
14
lent flow is present in most practical cases. Levich4 studied this

problem, and derived an-expression for the rate of coagulation as a

function of number of primary particles and the radius of coagulation.

He based this derivation on two assumptions: (1) that the coagulation

radius of the particles is very small as compared to the microscale of

turbulence, and (2) that colloidal particles are completely contained

within the turbulent eddy, the mechanism of contact occurring on a scale

less than the size of the eddy. These assumptions are equivalent to one

of isotropic turbulence, in which the particles are supposed to migrate

in a completely random way, similar to that of Brownian motion.

Under these conditions it was shown that the rate of coagulation,

as in the case of laminar mixing, is proportional to the cube of the

radius of coagulation r.

A comparison with the number of encounters produced by Bromwian

motion shows that, for larger particles, mixing is the determining fac-

tor in producing coagulation, whereas as particle size decreases Brownien

motion becomes the predominant factor.

Recently Swift and Friedlander15 theoretically and experimentally

have investigated the coagulation of hydrosols by Brownian motion and by

shear flow, and found it to satisfy Smoluehowski's equations. Fair and







-8-


Gemmell6 have developed a model of floc growth for orthokinetic coagu-

lation, and utilized high-speed digital computers to obtain numerical

results.

Flocculation by Polymers

In the discussion above, direct collision between particles and

bonding due to the Van der Waals' forces of attraction, have been

assumed. When long chain polymers are used in destabilization, either

alone or in conjunction with metal coagulants, a different mechanism

seems to prevail. Polymer molecules are adsorbed on particles at several

points of their structure, and since they are long chains, extend into the

solution. In 1951, Jenckel and Rumbach17 were the first to propose this

extended segment theory that soon received confirmation by additional

support in the investigations by Koral et a.,18 ottlieb,19 Peterson and

Kwei,20 and Fontana and Thomas.21 Simha, Frisch and Eirich22 23, 24

developed a theoretical model in which the polymer molecule appears to be

adsorbed on the particle surface at a fraction of the chain segments, while

the others extend or loop into the solution. Polymer adsorption is hin-

dered by inflexibility of the chain, polymer-polymer interactions on the

particle surface and difficulties in matching the spacing of the adsorption

sites on the particle with the length,of the chain segments on the polymer.

The polymer bridging mechanism was proposed first by Ruehrwein

and Ward? in 1952. This theory postulates that polymer chains are

partially adsorbed on a particle surface and extend into the solution

phase. Polymer bridges are formed when these extended segments sorb onto

vacant adsorption sites of other particles in the suspension, thus







-9-


initiating the formation of a floc. La Mer and coworkers25 assumed

that for polymer bridging to occur, a fraction no8 of the number concen-

tration no of particles in the system had to be initially covered by

adsorbed polymer molecules while the other fraction r, (1-e) remained

uncovered. The parameter 8 represents the fraction of the total area of

the system covered with polymer, and (1-0) the fraction that remains un-

covered and available for the adsorption of extended polymer chain seg-

ments upon the close approach of the particles at a rate determined by

von Smoluchowski's diffusion equation. In the aforementioned condition,

La Mer and Healy2 proposed that optimum flocculation should occur when

e = 0.5. Also, Healy and La Mer2 have introduced the parameters P

and I into the proposed mathematical flocculation model where p is the

number of adsorbed segments and T the total number of segments in the

polymer molecule. They also reestablish e as the fractionof adsorption

sites covered by the polymer as compared to the total number of adsorp-

tion sites on the system surface. Under the original La Mer's floccula-

tion theory it was assumed that an optimum floe size exists for each set

of conditions of the system. Consequently, it is possible to determine

mathematically a relationship between the floe radius R and the quanti-

ties 6 and no described before. no is in turn proportional to W, the

solids concentration for a fixed particle size distribution. When

equilibrium conditions are established, the rate of floc formation,

assumed to be proportional to the product no2 0 (1-0) should equal the

rate of disintegration. It is further assumed that the latter varies







- 10 -


directly with floc volume and inversely with floe area and the product

0 (1-9) The Kozeny-Carman equation is then used to relate R to the

rate of refiltration of the suspension through a cake formed in a first

filtration. This filter cake is formed from the flocculated particles

of radius R. By introducing in the Kozeny-Carman equation the condition

that when no flocculant is present R = Ro they arrive, after a few mathe-

matical steps, at the conclusion that the optimum polymer dosage is

related to the sum of first and second power terms of the solids concen-

tration of the suspension. However, in a more recent paper by Kane,

La Mer and Linford28 the second power term is neglected on the basis of

experimental evidence with silica suspensions. The contention that opti-

mum flocculation occurs when e = 0.5 is arrived at from a purely mathe-

matical standpoint and does not consider certain physical and chemical

details of the problem, which have been discussed by other investigators.
20, 21, 29, 30, 31, 32, 33 34 For a gelatin-quartz system, Kragh and

Langston35 found optimum flocculation to occur when only one third of the

saturation amount of gelatin had been adsorbed. Birkner.6 studied the

adsorption of labelled polymers on kaolinite and montmorillonite clays

and found this value to vary between 9 and 32 percent. A portion of this

amount of adsorbed polymer may be completely deposited on the surface

while the remainder either loops and extends into the solution phase or

forms interparticle bridges.

General Predictions

From the work which has been reviewed, it appears that in the pro-

cess of flocculation, surface areas of the systems are related in some way







- 11 -


to the polymer dosages required for their destabilization. In the case

of pure coagulation, when non-hydrolyzing coagulants are used, coagulant

dosage would not be expected to be.affected by particle size and sur-

face area. Surface areas increase directly with a decrease in average

particle size, as can be easily verified for a system of monodisperse

spheres. If we assume that the spheres have a radius r and a density p,

the mass m of each sphere is (4/3) i r3p The number n of spheres per

gram is 1/m and if 4 r2 = a is the surface area of a sphere, the

specific area of the system will be A = na. Hence A = 4 IT r2/(4/3)7 r3

or A = 3/rp which shows the inverse relationship postulated above.

For flocculation, however, the optimum polymer dosage should

depend to some extent on the surface area of the system and should there-

fore be a function of both particle size and suspension concentration.

The number of interparticle collisions depends on the number con-

centration of particles. For a given particle size and type it will then

be directly proportional to weight concentration. The time required for

the appearance of the first visible floc, depending on number of colli-

sions, should also vary inversely with the concentration for a given

particle size. A variation in size, in turn, should have a direct effect

on these times if weight concentration remains unchanged.















III. EXPERIMENTAL MATERIALS AND PROCEDURES


Most of the experiments were performed using latex suspensions.

Such suspensions of monodisperse particles are available in a wide range

of particle size and were found to be well suited for the destabiliza-

tion studies to be made. Other materials studied included four different

silica suspensions and two different kaolinite clay suspensions, all of

known particle size and distribution. A few tests were made using

samples of carbon black of known particle size and distribution and a

few with an ion exchange resin. Finally, a colored surface water was

flocculated with the cationic polymer in an attempt to extrapolate

results obtained with the other materials to those obtained by the de-

stabilization of the color colloid, of very much smaller particle size.


Suspending Medium

The suspending medium used throughout the study was deionized

water to which was added 50 mg/l of pure NaHCO3. This provided a syn-

thetic water having a concentration of 0.594 millimols/l, an ionic

strength of 0.0006, of moderate buffer capacity and containing no multi-

valent nations or anions. The concentration of Na+ was 14 ppm; of HCO,"

36 ppm.

Coagulants and Flocculants

Particle size effects were investigated with both inorganic coagu-

lants and with polymers. The two inorganic coagulants used were


- 12 -







- 13 -


calcium chloride and aluminum sulfate, the latter over two different pH

ranges. Aluminum sulfate has been designated as alum on the tables and

figures. They were prepared by simple solution of the reagents in

double deionized water. Fresh solutions of aluminum sulfate were pre-

pared daily.

Two types of polymers, a cationic material and a nonionic material,

were used. The cationic polymer is a linear homopolymer of diallyl-

dimethylammonium chloride, consisting of a linear chain of recurring

N-substituted piperidinium halide groups alternating along the chain

with methylene groups. The molecular weight of the sample used, which

was designated as CAT-FLOC 1015, was less than 50,000.

