DoE
EPA
United States
Department of Energy
Division of Solid f ye
Mining and PrepaM' 01
Pittsburgh PA 15213
FE 9001 1
U S Environmental Protection Agency
Office of Research and Development
Industrial Environmental Research
Laboratory
Research Triangle Park NC 2771 1
EPA 600 ; 79
January 1979
Surface Phenomena
in the Dewatering of Coal
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research rfports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of enrrgy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
Price: Paper Copy $9.00
Microfiche S3.00
-------
FE-9001-1
(EPA-600/7-79-008)
January 1979
Distribution Category UC-90b
Surface Phenomena
in the Dewatering of Coal
by
D.V. Keller, Jr., G.J. Stelma, and Y.M. Chi
Syracuse University
Department of Chemical Engineering and Materials Science
Syracuse, New York 13210
EPA/DoE Interagency Agreement No. DXE685AK
Program Element No. EHE623A
EPA Project Officer: David A.Kirchgessner DoE Project Officer: Albert F. Baker
Industrial Environmental Research Laboratory Division of Solid Fuel Mining and Preparation
Research Triangle Park, NC 27711 Pittsburgh, PA 15213
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
and
. U.S. DEPARTMENT OF ENERGY
Division of Solid Fuel Mining and Preparation
Pittsburgh, PA 15213
-------
FINAL REPORT
October 1, 1978
SURFACE PHENOMENA IN THE
DEWATERING OF COAL
by
Professor D. V. Keller, Jr., Director
G. J. Stelma, Graduate Research Assistant
and
Y. M. Chi, Graduate Research Assistant
for
DEPARTMENT OF ENERGY
CONTRACT NO. ET-75-G-01-9001
Contract Officer: Mr. Albert Deurbrouck
Pittsburgh Energy Research Center
Pittsburgh, Pennsylvania 15213
Department of Chemical Engineering and Materials Science
Syracuse University
Syracuse, New York 13210
MTS-6707-DK-FR-1078
-------
ABSTRACT
The influence of certain surfactants on the dewatering of fine coal
has been investigated. The surfactants investigated were found to have a
two-fold effect. They were found to effect the pressure differentials re-
quired for dewatering in addition to the residual water contents of the
coal beds attainable by this dewatering. Both effects were attributed to
surfactant adsorption.
Adsorption at the liquid-air interface resulted in a decrease in the
interfacial tension beween the two phases. The effect this decrease had on
the pressure differentials required for' dewatering was found to be in agree-
ment with that predicted by the capillary theory applied to the system.
Adsorption at the solid-liquid interface was correlated with the
complex behavior of the residual water contents as a function of sur-
factant addition. A comprehensive model for the adsorption of the sur-
factants onto the coal was presented, based on the Stern-Grahame theory
of adsorption at an electrical double layer. The model allowed for the
mode of physisorption to change as the amount of surfactant adsorbed in-
creased, and also for a phenomenon known as hemi-micellation. Using the
model, consistent and reasonable results were found for the specific sur-
face area of the coal and for the standard free energies of adsorption.
The model was also found to be appropriate when the heterogeneous nature
of the coal was considered. Furthermore, the hydrophobicity of the
molecular groups of the molecules, expected from the model to be con-
trolling the hydrophobicity of the interface, was also found to be in
agreement with that predicted by other means.
ii
-------
TABLE OF CONTENTS
PAGE
LIST OF TABLES iv
LIST OF FIGURES v
I. INTRODUCTION 1
Overview 1
Background 6
II. CAPILLARY THEORY OF DEWATERING 11
[II. COAL 16
IV. SURFACTANTS 21
V. ADSORPTION 27
Stern-Grahame Theory of Adsorption 27
Adsorption from Solution 32
Measurement of Concentration Change 36
VI. MATERIALS 38
Coal 38
Surfactants 39
1. Aerosol A-196 (anionic) 39
2. Triton X-114 (non-ionic) 39
3. Dodecyl Pyridinium Chloride (cationic) 42
4. Sodium Dodecyl Sulfate (anionic) 42
5. Aerosol-OT Cam'onic) 46
111
-------
VII. EXPERIMENTAL 49
Introduction 49
Pressure Dewatering (equilibrium) 51
Vacuum Dewatering (non-equilibrium) 56
Liquid-Air Interfacial Tension Measurements (YL«) 60
Liquid-Air Interfacial Tensions of Aqueous Surfactant
Solutions 63
Surfactant Adsorption Experiments 65
VIII. RESULTS 67
Pressure Dewatering (equilibrium) 67
Vacuum Dewatering (non-equilibrium) 87
Liquid-Air Interfacial Tension (y, .^Concentration Curves . . 92
Influence of Coal on the Interfacial Tension of Distilled
Water 97
Adsorption Isotherms 99
Standard Free Energy of Adsorption 108
IX. DISCUSSION 112
Interfacial Tension-Concentration Curves 112
Pressure Dewatering 114
The Coal-Surfactant Interaction 122
X. CONCLUSION ; .... 157
APPENDIX A 159
APPENDIX B 160
REFERENCES 166
iv
-------
LIST OF TABLES
TABLE NO. PAGE
I. GROUP HLB VALUES 26
II. DATA FOR PRESSURE DEWATERING EXPERIMENTS 69
III. WEIGHT OF COAL SAMPLES USED IN VACUUM DEWATERING EXPERIMENTS. 88
IV. CRITICAL MICELLE CONCENTRATION AND MAXIMUM SLOPE DETERMINED
FROM THE YLA-LN C CURVES OF EACH SURFACTANT 96
V. INFLUENCE OF COAL ON THE INTERFACIAL TENSION OF DISTILLED
WATER 98
VI. CRITICAL MICELLE CONCENTRATIONS OF AQUEOUS SURFACTANT
SOLUTIONS 113
VII. RATIO OF THE AVERAGE COAL PARTICLE RADIUS TO THE AVERAGE PORE
RADIUS OF THE COAL BEDS 118
VIII. DATA FOR CALCULATION OF THE SPECIFIC SURFACE AREA OF THE
COAL BASED ON THE CAPILLARY MODEL 121
IX. ADSORBED AREA OF SURFACTANT MOLECULES 127
X. SPECIFIC SURFACE AREA OF THE COAL DETERMINED ON THE BASIS OF
THE SURFACTANT ADSORPTION MODEL 144
XI. SURFACTANT EFFECTIVENESS 154
Bl. DATA FOR THE ADSORPTION ISOTHERM OF AEROSOL A-196 161
B2. DATA FOR THE ADSORPTION ISOTHERM OF TRITON X-114 162
B3. DATA FOR THE ADSORPTION ISOTHERM OF DODECYL PYRIDINIUM
CHLORIDE 163
B4. DATA FOR THE ADSORPTION ISOTHERM OF SODIUM DODECYL SULFATE. . 164
B5. DATA FOR THE ADSORPTION ISOTHERM OF AEROSOL-OT ^ . 165
-------
LIST OF FIGURES
FIGURE NO. PAGE
1. INTERFACIAL TENSIONS AT THE POINT OF THREE PHASE
CONTACT IN A CAPILLARY 12
2. DEWATERING CURVE FOR A BED OF PARTICLES-WITH A BROAD
SIZE DISTRIBUTION 14
3. DEWATERING CURVE FOR A BED OF PARTICLES WITH A NARROW
SIZE DISTRIBUTION 14
4. POSSIBLE ORGANIC STRUCTURE OF BITUMINOUS COAL 17
5. TYPICAL BEHAVIOR OF THE SOLUTION-AIR INTERFACIAL TENSION
OF AQUEOUS SURFACTANT SOLUTIONS AS A FUNCTION OF SURFACTANT
CONCENTRATION 22
6. EXAMPLE OF A SURFACTANT MOLECULE: SODIUM DODECYL SULFATE. . 22
7. POSSIBLE CROSS-SECTION OF A SPHERICAL MICELLE 24
8. POSSIBLE CROSS-SECTIONS OF BILAYER MICELLES 24
9. ILLUSTRATION OF THE ELECTRICAL DOUBLE LAYER 28
10. THE ELECTRICAL DOUBLE LAYER: VARIATION OF THE POTENTIAL
WITH DISTANCE, SHOWING THE INNER AND OUTER HELMHOLTZ
PLANES 29
11. STRUCTURAL FORMULA OF AEROSOL A-19G (SODIUM DICYCLOHEXYL
SULFOSUCCINATE) 40
12. -MOLECULAR MODEL OF A DICYCLOHEXYL SULFOSUCCINATE ION,
(SIDE AND TOP VIEW) 41
13. STRUCTURAL FORMULA OF TRITON X-114 (OCTYLPHENOXY POLYE-
THOXY ETHANOL) 43
VI
-------
FIGURE NO. PAGE
14. MOLECULAR MODEL OF TRITON X-114 43
15. STRUCTURAL FORMULA OF DODECYL PYRIDINIUM CHLORIDE 44
16. MOLECULAR MODEL OF A DODECYL PYRIDINIUM ION 44
17. STRUCTURAL FORMULA OF SODIUM DODECYL SULFATE 45
18. MOLECULAR MODEL OF A DODECYL SULFATE ION 45
19. STRUCTURAL FORMULA OF AEROSOL-OT (SODIUM DI (2-ETHYLHEXYL)
SULFOSUCCINATE) 47
20. MOLECULAR MODEL OF A DI (2-ETHYLHEXYL) SULFOSUCCINATE ION,
(SIDE AND TOP VIEW) 48
21. TEFLON 'DEWATERING CELL 52
22. PRESSURE DEWATERING APPARATUS 53
23. VACUUM DEWATERING APPARATUS 57
24. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH COAL
BED CONTAINING DISTILLED WATER 70
25. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH COAL
BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL A-196;
C = 1.32 X 10"6 MOLE FRACTION 71
26. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH COAL
BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL A-196;
C = 2.02 X 10"6 MOLE FRACTION 72
27. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH COAL
BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL A-196;
C = 1.42 X 10"4 MOLE FRACTION 73
28. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH COAL
BED CONTAINING AN AQUEOUS SOLUTION OF TRITON X-114;
C = 3.20 X 10"7 MOLE FRACTION 74
vii
-------
FIGURE NO. PAGE
29. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF TRITON X-114;
C = 1.45 X 10"6 MOLE FRACTION 75
30. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF DODECYL
PYRIDINIUM CHLORIDE; C = 4.75 X 10"5 MOLE FRACTION .... 76
31. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH COAL
BED CONTAINING AN AQUEOUS SOLUTION OF SODIUM DODECYL SUL-
FATE; C = 7.20 X 10"6 MOLE FRACTION 77
32. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF SODIUM DODECYL
SULFATE; C = 2.10 X 10"5 MOLE FRACTION 78
33. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH COAL
BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL-OT;
C = 3.00 X 10~6 MOLE FRACTION 79
34. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH COAL
BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL-OT;
C = 1.20 X 10"5 MOLE FRACTION 80
35. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH COAL
BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL-OT;
C = 1.85 X 10"5 MOLE FRACTION 81
36. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 X 60 MESH COAL
BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL-OT;
C = 2.00 X 10"5 MOLE FRACTION 82
viii
-------
FIGURE NO. PAGE
37. EFFECT OF AEROSOL A-196 ON THE AMOUNT OF AQUEOUS
SOLUTION RETAINED BY THE PRESSURE DEWATERED COAL BED
(ISOBARIC DATA; COAL SIZE: 35 X 60 MESH) 83
38. EFFECT OF TRITON X-114 ON THE AMOUNT OF AQUEOUS SOLUTION
RETAINED BY THE PRESSURE DEWATERED COAL BED. (ISOBARIC
DATA; COAL SIZE: 35 X 60 MESH) 84
39. EFFECT OF DODECYL PYRIDINIUM CHLORIDE ON THE AMOUNT OF
AQUEOUS SOLUTION RETAINED BY THE PRESSURE DEWATERED COAL
BED. (ISOBARIC DATA; COAL SIZE: 35 X 60 MESH) 85
40. EFFECT OF SODIUM DODECYL SULFATE ON THE AMOUNT OF AQUEOUS
SOLUTION RETAINED BY THE PRESSURE DEWATERED COAL BED.
(ISOBARIC DATA; COAL SIZE: 35 X 60 MESH) 85
41. EFFECT OF AEROSOL-OT ON THE AMOUNT OF AQUEOUS SOLUTION
RETAINED BY THE PRESSURE DEWATERED COAL BED. (ISOBARIC
DATA; COAL SIZE: 35 X 60 MESH) 86
42. EFFECT OF AEROSOL A-196 ON THE RESIDUAL AQUEOUS SOLUTION
CONTENT OF THE VACUUM DEWATERED COAL BED. (ISOCHRONAL
DATA AT AP*= 1 ATM; COAL SIZE: 35 X 60 MESH) 89
43. EFFECT OF TRITON X-114 ON THE RESIDUAL AQUEOUS SOLUTION
CONTENT OF THE VACUUM DEWATERED COAL BED. (ISOCHRONAL DATA
AT AP = 1 ATM; COAL SIZE: 35 X 60 MESH) 89
44. EFFECT OF DODECYL PYRIDINIUM CHLORIDE ON THE RESIDUAL
AQUEOUS SOLUTION CONTENT OF THE VACUUM DEWATERED COAL
BED. (ISOCHRONAL DATA AT AP = 1 ATM; COAL SIZE: 35 X 60
MESH 90
ix
-------
FIGURE NO. PAGE
45. EFFECT OF SODIUM DODECYL SULFATE ON THE RESIDUAL AQUEOUS
SOLUTION CONTENT OF THE VACUUM DEWATERED COAL BED. (ISO-
CHRONAL DATA AT AP = 1 ATM; COAL SIZE: 35 X 60 MESH). . . 90
46. EFFECT OF AEROSOL-OT ON THE RESIDUAL AQUEOUS SOLUTION
CONTENT OF THE VACUUM DEWATERED COAL BED. (ISOCHRONAL
DATA AT AP = 1 ATM; COAL SIZE: 35 X 60 MESH) 91
47. LIQUIDrAIR INTERFACIAL TENSION OF AQUEOUS SOLUTIONS OF
AEROSOL A-196 AT 25°C. [36] 93
48. LIQUID-AIR INTERFACIAL TENSION OF AQUEOUS SOLUTIONS OF
TRITON X-114 AT 25°C 93
49. LIQUID-AIR INTERFACIAL TENSION OF AQUEOUS SOLUTIONS OF
DODECYL PYRIDINIUM CHLORIDE AT 25°C 94
50. LIQUID-AIR INTERFACIAL TENSION OF AQUEOUS SOLUTIONS OF
TRITON X-114 AT 25°C 94
51. LIQUID-AIR INTERFACIAL TENSION OF AQUEOUS SOLUTIONS OF
AEROSOL-OT AT 25°C 95
52. LIQUID-AIR INTERFACIAL TENSION OF FIXED AMOUNTS OF
AQUEOUS SOLUTIONS OF AEROSOL A-196 IN CONTACT WITH AN
ADSORBENT COAL SAMPLE, TOGETHER WITH THE DATA TAKEN
WITHOUT THE COAL PRESENT, FOR DETERMINATION OF THE CHANGE
IN SOLUTION CONCENTRATION ON ADSORPTION 100
53. LIQUID-AIR INTERFACIAL TENSION OF FIXED AMOUNTS OF
AQUEOUS SOLUTIONS OF TRITON X-114 IN CONTACT WITH AN
ADSORBENT COAL SAMPLE, TOGETHER WITH THE DATA TAKEN
WITHOUT THE COAL PRESENT, FOR DETERMINATION OF THE CHANGE
IN SOLUTION CONCENTRATION ON ADSORPTION 100
-------
FIGURE NO. PAGE
54. LIQUID-AIR INTERFACIAL TENSION OF FIXED AMOUNTS OF
AQUEOUS SOLUTIONS OF DODECYL PYRIDINIUM CHLORIDE IN
CONTACT WITH AN ADSORBENT COAL SAMPLE, TOGETHER WITH
THE DATA TAKEN WITHOUT THE COAL PRESENT, FOR DETER-
MINATION OF THE CHANGE IN SOLUTION CONCENTRATION ON
ADSORPTION 101
55. LIQUID-AIR INTERFACIAL TENSION OF FIXED AMOUNTS OF
AQUEOUS SOLUTIONS OF SODIUM DODECYL SULFATE IN CON-
TACT WITH AN ADSORBENT COAL SAMPLE, TOGETHER WITH
THE DATA TAKEN WITHOUT THE COAL PRESENT, FOR DETER-
MINATION OF THE CHANGE IN SOLUTION CONCENTRATION ON
ADSORPTION 101
56. LIQUID-AIR INTERFACIAL TENSION OF FIXED AMOUNTS OF
AQUEOUS SOLUTIONS OF AEROSOL-OT IN CONTACT WITH AN
ADSORBENT COAL SAMPLE, TOGETHER WITH THE DATA TAKEN
WITHOUT THE COAL PRESENT, FOR DETERMINATION OF THE
CHANGE IN SOLUTION CONCENTRATION ON ADSORPTION 102
57. ADSORPTION ISOTHERM FOR AEROSOL A-196 ON 35 X 60 MESH
COAL FROM AQUEOUS SOLUTION AT 25°C 103
58. ADSORPTION ISOTHERM FOR TRITON X-114 ON 35 X 60 MESH
COAL FROM AQUEOUS SOLUTION AT 25°C 104
59. ADSORPTION ISOTHERM FOR DODECYL PYRIDINIUM CHLORIDE
ON 35 X 60 MESH COAL FROM AQUEOUS SOLUTION AT 25°C. ... 105
60. ADSORPTION ISOTHERM FOR-SODIUM DODECYL SULFATE ON
35 X 60 MESH COAL FROM AQUEOUS SOLUTION AT 25°C ..... 106
xi
-------
FIGURE NO. PAGE
61. ADSORPTION ISOTHERM FOR AEROSOL-OT ON 35 X 60 MESH
COAL FROM AQUEOUS SOLUTION AT 25°C 107
62. STANDARD FREE ENERGY OF ADSORPTION FOR AEROSOL A-196
ON 35 X 60 MESH COAL AT 25°C 109
63. STANDARD FREE ENERGY OF ADSORPTION FOR TRITON X-114
ON 35 X 60 MESH COAL AT 25°C 109
64. STANDARD FREE ENERGY OF ADSORPTION FOR DODECYL PYRIDIN-
IUM CHLORIDE ON 35 X 60 MESH COAL AT 25°C 110
65. STANDARD FREE ENERGY OF ADSORPTION FOR SODIUM DODECYL
SULFATE ON 35 X 60 MESH COAL AT 25°C 110
66. STANDARD FREE ENERGY OF ADSORPTION FOR AEROSOL-OT ON
35 X 60 MESH COAL AT 25°C Ill
67. RELATIONSHIP BETWEEN THE LIQUID-AIR INTERFACIAL TENSION,
YLA, OF THE AQUEOUS SOLUTION PRESENT IN THE COAL BED AND
THE PRESSURE DIFFERENTIAL, AP, AT WHICH THE RESPECTIVE
DEWATERING CURVE HAD ITS MAXIMUM SLOPE 115
68. GENERAL BEHAVIOR OF THE STANDARD FREE ENERGY OF ADSORP-
TION FOR SURFACTANT ADSORPTION ON COAL 130
69. POSSIBLE ADSORBED BILAYER OF SURFACTANT MOLECULES WITH
TWO HYDROCARBON CHAINS PER MOLECULE 134
70. ADSORBED LAYER OF DODECYL PYRIDINIUM IONS SHOWING AN
ALTERNATE ORIENTATION FOR THE MOLECULES 138
71. CLOSELY PACKED ADSORBED LAYER OF DODECYL PYRIDINIUM IONS. 138
72. POSSIBLE "L" CONFIGURATION OF THE ADSORBED TRITON X-114
MOLECULE AT THE POINT OF CROWDING 142
xii
-------
FIGURE NO. PAGE
73. CHANGE IN THE ADSORBED AREA PER TRITON X-114 MOLECULE
TO ACCOMODATE MORE ADSORBED MOLECULES 142
74. PORTION OF A HYPOTHETICAL REAL COAL SURFACE WITH TEN
POSSIBLE ADSORPTION SITES 150
75. PORTION OF A HYPOTHETICAL REAL COAL SURFACE WITH AD-
SORBED WATER MOLECULES 150
76. PORTION OF A HYPOTHETICAL REAL COAL SURFACE WITH AN
ADSORBED LAYER OF ANIONIC SURFACTANT MOLECULES 152
77. PORTION OF A HYPOTHETICAL REAL COAL SURFACE WITH THE
ADSORBED ANIONIC SURFACTANT MOLECULES IN THE HEMI-
MICELLE STATE 152
xin
-------
I. INTRODUCTION
Overvi ew
The presence of water in coal prior to its utilization as a source
of energy can, for the most part, be considered as an undesirable impurity.
The reasons are almost too numerous to list; however, one needs only to
consider that the purpose of the coal is energy conversion, and water
present during this conversion consumes energy for evaporation. Further-
more, coal is normally transported many millions of ton-miles per yearly
production at a considerable cost. If ten percent of this coal transport
cost were due to water, virtually millions of dollars are wasted moving
unnecessary water back and forth across the country just to have that
water, in the end, reduce the utilization efficiency of the coal. Cer-
tainly the total elimination of the water problem under the present con-
ditions appears to be beyond our grasp since water not only forms a
structural part of natural coal, but water also inundates most of the
coal mines, it is the basis of our beneficiation schemes and furthermore
the weather continually drenches the stockpiles and open transport facil-
ities. A more detailed inspection of the problem, on the other hand,
does point to certain areas where an understanding of the physio-chemical
phenomena of water adhesion to coal could in some degree improve the
situation, particularly in the beneficiation of fine coal. Water in
coal may exist in three forms [1]:
a. Surface water - that water which lies on the surface of
the coal and has a vapor pressure with its surroundings
which is very nearly equivalent to the equilibrium vapor
pressure of water.
-------
b. Capillary water - sometimes called "inherent water" or
structural water" is that water which is absorbed into
the capillary structure of the coal itself. The capillary
o
diameters of coal (average 40 A) vary in size according
to the age of the coal, i.e., the average pore diameter
in anthracite is exceedingly small and that for western
coal (Wyoming) is relatively quite large. The water
contents in this class vary from much less than one per-
cent to in excess of thirty percent, respectively. The
vapor pressure of water in the capillary system is less
than that of the surface system of (a) by virtue of the
Kelvin equation [2].
c. Chemical water - those water molecules held in chemical
structure such as
This water is usually involved with certain minerals in
coal and is rarely measured since total dehydration will
only occur at temperatures well above the decomposition
point of coal.
