PB81-20407?
The Equilibrium Fluoride Capacity of
Activated Alumina. Determination of the
Effects of pH and.Competing Ions
Houston Univ., TX
Prepared for
Municipal Environmental Research Lab,
Cincinnati, OH /
Hay 81
I
Department of Commerce
National Technical Information Service
-------
TECHNICAL REPORT DATA
(Pleae nut Inauctiaa on the reverie before completing)
1. REPORT NO.
EPA-600/2-81-082
2.
ORD Report
3. RECIPIENT'S ACCESSION NO.
POT 20A07 5
4. TITLE AND SUBTITLE
Equilibrium Floride Capacity of Activacted Alumina
8. REPORT DATE
May 1981
0. PERFORMING ORGANIZATION CODE
7. AUTHORISE
Gurinderjit Singh
Dennis A. Clifford
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Engineering Program
University of Houston
Houston, Texas 77004
10. PROGRAM ELEMENT NO.
BNC1A
11. CONTRACT/GRANT NO.
CR80607
12. SPONSORING AGENCY NAME AND AOORCSS
Municipal Environmental Research Laboratory - Cin. ,011
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT ANO PERIOD COVERED
F-lnal 7/7g _ T?/Qn
SH
14. SPONSORING AGENCY'COC
EPA/BOO/14
IS. SUPPLEMENTARY NOTES
Project Officer - Thomas J. Sorg
513/684-7370
IS. ABSTRACT
This report describes research on the aetermi-atlon of the equilibrium fluoride
adsorption capacity of small columns of acid pretreatad activated alumina (Alcoa F-l
grade). The experiraent&l observations verified the expectation that fluoride is very
favorably adsorbed in preference tc the common anions: sulfate, chloride and bicarbo-
nate. However, the adsorption capacities were found to be four to five times higher
than what has been reported in the early literature for municipal defluoridation ^
processes.
/
Fluoride adsorption capacity is significantly affected and is decreased with
an increase in pH beyond seven. The alumina selectivity sequence determined by
experiments was the same as has been reported in the early literature,
OF~>SO^=>C1~>HC02~. Although fluoride anions are preferred over sulfate ions, the
sulfate ions compete significantly at the levels found in ground water supplies.
Experiments with high ionic strength ranging up to 56 millimoles per liter (5600 ppm
as CaC03) indicate that the total adsorption capacity increases slightly. Fluoride
adsorption capacity decreases only slightly with the very significant increases in the
concentrations of the other anions. These equilibrium data will prove useful in
utilizing the maximum adsorption capacity of activated alumina in municipal defluorida'
tion processes* • - .
7. .
KEY WORDS ANO DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Water Treatment
Water 'Supply
Adsorption
Fluoride
Activated Alumina
Bench-Scale Treatment
13B
3. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
7/
20 SECURITY CLASS
Unclassified
22. PRICE
EPA form 222O-I (*-73]
-------
EPA 600/2-81-082
May 1981
THE EQUILIBRIUM FLUORIDE CAPACITY OF ACTIVATED ALUMINA
Determination of the Effects of
pH and Competing Ions
by
Gurinderjit Singh
Dennis A. CL ifford
Environmental Engineering Program
The University of Houston
Houston, Texas 77004
"Grant No. R806073010
Project Officer
Thomas J. Sorg
Drinking Water Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report: has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
FOKEWORD
The U.S. Environmental Protection Agency was created because of Increas-
ing public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled land
are tragic testimonies to the deterioration of our natural environment. The
complexity of that environment and the interplay of its components .require a
concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solution;
and it involves defining the problem, measuring its Impact, and searching for
solutions. The Municipal Environmental Research Laboratory develops new and
improved technology and systems to prevent, treat, and manage wastewater and
solid and hazardous waste pollutant discharges from municipal and community
sources, to preserve and treat public drinking water supplies, and to minimize
the adverse economic, social, health, and aesthetic effects of pollution.
This publication is one of the products of that research and provides a most
vital communications link between the researcher and the user community.
Pursuant to the provisions of the Safe Drinking Water Act of 1974 (Public
Law 93-523), many public water supplies in the U.S. may eventually have to be
treated to remove excess fluoride. The research described in this report
establishes the maximum adsorption capacity of activated alumina for the
removal of fluoride ions as influenced by pH and competing ions, viz., sulfate,
chloride, and bicarbonate.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
iii
-------
ABSTRACT
This report describes research on the determination of the equilibrium
fluoride adsorption capacity of small columns of acid-pretreated activated
alumina (Alcoa F-l grade). Mini-columns containing 1 g alumina adsorbent
were equilibrated over a period of 6 to 10 days with continuous flows of
constant pH fluoride solution. The adsorbed ions were then eluted with 1%
sodium hydroxide solution over a period of A hours. The alkaline regener-
ants were analyzed to determine the quantities of ions "adsorbed" (ion
exchanged) during the exhaustion step. As expected, fluoride was very
favorably adsorbed compared to the common anions: sulfate, chloride, and
bicarbonate. .However, the equilibrium fluoride adsorption capacity was
found to be twice as great as that reported in the recent literature for
municipal defluoridation processes.
Fluoride adsorption capacity is significantly affected and is decreased
with an increase in pH beyond 7. The alumina selectivity sequence determined
by experiments was the same as has been reported in the early literature:
OH~>F~>SO^>C1~>HC03. Although fluoride ions are preferred over sulfate ions,
the sulfate ions compete significantly at the levels found in groundwater sup-
plies. Fluoride adsorption capacity decreases only slightly with the very
significant increases (up to 56 millimoles/L) in the concentrations of the
other anions.
The laboratory research on fluoride removal was done in support of field
studies on the removal of fluoride using activated alumina columns in a mobile
pilot plant designed and constructed at the University of Houston. The Mobile
Drinking Water Treatment Research Facility was completed on April 9, 1980,
and transported to its first field location—Taylor, Texas, a small community
(population 13,000) with a high-fluoride (3.0 mg/1) groundwater supply. The
mobile facility containing activated alumina and ion exchange columns and
reverse osmosis and electrodialysis units may be transported to any U.S.
community with i\n inorganic contaminant problem.
This report .Is submitted in partial fulfillment of Grant Number R8 06073-
010 by the University of Houston under the sponsorship of the U.S. Environ-
mental Protection Agency. A second project report covering the design, con-
struction, and operation of the Mobile Drinking Water Treatment Research Facil-
ity will be published at a later date.
iv
-------
CONTENTS
Foreward ill
Abstract iv
Figures . . ...................... vl
Tables vii
Abbreviations and Symbols ix
Acknowledgment x
1. Introduction 1
2. Conclusions 6
3. Theory 7
4. Methods and Materials 12
5. Results 21
Effects of pH 26
Effects of sulfate Ions 26
Chloride effects 29
High ionic strength effects 29
6. Discussion 35
Effluent concentration histories 36
References 44
Appendices .. .................. 47
Mini-column Effluent Concentration Histories 51
Glossary 58
-------
FIGURES
Number Iage
1A Mini-column for capacity determination 14
IB Alumina-conditioning column 14
2 Typical alunina mini-column set-up for
adsorption run . . 15
3 Fluoride adsorption isotherm pH - 5.0 22
4 Fluoride adsorption isotherm pH - 6.0 23
5 Fluoride adsorption Isotherm pH - 7.3 24
6 Fluoride adsorption Isotherm pH - 8.0 25
7 pH effect on fluoride adsciption 27
8 Effects of ch.ioride and sulfate on
fluoride adsorption Isotherms 28
9 Mini-column effluent concentration histories
pH - 6.0, Effect of SO^" cone, on column kinetics ... 30
10 Mini-column effluent concentration histories
pH - 6.0, Effect of CI~ cone, on column kinetics .... 31
11 Mini-column effluent concentration histories,
pH - 7.1, Effect of ionic strength on column kinetics . 32
12 Anlon adsorption Isotherms, pH - 7.1 33
13 An ion adsorption Isotherms, pH - 7.1 34
14 Langmulr isotherm for pH 5, pH 6 38
15 Langmulr isotherm for pH 7, pH 8 39
16 Freundlich isotherm for pH 5, pH 6 40
17 Freundlich isotherm for pH 7, pH 8 41
18 Effect of pH on time for 90% adsorption Equilibrium ... 42
vl
-------
FIGURES
APPENDIX
Number Page
Al Mini-column effluent concentration histories,
pH - 5.0 5i.
A2 Mini-column effluent concentration histories,
pH - 5.0 52
A3 Mini-column effluent concentration histories,
pH - 6.0 53
A4 Mini-column effluent concentration histories,
pH - 6.0 54
A5 Mini-column effluent concentration histories,
pH - 7.0 t . 55
A6 Mini-column effluent concentration histories,
pH - 8.0 56
A7 Mini-column effluent concentration histories,
pH - 7.1 57
vii
-------
TABLES
Number Page
1 Maximum contaminant levels for fluoride 3
2 Full-scale packed bed defluoridation plants 4-5
3 Iso-electrlc point of activated alumina ... 10
4 Typical properties and specifications of activated
alumina, F-l type 13
5 Column data for determination of the effects of pH
on fluoride adsorption capacity 18
6 Column data for determination of the effects of
sulfate and chloride on fluoride adsorption capacity .... 19
7 Column data for determination of the effects of ionic
strength on fluoride adsorption capacity 20
8 Fluoride adsorption isotherm data at 22°C 37
APPENDIX
1A Time spent in hours for throughout T • 1 49
viii
-------
LIST OF ABBREVIATIONS AND SYMBOLS
g
a Separation factor between ions B and A, dimensionless
BV Bed volume
b. Langnuir constant related to adsorption energy, 1/mg
C Liouid chase concentration, mg anion/1
EBCT Empty Bed Contact Time
1 Ionic strength, moles/liter
K. Freundlich constant, mg/g
M Moles
n Freundlich constant, g/1
Q Rate of flow through column, ml/min
Q' Langtnuir ultimate solid phase adsorption capacity constant, tng/g
q Solid phase concentration, mg anions adsorbed/gm Al
r Correlation coefficient
t - Time in hours to reach 90 percent equilibrium of exhausting
mini-column
T Throughput, i.e., ratio of the total amount of fluoride which
comes in contact with the adsorption column during a speci-
fied amount of time to the total adsorptive capacity of the
column
T . Throughput at 90 percent approach to equilibrium
. y
TIC Total 'inorganic Carbon
TOC Total Organic Carbon
ix
-------
ACKNOWLEDGMENTS
Mr. Tom Sorg, Project Officer, is thanked and recognized for his helpful
cooperation throughout the course of this work and the concurrent work on the
design and construction of the Mobile Pilot Plant to be used for field research
on the removal of fluoride and other ions.