The nonionic polymer was a polyacrylamide, which was about 4 per-

cent hydrolyzed. The average molecular weight was between four and six

million. This polymer is designated as Separan NP-20.

Preparation of working solutions of the cationic polymer required

only careful weighing and a short period of rapid mixing, employing a

magnetic stirrer. The nonionic polymer was much more difficult to dis-

solve, requiring about two hours of relatively slow mixing for complete

solution.


Preparation of Suspensions

All dilute suspensions to be destabilized were prepared immedi-

ately before use. In the case of the latexes, carefully measured


A product of Peninsular Chemresearch, Inc., Gainesville, Fla.

A product of the Dow Chemical Company, Midland, Mich.







- 14-


volumes of stock suspension were pipetted into sufficient synthetic

water to yield dilute suspensions of the desired characteristics. Solid

materials were weighed out and added to the synthetic water with rapid

mixing. In all cases, mixing was continued for at least one hour to

insure uniformity. Six oneliter samples were then measured out into the

glass jars of the multiple stirrer, and a seventh jar of suspension

retained as a blank. The pH of each suspension was determined with a

pH meter after it had been added to the jars.


Jar Test Procedure

A multiple laboratory stirrer# permitting the simultaneous mixing

of six one-liter square jars was used throughout the investigation. The

tachometer indicating mixing rate was calibrated against a stop watch.

The jars were supported on a translucent base permitting light to pass

upward through the suspensions, and making it possible to readily note

the time of initial formation of visible floc particles. Coagulant

dosages to be used in each jar were prepared in advance in very small

glass beakers, and it was possible to dose all six jars within a very

few seconds.

Time and rate of mixing used depended upon the type of material

being studied. In case of the organic polymers, samples were mixed at

100 rpm for 20 minutes, followed by a 20-minute flocculation period at

15 rpm. When aluminum sulfate was used, the rapid mixing period was



Beckman Model 0, made by Beckman Instruments, Fullerton, Calif.

hipps and Bird, Richmond, Va.







- 15 -


three minutes at 100 rpm and flocculation 30 minutes at 10 rpm. Settling

periods were also varied depending upon the particular suspension being

studied and the coagulant and/or flocculant being used. The very large

but light floc masses resulting from the flocculation of the latex sus-

pensions required a settling time of eight hours, whereas a one-hour

settling period was used for the Cab-0-Sil suspensions and 30 minutes

for the other silica suspensions and all of the clays.

At the end of the settling period, 250 ml samples were withdrawn

from the supernatant in the jars by means of a suction apparatus connected

to a vacuum source and ending in a U-bent tube so as to permit sampling

at a uniform height of 1-1/2 inches above the bottom of the jar without

disturbing the settled floc. Part of the sample was used for measurements

of residual turbidity and pH and part for the electrophoretic mobility

determinations. Both series of observations were made as soon as possible

after sampling.


Turbidity Measurements

Residual turbidities have been used throughout the investigation

as a basis for comparing the efficiency of coagulation. Light transmit-

tance expressed as optical density was selected as the most convenient

method for determining turbidities of suspensions varying over a very

wide range. A photoelectric colorimeter was used to determine the opti-

cal densities of all the samples and of the original suspensions. Results

have been expressed as percent of initial turbidity. This procedure was



Lumetron Model 450 made by Photovolt Corp., New York, N. Y.







- 16 -


found to permit a better comparison between results obtained on suspen-

sions that differ so greatly in particle size and in concentration of

the dispersed phase. No standardization against the Jackson Candle

Turbidimeter was attempted due to the variety of materials investigated.

A 650 UA filter was used for most of the suspensions with a light path

of 75 mm. It was found possible to increase the sensitivity of the

determinations in case of the smaller particle size latex and silica

suspensions by using a 420 mi filter and a light path of 150 mm. In case

of the suspensions having high optical densities, the 650 mr filter and

a light path of 37.5 mm were used. When other than 650 mP filter and 75

mm light path were used, the data are shown in a footnote in the tables.


Electrophoretic Mobilities

Electrophoretic mobilities were determined by means of a Zeta-

Meter. The ocular micrometer grid in the microscope eyepiece was cali-

brated against a micrometer subdivided to 0.01 mm and found to be in

agreement with the values given in the Zeta-Meter Manual37 for the three

objective magnifications available, when measured along the counting

lines between vertical grid divisions. The distances between the count-

ing lines and the "zero line" were also calibrated with the micrometer

and found to be correct for a 4 mm diameter cell, since they correspond

to 0.147 of this diameter times each total magnification of the micro-

scope. This factor agrees with the position of the stationary layer in

a cylindrical cell as given by Kruyt.3 A mathematical derivation of


An instrument made by Zeta-Meter Inc., New York, N. Y.








- 17 -


the location of the stationary layer is given in the appendix.

For each sample, between five and twenty particles were timed in

each direction using the timer in the instrument. Readings were

recorded on a form prepared for this purpose and after average values

were found the mobilities were calculated. Most of the observations

were made at a voltage of 200 volts, but 300 volts was occasionally used

with particles of low mobility, and 50 volts in a few cases when working

with metal coagulants, since in these suspensions the increased ionic

strength permits more current to flow through the suspension, and the

chances of thermal overturn or gassing of the electrodes are increased.


pH Adjustment

The pH of the synthetic water containing 50 ppm of NaHCO3 was 8.1

and its buffer capacity was sufficient to maintain it at or very close

to that value in case of all of the suspensions being destabilized by

polymers. The series of experiments with alum at controlled pH values

required the previous preparation of titration curves for the water used

and the individual adjustment of each jar to a predetermined pH value

so as to obtain the final desired pH after the addition of each dosage

of the coagulant. pH adjustment was made with 0.1N HC1 or 0.1N NaOH as

required.


Sample Numberina

Since the jar test machine permits the simultaneous mixing of only

six samples, and to make it easier to recover the data from the tables

of results, jars have been given a number representing that of the Jar
/







18 -



test, followed by a dash and a number corresponding to the jar within

that test. Blank jars are numbered as zeros. Additional tests run con-

secutively under the same conditions of nature of colloid, suspension

concentration and type of coagulant, carry the letters A, B, C, or D

after the number of the test. To further identify the data, a different

plotting symbol has been used in the figures for the data corresponding

to each jar test number.















IV. EXPERIMENTS WITH LATEX SUSPENSIONS


General

While latex suspensions as such would rarely be encountered in

water treatment, they constitute an almost ideal model for the studies

to be carried out. They are probably the only materials presently avail-

able in which both particle size and particle size distribution are

accurately controlled within a very narrow range. In addition, the

particles themselves are perfectly spherical and ratios of particle

diameter to surface area and to weight'per particle may be calculated.

Furthermore, the range of sizes available covers almost completely the

range of particle sizes present as turbidity in water. They are all

prepared in similar polymerization systems and should, therefore, dis-

play similar characteristics in their interactions with destabilizing

agents. Finally, they are available from the manufacturer in the form

of a concentrated suspension of known solids content, which can be

diluted directly to any desired concentration in water.

A list of the monodisperse latex particles used in this investi-

gation with their most salient characteristics is given in Table 1. All

of the particles listed in Table 1 are polystyrene latexes, with one

exception indicated (*), which is a polyvinyltoluene latex. In addition

to the eight monodisperse sizes listed, a sample of styrene divinyl-

benzene copolymer latex with particle size ranging from 6-A14 was used

for comparative turbidity measurements, but not for coagulation or floc-

culation studies, since the larger particles settled out upon standing.


- 19 -







- 20 -


Table 1

Characteristics of the Monodisperse Latex Stock Suspensions#


Average Number Percent
Particle Standard of Solids Compositions of Solids. Percent
Diameter, Deviation, Measure- in Stabilizing
mP mg ments Sample Polymer Agent Inorganics

88 8.0 1164 10 92.59 5.56 1.85

126 4.3 328 10 97.09 2.43 0.48

264 6.0 577 3.5 98.23 1.13 0.64

365 7.9 438 4.0 98.93 0.41 0.66

557 10.8 373 10 98.78 0.26 0.95

796 8.3 85 10 99.04 0.25 0.71

1305 15.8 148 10 99.15 0.29 0.56

3490* 17.0 28 10 98.9 -- -

#After Lippie39


From Lippie's information, "the polystyrene latex is made from the
standard bulk polystyrene. The polyvinyltoluene latex particles were

prepared using the commercial vinyltoluene monomer which is comprised

of approximately 60 percent meta-vinyltoluene and 40 percent para-vinyl-

toluene."