An investigation of the dewatering of coal, in order to retain a prac-
tical stance, must be concerned only with the first class of water cited
above, surface water, since only the severest of thermal drying treat-;
ments will achieve a substantial degree of capillary dewatering.
The surface moisture on coal can vary from zero to greater than
thirty percent depending on the following variables [3]:
a. Rank of coal
b. Ambient relative humidity
-------
c. Mineral composition and concentration in the sample
d. Particle size distribution
e. Temperature
f. Mechanical treatments
g. Water impurities
h. Prior treatments, e.g., oxidation, chemical additives, etc.
all of which tend to reflect the surface states of the system and the
adhesion of water to coal.
For example, freshly mined anthracite coal which has a hydrophobic,
oil-like,surface [4] and a particle size distribution between 1.3 and 0.8 cm
will retain less than two percent water by simple mechanical agitation of
the mass on a vibrating screen since the void spaces between the particles
are too large to retain a significant amount of water by capillarity and
the coal surface tends to reject the water. On the other hand, well aged
Iowa coal with a particle size distribution of 0.25 mm x 0 would be most
difficult to dewater below twenty-five percent water even in a centrifuge,
since in this case we have a coal surface readily wetted by water (hydro-
philic) and a fine particle system which establishes a small interparticle
capillary system which will also aid in the retention of the water. The
effects of the interparticle capillary system on moisture retention in a
body have been under investigation by ceramists and soil engineers for
years [4]. Coal preparation engineers have also been faced with similar
problems in that shaker screens and vibrating screens will usually dewater
coal particles greater than or equal to one centimeter in diameter to
about five percent water, but for coal with size distributions below
-------
0.6 mm x 0,* and dependent on the amount of 0.07 mm x 0 material, the
dewatering process rarely achieves water contents below twelve percent.
Filtration and centrifugation which are usually used when large concen-
trations of 0.1 mm x 0 coal are present, incur similar problems even
though the forces on the liquid are higher. Discharge from these de-
vices rarely is below fifteen percent moisture under ambient conditions.
Deurbrouck [5] demonstrated that, in general, as the coal size was ,
reduced the pyritic sulfur (as the mineral Fe$2) release for eventual
separation from the relatively clean coal, was increased. Single stage
water froth flotation which has been commonly used in very fine coal
cleaning (0.6 mm x 0) has been recently expanded to a more efficient
system called "double froth flotation" [6]. In both cases the product
coal is less than 0.6 mm x 0 with a large concentration of very fine
material, the mass of which must be dewatered preferably to below ten
percent moisture to meet today's specifications. The tendency in the
industry today is to clean large quantities of 0.6 mm x 0 coal and bear
the expense of thermal drying.
The present investigation viewed the water problem from the stand-
point of the surface phenomena involved. The problem lies with the ad-
hesion of the Water to the solid surface and the molecular interactions
that occur along the coal-water interface. Furthermore,examination of
the list of variables "a" through "h" which influence surface moisture re-
vealed that the last one, i.e., prior treatments, and in particular "chem-
ical additives" seemed to be where the most control over this system could
be exercised. That is, if the simple addition of a chemical could to a
*This refers to the size distribution of particles which have passed through
a sieve whose hole size was 0.6 x 0.6 mm. The maximum size is 0.6 mm and
the minimum size is zero, i.e., sub-micron dust particles.
-------
large degree alter the affinity of water for the coal surface, then the
water retention might be readily decreased. Furthermore, if this could
be accomplished at a very low concentration of that chemical, then the
economics would be favorable for utilization. The above direction was
also influenced by the knowledge that certain classes of ch,emical com-
pounds are known to effect remarkable changes in interfacial properties.
Also, some attempts, reported in the literature, have been made to use
them for the improvement of coal dewatering as shall be discussed in de-
tail below. These compounds are known as surfactants. The present in-
vestigation is concerned with their effect on coal dewatering and the
reasons for which their influence is felt.
Surfactants were chosen specifically for two reasons. First, they
are known to decrease the interfacial tension of the air-water interface
drastically at extremely low concentrations. In terms of overcoming the
capillary forces of a bed of small granulated particles this interfacial
tension is proportional to the work required to displace the liquid. Any
improvement is worthwhile.
Secondly, similar molecules have been used by the mineral industry
in the area of ore flotation [7]. In this case the molecules are absorbed
at the surface of a mineral and render it hydrophobic such that through
the attachment of air bubbles the particles are caused to float to the
surface of the liquid where they are recovered. Such molecules change
a hydrophilic surface to hydrophobic.
-------
Background
Dewatering is a process whereby liquid held by capillary forces within
the interparticle voids of a bed of fine particles is displaced by the
application of desaturating forces. These forces can be applied, for ex-
ample, by centrifugation in which case they act on the body of the liquid
to pull the liquid out, or through the presence of a gas pressure differ-
ential across the bed, which if an over-pressure pushes the liquid out or
if a vacuum, the liquid is drawn out. In each case there is always a certain
amount of liquid retained in the bed. Investigations of dewatering may ad-
dress either the process of achieving final water content or studying the
final value irrespective of the process, or both.
In general, factors which can influence the dewatering are particle
size, particle shape, mode of particle packing, dimensions of the bed,
density and viscosity of the fluids (i.e., displacing gas and displaced
liquid), temperature, pressure gradient, rate of displacement, and the inter-
facial tensions at the liquid-air and solid-liquid interfaces [8]. Consid-
erable effort has been involved in the investigation of the above variables
for a variety of bed materials, but surprisingly little work has been ap-
plied specifically to coal.
Dewatering, fn general, can be considered from the viewpoint of the'
theory of flow through saturated porous media, specifically packed beds of
particles. Briefly, the theory utilizes the modified Darcy equation:
(1)
-------
where
Q = fluid flow
A = cross-sectional area of the bed
AP = pressure differential across the bed
n = viscosity of the fluid
L = bed thickness
K = bed permeability
2
= I/a = m ek
a = specific bed resistance
m = mean hydraulic radius
e = porosity
k = empirical constant
Equation (2) was developed by Kozeny [9] and confirmed by Carman [10]
for beds of many solids. The theory becomes applicable to dewatering when
two-phase flow is considered, where one is liquid (water) and the other is
air. This approach to dewatering has been considered by Browne!1 and Katz
[11] and Dombrowski and Brownell [12] for a variety of materials and by
Silverblatt and Dahlstrom [13] for beds of fine coal in particular.
However, as pointed-out by Gray [3] this semi-empirical approach to
dewatering is not satisfactory for a complete description of the situation.
A major objection was that the flow theory need not consider the actual
distribution of pore sizes of the packed bed, to be applied. It just re-
solves the entire distribution into a single "mean hydraulic radius", m,
and characterizes the bed by an overall permeability K. The flow theory,
therefore, describes flow "through" a bed not "from" a bed. Actually, as
dewatering proceeds, the system through which the liquid flows changes
according to the distribution of the pores of the bed yet to be emptied.
Initially dewatering is governed by the larger pores and the retention of
-------
8
the liquid by the smallest pores in addition to other modifying factors.
Oroe of these other factors which is not adequately considered by the
flow theory is the nature of the interfaces that are involved in the sys-
tem, i.e., the liquid-gas, liquid-solid, and solid-gas interfaces. This
is especially critical in dewatering, as there is contact of all three
phases at the point where the liquid is retracting from the solid. Further-
more, interfacial tensions are thermodynamic quantities and as such are
defined under equilibrium conditions which may not be reflected by con-
sideration of a dynamic flow situation.
For such reasons and especially because specific surface phenomena
are not appreciated by the flow treatment of dewatering it was decided,
like Gray, to approach dewatering from the viewpoint of capillary theory
which does allow the treatment of interfacial effects. This approach will
be treated in detail below.
Only a few attempts have even considered the role of specific inter-
facial phenomena involved in coal dewatering. Those investigations
[3,13-15] which did involve interfacial parameters most usually were
limited to the liquid-air interface and the effect of decreasing its
tension. Some research employed the use of a class of substances known
as surfactants in this endeavor; however, little attention was given to
their effect at the solid (coal)-Tiquid (water) interface. Some specu-
lations were made [3,13-15].
Silverblatt and Dahlstrorn [13] in 1954 investigated the effect of
liquid viscosity and the liquid-air interfacial tension on the moisture
content of a dowatered fine-coal filter bed. Their investigation involved
the addition of the surfactants Tcryitol-CW and Aerosol-OT in solution
to the feed coal. Thoy found that thn addition of those surfactants re-
-------
suited in a lower residual moisture content of the dewatered coal product,
and speculated that the effect was probably not due to the change in the
liquid-air interfacial tension, which they found to have a less pronounced
effect than viscosity on the dynamics of dewatering, but rather from some
kind of solid surface reaction between the surfactants and the coal.
Gray [3] in 1958 investigated the dewatering of fine coal from the
point of view of capillary theory and measured the equilibrium water con-
tents of saturated coal beds after incrementally applying a pressure dif-
ferential across the bed. The dewatering rate, the dewatering behavior
of various narrow size distributions, and the addition of several floc-
culants, oils, and wetting agents (surfactants) were investigated. Gray
found that dewatering was improved by each of the above additions; and
he alludes to the fact that the efficacy of several highly impure oils
in dewatering may be due to their content of surface active materials.
Furthermore, the wetting agents, which were assumed only to lower the
liquid-air interfacial tension of the liquid, also improved the final
moisture content of the beds. Gray also noted that the liquid-air inter-
facial tensions of the expelled solutions were higher than that originally
X
introduced to the bed and concluded that the wetting agents were removed
from the solution by strong adsorption to the coal surface.
Dolina and Kaminskii [14] in 1971 reported the effect of the addition
of a number of surfactants each on the residual moisture content of vacuum
filtered coal beds. They found that the residual water contents were as
likely to increase as well as decrease depending upon which surfactant
was used. Again the authors only speculate that the effect might have
something to do with adsorption of the surfactants by the coal, and not
-------
10
be wholly dependent on the decrease in the liquid-air interfacial tension.
Finally, Nicol [15] recently (1976) has observed improvement in de-
watering of coal with the addition of surfactants and also speculates on
a solid surface interaction.
However, no controlled experiments have been found in the literature
concerning the adsorption of surfactants by coal or an attempt at a correla-
tion with gross dewatering phenomena.
-------
11
II. CAPILLARY THEORY OF DEWATERING
If we consider a bed of fine particles, such as coal, less than a
millimeter in diameter, when partially immersed in water, the bed will
act as a bundle of capillary tubes and cause the water to be drawn up
into that system of pores. The removal of this water involves the ap-
plication of a force to initiate flow out of the system.
If we consider the general case for one capillary, the force per unit
area necessary to counteract the capillary pressure within a capillary of
radius r is given by the capillary equation
ZY, r cos 0
AP = - - (3)
where AP is the pressure differential across the capillary, Y,Q is the
interfacial tension of the liquid at the gas (air) interface and 0 is
the receding contact angle between the capillary wall and the liquid [16].
The contact angle, 0, is defined by Young's equation
YSG = YLG cos e * YLS (4)
at the point of three phase contact, as shown in Figure 1 where Yge is
the interfacial tension of the solid-gas interface and YIC is that of the liquid-
solid interface. There is considerable evidence in the literature [17]
to permit the assumption that the receding contact angle of v/ater on coal
is zero, and the capillary equation may, therefore, be written as
AP = ^ (5)
-------
12
>LG - -
_ —_- LIQUID i_ ~_
SOLID
FIGURE 1. INTERFACIAL TENSIONS AT THE POINT OF THREE PHASE CONTACT
IN A CAPILLARY.
-------
13
We are now going to assume that a bed of coal particles acts as a
bundle of straight capillaries so that the above equation may be applied.
Later we will examine the validity of this assumption. One recognizes that
capillaries of radius r, within the coal bed, will drain with the appli-
cation of an applied pressure differential, AP, just greater than 2y/r
while no drainage will occur when AP = 2y/r since the capillary and applied
forces are equal.
If there is a wide distribution of particle sizes in the bed there
will also be a wide variation of pore radii within the bed; and consequently,
a wide range of pressures will be necessary to dewater the entire system.
A plot of the amount of retained water in the coal bed as a function of
applied pressure differential, i.e., a dewatering curve, will reflect the
real pore radii distribution. If we assume, for instance, that the par-
ticles are uniform spheres and are arranged in a close packed structure,
then the radius, r., of the enclosed spherical void surrounded by the tetra-
hedral arrangement of particles is 0.225 rB> where rg is the solid particle
radius, so that the ratio of particle radius to pore radius is 4.44. If
there is a range of r.,, there will also be a range of rA; consequently, for
a broad particle size distribution, a dewatering curve would resemble Fig-
ure 2. However, if the particle size distribution is narrowed, as was es-
tablished for the experiments reported herein, then rA will be relatively
constant, and therefore a great amount of dewatering will occur in an ex-
tremely narrow pressure range as shown in Figure 3, which would be a step
function in an ideal case, where r was single valued.
Due to the inverse proportionality between r and AP in Equation
(5), a coal bed with uniform large pores, i.e., large particles, will de-
-------
14
Wt.%
RETAINED
LIQUID
APPLIED PRESSURE DIFFERENTIAL^
FIGURE 2. DEWATERING CURVE FOR A BED OF PARTICLES WITH A BROAD SIZE
DISTRIBUTION.
wr.%
RETAINED
LIQUID
APPLIED PRESSURE DIFFERENTIAL,AP
FIGURE 3. DEWATERING CURVE FOR A BED OF PARTICLES WITH A NARROW SIZE
DISTRIBUTION.
-------
15
water at lower values of applied pressure differential and a bed with
small pores at higher values of applied pressure differential. Equilibrium
dewatering curves like that of Figure 3, were, in fact, observed in the
experiments performed.
-------
16
III. COAL
Coal can be considered to be composed of two classes of constituents.
The first, the carbonaceous material, is a highly complex matrix of or-
ganic molecules with an empirical formula which may be roughly given as
Ci,cHg7OgNS for a typical low volatile bituminous coal [18]. The second
constituent comprises the inorganic mineral matter, for example, iron
pyrite, illite, silica, montmorillonite, and various other slates, clays,
and rock matter.
" Coal is the result of the natural application of heat and pressure,
without air, on the vegetation found in tropical prehistoric forests.
Coalification, i.e., the degree of transformation, appears to follow the
sequence beginning with peat, through sub-bituminous, bituminous, and semi-
anthracite to anthracite and eventually to graphite. The subcategories
are called the rank of coal. The main change through the series is in
the relative amount of carbon present, which increases in the order pre-
sented.
Based on evidence from standard analytical techniques a hypothetical
organic structure of low volatile bituminous coal can be imagined. An
element distribution and sketch of such a structural unit is shown in
Figure 4, as conceived by Bailey [18]. The structure consists of clusters
of condensed aromatic rings linked by cycloaliphatic rings, with peripheral
etherial oxygen and hydroxyl groups. The organic part of the coal surface
alone therefore can present three distinct regions.
On a macroscale, the organic matter in coal is also considered as
composed of a multiphase system called macerals [19] which are further
-------
17
CH.
FIGURE 4. POSSIBLE ORGANIC STRUCTURE OF BITUMINOUS COAL
-------
18
classified as fusain, clarain, durain, and vitrain. Although these
classes are based on morphological appearances they certainly reflect
gross modifications in their respective chemical systems relative to
hydrocarbon distributions as well as mineral matter contents.
The inorganic mineral matter content of coal also consists of
numerous unique phases, part distinct from the hydrocarbon system and
part directly associated with the complex molecular arrangements.
A common example of the mineral matter present in coal is contained
in a group called illite. Illite is considered a variation of muscovite,
which is a hydrated silicate of aluminum, with replacement atoms of po-
tassium or magnesium. The clay minerals possess complex crystal struc-
tures and atomic arrangements. For example, montmorillonite, also a
relative of illite, is a clay which is made of layers of silicon oxide
(SiO.) tetrahedrally arranged such as to share corners with octahedrally
arranged aluminum ions (Al ) which in turn have coordinated oxygen and
hydroxyl groups. Molecules of water lie in between the layers of sili-
cate sheets.
A prevalent non-oxide mineral found in coal is iron pyrite, FeS2,
an ionic solid, and a crystal modification of FeSp called marcasite.
All the crystalline solids found in coal can present their various
crystal planes, structures and compositions to their free surfaces.
A sample of coal then may be regarded as being an exceedingly hetro-
gencous material with surface characteristics which must be at least as
complex as the bulk chemistry.
Due to the fact that there is an abrupt discontinuity at a surface,
the balanced chemical bond configurations which exist in the bulk solid
-------
19
do not necessarily exist at the free surface. Charge neutrality is al-
tered, and bonds may be left "dangling" at the surface [20]. The surface
attempts to satisfy this imbalance and lower the overall free energy; how-
ever, certain surface potentials remain. These potentials are reflected
in the behavior of liquids in contact with the free surfaces and are
measured as a zeta potential [21], i.e., the potential difference be-
tween a point well removed from the surface, in field free space, and a
point in the liquid shear plane adjacent to that surface. Note should be
taken that the zeta potential only refers to the liquid adjacent to the
surface and not the potential determining ions or atoms in the solid sur-
face itself. Cambell and Sun [22] have determined the zeta potential of
whole coal and its lithotypes and found that it varies for each type from
-30 millivolts to 0 millivolts in a neutral aqueous solution. Silica [23]
is reported to have a negative and alumina a positive potential [23].
Pyrite has a negative zeta potential in neutral aqueous solution and its
change with pH has been found similar to that of coal [24]. The components
of the coal exhibit different signs and magnitudes of surface potential;
and therefore, the surface system of coal may be regarded as electro-
statically heterogeneous.
Coal should be considered as a solid gel consisting of a framework
enveloping many pores throughout the system. These pores or capillaries
vary in diameter from approximately 2 to 1000 angstroms in diameter with
o
an average of about 40 A depending on the rank of the coal. The system
p
generates between 100 - 200 m /g of surface area again depending on the
rank (age) of the coal. The possible interaction of the coal with any
other molecules, for example during an adsorption process, could well be
influenced by the pore distribution, that is, molecules larger than the
-------
20
pore diameter cannot penetrate the outer surface while smaller molecules
may envelope the whole system. The problem is carefully examined in a
recent paper [25] where the discrepancy of coal specific surface areas
determined by gas adsorption (BET) for different gases was considered. The
problem primarily lies in the fact that there is a distribution of pore
sizes which apriori is unknown.
In conclusion, therefore, the coal surface can be considered as being
chemically, electrostatically and topographically heterogeneous.
-------
21
IV. SURFACTANTS
Surfactants are ionic or polar derivatives of high molecular weight
*?
hydrocarbons. They are rather unique in that they are amphiphilic sub-
stances, that is one part of the molecule has an affinity for water,
which is termed hydrophilic, and another has an oil affinity, which is
j
termed hydrophobic. The hydrophilic portion is often referred to as
the "head" and the hydrophobic hydrocarbon chain as the "tail" of the
molecule. The name surfactant is just a convenient contraction of "sur-
face active agent".
The amazing property of surfactants is that as solutes in aqueous
solution, even at extremely small concentrations, they produce a drama-
tic decrease in the liquid-air interfacial tension of the solution, e.g.
72 dyn/cm for pure water to 30 dyn/cm for concentrations in water as
-4
small as 10 mole fraction. An example of a surface active agent is
common toilet soap or household detergents. A solution-air interfacial
tension curve as a function of concentration for typical aqueous sur-
factant solutions is illustrated in Figure 5.
Surfactant molecules are usually anisotropic in shape. Many are
usually considerable in length but limited in width. An example is
sodium dodecyl sulfate shown in Figure 6.
There are two main classes of surfactants: ionic and non-ionic.
s
The ionic class is composed of two sub-classes, anionic and cationic.
After ionization of a surfactant molecule in water the charge on the
hydrocarbon portion of the molecule establishes the nomenclature of
the parent molecule. For example, an anionic surfactant is sodium
-------
22
>LA
(dyn/cm)
In C
FIGURE 5. TYPICAL BEHAVIOR OF THE SOLUTION-AIR INTERFACIAL TENSION OF
AQUEOUS SURFACTANT SOLUTIONS AS A FUNCTION OF SURFACTANT
CONCENTRATION.
HYDROPHOBIC PORTION
HYDROCARBON CHAIN
HYDROPHIUC PORTION
HEAD ASSOCIATED
GROUP |ON
H-C-C-C-C-C-C-C-C-C-C-C-C-O-S-O
i i i i i i i i i i i i n ;
HHHHHHHHHHHH 0 I
No"
FIGURE 6. EXAMPLE OF A SURFACTANT MOLECULE: SODIUM DODECYL SULFATE,
-------
23
dodecyl sulfate, where
is the anionic portion. An example of a cationic surfactant is dodecyl
pyridinium chloride where
- N
is the cationic portion.
Non-ionic surfactants usually have hydroxyl groups and/or polyoxy-
ethylene chains as the hydrophilic portion of the molecule. Polyoxy-
ethylene p - tertoctylphenyl ether is an example:
CH3(CH2)7/' NVO(CH2CH20)10 H
where the (CH2CH2^in grouP is the polyoxyethylene chain.
Another unique property of surfactants is that as their number in
aqueous solution is increased a point is reached when it is energetically
favorable to form an aggregate, or precipitate , of the monomers. This
aggregate or colloidal particle is called a micelle. The concentration
at which this occurs is termed the "critical micelle concentratic-." and
is frequently abbreviated CMC. One example of the structure of a micelle
is shown in Figure 7. The hydrophobic hydrocarbon chains are directed
toward the inside of the micelle remote from the water-based phase, and
the polar head groups are on the outside of the particle as illustrated.
-------
24
FIGURE 7. POSSIBLE CROSS-SECTION OF A SPHERICAL MICELLE.
FIGURE 8. POSSIBLE CROSS-SECTIONS OF BILAYER MICELLES.
-------
25
Micellation is actually a phase separation process and the product is
not necessarily spherical. Included among the configurations possible
[26] is also a bilayer arrangement as shown in Figure 8. Micelle forma-
tion is a direct result of the dual nature of the molecules with their
antipathetic affinity for water. Another consequence of the dual
nature leads to adsorption of the molecules at the solution-air inter-
face. The hydrocarbon portions of the molecule oppose the solubility
of the whole molecule and try to alleviate the high hydrocarbon-water
interfacial energy by moving to the free surface where the hydrocarbon
portion of the molecule extends as far out of the water as is physically
possible. This is also the case at any other interface in contact with
the solution, e.g. any solid-liquid interface. When adsorption at in-
terfaces can no longer reduce the free energy of the system as the
solute concentration increases the hydrocarbon chains associate in the
bulk solution proceeding to "micellation".