Eric Rosenblum is acknowledged for his competent editing efforts on the
report; JoAnn Wardin for her excellent and efficient typing; and Sumeet Singh
for her valuable assistance in drafting all of the figures.
-------
SECTION 1
INTRODUCTION
EFFECTS OF FLUORIDE IN DRINKING WATER
Chronic effects of fluoride on the skeletal system have been documented
in several geographic areas where drinking water contains excessive fluoride.
Tliis form of chronic fluorosis was first described in India and has since been
reported in Ceylon, China, South Africa, Japan, Saudia Arabia, the United
States, Canada, and Europe [V.H.O., 1970].
Mottled tooth enamel is the most delicate index of chronic fluorosis in
humans. It is epidemic in areas where the drinking water contains 2 ppm F~
or more, and there is a precise relationship between the severity of mottling
and the concentration of fluoride in the drinking water [Sognnaes, 1953]. In
a 1953 report on the medical aspects of fluorosis in Bartlett, Texas, where
the drinking water contained 8 ppm of fluoride [Shimkin, et al., 1953], a
little difference was discovered between the "high fluoride" group and the
control group with respect to the number and types of specific disease symp-
toms elicited. However, the high fluoride group manifested a higher incidence
of mottled enamel and an Increased bone density of the spine and pelvis. They
also experienced a certain hrittleness and blotching of the fingernails,
hypertrophic changes in the spine and pelvis, and lenticular opacities in the
eye.
Dental tissues (including the skeleton) accumulate fluoride most rapidly
during formation and mineralization [National Research Council, 1971]. During
tooth formation, the cells of the dental tissues—particularly ameloblasts—
are very sensitive to fluoride. At relatively low doses (e.g., 2 ppm of fluo-
ride in the water) small spots of discoloration may form in the tooth surface.
These spots, or "mottling," vary in color from paper white to dark brown, the
brown stains usually accumulating when exposure to fluoride persists after the
tooth has erupted. At higher doses, the cells may be affected and the tooth
structure severely altered, so that the normally smooth surface becomes corru-
gated.
The effects of fluoride in drinking water on animals are analogous to
those in human beings. The ingestion of excessive amounts of fluoride by
young farm animals for long periods causes striking changes in the teeth
such as brown discoloratlons, pitting, and marked wear of the entire tooth.
The exposure of adult farm animals with completely erupted teeth to excessive
amounts of fluoride for long periods will produce chronic symptions of fluoro-
sis, except that the teeth will remain smooth. Bones and joints of animals
-------
may become enlarged, and extra bone growths or extroses may appear in differ-
ent parts of the skeleton. Some animals may become stiff, and Intermittent
lameness may occur [Greenwood, 1956].
MAXIMUM CONTAMINANT LIMITS FOR FLUORIDE
The U.S. Public Health Service (USPHS) Drinking Water Standards of 1962
set a mandatory limit for fluoride based on the annual average of maximum
daily ambient temperature. The fluoride-temperature relationship has been
established on the premise that the consumption of water increases with tem-
perature; therefore, the maximum allowable fluoride levels should decrease
with increased temperature. The U.S. Environmental Protection Agency (U.S.
EPA) subsequently adopted these standards [Federal Register, 1975] (Table 1)
when promulgating the National Interim Primary Drinking Water Regulation
(NIPDWR) pursuant to the Safe Drinking Water Act (PL 93-523).
In May 1972, the National Institute of Dental Research estimated that
community public water supply systems serving approximately 4.2 million people
exceeded the recommended maximum contaminant levels established for fluoride.
A summary published in January 1980 revealed that in the state of Texas, over
400 community water supplies exceeded their maximum NIPDWR limit of 1.6 mg/1
of fluoride.
FLUORIDE REMOVAL PROCESSES
Several method? of defluoridation have been utilized for municipal and
residential fluoride removal. These methods can be broadly categorized into
two kinds of processes: (1) precipitation methods and (2) packed bed adsorp-
tion methous.
In the first category, chemicals like lime (alone or in combination with
magnesium from dolomite), magnesium sulfate, magnesia, or calcium phosphate
are added to precipitate fluoride or to form hydroxide precipitates onto which
fluoride will adsorb and settle out of solution. Addition of other chemicals
—such as bentonites, fuller's earth, diatomaceous earth, silica gel, bauxite,
sodium silicate, sodium aluminate, and ferric salts—has also been attempted,
but these materials require a very low pH (less than 3.0) to effect fluoride
removal in reasonable quantities.
When both hardness and fluoride removal are desired and the water con-
tains sufficient magnesium, softening is the most feasible process. On the
other hand, fluoride removal with alum has been found to require three to five
times the usual alum dose of 50 ppm normally required for water clarification.
At LaCrosse, Kansas, raw water was treated with 225 ppm of alum to reduce the
fluoride content from 3.6 to 1.5 ppm [Gulp and Stoltenberg, 1958]. Maier
[1953] reported that water treatment at Bartlett would have required 900 ppn
alum to remove 7 ppm fluoride, and further dosage would have raised the sul-
fate content to a very undesirable level. Thus the total costs of alum treat-
ment of groundwater, including sludge disposal, would likely be prohibitive
for most small communities. Furthermore, when fluoride is the only contaminant
-------
TABLE 1. MAXIMUM CONTAMINANT LEVELS (MCLs) FOR FLUORIDE
(From National Interim Primary Drinking Water Standards, 1975)
Average Maximum Daily Temperature
Temperature
°F
53.7 and Below
53.8 to 58.3
58.4 to 63.8
63.9 to 70.6
70.7 to 79.2
79.3 to 90.5
Temperature
°C
12.0 and Below
12.1 to 14.6
14.7 to 17.6
17.7 to 21.4
21.5 to 26.2
26.3 to 32.5
MCL
mg/1
2.4
2.2
2.0
1.8 .
1.6
1.4
whose removal from drinking water is required, neither alum nor lime addition
may prove economical.
When fluoride alone must be removed from drinking water, fluoride adsorp-
tion on activated alumina, bone char or trlcalclum phosphate has historically
been considered the most reasonable, cost-effective removal method [Clifford,
1978]. However, an unanswered question remains as to which media is superior
overall for fluoride removal when capacity, service life, regenerant require-
ments, and media cost are taken into consideration.
Much confusion exists in the literature regarding the fluoride capacities
of the various media. The reason for this confusion car. be attributed to the
fact that the fluoride capacity of an adsorbent depends rather heavily on (1)
the pH of the feedwater and (2) the regeneration history of the adsorbent. In
addition, competing anions and cations and the total ionic strength of the
feedwater are expected to influence the fluoride adsorption or ion exchange
process. In general, one or TPore of these significant parameters have been
overlooked by earlier investigators [Fink and Lindsay, 1936; Swope and Hess,
1937; Savinelli and Black, 1958].
Clifford [1978] has collected and summarized data from municipal defluor-
idation installations dating back to 1937. Thia data are reproduced below as
Table 2.
-------
TABLE 2. FULL-SCALE PACKED-BED DEFLUORIDATION PLANTS
Process,
Location, Date,
(Reference)
Gross
Installed
Capacity
MGD
Process Description and Operation
Tri Calcium Phosphate 0.304
Climax, CO, 1937
(Warnsley & Jones, 1947)
Bone Char 0.30
Briton, SD, 1948
(Maier, 1953)
Activated Alumina 0.58
Bartlett, TX, 1952
*• (Kaier, 1953, 1970)
Bone Char 0.095
Ft. Irwin, CA, 1954
(Harmon, 1965)
Tri Calcium Phosphate 0.144
Apple Valley, CA, 1961
(Harmon, 1965)
Activated Alumina 0.72
Elsinor, CA, 1960
(Harmon, 1965)
Regeneration with NaOH then CO 2 neutralization
Regeneration operation considered complicated and time consuming
F~ removal operates "esceptionally well"
No mention of attrition losses
Bone char replaced original "Fluorex," Ca,(PO,)2, media
Fluorax losses were 42 percent/year
Regeneration with NaOH, then C02 neutralization
In 1960, this was the only F~ plant still using bone char
Plant abondoned in 1971
Regeneration with NaOH then H2S04 neutralization
F~ level - 0.5 - 2.0 ppm in plant effluent
Alkalinity, hardness, pH, S0$~, and F~ all variable in effluent
F~ concentration in raw water determines F~ capacity of alumina
Plant abandoned in 1977
Regeneration with NaOH, neutralization with
Semi-automatic regeneration triggered by manual grab samples
F- removal capacity decreases with each regeneration
Media had to be replaced annually
Regeneration with NaOH, then CC>2 neutralization
Spent regenerant disposal by evaporation pond
Ca3(P04>2 fines in treated water
Initially high attrition losses, later insignificant
Regeneration with NaOH then H2S04 neutralization
"Fortunately" treatment didn't remove H_S
Pretreatnent = lower pH from 9.9 to 7.4
Alumina "cementation" problem encountered
Abandoned after only a few years of operation
Replaced by a low F~ supply
-------
TABLE 2. FULL-SCALE PACKED-BED DEFLUORIDATION PLANTS
(Continued)
Process,
Location, Date
(Reference)
Gross
Installed
Capacity
MGD
Process Description and Operation
Activated Alumina
Gila Bend, AZ, 1978
(Rubel and Woosley, 1979)
0.75 Raw water pH adjusted to pH 5.5
Regeneration done in two steps
Upflow and downflow rinses with 1% NaOH solution followed by
neutralization with 2.5 pH adjusted raw water
Removal capacity observed 2000 grains/ft3
-------
SECTION 2
CONCLUSIONS
From these equilibrium experiments it can be concluded that the adsorption
of fluoride onto activated alumina la a function of pH, and that the optimum
pH for adsorption ranges between.5.0 and 6.0. While early literature reported
maximum fluoride adsorption capacities from 0.95 to 2.3 mg fluorid-2/gm alumina,
at these optltmm conditions, fluoride adsorption capacity reached 10.1 mg
fluoride/gm alumina at 6 ppm fluoride in the water. The pure-fluoride equil-
ibrum capacities found in these experiments are 20 to 50% higher than the best
column capacities reported in the literature. The differences are thought to
be due to kinetic limitations and sulfate competition in actual defluoridatlon
processes utilizing packed beds of activated alumina.