Bradford and Vanderhoff0 have discussed the measurement of latex

particles of the kind used by electron microscopy as well as the statis-

tical methods utilized for the determination of average diameters. Since

the weight concentrations appearing on the samples as received could have







- 21 -


changed, they were verified by careful weighing of the residue after

drying in the oven for 24 hours at less than 600C. Results obtained

were always within a fraction of 1 percent of the values on the labels

and it was found that suspensions could be prepared by pipetting

directly from the vials into the suspending medium.


Variation of Turbidity with Particle Size

It is well known that the intensities of light scattered and

transmitted by a suspension are a function of particle size. It was

found of interest for the present work to run a series of curves util-

izing the monodisperse latex suspensions over a range in concentrations

in accordance with those used in coagulation and flocculation experi-

ments with other materials and within the limitations of the instrument

used for turbidity measurements. Since the particles were of the same

shape and refractive index, the transmission of light could be expected

to depend mainly on size.

Monodisperse suspensions of nine particle sizes, each at six

different weight concentrations were prepared and the corresponding

optical densities read on the colorimeter. Optical densities were read

for three different sets of conditions: (1) 650 mA filter, light path

75 mm; (2) 650 mu filter, light path 37.5 mm, and (3) 420 mp filter,

light path 75 mm.

The results obtained are given in Tables 4, 5, and 6 respectively,

and Figure 2 shows the curves obtained for the first set of conditions.

The curves show that a particle diameter of about 0.8A gives minimum






22-
10




5








1






0 .5 1 2 5 '


50 ag/l.

10< 0 m/1.
0.2

Light Path = 75 mm.
e 0Wave Length = 450 mw.

0.1

0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Optical Density
Fig. 2 Effect of Particle Diameter upon the Optical Density of Mono-
disperse Latex Suspensions at Various Weight Concentrations.








- 23 -


light transmission for each weight concentration. It is interesting to

note that the optical density changes many fold when particle diameter

is varied over an order of magnitude. The use of the 420 mu filter, at

the other end of the visible spectrum, produces a similar set of curves

but maximum optical densities are observed in this case when the par-

ticle diameter is approximately 0.6 A. Some sensitivity is gained with

this filter when reading turbidities produced by the smaller sizes.

These curves make it easy to visualize how, depending on particle

size, a suspension with a lower concentration can produce a higher opti-

cal density reading than a more concentrated one of particles of dif-

ferent size.


Destabilization with Cationic Polymer

Initial work was begun on an exploratory basis since no previous

experience was available with this type of suspension. Electrophoretic

mobility determinations were helpful in finding the optimum polymer

dosages for the destabilization of the suspensions, since they cover only

a narrow range.

Once flocculation was obtained for a given combination of particle

diameter and suspension concentration, and as could be expected an opti-

mum polymer dosage was found to exist, the first part of the experimental

program was established on the basis of determining as accurately as

possible the optimum polymer dosages for different concentrations of the

same suspension. Between nine and eighteen jars were usually required

to determine each optimum dosage. It was soon found that a linear

relationship exists between suspension concentrations and the







- 24 -


corresponding optimum dosages for a given particle size. With the 88 mp

diameter latex a range of suspension concentrations of from 5-500 mg/1

was investigated. Similar studies were made with other particle

diameters. The results obtained are shown in Figures 21 to 46 where

both residual turbidities and electrophoretic mobilities have been

plotted against the corresponding polymer dosages for each combination

of particle size and suspension concentration investigated.

When the optimum polymer dosages as obtained from the curves in

the figures listed above were plotted against suspension concentrations

on log-log paper, three straight lines with slope 1:1 were obtained, as

can be seen on Figure 3, indicating the stoichiometric relationship

between concentration and dosage for each particle diameter.

When the data obtained are plotted on the basis of optimum poly-

mer dosages against particle diameter on log-log paper, a series of

parallel straight lines is also obtained, one for each suspension con-

centration. The slope of these lines is the same as that of a line

representing the variation of the surface area of a system of monodis-

perse spherical particles when the diameter of the spheres is varied.

Figure 4 shows the results obtained and also the variation in surface

area of a system having the specific gravity of the latex particles.

These results indicate that the optimum concentration of cationic poly-

mer necessary to destabilize the system is proportional to its surface

area. Calculations appearing on Table 51 lead to a figure of 0.34 mg of

polymer per square meter of latex system area as an optimum dosage.

This figure probably applies only to the particular polymer used.2' 41 42







- 25 -


-UUU


-fl I


I I I J -


11 II 11116 I


264 m,
dia.!s/1


I I II .1 1 I I.T I Il


20
88 m i
dia..
10







2
2 -- --


1 --. .--


0.05 0.1 0.2


0.5 1.0 2.0 5.0 10 20


Optimum Polymer Dosage, ag/1.

Fig. 3 Dosages of Cationic Polymer Required for Maximum Destabilization
of Three Monodisperse Polystyrene Latex Suspensions at Various
Concentrations.


l 'i l : : :


i 1


.1305 -


!-







- 26-


10o


500 mg/1.


200 mg/.

--_--_> --^-- _ --- --- -- -

Surface Areas of a Mono-
100 mg/1. disperse System of Spheres
with Specific Gravity 1.05











25 ag/1."
.50 mg/ .
_ ^- ---- - ... --- -- -
- S S -S ---- -- ---i
=::: -S~: 5;= :S-: -- = :=I


5



*
2







0.5




0.2



0.1



0.05


I II


1000 2000


Particle Diameter Millimicrons

Fig. 4 Effect of Particle Diameter upon the Optimum Cationic Polymer
Dosage Required for Maximum Destabilization of Monodisperse
Latex Suspensions of Given Weight Concentrations Variation
of System Area with Particle Size.


50 100 200


50

ti


20



10 &



5




2



1


1 0.5
00


50


""I


'







- 27 -


Polydisperse Suspensions

The proportionality found between surface area and optimum poly-

mer dosage in the monodisperse suspensions suggested the idea of experi-

menting with a few polydisperse latex systems. A suspension was pre-

pared containing two particle sizes, 88 mt diameter and 264& m diameter,

the first one at a concentration of 10 mg/l, the second one at.25 mg/l.

At these concentrations the total surface area of-each of the suspensions

would be roughly of the same magnitude. In Figure 5 it can be seen that

the results obtained show that the optimum dosage of polymer necessary

to destabilize the polydisperse system is approximately the sum of the

dosages required to destabilize each one of the monodisperse components.

Figure 6 shows similar results obtained with different concentrations of

the same suspensions. Figure 7 shows the results obtained when three

sizes of particles were mixed, the extreme diameters being in a 15:1

ratio. In this'-ase the dosage for the polydisperse system falls a

little short of the sum of the dosages required by the three components,

- and it is possible that some other interactions might affect the mecha-

nisms of the destabilization process. Kruyt points out that it is

easier to destabilize a polydisperse system than a monodisperse one, as

discussed in Chapter II.

Electrophoretic mobility determinations run on the same systems

are shown in Figures 8, 9, and 10. In all three cases it is seen that

the polymer dosages required to reverse the charge on the particles of

the polydisperse system is almost equal to the sum of those required by

each of the monodisperse components considered separately. As was the







- 28 -


0.6



0.5


100 150 200


250 300 350


400 450


SPolymer Dosage, Pg/1.

Fig. 5 Effect of Various Dosages of Cationic Polymer on the Destabili-
zation of a Polydisperse Polystyrene Latex Suspension Comprising
Two Discrete Particle Sizes, as Compared to the Effect on Each
of the Monodisperse Components Taken Separately (88 mu Particle
Diameter at 10 mg/1. and 264 am Particle Diameter at 25 mg/1.).


Polydisperse System
88 mp dia., 10 mg/l.
264 Io dia., 25 mg/1.





264 nW. dia.,
25 mg/l.


88 m/. d~a.,
10 mg/l.


0.4


0.1


**


I I


I I I I


I


Ct i. I







-29-


Polymer Dosage, Mg/1.