Griffin [27] and Davies [28] suggested a utilitarian way of classi-
fying surfactants based on the relative sizes and strengths of the hydro-
phi! ic and lipophilic (oil preference) groups of the molecules which is
called the Hydrophile-Lipophile Balance, abbreviated HLB. The HLB is
a number calculated for each surfactant as
HLB=z(values for hydrophilic groups)-E(values for hydrophobic groups) + 7
Some group values are given in Table I. The HLB system was devised to
organize much of the disjointed information about surfactants in order
to formulate a rational method of using them for controlling oil-water
-------
26
emulsions. Although detailed use of this concept was not applied in
this study, the HLB index does provide a way of comparing the relative
hydrophobicity of surfactants; and consequently, a partial measure of
their willingness to adsorb at interfaces and once adsorbed a measure
of the water affinity of its adsorbed film.
TABLE I
GROUP HLB VALUES [29]
HYDROPHILIC GROUPS
-S04 Na
-COO K
-COO Na
Sulfonate
-N (tertiary amine)
-COOH
-OH (free)
-0-
-OH (sorbitan ring)
HLB LIPOPHILIC GROUPS HLB
38.7 -CH- '
21.1 -CH2-
19.1 -CH3-
11.0 -CH=
*
0.475
9.4
2.1
1.9
1.3
0.5
-------
27
V. ADSORPTION
Stern-Grahame Theory of Adsorption
The surfactant molecules in a water-solid system are greatly influenced
by the charge state of the solid surface, i.e., the electrical double layer
which is formed in the solution.
According to the Stern treatment [30], the electrical double layer re-
gion that lies at the solid-liquid interface can be envisioned as a compact
and a diffuse type of ion arrangement. If the solid surface has a net elec-
tric charge then that field in an ionic solution will attract a balancing
charge from ions in that solution. The ions form two regions adjacent to
the surface. The first consists of adsorbed ions which is called the "Stern
layer". The second is a more diffuse layer which is known as the "Gouy
layer". The sum of the charges in all of the layers counter balances that
charge in the solid surface. A sketch of the arrangement is shown in Figure
9, where the solid is shown with a positive charge due to those ions of the
solid at its surface. Adsorbed at the surface sites are the negative ions
from the solution, i.e., the Stern layer. The charge density in the solu-
tion is presumed to decrease exponentially from the solid surface. Also
shown is the probable layer of shear in which a measureable potential is
obtained which is called the zeta potential.
Grahame [31] modified the Stern theory by subdividing the Stern layer
into two planes depending on the state of hydration of the adsorbed ions.
These are known as the inner and outer Helmholtz planes, and are shown in
Figure 10.
-------
28
STERN LAYER
GOUY LAYER
\ BULK SOLUTION -
LAYER OF SHEAR
FIGURE 9. ILLUSTRATION OF THE ELECTRICAL DOUBLE LAYER
-------
29
SOUP
POTENTIAL
IV INNE
INNER HELMHOLTZ PLANE
OUTER HELMHOLTZ PLANE
DIFFUSE REGION
0
DISTANCE FROM SOLID
FIGURE 10. THE ELECTRICAL DOUBLE LAYER: VARIATION OF THE POTENTIAL
WITH DISTANCE, SHOWING THE INNER AND OUTER HELMHOLTZ
PLANES.
-------
30
Thefmore specific treatment established by Grahame was used in the
analysis of the adsorption of the surfactant molecules at the coal-water
interface. According to this treatment adsorption of ions at an elec-
trical double layer can be described according to the following relation-
ship, for each type of ion adsorbed:
r = 2r C exp(-AG°DS/RT) (6)
where r is the number of moles of the ion adsorbed per square centimeter
of surface, r is the radius of the non-solvated ion, C is the concentra-
tion of the ion in solution, AGJL- is the standard free energy of ad-
sorption, R is the gas constant, and T is the absolute temperature.
The standard free energy of adsorption, AG^, is defined as
where (y°)s is the chemical potential of the i species in standard state
at the surface and y? is that chemical potential in the bulk solution at
standard state.
The standard free energy of adsorption can be considered [23] as the
summation of all the interaction energies involved in the adsorption; for
example,
-------
31
46che,n + Sc + (10)
where $ has been evaluated [32] to be about 0.6 kcal/mole for each -CH2~
component of the hydrocarbon chain and n is the number of components in
that chain.
-------
32
Adsorption From Solution
The experimental measurement in adsorption from solution is the change
in concentration of the solution from which molecules are adsorbed. Since
concentration changes are involved, one must be aware of the fact that more
than one component is present and able to adsorb. There exists, therefore,
a competition between components for adsorption sites on the solid.
Let us define adsorption from a two component solution on a more rig-
orous basis. Consider a solution of n, moles of solute and n~ moles
of solvent where
n1 + t\2 = n (11)
and the mole fractions are given by
nl
-
and X-j + Xg = 1. If to this solution we add w grams of a second phase,
e.g.,a solid, some fraction of each component of the solution will adsorb to
the solid surface, i.e., n* moles of component 1 per gram of solid and
n| moles of component 2 per gram of solid. The adsorbed molecules will
deplete the solution phase of their respective components to new values n.!
and n» such that
-------
33
Let us define
n-j + n2 = n (14)
where n is the total number of moles remaining in the depleted solution
phase, and
x;. ji
where X-j is the mole fraction of solute remaining in the solution phase.
Therefore, we can. write for the change in solution solute mole fraction on
adsorption
• ni ni
AX = X, - XT = -1-- 4 (16)
or
(17)
(n, + r\ + nw + nw) (n + rt)
.
-------
34
and accumulating terms and rearranging
i2 , , , , ,2 . , i s i s
n, + n,n, + n,n,w + n^n^w - n, - n,n9 - n,n,w - n,n7w
AX = -! - L2- - 1 1 . 2 ] - r1 - — - — - — (18)
Rearranging
(n, + n2) (n1 + n2 + nw + nw)
' s ' s
n-n^w - n,n,,w
AX = -2Jn - M- (19)
(^ + n2)n
1 s ' s
«n, - n,n
w nl + n2
(2Q)
M = n*(-^-r) - n!(-A-r) (21)
w nl + n2 1 + n2
- n« n-,
*(4-> - n|(4) (22)
n fc n
s ' s '
= n!X2 " n2Xl (23)
• • •
Also X2 = (1 - X.j), so in terms of X-j, the mole fraction of solute
remaining in solution,
AX _ _S/T v'\ «SV'
T" nl(1 • Xl} - n2Xl (24)
-------
35
The term n AX/w plotted as adsorption provides an isotherm composed of the
sum of the two individual isotherms of components 1 and 2. A problem exists
in separating the two isotherms. Since there are two unknowns, n! and
n|» without more information relating them no solution is possible.
However, in the case of very dilute solutions an approximation can be
i
made. When X, is very small the number of moles of component 1 adsorbed
is approximately
s „ nAX
nl - ~w~
s s '
Even if n« is large, the product n^ is small enough in relation
_ i
to n?(l - X-j) to be dropped. Furthermore
(26)
because (1 - X,) is approximately one. The situation is well approximated
I n
for values of X, < 10" , a range of concentrations which conveniently in-
cludes most surfactant behavior.
-------
36
Measurement of Concentration Change
The number of moles of solute adsorbed on a solid from dilute solu-
tion can be obtained from a measurement of the change in solute concen-
tration which in this case is the concentration of a specific surfactant
-4
at a very small concentration, i.e., usually much less than 10 mole
fraction. Any number of the various techniques are available, e.g., re-
fractonetry, colorimetry, titrimetric analysis, densitometric analysis,
the use of radio active tracers, gravimetric analysis or the measurement
of the solution- air interfacial tension.
The concentration of a surfactant in aqueous solution changes the
liquid-air interfacial tension of that solution in an orderly reproducible
manner. The liquid-air interfacial tension can be measured reproducibly
and the changes in interfacial tension with surfactant concentration are
usually on the order of 10 dyn/cm per order of magnitude change in mole
fraction which is sufficiently large to establish excellent precision as
the reproducible error in the interfacial tension is in the range of t
0.20 dyn/cm.
The adsorption process removes surfactant molecules from the bulk solu-
tion, thus causing a decrease in concentration and a subsequent increase in
the interfacial tension. By carefully measuring the change in the liquid-
air interfacial tension (Y^) versus surfactant concentration at a known
temperature and comparing these data with similar data obtained in the
presence of a known weight of coal of a known size range, surfactant ad-
sorption curves can be developed.
-------
37
Thus, the number of moles of surfactant adsorbed per unit weight of
coal, nf, at constant temperature (room temperature) was plotted as a
function of the surfactant concentration, x, of the aqueous solution into
which the coal was placed. This was reported as the room temperature sur-
factant adsorption isotherm.
The interfacial tension technique was also used by Garner, McKie,
and Knight [33] to determine the adsorption of aliphatic alcohols on
charcoal, and by Fowkes [34] in the adsorption of surfactants on cotton
fibers.
-------
38
VI. MATERIALS
Coal
The coal that was used for all the subsequent experiments was ob-
tained from a sample of Pittsburgh bed coal collected in the U.S. Bureau
of Mines (now U.S. Department of Energy) Bruceton coal mine, Pittsburgh,
Penna. The sample was prepared by the Bureau by stage crushing through
a roll crusher and a hammer mill to obtain a nominal size distribution
of 35 mesh (420 ym diameter) x 0.
The "as received" sample was wet-screened to give a sufficient
amount of 35 x 60 mesh (420 x 250 ym diameter) particles to conduct all
of the dewatering and adsorption experiments. It was recognized that
the phenomena under investigation were dependent on particle size dis-
tribution, but to examine this variable at this time was beyond the
scope of the study.
An analysis [35] of the 35 x 60 mesh product coal showed that it
contained 93% by weight coal and 7% mineral matter. The ash content of the
total was found to be 8.47 - 0.43%. The ash content of the coal product
from a gravity separation at 1.50 specific gravity was 4.62 - 0.05%.
The internal moisture content of the separated coal was 1.62 - 0.01%.
The internal moisture content of the total sample was 1.71 - 0.06%.
-------
39
Surfactants
Five surfactants were chosen for this investigation: three anionic,
one cationic and one non-ionic. The choice was arbitrary as there was
no well based understanding of the interaction of surfactants with coal,
j
yet it was felt that at least one representative from each class should
be examined for comparison. The two additional anionic surfactants were
chosen for reasons which became obvious as the investigation proceeded.
1. Aerosol A-196 (anionic)
Aerosol A-196 is the trade name for sodium dicyclohexyl sulfosuc-
cinate produced by the American Cyanamid Company. The empirical for-
mula is C-,gH^cSOyNa. The structural formula is shown in Figure 11.
Figure 12 shows a photograph of the molecular model of the anionic por-
tion from a side and top view. The main hydrocarbon portion of this
molecule consists of the two cyclohexyl groups which are joined to the
sulfosuccinate core. The molecular weight of Aerosol A-196 is 384.4
atomic mass units. The sample supplied to the laboratory was in the
form of an extruded solid. It contained approximately eighty-five
percent by weight of active ingredient with the balance as water. All
concentrations reported for solutions of this surfactant were corrected
to present the actual concentration of active ingredient only. Other
impurity concentrations were not determined.
2. Triton X-114 (non-ionic)
Triton X-114 is the trade name for a product of the Rohm and Haas
Company which is octylphenoxypolyethoxy ethanol. The empirical formula
1s CgH,,-^ y-(OCN2CH2)7_gOH. The expanded structural formula is
-------
40
CH2-COO
CH -COO
FIGURE 11. STRUCTURAL FORMULA OF AEROSOL A-196 (SODIUM DICYCLOHEXYL
SULFOSUCCINATE).
-------
TOP VIEW
SIDE VIEW
FIGURE 12. MOLECULAR MODEL OF A DICYCLOHEXYL SULFOSUCCINATE ION,
(SIDE AND TOP VIEW).
-------
42
shown in Figure 13 and Figure 14 is a photograph of a molecular model
where the polyoxyethyl ene chain is slightly coiled. The average mole-
cular weight of Triton X-114 is 536 atomic mass units. The product
contains a number of homologs containing either seven or eight units
of po1yoxyethyl ene (CH2CH20)n- The sample supplied to the laboratory
was reported to be one hundred percent active ingredient in the form
of a clear pale liquid. Any impurity concentrations were not determined,
3. Dodecyl Pyridinium Chloride (cationic)
The cationic surfactant dodecyl pyridinium chloride has an empiri-
cal formula of C,yH-0NCl. The structural formula is shown in Figure 15
and Figure 16 is a photograph of a molecular model of the cationic por-
O,
. The molecular weight is 283.9 atomic mass
units. The sample supplied to the laboratory was reported to be one
hundred percent active ingredient which was in the form of a straw
colored powder. Sample purity was not investigated.
4. Sodium Dodecyl Sulfate (anionic)
Another anionic surfactant used was sodium dodecyl sulfate,
also known as sodium lauryl sulfate. Use of this surfactant is fre-
quently reported in the literature. The empirical formula is
C-,0^25^4Na- The structural formula is shown in Figure 17 and Figure
18 is a photograph of a molecular model of the anionic portion. Its
molecular weight is 288.4 atomic mass units. The sample was supplied
to the laboratory in the form of a white powder. It was of laboratory
grade from the Fisher Scientific Company. Impurity content was not
analyzed.
-------
43
J~\
CH3-CH-,-CH2-CH2-CH2-CH
-------
44
CH3-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2 -CH2-CH2 -N
FIGURE 15. STRUCTURAL FORMULA OF DODECYL PYRIDINIUM CHLORIDE
FIGURE 16. MOLECULAR MODEL OF A DODECYL PYRIDINIUM ION.
-------
45
C H3 - CH2-CH2I2 -CH2-CH2-CH2 -SG4~] [Ma"1"
FIGURE 17. STRUCTURAL FORMULA OF SODIUM DODECYL SULFATE.
FIGUKL 10. MOLECULAR MODFL. OF A DOPLCYL SULFATC ION.
-------
46
5. Aerosol-OT (aniom'c)
Aerosol-OT, another aniom'c surfactant, was chosen also because
of its reputation as a commercial wetting agent. Aerosol-OT is the
trade name for sodium di(2-ethylhexyl) sulfosuccinate manufactured by
the American Cyanamid Company. The empirical formula is C2Q^^Qj^a.
The structural formula is shown in Figure 19. Figure 20 is a photo-
graph of a molecular model of the aniom'c portion from a side and top
view. This surfactant is essentially the same as Aerosol A-196 except
for the configuration of the hydrocarbon groups. The molecular weight
of Aerosol-OT is 444.5 atomic mass units. The sample supplied to the
laboratory was in the form of a soft waxy solid. The sample was re-
ported to be one hundred percent active ingredient, i.e., no water or
alcohol as a suspending medium. Purity of the sample was not investi-
gated.
-------
47
CH2-CHs
I
-COO-CH2-CH-CH2-CH2-CH2-CH,
| CH2-CH5
CH -COO-CH2-CH-CH2-CH2-CH2-CH,
so;] [NO*
FIGURE 19. STRUCTURAL FORMULA OF AEROSOL-OT (SODIUM DI (2-ETHYLHEXYL)
SULFOSUCCINATE).
-------
48
TOP VIEW
SIDE VIEW
FIGURE 20. MOLECULAR MODEL OF A DI (2-ETHYLHEXYL)
ION, (SIDE AND TOP VIEW).
•ULFOSUCCINATE
-------
49
VII. EXPERIMENTAL
Introduction
The purpose of the placement of surfactants on the surface of coal
was to modify its hydrophobicity and allow for more complete dewater-
ing of that modified coal. Let us now turn to the dewateringmeasure-
ments that were performed. Two techniques were utilized: one we call
equilibrium, in which the saturated coal bed was dewatered over many
hours with very small incremental changes in applied pressure differ-
ential across the bed, such that at any instant the dewatering process
could be stopped, and reinitiated without a change in dewatering rate
with pressure. This method, which has been also used for the testing
of soils, consisted of continuously recording the equilibrium retained
water content of a partially saturated bed of fine coal particles across
which a measured pressure differential existed and as the pressure dif-
ferential was increased in small increments a curve of retained water
content as a function of applied pressure differential was obtained.
The pressure differential was obtained by the application of air-pressure
to the top of the bed.
The second method, called non-equilibrium, relied on a fixed bed
pressure differential across the bed established by the application of
a vacuum such that the bed moisture could be measured at fixed times
of exposure. The non-equilibrium dewatering experiments that were
conducted were used for evaluating only the final water contents of a.
dewatered bed as a function of surfactant concentration in the water.
A gravity drained slurry was dewatered by the application of constant
vacuum from below for various time intervals. Hence the final water
-------
50
content was determined at certain exposures (pressure x time). This
technique was employed to increase and support other dewatering data
at certain surfactant concentrations.
Although the "non-equilibrium (vacuum) dewatering" technique was
developed by the author and the data accumulation directed by same, it
should be recognized at the outset that a large fraction of the data
collected in the experiments under this heading were done so by Chi [36]
of this laboratory. The details of the experiments and the results were
included as these data developed an exceedingly strong support for the
proposed molecular model of surfactant-coal surface interaction that
was the focal point of this dissertation.
-------
51
Pressure Dewatering (equilibrium)
The equilibrium dewatering experiments were performed on 30 gram
samples of the narrow size distribution coal which were formed into a
cylindrical bed approximately 2.5 cm deep and 4.5 cm in diameter. The
coal bed was formed from a slurry in a test cell, mixed and allowed to
settle and drain by gravity in a test cell apparatus which is shown
schematically in Figure 21. The cell was made from Teflon and designed
so that it could be readily disassembled. Teflon was chosen to aid in
cleaning the cell walls of any surfactant films and to reduce wall wet-
ting during the experiments. The bottom of the test cell was fitted
with a wire mesh support screen and a filter paper disc which was
changed for each experiment to prevent clogging. The filter paper was
purchased from the Carl Schleicher and Schuell Co. and was No. 595. A
dewatered coal bed could be removed intact in the form of a plug, since
the top and bottom cell caps were removable. This minimized any eva-
poration before the coal sample was weighed for final retained moisture
content. Approximately 98 to 99% by weight of the original coal charge
(^ 30 grams) remained intact for the weighing process.
The cell was supported on a torsion balance arrangement made from
stainless steel tubing which was shaped into a "T" configuration as
shown in Figure 22. Mass changes in the cell due to dewatering were
detected by a linear variable differential transformer (LVDT) which
was in contact with the end of the lever arm of the torsion balance at
the same point as the dewatering cell. The output of the LVDT was am-
plified and observed as the y coordinate of an x-y recorder. Before
-------
52
GAS INLET CONNECTION
liii
•O-RING
TOP CAP (TEFLON)
-TEFLON CELL
-O-RING
-FILTER PAPER
WIRE MESH
SUPPORT SCREEN
^-BOTTOM CAP (TEFLON)
DRAIN HOLE
FIGURE 21. TEFLON DEWATERING CELL
-------
CAPACITANCE MANOMETER
HEAD AND ELECTRONICS
PRESSURE
REGULATING
VALVES
AIR
TANK
PRESSURE INLET HOSE
(TYGON TUBING)
TORSION BALANCE
,, TOP VIEW
oo
0
O
O O
X-Y RECORDER
LVDT
AND ELECTRONICS
SHOWING "T"
CONFIGURATION
FIGURE 22. PRESSURE DEWATERING APPARATUS.
in
CO
-------
54
each experiment the system was calibrated by placing a twenty gram
weight on the cell. With this arrangement the system was sensitive to
weight changes of 0.05 gm or less than 0.17 \nt% retained water. Figure
22 also shows the gas pressure inlet line, which is connected to the
compressed dry air tank and pressure regulators and leak valves. The
pressure differential applied across the coal bed was measured by an
MKS Baratron capacitance manometer with differential pressure head.
One end of the pressure gauge was left open to the atmosphere, while
the other end read the pressure present over the coal bed. In this
way effects due to changes in ambient barometric pressure were elimi-
nated and the actual pressure differential across the coal bed was con-
tinuously recorded. The electronics of the manometer provide a 10 volt
full scale reading output. The output was read as the x coordinate on
the x-y recorder. With this arrangement the recorder was accurate to
0.25 mm Hg.
A dewatering experiment was conducted as follows. Thirty grams of
coal was placed in the cell. A slurry was formed and the solution
allowed to drain from the cell. This was repeated with fresh solution
until the expelled solution had approximately the liquid-air inter-
facial tension of the prepared solution. This procedure assured that
the concentration of the liquid phase in the bed was the known concen-
tration of the prepared solution, i.e. the coal was fully in equilibrium
with the surfactant solution of that Y,A value. The top cap of the cell
was then affixed and when drainage ceased, i.e., no weight change in
the cell occurred for about one hour, the x-y recorder was calibrated
and the x-y recorder scales adjusted accordingly. The pressure lines
-------
55
were opened and the leak valves adjusted so that the pressure above
the cell increased at an extremely slow rate of less than 0.25 mm Hg/hr.
The system was established so that when the applied pressure became
only equal to the capillary pressure, dewatering ceased. The system
could remain at equilibrium in this state for over a period of one day
or until the pressure over the coal was increased very slightly. When
no further dewatering was observed even after the pressure had risen
considerably or when air breakthrough took place, the experiment was
terminated. At that point the cell was removed and the dewatered coal
plug transferred to an open preweighed petri dish and immediately
weighed on an analytical balance. The coal was then allowed to air
dry at room temperature for over 24 hours. This time was greater than
that needed for the coal to dry to constant weight, but was chosen for
convenience and to allow the inherent moisture content to form in equi-
librium with the ambient humidity (the conditions under which the dry
coal was originally weighed). The sample was reweighed and the final
moisture content calculated on a weight percent basis as follows:
retained HpO
wtg retained H20 = wt% of dry Coa1 + wt% of retained H20 x 100% •
The accuracy obtained by using the analytical balance was - 0.005 wt%.
With the final wt% of water, the original weight of coal in the cell
and the calibration of the x-y recorder, the weight percent of retained
water was back calculated at various values of applied pressure differ-
ential AP, from the continuous dewatering curve of weight vs. AP. This
data then gave the equilibrium dewatering curves.
-------
56
Vacuum Dewatering (non-equilibrium)
The vacuum dewatering experiments were performed on five 15 gram
samples of the 35 x 60 mesh coal. Each sample was formed into a
cylindrical bed in a glass crucible, the bottom of which consisted of
a porous fritted glass disc. The formed beds were approximately 2.2 cm
deep and 3.7 cm. in diameter. The pore size of the disc was reported
as 4 to 5.5 microns. A schematic diagram of the apparatus is shown in
Figure 23. The crucible was connected securely to a vacuum flask by
means of a rubber stopper in which there was a hole slightly smaller
in diameter than the crucible bottom. This produced a vacuum-tight
fit. The vacuum connect nipple of the flask was connected, through
vacuum tubing and a stopcock, to an oil vacuum pump. The vacuum pump
_p
could evacuate the flask to a pressure of less than 10 Torr, so that
the pressure differential across the coal bed, with the stopcock open
and the pump in operation, was approximately 1 atmosphere.