The experimentally determined selectivity sequence was identical to that
reported in the literature, viz, (in order cf preference):
OH~ » F- » S02~ » Cl~ > HCOT
4 J
Fluoride adsorption capacity was not significantly reduced by substantial in-
creases in total dissolved solids concentration, except when sulfate was pres-
ent. Sulfate competition from (from 250 mg/1 So|~) will reduce the equilibrium
fluoride adsorption capacity of activated alumina by as much as 25% when pH *
7.1 and F~ = 5.7 mg/1.
Indications are that the rate of fluoride adsorption on alumina increases as
pH increases especially in the pH range of 6 - 8.
-------
SECTION 3
THEORY
ACTIVATED ALUMINA
"Activated alumina" is the common name for gamma aluminum oxide (y-Al^O.),
a porous adsorbent with a moderately high BCT surface area (150 - 300 m2/gj.
It adsorbs liquids and gases without changing form. I* is manufactured by
low temperature dehydration at 300 - 700°C of hydrous aluminum oxide, and dif-
fers from high temperature alpha aluulna (a-Al203> in that it readily takes up
water and dissolves in acids (both forms are soluble in strong alkalis). Acti-
vated aluminas are used as adsorbents in a variety of processes, principally
to dehydrate organic liquids and gases. Gamma activated alumina was one of
the earliest adsorbents used in inorganic ion chromatography [Kubll, 1947].
It was then known to be quite specific for hydroxide, fluoride, phosphate,
and silicate.
ADSORPTION
Activated alumina is usually grouped with silica, magnesia, and molecular
sieves as a "polar adsorbent." Its crystal structure is that of defect spinel
(MgAl204> and contains cation lattice discontinuities giving rise to localized
areas of negative charge. Since electroneutrallty considerations mandate the
balancing presence of areas of positive charge, sites for both cation and anlon
edsorption are available [Clifford et al., 1978], In an expijiment of nitro-
benzene adsorption on activated alumina, Hesse and Saut&r hypothesized that
both the mechanism of "van der Waals" adsorption and ion exchange adsorption
(i.e., salt transformation through the exchange of one ion for another) occur,
and that nitrobenzene is bound on surface locations which are not affected by
pH change. Kubli (1947] reported that the most important factors affecting
ion exchange adsorption are the nature of the pretreatment and the quantity of
mostly inorganic acids or bases in rhs alumina, while such factors as grain
sire and drying temperature are of minor importance. He reported that the
anion adsorption behavior of alumina treated with the base, Na2C03, was due to
the precipitation of metal carbonates or hydroxides and that anion adsorption
behavior depended on pretreatment with an acid such at HC1 or HC.IO^. He pro-
posed the following mechanism*, for anion exchange on alumina. *
-------
Adsorption-Neutralization of Acid 4X
Al-0^^^
J>A1'X + H,0
Al-0 ---
H + X
Exchange of Preferred Ion Y for X
The above mechanism can take place when the solubility of
Al-0>
Al-0
Al'Y
is lower than
Al-0,
-------
REGENERATION
According to Kubll, the elution of adsorbed ions from th'e column (regen-
eration) can be accomplished by three different means:
1. By alkali elution (reversal of the column)
2. By elution with anlons which form less soluble basic aluminum salts
3. By higher concentration of the anions originally bound to the column
Kubli also expanded the selectivity sequence of anion adsorption on alum-
ina originally established by Schwalb and Dattler [1937]. The sequence of
Important aninns in the expanded series is as follows:
OH~ » P0?~ » F~ » SO?" » Cl~ » NO"
44 3
Clifford et al [1978] have suggested a simplified series of chemical reac-
tions to explain the ion exchange adsorption of fluoride and the subsequent
regeneration of the packed bed of fluoride-exhausted alumina:
SIMPLIFIED PICTURE OF ALUMINA
ADSORPTION AND REGENERATION REACTIONS
1. NEUTRAL ALUMIKA
Alumina + HOH >• Alumina -HOH
2. ACIDIFICATION
Alumina-HOH + HC1 * Alumina-HC1 + HOH
3. ION EXCHANGE IN ACIDIC SOLUTION
Alumina-HC1 -f NaF >• Alumina-HF + NaCl
4. REGENERATION
Alumina «HF + 2NaOK >• Alumina-NaOH + NaF + HOH
5. ACIDIFICATION
Alumina-NaOr. + 2HC1 » Alumina-HC1 + NaCl + HOH
-------
CHARGE DEPENDENCE
Many researchers in the fields of soil and minerology have studied the
presence of electric charges on the oxide minerals suspended in aqueous solu-
tions. These charges comprise an important consideration when determining the
extent to which the solids to which they are attached can function as adsor-
bents in ion exchange columns [Amphlett, 1964; Bolt, 1965; Van Olphen, 1963].
The iscelectric point IEP (s) and the zero point of charge (ZFC) are convenient
references for predicting charge dependence behaviors [Parks, 1967].
Parks defined the ZPC as the pH at which solid surface charge from all
sources is zero. The IEP(s) is a ZPC arising from the interaction of the
solid with H+ and OH~, i.e., the interaction of the solid with water. For
other species, the ZPC will vary with the ionic composition of the system.
Data on pHx£ps and pHgpc for most minerals [Parks, 1965] are available else-
where. Values for the pHtEPs of activated alumina have been determined [Choi,
1979] and are presented in Table 3i aging in DI water raises the pHlEPs.
In general, the sign of the surface charge Itself will determine the
anionic or cationlc adsorption behavior. Thus aluminum oxide may be either
a cation exchanger (in alkaline medium) or an anlon exchanger (in acid medium),
and within a certain pH range, can act as an amphoteric ion exchanger [Zhabrovs,
1961]. While surface charge is by no means the only factor responsible for
electrolyte adsorption, the extent of adsorption decreases rapidly when the
sign of the oxide's surface charge is changed to that of the sorbing species.
However, Umland [1959] pointed out that in cases where the removal of ions
cannot be entirely attributed to either ion exchange or molecular sorption, it
may be caused by the precipitation of sparingly soluble bases and basic salts.
By comparison, adsorption caused by electrostatic attraction alone can
be considered not-specific or non-selective. Adsorption can also arise from
TABLE 3. ISOELECTRIC POINT OF ACTIVATED ALUMINA
Treatment pHIEPs
a. DI water wash, stored in DI
water for two days 6.2
b. DI water wash, stored in DI
water for seven days 5.8
c. DI water wash, 2N-HC1 wash,.
followed by DI wash, stored
in DI watet. for seven days 7.3
d. DI water wash, 2N-HC1 wash,
followed by DI wash, stored
in DI water for ten days 8.9
10
-------
electrostatic attraction augmented by hydrogen bonding, coordinate bonding, or
London Van der Waals bonding. Adsorption under the combined influence of ionic
and non-ionic bonding is called "specific adsorption," which can occur even
when the surface is uncharged.
At ZFC, anlon and cation adsorption is minimal [Parks, 1967]. Specific
adsorption reverses the sign of the surface charge and shifts the ZPC [Kings-
ton, 1972].
11
-------
SECTION 4
METHODS AND MATERIALS
OBJECTIVES
Adsorption capacity of activated alumina,for the fluoride Ion Is reported
to be affected by pH [Fink and Lindsay, 1936], fluoride concentration [Maier,
1953], and the relative concentration of competing Ions, especially sulfate
[Kubll, 1947; Umland, 1956; Hess and Sauter, 1947; Rubel, 1979). Nevertheless,
the literature reveals that most studies of fluoride adsorption by alumina have
been conducted either at uncontrolled pH values or in the presence of competing
buffers, thus complicating interpretation of the results. In this report,
experiments have been carried out specifically to determine, both quantita-
tively and qualitatively, the effect of numerous factors on fluoride concen-
tration, competing ions, Including su.'.fate, chloride, bicarbonate, and high
ionic strength.
PROCEDURES
Conditioning the Alumina
1. A 200-ml sample of F-l grade Alcoa activated alumina of mesh size
26-48, Lot 4630-18, was rinsed with water and >ec&nted several times until
all the suspended and fine particles were removed. (Typical properties of
this product are shown in Table 4.)
2. The washed sample of alumina was trickled down into a 2.1-mm ID
conditioning column (See Fig. IB) containing deionized water, in such a
way that no air bubbles remained or were entrapped in the settled alumina.
3. The column was eluted with tap water (which had been adjusted to pH
3) until the effluent pH dropped down to pH 5. The elutlon was concluded
with tap water whose pH had been adjusted to exactly 5.0 with 0.01 N
until equilibrium was attained at the column effluent.
rv,
4. The conditioned, acid-washed (pH 5) alumina was dried at IIO'C in
an oven for 48 hours.
Exhausting the Alumina with Standard F~ Solutions
1. 1.00 g of the dried conditioned alumina was placed in each of ten
12
-------
TABLE 4. TYPICAL PROPERTIES AND SPECIFICATIONS OF
ACTIVATED ALUMINA, F-l TYPE
Constituents and Properties
Content
A1203
Na20
Fe2°3
S102
Loss on ignition (1100°C)
Form
Contact surface area (sq.m./g)
Size (mesh)
Bulk density loose (lb/ft3)
Bulk density packed (lb/ft3)
Specific gravity
92.00Z
0.90Z
0.08Z
0.09S
6.501
Granular
210
26 - 48
52
55
3.3
11 mm ID standard wall pyrex mini-columns. No air bubbles were entrained in
the settled alumina. See Fig. LA.