Fig. 6 Effect of Various Dosages of Cationic Polymer on the Destabiliza-
tion of a Polydisperse Polystyrene Latex Suspension Comprising Two
Discrete Particle Sizes, as Compared to the Effect on Each of the
Konodisperse Components Taken Separately.(88 m Particle Diameter
at 20 rg/1. and 264 ma Partile Diameter at 50 mg/1.).







- 30 -


1? 1.0 .
0


.S 0.8 -
0.8
0


0,6 1305 Tr dia.,
100 -e//1.


0.4 -

264 mu dia.,
25 ,ag/1.
0.2 88 mw dia.,
S10 mg/l.

0
0 100 200 300 400 500 600
Polymer Dosage, Ag/l.

Fig. 7 Effect of Various Dosages of Cationio Polymer on the Destabili-
zation of a Polydisperse Polystyrene Latex Suspension Comprising
Three Disorete Particle Sizes, as Compared to the Effect on Each
of the Monodisperse Components Taken Separately.









+3

2: .






o
r-


170


210 250 290


330 370


410 450


Polymer Dosage, Pg/1.


Fig. 8 Effect of Various Dosages of Cationic Polymer on the Electro-
phoretic Mobilities of the Three Suspensions Compared for
Residual Turbidity in Figure 5.


Polymer Dosage, Ag/l.


Fig. 9 Effect of Various Dosages of Cationic Polymer on the Electro-
phoretic Mobilities of the Three Suspensions Compared for
Residual Turbidity in Figure 6.


120 200 280 360


440 520 600


Polymer Dosage, pg/l.

Fig. 10 Effect of Various Dosages of Cationic Polymer on the Electro-
phoretic Mobilities of the Four Suspensions Compared for
Residual Turbidity in Figure 7.


264 m dia.,
0 25g/.---
S/ y 88 mi dia., 10 mg/1.


3 I I I I I I
09- 4
3 --- --- I ---___ I __ 1 ___


B


a
1.4-



0 4
. H
P100
-a:
4'
0


1305 dia., I
0 100 mg/1.o
88 mO dia., 10 mg/1. *
-1 -264 mp dia. r / a Polydisperse system


-3 2 I I ., I


.1.1


41


P








- 32-


case with turbidities, some difference is noticed in the last case, in

which particles of three different sizes were present.


Observations on Particle Growth

When work was being done with the smallest size latex suspension

(88 mn diameter) it was early noticed that upon the addition of the poly-

mer under rapid mix conditions a large increase in turbidity was evident

prior to the appearance of visible floe. It was also noted that in those

jars receiving polymer dosages as much as 40-50 percent above or below

the optimum dosage for the suspension being studied, the turbidity of

the settled supernatant was much greater than that of the original sus-

pension before destabilization. An explanation for this behavior is

apparent from a study of Figure 2. When the destabilization process

begins and particles coalesce, new particles of increased size are

formed and the optical density of the suspension increases many fold.

After that portion of the suspension which has been sufficiently de-

stabilized has settled out, sufficient doublets, triplets, etc., remain

in suspension to yield optical density values substantially higher than

that of the original suspension.

A similar effect is noticed with the 264 mp diameter latex

except that it is much less marked, the size of the primary particles

being closer to that producing highest optical densities in Figure 2.

The effect was not found with the larger latex sizes. In all of them,

turbidity was reduced as soon as destabilization took place. In the case

of the 126 my suspension, values for settled turbidity were found to be








- 33 -


between those for the 88 mA and the 264 m4 suspensions, as would be

predicted from the explanation suggested.


Destabilization with Calcium Chloride

This series of experiments was planned with the idea of studying

the effect of particle size when the destabilization mechanism involves

coagulation rather than polymer flocculation. One would predict, on the

basis of the SchulzaHardy rule, that relatively high concentrations of

the bivalent calcium ion would be needed to counteract the strongly

negative latex particles.

Several tests were run with the smallest (88 mo diameter) latex.

No significant difference was found in coagulant dosage requirement for

suspension concentrations of 50 and of 100 mg/l. In those cases where

destabilization was not complete, the increased turbidity effect noticed

in the case of the cationic polymer was also found. Tests with two

other suspensions with particle sizes 264 mu and 557 mj at a concentra-

tion of 100 mg/l behaved similarly although for some unexplained reason

the 264 mP size was destabilized with lower dosages of coagulant than

the others. Figure 11 shows the results obtained with all four suspen-

sions. The behavior of the calcium ion in destabilizing the latex sus-

pensions was different from that of the cationic polymer. In the case

of the polymers an optimum dosage could be identified and determined.

However, in the case of the calcium ion, there was found for each sus-

pension a threshold dosage, above which higher dosages of the ion pro-

duced increasingly better coagulation. This was true for all suspend.

sions, independent of particle size and surface area.








400 ,,i




88 a~. dia. 100 mg/l.





300








S200


I I I




- oo



557 mi. dia.
100 mg/l.






264 niL. dia. -100 Tg/1.
0 1 2 3 4 5 6 7 8
Calcium Chloride, g/l.

Fig. 11 Effect of Calcium Chloride Upon the Stability of Monodisperse
Latex Suspensions of Various Concentrations and Particle Sizes.








- 35-


Observed times for the appearance of the first visible floe also

decreased with increasing coagulant dosages, being shorter in the

samples where residual turbidities were lower. These times are given

in Tables 37 to 40 in the appendix. In general, they were shorter the

higher the particle concentration, but the 264 mi size showed shorter

times than would be expected.


Destabilization with Nonionic Polymer

The effect of particle size on the destabilization behavior of

a nonionic polymer was studied with the same four types of latex sus-

pensions used in the experiments with calcium chloride. Reasonable

dosages of the nonionic polymer alone had no effect on the latex sus-

pensions. After a few trial experiments it was decided to run the tests

with nonionic polymer after the suspensions had been dosed with 2.0 g/1

calcium chloride. The polymer was added exactly two minutes after the

calcium chloride, while the jars were mixed at 100 rpm. This time

interval was found to be very critical and had to be carefully measured

to obtain reproducible results. After addition of the nonionic polymer

the procedure was the same as for the cationic polymer.

As can be seen in Figure 12, an optimum polymer dosage is again

obtained, and doubling the suspension concentration of the 88 ma size

required approximately double the polymer dosage for optimum destabili-
zation. Optimum polymer dosage for the 557 po suspension was approxi-

mately the value to be expected from extrapolation of the values for the

optimum dosages for the 88 ma suspension in two concentrations, 50 ppm

and 100 ppm. However, the optimum dosage for the 264 mn polymer was







-36-
400










N 300







0
43
U 200 .
200




.oo





;0 88 an dia., 100 mg/l.




557 ip dia.,
100 mg/l. 264 p dia., 100 mg/1




0 10 20 30 40 50 60 70
Nonionio Polymer Dosage, mg/1.

Fig. 12 Effect of Dosages of Nonionic Polymer on the Deatabilization of
Four Monodisperse Polystyrene Latex Suspensions of Different
Diameters and Concentrations, in the Presence of 2.0 g/1
Calcium Chloride.











found to be significantly lower than the value predicted on the basis

of total surface area. This behavior on part of the 264 m1 suspension

had been previously noted. In Figure 11, where results of the destablli-

zation of all four polymers with calcium chloride are shown, the same

behavior is evident. The results obtained, although not as conclusive

as those for the cationic polymer, suggest that particle size and con-

sequently surface area of the system also play an important role in the

destabilization by nonionic polymers. Times for the appearance of first

visible floc are given in Tables 41 to 44 in the appendix. In general,

they are shorter for the polymer dosages giving better turbidity

removal, and increase with overdosage.


Destabilization of Latexes with Alum at pH 7.5

A few experiments were run with alum at this pH value which
44
Black and Hannah found to be that of optimum destabilization for clay

suspensions. The results obtained here indicate the existence of opti-

mum alum dosages, high overdosage producing redispersion. Optimum

dosages were influenced by the concentration of the suspensions and by

their particle size. However, good floc was obtained over so wide a

range of coagulant dosages that accurate evaluation of the effect of

particle size on the process was not possible.


Effect of Particle Size and Mixing Times and Rates
While working with the cationic polymer and a small-size latex it

was noticed that similar stages of the destabilization process could be

reached in shorter times when mixing at higher rates than at slow ones.