The vacuum dewatering experiments were conducted in the following
manner. A known amount of coal (15 g) was placed in the crucible and
a quantity of distilled water was added to form a coal slurry. This
was allowed to gravity drain, and then, as much as possible of the re-
mainder of the water was removed by application of the vacuum. This
was repeated with more water until all coal particles of the sample
were readily wet by the water. At this point the actual dewatering
i
experiment commenced. A saturated bed was formed and the vacuum was
applied for a five minute interval after which the crucible was re-
moved and quickly weighed, without disturbing the bed of coal. Any
-------
57
COAL BED
GLASS CRUCIBLE
WITH
FRITTED GLASS DISC
RUBBER STOPPER
TO VACUUM PUMP
FIGURE 23. VACUUM DEWATERING APPARATUS
-------
58
excess water collected on the bottom of the crucible was wiped off be-
fore the weighing process. Immediately after being weighed the crucible
was placed back on the vacuum system for another five minute interval
and the process repeated. The procedure was continued for four intervals.
This procedure served to standardize each specific coal bed as to its
water retention as a function of time under a constant applied pressure
differential.
The entire procedure was then repeated, but instead of only dis-
tilled water, a prepared aqueous surfactant solution was used. A quantity
of the surfactant solution was added to the wet coal bed to form the
slurry. Again it was allowed to gravity drain and the remainder of the
liquid was removed by application of the vacuum. However, this was re-
\
peated with fresh solution until the recovered solution was of the same
concentration as that put in as determined by measurement of the liquid-
air interfacial tension of the solution. Again, at this point the actual
dewatering procedure was initiated. The bed was saturated with solution
and exposed to the vacuum for the five minute time intervals, after each
of which the crucible was weighed as before. The dewatering procedure
was then repeated again, only the next time the concentration of sur-
factant in the solution was increased to some predetermined value.
The dewatering procedure was conducted at each increased value of
surfactant concentration which was chosen as the experiment progressed.
The initial starting concentrations were chosen on the basis of infor-
mation from the pressure dewatering experiments, the liquid-air inter-
facial tension behavior of each surfactant, and the adsorption experiments.
-------
59
Only one coal bed was used for the entire series of experiments
with each surfactant. This was to insure that the only variable was
that of the increasing concentration of surfactant.
From these data, isochronal plots of the amounts of liquid (which,
due to the extremely small concentrations of surfactant, was essentially
water) retained by the coal beds, under a constant pressure differential,
as a function of surfactant concentration could be generated. These
curves were plotted at 5, 10, 15, and 20 minute cumulative exposure
times.
-------
60
Liquid-Air Interfacial Tension Measurements (YLA)
All liquid-air interfacial tension, y^. measurements were made
by the du Nouy ring method [37] which involves the determination of the
force required to detach a ring or loop of platinum-indium wire from
the surface of a liquid. A commercial apparatus, the Fisher Surface
Tensiomat Model 21, was used. The instrument makes use of a torsion
wire in determining the maximum pull on the ring.
The determination of Y, « for a liquid was accomplished in the
manner described in the instruction manual for the instrument [38].
The procedure can be briefly outlined as follows. The liquid under in-
vestigation was placed in an appropriate beaker, which was then placed
on the sample stage of the instrument, under the ring. The ring was
then submerged into the liquid and the instrument adjusted to give a
zero reading. The instrument was then manipulated to properly withdraw
the ring from the surface of the liquid. When detachment occurred the
"apparent" reading on the instrument dial was noted.
The "apparent" readings of the tensiometer must be multiplied by
a correction factor in order to obtain an absolute value. The correc-
tion factor takes into account the densities of the two phases, the
Interfacial tension between which is being measured, and the meniscus
volume, the volume of the distended liquid.
The correction factor was calculated for each reading accoroing to
the procedure outlined in the instrument instruction manual [38].
Certain values unique to the instrument used which were contained in
the correction factor were the ratio of the ring radius, R, to the
-------
61
radius of the wire, r, which was
R/r = 53.5697
and the circumference, c, of the ring which was 5.980 cm.
Also recorded with each y^ reading was the temperature of the
liquid being measured.
Before its initial use the instrument was carefully calibrated
according to the recommended procedure such that the dial of the in-
strument read directly in dynes per centimeter. An additional check
which was also performed periodically, was the determination of the
liquid-air interfacial tension of a sample of spectrographic grade
benzene. The YIA values of the benzene were found to be 28.20 - 0.09
dyn/cm at 25.0°C. An interpolated value from the 57th edition of the
Handbook of Chemistry and Physics [39] is 28.21 dyn/cm and rounded off
to the accuracy of the present instrument, which was 0.05 dyn/on, is
28.20 dyn/cm in excellent agreement.
To eliminate, or at least to prevent the spread of any surface
contaminants such as fingerprint oils etc. and especially any adhering
traces of the surfactants, which exhibit such a high degree of surface
activity at such small concentrations, the platinum-iridium ring and
all glassware involved in the measurement of the interfacial tensions
were thoroughly cleaned before each change in the test liquid.
The ring was thoroughly rinsed in distilled water and then heated
in the oxidizing portion of a gas flame.
-------
62
The glassware, inclusive of all solution transfer pipettes, was
thoroughly rinsed in distilled water. Then the glassware was immersed
in a warm, 60°C - 70°C, standard cleaning solution of sodium dichromate
in concentrated sulfuric acid, for at least 5 minutes. The glassware
was then rinsed in tap water and then finally in distilled water.
This same glassware cleaning procedure was also used for cleaning
the teflon cell used throughout the pressure dewatering experiments, the
crucible used in each vacuum dewatering experiment, and on a routine
basis for all other laboratory glassware that came into use.
-------
63
Liquid-Air Interfacial Tensions of Aqueous Surfactant Solutions
To determine the behavior of the liquid-air interfacial tension
as a function of concentration for aqueous solutions of each surfactant
under investigation, a concentrated stock solution of each surfactant
was initially prepared. The solutions were gravimetrically prepared
and stored in 100 ml stoppered, volumetric flasks. Distilled water was
used as solvent in all cases. Precise aliquots of these solutions were
then pipetted into premeasured quantities of distilled water, already
in the appropriate beaker used in determining YLA- For ultra low con-
centrations the use of a 10 yl pipette was necessitated. The quantities
of solution on which the interfacial tensions were measured were on the
order of 30 to 50 ml. It should be noted that all glass surfaces, e.g.,
of the pipettes, were fully saturated with the prepared solutions prior
to liquid transfer. This eliminated concentration variations due to any
surfactant adsorption on the glassware.
All surfactant solution-air interfacial tension values reported
were the result of ten readings each, over a period of about 30 minutes.
This procedure was followed to allow the surfaces to come to equili-
brium and to estimate and subsequently avoid any transient aging
effects (see Appendix A). Constant readings, though, were usually
achieved after allowing a freshly mixed solution to equilibrate for
about 5 minutes and then allowing about 3 minutes between individual
readings. The standard deviations of the YLA values reported from this
procedure were in the range of 0.20 dyn/cm.
The density of the surfactant solutions, needed to calculate the
-------
64
correction factor for the VIA readings, was taken as that of pure water
because of the extreme dilutions involved. The density of the solutions
was chosen as 0.9970 g/ml [40]. The density of the other phase con-
stituting the interface, which was ambient air,was chosen as 1.1845 x
10"3 g/cm3 [41].
-------
65
Surfactant Adsorption Experiments
The surfactant adsorption experiments were conducted on coal
samples of twenty grams each. A dilute solution containing a precisely
known amount of distilled water and surfactant was prepared to which the
coal sample was added, then the coal, water, surfactant slurry (with an
average value of about 30 wt% solids and about 50 to 100 ml solution)
was repeatedly thoroughly stirred and allowed to settle. This procedure
was followed to insure complete adsorption on all particles. The liquid-
air interfacial tension of the supernatant solution was then measured
as described above for the surfactant solutions alone. The concentra-
tion of the solution was then changed by dilution with precise amounts
of distilled water or by the addition of precise aliquots of the more
concentrated stock solution prepared for each surfactant. This was
repeated over as much of the same concentration range as investigated
above when coal was not present in the solution for each surfactant.
In the case of two surfactants tested the increase in the inter-
facial tension of the solutions, whose proportions of constituents
were mandated by physical experimental parameters, was not great enough
at the higher concentrations for the determination of a concentration
change. However, determination of the change in concentration was
possible by gravimetric analysis of the solution [36]. For this, most
of the solution was decanted from the beaker containing the coal slurry,
then weighed and allowed to evaporate at 100°C, then reweighed. This
process was repeated for a solution which was not exposed to the coal.
Comparison of the findings gave the concentration change. The solution
-------
66
was decanted, not filtered, because it was found that the filter paper
used sorbed too much of the surfactant.
-------
67
VIII. RESULTS
Pressure Dewatering (equilibrium)
The results of the pressure dewatering experiments are shown in
Figures 24-36, where the amount of solution, on a weight percent basis
with the coal, retained by the coal bed is plotted as a function of the
applied pressure differential across the bed. As explained in the Ex-
perimental section, the actual data was in the form of a continuous x-y
recorder trace of the weight loss versus the applied pressure differ-
ential, and that this information had to be transformed into a weight
percent versus applied pressure differential curve so that all the
curves could be compared on an equal basis. The data points shown in
Figures 24-36 are points at which the conversion from weight to weight
percent was calculated from the actual data to enable the construction
of the "normalized" curves shown. These then are only data transfer
points.
Table II lists the important data for the experiments. That is
the weight of the coal sample used in each experiment; the measured
solution-air interfacial tension of the dewatering solution; the con-
centration of the solution as derived from the respective curves cali-
brating the interfacial tension with surfactant concentration; and the
applied pressure differential at which the dewatering curve had its
steepest slope, as determined graphically from the curve.
Figures 37-41 are isobaric plots of the residual amount of solu-
tion left in the dewatered coal beds, at the given pressure as a
function of the surfactant concentration. The data was taken from the
third stage of the dewatering curves of Figures 24-36, where the
-------
68
retained amounts of solution are becoming asymtotic to some final value
as the pressure is increased. Isobaric plots of the dewatering data in
this region give an indication of how this portion of the dewatering is
affected by the surfactants. The data is plotted on the same concen-
tration scale as that for the solution-air interfacial tension plots of
each surfactant. The values of the isobaric residual water contents,
i.e. at zero concentration of surfactant, are also indicated
on the vertical axes of the graphs.
-------
TABLE II
DATA FOR PRESSURE DEWATERIN6 EXPERIMENTS
FIGURE
NUMBER
24
25
26
27
28
29
30
31
32
33
34
35
36
SOLUTION COAL SAMPLE
WEIGHT
Distilled Water
Aerosol A-196
Aerosol A-196
Aerosol A-196
Triton X-114
Triton X-114
Dodecyl Pyridinium Chloride
Sodium Dodecyl Sulfate
Sodium Dodecyl Sulfate
Aerosol -OT
Aerosol -OT
Aerosol -OT
Aerosol -OT
(g)
30.0887
30.3165
30.0242
30.0012
30.2416
30.2550
30.1446
30.0003
30.0073
30.3150
30.0090
30.0013
30.0043
SURFACTANT
CONCENTRATION
(mole fraction)
0
1.32 x 10"6
2.02 x 10"6
1.42 x 10"4
3.20 x 10"7
1.45 x 10"6
4.75 x 10"5
7.20 x 10"6
2.10 x 10"5
3.00 x 10~6
1.20 x 10"5
1.85 x 10"5
2.00 x 10"5
SOLUTION-AIR PRESSURE
INTERFACIAL TENSION DIFFERENTIAL
AT MAXIMUM
YLA(dyn/cm)
69.85
66.90
65.60
52.30
41.50
30.95
44.15
41.40
35.35
40.65
31.65
28.90
28.50
SLOPE
AP(max)(mmHg)
17.5
17.5
18.0
17.5
16.0
6.5
12.0
16.8
12.0
10.3
5.0
7.0
6.0
en
10
-------
70
40
35
fe
i30
25
20
'5
| i
.. ni
0 20 40 6O 80 100 120 140 160 ISO 200 220 240
APPLED PRESSURE DIFFERENTIAL,AP,(mmHg)
FIGURE 24. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING DISTILLED WATER.
-------
71
i40
o
i 35
30
I 25
o
<20
8
o 15
ui
1
£ 10
0 20 40 60 80 100 120 WO 160 180
APPUED PRESSURE DIFFERENTIALS P, (mmHg)
FIGURE 25. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL A-196;
c = 1.32 x 10^6 MOLE FRACTION.
-------
72
i i •" | •" i •" i' •' i" • i' • • i" • i • •:
I... I... I... I... I...-
20 40 60 80 100 120 140 160 180
APPLIED PRESSURE DIFFERENT! AL.AP, (mm Hg)
FIGURE 26. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL A-196;
c = 2.02 x 10"6 MOLE FRACTION.
-------
73
40
30
y 25
Q.
8 20
03
8 15
o
ui
10
§5
I
J_
20 4O 60 60 100 120 I4Q 160 180
APPLIED PRESSURE DIFFERENTIAL,AP, (mmHg)
FIGURE 27. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL A-196;
c = 1.42 x 10"4 MOLE FRACTION.
-------
74
40
J?
35
30
jr-
o
I
y
820
CD
b
o
10
I I I I I I . I I I . I I 1 I I I I I I I I I I I I I I 1 I I I I I I I I I I I I I I I I I I I l"
0 20 40 60 80 100 120 I4O 160 180 200 220 24O
APPLIED PRESSURE DIFFEREDAL.Ap.(mmHg)
FIGURE 28. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF TRITON X-114;
c = 3.20 x 10~7 MOLE FRACTION.
-------
75
~l I i I .11 i I i i i I I I I I I i i I | | i I i | i I | i i I | i I I i i i I i i i I i i i I i
0 20 40 60 80 -100 120 140 160 ISO 200 220 240
APPLJED PRESSURE DIFFERENTIAL,AP,(mmHg)
FIGURE 29. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF TRITON X-114;
c = 1.45 x 10"6 MOLE FRACTION.
-------
76
30
Z 15
20 4O 60 80 100 120 I4O I6O 180
APPUED PRESSURE DIFFERENTTAL.AP.tmro Hg)
FIGURE 30. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF DODECYL
PYRIDINIUM CHLORIDE; c = 4.75 x 10~5 MOLE FRACTION.
-------
77
20 40 60 80 100 ED MO 160 ISO
APPLIED PRESSURE DIFFERENTIAL,A P.(mmHg)
FIGURE 31. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF SODIUM
DODECYL SULFATE: c = 7.20 x 10"6 MOLE FRACTION.
-------
78
I I I I I I I I I T
I I I I I I I I I I I I I
L
0 20 40 60 80 100 120 WO 160 180
APPLIED PRESSURE DIFFERENTIAL,Ap,(mmHg)
FIGURE 32. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF SODIUM
DODECYL SULFATE: c = 2.10 x 10~5 MOLE FRACTION.
-------
79
111111111111111111
40 60 80 100 120 140 160 180 200 220
APPLIED PRESSURE DIFFERENTIAL,* P.(mmHg)
240
FIGURE 33. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL-OT;
c = 3.00 x 10"6 MOLE FRACTION.
-------
80
^31 I I I I I I 1 I I I I
H
§35
5
S
£30
ui
1-25
8
m
15
10
I.I.I
20 40 60 80 100 120 I4O 160 180
APPLED f'RESSURE DIFFERENnAL,Ap.(mmHg)
FIGURE 34. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL-OT;
c = 1.20 x 10~5 MOLE FRACTION.
-------
81
40
P
335
I
> 30
25
i-
?{ 25
o20
15
10
x>
I I I I I
20 40 60 80 100 120 140 160 180
APPLIED PRESSURE DIFFERENTIAL. A P. (mmHg)
FIGURE 35. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL-OT;
c = 1.85 x TO"5 MOLE FRACTION.
-------
82
. I... i... 11.. i., 1111, i.,, i .1.1, ,-:
20 40 60 80 100 120 140 160 180
A APPUED PRESSURE DIFFERENTlAL.AP,(mmHg)
FIGURE 36. PRESSURE DEWATERING CURVE OBTAINED FOR A 35 x 60 MESH
COAL BED CONTAINING AN AQUEOUS SOLUTION OF AEROSOL-OT;
C = 2.00 x 10~5 MOLE FRACTION.
-------
83
5.0
PURE
WATER
VALUE
icr3 icr4
MOLE FRACTION OF AEROSOL A-196
FIGURE 37. EFFECT OF AEROSOL A-196 ON THE AMOUNT OF AQUEOUS
SOLUTION RETAINED BY THE PRESSURE DEWATERED COAL
BED. (ISOBARIC DATA; COAL SIZE: 35 x 60 MESH).
-------
84
14.0
130
11.0
98.0
O '
§4.0
i3.o
20
1 ' i ""| ' ' ' l""|
AP« 30 mmHg
••— «^> ^^ Ap«60mmHg
i i i i 11 >
i i i I i 1111
PURE
VALUE
W8 KT7
MOLE FRACTION OF TRITON X-II4
ID"6
lO"9
FIGURE 38. EFFECT OF TRITON X-114 ON THE AMOUNT OF AQUEOUS
SOLUTION RETAINED BY THE PRESSURE DEWATERED COAL
BED. (ISOBARIC DATA; COAL SIZE: 35 x 60 MESH).
-------
85
I4°rtf
I3JO-
* I2ol-
^"- Y~"
yiijo -
§ 90
~T 1 I | I (Ml 1 1—I | I I I
— —— —"""" AP=30mmHg
"• ~~ •""" ~~ "™ APe4OnwnHa
P— —-„ AP«6OmmHg
°6.0
5.0
4.O
PURE
WATER
VALUE
FIGURE 39.
, ,_, i, ...I
AP*i25mmHg -
I CMC
I . , , . , , , I . ,1, i
KT7 KT6 K)-9
MOLE FRACTION OF DODECYL PYRIDINIUM CHLORIDE
EFFECT OF DODECYL PYRIDINIUM CHLORIDE ON THE AMOUNT
OF AQUEOUS SOLUTION RETAINED BY THE PRESSURE DE-
WATERED COAL BED. (ISOBARIC DATA; COAL SIZE:
35 x 60 MESH).
40mmHg
GOmmHg
•BOmrnHg
AP «l25mmHa
4O
J L
i
CMC
' "'I
_t_ > » i i t i il
1 — J_L
PURE
WATER
VALUE
FIGURE 40.
lO'6
MOLE FRACTION OF SODIUM DODECYL SULFATE
EFFECT OF SODIUM DODECYL SULFATE ON THE AMOUNT
OF AQUEOUS SOLUTION RETAINED BY THE PRESSURE DE-
WATERED COAL BED. (ISOBARIC DATA; COAL SIZE:
35 x 60 MESH).
5x10"*
-------
86
11.0
#
?
§
9O
AP» 40mmHg
AP«60mmHg
§ 7fvL_ __ . AP=80mmHg
AP=80mmHg
AP > 125 mmHg
"""* **• ^.Ap.250mmHg
rlil i i i I i i t 11
40
30
2O
•-O
i i i I i i ii
CMC
, .i , I i ..
PURE
WATER
VALUE
IO"6 K)
MOLE FRACTION OF AEROSOL-OT
FIGURE 41. EFFECT OF AEROSOL-OT ON THE AMOUNT OF AQUEOUS
SOLUTION RETAINED BY THE PRESSURE DEWATERED
COAL BED. (ISOBARIC DATA; COAL SIZE: 35 x 60
MESH).
-------
87
Vacuum Dewatering (non-equilibrium)
The results of the vacuum dewatering experiments [36] are plotted
in Figures 42-46. The data are presented in graphs which cover the
same range of concentration as presented in the data on the solution-
air interfacial tension of Figures 47-51 for perspective. The data
for each surfactant is plotted isochronally as stated previously and
gives the amount of retained solution as a function of surfactant con-
centration. The values of retained water content, i.e. at zero concen-
tration of surfactant, at each time value are indicated at the lowest
concentration listed on the graphs since a zero is not plottable on
a log scale. The data is presented for Aerosol A-196, Triton X-114,
dodecyl pyridinium chloride, sodium dodecyl sulfate, and Aerosol-OT
respectively. Table III lists the weight of the coal sample used
with each surfactant.
-------
TABLE III
WEIGHT OF COAL SAMPLES USED IN
VACUUM DEWATERING EXPERIMENTS
SURFACTANT COAL
SAMPLE WEIGHT
(g)
Aerosol A-196 15.2636
Triton X-114 15.6285
Dodecyl Pyridinium Chloride 15.4668
Sodium Dodecyl Sulfate 15.7705
Aerosol-OT 15.1406
-------
89
PURE
WATER
VALUE
10
MOLE FRACTION OF AEROSOL A-196
5x10"
FIGURE 42. EFFECT OF AEROSOL A-196 ON THE RESIDUAL AQUEOUS SOLUTION
CONTENT OF THE VACUUM DEWATERED COAL BED. (ISOCHRONAL
DATA AT AP = 1 ATM; COAL SIZE: 35 x 60 MESH).
a
I
UJ
OL
I
fe
13.0
IZO
11.0
IOO
9.0
7-0
c
6.0
I I I I 111 I I ] I I I 11
A.— — — — — — —
01 - t=20min
• A»j .- — —— — — — — ~~ ~—
5.0<
4.0
3.0
ao
t=5min
t = 10 min
t« ISrnin
1 = 20 min
i I I I I i i
I I i i i i ii
PURE
WATER
VALUE
IO'8 IO-T
M0l£ FRACTION OF TRITON X-II4
10"
FIGURE 43. EFFECT OF TRITON X-114 ON THE RESIDUAL AQUEOUS SOLUTION
CONTENT OF THE VACUUM DEWATERED COAL BED. (ISOCHRONAL
DATA AT AP = 1 ATM; COAL SIZE: 35 x 60 MESH).
-------
f
* 9.O
I 1
UJ
o:
6.0'
5dO
3X>
t = 5min
t = K)min
I=l5min~
t=aOnr«Sn"
_l . 1 I I . I ll
PURE
WATER
VALUE
K>'7 \0~* IO*
MOLE FRACTION OF OODECYL PYRIDINIUM CHIDRIOE
90
FIGURE 44. EFFECT OF DODECYL PYRIDINIUM CHLORIDE ON THE RESIDUAL AQUEOUS
SOLUTION CONTENT OF THE VACUUM DEWATERED COAL BED. (ISO-
CHRONAL DATA AT AP = 1 ATM; COAL SIZE: 35 x 60 MESH).
^iao
O II
111
tr.