2. The mini-columns were equilibrated by passing the standard fluoride
solutions through them at un average flow of 0.5-4.0 ml/nin. The effluent
fluoride concentration was checked from time to time. It took from five to
six days of continuous operation to achieve more than 90 percent equilibrium,
i.e., the effluent F~ concentration ^0.90 times influent F~ concentration.
See Fig. 2 for typical mini-column set up.
Regeneration of Fluoride-Spent Alumina
1. The equilibrated columns were disconnected and the solution in the
columns drained to a point Just above the surface of the alumina.
2. Each column was eluted with 100 ml one percent NaOH (.25 N) during
a period of approximately four hours followed by 500 ml deionized water wash.
3. The regenerant solution was collected and diluted to exactly one liter.
NOTE: It had been previously determined that about 95 percent of the fluoride
on the column could be recovered by this procedure.
13
-------
40
cm
I
j-— 11 mm ID Std. Wall
~|~] Pyrex Tubing
• • •£
. t
,
m urn ft
A
t(
/ 2
15
cm
V
t
6 cm
-V * U ml
"Activated Alumina to be Exhausted
6— Pyrex Wool
<—3/8" OD Tubing
<— Pinch Clamp
— Eye Dropper Tip
A. Mini Column for Capacity r>**formin.ilion
Activated Alumina
to be Conditioned
21 mm ID Standard Wall Pyrex Tubing
Coarse Pyrex Glass
10 mm OD
7/16" OD x 1/16" Wall, Gum Tubing
- Pinching Clamp
10 run OD Tip
B. Alumina Conditioning Column
Figure 1. A. Mini column for capacity o.at«rtnination.
B. Alumina conditioning column.
-------
Plastic Tubing
Tee Connection
For Influent Sample
Mini-Column
Class Wool
2 Liter Jar
for Effluent
5 Cal. Adsorbate
Solution Jar
Column Stand
1 gm Pretreated
Alumina
Figure 2. Typical Alumina Mini-Column Set Up for
Adsorption Run
(Represents one of ten columns used simultaneously)
15
-------
Analysis of Regenerant Solution
2_ 1. The regenerant solution was analyzed to determine the amounts of F~,
804 , Cl~, and HCOj adsorbed on the alumina. The amount of fluoride adsorbed
on the various alumina columns was thus determined as a function of pH, SOJ,
Cl~, and HCO-j concentrations, and ionic strength.
METHODS AND MATERIALS
Feedwaters Used
pH-adjusted Houston tap water was used as a background solution for con-
ditioning the alumina. Deionized water was used to make standard solutions
which were passed through the alumina columns for the fluoride capacity
d<» terminations.
Fluoride Source
Sodium fluoride (reagent grade) was used to make standard fluoride stock
solutions throughout the experiments.
pH Effects
To study pK effects on adsorption, influent fluoride standard solution
was adjusted with diluted ^SO^ (for lowering the pH) or with diluted NaOH
(for raising the pH). In addition, sodium carbonate solution was used as a
buffer while preparing the standard fluoride solutions to maintain constant
pH values of 7 and 8.
Standard Fluoride Concentrations Used
Fluoride concentrations of the standard solutions prepared with deionized
water were diluted to 2.00, 4.00, 6.00, 8.00, and 10.00 mg F~/l.
2— —
Concentrations of Competing Anions: SO, , Cl
To determine effects of competing ions, the fluoride concentration of all
the standard solutions was kept fixed at .3 meq F~/l (5.7 mg F~/l). The con-
centrations of competing Ions used weze 0.50, J.00, 5.00, 10.0, and 15.0 meq/I
in the influent standard solution. The pH was fixed at 6.0 by adding 1 1/2
drops of N/20 NaOH for each two-liter solution.
High Ionic Strength Solutions
The high ionic strength standard solution consisted of Na"*", F~, HCO^, Cl~,
and So|~ ions in deionized water. Some high ionic strength experiments were
16
-------
run at pH - 7.1 and F~ - 0.3 meq/1 (5.7 ppm) and others were run at pH - 7.1
and F~ * 0.5 meq/1 (9.5 ppm). To produce the background ionic strength solu-
tions, equivalent mixtures of Cl~, HCOj, and S0j£~ were utilized to yield solu-
tions of I - 3.8, 8.3, 14.3, 28.3, and 56.3 millimoles/liter.
See Tables 5, 6, and 7 for a complete analysis of the water for each
experimental run.
17
-------
TABLE 5. COLUMN DATA FOR DETERMINATION OF THE EFFECTS OF pH ON FLUORIDE ADSORPTION
CAPACITY IN pH ADJUSTED DEIONIZED WATER SOLUTIONS
oo
Fluoride Concentration Other Anions Present;
Run
Not
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
mg/1
2
2
4
4
6
6
8
8
10
10
2
2
4
4
6
6
8
8
10
10
2
4
6
8
10
2
4
6
8
10
* A "run" IB the
1 meq/1
S0| - 48
meq/1
.105
.105
.210
.210
.316
.316
.421
.421
.526
.526
.105
.105
.210
.210
.316
.316
.421
.421
.526
.526
.105
.210
.316
.421
.526
.105
.210
.316
.421
.526
exhaustion
tng/1 S0$
PH
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
7.3
7.3
7.3
7.3
7.3
8.0
8.0
8.0
8.0
8.0
S0*~
meq/1
.005
.005
.005
.005
.005
.005
.005
.005
.005
.005
.006
.006
.006
.006
.006
.006
.006
.006
.006
.006
2.500
2.438
2.500
2.500
2.250
.625
.625
.500
.687
.812
HC03
meq/1
-
_
_
-
-
_
-
-
-
-
_
-
—
-
-
_
-
-
-
-
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
Q
ml/min
4.18
3.63
2.13
3.09
4.01
2.45
3.54
2.11
3.91
2.67
.68
.93
.44
.77
.46
.39
.41
.36
.39
.76
1.47
1.61
.83
1.44
.54
.86
1.46
1.63
1.39
.83
t.90
hrs
104
87.2
90.5
64.7
24.9
62.6
33.0
51.1
21.3
40.0
244
244
224
137
152
155
131
112
110
53.5
67.7
41.3
52.1
30.1
67.1
53.9
27.2
19.3
26.3
31.9
T.9 m{
6.29
4.47
4.67
4.87
3.31
5.33
5.34
5.00
4.75
5.93
2.44
3.13
2.49
2.58
2.48
2.12
2.60
1.87
2.42
2.33
3.49
3.08
2.50
2.96
2.92
2.28
2.58
2.62
2.65
3.03
Fluoride Capacity
5 F~/g Al
8.27
8.51
9.85
9.85
10.89
10.36
10.49
10.40
10.53
10.80
8.20
8.70
9.63
9.91
10.30
10.30
9.85
10.50
10.70
10.5
3.44
5.20
6.25
7.03
7.54
2.45
3.72
4.35
4.82
5.28
meq F~/g Al
.435
.448
.513
.518
.573
.545
.552
.547
.550
.570
.431
.458
.507
.521
.542
.537
.518
.553
.563
.553
.181
.274
.329
.370
.397
.129
.196
.229
.254
.278
of one single column
- 50 mg/1
as CaC03
1 meq/1
HC03
• 61 mg/1
HC03 - 50
mg/1 as CaC03
-------
TAPLE 6. COLUMN DATA FOR DETERMINATION OF THE EFFECTS OF SULFATE AND CHLORIDE
ON FLUORIDE ADSORPTION CAPACITY AND af AT pH - 6.
Liquid
Phase.Concenfration
Run
No.
15
31
32
33
34
35
36
37
38
39
40
NOTES:
1.
2.
3.
4.
Q
ml/min
.46
.89
.78
.72
1.66
1.57
.88
.99
.59
.84
.62
c.9
hrs
152.0
104.0
96.3
85.4
34.1
36.1
75.7
66.8
111.0
79.1
107.0
T.9
2.48
3.69
2.99
3.85
2.74
3.02
2.21
2.11
2.13
2.19
2.07
S0j£~
meq/1
.006
.5
1.0
5.0
10.0
15.0
-
-
-
-
ci-
meq/1
-
-
-
-
-
.5
1.0
5.0
10.0
15.0
Phase
F-
meq/1
.542
.454
.451
.390
.371
.367
.542
.568
.563
.547
.578
Solid
Concentration
meq/1
.002
.106
.133
.290
.246
.320
ci-
meq/1
0.047
0.003
0.069
0.034
0.073
Total
.544
.560
.584
.680
.617
.687
.589
.571
.632
.581
.651
F
.208
7.14
11.4
22.4
50.2
57.34
19.2
631
136
264
545
A run is the exhaustion of one single column.
For runs 31 - 40, F~ concentration - 5.7 mg/1 (.300 meq/1).
For run 15, F~ concentration - 6.0 mg/1 (0.316 meq/1).
a^ - separation factor of F/C1 or F/SO,. See Appendix for calculation.
-------
TABLE 7. COLUMN DATA FOR DETERMINATION OF THE EFFECTS OF IONIC STRENGTH ON
FLUORIDE ADSOivPTION CAPACITY AT pH - 7.1
Fluoride
Other Anlons Present
Run
No.
*
41
42
43
44
45
46
47
48
49
50
Concentration
ng
F-/1
5.7
5.7
5.7
5.7
5.7
9.5
9.5
9.5
9.5
5.7
meq
F-/1
.3
.3
.3
.3
.3
.5
.5
.5
.5
.5
PH
7.1
7.1
7.1
7.1
7.1
7.1
7.1
7.1
7.1
7.1
SO?'
meq/1
1.0
2.0
4.0
8.0
16. 0
1.0
2.0
4.0
8.0
16.0
Cl"
meq/1
1.0
2.0
4.0
8.0
16.0
1.0
2.0
4.0
8.0
16.0
HCOo
meq/1
1.0
2.0
4.0
8.0
16.0
1.0
2.0
4.0
8.0
16.0
I
mmolea
1
3.8
8.3
14.3
28.3
56.3
4.0
8.5
14.5
28.5
56.5
Q
ml/min
1.22
.78
1.27
.76
1.11
1.21
.75
.72
.90
1.22
T.9
4.38
3.95
3.63
3.48
3.61
3.48
3.71
3.60
3.24
2.93
C.9
75.06
106.69
56.11
87.68
54.22
43. °7
f ,.45
69.21
48.00
29.97
Fluoride
mgF~/g Al
7.16
7.21
6.76
6.55
5.68
8.73
7.94
7.91
7.62
7.13
Capacity
meqF~/g Al
.377
.379
.356
.345
.299
.459
.418
.416
.401
.375
K>
o
* A run Is the exhaustion of one single column
-------
SECTION 5
RESULTS
COLUMN EFFLUENT HISTORIES
Effluent fluoride concentration versus volume of feedwater, I.e., column
effluent histories, were recorded for all the exhaustion experiments and are
shown In Figs. Al - A6. In the first two sets of early experiments (exhaus-
tion of 20 columns), pairs of columns were run In an Identical manner In
order to check the precision of the revolts. The repeatability proved satis-
factory.
The adsorption process for fluoride on activated alumina was observed to
be very slow. It took over 120 hours to reach equilibrium. It has been fur-
ther noted that the adsorption rate Is relatively fast In the beginning and
slowly decelerates to an asymptotic equilibrium. The area above the curves
shows the accumulated fluoride adsorbed onto alumina. The effluent rate was
not kept constant for any of the column pairs; to do so would have Involved
additional monitoring equipment, and the investigation of the kinetics of
adsorption was not the primary intention of these tests. Nevertheless, the
rate of flow was recorded for each column, and is depicted in column data,
Tables 5 to 7.
EFFECT OF FLUORIDE CONCENTRATION
Results indicate that solid phase adsorption capacity is dependent upon
the concentration of adsorbate in solution. Adsorption capacity increases
with adsorbate concentration, while the rate of increase in adsorption capa-
city at lower concentrations is much higher than the rate of Increase at
higher concentrations.
For the first four sets of experiments in which pH had been kept fixed
at 5, 6, 7, or 8, the influent fluoride concentration for each column was
kept at 2, 4, 6, 8, or 10 mg/1. Equilibrium fluoride concentration effects
are revealed in adsorption Isotherms (Figs. 3 to 6). As expected, the
total fluoride adsorption capacity of alumina increased with increasing equi-
librium fluoride concentration. Regarding the rate of adsorption, however,
Figs. Al - A6 illustrate that fluoride uptake slows down as available adsorp-
tion sites are filled. As equilibrium is approached, the number and avail-
ability of adsorption sites decrease thus the adsorption rate decreases.
The adsorption isotherms obtained are of the Langmuir type. Average
21
-------
11
10
e
o
o
CO
I
6
mg
8
10
Liquid Phase Fluoride Concentration
FIGURE 3. FLUORIDE ADSORPTION ISOTHERM
Adsorbate pH * 5.0, temperature - 22°C
22
-------
11
10
o
c <
i &
o •-»
* t?
•H B
O
=
-------
2 < 6
s g,
4) —
o
U
o
3
0)
CD
(0
o.
'6.2
00
2.0
4.0
6.0
C mg F"/l
8.0
10.0
Liquid Phase Fluoride Concentration
FIGURE 5. FLUORIDE ADSORPTION ISOTHERM
Adsorbate pH - 7.3, temperature - 22*C
24
-------
7
6 •
o
u
C
s.
"3
o
en
5 •
O.
E
3 •
6 _
mg F"
8
7.0
Liquid Phase Fluoride Concentration
FIGURE 6. FLUORIDE ADSORPTION ISOTHERM
Adsorbate pH • 8.0, temperature • 22°C
25
-------
fluoride adsorption capacity on activated alumina obtained from these experi-
ments was 10.2 mg/g at pH 5 and pH 6, five times greater than what has been
reported in the literature [Savinelli and Black, 1958] but in general agree-
ment with the results of Choi and Chen [1979] for essentially pure fluoride
solutions.
EFFECTS OF pH
The effects of pH on the adsorption of fluoride by activated alumina are
shown in Fig. 7. For a fixed fluoride concentration, adsorption remains
almost constant within the pH range 5.0 to 6.0 after which the adsorption
falls steeply until pH 7.5, then continues to decline more gradually as pH
is increased.
Experiments could be conducted only between pH 5 and 8. Below pH 5,
alumina dissolves to form aluminum complexes such as A1F2+, A1F£, AlFo, and
A1F£ (pK 6.16, 5.05, 3.91, and 2.71, respectively). Hence, below pH 5, the
column would never reach equilibrium until all the adsorbent dissolved to
form llgand compounds. On the other hand, above pH 8, predominant hydroxyl
ions are preferred over any other ions for adsorption onto alumina beds.
At pH 5 to 6, in the relative absence of hydroxyl and other competing
anions, conditions were most favorable for fluoride to occupy adsorption sites.
Equilibrium adsorption at pH 5 and 6 has been found to be 9.9 and 10.3 mg F~/g
Al, respectively, for adsorbate concentration of 6 mg F~/l.
As the pH was increased above 7, the hydroxyl anions, though present only
in small quantity (10~? moles/1), competed for adsorption sites and consider-
ably reduced the adsorption capacity of alumina at equilibrium by 30 percent.
EFFECTS OF S0*~ IONS
2-
Though sulfate (504 ) ions have been found to be less preferred than
fluoride ions in competition for adsorption sites, the presence of sulfate
had a pronounced effect upon fluoride adsorption, as reported in Fig. 8.
As.the sulfate ion adsorbate concentration increased from .5 to 15 meq/1,
SO/,~ ion adsorption increased from .1 to .3 meq S0j~/g Al, while fluoride
adsorption waa reduced from .46 F~/g Al to .38 meq F~/g Al—a reduction of
about 17 percent. By comparison, It should be noted with respect to the
removal of fluoride from municipal wa£er supplies that where fluoride contam-
inant levels are found in excess, S0£~~ let) concentration generally ranges
around 250 mg/1 or 5.2 n,c?q/l. Extrapolating from the above data, fluoride
adsorption under normal circumstances could be reduced by up to 26 percent.
26
-------
.6 .
.5 -
«
8
01
T O-
31 E
"O -I
Tl ^
c
.2
.1
4.0
5.0
6.0
7.0
8.0
pH
FIGURE 7. pH EFFECT ON FLUORIDE ADSORPTION
27
-------
.5
F~ Adsorption In Presence of Cl
OO
O
a
u
O
09
•O
4)
a
a
x
Ou
•o
O
to
.4
o
or
4)
e
.3
F Adsorption in Presence of SO,
2-
SO ~ Adsorption at 0.3 meq/1 F"
meq/1
10.0
FIGURE 8.
532 mg/1 Cl
2- 720 me/1 SO?"
Liquid Phase Concentration SO, or Cl I
EFFECTS OF CHLORIDE AND SULFATE ON FLUORIDE ADSORPTION ISOTHERMS
Adsorbate pH - 6.0, F~ Cone. - .3 meq/1 (5.7 rag F /I)
-------
CHLORIDE EFFECTS
The effects of the presence of chlorides (.5 meq/1 to 15 meq/1) on fluo-
ride adsorption was also tested, and the results are depicted in Fdg. 8. The
adsorbate acidity was fixed at pH 6 and fluoride concentration was kept con-
stant at .3 meq/1 throughout the experiment. As expected, no significant
change in fluoride adsorption was observed, demonstrating that the fluoride
ion is far more preferred for adsorption by alumina than chloride ions. This
result was predicted by Umland, whose preferential series presumed chloride
to be much less preferred than fluoride which in turn was only slightly less
preferred than hydroxide and phosphate for anion adsorption.
A comparison of Figs. 9 and 10 reveals that, whereas variation in SO^
concentration affected the fluoride adsorption kinetics to a noticeable extent,
variation iu chloride concentration did not. This is contrary to the results
of Choi and Chen [1979].
Referring to Table 6, it should be noted that although chloride adsorp-
tion remained relatively constant with increasing chloride concentration,
the amount of fluoride adsorbed slightly increased so that the sum of anionic
species adsorbed onto alumina increased with increasing anionic concentration.
Furthermore, the total anionic adsorption of the sulfate plus fluoride solu-
tion was almost the same as that of chloride plus fluoride solution, suggest-
ing that increasing the concentration of a much less preferred ion might
increase the adsorption of the more preferred ion. However, this effect
needs to be researched in greater detail.
HIGH IONIC STRENGTH EFFECTS
Fluoride contaminated groundwater found in nature tends to contain levels
of dissolved solids, consisting of a variety of cations and anions. The
anions are mostly sulfate, chloride, bicarbonates (or carbonates), and nitrate.
Another set of experiments was conducted to determine the effect of high
ionic strength adsorbate solution on fluoride removal by activated alumina
adsorption. Adsorbate solutions were prepared in deionized water with equiva-
lents of sulfate, chloride, and bicarbonate anions. The sum of chloride,
sulfate, and bicarbonate concentration (I Cl~ + 50^ + HCOp was kept at 3,
. 6, 12, 24, or 48 meq/1. The ionic strengths of these solutions were 4.0,
8.5, 14.5^ 22.5, or 56.5 mmoles/1.
The effect of ionic strength on fluoride adsorption is reported in Figs.
11, 12&13. Fluoride adsorption decreased with increasing ionic strength. An
analysis of species adsorbed onto the alumina indicated that the reduction in
fluoride adsorption was caused by the S0£~ Jon species, while chloride and
bicarbonate had an insignificant effect.
29
-------
6.0
4.5
'if
o
§
o
3.0
w
1.5
0.0
3 raeq/1 F + .5 meq/1
S0?~ In Feed, t 9 -
104 hrs
3 meq/1 F + 1 meq/1
SO?" In Feed, t
96 hrs
3 meq/1 F~ + 5 meq/1 SO
In Feed, t
3 meq/1 F -I- 10 meq/1 SOT in Feed,
1500 3000 4500 6000
BED VOLUMES (BV) OF EFFLUENT
7500
9000
FIGURE 9. MINI-COLUMN EFFLUENT CONCENTRATION HISTORIES, pH - 6.0
EFFECT OF S0^~ CONCENTRATION ON COLUMN KINETICS
Q - 1.1 ml/mln, EBCT • .88 mln, BV - 1.0 ml
t _ » time to 90% of equilibrium in hours
-------
~ A.5
60
e
u
§
u
3.0
w
1.5
0.0
90 hrs
Feed Solutions Contain
.3 meq/1 F- -I- Cl~
O .5 meq/1 Cl~
Q 1 meq/1 Cl~
& 5 meq/1 Cl~
G| 10 meq/1 Cl"
V 15 meq/1 Cl"
1500
3000
4500
6000
7500
BED VOLUMES (BV) OF EFFLUENT
9000
FIGURE 10. MINI-COLUMN EFFLUENT CONCENTRATION HISTORIES, pH - 6.0
EFFECT OF Cl~ CONCENTRATION ON COLUMN KINETICS
Q -0.78ml/mln, EBCT - 1.2 mln, BV - 1.0 ml
t _ - time to 90X of equilibrium in hourK
-------
6.0
U)
.9
- 106 hra
(See Fig. A 7 for 9.5 ppm F )
5.7 ppm f Feed Solutions Contain
O E HC03 + SOT + Cl - 3 meq/1, I - .0038 M
HCO~ + SOT + Cl~ - 6 meq/1, I - .0083 M
0 E HCO~ -f S0,~ + Cl~ - 12 meq/1, I - .0143 M
Ki E HCOj + SO|~ -»- Cl - 24 meq/1, I - .0283 M
. 0E HC03 f sofr* + Cl",- 48 meq/1.. I - .0563 M
1500
3000 4500 6000
BEN VOLUMES (3V) 0V EFFLUENTS
7500
9000
FIGURE 11. MINI-COLUMN KFFLUiZNT roNCENTRATION HISTORIES, pH
EFFECT OF IONIC STRENGTH ON COLUMN KINETICS
Q - 1.0 ml/min, EBCT - i min, BV - 1.0 ml
t _ • time to 90% of equilibrium in hours
See Fig. A 7 for 9.5 ppm F*
7.1
-------
tit
o
M
O
O
v.
-------
o
+
'V1
+
o er
o.
u
o
0)
n)
Ou
O
to
.7
.6
.5
.4
0
2.0
100
4.0
20G
6.0
300
8.0
400
10.0
500
12.0
600
14.0
700
mgCaC03/l
.2-
HCO.
Liquid Phase Cone, of SO ~, Cl~, n^v/
FIGURE 13. ANIONS ADSORPTION ISOTHERM
Adsorbate pH - 7.1, F cone. » .3, .5 meq/1
-------
SECTION 6
DISCUSSION
The exact phenomenon by which fluoride is removed from water by acti-
vated alumina has yet to be conclusively described. Many authors [Savinelli
and Black, 1958] regarded fluoride adsorption as an ion exchange phenomenon
while others [Kingston, 1972; Parks, 1967; Muljadi, 1966] considered it typi-
cal of adsorption behavior by metal oxides, an aqueous surface phenomenon.
However, the information derived from these experiments—concerning the depen-
dence of fluoride adsorption upon initial fluoride concentration, pH, com-
peting anions, and ionic strength—may provide a new foundation for future
studies of the kinetics of activated alumina adsorption.
ACTIVATED ALUMINA F~ ADSORPTION CAPACITY
In general, the literature (most notably, Churchill, 1936; Fink and
Lindsay, 1936; Swope and Hess, 1937; Zabban, 1967; Maier, 1953) reports much
lower capacities of F~ adsorption on activated alumina than were here obtained.
Savinelli and Black, in their bench scale experiments, showed a capacity
of 8.8 mg/g Al (3400 grains/cu ft*) at pH 5.6 for Ce - 10 ppm F~. In recent
articles, Yeun C. Wu [1978], and Rubel and Uoosley [1979] reportedly attained
an optimum adsorption capacity for fluoride on activated alumina of 2620
grains/cu ft Al (6.8 mg/g), and 3200 grain/cu ft Al (8.3 mg/g) respectively.
In Rubel and Woosley's pilot scale column experiments, the F~ concentration
was 6.0 mg/1 and the pH was controlled between 5 and 6. The experiments
reported here indicate that, in the absence of competing ions and when
acidity is strictly controlled between pH 5.0 and pH 6.0, adsorption capacity
can reach as high as 10.1 gm/g Al (3894 grain/ft-*) for an influent F~ concen-
tration of 6 ppm (a high but not uncommon concentration for contaminated under-
ground municipal water supplies). Tne significantly lower values for the
maximum adsorption capacity reported in the early literature may be attributed
to a number of factors, such as uncontrolled pH, equilibria obtained from batch
reactors, or studies conducted In continuous flow columns which were termin-
ated for practical considerations at some arbitrary fluoride concentration
breakthrough and never allowed to reach equilibrium. It is not being sug-
gested here that in field operation, one should always expect these maximum
capacities to be achieved. The equilibrium capacities are offered as a goal
for ideally efficient processes.
*Note: Above conversions based on packed alumina density of 55 Ib/ft .
35
-------
'*FLUORIDE ADSORPTION ISOTHERMS
The shapes of the isotherms describing fluoride adsorption equilibria
were largely determined by such factors ar? the relative affinity of the solute
for the solid surface; the number of sites available for adsorption; and
the interactions of the adsorbed molecules. The F~ adsorption isotherms
obtained are typical of weak base synthetic resins. They satisfy the Lang-
Inulr mathematical model well compared to the Freundlich model; the average
correlation coefficients r's are .994 and .971, respectively. Other constants
and isotherms are reported in Table 8, and Figs. 14 through 17. In his explan-
ation of the selectivity coefficient for ion exchange, Clifford [1978] reiter-
ated that the Donan membrane equilibrium theory, the law of mass action, and
the Langmuir isotherm assumptions all yield the same equilibria expressions.
For binary isotherms, the Langmuir multicomponent equilibria theory yields a
separation factor constant, a, such that
b,C,
Q'bBcB
where Q1 - Langmuir ultimate solid phase adsorption (or Ion exchange) capacity
b. • Langmuir constant related to adsorption energy
Taking OH~ and F~ Ions as the two components, it may be noted that the rela-
tive affinity of alumina for OH~ ions is so disproportionately large that the
separation factor approaches infinity. As a result, initial acid condition-
ing and regeneration of the alumina after exhaustion are essential.
EFFLUENT CONCENTRATION HISTORIES (Appendix Figs. Al - A6)
Further analysis of effluent concentration histories obtained for each
mini-column sheds light on several important kinetic relationships.
Considering throughput at 90 percent equilibrium (T.g - meq F~ ions
passed/meq F~ capacity on Al) and t.g time for 90 percent exhaustion (both
of which are flow rate dependent for a given pH and ionic strength—Tables 5
through 7), t.g is inversely proportional to the influent F~ concentration
(Fig. 18) while T.g throughput is Directly proportional to the rate of load-
Ing. Since a slower flow rate maximizes adsorption, while the practial
limitations of treatment plants (such as reactor size and effluent demand)
press for a faster flow, a compromise must be accomplished based upon the
rate of adsorption, which requires a detailed knowledge of kinetics. Ne\ar-
theless, these experimental results do suggest that the kinetics of fluoride
36
-------
TABLE 8. FLUORIDE ADSORPTION ISOTHERM DATA AT 22°C
pH
5
6
7
8
Ce
mgF-/l
2
4
6
8
10
2
4
6
8
10
2
4
6
8
10
2
4
6
8
10
! Correlation K n Correlation
mg/g 1/mg r mg/g g/1 r
11.62 1.320 .9867 7.778 6.683 .543
11.24 1.539 .9899 7.873 7.475 .958
10.71 0.236 .9999 2.527 2.037 .994
7.41 0.239 .9995 1.832 2.122 .992
-------
.12
.11
00
B
41 .10
.09
pH - 5.0
1/Q' - .086 g/mg*
1/bjQ* = .065 g/1
pH - 6.0
1/Q' - .089 g/mg
1/biQ1 - .058 g/1
.08
.1
.2
1/C
.3
1/mg
.5
FIGURE 14. LANGMUIR ISOTHERM FOR pH 5, pH 6
-------
CO
VO
.3
60
O
.2
pH - 8.0
i/Q
1/b
.1
a
pH - 7.0
1/Q' - .093 g/mg
1/b Q' - .39 g/1
.2
.3
1/C 1/mg
.4
.5
FIGURE 15. LANGMUIR ISOTHERM FOR pH 7, pH 8
-------
1.0
.8
eo
00
e
.6
00
o
•-I
.A
pH - 6.0
log K - .986 mg/g
1/n = .134 1/g
891 mg/g
1/n - .149 1/g
.3
.4
.5
.8
.6 .7
Log Ce mg/1
FIGURE 16. FREUNDLICH ISOTHERM FOR pH 5, pH 6
.9
1.0
-------
1.0
.8
oo
£
.6
or
00
.4
pH - 7.0
log X =• .403 mg/g
1/n - .491 1/g
pH - 8.0
log K - .263 mg/g
1/n - .47 1/g
.21—V
.5
.7
.8
.6
log C mg/1
FIGURE 17. FREUNDLICH ISOTHERM FOR pH 7, pH 8
.9
1.0
-------
O O 2
ppm
4 ppm
6 ppm
10 ppm
Cone, of F~ Feed
FIGURE 18. EFFECT OF pH ON TIME FOR 90* ADSORPTION EQUILIBRIUM
T • Throughput for given data point
T « number shown adjacent to each point
A2
-------
adsorption on alumina (26 - 48 mesh particles) are very slow—48 hours to reach
50 percent exhaustion of the column and 120 hours to 90 percent exhaustion.
Also, in the pH range of 6 - 8, rate of F adsorption increases, as pH increases.
COMPETING IONS
In concentrations at which they are normally found in drinking water sup-
plies, none of th« anlons tested (S0$~, Cl~, HCOj) except for OH~ significantly
decreased the fluoride adsorption capacity of activated alumina. However, at
higher ionic strengths,-fluoride adsorption capacity was slightly affected,
primarily because of SO^ competing ions, indicating that adsorption onto
alumina is "fluoride specific." Tills conclusion is in agreement with the
reported preferential sequence of adsorption onto alumina:
OH" » F~ » S0*~ » Cl" » HCO~
The exact values for separation factors between any two anlons in the series
may be derived after tests for matched pairs of radicals determine the Lang-
muir mathematical model parameters, i.e., after the binary adsorption isotherms
are established.
-------
REFERENCES
1. Activated and Catalytic Alumina, Alcoa Product Data Chemicals, Section
GB-2A, Aluainum Company of America, July 14, 1969.
2. Amphlett, C.B. Inorganic Ion exchangers (New York, Elsevier Publishing
Company, 1964), p. 90, Fig. 25.
3. Bolt, G.H. and Page, A.L. "Ion Exchange Equations Based on Double Layer
Theory," Soil Science, 99, 357 (1965).
4. Choi, Won-Wook and Chen, Kenneth Y., "The Removal of Fluoride from Waters
by Adsorption," Journal AtfHA, Vol. 71, No. 10, p. 562, October, 1979.
5. Churchill, II.F., U.S. Patent 2,059,552, November 3, 1936.
6. Clifford, Dennis A., "Mobile Pilot-Scale Evaluation of Reverse Osmosis,
Ion-Exchange, and Activated Alumina Adsorption for the Treatment of Small
Community Water Supplies," A research proposal submitted to US EPA,
March, 1978.
7. Clifford, D.A.; Matson, J.V.; Kennedy, Ralph, "Activated Alumina: Redis-
covered 'Adsorbent1 for Fluoride, Humic Acid, and Silica," Industrial
Water Engineering, p. 6, December, 1978.
8. Clifford, D.A., "Nitrate Removal from Water Supplies by Ion Exchange,"
US EPA-600/2-78-952, June. 1978, p. 34-37.
9. Culp, R.L. and Stoltenberg, H.A., "Fluoride Reduction at LaCrosse, Kansas,'
J. AWWA, Vol. 50, No. 3, p. 423. March, 1958.
10. Fink, G.J. and Lindsay, F.K., "Activated Alumina for Removing Fluorides
from Drinking Water," Ind. and Eng. Chem., Vol. 28, No. 8, p. 947 (1936).
11. Greenwood, D.A., "Some Effects of inorganic Fluoride on Plants, Animals,
and Man," Fifteenth Annual Faculty Research Lecture, The Faculty Associa-
tion, Utah State Agricultural College, Logan, Utah, 1956, pp. 12-23.
12. Hesse, G. and Sauter, 0., "On the Independence of Exchange Adsorption and
Van der Waals Adsorption of Aluminum Oxide," Naturwiss., Vol. 34, p. 250
(1947).
44
-------
13. Hingston, F.J.; Posner, A.M. ; and Quirk, J.P., "Anion Adsorption by Geo-
thite and Cilinite: I. The Role of r.he Portion in Determining Adsorp-
tion Envelopes," J. Soil Science, Vol. 23, No. 2, p. 177, June, 1972.
14. Rubli, H.A., "Contribution to the Knowledge of Anion Separation by Means
of Adsorption on Alumina," Helv. Chem. Acta, Switzerland, 3:453 (1947).
15. Maier, F.J., "Defluoridation of Municipal Water Supplies," J. AWWA, Vol.
45, August, 1953, p. 879.
16. Muljadi, D.; Posner, A.M.: and Quirk, J.P., "The Mechanism of Phosphate
Adsorption by Kaolinite, Gibbsite, and Psuedoboahmite I," J. Soil
Science, Vol. 17, p. 212, September (1966).
17. Committee on Biological Effects of Atmospheric Pollutants, Division of
Medical Sciences National Research Council, Biological Effects of Atmos-
pheric Pollutants: Fluorides, National Academy of Sciences, Washington,
D.C., 1971, pp. 209-214.
18. Parks, G.A. , "The Isoelectric Points of Solid Oxides, Solid Hydroxides,
and Aqueous Hydroxo Complex Systems," Chem. Rev., 65:117 (1965).
19. Rubel, F. , Jr. and Woosley, R.D., "The Removal of Excess Fluoride from
Drinking Water by Activated Alumina," J. AWWA, 71:1:45 (January, 1979).
20. Savinelli, E.A. and Black, A.P., "Defluoridation of Water with Activated
Alumina," J. AWWA 50:1:33 (January, 1958).
21. Schwalb, G.M. and Dattler, G. , "Inorganic Chroma tography," Agnew. Cheraie,
Vol. 50, No. 33, p. 691 (August, 1937).
22. Shimkin, M.B.; Arnold, F.A. , Jr.; Hawkins, J.W. ; and Dean, H.T., "Medical
Aspects of Fluorosis," American Association Advancement Sciences, 1953.
23. Sognnaes, R.F., "The Problem of Providing Optimum Fluoride Intake for
Prevention of Dental Caries," A Report of the Committee on Dental Health,
Publication 294, November, 1953, p. 4.
24. Swope, H.G. and Hess, R.H., "Removal of Fluorides from Natural Waters
by Deflurite," Ind. and Eng. Chem., Vol. 29, No. 4, p. 424 (1937).
25. Texas Department of Health, "Community Water Systems in Texas Which Exceed
the MCL for Fluroides as Set by the National Interim Primary Drinking
Water Regulations, 18 pages, Texas Department of Health, Austin, Texas,
March, 1977.
26. Umland, F. , "The Interaction of Electrolyte Solutions and a-Al203: IV:
Development of a Formal Ion Exchange Theory for the Adsorption of
Electrolytes from Awueous Solutions," Z. Electrochem. , V. 60, p. 711
(1956).
-------
27. Federal Register, U.S. EPA, "National Interim Primary Drinking Water
Regulations," Washington, D.C., December 24, 1975.
28. U.S. Public Health Service, Drinking Water Standards, 1962.
29. Van Olphen, H., An Introduction to Clay Colloid Chemistry, Interscience,
New York, 1963.
30. World Health Organization, "Toxic Effects of Larger Doses of Fluoride,"
Fluorides and Human Health, WHO Monograph Series No. 59, Geneva: WHO,
1970, pp. 238-239.
31. Zabban, Walter and Jevett, N.W., "The Treatment of Fluoride Wastes,"
Proc. 22nd Ann. Purdue Industrial Waste Conference, Engrs. Bull. No. 129
(1967).
32. Parks, G.A., "Aqueous Surface Chemistry of Oxides and Complex Oxide
Minerals. Equilibrium Concepts in Natural Water Systems," Advances in
Chemistry Series, 67, Robert C. Gould (Editor), American Chemical Soceity
Publications, Washington, D.C., 1967.
33. Wu, Yeun C., "Activated Alumina Removes Fluoride I0ns from Water," Water
and Sewage Works, Vol. 125, No. 6, p. 76, June, 1978.
34. Zhabrova, G.M. and Eorov, E.V., "Sorption and Ion Exchange on Amphoteric
Oxides and Hydroxides," Russian Chemical Reviews, Vcl. 30, No. 6, p. 338,
1961.
-------
APPENDICES
MODIFIED TURBIDIMETRIC METHOD FOR SULFATE DETERMINATION
Reference: Standard Methods for the Examination of Water and Waatevater,
14th Edition, p. 344.
1. Place 100 ml sample in 300 ml Erlenmeyer flask.
2. Add 5 ml conditioning reagent.
3. Add one "scoop" (0.2 - 0.3 ml) of reagent-grade barium chloride.
4. Shake by hand, swirling occasionally, for one minute.
5. Allow 4 additional minutes for turbidity to develop with no additional
agitation.
6. Set 10 ppm sulfate to read "40" on 0 - 100 scale of Hach turbidimeter
(Model 2100A) using 25 ml sample and no spacer in the reading chamber.
7. Read turbidity of all standards and samples after exactly four minutes
of turbidity development following initial one minute agitation period.
8. Plot NTU versus ppm sulfate and read off samples.
Notes: a. Linear range is 2 - 10 ppm sulfate
b. Standard typically, 2, 4, 6, 8, and 10 ppm sulfate
c. See Standard Methods for Preparation of reagents and
standards.
FLUORIDE DETERMINATION PROCEDURE WITH ELECTRODES USING 901 IONALYZER
Reference: Orion Analyzer Instruction Manual, Fluoride Electrode Model
96-09.
1. Prepare 10 ppm and 2 ppm standard fluoride solutions from the 100
ppm stock solution.
2. Add 25 ml TISAB II or IV to each 25 ml of above standard and samples
to be tested.
3. Set the standard value switch to 10. Set the slope switch to the
slope of the electrode, that is, 56.0 mV/decade.
4. Place electrode in 50 ml of 10 ppm standard solution. Press Clear/
Read mV and wait for a stable reading. Press Set Concentration.
Check the slope by placing electrode in 2 ppm standard solution.
5. Place the electrode in 50 ml of sample solution, wait for a stable
reading, and record sample concentration. Similarly repeat step for
other samples.
Notes: a. For every change of standard or sample solution, electrode
• should be rinsed and fillter dried before placing it in
the solution.
47
-------
b. TISAB II and TISAB iV ionic strength adjusters were pre-
pared as per Orion lonalyzer Instruction Manual, Fluoride
Electrodes Model 94-09, Model 96-09.
c. TISAB IV was used only for testing NaOH-eluted samples.
because they contained high concentrations of Al.
POTENTIOMETRIC TITRATION METHOD FOR CHLORIDE
Reference: Clifford, Dennis A., "Nitrate Removal from Water Supplies by
Ion Exchange," EPA 600/2-78-052, June, 1978, p. 250.
1. Make sample to be titrated up to approximately 50 ml In ft 150 ml
beaker with teflon-coated magnetic stirring bar.
2. Titrate with .0141 N AgN03 (. nO meq Cl/ml) to +275 mV end point.
This was previously determined to coincide with the inflection point
in the ml tltrant adder versus mV plot. Potential due to increase
in Ag+ ion was measured using double junction (nitrate-external)
calomel, reference electrode (Orion 90-02-00) with Ag/AgS solid
state specific Ion electrode (Orion 94-10A).
Notes: a. Sensitivity is 125 mV/ml titrant added at inflection point
for .0141 N AgN03.
b. AgN03 standardized against 1000 ppm NaCl.
BICARBONATE DETERMINATION WITH TOC ANALYZER
Reference: Beckman Model 95 TOC Analyzer Instruction Manula, for Instru-
ment operations.
1. Prepare 10, 20, 40, and 100 ppm NaHC03 standards from freshly pre-
pared NaHCC>3 stock solution.
2. Rinse syringe at least twice by discharging contents elsewhere (not
back into beaker). Draw sample up slowly.
3. Inject into inroganic channel of TOC analyzer and set recordings to
standardize.
4. Make at least duplicate Injections or repeat until reproducible peaks
are obtained. .
5. Repeat Steps 2, 3, and 4 for all unknown samples and from peak heights
calculate HCO^ in samples.
TYPICAL CALCULATIONS
For Run No. 34:
F
Is
xs
t.3711110]
[.300][.246]
48
-------
TABLE Al
TIME SPENT IN HOURS FOR
THROUGHPUT T » 1
Run No.
1
M
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
35
19
20
21
22
23
24
25
Hours
16.49
19.54
19.26
13.28
7.54
11.75
6.17
10.26
4.49
6.74
100.49
77.96
82.39
53.63
62.19
73.37
50.05
60.76
45.73
23.03
19.50
13.48
20.02
10.17
23,27
Run No.
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
A3
44
45
46
47
48
I*
50
Hours
23.74
10.62
7.41
7.22
10.60
28.34
32.12
20.09
12.42
12.98
34.22
31.87
53.01
26.17
51.79
17.16
26.99
15.57
25.20
14.96
12.64
18.57
19.26
14.85
10.24
-------
TABLE A2 CONVERSION FACTORS
1.00 milliequivalent fluoride - 19.00 mg fluoride
1.00 gram - 15.43 grains
1.00 gram alumina - 1.067 ml alumina (See Note 2)
(1.134) 3
•1.00 gram alumina - f8'888n2r$s}ft alumlna
1.00 mg F~7g alumina - 409.5 g F~/ft alumina
(385.3) _ 3
1.00 meq F /g alumina - 77ft? grains F /ft axumlna
(7320) _ 3
1.00 meq F/ml alumina - 8303 grains F /ft alumina
Notes: (1) Alumina is dilute-H2SO^-treated Alcoa F-l grade
activated alumina, 26x48 mesh, conditioner ac-
cording to the procedure given on page 12.
(2) Alumina bulk densities are based on
treated, loosely packed, conditioned alumina whose
weight was determined after drying for 48 hrs at
110°C. The experimental bulk density of 1.00 g/
1.067 ml corresponds to 58.5 lbs/ft3, i.e., 6.4%
higher than the 55 Ibs/ft3 (manufacturer's data)
reported for packed alumina in Table 4, p. 13.
The values in parentheses are based on 55 Ibs/ft3.
50
-------
10.00 ppm F in Feed
t Q - 21 hrs
• 7
o—"—o—o-
6.0 ppm P in Feed
t - 24 hrs
^ _ 4.0 ppm F in Feed
«—- t . - 90 hrs
* J
2.0 ppm F in Feed
5,000 10,000 15,000 20,000
BED VOLUMES (BV) OF EFFLUENT (ml)
25,000 30,000
FIGURE Al MINI-COLUMN EFFLUENT CONCENTRATION HISTORIES, pH «• 5.0
EFFECT OF CONCENTRATION ON COLUMN KINETICS
Q - 3.55 ml/min, EBCT - ,28 rain
BV - 1.0 ml, t o - time to 90% of equilibrium in hours
-------
in
ppm F in Influent
t _ - 64 hrs
ppm F i.i Infli-ent
- 87 hrs
5.000 10,000 15.000 20,000
BED VOLUMK (BV) OF EFFLUENTS (ml)
25,000 30,000
FIGURE A2 MINI-COLUMN EFFLUENT CONCENTRATION HISTORIES, pll = 5.0
EFFECT OF CONCENTRATION ON COLUMN KINETICS
Q - 2.79 ml/min, EBCT = .35 mln, BV - 1.0 ml
t _ = time to 90% of equilibrium in hours
-------
10.0 ppm F in Feed
8.0 ppm F in Feed
2,500 5,000 7,500 10,000 12,500 15,000
BED VOLUME (BV) OF EFFLUENT
FIGURE A3 MINI-COLUMN EFFLUENT CONCENTRATION HISTORIES, pH - 6.0
EFFECT OF CONCENTRATION ON COLUMN KINETICS
Q - 0.5 ral/min, ETCT - 2 min, BV - 1.0 ml
t „ = time to 90% of equilibrium in hours
*
-------
10.0 ppm F In Feed
8.0 ppm F in Feed
t = 112 hrs
6.0 ppm F in Feed
t = 155 hrs
"0 ppm F in Feed
9 « 137 hrs
„ _ . „ , t «= hrs
2.0 ppm F in Feed .9
2,000 4,000 6,000 8,000
BED VOLUMES (BV) OF EFFLUENT
10,000 12,000
FIGURE A4 MINI-COLUMN EFFLUENT CONCENTRATION HISTORIES, pll - 6.0
EFFECT UF CONCENTRATION ON COLUMN KINETICS
Q - 0.5 ml/rain, EBCT - 2 min, BV - 1.0 ml
t 0 • time to 90% of equilibrium in hours
• V
-------
10
9
8
oo
e
u
I 5
'- 4
u
§ '
h.
W
2
1
0
10.0 ppm F in Feed
t . = 67 hrs
8.0 ppm F in Feed
6.0 ppm F In Feed
t - 52 hrs
4.0 ppm F in Feed
t « 41 hrs
.9
2.0 ppm F in Feed
67 hrs
2,000 4,000 6,000 8,000 10,000
BED VOLUMi7.S (BV) OF EFFLUENT
12,000
FIGURE A5 MINI-COLUtfN EFFLUENT CONCENTRATION HISTORIES, pH - 7.0
EFFECT OF CONCENTRATION ON COLUMN KINETICS
Q - 1.78 ml/min, EBCT - .84 min, BV - 1.0 ral
t Q ~ time to 90% of equilibrium in hours
. 7
-------
1.1
10.0 ppm F In Fe«d
t Q - 31 hrs
» *
8.0 ppsi F in Feed
t - 26 hrs
6.0 ppm F in Feed
t = 19 hrs
• y
4.0 ppm F in Feed
2.0 ppm F in Feed
2,000 4,000 6.00C 8.000
BED VOLUMES (BV) OF EFFLUENT
10.000
12,000
FICURE A6 MINI-COLUMN EFFLUENT CONCENTRATION HISTORIES, pH - 8.0
EFFECT OF CONCENTRATION ON COLUMN KINETICS
Q - 1.23 ml/rain, EBCT - .81 irdn. BV - 1.0 ml
t Q *• time to 90% of equilibrium in hours
• 7
-------
10.0
In
7.5
60
B
O
25
8
5.0
u.
2.5
0.0
t.g = 68 bra
t 9 = 69 hrs
t ~ = 48 hra
9.5 ppm F Feed Solutions Contain
HCO~ + S0*~ + Cl~ - 3 meq/1, I
.0040 M
.0085 M
2-
2-
+ Cl =6 meq/1, I
+ Cl~ = 12 meq/1, I - .0145 M
+ Cl~ = 24 meq/1, I - .0285 M
EHC03 + SO + Cl - 48 meq/1, 1 = .0565 M
2000 4000 6000 8000
BED VOLUMES (BV) OF EFFLUENTS
10,000
12,000
FIGURE A7 MINI-COLUMN EFFLUENT CONCENTRATION HISTORIES, pH - 7.1
EFFECT OF IONIC STRENGTH ON COLUMN KINETICS
Q = 1 ml/mln, EBCT - 1 mln, BV = 1.0 ml
t „ = time to 90% of equilibrium In hours
-------
GLOSSARY
"as CaCC>3": Normality (N) can be converted to calcium carbonate equivalents.
There are 50 mg of CaCC>3 per milliequivalent.
bed: Alumina resin contained in a column. Water to be defluoridated by
alumina adsorption process is passed downward through the column
breakthrough: The point at wbic.i F~ concentration in the effluent reaches
the MCL or some predetermine' concentration.
elution: Application of NaOH soiut'.on to an exhausted alumina bed so as to
"elute" or drive off the adsorbed anions.
exhaustion: The step in an adsorption cycle in which the undesirable F~ ions
are removed from the water, and are adsorbed on alumina bed.
ion-exchange: A physicochemical process in which ions in the water being
treated replace and are exchanged for ions in a solid phase (alumina).
isotherm: A constant temperature plot of resin phase concentration of an
ion versus the water phase concentration of that ion, at the equilibrium
slope.
milliequivalent (meq): 1/1000 of a gram equivalent of ion.
regeneration: Restoration of the activated alumina after exhaustion to
reclaim its original adsorption capacity.
selectivity: A measure of the relative affinity for one ion over another
exhibited by the alumina. Selectivity is measured by the separation
factor.
separation factor (binary): The ratio of the distribution of ions between
the solid phase and the liquid phase at equilibrium.
A = distribution of A ions between solid & liquid phase
aB ° distribution of B ions between solid & liquid phase
VXA VQT * VCA
" yB/XB " VQT X VCB
superficial detention time (t): The time a particle of feed water spends in
the empty alumina bed assuming plug flow. It is calculated as the empty
58
-------
bed volume divided by the feed flow rate.
throughput: Ratio of the total amount of fluoride which comes in contact with
the adsorption column during a specified amount of time to the total
adsorptive capacity of the column.
59
-------
------- |