-38-


This is in agreement with results obtained by La her26 and by Tenney.45

Birkner,6 working with clay suspensions, found a strong similarity be-

tween two curves representing residual turbidities versus total number

of paddle revolutions on suspensions mixed at 20 and at 100 rpm. How-

ever, residual turbidities were always lower at the lower rate.

For the present study it was decided to investigate the effect

of particle size under varied conditions of mixing varying both rates

and times. The mixing velocities which Birkner used, normally 20 rpm

and 100 rpm,were selected. Three monodisperse latex suspensions of the

same concentration, 100 mg/l, and having particle diameters of 88, 264

and 1305 mp respectively, were mixed at each rate for predetermined

lengths of time, settled and tested for turbidity and electrophoretic

mobility. The results obtained can be seen in Figures 13 and 14. In

both cases, the process is faster for the smaller particle sizes,

showing the effect of the number of interparticle collisions on the

kinetics of the process. At both mixing rates there is a marked simi-

larity between the shape of the curves, indicating that the kinetics

of the process are related to the total number of revolutions rather

than to the rate of mixing. Better destabilization is obtained at the

lower mixing rate with all three sizes but especially with the largest

two. Higher shear forces in the fluid seem to have a more marked effect

on the floe formed between. larger particles.

Eleotrophoretic data show similar values for each of the suspen-

sions independent of the time or rate of mixing. The values obtained

were comparable to those found at polymer dosages for optimum turbidity









-39-


0 r-4
0 r


o0 0



S. 0
.4o


Q*rS





0 (0
.U)







. 0
ors









10(U)
crl



o 0 4










0 9! 0
ti P1"











0 CY
OEl






hno





r .




.4-43


.t4pTpq.zu, TETeV4u o 4uesaid AW Wpyqmni TnpPSoa









- 40 -


Altprcqn.L xe!VUI Jo uoojaJGd '*4pTqjnl, tnpyTsany


a,
0 ..

.cR
to




C"U U)


I b
ca o





0: go
U) A.G



HC +' b



to






1o 0
U)H t..







4)
S0 U )






o. o

mC.
nO3



V) 0L


.1 0
0 CO4

"IN








4re
ti-S
ni*- a







- 41 -


removal for the first series of experiments. It is concluded that poly-

mer adsorption is very rapid and thus immediately establishes the

mobility of the particles remaining in suspension, while interparticle

bridging, being slower, is the rate controlling process.


Observations on Times for First Visible Floe

As indicated in Chapter III, times for the appearance of the

first visible floe were recorded for most jars almost from the beginning

of the investigation. These times were determined by careful visual

observation of the jars during the rapid mixing period.

The number of interparticle collisions determines the rate of

formation of floc. As discussed in Chapter II, the number of collisions

is proportional to the particle concentration and is affected by par-

ticle size. It also depends on the rate of mixing, since for higher

rates the effect of Brownian motion is secondary.

Figure 15 shows the inverse relationship observed between suspen-

sion concentrations and times for first visible floc for three monodis-

perse latex suspensions. It can be seen from this figure why it was

necessary to increase the mixing times when the destabilization of sus-

pensions of large size at low concentrations was attempted.

Figure 16 shows a plot of the times observed for first visible

floc against the surface areas of the systems involved. Points taken

from seven sizes of monodisperse suspensions at different concentrations

and from three polydisperse ones give an almost linear plot over several

cycles on log-log paper. A possible explanation for this could be that

under the turbulent but anisotropio mixing conditions predominating in






- 42 -


100



50- -- -




20

1305 -

CD 1





0

-P 0







0.5




0.2
0.1
- I- L --- -I I-- --- -

o -S 1111^--





















5 10 20 50 100 200 500
Suspension Concentration, mg/1.

Fig. 15 Effect of Particle Concentration (Expressed as Weight Concentra-
tion) on the Approximate Times of Mixing Required for the Appear.
ance of the First Visible Floe for Three Monodisperse Latex Sun.
pensions, at Optimum Dosages of Cationio Polymer (Mixing Rate
100 rpm).






- 43-


I


.1


0.5 1 2 5 10 20
Time for First Visible Floe, Minutes


__-.._--. ,, ,,1


___ _ --. I__ _ --. __I-. --.

--I. I ~ ' -- - -- - --ll l l


50 100


Fig. 16 Plot of Surface Areas of Several Latex Suspensions Versus Approxi-
mate Times of Mixing Required for the Appearance of the First
Visible Floe (Mixing Rate 100 rpm).
0 88 mp particle diameter Q 557 m7 particle diameter
0 126 mp particle diameter A 1305 particle diameter
+ 264 mp particle diameter 3490 ma particle diameter
0 365 w1 particle diameter r Polydisperse suspensions


1000


A W
0
-H


20



10

5



I 2

1


0.5



0.2


0.1
0


-t~m







-44-


the jars it is the area of the particles projected on a plane perpendicu-

lar to their direction of travel that determines the number of collisions.

This projected area is proportional to the area of the system, since both

depend on the square of the diameter of the particles.

Another interesting observation can be made from Figures 13 and 14

where residual turbidities are plotted against paddle revolutions for

three different sizes. If, following Packham's6 techniques, the number

of revolutions for 50 percent turbidity removal are read from the

figures and compared they are not far from being in an inverse relation-

ship to the surface areas of each of the three suspensions involved. It

is concluded that not only initial floc formation,but also subsequent

kinetic steps in the flocculation process are determined by the same

characteristics of number of particles, particle size and probably sur-

face area of the original suspension.













V. EXPERIMENTS WITH SILICA SUSPENSIONS


General

It seems logical to assume that shape and chemical character-

istics of the particle surface can modify to some extent the particle

size effects observed with latex suspensions. This consideration makes

it advisable to investigate the behavior of several types of suspen-

sions.

Kane et al.3, 28, 47 have recently studied the flocculation of

silica suspensions. The concentrations used were much higher than those

normally encountered in water treatment. They state that considerably

more polymer was necessary to reach the optimum dosage in an amorphous

silica system than in a crystalline one despite the larger particle size

of the former.

In the present study silica samples in both the amorphous and

crystalline forms were investigated.


Experimental Materials

Four types of silica particles were used, and a brief description

of each one follows.

Min-u-Sil* is a crystalline silica containing 99.90 percent

silicon dioxide. Four sizes are available, but only the smallest, identi-

fied as "51", was used in this investigation, since it was noticed that

some fractions of the larger sizes settle too rapidly after agitation.



*A product of Pennsylvania Glass and Sand Corp., Pittsburgh, Pa.


45-








- 46-


UCAR-Submicron is an amorphous silica powder containing 95 per-

cent silicon dioxide and showing on a photomicrograph spherical par-

ticles ranging from 20 to 250 m in diameter.

Cab-O-Sil is an extremely fine silica with spherical particles

and a 99.0 percent silica content, available in three grades. The M-5

grade used in this study is a fluffy white powder of very low bulk

density.

Ludox* is a colloidal type of silica available in five grades

characterized by extremely small spherical particles surface-hydroxylated

and alkali stabilized to introduce negative charges onto the silica

surface.

Some physical characteristics of these silicas, as given in the

producers' literature, are listed in Table 2 which follows.















*A product of Union Carbide Corp., Niagara Falls, N. Y.

A product of Cabot Corp., Boston, Mass.

*A product of E. I. du Pont de Nemours & Co., Inc., Wilmington,








- 47 -


Table 2
Physical Characteristics of the Four Silicas Investigated


Average Surface
Particle A ea Specific
Silica Size m /g Gravity

Min-u-Sil "S5" l.i 2.06 2.65

UCAR-Suhbicron 120 U. 25.9 --

Cab-0-Sil M-5 15 mq 200 2.2

Ludox HS 12 ma 230 1.21*

in 30.1 percent SiO2 water suspension


Destabilization with Cationic Polymer

A series of tests were run wth the first three types of silicas

listed, at suspension concentrations of 100, 200 and 300 mg/1. Residual

turbidities and electrophoretio mobilities for various polymer dosages

are shown in Figures 4j to 55 indicating the existence of an optimum

polymer dosage in each case. These optimum dosages vary stoichiometrio-

ally with the concentration of the suspensions, as can be seen in

Figure 17. From this figure it is apparent that the suspensions dwth

smaller particle size and higher surface area require higher dosages of

polymer for optimum destabilization. The amounts of polymer required

by the Cab-O-Sil and the MLn-u-Sil suspension are proportional to their

. relative surface areas but the optimum dosage for the UCAR suspension is

50 percent lover than it should be on the same basis. Some different













2.5




S2.0 Cab-O-1 H-5




A 1.5 -


1 .
1'0 UCAR-Submincron




0.5

Min-uSil "5 micron"

00
0 .100 200 300

Suspension Concentration, mg/1.