K>O
9.O
o
60
5X)
40
• ' i
l=5min
* 11
I I I I I 1111
PURE
WATER
VALUE
(CMC
I I I I I T I
O'6 ID'9
MOLE FRACTION OF SODIUM DODECYL SULFATE
5x10"
FIGURE 45. EFFECT OF SODIUM DODECYL SULFATE ON THE RESIDUAL AQUEOUS
SOLUTION CONTENT OF THE VACUUM DEWATERED COAL BED. (ISO-
CHRONAL DATA AT AP = 1 ATM; COAL SIZE: 35 x 60 MESH).
-------
91
13.0
12.0
110
~ 9.0
i 8j°
oc 7.0
v> 5.Q
O
I
4.0<
3.0
y/r\
PURE
WATER
VALUE
10 IO"6 10°
MOLE FRACTION OF AEROSOL-OT
CMC -
U I I i i n
10
FIGURE 46 EFFECT OF AEROSOL-OT ON THE RESIDUAL AQUEOUS SOLUTION CONTENT
OF THE VACUUM DEWATERED COAL BED. (ISOCHRONAL DATA AT AP =
1 ATM; COAL SIZE: 35 x 60 MESH).
-------
92
Liquid-Air Interfacial Tension (YIA) - Concentration Curves
It has been stated that it was necessary to determine the liquid-
air interfacial tension of the aqueous surfactant solutions as a function
of concentration so that this could be used as a convenient analytical
procedure for determination of the changes in solution concentration on
adsorption. Figures, 47-51 are the experimentally determined curves for
each of the five surfactants studied in this investigation at room tem-
perature, 25°C. The concentration scales are logarithmic to encompass
the full range of concentrations investigated. Listed in Table IV are
the critical micelle concentrations (CMC) determined by the radical
change in slope of the curves at the higher concentrations. It is at
this point that the surfactants begin a form of precipitation or aggrega-
tion phenomena as described above. Also reported in Table IV are the
values of the maximum slope
d InC max
of each curve for later use in determing the area of the adsorbed sur-
factant molecules.
-------
93
(dyn/cm)
I I I I Mil 1 1 I I I I | I.
35
5x10"' 10'
M0l£ FRACTION OF AEROSOL A-196
FIGURE 47. LIQUID-AIR INTERFACIAL TENSION OF AQUEOUS SOLUTIONS OF
AEROSOL A-196 AT 25°C. [36]
fcyn/cm)
10
I0'8 IO'
MOLE FRACTION OF TRITON X-II4
FIGURE 48. LIQUID-AIR INTERFACIAL TENSION OF AQUEOUS SOLUTIONS OF
TRITON X-114 AT 25°C.
-------
94
MOLE FRACTION OF DOOECYL PYRIDIN1UM CHLORIDE
FIGURE 49. LIQUID-AIR INTERFACIAL TENSION OF AQUEOUS SOLUTIONS OF
DODECYL PYRIDINIUM CHLORIDE AT 25°C.
iff
KT* O'5 0'
MOLE FRACTION OF SODIUM DODECYL SULFATE
5x10
FIGURE 50. LIQUID-AIR INTERFACIAL TENSION OF AQUEOUS SOLUTIONS OF
SODIUM DODECYL SULFATE AT 25°C.
-------
95
'LA
(dyn/cm)
65
60
55
50
45
40
35
30
25
20
to-
I
I I I ! I
CMC
I .1.1..
IO'7 10**
MOLE FRACTION OF AEROSOL-OT
10'
FIGURE 51. LIQUID-AIR INTERFACIAL TENSION OF AQUEOUS SOLUTIONS OF
AEROSOL-OT AT 25°C.
-------
96
TABLE IV
CRITICAL MICELLE CONCENTRATION AND MAXIMUM SLOPE
DETERMINED FROM THE YlnC CURVES OF EACH SURFACTANT
SURFACTANT
CMC
ILA|
d InClmax
(mole fraction) (dyn/cm per order of magnitude
increase in concentration)
Aerosol A-196
Triton X-114
Dodecyl Pyridinium Chloride
Sodium Dodecyl Sulfate
Aerosol -OT
1.50 x 10"3
2.10 x 10"6
7.60 x 10"5
1.30 x 10"4
3.20 x 10"5
- 6.06
- 7.29
-11.93
, - 7.00
- 6.45
-------
97
Influence of Coal on the Interfacial Tension of Distilled Water
Before conducting the adsorption experiments the liquid-air inter-
facial tension, YLA» of twenty-eight milliliters of distilled water
^YLA = 70>5° ' °-20 dyn/cm) in which there was approximately twenty
grams of coal was monitored [36] for five days to determine whether or
not the coal or any of its components were dissolving into the water,
which might alter the subsequent interfacial tension measurements. The
Y|_£ value of the water over the slurry remained constant over this time.
The settled mixture was agitated regularly to insure homogeneity in the
liquid phase. The results are given below in Table V.
The coal sample apparently presented no significant impurities to
the contacting liquid phase which could cause a significant YLA varia-
tion. One should note, however, that traces of calcium (or other alkali
or alkaline earth elements) might effect reactions with the small con-
centrations of surfactants in this study. This would be particularly
the case at surfactant concentrations below 10" mole fraction. The
Y, A vs- concentration data, however, did not indicate major inflections,
which would be expected, if the surfactants were caused to precipitate.
-------
98
TABLE V
INFLUENCE OF COAL ON THE
INTERFACIAL TENSION OF DISTILLED WATER
DAY
1
2
3
4
5
TEMPERATURE
o
( c)
25.5
25.5
25.0
25.0
25.5
YLA
(dyn/cm)
70.50
71.00
70.65
70.75
70.65
mean 70.70 ±0.20
H20 REMAINING
IN BEAKER (g)
27.6730
27.0302
26.3224
25.1628
24.2422
Weight of coal = 20.0099 g.
Original weight of water = 28.0000 g.
-------
99
Adsorption Isotherms
The decrease in surfactant concentration occurring due to the ad-
sorption of the surfactant on the coal was determined by measurement
of the liquid-air interfacial tension.y^.of the solution over the coal
slurry. The "missing" number of surfactant molecules necessary to de-
crease the concentration to some final value were presumed to be adsorbed
on the coal surface. Figures 52-56 are plots of the change in inter-
facial tension as a function of the original concentration of surfactant
in solution when twenty grams of coal were present in that solution. In
order to compare these data with those of the standard liquid-air inter-
facial tensions, the points from the adsorption experiments are shown
superimposed on their respective interfacial tension-concentration
curves without coal. A smooth curve cannot be drawn through these
points since they are not normalized. That is, each point provides
only the numerical determination of the change in concentration of the
solution, a value which must be multiplied by the total number of moles
of solution present and divided by the weight of the adsorbent coal,
such that a comparison to other values may be made on an equal basis.
Normalization in this manner provides a value for the number of moles
of surfactant adsorbed per gram of coal and when this value is plotted
against the overall concentration of surfactant present in the coal-
water-surfactant solution an adsorption isotherm is generated.
Tables B1-B5 shown in Appendix B provide the data for each sur-
factant in order to determine the adsorption isotherms. Figures 57-61
are the adsorption isotherms.
-------
100
SxICT7
D WITH COAL
O WITHOUT COAL
KT9 JO"4
MOLE FRACTION OF AEROSOL A-196
FIGURE 52. LIQUID-AIR INTERFACIAL TENSION OF FIXED AMOUNTS OF AQUEOUS
SOLUTIONS OF AEROSOL A-196 IN CONTACT WITH AN ADSORBENT
COAL SAMPLE, TOGETHER WITH THE DATA TAKEN WITHOUT THE COAL
PRESENT, FOR DETERMINATION OF THE CHANGE IN SOLUTION CON-
CENTRATION ON ADSORPTION.
75
70 r
65^-
60^-
55 —
50 —
>u
Cdyn/tm)45
40
35
30F-
25
cr*
o WITH COAL
O WITHOUT COAL
I .... , , , l,,.,l
I i ..J
I i ,r
10-=
MOLE FRACTION OF TRITON X-II4
FIGURE 53. LIQUID-AIR INTERFACIAL TENSION OF FIXED AMOUNTS OF AQUEOUS
SOLUTIONS OF TRITON X-114 IN CONTACT WITH AN ADSORBENT
COAL SAMPLE, TOGETHER WITH THE DATA TAKEN WITHOUT THE COAL
PRESENT, FOR DETERMINATION OF THE CHANGE IN SOLUTION CON-
CENTRATION ON ADSORPTION.
-------
101
'LA
(dyn/cm) 55 —
\OT 5x KT
MOLE FRACTION OF DODECYL PYRIDIMUM CHLORIDE
FIGURE 54. LIQUID-AIR INTERRACIAL TENSION OF FIXED AMOUNTS OF AQUEOUS
SOLUTIONS OF DODECYL PYRIDINIUM CHLORIDE IN CONTACT WITH
AN ADSORBENT COAL SAMPLE, TOGETHER WITH THE DATA TAKEN
WITHOUT THE COAL PRESENT, FOR DETERMINATION OF THE CHANGE
IN SOLUTION CONCENTRATION ON ADSORPTION.
70
65
60
55
50
1*
(dyn/cm) 45
40
35
30
SxlO'8 KT7
QWITH COAL
O WITHOUT COAL
IO'5 IO'4
MOLE FRACTION CF SODIUM DODECYL SULFATE
5xKT*
FIGURE 55 LIQUID-AIR INTERFACIAL TENSION OF FIXED AMOUNTS OF AQUEOUS
SOLUTIONS OF SODIUM DODECYL SULFAJE IN CONTACT WITH AN AD-
SORBENT COAL SAMPLE, TOGETHER WITH THE DATA TAKEN WITHOUT
THE COAL PRESENT, FOR DETERMINATION OF THE CHANGE IN SOLU-
TION CONCENTRATION ON ADSORPTION.
-------
102
(dyn/cm)
70
65
60
55
50
45
40
35
30
25
I I I ' I I
Q WITH COAL
O WITHOUT COAL
2D
i i
,0-8
D"7 O"6
MOLE FRACTION OF AEROSOL-OT
O
,-s
ll I I ! I I I
10"
FIGURE 56. LIQUID-AIR INTERFACIAL TENSION OF FIXED AMOUNTS OF AQUEOUS
SOLUTIONS OF AEROSOL-OT IN CONTACT WITH AN ADSORBENT COAL
SAMPLE, TOGETHER WITH THE DATA TAKEN WITHOUT THE COAL
PRESENT, FOR DETERMINATION OF THE CHANGE IN SOLUTION CON-
CENTRATION ON ADSORPTION.
-------
103
8
fi 10*
5 -a
CO?
o
g
^
fe
IO"7
• • ' I""! ' ' ' I"11! ' ' 'I111:
1
CMC
, . . I....I . . . I....I I ... I..I!
IO'5 10-* IO'3
M0l£ FRACTION OF AEROSOL A-196
io-2
FIGURE 57. ADSORPTION ISOTHERM FOR AEROSOL A-196 ON 35 x 60 MESH COAL
FROM AQUEOUS SOLUTION AT 25°C.
-------
104
UJ
0.
£10*
CO
y
o
2
1
CMC
_J •
iO'7 \or*
MOLE FRACTION TRITON X-II4
FIGURE 58. ADSORPTION ISOTHERM FOR TRITON X-114 ON 35 x 60 MESH COAL
FROM AQUEOUS SOLUTION AT 25°C.
-------
105
u.
o
:IO'4
.35
o
o
o
IO'7
I i , . ...I i . . I ,TT.
I(T5 10"* KT3
MOLE FRACTION DODECYL PYRIDINIUM CHLORIDE
FIGURE 59. ADSORPTION ISOTHERM FOR DODECYL PYRIDINIUM CHLORIDE ON
35 x 60 MESH COAL FROM AQUEOUS SOLUTION AT 25°C.
-------
106
10'
or
o
Q
UJ
m
I »«
§1
§ KT7
10*
il
(CMC
. I ....I I
IO'7
IO'6 KT9 XT' 2x10-
MOLE FRACTION OF SODIUM DODECYL SULFATE
FIGURE 60. ADSORPTION ISOTHERM FOR SODIUM DODECYL SULFATE ON 35 x 60
MESH COAL FROM AQUEOUS SOLUTION AT 25<>C.
-------
107
CMC
K)'6 IO'5
MOLE FRACTION OF AEROSOL-OT
10
,-4
FIGURE 61. ADSORPTION ISOTHERM FOR AEROSOL-OT ON 35 x 60 MESH COAL
FROM AQUEOUS SOLUTION AT 25°C.
-------
108
Standard Free Energy of Adsorption
The standard free energy of adsorption as a function of concentra-
tion for each surfactant adsorption isotherm was generated through the
use of Equation (6)
r = 2r C exp [- AG°DS/RT]
In calculating the values for AG?DS, R was taken as 1.987 cal/mole°K
and T as 298°K. The radii, r, of the adsorbed molecules were determined
from the areas per molecule found through the use of the slopes of the
interfacial tension curves for each surfactant. The areas were assumed
2
to be circular so that A = ttr .
2
To determine a value for r, which is in units of moles/cm , the
values of the moles of surfactant adsorbed per gram of coal had to be
divided by a specific surface area of the coal. The value used was
2
500 cm /gm for reasons which will become apparent in the discussion
section. In any case, the shape of the curve remains constant regard-
less of the value chosen.
Figures 62-66 are the plots of the standard free energy of adsorp-
tion generated from each data point of the adsorption isotherms of
each surfactant.
-------
109
(keel/mote) «scn —
5.30
SxICT7 KT6
IOT5 I0'4
MOLE FRACTION OF AEROSOL A-196
KT3
FIGURE 62. STANDARD FREE ENERGY OF ADSORPTION FOR AEROSOL A-196 ON 35 x 60
' MESH COAL AT 25°C.
650
6.40
-AG-ADS
(kcol/mole)
6.30
6.20
IO
,-B
CMC
IO'7 IO'6
MOLE FRACTION OF TRITON X-II4
FIGURE 63. STANDARD FREE ENERGY OF ADSORPTION FOR TRITON X-114 ON 35 x 60
MESH COAL AT 25°C.
-------
110
6.70
660
6.50
6.40
(keel/mote) 53°
620
6.10
6.00
5.90
CMC
i i i i i i 11 i i i i i i li I
i i i i i r
ID'6 KT5 ID'4
MOLE FRACTION OF DODECYL PYRIDINIUM CHLORIDE
FIGURE 64. STANDARD FREE ENERGY OF ADSORPTION FOR DODECYL PYRIDINIUM
CHLORIDE ON 35 x 60 MESH COAL AT 25°C.
6.40
630
6.20
-AGAOS 6.10
(keel/mole)
6.00
5.90
5.80
5.70
icr7
-T r-rp^i
JO"6 10'* IO"4
MOLE FRACTION OF SODIUM DODECYL SULFATE
5x10'
FIGURE 65. STANDARD FREE ENERGY OF ADSORPTION FOR SODIUM DODECYL SULFATE
ON 35 x 60 MESH COAL AT 25°C.
-------
Ill
6.50
6.00
-AG« 55°
(keel/mote)
5.00
5x10" IO'T
\
\
*
i
, .,,... . , .
10" KT
MOLE FRACTION OF AEROSOL-OT
KT
FIGURE 66. STANDARD FREE ENERGY OF ADSORPTION FOR AEROSOL-OT ON 35 x 60
MESH COAL AT 25°C.
-------
112
IX. DISCUSSION
Interfacial Tension - Concentration Curves
The data for the liquid-air interfacial tension, YL/•> as a function
of concentration for each surfactant were not only used as calibration
curves to determine unknown concentrations of surfactants, but also to
obtain the limits of the range of concentration to be investigated for
each surfactant. The lower limit was governed by the onset of observable
surface activity as evidenced by a decrease in y... The upper limit,
however, was limited by the CMC, since at this point the system under in-
vestigation becomes multiphase.
Data past the CMC in each case was limited, therefore, since a pre-
cise determination of the CMC was not the objective. Precise determination
of the CMC by the surface tension log plot method is dependent upon the
extent of the data points used in determining the intersection of the best
fit straight lines for the data above and below the change in slope of the
curves [42]. This and other techniques for determination of CMC's of aqueous
surfactant systems have been reviewed and evaluated by Murkerjee and Mysels
[42].
Nevertheless, comparison of the CMC data with values reported in the
literature shows good agreement. Table VI lists the CMC's determined in
\
this investigation along with literature values.
Examination of the YIA - InC curve for sodium dodecyl sulfate shown
in Figure 50 reveals evidence of a minimum slightly below the CMC. Such a
minimum has been observed by other investigators [43-47] and has been
determined [46] to be due to the presence of some lauryl alcohol impurity,
common in most commercially available samples of sodium dodecyl sulfate,and
-------
113
is reported to disappear when highly purified sodium dodecyl sulfate was
used [47].
TABLE VI
CRITICAL MICELLE CONCENTRATIONS
OF AQUEOUS SURFACTANT SOLUTIONS*
SURFACTANT
CMC
(mole fraction)
THIS WORK OTHERS
Aerosol A-196 1.50x10
Triton X-114 2.10x10
Dodecyl Pyridinium Chloride 7.60x10
Sodium Dodecyl Sulfate 1.30x10
Aerosol-OT 3.20x10
-3
-6
-5
-4
-5
3.02xlO"6[48]
5.04xlO"5[49]; 2.53xlO"4[50]
1.44xlO"4[51]; 1.52xlO"4[52]
4.51xlO"5[53]; 2.84xlO"5[54]
Room temperature values, 25°C.
-------
114
Pressure Dewatering
Examination of the dewatering curves of Figures 24-36 reveals that de-
creasing the interfacial tension of the water in the coal beds at the
liquid-air interface through the addition of the surfactants reduced the
pressure differentials necessary to accomplish the major portion of the
dewatering, which is a trend expected for a capillary system.
Figure 67 is a plot of the interfacial tensions of the solutions ex-
pelled from the coal beds versus the respective differentials AP(max) at
which the dewatering curve had the maximum slope. The data were listed in
Table II. According to the capillary theory discussed above the relation-
ship should be linear if the pore size of each coal bed was the same. The
capillary equation, Eq.(5) , in this case can be rewritten as:
(27)
The solid line drawn through the data points of Figure 67 is the result
of a linear regression analysis of the data, inclusive of the natural data
point at the origin. The correlation coefficient of the data, which is a
measure of the "degree of fit" of the given points to the straight line,
was calculated to be 0.894 which indicates a reasonably good correlation
considering the complexity of the system under study.
If then the coal beds can be regarded as a bundle of capillary tubes,
then the slope of this line is half of an average effective pore radius,
i.e., a pore radius of 42.18 ym.
The least squares line, however, gives an intercept on the Y axis
-------
115
70.00
£00 10.00 15.00
AP (moxHmmHg)
2QOO
FIGURE 67. RELATIONSHIP BETWEEN THE LIQUID-AIR INTERFACIAL TENSION,
Y OF THE AQUEOUS SOLUTION PRESENT IN THE COAL BED AND
THE PRESSURE DIFFERENTIAL, AP, AT WHICH THE RESPECTIVE
DEWATERING CURVE HAD ITS MAXIMUM SLOPE. ( ) LINEAR
REGRESSION LINE, (—) BEST FIT LINE PASSING THROUGH THE
ORIGIN.
-------
116
of 8.73 tiyn/cm. According to the theory though, the line must pass through
the origin. It should be noted, however, that the equation as applied as-
sumes that the bed of particles can be regarded as a bundle of capillary
tubes of uniform circular cross section; however, in the original deriva-
tion from the Laplace equation [55] in which the pressure difference is
determined across a curved interface, the surface of a liquid in a capil-
lary, is related to the liquid surface tension and the harmonic mean of the
two principal radii of curvature of the interface:
AP = Y(F- + ir}
Kl K2
Only for the case of a spherical interface, e.g.,the hemispherical meniscus
1n the capillary, does R, = R2 and therefore give
(29)
Furthermore, even if the approximation of circular capillaries could
be applied to uniformly packed smooth spheres this need not at all be
appropriate for coal particles which are neither smooth nor spherical.
Also, the intercept is only an error of about 10% on the YIA scale,
which for a real system with the above assumptions is not unreasonable.
Furthermore the values of AP(max) were chosen as the representative
number for each experiment. The slopes of all the dewatering curves were
however all askew, with the largest slopes always at the lower AP values.
The solid line of Figure 67 has only to be offset up the AP axis by 3.2
-------
117
mm Hg for the line to pass through the origin. The determination of the
AP(max) value is subject to interpretation. In this case the actual maxi-
mum slope was chosen. The other choices however would give values of AP
higher than the AP(max) and values 3.2 mmHg higher are still contained
on the steeply sloping portion of each dewatering curve.
Therefore the fact that the data produce an intercept value is not
unreasonable. However to provide a comparison curve, the dashed line of
Figure 67 was drawn so as to be the best fit straight line for the data
points but passing through the origin. This slope gives a slightly larger
pore radius of 57.42 ym.
At this point we can determine a relationship between the effective
pore size determined to be in the coal bed and the known particle size of
the coal present, by obtaining the ratio of average particle radius to
average pore radius.
The actual coal particles size used in these experiments was fixed to
range from 420 m to 250 ym in diameter. Two values for an average par-
ticle diameter can be calculated: an arithmetic average and a logarithmic
average, since when coal fractures it is known to produce a logarithmic
size distribution of the resulting particles. The arithmetic average is
335 ym and the log average is 342 ym which give particle radii of 167.5
and 162 ym respectively. TableVII gives the values of the ratio of particle
radii to pore radii calculated for both averages and both pore radii. For
uniform spheres arranged in a close packed structure, the ratio would be
4.44. Haines [56] has calculated that for spheres, depending on their
packing,the ratio can vary from 2.0 to 6.4. Gray [3] in his study of
Betteshanger coal of a similar size distribution obtained a value of about
3.00. The preferred value of TableVII would be 3.84. However, by multi-
-------
118
TABLE VII
RATIO OF THE AVERAGE COAL PARTICLE RADIUS
TO THE AVERAGE PORE RADIUS OF THE COAL BEDS
AVERAGEJ'ARTICLE
DIAMETER
(ym)
*
335
324**
335
324
RADIUS
(ym)
167.5
162
167.5
162
AVERAGE PORE
RADIUS
(ym)
42.18
42.18
57.42
57.42
RATIO
AVERAGE PARTICLE RADIUS
AVERAGE
3
3
3
2
2
PORE RADIUS
.97
.84***
.00
.92 1
.82 J
, With
Intercept
Gray's value
Without
Intercept
* Arithmetic average
** Logarithmic average
*** Preferred value
-------
119
plying the smallest ratio with the smallest pore radii and the largest
ratio with the largest pore radii and then doubling the values we can ob-
tain a particle size distribution indicative of the spread of these values.