Fig. 17 Dosages of Cationio Polymer Required for Maximum Destabilization
of Various Silica Suspensions.











characteristic of the surface or a difference between the figure given

by the manufacturer for the surface area and the external area really

available to the polymer, could account for this difference. Rough cal-

culations indicate that at optimum dosage approximately 27 mg of polymer

per square meter of surface area are required by the Cab-0-Sil and

Min-u-Sil suspensions. Using this figure to determine polymer dosage,

a 100 mg/l Ludox silica suspension was flocculated successfully. How-

ever, the UCAR-Sulnicron suspension required only 13 mg of polymer per

m2 of surface area.

Destabilization with Alum at pH 4.0

This experimental series was designed to investigate the particle

size effect in a region of the [A12(SO)]J pH domain where, according

to Matijevic' and Stryker, a counter ion of charge +3 predominates and

a much higher critical coagulation concentration should exist.

Prior to .jar testing, titration curves were run in the suspending

medium so as to determine initial pH values such that upon the addition

of the coagulant dosage the suspensions would each have the desired

final pH values. Once the initial pH values were found experimentally

as described, each jar was adjusted individually and upon addition of

the coagulant, the experiments were carried out in the usual manner.

Suspensions of Min-u-Sil "54" and of UCAR-Suabicron at 200 mg/1 and at

400 mg/1 concentration were destabilized. The results obtained can be

seen in Figures 56 to 59. Alum dosages required appeared to be com-

pletely independent of the concentration of the suspensions. An opti-

mum dosage does not exist, but rather a critical coagulation


- 49-







-50-


concentration or threshold coagulation dosage, and turbidity removal

increases once this value is exceeded. Electrophoretic mobilities reach

a zero value at approximately this alum dosage and become slightly

positive at higher dosages. A difference was noted in the behavior of

the two silicas. The UCAR-Submicron required less coagulant than the

MIn-u-Sil "NS". Since the former has a much higher surface area, this

difference is not believed to be due to particle size but rather to some

characteristic of the surface. It had been noticed before, when working

with the cationic polymer, that the UCAR-Submicron, as compared to other

silicas, required a much lower polymer dosage for optimum removal than

corresponded to its surface area.

Better turbidity removal, based on percent comparisons, is ob-

tained for the more concentrated suspensions. This is in agreement with

the results obtained by others36 and in other parts of this study.

One interesting point to notice is that high dosages of alum are

required at this pH value. These results are in accordance with the

assumption made that a counter ion of charge +3 was predominating in the

solution at the pH value investigated.

Deatabilization with Alum in pH Ranae 4.6-5.4

This experimental series was initially designed to investigate

the particle size effects at a final pH value of 5.2, but control of

this variable at this range of pH values proved to be much more diffi-

cult than at pH 4.0. A complete change in the alum species predominating

seems to occur between the pH values selected for these two experimental

series, as could be expected from the work of Matijeviic Black and








- 51 -


Chen and Black and Hannah also find a transition region at around

pH 4.4. Stumm and Morgan describe the hydrolytic reactions of the

aluminum ion and cite the work of Matijevic et al.5 where the formation

of a soluble polynuclear tetravalent aluminum complex of the formula

Al(OH)204+ is postulated. This work has bean recently followed by

other studies in which even more highly polymerized hydrolyzed ions are

proposed. Matijevic'and Stryker cite the work by Biederman50 where

the existence of the species Al7(OH)17 and A13(OH) + is suggested

to account for potentiometric titration results.


Experimental

The same types of silica suspensions and concentrations as in the

preceding experimental series were used. pH values were also adjusted

in a similar manner, before addition of the coagulant. It was immediately

noted that alum dosages roughly one order of magnitude lower were suf-

ficient to bring about good turbidity removal at this pH range. Once

more the existence of optimum dosages of coagulant was found, and dosages

higher than the optimum produced increasing residual turbidities. Par-

ticle size and concentration show the expected effect, higher system

areas requiring higher coagulant dosages for optimum turbidity removal.

All of these observations agree with predictions based upon the existence

of highly polymerized hydrolyzed species and the flocculation mechanisms

of the polymers. The effect of particle size in this case is not so

marked as with high polymers. It seems to fall between the case of the

pure metal coagulant and that of the long chain polymers. The results

obtained are shown in Figures 60 to 63.















VI. EXPERIMENTS WITH KAOLINITE CLAYS


General

Clays are the most common cause of turbidity in water. They

constitute, therefore, a "practical" model in contrast to the latexes.

However, from the standpoint of the present study they have the dis-

advantage that their laminar structure does not permit a separation into

closely monodisperse fractions that would make it possible to detect

more precisely the effect of differences in particle size.

Clay minerals have a two-layer structure, as do the kaolinites,

or a three-layer structure that can be either expanding as with mont-

morillonite, or nonexpanding as in illites. Kaolinites are in general

nonexpandable and according to van Olphen51 the external surface area is

the only area which is accessible to any molecule or ion added to the

system. This fact makes them more desirable for the present study.


Previous Studies

Langelier and others investigated the effect of particle size on

alum coagulation of a silty clay loam. Starting from a concentrated

stock suspension, turbidity and coagulant demands were determined on

samples taken from the unsettled stock and from its supernatant after

allowing it to settle for various lengths of time up to 25 hours. All

samples were then diluted to the same optical turbidity and tested for

coagulant demand. They concluded that the large particles exerted

little or no effect on coagulant demand, some of them settling


- 52 -







- 53 -


independently of the flocs; that settling of the intermediate size par-

ticles (1.5 P to 5.0 W) reduced the turbidity much more than it lowered

the coagulant demand, and that when size distributions were not altered

by settling, the decrease in coagulant demand was directly proportional

to that in turbidity. They also found that the coagulant demand in-

creased with increasing percentage of smaller particles in the suspen-

sion. In these experiments particle size was determined on the basis of

Stoke's law computations. A fractionation of the suspension on this

same basis led these investigators to conclude that to obtain good floe,

it was necessary to employ suspensions containing particles from the

fine fraction.

It should be noted that this work was done under uncontrolled pH

conditions, but probably in a region where highly polymerized hydrolysis

products of the aluminum ion were present.


Experimental Materials

For the present investigation two kaolinite clay samples desig-

nated as 30-4 and 30-5 were used. These kaolinite samples have particle

size distributions as shown on Figure 18. Sample 30-4 is somewhat

coarser than sample 30-5, and by subdividing the areas under the size

distribution curves into ten equal-weight fractions, average equivalent

spherical diameters of 0.69 p and 0.26 p are obtained respectively for

each of them. The ratio of these two figures is roughly 2.6:1. The

exchange capacity of these kaolinites is in the range of 2.5 to 4.0



Obtained from Particle Information Service, Los Altos, Calif.



































SaRpI. 30-4






sample 30:-


0 10 20 30 40 50 60 70 80 90 100

Percent Finer by Weight

Fig. 18 Particle Size Distribution of the Two Kaolinite Clay Samples
Investigated.


2




1.0
0




i 0.5






0.2


0.1


0







- 55-


milliequivalents per 100 grams.52


Destabilization with Cationic Polymer

Five different concentrations of each of the two clay samples were

studied. In each case the procedure was to weigh the amount of sample

required for seven liters of suspension, mix for two hours with a mag-

netic stirrer in a plastic container, and proceed as described in other

jar test experiments.

The results obtained appear in Figures 64to 73. They resemble

those found with the latexes and silicas, although in a general way the

experimental points do not result in as well-defined curves as with these

materials. As can be seen in Figure 19, the amounts of polymer required

for optimum destabilization of the two samples vary in direct proportion

to the concentration of the respective suspensions. The finer sample

30-5 required on the average, polymer dosages 2.2 times higher than those

required by sample 30-4 at the same weight concentrations. This ratio

is not too far from the one between average equivalent spherical

diameters of the two samples.

It is of interest to note that the range of polymer dosages over

which good removal is obtained is in general much wider than in the case

of latex suspensions. This might be due, in part, to the polydispersity

of the clay particles, and also to their characteristic shape and surface

structure.

It was also found that optimum turbidity removal occurred in a

region of reduced but still negative electrophoretic mobilities. It was

always necessary to overdose a suspension in order to reverse the






56 -
600



500



400
SSample 30-5


8 300
ioo /



Sample 30-4


100
0 --/- ---- --




0 25 50 75 100 125 150 175 200
Suspension Concentration, mg/1.
Fig. 19 Dosages of Cationic Polyn1 r Required for Maximum Destabilization
of Two Kaolinite Clay Suspensions with Different Particle SiLe
Distributions.







57 -



electrophoretic mobility of the particles from negative to positive.


Times for First Visible Floc

As was done before with the latex suspensions, the .times for the

appearance of the first visible floc were observed. For both kaolinite

samples, these times were found to diminish with increasing suspension

concentrations, in agreement with the results obtained with the latexes.

However, the differences between the two samples were not significant.















VII. EXPERIMENTS WITH OTHER MATERIALS


General

In addition to the investigations.with latex, silica and clay

suspensions, some work was attempted with three carbon black samples,

one ion exchange resin and a colored water devoid of turbidity. This

was done in order to determine, if possible, the effect of particle size

and surface characteristics of particles different from those previously

studied.

Unfortunately, these studies were found to present difficulties

that would have required the scope of this investigation to be greatly

expanded in order to achieve completely quantitative results. Accord-

ingly, the observations to be described are somewhat preliminary and

semiquantitative in nature. They are nonetheless considered of some

interest.


Carbon Black Suspensions

Three carbon black samples designated as 2-3 Sterling R, 2-6

Carbolac 1, and 2-9 Sterling FT were used. Particle size distributions

for these samples appear in Figure 20 which follows.


- 58 -


Obtained from Particle Information Service, Los Altos, Calif.








- 59 -


0 200 400 0 8 16 24
Diameter, mn Diameter, mL

Fig. 20 Particle Size Distributions of the Three Carbon Samples
Investigated.



These size distributions show rather wide ranges of sizes but at

the same time a significant difference in average size especially be-

tween samples 2-6 and 2-9.

These carbon black samples are available in the form of a dry

powder. Table 3 shows also the principal characteristics of the size

distributions of these samples, as furnished by the distributors.







- 60 -


Table 3

Characteristics of Carbon Black Samples



Sample No. 2-3 2-6 2-9

Designation Sterling R Carbolac 1 Sterling FT

Average particle
diameter, 1 0.0797 0.0106 0.1796

Distribution
within limits 98% between 97% between 99% between
0.01. and 0.21h 0.005P and 0.0171 0.09P and 0.301

Surface area
m2/g 22 1000 15.7




Experimental

Several methods were tried to suspend the particles without loss

of some fraction, since they are very difficult to wet. Pouring the

weighed sample into the vortex of the Waring blendor was found to be the

most convenient method, but still some loss was experienced by adsorption

on the glassware and at the liquid-air interface. Coating of the glass-

ware with a silicone did not help.

The use of a surfactant was not attempted since it would have

involved going into a previous study to determine the effects of changing

the characteristics of the carbon surface.

Difficulties were also experienced with turbidity measurements,

since very low concentrations of carbon produce high optical densities,

and no satisfactory results were obtained by nephelometry, by gravimetric







- 61 -


methods or even by refiltration rates.

Several jar tests were run using a cationic polymer on each of

the three samples at a few different concentrations. Results obtained

were not too consistent, due probably to the difficulties described

above. Some general observations may be made.

Tests run on sample 2-3 that had a fairly wide range of particle

size, always produced good floc with fairly good turbidity removal at

polymer dosages of only 20 to 50 tg/l. Two carbon concentrations, 40

and 60 mg/1,were investigated.

Tests run on sample 2-6, finest of all, also produced good floc

in all jars, with good turbidity removal and at an optimum polymer

dosage of 70 Pg/1. The suspension concentration employed was 40 mg/1.

Electrophoretic mobilities were quite negative for this polymer dosage.

Dosages eight or ten times larger were required for charge reversal.

At these positive mobilities good flocs were obtained but turbidity

removal was poor. It was possible to secure reproducible results with

this sample and optimum polymer dosage could be accurately determined.

Tests run on sample 2-9, of larger particle size, produced no

floc with polymer doses which successively lowered and finally reversed

electrophoretic mobilities. Some turbidity removal was found, neverthe-

Sless, and optimum dosages were found for each of three suspension con-

centrations. Three concentrations were used; 10, 40 and 60 mg/1.

Respective optimum dosages were approximately 20, 40 and 50 1g/1.

Results were not too consistent.

Good floc and good turbidity removal were obtained, however,







- 62 -


when one-third by weight of the fine sample 2-6 was mixed with two-thirds

of the coarse 2-9 that did not flocculate when alone.

One experiment was run with alum on sample 2-9 at nearly neutral .

pH, and good floc was formed at dosages as low as 5 mg/l. Electro-

phoretic mobilities were reversed at 10 mg/l and at all higher dosages.

It is of interest to point out that in the carbon black systems

the ratio of polymer dosage required for destabilization to surface

areas of the systems is very low. This is probably because only a small

fraction of this area is really available to the polymer to adsorb.


Ion Exchange Resin

A sample of a fine particle size ion exchange resin in the form

of an essentially water-free powder was obtained. According to the pro-

ducer this is a sulfonated polystyrene cation exchange resin in the

sodium form.

Separation of the sample into size fractions was attempted by

settling and Stoke's law calculations. After a few small-scale attempts

it was found that the errors involved in fractionation of the sample and

the uncertainties in percentages settled and in suspension were too great

to make it possible to determine the effect of the particle size on

destabilization. A few tests run on suspensions of the unfractionated

resin produced, together with sparse floc, the settling of a few indi-

vidual particles of the larger fraction, while the very fine ones

remained in suspension. Electrophoretic mobility determinations showed



Amberlite CO-120, Type III, a product of Rohm & Haas Company,
Philadelphia, Pa.











a wide range of values for the individual particles. No further

attempts were made since the described results were not too promising

for the purpose of this investigation.


Colored Water

Color in water is believed to be present as a colloid of extremely

fine particle size. mack and Christman5 found that the majority of the

particles of the color are in the 4.8 to 10.0 m1 range. Black and

Willems,5 working with a cationic polyelectrolyte previously used by'

Cohen t ai.,55 found that very high dosages were fairly effective for

color removal but only in the acid range.

Since the particles of the color colloid are so small in size,

the surface areas of the systems involved must be very high even at low

concentrations. To destabilize a system of this type, very high polymer

dosages would be predicted in accordance with the results obtained

throughout this study. A short experimental series was run on a sample

of colored water previously used in the laboratory by other investiga-

tors.53' 5 After some exploratory tests with increasing polymer dosages,

an increased turbidity effect similar to that described for the smallest

latex particles was noticed first, followed by actual formation of floc.

Polymer overdosage caused redispersion. Fairly good flocculation was

obtained at a polymer dosage range of from 20 to 30 mg/l with optimum

removal occurring at 25 mg/l of polymer. Color reduction was not com-

plete; approximately 36 percent of the original color remaining after


From Hatchet Creek, a tributary of Newnans Lake near Gainesville,


- 63-












destabilization at the optimum polymer dosage.

The results obtained show, at least qualitatively, that the

mechanisms involved in this case are probably similar to those studied

for the other colloids.












VII. DISCUSSION AND CONCLUSIONS


Discussion

Most of the results obtained in the present investigation agree

with the general predictions outlined in Chapter II, and either consti-

tute a proof of the correctness of the arguments relating particle size

or furnish additional evidence supporting them.

The simple series of observations on the variation of the optical

densities of latex suspensions with particle size demonstrates that

optical turbidity may not be used as a safe guide in predicting the

dosage of flocculant required for optimum destabilization.