The distribution obtained is 456 ym to 238 urn, which is almost identical
to the distribution known to be in the bed, i.e. 420 pm x 250 pro.
The addition of the surfactants to the water contained in a coal bed,
therefore, serves to lower the interfacial tension of the water at the
liquid-air interface, which in turn lowers the applied pressure differen-
tial required to effect dewatering. And, furthermore, the relationship
between the interfacial tension of the liquid of the bed and the pressure
differentials is consistent with that predicted by the application of a
capillary theory to the system.
If the coal beds are reasonably well approximated by a bundle of
capillary tubes with an average radius of 42.18 vm then it is possible
to calculate the surface area of those capillaries if their contained vol-
ume were known. This volume would be the volume of the water contained
in a completely saturated bed. The dewatering data however give only a
value of the amount of water contained by the bed in a partially saturated
or capillary state. This value can be obtained from the weight percent
retained water at AP = 0 and the weight of the respective coal bed. An
area calculated with this data will nevertheless give a reasonable estimate.
The area of the capillaries per gram of coal is related to the contained
volume by
= 2V (30)
rw
-------
120
where A = specific surface area
V = volume of capillary
r = radius of capillary
w = weight of coal
Table VI11 lists the appropriate data and the calculated areas for each
bed tested from each dewatering experiment. The volume of the contained
solutions were calculated assuming their density was 1 gm/ml, due to their
extremely dilute concentration of surfactant. The mean value of specific
2
surface area was found to be 284 cm /gm. Gray [3] found an area of
2
290 cm /gm for Betteshanger coal beds of similar size distribution. He
found that his value, also determined on a partially saturated coal bed,
was 17% lower than a value obtained by permeability methods from flow ex-
periments in which the beds were fully saturated. A correction of 17%
brings the mean value of the specific surface area of the coal determined
2
in the present investigation to 342 cm/gm. The value is expected to be
lower than the actual surface area of the coal because of the assumption
of circular capillaries and the fact that the areas between touching par-
ticles is not included.
-------
121
TABLE VIII
DATA FOR CALCULATION OF THE
SPECIFIC SURFACE AREA OF THE COAL BASED ON THE CAPILLARY MODEL
DATA FROM
FIGURE
284
r =
V =
A'=
A =
24
25
26
27
28
29
30
31
32
33
34
35
36
f-i£
42.18
Li
A'
w
Wt. % @ AP=0 VOLUME OF SOLUTION Wt. COAL
V W
(Wt. X) (cm3) (g)
36.1 17.0
38.0 18.6
38.6 18.9
35.5 16.5
37.4 18.1
38.9 19.3
39.4 19.5
35.1 16.2
35.7 16.6
35.4 16.6
42.4 22.1
38.0 18.4
35.5 16-5
2
; 17% lower than 342 ~-
-4
x 10 ym
2V
A = —
M rw
30.0887
30.3165
30.0242
30.0012
30.2416
30.2550
30.1446
30.0003
30.0073
30.3150
30.0090
30.0013
30.0043
SPECIFIC SURFACE
AREA
A
(cm2/g)
268
291
299
261
284
302
306
256
263
260
349 ~
291
261
mean = 284 ± 27
-------
122
The Coal-Surfactant Interaction
Let us now consider the dewatering data only in the latter stages of
dewatering and its relationship to the retention of liquid on the coal pro-
duct. This data is best examined on the isobaric plots of percent water
retention versus concentration as shown in Figures 37-41 and the non-
equilibrium dewatering experiments of Figures 42-46. Examining the two
sets of data for anionic Aerosol A-196, Aerosol-OT and sodium dodecyl sul-
fate, it is observed that the amount of solution retained exhibited a mini-
mum in retained liquid in all three cases as the concentration of surfac-
tant is increased. Also, after this minimum, the amount of solution re-
tained on the coal increased considerably. This is evident particularly
on the non-equilibrium dewatering studies. Cationic dodecyl pyridinium
chloride, on the other hand, only began to show a decrease in the residual
solution content after a certain concentration was reached after which it
remained constant. Nonionic Triton X-114 behaved almost in the same fashion
excep't only a slight minimum was observed.
In all five cases the interfacial tensions, plotted over the same con-
centration range, decreased monatonically with increasing concentration of
surfactant to the critical micelle concentration. (The problem of sodium
dodecyl sulfate has already been discussed). No minimums or abrupt changes
were observed at the concentrations indicated by the residual solution con-
tent data.
Therefore, the effect of the addition of surfactants on the residual
amount of solution retained by the coal beds cannot be related simply or
directly to the lowering of the liquid-air interfacial tensions.
-------
123
Let us now examine the effect of the surfactants in more detail. As
the concentration of the surfactants was increased the amount of solution
retained in the coal beds began to decrease at some point in all five cases.
This improvement in dewatering can only be explained, if the coal surface
became more hydrophobic as the number of surfactant molecules present in
the coal-water interface system was increased. The adsorption from solu-
tion experiments indicate that each of the surfactant molecules were ad-
sorbed to some extent onto the coal surface, and that amount adsorbed in-
creased with increasing surfactant concentration in solution. For the coal
surface to become more hydrophobic as the number of surfactant molecules
adsorbed increased, the molecules must have adsorbed with their head groups
adjacent to the solid, leaving the hydrophobic hydrocarbon chains extended
out into the aqueous phase. Looking at the solid from the liquid, one
could imagine a surface becoming more and more dense with hydrophobic hy-
drocarbon chains.
This preferential orientation of the adsorbed molecules can explain
the coal surface becoming more hydrophobic as the concentration of sur-
factant in solution was increased but seems to be at a loss to explain a
minimum, then a reversal, and subsequent increase in the amount of solution
retained by the bed in the case of the anionic surfactants. The situation
could be reconciled if at the reversal concentrations the surfactant mole-
cules began to adsorb the other way around, presenting their hydrophilic,
water soluble, head groups to the water phase, thus making the surface be-
come hydrophilic.
If a situation was known in which the surfactant molecules did aggregate,
hydrocarbon chain to hydrocarbon chain, presenting their head groups to the
-------
124
water phase at a certain concentration, then the change in adsorption be-
havior might be explained. Fortunately, this is known to happen and has al-
ready been described, i.e., micellation; one of the unique properties ex-
hibited by these molecules in bulk aqueous solution. Micellation, however,
was found to occur in the bulk surfactant solutions at considerably higher
concentrations (CMC's) than those exhibited by the respective minimums when
the coal was present. The coal surface may, therefore, be regarded as a
form of heterogeneous nucleation site for micelle-like aggregation of the
surfactant molecules on the coal surface. If one layer, i.e., some effec-
tive monolayer, was adsorbed in the first orientation, it would appear sim-
ilar to a half micelle in that it is accumulated into a two dimensional
sheet across the surface of properly oriented monomers. The excess concen-
tration in the bulk liquid which is required to serve as the driving force
for micellation has been reduced by the preferential orientation of the
adsorbed layer.
A type of surfactant adsorption phenomena such as this was treated by
Fuerstenau and coworkers [23,32,57] and is called "hemi-micelle" formation.
Fuerstenau investigated the adsorption of sodium alkyl sulfonates from
aqueous solution onto alumina. The hemi-micelle theory was based on in-
formation from adsorption isotherms that were correlated with electro-
phoretic studies. The adsorption data was analyzed by the Stern-Grahame
Theory and exhibited sharp slope changes when plotted on log-log plots.
The electrophoretic studies gave information about the potentials in the
electric double layer, specifically the zeta potential. The existence of
the hemi-micelles was basically proposed to explain the zeta potential sign
reversal at the same concentration as a slope change in the adsorption iso-
therms. Furthermore, Fuerstenau proposed a model for the adsorption be-
-------
125
havior of the surfactants on the alumina, based also on the change in the
standard free energy of adsorption as a function of surfactant concen-
tration in solution, obtained from the adsorption isotherms and the Stern-
Grahame adsorption theory. It was concluded that the adsorption progressed
through three distinct regions in which the mode of adsorption was con-
trolled by one or another of the components of the total standard free
energy of adsorption, AG°, becoming more or less dominant. It should be
noted here that the solid under investigation, viz., alumina, has a much
more well defined surface chemistry than that of heterogeneous coal.
In light of this approach, which seems to reconcile the dilemna of the
surfactant adsorption on coal in reverse orientation, let us examine evi-
dence for each surfactant more critically with the assumption that the
minimums in the residual solution contents are, in fact, indicative of
some effective monolayer coverage or at least the point at which it is
energetically favorable to adsorb in the next layer of opposite orienta-
tion. With this in hand one can then calculate an effective area covered
by the adsorbed molecules, i.e., an effective specific surface area of the
coal accessible to the surfactants.
The minimum in residual content of solution in the coal bed for Aero-
sol A-196 occurred at about 2.0 X 10"6 mole fraction. From the adsorption
isotherm the number*of moles of surfactant adsorbed per gram of coal at this
concentration was 1.25 X 10~7 moles/g, and if this number is multiplied by
Avogadro's number, 6.023 X TO23 molecules/mole, we obtain 7.5 X 10 mole-
cules/g.
The lowering of the liquid-air interfacial tension as a function of
concentration is the result of the adsorption of the surfactant molecules
at the liquid-air interface, as stated previously in section IV. SURFACTANTS,
-------
126
above, from which we can obtain a value for the area per adsorbed mole-
cule. This adsorption is treated by means of the Gibbs Equation [58]
where the number of molecules adsorbed per square centimeter of inter-
face, r, is related to the slope of the interfacial tension versus the
log of the concentration curves as
r = =1. dY
r
kT d In C
Where k is the Boltzmann constant, T the absolute temperature, y
the liquid-air interfacial tension, and C the surfactant concentration
in mole fraction. The maximum slope represents the closest packing of
the molecules at the liquid-air interface. The slopes for each surfactant
were reported in Table IV with the calibration curves of Figures 47-51.
The area per molecule, a , is obtained from the inverse of r at the value
of the maximum slope:
- UT
o ~ ~kT dy
The areas per molecule arrived at in this manner are given in Table IX,
along with the values observed by other authors as cited in the literature.
We now assume that the area the adsorbed Aerosol A-196 molecule oc-
cupies at the liquid-air interface is essentially the same as it might
occupy at the solid (coal) - liquid interface. This is reasonable since
the molecules are most probably oriented at the coal-liquid interface as
they are in the liquid-air interface. Furthermore, the molecules investi-
gated had sufficient steric hinderance as to limit their configurations.
-------
127
TABLE IX
ADSORBED AREA OF SURFACTANT MOLECULES
SURFACTANT AREA PER ADSORBED MOLECULE
THIS WORK OTHERS
Aerosol A-196 68
Aerosol - OT 64 70 [59]; 67.[60]*
Sodium Dodecyi SuKate 59 52 t61^
Dodecyl Pyridinium Chloride 34 35 [62]
Triton X-114 56 50 [48]
MOLECULE
CH3 9 9.6 [63]**
Polyoxyethylene chain
/ -I i- \ i t~t ***
(n = 7.5) 1/7
83
*area of the similar molecule Aerosol-OTN (Sodium di-n-octyl sulfosuccinate)
o
**calculated on the basis of a 3.5A distance between CH3 groups
o
***calculated from n = 2A [63]
-------
128
Op
The area of the Aerosol A-196 molecule was found to be 68A . Thus
multiplying this by the number of molecules adsorbed per gram of coal we
2
obtain a value of 512 cm /gm for the effective specific surface area of
the coal size under investigation.
A similar calculation can be conducted for Aerosol-OT, a molecule
quite similar to that of Aerosol A-196. The concentration at which Aerosol-OT
exhibited the dewatering minimum was chosen as 1.0 x 10~ mole fraction from
the isobaric plots shown in Figure 41 since these plots represent an equi-
librium situation. The minimums of the non-equilibrium vacuum dewatering
curves of Figure 46 appear to be approaching this equilibrium value, as the
time of exposure was increased. There was a problem in the determination
of the number of moles of surfactant adsorbed at this concentration, how-
ever, since it was not possible to obtain adsorption data for this surfactant
in this range. Recognizing that all the adsorption isotherms for all of the
surfactants were relatively linear when subjected to a log-log plot, a
value for the number of moles adsorbed was therefore estimated from a linear
interpolation of the data that was available, cf. Figure 61. The value ob-
tained was 1.25 X 10 moles/g or 7.5 X 10 molecules/g which was the same
as the number of molecules/gm of Aerosol A-196. The area of the Aerosol-OT
molecule was determined to be 64A , similar to Aerosol A-196 and in good
09
agreement with that found in the literature [59] i.e., 70A . This is also
reasonable since the two molecules are almost equal size and shape as illus-
trated by the molecular models shown in Figures 12 and 20. Using our value, an
2
effective specific surface area of the coal of 482 on /gm is obtained. Using
the literature value of 70A , which is also reported [59] as being a value calcu-
2
lated from a molecular model, gives an area of 527 cm /gm, both values being in good
agreement with that of Aerosol A-196, considering the heterogeneity of .the coal
-------
129
surface.
Before continuing with this study another relation between the concen-
trations of the observed minimums and the adsorption data should be con-
sidered, i.e., the variation of the standard free energies of adsorption
with concentration as shown in Figures 62-66. For Aerosol A-196 the con-
centration in question was 2.0 X 10"6 mole fraction. In Figure 62 this
was also where the standard free energy rose to a certain value and remained
fairly constant with increased concentration. Fuerstenau [23] concluded
that hemi-micelle formation is accompanied by the standard free energy of
adsorption becoming sharply more negative then leveling off. The standard
\
free energy of adsorption for Aerosol A-196 remained constant to about
3.0 X 10" mole fraction beyond which it dropped towards more positive
values indicating a loss of driving force for adsorption. The vacuum de-
watering data indicated a severe increase in the amount of retained solu-
tion also beyond this concentration. These concentrations, however, are
approaching the CMC and micellation in the bulk, precipitated by the early
hemi-micellation, may be beginning to compete with the surface phase for
molecules. Extended three-dimensional micelles may also be forming at the
interface. ^
Nevertheless, the added correlation of the behavior of the standard
free energies more fully supports the model proposed by Fuerstenau for sur-
factant adsorption on a well defined crystal,alumina. Therefore, let us
adopt and elaborate on this model which may be appropriate even for the
heterogeneous surfaces found in coal.
Let us now construct a general standard free energy of adsorption
curve for the surfactant adsorption on coal. Figure 68 shows this curve
along with schematic interpretations of the molecules at the interface.
-------
130
-464ns
Ln MOLE FRACTION CONCENTRATION
FIGURE 68. GENERAL BEHAVIOR OF THE STANDARD FREE ENERGY OF ADSORPTION
FOR SURFACTANT ADSORPTION ON COAL.
-------
131
At extremely low Surfactant concentrations the free energy of adsorption
should be approximately constant; the assumptions of Langmuirian type ad-
sorption prevailing. (The Stern-Grahame adsorption equation is actually
a form of the Langmuir if the adsorption energy is assumed to remain con-
stant). Specifically, there are no interactions between adsorbing mole-
cules and only the coulombic attraction potentials are operative. This is
stage 1 of Figure 68. In addition to any of the specific surface attrac-
tive potentials, there is always the hydrophobic repulsion of the hydro-
carbon chains by the bulk water molecules driving the surfactant mole-
cules to any interface. This hydrophobic effect has been treated by Tanford
[26] and is the result of the water molecules preferring to retain their
mutual hydrogen bonding rather than attempt coping with the hydrocarbon
chains which offer little bonding capability with the water. It should
also be noted that the water molecules and hydroxyl and hydronium ions of
the water are also adsorbed on the coal surface. If the coulombic attrac-
tion is between the positive solid coal surface and an anionic surfactant,
then the surfactant adsorbes by an ion exchange with, for example, a more
weakly adsorbed hy.droxyl ion. For a cationic surfactant the attraction
would be to the adsorbed layer of negative water ions (-OH) and may result
in the surfactant adsorbing in the outer Helmholtz plane.
As the number of molecules adsorbed is increased, the electric double
layer is being neutralized and the number of adsorption sites is being re-
duced which results in a more positive trend in free energy. This is shown
as stage 2 of Figure 68. A concentration is then reached, stage 3, at
which the adsorbed molecules are close enough for the Van der Waals at-
traction between hydrocarbon chains to become a significant factor.
-------
132
the water molecules surrounding them and further lower the overall free
energy of the system through an elimination of the high energy hydrocarbon-
water interface. Due to the great number of -CH?- groups per surfactant
molecule the Van der Waals attraction, on a per molecule basis, is large
compared to the coulombic repulsion of the head groups. This has been
found to be the case in studies of the standard free energy of bulk liquid
micellation [64], in which for example, with dodecyltrimethyl ammonium
bromide, the hydrocarbon contribution favored micellation by -14.0 RT
and the coulombic contribution opposed it by only +0.84 RT. Therefore,
in stage 3, the standard free energy of adsorption increases (i.e. tends
toward lower negative values) which favors adsorption.
The beginning of stage 4 occurs when the number of adsorption sites in
the first layer become exhausted, and coulombic repulsion due to the in-
creased density of like charges of the adsorbed head groups, now can no
longer be offset by the hydrocarbon association. Adsorption in the subse-
quent layer above the first monolayer now begins with the molecules still
bonding hydrocarbon to hydrocarbon but arranged so as not to increase the
number of like charges in the first layer, i.e., they adsorb in reverse
orientation. This would be the onset of hemi-micelle formation as des-
cribed by Fuerstenau. The free energy should be constant since adsorption
is now occuring on a new surface, one of hydrocarbon chains. Only one type
of adsorption mode is operative, i.e., hemi-micelle formation. The actual
value of the free energy, therefore, should be similar to the standard
free energy of bulk micellation and even somewhat less, i.e., more nega-
tive, due to the existence of the nucleated phase.
-------
133
Stage 5 is initiated when the number of adsorption sites again be-
comes exhausted, the repulsion of the like head charges building up in
the outermost layer becomes dominant and micellation in the bulk liquid
is being approached. Therefore,in state 5 the standard free energy be-
comes more positive.
*•
If we now re-examine the standard free energy of adsorption of Aerosol
A-196 we can deduce, in light of the above treatment, that the data re-
flects the adsorption stages 3,4, and 5 quite clearly. The free energy
curve for Aerosol-OT was constructed assuming the above treatment to be
appropriate.
The above model was presented envisioning the form of the hemi-micel-
ation as a bilayer arrangement, which as stated previously, is one of the
micellar forms possible in a bulk liquid situation. However, as pointed
out by Tanford [65] this form is not only possible, but actually optimal
specifically for surfactants possessing two hydrocarbon chains per molecule;
a category which includes both Aerosol A-196 and Aerosol - OT. Furthermore,
Saleeb and Kitchener [59] have speculated on this type of behavior for the
adsorption of Aerosol - OT on carbon blacks. An illustration of the pos-
sible bilayer for these surfactants is shown in Figure 69.
Let us now return to examining the data for each of the specific sur-
factants. The last anionic surfactant to be considered is sodium dodecyl
sulfate. Examination of the standard free energy curve for this surfactant
reveals that it consists of adsorption stages 2, 3, 4 and 5 up to about
10~5 mole fraction surfactant. The data from 2.0 X 10"5 mole fraction
and higher may be suspect since this is the concentration range in which
there was a minimum in the solution-air interfacial tension calibration
curve, which was discussed above. However, assuming the free energy model
-------
134
HYDROCARBON TAILS
IONIC HEAD GROUP
FIGURE 69. POSSIBLE ADSORBED BILAYER OF SURFACTANT MOLECULES WITH
TWO HYDROCARBON CHAINS PER MOLECULE.
-------
135
to be correct, a concentration of 2.0 X 10~6 mole fraction should be the
beginning of the hemi-micelle formation. The vacuum dewatering data of
Figure 45 are in reasonable agreement, as evidenced by the wide minimum
that appears in this range. The existence of such a minimum is consistent
with the behavior of the other anionic surfactants. Using these data and
proceeding as before, we find the effective specific surface area of the
2
coal to be 533 cm /gm which is in excellent agreement with the values
obtained for Aerosol A-196 and Aerosol-OT.
Therefore agreement seems to exist among the anionic surfactants as
to an effective specific surface area of the coal calculated on the basis
of the "hemi-micelle" model.
Let us now examine the data for the cationic surfactant, dodecyl pyri-
dinium chloride. The vacuum dewatering data of Figure 44, shows that the
residual solution contents exhibited no minimum, but rather began to de-
crease at about 2.0 X 10 mole fraction and level off at about 7.0 X 10
mole fraction. The standard free energy of adsorption, shown in Figure 64,
is just about constant between these concentrations. This would tend to
reflect a constant mode of adsorption. Since the residual solution con-
tents decreased, the coal surface became more hydrophobic and therefore
the hydrocarbon chains of the adsorbed molecules must be pointed into the
liquid. However, the ionic end of the molecule is positive as is the coal
solid surface, and not negative as in the anionic cases, which seems to be
an incompatable situation. An explanation has already been stated in the
description of the general adsorption standard free energy curve. The
dodecyl pyridinium ion may be adsorbed in the outer Helmholtz plane separ-
ated from the coal surface by an adsorbed layer of hydroxyl ions and hydra-
tion water. This is also reasonable because dodecyl pyridinium chloride
-------
136
is derived from a quarternary ammonium compound and such derivatives
according to Hummel [66] "tenaciously retain water" as evidenced by his
infra-red absorption experiments.
This assumption is also supported by the electrokinetic studies of
Cambell and Sun [22] who found that the adsorption of a cationic surfac-
tant, dodecylamine acetate, on coal completely reversed the polarity of
the zeta potential while adsorption of the anionic surfactants effected
little change. Similar electrokinetic behavior has also been reported by
Saleeb and Kitchener [59] in the adsorption of Aerosol-OT and cetyltrimethyl
ammonium bromide on carbon blacks. The charge on molecules adsorbed in the
outer Helmholtz plane would more likely have a direct effect on influencing
the measured potential at the plane of shear, i.e., the zeta potential,
than molecules adsorbed at the inner plane. Let us restate the electro-
kinetic findings. When coal, which initially has a negative zeta potential
and therefore a positive surface, adsorbs a cationic surfactant, presumably
In the Outer Helmholtz plane, the zeta potential changes greatly even
enough to change polarity. In adsorbing an anionic surfactant little
change in the zeta potential is observed. One would expect therefore that
a solid which initially had a positive zeta potential and hence a negative
surface would adsorb an anionic surfactant in the Outer Helmholtz plane
effecting a great change in the zeta potential even as far as changing
polarity. This is exactly what Fuerstanau [23] has observed for alumina
which does have a positive zeta potential of + 60 mv in neutral aqueous
solutions.
With the ionic head of the dodecyl pyridinium ion fixed in the outer
Helmholtz plane, an additional adsorption orientation is possible, that
-------
137
with the hydrocarbon chains adjacent to the less water avid areas of
the coal surface and not out into the solution, as illustrated in Figure
70, since they would like to get as far out of the water as possible. And,
in addition, the hydrocarbon chains can also associate with the aliphatic
portion of the coal. The net orientation, however, must be as first sup-
posed. If we then assume that at 2.0 X 10"6 mole fraction the number of
molecules adsorbed begins to be dense enough to show a net hydrocarbon
shield and cover the hydrophilic groups then we can calculate another ef-
fective specific surface area for the coal. With the molecules able to
adsorb in two orientations in the first layer, no minimum in the retained
solution contents would be expected to occur. No great increase in the
adsorption free energy due to hydrocarbon chain association would occur as
before since now the hydrocarbon chains can associate with the solid sur-
face, and with some oriented up and some oriented down the number of
associations possible per molecule is decreased. For example, half the
total number in the upper plane and half in the lower plane. Therefore,
there is insufficient association on the liquid side of the outer Helmholtz
plane to force formation of a hemi-micelle.
The standard free energy curve reflects this model, i.e., a constant
value indicative of coulombic attraction which then monotonically decreases as
the concentration increases.
The specific surface area is then calculated to be 2.0 X 10" mole
fraction, i.e., 2.75 X 10"7 moles or 1.7 X 1017 molecules of dodecyl py-
ridium chloride adsorbed per gram of coal. The value obtained for the area
ft «
of the molecule was 34A2. These values give an area for the coal of 563 cm /
gm
, also in agreement but somewhat larger than those of the anionic sur-
-------
138
OUTER HELMHOLTZ
PLANE
(*)
INNER HELMHOLTZ
PLANE _OH-
ALTERNATE
ADSORBED
ORIENTATION
-OH" —HQH
COAU+)
FIGURE 70. ADSORBED LAYER OF DODECYL PYRIDINIUM IONS SHOWING AN
ALTERNATE ORIENTATION FOR THE MOLECULES.
FIGURE 71. CLOSELY PACKED ADSORBED LAYER OF DODECYL PYRIDINIUM IONS
-------
139
factants. Since the residual solution contents never increased again past
7 X 10 mole fraction the net number of hydrocarbon chains facing the
solution must have remained constant. Using the area of 500 cm2/gm for
the coal surface and values from the adsorption isotherm, we can back
calculate the size of the molecule adsorbed at 7 X 10~6 mole fraction.
This calculation gives an area per molecule of only about 9A2. Oddly
enough this is approximately the area of the methyl group at the end of a
hydrocarbon chain. Solidified, hydrocarbon chains have been determined
o o9
[63] to be 3.5A apart. This diameter gives a circular area of 9.6A which
i
appears to indicate that the molecules of the adsorbed surfactant have
compacted themselves so as to resemble the packing of solid hydrocarbons,
something like this is shown in Figure 71. Beyond this concentration more
molecules are ".dsorbed but the net orientation must be such that the
number of hydrophobic sites remained constant. A net pair-wise adsorp-
tion as described above would be reasonable.
Let us now consider the last surfactant investigated, ie., nonionic
Triton X-114. The isobaric plots of Figure 38 show that the retained
solution content began to decrease as the concentration of surfactant in the
solution exceeded about 3.0 X 10"7 mole fraction. The vacuum dewatering
data of Figure 43 are in agreement with this and furthermore show some-
thing of a small minimum at 9.5 X 10"7 mole fraction. The adsorption
isotherm shows that the amount of surfactant adsorbed continues to in-
crease through these concentrations. The standard free energy of ad-
sorption curve shows a constant value up to about 5.0 X 10 mole
fraction after which it declines. A constant free energy again implies
a constant mode of adsorption, Langmuir type adsorption as before.
-------
140
The Triton X-114 molecule is not ionic, but it does have polar
groups. It is not only highly polar due to the large number of ethylene
oxide units but also because of the hydrophilic hydnoxyl group con-
tained at one end: all of which leads to the hydrophilic portion of the
molecule having a negative polarity. Since the molecules of Triton
X-114 do adsorb, causing the coal surface to become more hydrophobic
Therefore, the molecules must have arranged themselves such that the
hydrocarbon chains were extended away from the coal surface and, as
implied, the polyoxyethylene chain must have adsorbed onto the coal
surface. The constant free energy of adsorption is then the free energy
of the polar interaction. At low concentrations, the Triton X-114 mole-
cules probably stretch out allowing the entire length of the polyoxy-
ethylene chain to adsorb onto the coal surface. Schick [67, 68] has
investigated molecules of this type adsorbed at the aqueous solution-
air interface, and finds that the polyoxyethylene chains must be coiled
in the water phase at the interface to explain their apparent adsorbed
area. If stretched straight the Triton X-114 molecule should have an
end cross sectional area similar to the dodecyl pyridinium ion because of
°2
the similar aromatic centers which would be 34A . The area obtained for
the Triton X-114 molecule from the Gibbs equation and the interfacial
°2
tension data of Figure 48 was 56A . Schick found similar discrepencies
and, hence, the coil theory. Furthermore, the molecular model of Figure
14 was found to be quite flexible. This implies that the molecule can
contort itself to give an area appropriate to the prevailing circumstances,
I.e., it can present a variable adsorbed surface area per molecule. This
makes a determination of a specific surface area of the coal an arbitrary
matter. However, we can make some sound estimates as to the adsorption
-------
141
process. If the ethylene oxide chains are adsorbing, then no improvement
in hydrophobicity of the coal surface would be expected to occur until
the molecules began to impinge upon one another. The length per ethylene
o
oxide unit in a polyoxyethylene chain has been reported as 2A [63]. The
average length of the polyoxyethylene chain in Triton X-114 is 7.5 units
o
giving a length of 15A. Using this as an effective diameter we can deter-
mine the effective circular area occupied by the chain. This approach is
similar to Onsager's covolume concept for highly asymmetric molecules [69].
°?
The area obtained is 177A . Let us assume that this is the area occupied
by the adsorbed molecules of Triton X-114 at the point of crowding, i.e.
when the molecule could be in an "L" configuration as shown in Figure 72.
If we choose 4.0 X 10" mole fraction as the concentration at which this
-8
happens, then from the adsorption isotherm about 5.0 X 10 moles or
3.0 X 1016 molecules are adsorbed per gram of coal. Multiplying this by
Oy 2
177A we arrive at a value of 533 cm /gm for the specific surface area of
the coal. Again in excellent agreement with the values determined with
the other surfactants.
We can also consider the slight minimum in the retained solution con-
tents at 9.5 X 10"7 mole fraction since there is also an indication of a
change in the standard free energy curve at this concentration. Assuming
that this may be an attempt at hemi-micelle formation, we can determine
2
an area of the adsorbed molecules using the value of 500 cm /gm for the
specific surface area of the coal and see if the value is reasonable. At
7 ?fi
9.5 X 10"7 mole fraction of Triton X-114, 1.0 X 10' moles or 6.0 X 10
molecules are adsorbed per gram of coal. This is approximately two times
the number observed previously. Dividing 500 cm2/gm by this value we
-------
142
FIGURE 72. POSSIBLE "L" CONFIGURATION OF THE ADSORBED TRITON X-114
MOLECULE AT THE POINT OF CROWDING.
FIGURE 73. CHANGE IN THE ADSORBED AREA PER TRITON X-114 MOLECULE
TO ACCOMODATE MORE ADSORBED MOLECULES.
-------
143
obtain an area per molecule of 83A2 which is approximately half the
area of that determined before. Doubling the number of molecules ad-
sorbed decreased the area occupied by each by half. The molecules seem
to have stood up more as is illustrated in Figure 73. The area obtained
is reasonable since the smallest area possible would be about 34A2 and
the area at the liquid-air interface was only 56A2.
At this point, let us summarize. A specific qualitative model for
the adsorption behavior of the surfactants onto the coal surface was
formulated, and using this model, values for the specific surface area of
the coal occupied by the adsorbed surfactant molecules were determined.
The values obtained, which are summarized in Table X, were found to be
consistant in the case of all five surfactants involved. Therefore, the
proposed model, which was adopted from an explanation of surfactant ad-
sorption on a rather homogeneous and well defined crystal, viz., alumina
seems appropriate for the adsorption behavior of surfactants on complex,
heterogeneous coal.
The model is successful in producing at least a consistent set of
results. Let us now consider these results. What we have found con-
sistency in is a specific surface area of the coal. The topographical
heterogeneity of coal has already been introduced in section III. COAL.
This must be critically taken into account when determination of a
specific surface area is considered. Methods by which specific surface
areas are determined rely on some observable property dependent upon the
extent of the surface, which is theoretically relatable to the actual sur-
face area. Surface areas determined by different methods should, in
general, be expected to be different. The main difference between methods
-------
144
TABLE X
SPECIFIC SURFACE AREA OF THE COAL
DETERMINED ON THE BASIS OF THE SURFACTANT ADSORPTION MODEL
SURFACTANT AREA PER ADSORBED
MOLECULE
Aerosol A - 196
Aerosol - OT
Sodium Dodecyl Sulfate
Dodecyl Pyridinium Chloride
Triton X-114
X2
68
64
70
59
34
9
177
83
SPECIFIC SURFACE AREA
OF THE COAL
(em2/ g)
512
482
527
533
563
500
533
500
-------
145
and their results is in the extent to which the surface irregularities can
be considered [70].
One estimate of the specific surface area of the coal is a calcu-
lation of the geometric surface area, assuming the coal particles are
spherical. Assuming the particles to have an average diameter of 324 ym,
ie., the log average value used above in the treatment of capillary theory,
and that the density of the coal was 1.3 g/cm, the specific surface area
2
obtained is 142 cm /g. This value would be indicative of the lowest value
expected, since no surface irregularities were considered. Even assuming
that all the coal particles were of the smallest diameter known to be
2
present, i.e., 250 ym, only brings the value to 185 cm /gm.
We have already determined another value for the specific surface area
of the coal based on a method which would tend to reflect a more accurate
description of the situation. This was the value generated from the data
of the dewatering experiments. Although the bed was assumed to be
approximated by a bundle of cylindrical capillary tubes, the volume used
in calculating the area includes the volume occupied in any accessible
2
surface indentation or irregularity. The value obtained was 284 cm /gm,
which was determined for a partially saturated bed. If empirically cor-
2
rected for complete liquid saturation in a "flow" situation it was 342 cm /gm.
Even the "flow" value would be expected to be low since theoretically this
area is assumed to be the area seen by the contacting liquid as it flows
through the pore system established by the packed particles. Stagnant and
blind passages are not included, nor the area between touching particles;
nor would the area reflect the situation for particles of complex shape in
this simple treatment.
-------
146
Another method of surface area determination is by gas adsorption.
As stated previously, typical values of the specific surface areas of
2
coal determined by gas adsorption are in the range of 100 - 200 m /gm,
four orders of magnitude higher than the values above! Nitrogen [71] is
a typical gas used in this determination and its molecular area is only
°2 °?
about 16A . In this case, any space on the order of 16A will be included
in the determination. This includes much of the internal pore area of each
coal particle.
The area obtained by the surfactant adsorption of the present work
2
was found to be about 500 cm /g, which more nearly reflects the geometric
and capillary areas than that found by gas adsorption. The only reason-
able conclusion that this allows is that the surfactant molecules must be
too large to "see" most of the area of the coal contained in the internal
pores of its gel structure and that they adsorb only on the accessible
irregularities of the "external" surface of the particles.
Further support for this conclusion comes from the work of Saleeb and
Kitchener [59] who investigated the adsorption of Aerosol-OT and cetyl
trimethyl ammonium bromide on carbon blacks. Such carbon blacks have no
internal pores to speak of and are usually used as reference surfaces in
gas adsorption studies. The particle diameters used were less than a micron.
°2
The surface areas, therefore, determined form nitrogen gas (area = 16A )
adsorption are the external surface areas of all the particles, and should
be equally accessible to the larger si/rfactant molecules. From a knowl-
2
edge of the nitrogen specific surface area (about 10 m /gm) and the number
of moles adsorbed at the CMC, beyond which no more adsorption was found,
they determined the area of the adsorbed molecules. They found excellent
-------
147
agreement with the areas of molecular models of the surfactants oriented
vertically. Their value for Aerosol-OT of 70A2 is already listed in
Table IX.
Further support can be obtained by noting that the free energies of
heini-micellation should be similar to the standard free energies of micel-
lation of surfactants in the bulk liquid. The standard free energies of
micellation of a few surfactants: sodium dodecyl sulfate, sodium decyl
sulfate, dodecyl trimethyl ammonium bromide, and dodecyl pyridinium bromide,
for example, have been reported as -5.3 kcal/mole [73], -4.5 kcal/mole
[73], -7.8 kcal/mole [64], and -5.03 kcal/mole [74] respectively in agree-
ment with the magnitudes of hemi-micellation determined "in the present work
2
using a value of 500 cm /gm for the effective specific surface area of
the coal.
The standard free energies of hemi-micellation should even tend
towards more negative values than those of bulk liquid micellation, due
to the presence of the highly oriented adsorbed first layer, which would
tend to reduce the entropy of the formation of hemi-micelles. This was
found to be the case in genera';. An example is the standard free energy
at the point of hemi-micellation calculated for sodium dodecyl sulfate,
shown in Figure 65. The value was about -6.1 kcal/mole, approximately
1 kcal/mole more negative than the value for bulk micellation of -5.3
kcal/mole cited above.
The fact that better agreement with the theory is expected for
;
energy values more negative than those of bulk micellation further sup-
ports the magnitude of the specific surface area derived from that theory.
i
According to Equation (6) , the Stern-Grahame adsorption equation, a
-------
148
trend towards more negative energies favors a smaller specific surface
area, not larger by orders of magnitude.
Additional evidence supporting the order of magnitude obtained for
the specific surface area of the coal comes from an examination of the
standard free energies of adsorption calculated for each surfactant ad-
sorption isotherm. To calculate the standard free energy of adsorption
curves plotted in Figures 62 - 66 for each surfactant, a value for the
specific surface area accessible to the molecules had to be assumed.
Therefore, the actual magnitude of the energies obtained was determined
by this value. These values, however, are only sensitive to the order of
magnitude of the surface area chosen, because of the logarithmic re-
lationship between the two variables as contained in Equation (6). The
2
value used, chosen by hindsight, was 500 cm /g.
The magnitudes of the standard free energies of adsorption calcu-
lated from this value ranged about -5 to -6 kcal/mole, which is in the
region of physical adsorption, in agreement with the supposition of the
adsorption model presented. The region of chemisorption is assumed to
be considerably higher, e.g., in excess of -10 kcal/mole and often larger
than-20 kcal/mole [72].
The curves of Figures 62-66 also show that the magnitude of the
standard free energy of adsorption for each surfactant does not vary great-
ly through the concentration range investigated, as would be expected for
a situation in which, through this range, the adsorption process was only
changing modes of physical adsorption, the energy of each being small and
approximately the same.
The standard free energy of adsorption of sodium dodecyl sulfonate
on alumina reported by Fuerstenau [23], ranged from -3.6 to -5.0 kcal/mole,
-------
149
with the higher value at the point of hemi-micellation, in excellent
agreement.
Therefore, the proposed adsorption model is supported not only by
consistent results but by the actual value of these results calculated
on the basis of the model.
Let us now consider why the final retained solution contents of the
coal bed show such a large increase for additions of Aerosol A-196, Aero-
sol -OT, and sodium dodecyl sulfate, the anionic surfactants, at the point
of hemi-micelle formation, a case which, in fact, makes the coal surface
even more hydrophilic than it appeared originally.
The adsorption model, as presented, although sufficiently satis-
factory to explain the observed adsorption phenomena, however, is too
simple to be all encompassing. For example, the coal surface during the
development was regarded as being homogeneous in net positive charge, yet
as explained earlier, coal is a heterogeneous material. The measured
zeta potential is actually an average of all of the local potentials at
the plane of shear. In local areas some patches may be positive and
some negative but arranged as to give a net negative zeta potential to
whole coal.
Let us now envision the adsorption of the surfactants on a surface
more like that of real coal. Let us assume, as illustrated in Figures
74 - 77 that a typical section of the whole coal surface is made of ten
adsorption sites. Let six of these sites represent the aromatic ring
structure, two represent a mineral, and two represent aliphatic hydro-
carbon areas, as shown in Figure 74. The potential of the aromatic sites
is positive, the mineral negative, and the aliphatic areas neutral.
-------
150
COAL 1
0
MINERAL
/ %
COAL
/•j_\
COAL
CH2
FIGURE 74. PORTION OF A HYPOTHETICAL REAL COAL SURFACE WITH TEN
POSSIBLE ADSORPTION SITES.
COAL
lAi H_+
MINERAL
(+)
OH
"COAT
CH2
FIGURE 75. PORTION OF A HYPOTHETICAL REAL COAL SURFACE WITH AD-
SORBED WATER MOLECULES.
-------
151
Figure 75 shows a possible situation when this coal surface is in
contact with water alone. The aromatic portions have attracted the OH"
ions from water and/or the negative poles of the undissociated water
molecules. The hydronium ions and the other pole of the water mole-
cules on the other hand have been attracted to the mineral areas. The
aliphatic sites, cyclic or otherwise, are hydrophobic and do not attract
the water. In this example, therefore, 6/10 sites are hydrophilic and
4/10 are hydrophobic, which appears to describe the coal used in these
experiments.
In Figure 76 the coal surface is shown with an adsorbed monolayer of
anionic surfactant. The aromatic sites have adsorbed as we have already
assumed. The mineral sites, with their opposite charge, have forced the
ions to be adsorbed in the Outer Helmholtz Plane in much the same way as
was speculated for the adsorption of a cationic surfactant on the positive
areas of the coal, and are possibly oriented as shown. The aliphatic
sites have adsorbed the hydrocarbon chains of the surfactant molecules
since this area is governed by the Van der Waals attraction between hydro-
carbon chains and the hydrophobic effect. In this situation only 4/10
of the surface sites are hydrophilic to the water phase and 6/10 hydro-
phobic. Therefore, the average surface has become slightly more hydro-
phobic than before.
Figure 77 shows the case when adsorption occurs in the second layer,
i.e., the point of hemi-micellation as described above. The portions of
the surface of Figure 76 where hydrocarbon chains were extended to the
water phase have adsorbed the hydrocarbon chains of the surfactants.
Areas with head groups outwards cannot attract hydrocarbon chains and,
-------
152
COAL
MINERAL
COAL
COAL
CH0
FIGURE 76. PORTION OF A HYPOTHETICAL REAL COAL SURFACE WITH AN
ADSORBED LAYER OF ANIONIC SURFACTANT MOLECULES.
Jill
COAL MINERALJ
(-)
FIGURE 77. PORTION OF A HYPOTHETICAL REAL COAL SURFACE WITH THE
ADSORBED ANIONIC SURFACTANT MOLECULES IN THE HEMI-
MICELLE STATE.
-------
153
In addition, must repel the like charges so that no second layer can
be formed. Therefore, in this case 10/10 of the sites have become hydro-
pMlic and none are left hydrophobic. All the sites which were made
hydrophobic by the adsorption of the surfactants at the lower concen-
trations are changed at'the point of hemi-micellation Therefore, as
this example illustrates, it is possible to make the coal surface much
more hydrophilic than it can be made hydrophobic by the addition of
these surfactants.
Up to this point we have attempted to correlate the complex de-
watering behavior induced by the addition of the surfactants with an
overall adsorption model. The model, which incorporates changes in the
mode of physisorption, seems appropriate enough to explain the changes
in retained solution content of the coal beds, as a function of the con-
centration of the surfactant present. We have established that all the
surfactants investigated, improved the residual solution content at some
point. We have also found that should hemi-micellation on the coal sur-
face occur, the coal becomes drastically more hydrophilic. With this in
mind let us compare the maximum degree to which each surfactant reduced
the liquid retention of the coal beds. Two ways of comparing the sur-
factants with respect to this are: (1) by the maximum percentage decrease
in retained liquid content effected by addition of the surfactant over
the pure water case, and (2) by the lowest retained liquid content value
obtained with each surfactant. To compare the surfactants this way on
as equal a basis as possible it was decided to use the 20 min. - 1 atm
exposure data of the vacuum dewatering experiments of each surfactant
presented in Figures 42 - 46. Table XI lists the values determined for
each surfactant according to both methods. As evidenced in the table
-------
154
TABLE XI
SURFACTANT EFFECTIVENESS
SURFACTANT Max. % decrease Lowest wt. % retained
over pure water (wt. %)
Dodecyl Pyridinium
Chloride
Aerosol -OT
Triton X-114
Aerosol A-196
Sodium Dodecyl Sulfate
-47
-45
-42
-35
-26
3.2
2.2
2.8
4.2
4.8
-------
155
both methods rank the surfactants in the same order with the exception
of cationic dodecyl pyridinium chloride.
Let us now examine this ordering more closely for the two similar
surfactants Aerosol-OT and Aerosol A-196. From Table XI we can see that
an effective monolayer of Aerosol-OT seems more hydrophobic than an ad-
sorbed monolayer of Aerosol A-196. As pointed out in Section VI. MATERIALS
(Surfactants), these molecules differ only in their hydrocarbon chain
groups, which are the entities the adsorption model assumes to be in
contact with the water phase at this point. Also as stated earlier, a
convenient and concise way of comparing the relative hydrophobicities of
the surfactants is by the HLB index and group numbers, a partial listing
of which was given in Table I. Using Table I we can calculate the effec-
tive hydrophobic number appropriate for the two surfactants, assuming
that the number for the -CH«- groups is additive irrespective of the chain
being cyclic or branched. The number for Aerosol-OT was calculated for
only the two ethyl hexyl groups and for Aerosol A-l 96 for the two cyclo-
hexyl groups. The numbers are, respectively -7.6 and -5.7. The number
for Aerosol A-196 is 25% less than that of Aerosol-OT. This would imply
that the monolayer of Aerosol A-196 was 25% less hydrophobic than that
of Aerosol-OT. From TableXI, the maximum percentage decrease in retained
liquid content for Aerosol A-196 is 22% less than that of Aerosol-OT, in
remarkable agreement; this, however, may be fortuitous.
Due to the dissimilarity between the other surfactants a detailed
comparison such as above was not considered as valid and was not attempted.
Furthermore, use of the HLB index and group numbers may not be exactly
appropriate in determining the relative hydrophobicity of adsorbed mono-
-------
156
layers, since the group numbers are for parts of a molecule which are in
a free state; and, it is not obvious how adsorption of a group alters its
hydrophobicity. Nevertheless, additional agreement with the relative
order of the surfactants as listed in TableXI.was found using the HLB
as a guide. One example, using the total HLB index for each surfactant
is also in agreement. This is the case of Triton X-114 and sodium dodecyl
sulfate which have HLB index numbers of 12.4 and 40 respectively. The
high number, implying more hydrophilic, for sodium dodecyl sulfate comes
from the hydrophilic group number for the SO. group of the molecule which
according to Table I is the most hydrophilic group listed. This may be
the reason why sodium dodecyl sulfate was the poorest dewatering agent
shown in both columns of Table XI.
The fact that dodecyl pyridinium chloride had the maximum percentage
decrease and only a fair value for the lowest retained amount of solution
may be indicative of the way in which it was assumed to be adsorbed, i.e.,
in the outer Helmholtz plane, on top of an adsorbed water layer- The
correct orientation of the molecule would allow for the improvement of
dewatering, but the underlying water layer would prevent the lowest value
of retained water from being achieved.
Therefore support for the proposed surfactant adsorption model also
comes from a more detailed examination of the molecular structure in
relation to the hydrophobic nature of each adsorbed molecule.
-------
157
X. CONCLUSION
The addition of the surfactants investigated has been found to have
a two-fold influence on the dewatering of coal. They were found to effect
the pressure differentials required for dewatering in addition to the
residual water contents of the coal beds attainable by this dewatering.
Both effects are the result of surfactant adsorption. Adsorption
at the liquid-air interface, results in a decrease in the interfacial
-».
tension between the two phases. The effect this decrease had on the
pressure differentials required for dewatering was found to be in agree-
ment with that predicted by the capillary theory applied to the system.
Adsorption at the solid-liquid interface was correlated with the
complex behavior of the residual water contents as a function of sur-
factant addition. A comprehensive model for the adsorption of the sur-
factants onto the coal was presented, based on the Stern-Grahame theory
of adsorption at an electrical double layer. The model allowed for the
mode of physisorption to change as the amount of surfactant adsorbed in-
creased, and also for a phenomenon known as hemi-micellation. Using the
model, consistent and reasonable results were found for the specific sur-
face area of the coal and for the standard free energies of adsorption.
The model was also found to be appropriate when the heterogeneous nature
of the coal was considered. Furthermore, the hydrophobicity of the
molecular groups of the molecules, expected from the model to be con-
trolling the hydrophobicity of the interface, was also found to be in
agreement with that predicted by other means.
Therefore, a complete knowledge of the adsorption behavior of the
-------
158
surfactants in the coal-water system is the key in determing the cri-
teria for optimum utilization in a dewatering situation.
Finally, an ideal surfactant for aiding the dewatering of coal
should possess the following characteristics:
1. The molecules should adsorb on the coal oriented with the hydro-
carbon tails extended to the water phase.
2. The hydrocarbon tail should be as hydrophobic as possible, and
still retain a degree of water solubility of the whole molecule. (An
increase in the number of hydrocarbon tails per molecule may be necessary).
3. The head group should be the least hydrophilic group possible,
and still retain the solubility of the whole molecule.
4. The adsorbed molecule should have a large effective area so that
fewer molecules are required to affect the interfaces.
5. The surfactant should decrease the interfacial tension at the
liquid-air interface as much as possible at the lowest concentration.
6. The molecules should not be prone to henri-micellation or if
they are, it should occur at a sufficiently high concentration, to be
avoided.
It should be recognized that these conclusions may be strongly de-
pendent on the rank of the coal under investigation as well as the nature
of the mineral matter therein. Therefore it is recommended that the data
and conclusions generated herein be examined next as a function of these
variables.
-------
159
APPENDIX A
Surfactant Adsorption Kinetics
The lowering of the liquid-air interfacial tension of aqueous sur-
factant solutions has been attributed to the adsorption of the solute
molecules or ions at the interface. A finite time is required for the
surface system to develop and establish equilibrium with the two bulk
phases. In the case of aqueous surfactant solutions the time required to
establish equilibrium can vary from less than a second to a few days [75],
i
depending primarily upon the specific surfactant molecule and its concen-
tration in the solution. This effect is often referred to as surface aging.
There exists some controversy in ascribing the rate process responsible
for this effect. One postulate [76] is that the ions arriving at the in-
terface first, establish an electrical double layer, which sets up a
potential barrier to inhibit the introduction of additional ions. Other
theories are concerned with a pre-existing surface phase in which molecular
penetration and reorientation is considered [77].
Regardless of the appropriate theory, the existence of an aging effect
in the interfacial tension of surfactant solutions should be recognfzed and
taken into account whenever an application of surfactants is considered.
The precautions taken in the present work to alleviate any problems
with this, have already been stated. Furthermore, it was assumed that ad-
sorption of the surfactant molecules at the solid-liquid interface was also
time dependent, and at least of the same order of magnitude as that found
at the liquid-air interface. Therefore the time which was allowed for
the surfactants to adsorb onto the coal was much, much greater than the
longest time found for a solution surface to come to equilibrium. The
time allowed was usually on the order of an hour.
-------
160
APPENDIX B
TABLES OF DATA FOR THE ADSORPTION ISOTHERMS
-------
TABLE B 1
DATA FOR THE ADSORPTION ISOTHERM OF AEROSOL A-196
SURFACTANT
CONCENTRATION
x
(mole fraction)
9.37 x 10"7
1.94 x 10"6
3.86 x 10"6
9.49 x TO"6
1.85 x 10"5
4.88 x 10"5
8.55 x 10"5
1.14 x 10"4
1.37 x 10~4
1.71 x 10"4
1.85 x 10""4
1.96 x 10"4
1.14 x 10"3
MOLES OF
SOLUTION
n
1.95
1.95
1.96
2.00
2.05
1.94
2.22
2.49
2.77
3.32
3.59
3.87
3.51
WEIGHT OF
COAL SAMPLE
w
(g)
20.0098
20.0098
20.0098
20.0098
20.0098
20.0200
20.0200
20.0200
20.0200
20.0200
20.0200
20.0200
19.9662
CONCENTRATION
CHANGE OF SOLUTION
ON ADSORPTION
AX
3.22 x 10-7
1.26 x 10"6
2.76 x 10"6
5.77 x 10"6
1.04 x 10"5
3.71 x 10"5
5.29 x 10"5
6.33 x 10"5
6.86 x 10"5
6.44 x 10"5
6.94 x 10"5
6.14 x 10"5
1.30 x 10"4*
AMOUNT ADSORBED
PER GRAM OF COAL
(moles/g)
3.13 x 10'8
1.22 x 10"7
2.71 x 10"7
5.76 x 10"7
1.07 x 10"6
3.60 x 10"6
6.58 x 10"6
7.87 x 10"6
9.49 x 10"6
1.07 x TO"5
1.25 x 10"5
1.19 x 10"5
2.29 x 10"5
* determined grav1metr1cally
-------
TABLE B 2
DATA FOR THE ADSORPTION ISOTHERM OF TRITON X-114
SURFACTANT
CONCENTRATION
x
(mole fraction)
1.89 x 10"8
3.39 x 10"8
9.44 x TO"8
1.88 x 10"7
2.82 x 10"7
3.73 x 10'7
4.67 x 10"7
9.27 x 10"7
1.38 x 10"6
1.71 x 10"6
2.25 x TO"6
3.30 x 10"6
5.27 x TO"6
MOLES OF
SOLUTION
n
2.78
2.78
2.78
2.79
2.79
2.80
2.81
2.83
2.29
2.31
2.33
2.39
2.50
WEIGHT OF
COAL SAMPLE
w
(g)
20.1948
20.1948
20.1948
20.1948
20.1948
20.1258
20.1258
20.1258
20.1258
20.1258
20.2047
20.2047
20.2047
CONCENTRATION
CHANGE OF SOLUTION
ON ADSORPTION
AX
1.64 x 10"8
3.07 x 10"8
8.28 x 10"8
1.69 x 10"7
2.54 x 10~7^
3.31 x 10"7
4.10 x 10"7
7.40 x 10"7
1.11 x 10"6
1.37 x 10"6
1.68 x 10"6
2.22 x 10"6
3.57 x 10"6
AMOUNT ADSORBED
PER GRAM OF COAL
nl
(moles/g)
2.26 x 10"9
4.22 x 10"9
1.14 x 10"8
2.33 x 10"8
3.51 x 10'8
4.61 x 10"8
5.73 x 10"8
1.04 x 10"7
1.26 x 10"7
1.57 x 10"7
1.94 x 10"7
2.63 x 10"7
4.42 x 10"7
ro
-------
TABLE B 3
DATA FOR THE ADSORPTION ISOTHERM OF DODECYL PYRIDINIUM CHLORIDE
SURFACTANT
CONCENTRATION
x
(mole fraction)
1.41 x 10"6
3.46 x 10"6
-6
9.19 x 10 °
1.53 x 10"5
3.02 x 10"5
3.91 x 10"5
5.93 x 10"5
9.64 x 10"5
1.40 x 10"4
2.21 x 10"4
MOLES OF
SOLUTION
n
2.82
2.89
2.79
' 2.81
2.83
2.85
2.89
1.77
1.83
1.94
WEIGHT OF
COAL SAMPLE
w
(g)
20.3999
20.3999
20.4030
20.4030
20.4030
20.4030
20.4030
20.4018
20.4018
20.4018
CONCENTRATION
CHANGE OF SOLUTION
ON ADSORPTION
AX
1.42 x 10"6
3.32 x 10"6
-K
8.63 x 10 °
1.29 x 10"5
2.23 x 10"5
2.86 x 10"5
4.02 x 10"5
5.59 x TO"5
8.94 x 10"5
2.12 x 10"4
AMOUNT ADSORBED
PER GRAM OF COAL
"f
(moles/g)
1.96 x 10"7
4.71 x 10"7
-6
1.18 x 10 °
1.77 x 10'6
3.09 x 10"6
3.99 x 10"6
5.69 x 10"6
4.85 x 10"6
8.02 x 10"6
2.02 x 10"5
00
-------
TABLE B 4
DATA FOR THE ADSORPTION ISOTHERM OF SODIUM DODECYL SULFATE
SURFACTANT
CONCENTRATION
x
(mole fraction)
5.67 x 10'7
1.12 x 10"6
2.25 x 10"6
2.93 x 10"6
3.93 x 10"6
5.03 x 10"6
6.68 x 10"6
1.59 x 10"5
2.05 x 10"5
2.98 x 10"5
3.83 x 10"5
3.90 x 10"5
4.98 x 10~5
5.27 x 10"5
5.76 x 10"6
6.10 x 10"5
MOLES OF
SOLUTION
n
2.78
2.79
2.80
2.80
2.81
*2.83
2.84
1.36
1.39
1.44
1.50
1.42
1.44
1.50
1.50
1.55
WEIGHT OF
COAL SAMPLE
W
(g)
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
20.0098
CONCENTRATION
CHANGE OF SOLUTION
ON ADSORPTION
AX
4.27 x 10"7
5.70 x 10'7
1.15 x 10"6
1.58 x 10"6
1.82 x 10"6
2.73 x 10"6
2.93 x 10'6
1.09 x 10"5
1.41 x 10"5
2.00 x 10"5
2.63 x 10"5
3.10 x 10"5
3.86 x 10"5
4.12 x 10'5
4.36 x 10"5
4.75 x 10"5
AMOUNT ADSORBED
PER GRAM OF COAL
"i
(moles/g)
5.93 x 10"8
7.95 x 10"8
1.61 x 10"7
2.21 x 10"7
2.55 x 10"7
3.86 x 10"7
4.16 x 10"7
7.41 x 10"7
9.79 x 10'7
1.44 x 10"6
1.97 x 10"6
2.20 x 10"6
2.78 x 10"6
3.09 x 10"6
3.27 x 10"6
3.68 x 10~6
CT1
-------
TABLE B 5
DATA FOR THE ADSORPTION ISOTHERM OF AEROSOL-OT
SURFACTANT
CONCENTRATION
x
(mole fraction)
1.09 x 10"7
1.09 x 10"7
2.18 x 10"7
2.18 x 10"7
2.85 x 10"5
MOLES OF
SOLUTION
n
2.78
2.78
2.79
2.79
2.76
WEIGHT OF
COAL SAMPLE
w
(g)
20.3994
20.3992
20.3994
20.3992
20.0662
CONCENTRATION
CHANGE OF SOLUTION
ON ADSORPTION
AX '-
1.09 x 10"7
1.09 x 10"7
1.38 x 10"7
1.63 x 10"7
-fi*
1.54 x 10 b
AMOUNT ADSORBED
PER GRAM OF COAL
s
(moles/g)
1.49 x 10"8
1.49 x 10'8
1.89 x 10"8
2.23 x 10"8
2.12 x 10'7
*determ1ned gfavimetrically
CTi
en
-------
166
REFERENCES
1. Lowery, H.H., ed., Chemistry of Coal Utilization, John Wiley &
Sons, Inc., New York, 1963, pp. 204, 307.
2. Adamson, A.W., Physical Chemistry of Surfaces, Third Edition,
John Wiley & Sons, Inc., New York, 1976, p. 51.
3. Gray, V.R., J. Inst. Fuel, 31_, 96 (1958); also cf. Discussion,
Ibid., 32, 72 (1959).
4. Taggart, A.F., Handbook of Mineral Dressing, John Wiley & Sons,
Inc., New York, 1945.
5. Deurbrouck, A.W., U.S. Bureau of Mines Report of Investigation
No. 7633 (1972).
6. Miller, K.V., U.S. Bureau of Mines Report of Investigation No. 7822
(1973).
7. reference (2) p. 464.
8. Wakeman, R.J., Filtration Post-Treatment Processes, Elsevier
Scientific Publishing Co., Amsterdam, 1975.
9. Kozeny, J.S.B., Akad, Wiss, Wein, Math-Naturw., Klasse 136
(Abt. Ila), 271 (1927); Wasserkr. u. Wasserwirtz, 22, 67, 86
(1927).
10. Carman, P.C., Trans. Inst. Chem. Engrs., 15, 150 (1937).
11. Browne!!, I.E., and Katz, D.L., Chem. Engng. Progr., 43, 601
(1947); and ibid., 43, 703 (1947).
12. Dombrowski, H.S., and Browne!!, I.E., Industr. Engng. Chem., 46,
1207 (1954).
13. Silverblatt, C.E., and Dahlstrom, D.A., Industr. Engng. Chem., 46,
1201 (1954).
14. Dolina, L.F., and Kaminskii, V.S., Coke and Chemistry (U.S.S.R.)
(translation of Koks i Khimiya), No. 10, 16 (197!).
15. Nicol, S.K., Aus. I.M. & M. Proceedings, No. 260, 37 (1976).
16. reference (2) p. 339, 345.
17. Bailey, R., and Gray, V.R., J. Appl. Chem., 8_, 197 (1958).
*
18. Bailey, M.E., J. Chem. Education, 51_, 446 (1974).
19. reference (1) Chapter .1.
-------
167
20. reference (2) Chapter V.
21., reference (2) Chapter IV.
22. Campbell, J.A.L., and Sun, S.C., "An Electrokinetic Study of
Bituminous Coal Froth Flotation and Flocculation," Special
Research Report SR - 74, Penn. State Univ. May 23, 1969.
23. Fuerstenau, D. W., and Raghavan, in Flotation. Fuerstenau, M.C. ed.,
AIMME, New York, 1976, p. 21.
24. Baker, A.F., and Miller, K.J., U.S. Bureau of Mines Report of
Investigation No. 7518 (1971).
25. Thomas, Jr., J.,and Damberger, H.H., "Internal Surface Area, Moisture
Content, and Porosity of Illinois Coals: Variations with Coal Rank",
Illinois State Geological Survey Circular 493 (1976).
26. Tanford, C., The Hydrophobic Effect: Formation of Micelles and
Biological Membranes, John Wiley & Sons, Inc., New York, 1973.
27. Griffin, W.C., J. Soc. Cosmetic Chem., 1_, 311 (1949); ibid., 5_,
249 (1954).
28. Davies, J.T., Proc. 2nd Int. Congr. Surface Activity, London,
Vol. I, p. 426.
29. reference (2) p. 506.
30. Stern, 0., "The Theory of the Electrolyte Double Layer", Electrochem.,
30 (1924).
31. Grahame, D.C., Chem. Rev., 41_> 441 (1947).
32. Wakamatsu, T., and Fuerstenau, D.W., Advan. Chem. Series 79, 161
(1968).
33. Garner, W.E., McKie, D., and Knight, B.D.J.G., J. Phys. Chem.,
3T_» 641 (1927).
34. Fowkes, P.M., J. Phys. Chem., 57, 98 (1953).
35. Analysis done by Otisca Industries, Ltd., LaFayette, New York.
36. Chi, Y.M., Master's Thesis in progress, Syracuse Univ., 1978.
37. reference (2) p. 21.
38. Instruction Manual, Fisher Surface Tensiomat Model 21, Catalog No.
14-814, Instrument Division, Fisher Scientific Co.
-------
168
39. Handbook of Chemistry and Physics, 57th Edition, CRC Press,
1976-1977.
40. reference (39) p. F-ll.
41. reference (39) p. F-10.
42. Murkerjee, P., and Mysels, K.G., Critical Micelle Concentrations
of Aqueous Surfactant Systems, NSRDS- NBS 36, 1971, (Nat. Stand.
Ref. Data Ser., Nat. Bur. Stand., U.S.).
43. Vijayendran, B.R., J. Colloid Interface Sci., 6(3, 418 (1977).
44. Miles, G.D., J. Phys. Chem., 49_, 71 (1945).
45. Brady, A.P.,. J. Phys. and Colloid Chem., 53, 56 (1949).
46. Nilsson, G., J. Phys. Chem., 61_, 1135 (1957).
47. Rehfeld, S.J., J. Colloid Interface Sci., 31_, 46 (1969).
48. Rohm and Haas Surfactants and Dispersants, "Handbook of Physical
Properties", Rohm and Haas Co., Philadelphia, Pa., C5-16, July 1976.
49. Cushman, A., Brady, A.P., and McBain, J.W., J. Colloid Sci., 3_,
425 (1948).
50. Weiner, N.D., and Zografi, 6., J. Pharm, Sci., 54_, 436 (1965).
51. Elworthy, P.H., and Mysels, K.J., J. Colloid Sci., 2]_, 331 (1966).
52. Miura, M., and Matsumoto, T., J. Sci. Hiroshima Univ., Ser. A-II,
2]_, 51 (1957).
53. Williams, E.F., Woodberry, N., and Dixon, J.K., J. Colloid Sci.,
12, 452 (1957).
54. Aerosol Surface Active Agents, American Cyanamid Company, Process
Chemicals Department, Wayne, N.O., PRC 102 (2-1227 cM 5/72).
55. reference (2) p. 4.
56. Haines, W.B., J. Agric. Sci., T7, 264 (1927); ibid., 20_, 97 (1930).
57. Dick, S.6., Fuerstenau, D.W., and Healy, T.W., 0. Colloid Interface
Sci., 37_, 595 (1971).
58. reference (2) P- 68.
59. Saleeb, F.Z., and Kitchener, J.A., 0. Chem. Soc. (Lond.) 911 (1965).
60. Salley, D.J., Weith, A.J., Argyle, A.A., and Dixon, J.K., Proc.
Roy. Soc., 20, 519 (1950).
-------
169
61. Tajima, K., Muramatsu, M., and Sasaki, T., Bull. Chem. Soc. Japan,
43, 1991 (1970).
62. Ginn, M.E., in Catiom'c Surfactants Jungermann, E. , ed., Marcel
Dekker, Inc., New York, 1970, p. 364.
63. Rosch, M., in Nom'onic Surfactants, Schick, M.J., ed., Marcel
Dekker, Inc., New York, 1967, p. 753.
64. Anacker, E.W. , in Cationic Surfactants. Jungermann, Ed., ed. ,
Marcel Dekker, Inc., New York, p. 217.
65. reference (26) p. 78.
/
66. Hummel, D. , Identification and Analysis of Surface-Active Agents,
Interscience, New York, 1962, p. 109.
67. Schick, M.J., J. Colloid Sci., 17., 801 (1962).
68. Schick, M.J., J. Am. Oil Chemists' Soc., 40, 680 (1963).
69. Onsager, L., Ann. N.Y. Acad. Sci., 51_, 627 (1949).
70. reference (2) p. 524.
71. reference (2) p. 569.
72. reference (2) p. 552.
73. Elworthy, P.H., Florence, A.T., and MacFarlane, Sol utilization
by surface-active agents, Chapman and Hall Ltd., London, 1968,
p. 52.
74. reference (64) p. 222.
75. grhvp-tTj a M , anri Pprry. .i.w.. ' Surface Active Agents, Interscience
Publishers, Inc., New York, 1949, p. 286.
76. Doss, Kolloid-Z., 86, 205 (1939); ibid., 87, 272 (1939).
77. Ross, J. Am. Chem. Soc., 67, 990 (1945); J. Phys. Chem ,49_, 377
(1945); Dognon and Gougerot, J. Chim, Phys. 40, 127 (1943J;
Addi son, Nature, 156, 600 (1945).
-------
170
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
i. REPORT NO.
FE-9001-1 (EPA-600/7-79-008)
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Surface Phenomena in the Dewatering of Coal
5. REPORT DATE
January 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO,
D. V. Keller, Jr., G. J. Stelma, and Y. M. Chi
DoE FE-9001-1
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Syracuse University
Department of Chemical Engineering and
Materials Science
Syracuse, New York 13210
10. PROGRAM ELEMENT NO.
EHE623A
11. CONTRACT/GRANT NO.
EPA Inter agency Agreement
DXE685AK
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF FLEPORT AND PEI
Final; 6/75 - 11/78
RIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES Project officersi D. A. KJTchgessner (lERL-RTP); andR.E.Hucko
(DoE), Div. of Solid Fuel Mining and Preparation, Pittsburgh PA 15213.
16. ABSTRACT Tne j^p^ gives results of a. study of the influence of certain surfactants
on the dewatering of fine coal. The surfactants were found to have a two-fold effect:
they were found to affect the pressure differentials required for dewatering, in
addition to the residual water contents of the coal beds attainable by this dewatering.
Both effects were attributed to surfactant adsorption. Adsorption at the liquid/air
interface resulted in a decrease in the interfacial tension between the two phases.
The effect this decrease had on the pressure differentials required for dewatering
agreed with that predicted by the capillary theory applied to the system. Adsorption
at the solid/liquid interface was correlated with the complex behavior of the residual
water contents as a function of surfactant addition.
* US. GOVERNMENT PRINTING OFFICE: 1979-«40-082/ 78S
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Coal
Dewatering
Surfactants
Pollution Control
Stationary Sources
13B
08G,21D
13H,07A
UK
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
186
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
$9.00
EPA Form 2220-1 (9-73)
------- |