The results obtained for the destabilization of latex suspensions

with a cationic polymer are conclusive and clearly show quantitative

relationships between the optimum polymer dosage and the surface area

of the suspensions, which are in turn related to the size of the par-

ticles involved. The availability of monodisperse suspensions in a

wide range of particle sizes has made it possible to demonstrate the

quantitative relationships which have been found to characterize the

destabilization of such suspensions. Also, the experiments run with

polydisperse suspensions and with their component monodisperse frac-

tions provide convincing evidence of the relationship between system

area and optimum polymer dosage. All these results make it possible

to state that the amount of cationic polymer required for the optimum

destabilization of a colloidal suspension is directly proportional to

the area of the system, other factors remaining unchanged. Although


- 65 -







- 66 -


suspension concentrations used were never too high, the fairly wide

range of two orders of magnitude covered with the 88 mr diameter latex

did not show any deviation from this proportionality. The fact that

a polymer of only one molecular weight has been used throughout the

investigation is not believed to significantly limit the validity of

this statement. The results obtained on silica and clay suspensions

indicate the applicability of these findings to colloids of different

types. When one compares the effect of particle size on flocculation

by high polymers with those observed when destabilizing the suspensions

with metal coagulants, the observed differences are those which would

be expected to be found, based upon what is presently known.

Of some interest also are the intermediate results obtained with

alum at pH values where the existence of polymerized complexes has been

rather well established by a number of investigators.

From the discussion advanced in Chapter II concerning different

models for coagulation and flocculation under static and nonstatic

conditions of the suspending medium, and based upon a comparison with

the experimental results related to the kinetics of the flocculation

mechanisms discussed in Chapter IV, it appears that none of the theoret-

ical models is really applicable to the actual mixing conditions

existing in the jars. There is no doubt that at the mixing speeds

employed turbulent conditions should prevail at the normal viscosity

of water. However, during the mixing of clay suspensions, close

observation of the jars before addition of the destabilizing agent

shows the existence of the phenomenon of streaming birefringence,







- 67 -


indicating that the kaolinite plates travel in a parallel orientation,

and this fact is contrary to the assumption of isotropic turbulence in

the model discussed by Levich.

From the inverse linear relationship noticed in Figure 16 between

surface areas and times for first visible floc, it might be thought that

it is the projected area of the particles, perpendicular to their

general direction of motion, that determines their collision probability,

If a model based on an assumption of anisotropic turbulence could be

established, in which the particles would travel in approximately

parallel paths but subject to conditions of alternate relative accelera-

tion and deceleration due to mixing, these empirical results might be

theoretically supported, since the surface areas of the systems are

proportional to the projected areas on a plane, both depending on the

square of the radii of the particles. If in the derivation made in

Chapter II for the case of gradient coagulation the velocity of the

particles approaching the sphere of attraction is not assumed to be

related to a gradient but independent of the values of z, and the

corresponding term is taken out of the integral, a final expression is

obtained in which the probability of collision is proportional to the

area of the circle and in turn to the system area. These are, of

course, only conjectures, which serve to suggest possible explanations

for the observations made.

During the experiments it was also observed qualitatively that

as a rule the small particle size fractions of polydisperse suspensions

helped to remove the larger ones by entrainment in the floc, as earlier







- 68 -


suggested by Langelier. Polydispersity appears to facilitate floccula-

tion.

One more observation was that for the unchanged conditions of

mixing and the same molecular weight of polymer, the final size of the

latex floc seemed to vary in inverse proportion with the diameter of

the primary suspended particles. The relative size of the polymer

molecules and of the particles might be expected to exert an influence

since inertia forces increase with the volume of the particles, whereas

polymer adsorption depends on the number of available sites on the sur-

face and the number of segments in the molecule.

During the early stages of the investigation the observations Of

the electrophoretic mobility of the particles were of great help in

identifying zones of overhand underdosage of the systems. In general,

it was found that optimum turbidity removal was obtained at lower

dosages than those required for mobility reversal. In the unflocculated

latex suspensions, mobilities were found to increase with particle size.

When particles of different chemical nature were flocculated with a

cationic polymer it was observed that for equal total surface areas,

different amounts of the same polymer were required for optimum de-

stabilization of each system. An explanation is that a different type

of polymer-to-surface bonding might predominate in each case. It is

also possible that all the surface area is not available to the polymer

molecule, and that the fraction available is different for each type of

particle.








- 69 -


Conclusions

The following principal conclusions can be drawn as a result of

this investigation.

First, in flocculation with a cationic polymer, the dosages that

produce optimum turbidity removal increase directly with the surface

area of the system, which in turn is inversely proportional to the size

of the suspended particles.

Second, in the case of a nonionic polymer used in conjunction

with a metal coagulant, particle size has an effect on the polymer

dosages for optimum destabilization, the smaller particles requiring

higher dosages.

Third, in coagulation with a nonhydrolyzing metal coagulant no

optimum dosages are observed, and particle size has no marked effect

on required coagulant dosages.

Fourth, with a coagulant such as alum,which is polymerized

through hydrolysis at certain pH values, particle size has an effect

on the required dosages, although the relationship has not been so

well defined as with high polymers.

Fifth, particle size has a definite effect on the kinetics of

the destabilization process, making it faster the smaller the size,

other factors remaining unchanged.

Sixth, suspensions having the same weight concentration but

differing in particle size may exhibit wide differences in optical

turbidity.

Seventh, the smaller size particles always produced larger

masses of floc under the conditions investigated.

































APPENMCES






















APPPIENX A






Figures 21 to 73 Dependence of Residual Turbidity

and Electrophoretic Mobility on Coagulant or Floccu-

lant Dosage.for Latex, Silica, and Clay Suspensions.






- 72 -


/ U I \

400


o 300 /



200
I

100



70 80 90 100 110 120 130 140
Polymer Dosage, pg/1.




| 0 2



-3 -- I I I i 1
-1 -



0 70 80 90 100 110 120 130 140
Polymer Dosage, ,lg/1.

Fig. 21 Effect of Various Dosages of Cationic Polymer on the Destabili-
zation of a Monodisperse Polystyrene Latex Suspension. Particle
Diameter 88 nm, Suspension Concentration 5 mg/1.







S-73-
S500
E' 0 '0
*-
' 400

74
0

300
41.\









0o

0 I I 0 I I
125 150 175 200 225 250 275 300
Polymer Dosage, pig/1.

+2

;1 +1 *
I | I I I I




o +
00




-3 I I I I I
125 150 175 200 225 250 275 300
Polymer Dosage, pg/1.

Fig. 22 Effect of Various Dosages of Cationic Polymer on the Destabili-
zation of a Monodisperse Polystyrene Latex Suspension. Particle
Diameter 88 mr, Suspension Concentration 10 mg/l.







74-

500f




400


*r
300
H


t 00



i00




0 0 I I I I I I




+o r
+2









+ O




".
10 0




4, 0

















4 -3 I I I I I A
4 200 300 400 500 600 700 800 900
Polymer Dosage, Ag/l.











Fig. 23 Effect of Various Dosages of Cationic Polymer on the Destabili-
zation of a Monodisperse Polystyrene Latex Suspension. Particle
Diameter 88 mo, Suspension Concentration 20 ag/1.
43 0






200 300 400 500 600 700 800 900
)' Polymer Dosage, ~Lg/l.

Fig. 23 Effect of Various Dosages of Cationic Polymer on the Destabili-
zation of a Monodisperse Polystyrene Latex Suspension, Particle
Diameter 88 um, Suspension Concentration 20 mg/i.







- 75 -


. 400



S300



S200



100



Sc o -- i -- -- i -- ^'"^ --
S 0 100 200 300 400 500 600 700
Polymer Dosage, cig/l.



+1



X-S

o -2 -
0 -) 0lll
.. -3 I --
0 100 200 300 400 500 600 ?00
w Polymer Dosage, pg/l.

Fig. 24 Effect of Various Dosages of Cationic Polymer on the Destabili-
zation of a Monodisperse Polystyrene Latex Suspension. Particle
Diameter 88 mP, Suspension Concentration 25 mg/1.






- 76-


S' 400



S300



200 -



100 -



0 I I I
200 400 600 800 1000 1200 1400 1600
Polymer Dosage, Pg/1.



+? I I I ---\ -
+1



4 o0-

o u-2 a.


o 200 400 600 800 1000 1200 1400 1600
SPolymer Dosage, Lg/l.

Fig. 25 Effect of Various Dosages of Cationic Polymer on the Destabili-
zation of a Monodisperse Polystyrene Latex Suspension. Particle
Diameter 88 mp, Suspension Concentration 50 mg/1.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs