Removal of viruses from water by magnetic filtration

MISSING IMAGE

Material Information

Title:
Removal of viruses from water by magnetic filtration
Series Title:
Florida Water Resources Research Center Publication Number 40
Physical Description:
Book
Creator:
Bitton, Gabriel
Gifford, George E.
Pancorbo, Oscar C.
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Notes

Abstract:
The process of magnetic filtration was applied to the removal of T2 bacteriophage, MS2 bacteriophage, and poliovirus type 1 (Sabin) from water and wastewater. Since the effectiveness of this process is governed by the adsorption of the viruses onto the magnetic iron oxide seed, magnetite, those factors influencing adsorption were investigated. It was found that the adsorption of the three viruses studied to magnetite at a concentration of 500 ppm increased in the presence of cations, was not affected when the pH was varied from 5 to 9, and could be described by the Freundlich adsorption isotherm. Viruses were suspended in two wastewater effluents and subsequently removed by magnetic filtration. It was shown that organic color interfered with the adsorption process. The interference was reduced by the addition of CaC12, by the dilution of the wastewater with distilled water or by removing the organic color with activated carbon. Competition for adsorption was also exerted by 100 ppm of egg albumin, casein and dextran resulting in a significant reduction in the removal of poliovirus. Additionally, the infectivity of the magnetite-bound viruses was studied. Poliovirus, adsorbed to magnetite, was 100% infective whereas T2 phage displayed a lower and somewhat variable infectivity. Attempts were made to desorb poliovirus from magnetite by resuspension of the magnetite pellet in a variety of possible eluents. The best eluent was an isotonic, 10% solution of fetal calf serum buffered at the pH of 9 with Tris buffer. Preliminary results of the concentration of poliovirus by magnetic filtration are also presented and discussed. The experimental data obtained in this work show that magnetite is a good adsorbent towards viruses and that magnetic filtration can be effectively used for the removal of viruses from water and wastewater.

Record Information

Source Institution:
University of Florida Institutional Repository
Holding Location:
University of Florida
Rights Management:
All rights reserved by the source institution and holding location.
System ID:
AA00001514:00001


This item is only available as the following downloads:


Full Text



WATER RESOURCES

research center

Publication No. 40
Removal of Viruses From Water by
Magnetic Filtration
By
Gabriel Bitton, George E. Gifford
and Oscar C. Pancorbo


Department of Environmental Engineering Sciences
University of Florida
Gainesville


UNIVERSITY OF FLORIDA


h1j~4, j



'I
'w~sI
~.;















Publication No. 40

Removal of Viruses From Water by
Magnetic Filtration

By

Gabriel Bitton, George E. Gifford
and Oscar C. Pancorbo

Department of Environmental Engineering Sciences
University of Florida
Gainesville























Removal of Viruses From Water by
Magnetic Filtration

By

Gabriel Bitton, George E. Gifford
and Oscar C. Pancorbo


PUBLICATION NO. 40

FLORIDA WATER RESOURCES RESEARCH CENTER

RESEARCH PROJECT TECHNICAL COMPLETION REPORT

OWRT Project Number A-030-FLA

Annual Allotment Agreement Numbers

14-31-0001-5009
14-34-0001-6010

Report Submitted: October, 1976

The work upon which this report is based was supported in part
by funds provided by the United States Department of the
Interior, Office of Water Research and Technology/
as Authorized under the Water Resources
Research Act of 1964 as amended.












TABLE OF CONTENTS

Page

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

LIST OF FIGURES .................................................. v

ABSTRACT ......................................................... vi

PUBLICATIONS ..................................................... vii

THESIS ........................................................... vii

INTRODUCTION ..................................................... 1
REVIEW OF LITERATURE .............................................. 1
Adsorption of Viruses to Surfaces........................... 1
Implication of Sorptive Phenomena in the Removal of Viruses. 2
Use of Adsorbents in the Concentration of Viruses ........... 4
The Process of High Gradient Magnetic Separation............ 5

MATERIALS AND METHODS ............................................ 7

Viruses Used and Their Assays............................... 7
Magnetic Filtration of Adsorbed Viruses..................... 9
Wastewater Effluents Used in Adsorption Experiments......... 13
Organics Used in Adsorption Experiments..................... 15
Infectivity of Virus Adsorbed to Magnetite.................. 15
Elution of Adsorbed Poliovirus from Magnetite............... 15
Concentration of Poliovirus by Magnetic Filtration.......... 15
Statistical Treatment of Data............................... 16

RESULTS AND DISCUSSION........................................... 19

Effect of Cation Valency and Concentration.................. 19
Effect of pH ................................................ 30
Kinetics of the Adsorption Process.......................... 30
Effect of Magnetite Concentration........................... 34
Effect of Initial Virus Concentration....................... 34
Effect of Wastewater Effluents on the Adsorption of Viruses
to Magnetite ................................................ 39
Effect of Organics on the Adsorption of Poliovirus to Magne-
tite ........................................................ 51
Infectivity of Virus Adsorbed to Magnetite.................. 51
Elution of Poliovirus from Magnetite........................ 53
Concentration of Poliovirus by Magnetic Filtration.......... 56













TABLE OF CONTENTS -- Continued

Page

CONCLUSIONS ........................................................ 58

ACKNOWLEDGEMENTS................................................... 61

REFERENCES ......................................................... 62

APPENDIX ........................................................... 69












LIST OF TABLES


Table Page

1 Properties of viruses used in this study................ 8

2 Characteristics of the secondary effluents used in the
magnetic filtration of viruses.......................... 14

3 Effect of monovalent cations (Na+ and K+) on the removal
of bacteriophage T2 by magnetic filtration.............. 20

4 Effect of a monovalent cation (Na+) on the removal of
Poliovirus Type I (Sabin) by magnetic filtration........ 21

5 Effect of divalent cations (Ca++ and Mg++) on the
removal of bacteriophage T2 by magnetic filtration...... 22

6 Effect of divalent cations (Ca++ and Mg++) on the
removal of bacteriophage T4R(II) by magnetic filtration. 23

7 Effect of divalent cations (Ca++ and Mg++) on the
removal of bacteriophage MS2 by magnetic filtration..... 24

8 Effect of a divalent cation (Ca++) on the removal of
Poliovirus Type I (Sabin) by magnetic filtration........ 25

9 Effect of a trivalent cation (Al +++) on the removal of
bacteriophage T2 by magnetic filtration................. 26

10 Effect of a trivalent cation (Al ..) on the removal of
bacteriophage T4R(II) by magnetic filtration............ 27

11 Effect of a trivalent cation (Al ..) on the removal of
bacteriophage MS2 by magnetic filtration................ 28

12 Effect of a trivalent cation (Al+++) on the removal of
Poliovirus Type I (Sabin) by magnetic filtration........ 29

13 Effect of time of shaking on the removal of bacterio-
phage T2, bacteriophage MS2 and Poliovirus Type I
(Sabin) by magnetic filtration.......................... 33

14 Effect of a divalent cation (Ca++) on the removal of
bacteriophage T2 by magnetic filtration in the presence
of dome water........................................... 40

15 Removal of bacteriophage T2 by magnetic filtration in
the presence of dome water .............................. 41












LIST OF TABLES -- Continued

Table Page

16 Removal of Poliovirus Type I (Sabin) by magnetic filtra-
tion in the presence of dome water........................ 42

17 Removal of Poliovirus Type I (Sabin) by magnetic filtra-
in the presence of activated sludge effluent.............. 44

18 Interference by egg albumin, casein and dextran with the
removal of Poliovirus Type I (Sabin) by magnetic filtra-
tion ...................................................... 52

19 Infectivity of bacteriophage T2, bacteriophage MS2 and
Poliovirus Type I (Sabin) absorbed to magnetite........... 54

20 Elution of Poliovirus Type I (Sabin) from the magnetite
surface ................................................... 55

21 Concentration of Poliovirus Type I (Sabin) by magnetic
filtration ................................................ 57












LIST OF FIGURES


Figure Page

1 Particle size distribution of magnetite (Fisher) as mea-
sured with the Coulter Counter Industrial Model B (Coul-
ter Electronics, Illinois) ............................... 10

2 High gradient magnetic filter used to separate the virus-
magnetite complex from the water suspension.............. 11

3 High gradient magnetic filter supplied by Sala Magnetics,
Cambridge, Mass .......................................... 17

4 Effect of pH on the removal of bacteriophage T2, bacterio-
phage MS2 and Poliovirus by magnetic filtration.......... 31

5 Effect of magnetite concentration on the removal of bac-
teriophage T2, bacteriophage MS2 and Poliovirus Type I
(Sabin) by magnetic filtration ........................... 35

6 Freundlich isotherms for the adsorption of bacteriophage
T2, bacteriophage MS2 and Poliovirus Type I (Sabin) to
magnetite: variable virus concentration................. 37

7 Freundlich isotherms for the adsorption of bacteriophage
T2 to magnetite in distilled water and dome water: vari-
able magnetite concentration............................. 45

8 Freundlich isotherms for the adsorption of bacteriophage
MS2 to magnetite in distilled water and activated sludge
effluent: variable magnetite concentration.............. 47

9 Freundlich isotherms for the adsorption of Poliovirus
Type I (Sabin) to magnetite in distilled water and waste-
water effluents: variable magnetite concentration....... 49












ABSTRACT


The process of magnetic filtration was applied to the removal of
T2 bacteriophage, MS2 bacteriophage, and poliovirus type 1 (Sabin)
from water and wastewater. Since the effectiveness of this process
is governed by the adsorption of the viruses onto the magnetic iron
oxide seed, magnetite, those factors influencing adsorption were
investigated.

It was found that the adsorption of the three viruses studied
to magnetite at a concentration of 500 ppm increased in the pre-
sence of cations, was not affected when the pH was varied from 5 to 9,
and could be described by the Freundlich adsorption isotherm.

Viruses were suspended in two wastewater effluents and subse-
quently removed by magnetic filtration. It was shown that
organic color interfered with the adsorption process. The inter-
ference was reduced by the addition of CaC12, by the dilution of
the wastewater with distilled water or by removing the organic
color with activated carbon. Competition for adsorption was also
exerted by 100 ppm of egg albumin, casein and dextran resulting
in a significant reduction in the removal of poliovirus.

Additionally, the infectivity of the magnetite-bound viruses was
studied. Poliovirus, adsorbed to magnetite, was 100% infective
whereas T2 phage displayed a lower and somewhat variable infec-
tivity.

Attempts were made to desorb poliovirus from magnetite by re-
suspension of the magnetite pellet in a variety of possible
eluents. The best eluent was an isotonic, 10% solution of fetal
calf serum buffered at the pH of 9 with Tris buffer. Preliminary
results of the concentration of poliovirus by magnetic filtration
are also presented and discussed.

The experimental data obtained in this work show that magnetite
is a good adsorbent towards viruses and that magnetic filtration
can be effectively used for the removal of viruses from water and
wastewater.










PUBLICATIONS


1. G. Bitton, 0. Pancorbo and G. E. Gifford. 1976. Adsorption of
Poliovirus Type 1 to Magnetite and Subsequent Removal by Magne-
tic Filtration. Abstract, Amer. Soc. for Microbiology Ann.
Meeting, Atlantic City, N.J.


2. G. Bitton, J. L. Fox and H. G. Strickland. 1975.
Algae from Florida Lakes by Magnetic Filtration.
30:905-908.

3. G. Bitton, 0. Pancorbo and G. E. Gifford. 1976.
ing the Adsorption of Poliovirus to Magnetite in
water. Water Res., Vol. 10 (in press).


Removal of
Appl. Microbiol.


Factors Affect-
Water and Waste-


4. G. Bitton, J. L. Fox, G. E. Gifford and 0. Pancorbo. 1976.
Utilisation d'electroaimants dans 1'elimination des virus et
des algues presents dans les eaux de surface et les eaux uses.
Annales de l'ACAS, 43:133.


THESIS


Pancorbo, 0. C. 1976.
wastewater by magnetic
Dept. of Environmental
Florida, Gainesville.


The removal of viruses from water and
filtration. 115 pp. M.S. thesis,
Engineering Sciences, University of












INTRODUCTION


Viruses are obligatory intracellular parasites ranging in size
from 100 A to 4000 A in diameter (Pollard, 1953). They contain a
single type of nucleic acid, either DNA or RNA, surrounded by a pro-
tein coat. The ionization of certain protein coat groups results in
viruses acquiring a negative charge at neutral pH. The interaction
of viruses with their environment is determined by the nature of
their protein coat and is influenced by their ability to adsorb to
surfaces. Bitton (1975) reviewed the adsorption of viruses to sur-
faces and its importance in the removal of viruses by water treat-
ment processes and in the concentration and recovery of viruses.
Sproul et al. (1969) proposed that probably the most important
mechanism for virus "inactivation" in waste water treatment plants
is adsorption.

Among the various adsorbents studied, iron oxides have been
shown to display excellent sorptive capacity towards viruses (Bitton
and Mitchell, 1974; Pearson and Metcalf, 1974; Rao et al., 1968;
Warren et al., 1966). Using magnetite as an adsorbent, Bitton and
Mitchell (1974) removed the bacteriophage T7 by a "magnetic filtra-
tion technique." This process consists of adsorbing the virus onto
the magnetic iron oxide seed, magnetite, and subsequently removing
the virus-seed mixture from the water by pouring through a filter
placed in a background magnetic field.

The purpose of this study was to delineate further the effec-
tiveness of magnetic filtration towards the removal of viruses,
mainly those of animal origin, from water. Factors known to
influence the adsorption of viruses onto surfaces and that could,
therefore, affect the adsorption of viruses onto magnetite were
investigated. These included pH, electrolyte valency and concen-
tration, magnetite concentration, virus type and concentration, time
allowed for adsorption, and concentration and type of organic sub-
stances in the suspending medium. It was felt that the knowledge
gained from this study could be applicable in the removal of patho-
genic enteric viruses present in water and wastewater.

REVIEW OF LITERATURE

Adsorption of Viruses to Surfaces

Viruses are able to adsorb to biological (Tolmach, 1957) and non-
biological surfaces (Bitton, 1975). The adsorption process is usually
influenced by such factors as the type of virus, nature of the surface
involved, pH, and electrolyte and organic matter content of the sus-
pending medium.

Due to their widespread occurrence in soil and aquatic environ-
ments and to their large surface area, clay minerals have been studied
for their sorptive capacity towards viruses. The adsorption of entero-












viruses to such clays as kaolinite, illite, montmorillonite, and
bentonite has been shown to be significant and to increase in the
presence of cations with trivalent cations being more effective
than divalent or monovalent cations (Carlson et al., 1968; Schaub
and Sagik, 1975; Schaub et al., 1974). Virus sorption onto clays
has also been found to be independent of pH over the range of 3 to
9 (Bartell et al., 1960; Schaub et al., 1974) and to conform to
the Freundlich isotherm (Schaub et al., 1974). Organic materials
such as egg and bovine albumin, and fetal calf serum were also
shown to interfere significantly with the adsorption of viruses
to clays (Carlson et al., 1968; Schaub and Sagik, 1975). Varia-
bility in the adsorptive capacity of different clays towards viruses
was proposed by Carlson et al. (1968) to result from differences
in clay surface exchange capacity which is determined by the sur-
face charge density and clay particle geometry.

Activated carbon is extensively used in water and wastewater
treatment and consequently, its ability to adsorb viruses has been
investigated. Cookson (1969) studied the adsorption of T4 bacterio-
phage on activated carbon and determined that the adsorption rate
was influenced by the pH and the ionic strength of the medium. It
was proposed that the adsorption process involved an electrostatic
interaction between amino groups on the virus and carboxyl groups
on the activated carbon. At very high ionic strengths as well as
at low pH's, the tail fibers were unavailable for adsorption (Lauf-
fer and Bendet, 1962) and this decreased the rate of reaction. The
adsorption was described as a diffusion-limited process (Cookson,
1967), conformed to the Langmuir isotherm and did not inactivate
the virus (Cookson and North, 1967). Gerba et al. (1975) found
that the adsorption of poliovirus type 1 (strain LSc) to activated
carbon obeyed also the Freundlich isotherm and was reduced in the
presence of wastewater effluents. Desorption of viruses from an
activated charcoal column occurred when the wastewater pH was in-
creased or when glycine buffer adjusted to pH 11.5 was used to elute
the virus.

The adsorption of viruses onto such surfaces as glass, celite,
nitrocellulose, aluminum oxide and gold has been reported and was
found also to be influenced by the presence of cations and organic
materials (Boche and Quilligan, 1966; Shepard and Woodend, 1951;
Valentine and Allison, 1959).

Implication of Sorptive Phenomena in the Removal of Viruses

Pathogenic enteric viruses mainly from human and animal origin
are constantly present in our water systems and constitute a serious
public health hazard. Consequently, a great deal of research has
been devoted to finding efficient water treatments for their elimi-
nation. The removal of viruses by water (Berg, 1973a; Berg, 1973b;
Berg, 1975; Sproul, 1972) and wastewater (Berg, 1973a; Berg, 1973b;












Berg, 1975; Grabow, 1968) treatment processes usually involves stor-
age, biological and tertiary treatment followed by disinfection with
chlorine, iodine, bromine or ozone.

Among the various secondary (biological) wastewater treatment
processes,the activated sludge system has been the most intensively
studied for the removal of viruses. Clarke et al. (1961) found that
the adsorption of coxsackie A9 virus and poliovirus type 1 (Mahoney)
to activated sludge conformed to the Freundlich isotherm and resulted
in the removal of about 90% of the viruses. The virus-sludge complex
was shown to be very stable since only a small percentage of adsorbed
viruses could be recovered. However, the removal or inactivation of
viruses in stabilization ponds has received little attention. Data
from laboratory experiments did not show evidence of active removal
of poliovirus and reovirus by algae, but there was a significant
inactivation of viruses by sunlight at the surface of sewage matura-
tion ponds (Malherbe and Strickland-Cholmley, 1967). Sobsey and
Cooper (1973) observed that the reduction of poliovirus in stabili-
zation pond water was due to adsorption to solids.

Tertiary wastewater treatment processes are designed to further
polish the quality of secondary effluents with regards to suspended
solids and nutrients containing nitrogen and phosphorus. Among the
various tertiary wastewater treatment processes, activated carbon
has been shown to remove viruses very poorly. The factors influencing
the adsorption of viruses to activated carbon have already been dis-
cussed. Sproul et al. (1969) reported that virus removal from waste-
water by activated carbon is not dependable. Coagulation and floc-
culation using cationic polyelectrolytes as prime coagulants or as
coagulant aids in the presence of alum: has also been studied and
was found to be effective in the removal of viruses (Amirhor and
Engelbrecht, 1974; Chaudhuri and Engelbrecht, 1972; Thorup et al.,
1970). This process was shown to be salt dependent (Thorup et al.,
1970) and unfortunately, organic matter interfered with virus re-
moval (Amirhor and Engelbrecht, 1974; Chaudhuri and Engelbrecht,
1972). Even though sand filtration plays a role as a tertiary treat-
ment process, its standard and more important use in potable water
treatment is the primary reason for the intensive investigation of
its role in virus removal. Dieterich (1953) found that adsorption
was the primary mechanism responsible for the removal of a bacterio-
phage during sand filtration and that egg albumin significantly inter-
fered with virus removal. Cations, on the other hand, were reported
to enhance the retention of viruses by sand columns (Lefler and Kott,
1974). Although sand is a poor adsorbent due to its small surface
area (Dieterich, 1953), significant virus removal can be achieved by
combining sand filtration with alum flocculation (Gilcreas and Kelly,
1955; Robeck et al., 1962). The removal of viruses from water by
diatomaceous-earth filtration (Brown et al., 1974a; Brown et al.,
1974b) and by the use of coal as sorbent (Oza and Chaudhuri, 1975;
Oza et al., 1973) was also investigated, and the results were encou-












raging. However, more research on the effectiveness of these two
treatments is in order.

Use of Adsorbents in the Concentration of Viruses

Viruses occur in natural water bodies in very low concentrations
and therefore, their detection is dependent upon their adequate con-
centration. Various virus adsorbents have been investigated for
their potential application in the concentration of viruses from di-
lute suspensions (Hill et al., 1971). The membrane-adsorption tech-
nique has been demonstrated to effectively concentrate viruses.
Wallis and Melnick (1967a) showed that the adsorption of viruses on
membrane filters was enhanced in the presence of cations and inhibited
in the presence of proteinaceous matter such as cell extracts or
serum. The optimal pH for adsorption was found to be 5. This method
was further used in the concentration of viruses from water (Hill et
al., 1974) and wastewater (Wallis and Melnick, 1967b). Synthetic
insoluble polyelectrolytes (Wallis et al., 1970, 1971), cationic
exchange resins (Muller and Rose, 1952; Puck and Sagik, 1953) and
anionic exchange resins (LoGrippo, 1950; Puck and Sagik, 1953) have
also been found to be effective virus adsorbents and consequently,
to possess the ability to concentrate and purify viruses. Finally,
precipitates of aluminum hydroxide, aluminum phosphate, and calcium
phosphate have been used effectively to concentrate viruses (Wallis
and Melnick, 1967c, 1967d). Viruses may also be concentrated by a
variety of techniques which do not involve adsorption onto surfaces
(Hill et al., 1971; Shuval et al., 1967) but these methods will not
be discussed in this review.

Among the various adsorbents used for virus concentration and
removal, iron oxides have received little attention. Warren et al.
(1966) reported for the first time the adsorption of influenza virus
to an iron oxide, hematite. Concentration and purification of this
virus were achieved by adsorption onto hematite and subsequent elu-
tion with a 10% sodium phosphate solution at a pH between 7.5 and 8.5
resulting in 10-fold concentrates. Rao et al. (1968) later used an
iron oxide (M.O. 2530) column to adsorb and concentrate a strain of
coxsackie virus A9. It was found that concentrations as high as 2 x 106
pfu/ml of this virus could be removed, without detecting any virus
in the filtrate, by 25 grams of the iron oxide. Concentration of the
coxsackie virus A9 by adsorption to the iron oxide and subsequent
elution with foetal calf serum resulted in 5-fold concentrates with
recoveries ranging from 55% to 95%, depending on the initial virus
concentration. Pearson and Metcalf (1974) studied the adsorption of
several enteroviruses to the same iron oxide, M.O. 2530, and found
virus recovery most effective when elutions were made from thin
layers of iron oxide under alkaline pH and in the presence of 3%
isotonic beef extract. Unfortunately, the use of columns or thin
layers of iron oxide may ultimately result in the clogging of the












filter (Rao et al., 1968). Due to the open matrix of the filter
used in High Gradient Magnetic Separation (HGMS), the problem of
clogging is greatly minimized.

The Process of High Gradient Magnetic Separation (HGMS)

Convential magnetic separators have been used for many years to
remove strongly magnetic iron-bearing particles larger than 100 mi-
crons in size from non-magnetic media. These devices are used to
remove magnetic impurities in a variety of feeds, and to concentrate
magnetic materials such as iron ores for their beneficiation (Ober-
teuffer, 1973).

Recently, a great deal of research has been devoted to the
development of "High Gradient Magnetic Separation." This process,
unlike convential magnetic separation, is able to separate weakly
paramagnetic materials of micron size by maximizing the magnetic
force on such particles. The magnetic force on a particle is given
by the equation,

Fx = VM dH (1)

where V is the volume of the particle, M is the magnetic moment of
the particle in the field H, and dH is the field gradient in the x
direction (Oberteuffer, 1973). In the case of paramagnetic materials,

M = xH (2)
and
Fx = V XH d (3)

where X is the magnetic susceptibility. From equation 3, it can bedH
seen that both a large magnetic field H and a large field gradient dx
will yield a maximum magnetic force Fx. A large magnetic force is
required in order to overcome competing forces (e.g., gravitational)
that oppose the magnetic separation of particles (Oberteuffer, 1973).
A high gradient magnetic separator consists of a filter containing a
ferromagnetic matrix (usually stainless steel wool) placed in a
strong magnetic field. Such a matrix provides a very large number
of trapping sites for susceptible particles and enables high flow
rates (50-150 gal. min.-1 ft.-2) due to its loosely packed structure.
With the magnetic field on, the magnetic particles in the feed slurry
are trapped at the points and edges of the matrix fibers while non-
magnetic constituents of the slurry pass through the filter easily.
Due to the low residual magnetization of the matrix fibers, magnetic
particles trapped in the matrix during the feed mode can be easily
washed out by turning the magnetic field off and backflushing (Mit-
chell et al., 1975a, 1975b).

Several industrial applications have been found for high gradient
magnetic separation. In the clay industry, for example, this process
is used to clean kaolin from micron size iron stained titaniferous












materials (Oder, 1973). Such impurities can seriously reduce the
use of kaolin as a paper coating material. High gradient magnetic
separation has also been used in the beneficiation of semitaconites
which are made up of very small particles of weakly magnetic iron
oxides intermixed with gangue material such as slate and chert.
Kelland (1973) has used successfully the high gradient magnetic
separation technique to separate the iron minerals from the gangue
and this process resulted in an increase in the iron grade (iron
content) of the semitaconites. This technique has also been applied
to the desulfurization of coal. Trindade and Kolm (1973) obtained a
50 to 75% reduction in the sulfur content of Brazilian coal.

High gradient magnetic separation has been shown to be effec-
tive in the removal of various pollutants from water and wastewater
(DeLatour and Kolm, 1976; Mitchell et al., 1975a). Removal may be
achieved in one of two ways depending upon the nature of the pollu-
tant. For magnetic contaminants, such as the fine iron oxide par-
ticles found in steel mill effluents and boiler feed waters, high
gradient magnetic separation may be used alone to obtain their
removal. However, for the more common non-magnetic pollutants
found in natural waters, a magnetic seeding technique is required.
This technique involves the adsorption of non-magnetic contaminants
onto a magnetic seed, magnetite, and subsequent passage of the mix-
ture through a high gradient magnetic separator. The magnetite
and adsorbed pollutants are trapped on the matrix of the separator
and can be easily washed out when the magnetic field is turned off.
This process has been called "magnetic filtration" and it has been
shown to reduce effectively coliforms, color, turbidity (DeLatour,
1973; Mitchell et al., 1975a, 1975b), phosphate (Bitton et al., 1974;
DeLatour, 1973), algae in lake water (Bitton et al., 1975) and the
phage T7 (Bitton and Mitchell, 1974). Bitton and Mitchell (1974)
found that the removal of the bacteriophage T7 by this process was
dependent on the presence of cations and independent of virus con-
centration in the range of 30 pfu/ml to 14 x 103 pfu/ml. Although
having many similar properties, bacteriophages and animal viruses
are, nevertheless, inherently different. Therefore, an animal virus
study should be included when assessing the virus removing effec-
tiveness of any treatment. Unfortunately, no research has previously
been done on the removal of animal viruses by magnetic filtration.
Additionally, little is known about the effect of such factors as
pH and the presence of organic materials on the removal of viruses
by this process. The research reported in this report attempts to
fill in these gaps in theknowledge of virus removal by magnetic
filtration. Finally, it can be said that high gradient magnetic
filtration is an efficient process capable of removing a wide variety
of pollutants from wastewater. Consequently, its use either as a
replacement for small sewage treatment facilities or in large facili-
ties as an advanced treatment process is not far off (Mitchell et al.,
19-4, 1975a, 1975b).












MATERIALS AND METHODS

All virological work was performed aseptically and in the case
of poliovirus, all work was done in a room previously sterilized by
ultraviolet irradiation.

All glassware which came into contact with the virus was soaked
overnight in Haemo-Sol (Scientific Products) and subsequently rinsed
in tap water and distilled water. Glassware was sterilized by auto-
claving.

Viruses Used and Their Assays

The viruses used in this study were the T2 bacteriophage, T4R(II)
bacteriophage, MS2 bacteriophage and poliovirus type 1 (Sabin strain).
Table 1 shows some important properties of these viruses. Bacterio-
phages were used in this study to obtain preliminary data on a par-
ticular phase of research and to resolve any difficulties in the
experimental procedure prior to initiating experiments with polio-
virus. Representatives from two different groups of bacteriophages
were used: MS2 phage representing the group of small, RNA phages
and T2 and T4R(II) phages representing the large, DNA phages posses-
sing a contractile tail. These two types of bacteriophages were
selected in an attempt to compare their behavior during magnetic
filtration with that of poliovirus and to determine which is a
better model of enterovirus removal by magnetic filtration. The
bacteriophage MS2 should be a better model of poliovirus behavior
than T2 and T4R(II) bacteriophages due to its greater similarity to
poliovirus in size, particle mass and nucleic acid type (Table 1).
Preliminary investigations showed that T2 and T4R(II) bacterio-
phages behaved similarly during magnetic filtration, and therefore,
subsequent experimentation was only carried out with T2 phage. Polio-
virus type 1 (Sabin) was used as a representative human enteric virus.
It is generally agreed that the use of an animal virus in a viral
adsorption study increases the reliability of the results obtained.

The T2 and T4R(II) bacteriophage stocks and their host Esche-
richia coli B were obtained from Dr. Donna H. Duckworth, Department
of Immunology and Medical Microbiology, University of Florida. The
phages were refrigerated at 4C in.1-9medium (Adams, 1959) and the
host, E. coli B, was maintained at room temperature on 2% nutrient
agar (Difco) slants. A complete list of all media and solutions used
in this study including their composition and source is presented in
the Appendix. These T-even bacteriophages were assayed by the plague
technique as described by Adams (1959) using the double layer plating
procedure.

The MS2 bacteriophage stock was obtained from Dr. Anthony
Pfister, Department of Immunology and Medical Microbiology, Univer-
sity of Florida. The host Escherichia coli C3000 (ATCC #15597) was












Table 1. Properties of viruses used in this study


Virus Nucleica Isoelectricb Sizec (mp) Dry Particled pH Stabilitye
Acid Point Mass Range
(x 10-16 g)


T2 phage DNA 4.2 head 65 x 95 3.3 5.0 9.0
tail 25 x 100

T4 phage DNA head 65 x 95 3.3 5.0 9.0
tail 25 x 100

MS2 phage RNAk 3.9 25 0.06 -

Poliovirus RNA 4.5 and 7.0 27 0.14 3.6 8.4



a From Davis et al. 1973.
b
From Mandel, 1971; Overby et al. 1966; Sharp et al. 1946.
c From Overby et al. 1966; Schwerdt and Schaffer, 1955; Williams and Fraser, 1953.
From Kellenberger, 1962; Overby et al. 1966; Schwerdt and Schaffer, 1955; Taylor et al. 1955.

e From Bachrach and Schwerdt, 1952; Putnam, 1953; Sharp et al. 1946.













obtained from Miles Laboratories, Kankakee, Illinois. The phage was
refrigerated at 40C in Tris buffer (see Appendix) and the host, E.
coli C3000, was maintained at room temperature on 2% nutrient agar
(Difco) slants. Assay of this virus was by the plaque technique
using the double layer procedure (Adams, 1959).

The Poliovirus type 1 (Sabin strain) stock suspension used
in this study was prepared by infecting a monolayer culture of AV3
cells in a 32 ounce bottle. After a 1 hour adsorption period with
tilting at 15 minute intervals, 40 ml of Eagle's Minimal Essential
Medium (MEM) plus 10% fetal calf serum (see Appendix) were added.
After two days of incubation at 379C, the overlay medium containing
poliovirus was decanted, centrifuged for 15 minutes at 1500 rpm to
remove debris and then distributed to 1 ml ampules which were im-
mediately frozen at -700C. The poliovirus was assayed by the
plaque assay technique on AV3 (human amnion) cell monolayers.
This assay was performed by inoculating a drained AV3 monolayer
with 0.2 ml of the virus suspension which had been diluted in
Eagle's MEM + 5% calf serum + 0.03 M Hepes buffer (see Appendix)
to yield 100-300 plaques per bottle. Following inoculation, a one
hour adsorption period with tilting at 15 minute intervals was
allowed. The infected cells were then overlayed with 4 ml of methyl
cellulose overlay (see Appendix). After incubation at 37C for 48
hours, the methyl cellulose overlay was decanted and the monolayer
was stained with crystal violet (see Appendix). Plaques were sub-
sequently counted using an Omega Enlarger B22 (Simmon Bros., Wood-
side, N.Y.).

Magnetic Filtration of Adsorbed Viruses

The salts used in this study were NaCl (Fisher Scientific,
Fairlawn, N.J.), CaC12 (Fisher) and A12(SO4)3 18 H20 (Mallinc-
krodt, St. Louis, Mo.). The magnetite was supplied by Fisher
Scientific (catalog no. 1119) and its particle size distribution
was determined using a Coulter Counter Industrial Model B (Coulter
Electronics, Illinois). Figure 1 shows that the particle size
varied from 3 to 30 p with a maximum count between 3 and 5 p.

The adsorption of viruses to magnetite and the subsequent
removal of the magnetite-virus complex from the water suspension
by filtration through a magnetic separator was undertaken as
described below. Appropriate volumes from concentrated stock
solutions of magnetite, salt, and virus were added successively in
batch experiments to samples of water resulting in a final volume
of 100 ml. The pH of each solution was measured using a Beckman
Expandomatic SS-2 pH meter and the resulting pH's were usually
between 6 and 7. If required, pH adjustments were made using 0.1





















4



o


0
..-J



2





1 I 111- I I I II- II
0 10 20 30
Particle diameter (vj)


Figure 1. Particle size distribution of magnetite (Fisher) as measured with
the Coulter Counter Industrial Model B (Coulter Electronics, Illinois)







-11-


Figure 2. High gradient magnetic filter used to separate the
virus-magnetite complex from the water suspension

This filter consists of a stainless steel wool ma-
trix (compaction equal to lOg/50 cm3) placed in a
background magnetic field of 2000 gauss.








-12-


OWN A












fi-ft







-13-


N NaOH or HC1. These mixtures were then shaken at room temperature
at approximately 170 oscillations/min and poured through a filter
placed in a background magnetic field of 2000 gauss (see Figure 2).
The filter was made of a matrix of stainless steel wool (supplied
by Dr. E. Maxwell, Francis Bitter National Magnet Laboratory, M.I.T.,
Cambridge, Mass.) having the compaction of 10 g/50 cm3 and magnetized
in a background magnetic field. Each filtration experiment was
carried out in triplicate and one control without magnetite was
included in each experimental condition. Samples were removed and
assayed for virus immediately after the shaking period (i.e., just
prior to filtration) and after filtration through the magnetic
separator. The controls without magnetite were included in order
to account for loss of infectivity due to adsorption to containers
or inactivation. The work described above was all done aseptically.

Wastewater Effluents Used in Adsorption Experiments

The wastewater effluents used in the adsorption experiments were
"dome water" and an activated sludge effluent. "Dome water" is a
secondarily treated effluent sampled at the edge of a cypress dome,
located north of Gainesville, Florida. This effluent originated
from a package treatment plant and became highly colored after
standing in the dome site. When required, the removal of these
coloring materials was undertaken by filtration through an activated
carbon column, 11 cm in length and 3 cm in diameter. The column con-
tained 10 cm of Filtrasorb 400 (Calgon Co.) and 1 cm of fine particles
of Hydrodarco B (Atlas Powder Co., N.Y.). The flow rate through the
activated carbon column was 0.37 ml/min.

The activated sludge effluent was sampled, after secondary
settling and prior to chlorination, at the University of Florida
campus sewage treatment plant.

The effluents were immediately brought back to the laboratory
and subsequently clarified by filtration through a Whatman filter
no. 41, sterilized by autoclaving and stored at 4C until used.
Total organic carbon, turbidity, color, pH and conductivity were
determined in the laboratory using a Beckman 915 total organic
carbon analyzer, a Hach model 2100A turbidimeter, a Bausch and
Lomb Spectronic 88, a Beckman Expandomatic SS-2 pH meter and a
conductivity bridge Model RC 16B2, respectively. The values
for the various characteristics of the effluents are shown in
Table 2.












Table 2. Characteristics of the secondary effluents used in the magnetic filtration of viruses



Parameter Dome Wateri Dome Dome Water #1 Activated
Water #lb After Activated Sludge
Carbon Treatment Effluentc


6.20 7.65


7.10


8.25
(adjusted to 7.10
with 0.1 N HC1)


7.40 7.73


Color (mg/l as Pt)

Turbidity (JTU)

TOC (mg/1)

Conductivity
(pmho/cm at 250C)


350.0 750.0

3.0 4.0

26.5 37.0

227.3 385.0


575.0

4.0

31.0

385.0


320.9


30.0 35.0

0.6 2.0

4.0 12.0

385.0 472.7


a Various batches of dome water were collected and the values given for each parameter
represent the high and low measurements obtained.


This is the first batch of dome water collected.


c Various batches of activated sludge effluent were collected and the values
parameter represent the high and low measurements obtained.


given for each







-15-


Organics Used in Adsorption Experiments

The interference by organic with the adsorption of poliovirus
to magnetite and consequently, with the removal of the virus by
magnetic filtration was investigated. The organic substances
used were egg albumin (cat. no. B255), casein (cat. no. B337), both
supplied by Difco Laboratories (Detroit, Mich.) and dextran-5 x 105
MW (cat. no. D5251) supplied by Sigma Chemical Co. (St. Louis, Mo.).
Stock solutions (0.1%) of these substances were prepared in dis-
tilled water and were then sterilized by autoclaving and stored
at 40C.

Infectivity of Virus Adsorbed to Magnetite

In order to determine the infectivity of virus adsorbed to
magnetite, an experiment was undertaken as described below. Viruses
were shaken (, 170 oscillations/min.) for 20 min. in the presence
of 500 ppm of magnetite and 1610 ppm of CaC12. Low virus concen-
trations were used so that samples might be plated directly with-
out requiring dilution. Samples were removed and directly plated
immediately after addition of the virus and after 20 min. adsorp-
tion period. Following the adsorption period, the magnetite was
pelleted by placing the mixtures next to a magnet and the supernatants
were subsequently assayed for virus. Controls without magnetite were
studied in order to account for loss of infectivity due to adsorption
to containers or inactivation.

Elution of Adsorbed Poliovirus from Magnetite

Attempts were made to elute the adsorbed poliovirus from mag-
netite. The procedure employed is described below. Poliovirus
was shaken (, 200 oscillations/min.) for 20 min. in the presence
of 500 ppm of magnetite and 1610 ppm of CaC12. Ten ml aliquots were
then removed and distributed into sterile test tubes. The magnetite
in each test tube was then pelleted by placing the mixture next to a
magnet. Supernatants were removed for assay and then discarded. The
magnetite pellets were resuspended in 10 ml of the eluents shown in
Table 20 (the composition and source of these eluents appear in the
Appendix). After a 1 hour elution period with periodic shaking at
40C (except for glycine buffer which was allowed only 1 min of con-
tact time), the magnetite was pelleted again. The supernatants were
then assayed for desorbed viruses.

Concentration of Poliovirus by Magnetic Filtration

Recovery of poliovirus from 4 liters of distilled water contain-
ing initial virus concentrations ranging from 1.0 x 103 to 1.3 x 103
pfu/ml was undertaken as described below. Poliovirus was shaken
(A 200 oscillations/min.) for 20 min. in the presence of 1000 ppm
of magnetite and 1610 ppm of CaC12. The mixtures were subsequently







-16-


poured through the magnetic filter in order to separate the polio-
virus-magnetite complex from the water suspension. Poliovirus was
then eluted from the magnetite surface with a small volume of 10%,
isotonic fetal calf serum, Tris buffered, pH = 9. The procedure
used for the recovery of poliovirus varied with each experiment
performed.

In a first experiment, 100 ml of the eluent was passed twice
through the filter with the filter in the magnetic field and once
(the last passage) with the filter outside the magnetic field. The
eluate; (separated from any magnetite) was assayed for virus after
each passage through the filter. The Sala magnet pictured in
Figure 3 was used in this experiment (and only this experiment).

In a second experiment, 50 ml of the eluent was introduced
into the filter and then the filter was shaken manually for 1 min
and then shaken mechanically overnight at 40C. After each shaking
period, the eluate was separated from magnetite and then assayed
for viruses.

In the third experiment, the matrix of the filter was removed
(along with all the magnetite) and placed in a sterile beaker con-
taining 100 ml of the eluent. The matrix was then sonicated in an
ice bath for 10 min. using the standard probe of a Branson Sonifier
(Danbury, Conn.) model S-75 which has a power output of 75 watts and
an ultrasonic frequency of 20 kc/sec. The eluate was then assayed
for viruses.

Statistical Treatment of Data

The two statistical procedures used to treat data were: 1. the
small-sample, t test for comparing two means, and 2. linear regres-
sion analysis to determine the least squares regression line for a
set of points and the coefficient of determination r2 which is a
measure of the strength of the relationship represented by the regres-
sion line. The Hewlett-Packard Calculator Model 9810A and Statistics
PackageV-6 were used to perform the statistical analysis.







-17-


Figure 3. High gradient magnetic filter supplied by Sala Magnetics,
Cambridge, Mass.

This filter was used in the concentration of Poliovirus,
experiment 1.
















































I/.eW


-15-







-19-


RESULTS AND DISCUSSION

The purpose of this study was to determine the effective-
ness of magnetic filtration in removing viruses from water.
Consequently, factors that could influence the adsorption of
viruses onto magnetite were investigated. These included
cation valency and concentration, pH, time allowed for adsorp-
tion, magnetite concentration, virus type and concentration and
organic substances in the suspending medium.

Effect of Cation Valency and Concentration

The effect of monovalent (Na+, K+), divalent (Ca++, Mg++)
and trivalent (Al +++) cations on virus adsorption to magnetite
and on the subsequent virus removal by magnetic filtration was
studied (Tables 3-12). It was observed that, in the presence of
any salt under consideration, the removal of poliovirus and MS2
phage approached or exceeded the 99% level when the salt concen-
tration was at or above 20 ppm. The percent removal of T2 phage
was below that of poliovirus or MS2 phage for any cation used and
at any concentration. The highest removal of T2 phage (95.9%)
was achieved in the presence of 16100 ppm of CaC12 (see Table 5).
The removal obtained for the T4r (II) phage in the presence of
1610 ppm of CaC12 or 1380 ppm of MgC12 (Table 6) was similar to
that found for the T2 phage (Table 5). In the presence of 100
ppm of alum (see Tables 9 and 10), however, the removal of the
T4r (II) phage (99.2%) greatly exceeded that displayed by the
T2 phage (85.4%). There was no significant difference in the
removal of the phage T2 in the presence of K* or Na+ cations (Table
3). Calcium cations (Ca++) were found more effective than Mg++
cations at ionic strengths (p) equal to 4.35 x 10-2 and 4.35 x
10-1 (see Tables 5 and 6).

It has been shown by other workers that the electrolyte
content of the suspending medium is important for the adsorption
of viruses to particles such as iron oxides, clays, activated
carbon, polyelectrolytes, sand and soil (Bitton and Mitchell, 1974;
Carlson et al., 1968; Cookson, 1969; Thorup et al., 1970; Lefler
and Kott, 1974; Drewry and Eliassen, 1968). Furthermore, it is
known that, on a concentration or molar basis, trivalent cations
are more efficient than divalent cations which in turn are more
efficient than monovalent ones. Viruses are colloidal particles
having a net negative surface charge at neutral pH. Most of the
adsorbents studied also carry a net negative surface charge at
neutral pH. An increase in the ionic strength of the suspending
medium leads to a reduction of the thickness of the double-layer
around the particles which are then bound by attractive forces,
namely, London-Van der Waals forces (Clark et al., 1971; Osipow,
1962). The adsorption of viruses to magnetite, as seen in Tables
3 through 12, follows the trends discussed above with the exception







-20-


Table 3. Effect of monovalent cations (Na+ and K+) on the
removal of bacteriophage T2 by magnetic filtration


Ionic Strength (vi) Concentration (ppm) % Removal of T2a
of the Salt Used of the Salt Used by Magnetic Filtrationb
NaCl or KC1 NaCl KC1 NaCl KC1

4.35 x 10-4 25.4 32.4 62.1 46.8

4.35 x 10-3 254 324 56.1 55.9

4.35 x 10-2 25-1, 3240 72.9 79.5

4.35 x 10-1 25400 32400 65.2 61.0



a
The initial bacteriophage T2 concentration was 2000 pfu/ml.
b The bacteriophage T2 was shaken (,130 oscillations/min) for
20 min in the presence of 300 ppm of magnetite and various
concentrations of NaCl or KCI. These mixtures were subse-
quently poured through the magnetic filter.







-21-


Table 4. Effect of a monovalent cation (Na+) on the removal of
Poliovirus Type I (Sabin) by magnetic filtration


Ionic Strength
of NaCl (p)


4.35 x 10-4

4.35 x 10-3


4.35 x 10-2

4.35 x 10-1


Concentration
of NaCl (ppm)


25.4


2540

25400


% Removal of Poliovirusa
by Magnetic Filtrationb


93.0

98.7

99.1

98.9


a The initial Poliovirus concentration was 15 x 103 pfu/ml.

b The Poliovirus was shaken (% 200 oscillations/min) for 20 min
in the presence of 500 ppm of magnetite and various concentra-
tions of NaCI. These mixtures were subsequently poured through
the magnetic filter.







-22-


Table 5. Effect of divalent cations (Ca"* and Mg+") on the removal
of bacteriophage T2 by magnetic filtration


Ionic Strength (v) Concentration (ppm) % Removal of T2a by
of the Salt Used of the Salt Used Magnetic Filtrationb

CaC12 or MgC12 CaC12 MgC12 CaC12 MgCl2


4.35 x 10-4 16.1 13.8 26.1 35.4

4.35 x 10-3 161 138 82.1 80.7

4.35 x 10-2 1610 1380 93.1 81.2

4.35 x 10-1 16100 13800 95.9 82.3


a
The initial bacteriophage T2 concentration was 2000 pfu/ml.

b The bacteriophage T2 was shaken (% 130 oscillations/min) for
20 min in the presence of 300 ppm of magnetite and various
concentrations of CaC12 or MgC12. These mixtures were sub-
sequently poured through the magnetic filter.







-23-


Table 6. Effect of divalent cations (Ca+" and
of bacteriophage T4R(II) by magnetic


Mg++) on the removal
filtration


Ionic Strength (v) Concentration (ppm) % Removal of T4R(II)a
of the Salt Used of the Salt Used by Magnetic Filtrationb

CaC12 or MgC12 CaC12 MgCl2 CaC12 IgC12

4.35 x 10-4 16.1 13.8 62.6 63.3

4.35 x 10-3 161 138 60.1 69.8

4.35 x 10-2 1610 1380 85.1 71.2

4.35 x 10-1 16100 13800 92.5 76.4


a The initial bacteriophage T4R(II) concentration was 2000 pfu/ml.

b The bacteriophage T4R(II) was shaken (% 130 oscillations/min)
for 20 min in the presence of 300 ppm of magnetite and various
concentrations of CaC12 or MgC12. These mixtures were subse-
quently poured through the magnetic filter.







-24-


Table 7. Effect of divalent cations (Ca++ and Mg+") on the
removal of bacteriophage MS2 by magnetic filtration



Ionic Strength (p) Concentration (ppm) % Removal of MS2a by
of the Salt Used of the Salt Used Magnetic Filtration b

CaC12 or MgCl2 CaC12 MgC12 CaC12 MgC12


4.35 x 10-2 1610 1380 99.6 99.6



a The initial bacteriophage MS2 concentration was 7000 pfu/ml.

b The bacteriophage MS2 was shaken (% 130 oscillations/min) for
20 min in the presence of 500 ppm of magnetite and the concen-
tration of CaC12 or MgCl2 shown above. These mixtures were
subsequently poured through the magnetic filter.







-25-


Table 8. Effect of a divalent cation (Ca+") on the removal of
Poliovirus Type I (Sabin) by magnetic filtration


Ionic Strength Concentration % Removal of Poliovirusa
of CaC12 (P) of CaC12 (ppm) by Magnetic Filtrationb


4.35 x 10-4 16.1 16.0

4.35 x 10-3 161 99.8

4.35 x 10-2 1610 99.6

4.35 x 10-1 16100 99.1



a The initial Poliovirus concentration was 15 x 103 pfu/ml.

b The Poliovirus was shaken (, 200 oscillations/min) for 20
min in the presence of 500 ppm of magnetite and various
concentrations of CaC12. These mixtures were subsequently
poured through the magnetic filter.







-26-


Table 9. Effect of a trivalent cation (Al"++) on the removal of
bacteriophage T2 by magnetic filtration


Ionic Strength of Concentration of % Removal of T2a by
A12(S04)3 18H20 () Al2(SO4)3 18H20 (ppm) Magnetic Filtrationb



4.35 x 10-4 20 64.8

13.05 x 10-4 60 43.6

21.75 x 10-4 100 85.4



a The initial bacteriophage T2 concentration was 5600 pfu/ml.

b The bacteriophage T2 was shaken (% 130 oscillations/min) for
20 min in the presence of 300 ppm of magnetite, various
concentrations of A12(SO4)3 18H20 and pH's adjusted to
6.0 + 0.1 with l.ON NaHCO3. These mixtures were subsequently
poured through the magnetic filter.







-27-


Table 10.


Effect of a trivalent cation (Al+++) on the removal of
bacteriophage T4R(II) by magnetic filtration


Ionic Strength of Concentration of % Removal of T4R(II)a
A12(SO4)3 18H20 (N) Al2(S04)3 18H20 by Magnetic Filtrationb
(ppm)


4.35 x 10-4 20 41.7

13.05 x 10-4 60 99.5

21.75 x 10-4 100 99.2



a The initial bacteriophage T4R(II) concentration was 3000 pfu/ml.

b The bacteriophage T4R(II) was shaken (, 130 oscillations/min)
for 20 min in the presence of 300 ppm of magnetite, various
concentrations of A12( SO4)3 18H20 and pH's adjusted to 6.0
+ 0.1 with l.ON NaHCO3. These mixtures were subsequently
poured through the magnetic filter.







-28-


Table 11.


Effect of a trivalent cation (Al+++) on the removal of
bacteriophage MS2 by magnetic filtration


Ionic Strength of Concentration of % Removal of MS2a by
A12(SO4)3 18H20 (1) A12(S04)3 18H20 Magnetic Filtration
(ppm)


4.35 x 10-4 20 99.6

13.05 x 10-4 60 100.0

21.75 x 10-4 100 100.0



a The initial bacteriophage MS2 concentration was 3700 pfu/ml.

b The bacteriophage MS2 was shaken (% 130 oscillations/min)
for 20 min in the presence of 500 ppm of magnetite, various
concentrations of A12(S4)3 18H20 and pH's adjusted to
6.0 + 0.1 with 0.1N NaOH. These mixtures were subsequently
poured through the magnetic filter.







-29-


Table 12.


Effect of a trivalent cation (Al+++) on the removal
of Poliovirus Type I (Sabin) by magnetic filtration


Ionic Strength of Concentration of pHa of % Removal
A12(S04)3 18H20 A12(Sl )3 18H20 Solution of Poliovirusb by
(0) (ppm) Magnetic Filtrationc


4.35 x 10-4 20 4.64 81.6

13.05 x 10-4 60 4.40 98.5

21.75 x 10-4 100 4.29 90.4

4.35 x 10-3 200 4.14 91.7



a The pH's reported represent the actual measured pH's of the
solutions; no pH adjustments were made.

b The initial Poliovirus concentration was 15 x 103 pfu/ml.

c The Poliovirus was shaken (u 200 oscillations/min) for 20
min in the presence of 500 ppm of magnetite and various
concentrations of A12(S04)3 18H20. These mixtures were
subsequently poured through the magnetic filter.







-30-


of the 93% removal of poliovirus and 62.1% removal of T2 phage
obtained in the presence of only 25.4 ppm of NaCl (Tables 3 and
4). The adsorption of poliovirus was also found to decrease
when the alum concentration was increased to above 60 ppm (see
Table 12). As no pH adjustments were made when adding the alum,
this decrease in sorption was probably due to the lowering of
the pH by 100 and 200 ppm of alum to pH values of 4.29 and 4.14,
respectively. These values are below the isoelectric point of
4.5 reported by Mandel (1971) for poliovirus type 1. We also
noted that in the pH experiment described in the next section,
the adsorption of poliovirus was also weak below pH 5. One
may also add that, among the three bacteriophages studied,
MS2 is the only one which has an adsorption pattern similar to
that of poliovirus.

Effect of pH

The pH of the suspending medium has been reported to be
significant in the adsorption of viruses to such surfaces as
soil, membrane filters and synthetic insoluble polyelectrolytes
(Reece, 1967; Wallis and Melnick, 1967a; Wallis et al., 1971).
In order to simulate the pH of most natural waters, the adsorp-
tion of T2 phage, MS2 phage and poliovirus to magnetite was
studied in the pH range of 4 to 9. Figure 4 shows that, above
pH 5, the removal of poliovirus was in the 98-99% level and did
not vary significantly. However, below pH 5, a decrease in
removal was observed for this virus. For the phages T2 and MS2,
the removal (98-99% level) did not vary significantly in the pH
range of 4 to 9. Bacteriophages T2 and MS2, and poliovirus type
1 have isoelectric points of 4.2, 3.9 and 4.5, respectively (see
Table 1). Consequently, the studied pH 4 was above the isoelec-
tric point of MS2, only slightly below the isoelectric point of
T2 and significantly below the isoelectric point of poliovirus.
It is postulated that the adsorption of poliovirus onto magne-
tite was inhibited by lowering the pH of the suspending medium
sufficiently below the isoelectric point of the virus. Unfortu-
nately, no data was collected on the electrophoretic mobility
of magnetite particles in order to draw more definite conclusions.

Kinetics of the Adsorption Process

The kinetics of the adsorption process was investigated and
it was found that 10 to 20 minutes of shaking were sufficient to
allow an optimum removal of poliovirus (see Table 13). For the
phage MS2, equilibrium was reached with as little as 5 to 10 min-
utes of shaking (Table 13). On the other hand, the removal of
T2 phage increased with time in the range of shaking times studied
and did not reach an equilibrium value (see Table 13). From the
above results, it was decided to allow 20 minutes shaking time
for the adsorption experiments. This period of shaking was suf-
























Figure 4. Effect of pH on the removal of bacteriophage T2, bacteriophage MS2 and Poliovirus by
magnetic filtration

Bacteriophage T2 (5 x 103 pfu/ml), bacteriophage MS2 (5 x 103 pfu/ml) and
Poliovirus Type I (15 x 103 pfu/ml) were each shaken (\ 170 oscillations/
min) for 20 min in the presence of 500 ppm of magnetite, 1610 ppm of CaC12
and varying pH (adjusted with 0.1N NaOH and 0.1N HC1). These mixtures
were subsequently poured through the magnetic filter.
















100












LI

*- 90

I-
0


0
E







80


6 7 8 9







-33-


Table 13. Effect of time of shaking on the removal of bacterio-
phage T2, bacteriophage MS2 and Poliovirus Type I
(Sabin) by magnetic filtration



Time of Shakinga % Removal by Magnetic Filtrationb
(min) T2c MS2d Pol iovirusl


1 56.3 90.4 66.8

5 80.9 99.2 94.7

10 83.2 100.0 100.0

20 95.3 98.5 99.5

30 97.1 99.8 98.9


The

The
500

The

The

The


flasks were shaken at 170 oscillations/min.

magnetic filtration was undertaken in the presence of
ppm of magnetite and 1610 ppm of CaC12.

initial bacteriophage T2 concentration was 7 x 103 pfu/ml.

initial bacteriophage MS2 concentration was 4 x 103 pfu/ml.

initial Poliovirus concentration was 3 x 103 pfu/ml.







-34-


ficient for the optimum removal of poliovirus and MS2 phage but
apparently not for T2 phage. Nevertheless, a longer period of
shaking was not allowed for the phage T2 in order to avoid the
large inactivation of the virus which would occur when shaken for
periods longer than 20 minutes.

Effect of Magnetite Concentration

The concentration of magnetite, used as a seed material for
magnetic filtration, is an important factor to consider in light
of the economic feasibility of the process. Figure 5 shows that
300 ppm of magnetite was sufficient for the optimum removal of
the phage MS2 (6 x 103 pfu/ml) and poliovirus (8 x 103 pfu/ml)
and 500 ppm is sufficient for the phage T2 (4.5 x 103 pfu/ml).
Consequently, adsorption experiments were carried out in the
presence of 500 ppm of magnetite in order to insure optimal
conditions for adsorption and subsequent removal by magnetic
filtration.

Effect of Initial Virus Concentration

The Freundlich adsorption isotherm (Fair et al., 1968) has
often been used to show that removal of a solute (e.g., virus)
from an aqueous solution by solid media was an adsorption pro-
cess. This empirical relation is expressed as

y/m = kcl/n, (4)

where, for the specific case in which the adsorbate is a virus,
y/m is the quantity of virus removed per unit weight of adsor-
bent (e.g., magnetite) and c is the concentration of virus
remaining in solution at equilibrium. The constant k has been
described as a measure of the surface area of the solid phase
and the constant n as an indicator of the intensity of adsorp-
tion (Reece, 1967). The Freundlich equation is usually used in
the logarithmic form,

log (y/m) = log k + 1/n log c, (5)

which indicates a linear variation of log (y/m) with log c. Con-
sequently, a double-logarithmic plot of data conforming to the
Freundlich isotherm should give a line with slope 1/n and log k
as the y-intercept.

The interaction of viruses with such surfaces as activated
carbon, activated sludge, stabilization pond solids and soil
(Gerba et al., 1975; Clarke et al., 1961; Sobsey and Cooper, 1973;
Drewry and Eliassen, 1968) has been shown to obey the Freundlich
isotherm. This general tendency of virus-surface systems to con-
form to the Freundlich isotherm shows the importance of adsorption



















Figure 5. Effect of magnetite concentration on the removal of bacteriophage T2, bacteriophage MS2
and Poliovirus Type I (Sabin) by magnetic filtration

Bacteriophage T2 (4.5 x 103 pfu/ml), bacteriophage MS2 (6 x 103 pfu/ml) and
Poliovirus (8 x 103 pfu/ml) were each shaken (, 170 oscillations/min) for
20 min in the presence of 1610 ppm of CaC12 and various concentrations of
magnetite. These mixtures were subsequently poured through the magnetic
filter.



C1
c-f












100


.. 60 1
>



0
q--/


|f

E 40



T2 phageo------0

20( MS2 phage g
Pol iovirus"


0 I I I I I I I

200 400 600 800 1000

Magnetite concentration (ppm)







-37-


Figure 6. Freundlich isotherms for the adsorption of bacterio-
phage T2, bacteriophage MS2 and Poliovirus Type I
(Sabin) to magnetite: variable virus concentration

Bacteriophage T2, bacteriophage MS2 and
Poliovirus were each shaken (% 170 oscilla-
tions/min) for 20 min in the presence of
500 ppm of magnetite and 1610 ppm of CaC12.
These mixtures were subsequently poured
through the magnetic filter.

Linear regression analysis yielded the least
squares regression lines drawn:

Slope y-Intercept r2

T2 phage 1.035 1.354 0.99

MS2 phage 1.166 2.314 0.98


Poliovirus 0.736


2.743 0.99








-38-


6 1-


S(0)


4 k-


/
'*


-4-)

-4-.)



E





0

U)




~0
__


3 k-


T2 phageO----..

MPolS2 pihage 0

Pol iovirusO in0


II I I I
I 1 2 3 4 5


Loglo virus concentration in effluent (pfu/ml)


/
/
/


5 I-







-39-


in removing viruses from the water suspension. In order to
determine the nature of the removal process by magnetite, an
experiment was performed in which the T2 phage, MS2 phage and
poliovirus concentration was varied but the magnetite concen-
tration was held constant. Figure 6 shows that the results
obtained for the three viruses conform to the Freundlich iso-
therm. The Freundlich isotherm for the T2 phage was found to
be below the isotherms for poliovirus and the MS2 phage (Figure
6). This could be explained by the larger size of the T2 phage in
comparison to poliovirus and the phage MS2 (see Table 1). Due
to its larger size, less T2 phage can be adsorbed per mg of
magnetite when compared to a smaller virus such as poliovirus
or the MS2 phage. Also, notice that the y-intercept ("log k"
in the Freundlich equation) is less for the T2 phage (1.354)
than for poliovirus (2.743) or the phage MS2 (2.314). Since
the constant k in the Freundlich equation has been described
as a measure of the surface area of the adsorbent, then a
decrease in its value, as seen for the phage T2, may indicate
a reduction of the surface area available to viruses for adsorp-
tion onto magnetite. In this case, the surface area is not
decreased but the size of the virus is increased. The dif-
ference in the adsorption pattern may also be due to the fact
that T2 phage has a tail whereas MS2 phage and poliovirus are
tailless.

Effect of Wastewater Effluents on the Adsorption of Viruses to
Magnetite

Wastewater effluents contain organic materials which may
compete with viruses for adsorption onto solids (Amirhor and
Engelbrecht, 1974; Carlson et al., 1968; Dieterich, 1953; Gerba
et al., 1975) and may lead to a decreased removal of the infective
particles. Consequently, experiments were undertaken to deter-
mine if the adsorption of viruses to magnetite was influenced
by organic present in two wastewater effluents, "dome water"
and University of Florida campus activated sludge effluent (see
Table 2 for characteristics of these effluents). Table 14
shows that, in the absence of CaC12, dome water strongly inter-
feres with the removal of the phage T2 (0.0% removal) and that
this interference could be reduced (but not completely eliminated)
with the addition of CaC12. It can be seen in Table 15 that the
removal obtained in the presence of dome water and 1610 ppm of
CaC12 for the phage T2 can be increased by diluting the waste-
water effluent (and the organic present) in distilled water.
Removal of the organic color present in dome water with an acti-
vated carbon treatment greatly reduced the interference exerted
by this effluent (94.9% removal of T2 -- see Table 15). However,
the removal of T2 phage found in distilled water (98.3%) could
not be obtained in dome water regardless of previous treatment.
Apparently, the activated carbon did not remove all the interfer-
ing substances from the water. Table 16 shows clearly that dome






-40-


Table 14.


Effect of a divalent cation (Ca++) on the removal of
bacteriophage T2 by magnetic filtration in the pre-
sence of dome water


Ionic Strength Concentration % Removal of T2a
of CaCl2 (p) of CaC12 (ppm) by Magnetic Filtrationb
in the presence of Dome
Water

0 0 0.0

4.35 x 10-4 16.1 16.4

4.35 x 10-3 161 44.8

4.35 x 10-2 1610 76.0

4.35 x 10-1 16100 86.9


a The initial bacteriophage T2
pfu/ml.


concentration was 5.6 x 103


b The bacteriophage T2 was shaken (% 130 oscillations/min)
for 20 min in the presence of dome water, 500 ppm of magne-
tite and various concentrations of CaC12. These mixtures
were subsequently poured through the magnetic filter.

c The dome water was sampled at a cypress dome located north
of Gainesville, Florida.







-41-


Removal of bacteriophage T2 by magnetic filtration
in the presence of dome water


Description


% Removal of T2a by
Magnetic Filtrationb


Dome Waterc


Dome Water diluted 1:2
in Distilled Water

Dome Water diluted 1:4
in Distilled Water

Activated Carbon Filtered
Dome Water

Distilled Water


74.3

75.7


88.6


94.9


98.3


a The initial bacteriophage T2 concentration was 4 x 103 pfu/ml.

b The bacteriophage T2 was shaken (% 130 oscillations/min) for
20 min in the presence of 500 ppm of magnetite, 1610 ppm of
CaC12 and the various solutions described above. These mix-
tures were subsequently poured through the magnetic filter.

c The dome water was sampled at a cypress dome located north
of Gainesville, Florida.


Table 15.







-42-


Removal of Poliovirus Type I (Sabin) by magnetic
filtration in the presence of dome water


Description


% Removal of Poliovirusa
by Magnetic Filtrationb


Dome Waterc

Dome Water + 1610 ppm CaC12

Distilled Waterd

Distilled Water + 1610 ppm
CaCl2

Activated Carbon Filtered
Dome Water

Activated Carbon Filtered
Dome Water + 1610 ppm CaC12


a The initial Poliovirus concentration was 15


x 103 pfu/ml.


b The Poliovirus was shaken (n 200 oscillations/min) for 20
min in the presence of 500 ppm of magnetite and the various
solutions described above. These mixtures were subsequently
poured through the magnetic filter.

c The dome water was sampled at a cypress dome located north of
Gainesville, Florida.

d The distilled water was adjusted to the same conductivity as
the dome water (conductivity equal to 385.0 pmho/cm at 250C)
with CaC12.


Table 16.


96.3

99.4

99.6


99.2


99.1







-43-


water also interferes with the removal of poliovirus. This
interference was significantly reduced by adding 1610 ppm
of CaC12 (96.3% removal) or by removing the organic color
with activated carbon treatment (99.2% removal). The interfer-
ence of dome water with the adsorption of viruses on magnetite
is probably due to the fulvic acid fraction released in the
water by the decomposition of cypress needles and other leaves
within the cypress dome. It is generally known that fulvic
acids are soluble in water whereas humic acids are not although
some low molecular fractions may be (Flaig, 1960; Prakash and
Rashid, 1968). It has been found that dome water also inter-
feres with the adsorption of poliovirus on soil and may be
efficient for the elution of soil-adsorbed viruses (Bitton,
et al., 1976).

The other effluent used in this study was the University
of Florida campus activated sludge effluent. The interference
exerted by the organic present in this effluent was slight.
When this effluent was used as the suspending medium for polio-
virus, the removal was 96.7% in the absence of any salt and was
enhanced to 99.4% with the addition of 1610 ppm of CaC12 (Table
17). This removal was similar to the one observed in distilled
water adjusted to the conductivity of the activated sludge
effluent (385 pmho/cm at 250C) or containing 1610 ppm of CaCl2
(Table 17).

It is possible to compare the two wastewater effluents, dome
water and campus activated sludge effluent, for their inter-
ference with the adsorption of poliovirus onto magnetite (Tables
16 and 17). The interference was mostly apparent with the dome
water and was completely eliminated by filtering the dome water
through an activated carbon column. In the presence of CaC12,
the campus activated sludge effluent displayed a better virus
removal (99.4%) than the dome water (96.3%). The main difference
between the two types of effluents (see Table 2) was the color
content. The dome water had a color ranging from 350.0 to 750.0
units whereas the campus activated sludge effluent had a color
ranging from 30.0 to 35.0 units. Therefore, it could be postu-
lated that the organic color was responsible for the observed
interference. It should be added that dome water is not a typical
wastewater effluent due to its high organic color. Consequently,
the results of the campus activated sludge effluent should be
considered when assessing the virus removing potential of magnetic
filtration in the presence of wastewater effluents.

Additional experiments were carried out in which the virus
concentration was held constant and the magnetite concentration
was varied. The data in Figures 7, 8 and 9 shows that, in the
presence or the absence of secondary wastewater effluents, the
adsorption of bacteriophage T2, bacteriophage MS2 and poliovirus







-44-


Removal of Poliovirus Type I (Sabin) by magnetic
filtration in the presence of activated sludge effluent


Description


% Removal of Poliovirusa
by Magnetic Filtrationb


Activated Sludge Effluentc

Activated Sludge Effluent
+ 1610 ppm CaCl2

Distilled Waterd

Distilled Water + 1610 ppm
CaC12


96.7

99.4


99.4

99.6


a The initial Poliovirus concentration was 8 x 103 pfu/ml.

b The Poliovirus was shaken (% 200 oscillations/min) for 20
min in the presence of 500 ppm of magnetite and the various
solutions described above. These mixtures were subsequently
poured through the magnetic filter.

c The activated sludge effluent was sampled at the University
of Florida campus sewage treatment plant.

d The distilled water was adjusted to the same conductivity as
the activated sludge effluent (conductivity equal to 385.0
vimho/cm at 250C) with CaC12.


Table 17.

























Figure 7. Freundlich isotherms for the adsorption of bacteriophage T2 to magnetite in distilled
water and dome water: variable magnetite concentration

Bacteriophage T2 (4.5 x 103 pfu/ml) was shaken (r 130 oscillations/min) for 20 min in
the presence of 1610 ppm of CaC12, various concentrations of magnetite (ranging from
100 ppm to 1000 ppm) and distilled water or dome water. These mixtures were subse-
quently poured through the magnetic filter.

Linear regression analysis yielded the least squares regression lines drawn:


slope y-intercept r2

Distilled Water 0.595 2.889 0.86
Dome Water 0.718 1.167 0.99




















C0.



(E3

40 1
- 9

E -.
S4- -






C:



ro
0.
Pl-
s- 3
C)L
S--- Distilled Water *------4
-Q -
o Dome Water 0- 0


.J
>11



0 1 2 3 4

Log10 bacteriophage T2 concentration in effluent (pfu/ml)
























Figure 8. Freundlich isotherms for the adsorption of bacteriophage MS2 to magnetite in distilled
water and activated sludge effluent: variable magnetite concentration

Bacteriophage MS2 (6.3 x 103 pfu/ml) was shaken (, 130 oscillations/min) for 20 min in
the presence of 1610 ppm of CaC12, various concentrations of magnetite (ranging from
100 ppm to 1000 ppm) and distilled water or activated sludge effluent. These mixtures
were subsequently poured through the magnetic filter.

Linear regression analysis yielded the least squares regression lines drawn:


slope y-intercept r2

Distilled Water 0.769 3.082 0.99
Activated Sludge Effluent 1.021 1.272 0.95
























(0
F--
4-
0)

CY)
E
5-
CLJ

(1)
.0


V)

CL)
.0




0
-1


Distilled Water ---

Activated Sludge Effluent ----


0 1-/-1L


Log10 bacteriophage MS2 concentration in effluent (pfu/ml)


4L


'000
Ile


[o]
























Figure 9. Freundlich isotherms for the adsorption of Poliovirus Type I (Sabin) to magnetite in
distilled water and wastewater effluents: variable magnetite concentration

Poliovirus (8 x 103 pfu/ml) was shaken (n 200 oscillations/min) for 20 min in the
presence of 1610 ppm of CaC12, various concentrations of magnetite (ranging from 100
ppm to 1000 ppm) and distilled water, dome water or activated sludge effluent. These
mixtures were subsequently poured through the magnetic filter.

Linear regression analysis yielded the least squares regression lines drawn:


slope y-intercept r2

Distilled Water 0.715 2.897 0.98
Activated Sludge Effluent 1.395 1.011 0.94
Dome Water 2.590 -3.313 0.99











[*]


[*]


[o]


[o]


Distilled Water 0-----*
Activated Sludge Effluent 0 0
Dome Water 0- O
I I
2 3
Logo0 Poliovirus concentration in effluent (pfu/ml)


4 _-







-51-


onto magnetite conformed to the Freundlich isotherm. However,
in the presence of wastewater effluents, the Freundlich iso-
therms had a lower y-intercept ("log k" in the Freundlich iso-
therm equation) than that found in distilled water. For example,
in Figure 7, for the phage T2, the y-intercept of the Freundlich
isotherm was as high as 2.889 in the absence of organic materials
and decreased to 1.167 in the presence of dome water. For the
phage MS2 (see Figure 8), the y-intercept was 3.082 in distilled
water and dropped to 1.272 in the presence of activated sludge
effluent. Figure 9 shows that, for poliovirus, the y-intercept
was as high as 2.897 in distilled water and decreased to 1.011
in the presence of campus activated sludge effluent and -3.313
in the presence of dome water. This consistent drop in the
value of the y-intercept ("log k") when the viruses are suspended
in wastewater effluents can be explained in the following manner.
The constant k in the Freundlich isotherm equation has been des-
cribed as a measure of the surface area of the solid phase (Reece,
1967) and therefore, a decrease in its value may indicate a
reduction in the surface area of magnetite available to viruses
for adsorption. This reduction is due to the occupation of
adsorption sites on the magnetite by competing organic present
in wastewater.

Effect of Organics on the Adsorption of Poliovirus to Magnetite

A variety of organic substances such as bovine albumin,
egg albumin meat infusion broth and cell extracts have been
shown to compete with viruses for adsorption onto solids (Carl-
son et al., 1968; Shepard and Woodend, 1951; Wallis and Melnick,
1967a). In order to determine if organic would also interfere
with the adsorption of poliovirus to magnetite, an experiment
was undertaken using egg albumin, casein and dextran. Table 18
shows that 100 ppm of these materials significantly interfered
with the removal of poliovirus. Casein is known to be a highly
efficient eluent of virus from cellulose membranes (Wallis and
Melnick, 1967b) and therefore, its complete interference with
the removal of poliovirus (0.0%) is not surprising (Table 18).
Moreover, it should be noted that 100 ppm of these organic
materials is an extremely high concentration. For example,
Carlson et al. (1968) showed that as little as 2ppm of egg
albumin reduced the adsorption of the phage T2 to the clay
Kaolinite 4 from 93% to 27%. It can be concluded that a
reduction in the removal of viruses by magnetic filtration in
the presence of high concentrations of an interfering substance
is to be expected.

Infectivity of Viruses Adsorbed to Magnetite

The previous sections have dealt with the factors influenc-
ing the adsorption of viruses to magnetite. It was learned that







-52-


Interference by egg albumin, casein and dextran with the
removal of Poliovirus Type I (Sabin) by magnetic fil-
tration


Description


% Removal of Poliovirusa
by Magnetic Filtrationb


100 ppm Egg Albumin 42.4

100 ppm Casein 0.0

100 ppm Dextran (5 x 105 MW) 7.1

Control 99.6


a The initial Poliovirus concentration was 1.0 x 104 pfu/ml.

b The Poliovirus was shaken (% 200 oscillations/min) for 20
min in the presence of 500 ppm of magnetite, 1610 ppm of
CaC12 and the various solutions described above. These
mixtures were subsequently poured through the magnetic filter.


Table 18.







-53-


under appropriate conditions, viruses can be efficiently ad-
sorbed to magnetite and subsequently removed by magnetic fil-
tration. Experiments were then performed to determine the
infectivity of the adsorbed virus. We used low concentrations
of viruses so that samples might be directly plated and there-
by, avoid dilutions which could lead to the desorption of
viruses. Table 19 shows that T2 phage (97.5%), MS2 phage
(100.0%) and poliovirus (98.9%) were effectively adsorbed to
magnetite. The infectivity of adsorbed poliovirus and MS2
phage was >100.0%. The elevated titer (>100.0%) of the adsorbed
virus could have resulted from virus disaggregation when shaken
for 20 minutes. Schaub and Sagik (1975) have also observed
that enteric viruses adsorbed to either Montmorillonite clay or
other naturally occurring solids displayed an elevated titer.
These authors proposed that the clay particles may function in
allowing the virus to establish better proximity to the cells.
Table 19 also shows that the infectivity of adsorbed T2 phage
was lower than that displayed by poliovirus or-MS2 phage and
much more variable (57.8% and 93.5% infective). Puck and Sagik
(1953) have shown that T2 phage, after its attachment to a
cationic resin, was split into its DNA and protein fractions
and therefore, could not be recovered in active form. It
appears that upon adsorbing to this surface, the T2 phage ejac-
ulates its DNA much as it does ordinarily at the surface of its
host cell. It is postulated that the reduced infectivity of T2
phage may have resulted from the ejaculation of its DNA upon
adsorbing to magnetite. Furthermore, the variability found in
infectivity may be due to the way in which the phage is adsorbed.
If the phage is adsorbed by the tail to the magnetite, then
ejaculation and inactivation may occur. However, if adsorption
occurs on the phage head, then inactivation does not occur.
Finally, it can be said that viruses adsorbed to magnetite are
infective. Moore et al. (1975) have studied the association of
poliovirus and phages T2, T7 and f2 with clays and suspended
solids. They found that all the viruses tested, with the excep-
tion of f2 phage, were infective in the adsorbed state. In our
study, we found that the tailless RNA phage, MS2, displayed a
similar behavior as poliovirus type 1 and these two viruses were
100% infective in the adsorbed state. In view of these findings,
it should be clear that the safe treatment and disposal of virus-
laden magnetite is critical in controlling the spread of disease.

Elution of Poliovirus from Magnetite

Prior to initiating experiments on the concentration of
poliovirus by magnetic filtration, it was important to deter-
mine which solution would be best for the elution of poliovirus
from magnetite. Table 20 shows the eluents used and the percent












Table 19.


Infectivity of bacteriophage T2, bacteriophage MS2 and Poliovirus Type I (Sabin)
adsorbed to magnetite


Virusa Direct Plate Direct Plate % Infectivityb Supernatant % Virus
@ T = 0 min @ T = 20 min of Adsorbed (pfu/ml) Adsorbedc
(pfu/ml) (pfu/ml) Virus

T2 phage 1190 700 57.8 30 97.5
790 740 93.5 20 97.5
MS2 phage 717 800 >100.0 0 100.0

Poliovirus 1200 1335 >100.0 13 98.9


a The bacteriophage T2, bacteriophage MS2 and Poliovirus were each shaken (% 170 oscillations/
min) for 20 min in the presence of 500 ppm of magnetite and 1610 ppm of CaC12. Samples were
removed and assayed by direct plating (no dilutions made) immediately after addition of the
virus (T = 0 min) and after the adsorption period (T = 20 min). Following the adsorption
period, the magnetite was pelleted by placing the mixtures next to a magnet and then the
supernatants were assayed for virus. Controls without magnetite were also studied and they
showed no significant inactivation of the three viruses during the 20 minutes of shaking.
b The % infectivity of adsorbed virus was determined using the following equation:

% Infectivity = Direct Plate @ T = 20 (pfu/ml) Supernatant (pfu/ml)
Direct Plate @ T = 0 (pfu/ml) Supernatant (pfu/ml)
c The % virus adsorbed is based on the percentage of the direct plate at T = 0 min.







-55-


Table 20. Elution of Poliovirus Type I (Sabin) from the magne-
tite surface


Eluentb Used % Recoveryc of
Adsorbed Poliovirusa


Distilled Water 7.1

1610 ppm CaC12 1.8

Tris Buffer, pH = 9 9.9

3%, isotonic Beef Extract, 3.2
isotonic buffered, pH = 9

10%, isotonic Fetal Calf Serum, 77.9A
isotonic buffered, pH = 9

Eagles Minimal Essential Medium (MEM) + 59.1B
5% Calf Serum + 0.03 M Hepes Buffer, pH = 7.0

Glycine Buffer, 0.05 M, pH = 11.3 39.5C

Dome Water,d Tris buffered, pH = 9 8.8


a The Poliovirus was shaken (% 200 oscillations/min) for 20 min in
the presence of 500 ppm of magnetite and 1610 ppm of CaC12.
Aliquots were then removed and the magnetite was pelleted by
placing the mixture next to a magnet. Supernatants were assayed
and showed that 99% of the initial Poliovirus concentration (1.0
x 104 pfu/ml) was adsorbed to magnetite.

b Elution was undertaken as follows: the magnetite pellets were
resuspended in the indicated eluents for 1 hour at 40C (except
for glycine buffer which was allowed only 1 min of contact time),
were then pelleted again and finally, the supernatants were
assayed for viruses.

c The % recovery values displaying different superscript capital
letters are significantly different at the 0.01 level for a two-
tailed test. If no superscript capital letter appears, then a
test of significance was not performed.

d The dome water was sampled at a cypress dome located north of
Gainesville, Florida.







-56-


recoveries obtained. The composition and source of the eluents
used appear in the Appendix. The 1610 ppm CaC12 solution was the
medium in which adsorption initially took place and as expected,
the recovery of adsorbed poliovirus using this solution as eluent
was very low (1.8%). The best eluent was an isotonic, 10% solution
of fetal calf serum, Tris buffered, pH = 9 and it enabled the
desorption of 77.9% of poliovirus from the magnetite surface
(Table 20). These results are not surprising since Rao et al.
(1968) also used a similar solution to effectively elute entero-
viruses from an iron oxide. However, the low recovery obtained
with an isotonic, 3% solution of beef extract, Tris buffered at
pH 9 (3.2% recovery) was not expected. This solution has been
shown to be an efficient eluent when desorbing viruses from an
iron oxide surface (Pearson and Metcalf, 1974; Rao et al., 1968).
Rao et al. (1968) has stated that not all lots of beef extract
have the same virus eluting capacity. Thus,it appears that the
lot of beef extract used in this study probably had a low virus
eluting capacity. Generally, the best eluents are proteinaceous
materials adjusted to an alkaline pH (8-9). In many cases, good
eluents also effectively interfere with the initial adsorption of
viruses to surfaces. Dome water, which had previously been shown
to interfere with the adsorption of poliovirus to magnetite, was
studied to determine its efficiency in eluting poliovirus from
magnetite. Table 20 shows that dome water, Tris buffered, pH =
9 could only elute 8.8% of adsorbed poliovirus. Therefore, in
this case, a good interfering substance (dome water) was shown
not to be a good eluent. Experiments with glycine buffer, 0.05M,
pH = 11.3 were performed allowing only 1 minute of contact time
(see Table 20). This was done in order to avoid the inactivation
of the virus due to the high pH. In fact, when 1 hour of contact
time was allowed, no infectivity could be detected in the eluate
as compared to 39.5% recovery of adsorbed poliovirus when the
contact time was 1 minute. From the results presented, it was
decided to use the 10%, isotonic fetal calf serum, Tris buffered,
pH = 9 solution as the eluent in the concentration experiment
described in the next section.

Concentration of Poliovirus by Magnetic Filtration

Viruses occur in natural water bodies in very low concentra-
tions and therefore, their detection is dependent upon their
adequate concentration. Various virus adsorbents have been
used to effectively concentrate viruses and these have been
discussed in the literature review section. Since magnetite
can adsorb, under appropriate conditions, over 99% of
poliovirus, we have carried out a preliminary experiment to
investigate the application of magnetic filtration to the con-
centration of viruses. Table 21 shows the results obtained
when poliovirus was concentrated using this technique. We
studied the recovery of viruses under various conditions and it












Table 21. Concentration of Poliovirus Type I (Sabin) by magnetic filtration


Experiment Initiala Virus Volume of Filtrate Virus % Recoveryc Concentration
No. Conc. (pfu/ml) Sample Conc. (pfu/ml) Factor

1 1.0 x 103 4 liters 0 1st pass: 0.6 40
2nd pass: 4.8
3rd pass: 6.5

2 1.3 x 103 4 liters 17 1 min: 11.7 80
overnight: 50.7

3 1.2 x 103 4 liters 0 >100.0 40


a Recovery of Poliovirus from 4 liters of distilled water containing the initial virus con-
centrations shown was undertaken as described below. Poliovirus was shaken (% 200 oscilla-
tions/min) for 20 min in the presence of 1000 ppm of magnetite and 1610 ppm of CaC12. The
mixtures were subsequently poured through the magnetic filter in order to separate the
Poliovirus-magnetite complex from the water suspension. Poliovirus was then eluted from
the magnetite surface with 100 ml (or 50 ml) of 10% isotonic fetal calf serum, Tris buffered,
pH = 9. The procedure for eluting the Poliovirus varied with each experiment and appears
in the Materials and Methods section.


The headings 1st pass, 2nd pass, 3rd pass, 1 min and overnight refer to
assayed for viruses. In experiment 1, the eluate was assayed after the
pass through the filter. In experiment 2, the eluate was assayed after
for 1 min and after shaking mechanically overnight at 40C.

c The % recovery values were obtained by using the following equation:


% Recovery =


when the eluate was
1st, 2nd, and 3rd
shaking manually


(7)


Eluate Virus Conc. (pfu/ml) 1
[Initial Virus Conce (pfu/ml) Filtrate Virus Conc. (pfu/ml)] Conc. Factor







-58-


was found that the number of recovered virus particles from 4
liters of water increased with the degree of contact between
the eluent used and the magnetite-virus complex. For example,
in experiment 1 (see Table 21), recovery increased with the
number of passes of the eluent through the filter. Since the
flow rate through the filter was very high, there was not enough
time for effective contact between eluent and virus. Consequently,
the highest recovery achieved was 6.5%. In experiment 2 (Table
21), by adding the eluent into the filter and shaking, the
contact was increased beyond that of experiment 1 and the re-
coveries achieved increased accordingly to 11.7% and 50.7% for
1 minute shaking and overnight shaking, respectively. Sonica-
tion has been used effectively to desorb microorganisms, includ-
ing viruses, from surfaces (Puleo et al., 1967; Tschider et al.,
1974). In experiment 3 (Table 21), sonication of the matrix
afforded enough contact between eluent and virus such that a
recovery of >100.0% was obtained. The elevated titer (>100.0%)
could have resulted from disaggregation of the virus as explained
before or simply from plating error. The concentration factors
in the experiments described above were 40 and 80 and the initial
virus concentrations ranged from 1.0 x 103 to 1.3 x 103 pfu/ml.
In order to fully assess the effectiveness of magnetic filtra-
tion in concentrating viruses, the concentration of lower initial
virus titers (e.g., <5 pfu/ml) in natural waters should be at-
tempted. Finally, it can be said that magnetic filtration
appears promising as a concentrating technique. It is an easy
procedure and since it does not require a long time to perform,
inactivation of virus does not occur.

CONCLUSIONS

Based on the findings of this investigation, the following
conclusions may be drawn:

1. The adsorption of bacteriophage T2, bacteriophage MS2
and poliovirus type 1 (Sabin) to magnetite, and the subsequent
removal by magnetic filtration increased in the presence of
cations. Furthermore, on a concentration or molar basis, tri-
valent cations were more efficient than divalent cations which
in turn were more efficient than monovalent ones.

2. The adsorption of the three viruses to magnetite remained
constant in the pH range of 5 to 9.

3. Removal by magnetic filtration in the 98-99% level was
achieved with as little as 300 ppm of magnetite for poliovirus
and MS2 phage, and 500 ppm of magnetite for T2 phage.

4. The adsorption of the three viruses to magnetite conformed
to the Freundlich adsorption isotherm. However, the isotherm for







-59-


the T2 phage was below the isotherms for poliovirus and MS2
phage. This indicates that less T2 phage is adsorbed per mg of
magnetite due to the larger size of this virus in comparison to
poliovirus or MS2 phage.

5. Interaction between magnetite and virus was hindered in
a highly colored wastewater effluent. However, this interfer-
ence was reduced by the addition of CaC12, by the dilution of
the wastewater with distilled water or by removing the organic
color with activated carbon.

6. The interference exerted by an activated sludge effluent
on the removal of poliovirus by magnetic filtration was slight.
This interference was completely eliminated by the addition of
1610 ppm of CaC12 (99.4% removal). This effluent is more typical
than the highly colored effluent described above if one considers
the effect of wastewater effluents upon the removal of viruses by
magnetic filtration.

7. Competition for adsorption was also exerted by 100 ppm
of egg albumin, casein and dextran resulting in a significant
reduction in the removal of poliovirus. This finding is not
unexpected since 100 ppm of these organic materials is an
extremely high concentration.

8. Poliovirus and MS2 phage, adsorbed to magnetite, were
100% infective. The infectivity of adsorbed T2 was lower than
that of poliovirus or MS2 phage and much more variable. The
reduced infectivity of T2 phage may result from the ejaculation
of its DNA upon adsorbing to magnetite.

9. From the results of the infectivity of adsorbed virus,
the Freundlich isotherms and other experiments, it is clear
that MS2 phage is a better model of poliovirus removal by
magnetic filtration than T2 phage. Consequently, it is recom-
mended that future research in this area, involving only a
bacteriophage, should be performed with an RNA phage such as
MS2.

10. Poliovirus was effectively eluted from the magnetite
surface with an isotonic, 10% solution of fetal calf serum, Tris
buffered, pH = 9 (77.9% recovery of adsorbed virus).

11. A 40-fold concentration of poliovirus was achieved using
magnetic filtration. Recovery of the virus (initial concentration
equal to 1.2 x 103 pfu/ml) from 4 liters of distilled water was
>100.0% when sonication of the matrix, in the presence of the
eluent described above, was undertaken. More research is needed
on the concentration of lower virus titers in natural waters by
this technique.







-60-




The experimental data obtained in this work show that magne-
tite is a good adsorbent towards viruses and that magnetic
filtration can be effectively used for the removal of viruses
from water and wastewater. Magnetic filtration also appears
promising as a concentrating technique.







-61-




ACKNOWLEDGMENTS

The authors wish to thank Dr. William H. Morgan, Director
of the Florida Water Resources Research Center, for his help
and patience during the course of this investigation. We are
also indebted to Michael Duke for his valuable help in this
work.







-62-


REFERENCES


Adams, M. H. 1959. Bacteriophages. Interscience Publishers,
Inc., New York, N.Y.

Amirhor, P. and Engelbrecht, R. S. 1974. Interaction of virus
and polyelectrolyte in wastewater effluent. 7th Int. Conf. Water
Poll. Res., Paris.

Bachrach, H. L. and Schwerdt, C. E. 1952. Purification studies
on Lansing poliomyelitis virus: pH stability, CNS extraction
and butanol purification experiments. J. Immun., 69:551-561.

Bartell, P., Pierzchala, W. and Tint, H. 1960. The adsorption
of enterioviruses by activated attapulgite. J. Am. Pharm. Assoc.
Sci. Ed., 49:1-4.

Berg, G. 1973a. Removal of viruses from sewage, effluents and
waters. 1. A review. Bull. Wld. Hlth. Org., 49:451-460.

Berg, G. 1973b. Removal of viruses from sewage, effluents and
waters. 2. Present and future trends. Bull. Wld. Hlth. Org.,
49:461-469.

Berg, G. 1975. Microbiology-detection, occurrence, and removal
of viruses. J. Wat. Pollut. Control Fed., 47:1587-1595.

Bitton, G. 1975. Adsorption of viruses onto surfaces in soil and
water. Water Res., 9:473-484.

Bitton, G., Fox, J. L. and Strickland, H. C. 1975. Removal of
algae from Florida lakes by magnetic filtration. Appl. Microbiol.,
30:905-908.

Bitton, G., Masterson, N. and Gifford, G. E. 1976. Effect of a
secondary treated effluent on the movement of viruses through a
cypress dome soil. J. Env. Qual. (in press).

Bitton, G. and Mitchell, R. 1974. The removal of Escherichia coli
bacteriophage T7 by magnetic filtration. Water Res., 8:549-551.

Bitton, G., Mitchell, R., De Latour, C. and Maxwell, E. 1974. Phos-
phate removal by magnetic filtration. Water Res., 8:107-109.

Boche, R. D. and Quilligan, J. J., Jr. 1966. Adsorption to glass
and specific antibody inhibition of 125I-labeled influenza virus.
J. Immun., 97:942-950.







-63-


Brown, T. S., Malina, J. F., Jr. and Moore, B. D. 1974a. Virus
removal by diatomaceous-earth filtration part 1. J. Am. Wat.
Wks. Assoc., 66:98-102.

Brown, T. S., Malina, J. F., Jr. and Moore, B. D. 1974b. Virus
removal by diatomaceous-earth filtration part 2. J. Am. Wat.
Wks. Assoc., 66:735-738.

Carlson, G. F., Jr., Woodard, F. E., Wentworth, D. F. and Sproul,
0. J. 1968. Virus inactivation on clay particles in natural waters.
J. Wat. Pollut. Control Fed., 40:R-89-106.

Chaudhuri, M. and Engelbrecht, R. S. 1972. Removal of viruses
from water by chemical coagulation and flocculation. J. Am. Wat.
Wks. Assoc., 62:563-567.

Clark, J. W., Viessman, W., Jr. and Hammer, M. J. 1971. Water
Supply and Pollution Control, 2nd ed., International Textbook
Company, N.Y.

Clarke, N. A., Stevenson, R. E., Chang, S. L. and Kabler, P. W.
1961. Removal of enteric viruses from sewage by activated
sludge treatment. Am. J. Publ. Hlth., 51:1118-1129.

Cookson, J. T., Jr. 1967. Adsorption of viruses on activated
carbon. Adsorption of Escherichia coli bacteriophage T4 on
activated carbon as a diffusion-limited process. Environ. Sci.
Tech., 1:157-160.

Cookson, J. T., Jr. 1969. Mechanism of virus adsorption on acti-
vated carbon. J. Am. Wat. Wks. Assoc., 61:52-56.

Cookson, J. T., Jr. and North, W. J. 1967. Adsorption of viruses
on activated carbon. Equilibria and kinetics of the attachment
of Escherichia coli bacteriophage T4 on activated carbon. Environ.
Sci. Tech., 1:46-52.

Davis, B. D., Dulbecco, R., Eisen, H. N., Ginsberg, H. S. and Wood,
W. B. 1973. Microbiology. Harper and Row, Hagerstown, Maryland.

De Latour, C. 1973. Magnetic separation in water pollution control.
IEEE Transactions on Magnetics, MAG-9:314-316.

De Latour, C. and Kolm, H. 1976. High gradient magnetic separation:
A water treatment alternative. J. Am. Wat. Wks. Assoc., 68:325-327.

Dieterich, B. H. 1953. The behavior of bacteria virus in contact
with ordinary and uniform filter sand. Master's thesis, Harvard
Univ., Cambridge, Mass.







-64-


Drewry, W. A. and Eliassen, R. 1968. Virus movement in ground-
water. J. Wat. Pollut. Control Fed., 40:R257-271.

Fair, G. M., Geyer, J. C. and Okun, D. A. 1968. Water and Waste-
water Engineering, Vol. 2. Water purification and wastewater
treatment and disposal. John Wiley and Sons, Inc. New York.

Flaig, W. 1960. Chemie der humusstoffe. Acta Chem. Fenn. A,
33:229-251.

Gerba, C. P., Sobsey, M. D., Wallis, C. and Melnick, J. L. 1975.
Adsorption of poliovirus onto activated carbon in wastewater.
Environ. Sci. Tech., 9:727-731.

Gilcreas, F. W. and Kelly, S. M. 1955. Relation of coliform-
organism test to enteric virus pollution. J. Am. Wat. Wks.
Assoc., 47:683-694.

Grabow, W. 0. K. 1968. The virology of wastewater treatment.
Water Res., 2:675-701.

Hill, W. F., Jr., Akin, E. W. and Benton, W. H. 1971. Detection
of viruses in water: A review of methods and application.
Water Res., 5:967-995.

Hill, W. F., Jr., Akin, E. W., Benton, W. H., Mayhew, C. J. and
Jakubowski, W. 1974. Apparatus for conditioning unlimited quan-
tities of finished waters for enteric virus detection. Appl.
Microbiol., 27:1177-1178.

Kelland, D. R. 1973. High gradient magnetic separation applied
to mineral beneficiation. IEEE Transactions on Magnetics, MAG-9:
307-310.

Kellenberger, E. 1962. Vegetative bacteriophage and the matura-
tion of the virus particles. Adv. Virus Res., 8:1.

Lauffer, M. A. and Bendet, I. J. 1962. Comments on "Bio-physical
properties of bacteriophage T2." Biochim. Biophys. Acta, 55:211-
214.

Lefler, E. and Kott, Y. 1974. Virus retention and survival in
sand. In Virus Survival in Water and Wastewater Systems, edited
by Malina, J. F., Jr. and Sagik, B. P., Center for Research in
Water Resources, University of Texas, Austin.

LoGrippo, G. A. 1950. Partial purification of viruses with an
anion exchange resin. Proc. Soc. Exp. Biol. Med., 74:208-211.







-65-


Malherbe, H. H. and Strickland-Cholmley, M. 1967. Survival of
viruses in the presence of algae. In Transmission of Viruses by
the Water Route, edited by Berg, G., Interscience, New York.

Mandel, B. 1971. Characterization of type 1 poliovirus by electro-
phoretic analysis. Virology, 44:554-568.

Mitchell, R., Bitton, G, De Latour, C. and Maxwell, E. 1974. Magnetic
separation: A new approach to water and waste treatment. 7th Int.
Conf. Water Poll. Res., Paris.

Mitchell, R., Bitton, G. and Oberteuffer, J. A. 1975a. High gra-
dient magnetic filtration of magnetic and non-magnetic contaminants
from water. Separation and Purification Methods, 4(2):267-303.

Mitchell, R., Chet, I., Bitton, G., Oberteuffer, J. and Harland,
J. R. 1975b. High gradient magnetic separation of microorganisms
and other non-magnetic suspended solids from waste water. 48th
Annual Conf. Water Poll. Cont. Fed., Miami Beach, Florida.

Moore, B. E., Sagik, B. P. and Malina, J. F., Jr. 1975. Viral
association with suspended solids. Water Res., 9:197-203.

Muller, R. H. and Rose, H. M. 1952. Concentration of influenza
virus (PR8 strain) by a cation exchange resin. Proc. Soc. Exp.
Biol. Med., 80:27-29.

Oberteuffer, J. A. 1973. High gradient magnetic separation.
IEEE Transactions on Magnetics, MAG-9:303-306.

Oder, R. R. 1973. In High Gradient Magnetic Separation Symposium,
edited by Oberteuffer, J. A. and Kelland, D. R., M.I.T., Cambridge,
Mass.

Osipow, L. I. 1962. Surface Chemistry. Robert E. Krieger Publish-
ing Company, Huntington, N. Y.

Overby, L. R., Barlow, G. H., Doi, R. H., Jacob, M. and Spiegelman,
S. 1966. Comparison of two serologically distinct ribonucleic
acid bacteriophages. J. Bacteriol., 91:442-448.

Oza, P. P. and Chaudhuri, M. 1975. Removal of viruses from water
by sorption on coal. Water Res., 9:707-712.

Oza, P. P., Sriramulu, N. and Chaudhuri, M. 1973. Preliminary
investigation on the use of coal for removing viruses from water.
Proc. Symp. on Environmental Pollution, edited by Saraf, R. K.,
CPHERI, Nagpur, India.







-66-


Pearson, F. and Metcalf, T. G. 1974. The use of magnetic iron
oxide for recovery of virus from water. Completion report, Project
No. A-014-NH, under Grant No. 14-31-0001-3029, Water Resources
Research Center, Univ. of New Hampshire, Durham.

Pollard, E. C. 1953. The Physics of Viruses. Academic Press, New
York.

Prakash, A. and Rashid, M. A. 1968. Influence of humic substances
on the growth of marine phytoplankton: Dinoflagellates. Limnol.
Oceanog., 13:598-606.

Puck, T. T. and Sagik, B. 1953. Virus and cell interaction with
ion exchangers. J. Exp. Med., 97:807-820.

Puleo, J. R., Favero, M. S. and Tritz, G. J. 1967. Feasibility
of using ultrasonics for removing viable microorganisms from
surfaces. Contamination Control Journal, 6:58-67.

Putnam, F. W. 1953. Bacteriophages: nature and reproduction.
Adv. Protein Chem., 8:175-284.

Rao, V. C., Sullivan, R., Read, R. B. and Clarke, N. A. 1968.
A simple method for concentrating and detecting viruses in water.
J. Am. Wat. Wks. Assoc., 60:1288-1294.

Reece, R. E. 1967. Virus sorption on natural soils. Master's
thesis, University of Arkansas, Fayetteville.

Robeck, G. G., Clarke, N. A. and Dostal, K. A. 1962. Effectiveness
of water treatment processes in virus removal. J. Am. Wat. Wks.
Assoc., 54:1275-1290.

Schaub, S. A. and Sagik, B. P. 1975. Association of enterioviruses
with natural and artificially introduced colloidal solids in water
and infectivity of solids-associated virions. Appl. Microbiol.,
30:212-222.

Schaub, S. A., Sorber, C. A. and Taylor, G. W. 1974. The associa-
tion of enteric viruses with natural turbidity in the aquatic
environment. In Virus Survival in Water and Wastewater Systems,
edited by Malina, J. F., Jr. and Sagik, B. P., Center for Research
in Water Resources, University of Texas, Austin.

Schwerdt, C. E. and Schaffer, F. L. 1955. Some physical and chemical
properties of purified poliomyelitis virus preparations. Annals
N.Y. Acad. Sci., 61:740-753.

Sharp, D. G., Hook, A. E.,Taylor, A. R., Beard, D. and Beard, J. W.
1946. Sedimentation characters and pH stability of the T2 bacterio-
phage of Escherichia coli. J. Biol. Chem., 165:259-270.







-67-


Shepard, C. C. and Woodend, W. G. 1951. The adsorptive
behavior of bacteriophage T2 with certain inorganic materials.
J. Immun., 66:385-393.

Shuval, H. I., Cymbalista, S., Fatal, B. and Goldblum, N. 1967.
Concentration of enteric viruses in water by hydro-extraction and
two-phase separation. In Transmission of Viruses by the Water
Route, edited by Berg, G., Interscience, New York.

Sobsey, M. D. and Cooper, R. C. 1973. Enteric virus survival
in algal-bacterial wastewater treatment systems-I-Laboratory
studies. Water Res., 7:669-685.

Sproul, 0. J. 1972. Virus inactivation by water treatment.
J. Am. Wat. Wks. Assoc., 64:31-35.

Sproul, 0. J., Warner, M., LaRochelle, L. R. and Brunner, D. R.
1969. Virus removal by adsorption in wastewater treatment
processes. Adv. Water Pollut. Res., 14th Int. Conf. Water Pollut.
Res., Prague.

Taylor, N. W., Epstein, H. T. and Lauffer, M. A. 1955. The
particle weight, hydration and shape of the T2 bacteriophage of
Escherichia coli. J. Am. Chem. Soc., 77:1270-1273.

Thorup, R. T., Nixon, F. P., Wentworth, D. F. and Sproul, 0. J.
1970. Virus removal by coagulation with polyelectrolytes. J.
Am. Wat. Wks. Assoc., 62:97-101.

Tolmach, L. J. 1957. Attachment and penetration of cells by
viruses. Adv. Virus Res., 4:631-110.

Trindade, S. C. and Kolm, H. H. 1973. Magnetic desulfurization
of coal. IEEE Transactions on Magnetics, MAG-9:310-313.

Tschider, S. R., Berryhill, D. L. and Schipper, I. A. 1974.
Membrane concentration of infectious bovine rhinotracheitis virus
from water. Appl. Microbiol., 28:1030-1032.

Valentine, R. C. and Allison, A. C. 1959. Virus particle adsorp-
tion. I. Theory of adsorption and experiments on the attachment
of particles to non-biological surfaces. Biochim. Biophys. Acta,
34:10-23.

Wallis, C. and Melnick, J. L. 1967a. Concentration of entero-
viruses on membrane filters. J. Virology, 1:472-477.

Wallis, C. and Melnick, J. L. 1967b. Concentration of viruses
from sewage by adsorption on Millipore membranes. Bull. Wld. Hlth.
Org., 36:219-225.







-68-


Wallis, C. and Melnick, J. L. 1967c. Concentration of viruses
on aluminum and calcium salts. Amer. J. Epidemiol., 85:459-468.

Wallis, C. and Melnick, J. L. 1967d. Concentration of viruses
on aluminum phosphate and aluminum hydroxide precipitates. In
Transmission of Viruses by the Water Route, edited by Berg, G.,
Interscience, New York.

Wallis, C., Melnick, J. L. and Fields, J. E. 1970. Detection
of viruses in large volumes of natural waters by concentration
on insoluble polyelectrolytes. Water Res., 4:787-796.

Wallis, C., Melnick, J. L. and Fields, J. E. 1971. Concentration
and purification of viruses by adsorption to and elution from
insoluble polyelectrolytes. Appl. Microbiol., 21:703-709.

Warren, J., Neal, A. and Rennels, D. 1966. Adsorption of myxo-
viruses on magnetic iron oxides. Proc. Soc. Exp. Biol. Med.,
12:1250-1253.

Williams, R. C. and Fraser, D. 1953. Morphology of the seven
T-bacteriophages. J. Bacteriol., 66:458-464.







-69-


APPENDIX

COMPOSITION OF MEDIA AND SOLUTIONS USED

Media Used in T2 and T4r(II) Phage Work

All of these media were described by Adams (1959).

1. Basal medium for phage plating:
8 grams nutrient broth, dehydrated (Difco)
5 grams NaCl (Fisher)
Make up to 1 liter with distilled water.

2. Nutrient agar for plates:
Solution 1 above plus 15 grams per liter of
Bacto-Agar (Difco). Between 25 and 30 ml are
poured per plate and the plates are dried in
an incubator at 370C overnight before use.

3. Soft agar overlay:
Solution 1 above plus 7 grams per liter of
Bacto-Agar (Difco). This solution is distri-
buted in 2.5 ml amounts to test tubes. Prior
to assaying, the soft agar is melted in a
boiling water bath and then held in a 460C
water bath.

4. M-9 medium for storing phages:

Solution I: 3 grams KH2PO4 (Fisher)
6 grams Na2HPO4, anhydrous (Fisher)
1 gram NH4Cl (Fisher)
Make up to 900 ml with distilled water.

Solution II: 4 grams glucose (Fisher as dextrose)
100 ml of distilled water.

Solution III. 1.3 grams MgSO4 (Fisher)
100 ml of distilled water.
Autoclave each solution separately. Prior to use,
add sterilely9 parts of solution I to one part of
Solution II and 0.1 part of Solution III.

5. Agar slants for maintaining the host Escherichia coli
B:
2 grams nutrient agar (Difco)
100 ml distilled water.

Distribute 5 ml into 20 test tubes and autoclave.
Allow to harden slanted after autoclaving.

All media described above were kept refrigerated at 40C and
usually used within 2 days of preparing.







-70-


Media Used in MS2 Phage Work

The L medium, L agar and L soft agar were described by
Overby et al. (1966).

1. L medium:
10 grams tryptone (Difco)
1 gram glucose (Fisher as dextrose)
5 grams yeast extract (Difco)
10 grams NaCI (Fisher)
0.22 grams CaC12 (Fisher)
Make up to 1 liter with distilled water.

2. L agar for plates:
L medium above plus 15 grams per liter of
Bacto-Agar (Difco). Between 25 and 30 ml are
poured per plate and the plates are dried in
an incubator at 370C overnight before use.

3. L soft agar overlay:

L medium above plus 10 grams per liter of
Bacto-Agar (Difco). This solution is distributed
in 2.5 ml amounts to test tubes. Prior to assaying,
the soft agar is melted in a boiling water bath
and then held in a 460C water bath.

4. Tris buffer (0.05 M Tris, 0.1 M NaCl, pH 7.6) for
suspending (storing) phages as recommended by Miles
Laboratories:
1.39 grams Trizma base (Sigma)
6.06 grams Trizma HCI (Sigma)
5.85 grams NaCl (Fisher)
Make up to 1 liter with distilled water. The resulting
pH is 7.6 at 250C.

The host Escherichia coli C3000 was maintained on 2% nutrient
agar (Difco) slants as described for E. coli B. All media described
above were kept refrigerated at 40C and usually used within 2 days
of preparing.

Media Used in Poliovirus Type I (Sabin) Work

1. Gey's Balanced Salt Solution (BSS) is the common diluent for
cell culture:
Gey's A (10x) : 70 grams NaCl
3.7 grams KC1
3.01 grams Na2HPO4 12H20
0.237 grams KH2PO4
100 ml 0.1% phenol red
10 grams glucose
900 ml glass distilled water
5 ml chloroform as a preservative







-71-


This stock solution of Gey's
temperature unautoclaved and
autoclaved when needed.


Gey's B (20x):




Gey's C (20x):


A is stored at room
is diluted 1:10 and


0.42 grams MgC12 6H20
0.14 grams MgSO4 7H20
0.34 grams CaCl2
100 ml glass distilled water

2.25 grams NaHC03
100 ml glass distilled water
Bubble CO2 into Gey's C until pH
is less than 7.6. Dispense and
tightly cap.


Gey's B and C are autoclaved without further dilution.

To make the complete Gey's Balanced Salt Solution (BSS) add:
90 parts Gey's A (lx)
5 parts Gey's B (20x)
5 parts Gey's C (20x)


2. Hepes buffer (1 M) stock solution:
47.7 grams Hepes
190 ml Gey's A (Ix)
10 ml Gey's B (20x)
16 ml 2 M NaOH (8g/100 ml)


Dispense and autoclave.

3. Streptomycin-penicillin (1000x) stock solution:

Solution I: 1.0 gram streptomycin
8 ml Gey's A (lx)

Solution II: 106 units of penicillin
4 ml of Solution I.

Solution II contains 125 mg of streptomycin and 2.5 x 105
units of penicillin per ml which is 1000x of what is
required. Therefore, it must be diluted 1:1000 in the
final solution.

4. Eagle's Minimal Essential Medium (MEM) using Gey's BSS
+ 10% fetal calf serum:
300 ml Gey's A (lx)
20 ml Gey's B (20x)
20 ml Gey's C (20x)
8 ml MEM essential amino acids (50x) (International
Scientific, Cary, Ill.)







-72-


4 ml vitamins (100x) (International Sci.)
4 ml glutamine (100x) (International Sci.)
0.4 ml strep.-pen. stock (lO00x)
40 ml fetal calf serum (International Sci.)

5. Eagle's MEM using Gey's BSS + 5% calf serum + 0.03 M
Hepes buffer at pH = 7: This solution was used to make
most virus dilutions.

The solution is made by substituting, in Solution 4
above, 20 ml of calf serum (International Scientific)
+ 12 ml of 1 M Hepes buffer stock solution + 8 ml of
Gey's A (lx) for 40 ml of fetal calf serum.

6. Solutions required for the removal of AV3 cells from
glass (trypsinization):


Solution I:


Solution II:


Pre-trypsin wash:
This solution removes all traces of serum
(which contains trypsin inhibitors) as well
as Ca++ and Mg++ ions.
300 ml Gey's A (lx)
5 ml Gey's C (20x)

Dispense and autoclave.

1% versene (EDTA) stock in Gey's A:
2.0 grams ethylene diamine tetraacetate (EDTA)
10 ml 2 M NaOH (8g/100 ml)
20 ml Gey's A (10x)
170 ml glass distilled water


Dispense and autoclave.


Solution III:


Solution IV:


2.5% trypsin stock:
1.0 gram trypsin (Difco, 1:250)
100 ml glass distilled water


Sterilize by cold filtration. Dispense in 5
ml aliquots to screw-capped test tubes and
freeze for storage.

Standard versene-trypsin solution:
This solution is used to remove the AV3
cells from the 32 ounce bottles in which
they have been growing prior to their dis-
tribution to plaque bottles.







-73-


100 ml Gey's A (lx)
4 ml Gey's C (20x)
4 ml stock trypsin (2.5%)
4 ml stock versene (1% in Gey's A)

This solution is good for only one day.

7. Methyl cellulose overlay for AV3 cells (1% methyl cellulose
+ 5% fetal calf serum):

Solution I: 300 ml glass distilled water
6 grams methyl cellulose (1500 centipoise)

Autoclave and then allow to cool to room temperature,
shaking vigorously every hour to avoid layering.
Refrigerate.

Solution II: 2x Eagle's MEM:
350 ml glass distilled water
120 ml Eagle's MEM (10x) with Hanks' salts
(International Scientific)
50 ml Gey's C (20x)
60 ml fetal calf serum (International Sci.)
25 ml Hepes buffer (1M) stock solution
12 ml glutamine (100x) (International Sci.)
1.2 ml strep.-pen. stock (1000x)

Combine equal amounts of Solution I and II to make the
methyl cellulose overlay.

8. Crystal violet:

This stain is used to make the plaques on the cell mono-
layer visible to the naked eye.

Solution I: 20 grams crystal violet
200 ml absolute ethanol

Allow this solution to sit overnight.

Solution II: 8 grams ammonium oxalate
800 ml distilled water

Mix Solution I and II and dilute 1:10 with tap water.

Solutions Used for the Elution of Poliovirus from Magnetite

The eluents used are listed below along with their composition.

1. Distilled water (Carlson et al., 1968; Puck and Sagik,
1953)







-74-


2. 1610 ppm CaC12

3. Tris buffer, pH = 9

0.076 g Trizma HC1 (Sigma, St. Louis, Mo.)
0.547 g Trizma base (Sigma)
100 ml distilled water

4. 3%, isotonic beef extract, Tris buffered, pH = 9
(Pearson and Metcalf, 1974; Rao et al., 1968)

0.076 g Trizma HC1 (Sigma)
0.547 g Trizma base (Sigma)
0.87 g NaCl (Fisher)
3.0 g beef extract (Difco)
100 ml distilled water

5. Eagle's Minimal Essential Medium (MEM) + 5% calf serum
+ 0.03 M Hepes buffer, pH = 7.0 (Pearson and Metcalf,
1974)

The composition of this solution can be found in this
appendix in the section dealing with the media used
in poliovirus work.

6. 10%, isotonic fetal calf serum, Tris buffered, pH = 9
(Rao et al., 1968; Wallis and Melnick, 1967a, 1967b;
Wallis et al., 1971; Wellings et al., 1974)

0.076 g Trizma HCI (Sigma)
0.547 g Trizma base (Sigma)
0.87 g NaCl (Fisher)
10 ml fetal calf serum (International Scientific, Inc.)
90 ml distilled water

7. Glycine buffer, 0.05 M, pH = 11.3 (Hill et al., 1974)

Solution I: 0.375 g glycine (Fisher)
100 ml distilled water

Solution II: 10 N NaOH prepared:
40 g NaOH (Fisher)
100 ml distilled water

Within 2 hours of using, add 0.54 ml of Solution II to
100 ml of Solution I.


8. Dome water, Tris buffered, pH = 9







-75-




0.076 g Trizma HC1 (Sigma)
0.547 g Trizma base (Sigma)
100 ml dome water (see composition in Table 2)

All the solution above were sterilized by autoclaving. Serum,
when required, was added sterilely after autoclaving.




Full Text

PAGE 2

Publication No. 40 Removal of Viruses From Water by Magnetic Filtration By Gabriel Bitton, George E. Gifford and Oscar C. Pancorbo Department of Environmental Engineering Sciences University of Florida Gainesville

PAGE 3

Removal of Viruses From Water by Magnetic Filtration By Gabriel Bitton, George E. Gifford and Oscar C. Pancorbo PUBLICATION NO. 40 FLORIDA WATER RESOURCES RESEARCH CENTER RESEARCH PROJECT TECHNICAL COMPLETION REPORT OWRT Project Number A-030-FLA Annual Allotment Agreement Numbers 14-31-0001-5009 14-34-0001-6010 Report Submitted: October, 1976 The work upon wh ich th is report is based was supported in part by funds provided by the United States Department of the I nterior, Office of Water Research and TechnologyI' as Authorized under the Water Resources Research Act of 1964 as amended.

PAGE 4

TABLE OF CONTENTS LIST OF TABLES .................................................. LIST OF FIGURES ................................................. ABSTRACT ........................................................ PUBLICATIONS .................................................... THESIS .......................................................... INTRODUCTION .................................................... REVIEW OF LITERATURE ............................................ Adsorption of Viruses to Surfaces .......................... Implication of Sorptive Phenomena in the Removal of Viruses. Use of Adsorbents in the Concentration of Viruses .......... The Process of High Gradient Magnetic Separation ........... MATERIALS AND METHODS ........................................... Viruses Used and Their Assays .............................. Magnetic Filtration of Adsorbed Viruses .................... Wastewater Effluents Used in Adsorption Experiments ........ Organics Used in Adsorption Experiments .................... Infectivity of Virus 'Adsorbed to Magnetite ................. Elution of Adsorbed Poliovirus from Magnetite .............. Concentration of Poliovirus by Magnetic Filtration ......... Statistical Treatment of Data .............................. RESULTS AND DISCUSSION .......................................... Effect of Cation Valency and Concentration ................. Effect of pH ............................................... Kinetics of the Adsorption Process ......................... Effect of Magnetite Concentration .......................... Effect of Initial Virus Concentration ...................... Effect of Wastewater Effluents on the Adsorption of Viruses to Magnetite ............................................... Effect of Organics on the Adsorption of Poliovirus to Magne-ti te ....................................................... Infectivity of Virus Adsorbed to Magnetite ................. Elution of Poliovirus from Magnetite ....................... Concentration of Poliovirus by Magnetic Filtration ......... i iii v vi vii vii 1 1 1 2 4 5 7 7 9 13 15 15 15 15 16 19 19 30 30 34 34 39 51 51 53 56

PAGE 5

TABLE OF CONTENTS --Continued Page CONCLUSIONS. .. . 58 ACKNOWLEDGEMENTS. . . . . . . . 61 REFERENCES. . . 62 APPENDIX. . . 69 ii

PAGE 6

Table 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 LIST OF TABLES Properties of viruses used in this study ............... Characteristics of the secondary effluents used in the magnetic filtration of viruses ......................... Effect of monovalent cations (Na+ and K+) on the removal of bacteriophage T2 by magnetic filtration ............. Effect of a monovalent cation (Na+) on the removal of Poliovirus Type I (Sabin) by magnetic filtration ....... Effect of divalent cations (Ca++ and Mg++) on the removal of bacteriophage T2 by magnetic filtration ..... Effect of divalent cations (Ca++ and Mg++) on the removal of bacteriophage T 4R(II) by magnetic filtration. Effect of divalent cations (Ca++ and Mg++) on the removal of bacteriophage MS2 by magnetic filtration ..... Effect of a divalent cation (Ca++) on the removal of Poliovirus Type I (Sabin) by magnetic filtration ....... Effect of a trivalent cation (Al+++) on the removal of bacteriophage T2 by magnetic filtration ................ Effect of a trivalent cation (Al+++) on the removal of bacteriophage T4R(II) by magnetic filtration ........... Effect of a trivalent cation (Al+++) on the removal of bacteriophage MS2 by magnetic filtration ............... Effect of a trivalent cation (Al+++) on the removal of Poliovirus Type I (Sabin) by magnetic filtration ....... Effect of time of shaking on the removal of bacteriophage T2, bacteriophage MS2 and Poliovirus Type I (Sabin) by magnetic fil tration ......................... Effect'iof:'a divalent cation (Ca++) on the removal of bacteriophage T2 by magnetic filtration in the presence of dome water .......................................... Removal of bacteriophage T2 by magnetic filtration in the presence of dome water ............................. iii 8 14 20 21 22 23 24 25 26 27 28 29 33 40 41

PAGE 7

Table 16 17 18 19 20 21 LIST OF TABLES --Continued Removal of Poliovirus Type I (Sabin) by magnetic filtra-tion in the presence of dome water ....................... Removal of Poliovirus Type I (Sabin) by magnetic filtra-in the presence of activated sludge effluent ............. Interference by egg albumin, casein and dextran with the removal of Poliovirus Type I (Sabin) by magnetic filtra-tion ..................................................... Infectivity of bacteriophage T2 bacteriophage MS2 and Poliovirus Type I (Sabin) absorbed to magnetite .......... Elution of Poliovirus Type I (Sabin) from the magnetite surface .................................................. Concentration of Poliovirus Type I (Sabin) by magnetic fil tration ............................................... iv 42 44 52 54 55 57

PAGE 8

LIST OF FIGURES Figure 1 Particle size distribution of magnetite (Fisher) as measured with the Coulter Counter Industrial Model B (Coul-ter Electronics, Illinois)............................... 10 2 High gradient magnetic filter used to separate the virus-magnetite complex from the water suspension .............. 11 3 High gradient magnetic filter supplied by Sala Magnetics, Cambridge, Mass.......................................... 17 4 Effect of pH on the removal of bacteriophage T2, bacterio-phage MS2 and Poliovirus by magnetic filtration .......... 31 5 Effect of magnetite concentration on the removal of bacteriophage T2, bacteriophage MS2 and Poliovirus Type I (Sabin) by magnetic filtration........................... 35 6 Freundlich isotherms for the adsorption of bacteriophage T 2 bacteriophage MS2 and Poliovirus Type I (Sabin) to magnetite: variable virus concentration ................. 37 7 Freundlich isotherms for the adsorption of bacteriophage T2 to magnetite in distilled water and dome water: variable magnetite concentration ............................. 45 8 Freundlich isotherms for the adsorption of bacteriophage MS2 to magnetite in distilled water and activated sludge effluent: variable magnetite concentration .............. 47 9 Freundlich isotherms for the adsorption of Poliovirus Type I (Sabin) to magnetite in distilled water and waste-water effluents: variable magnetite concentration ....... 49 v

PAGE 9

ABSTRACT The process of magnetic filtration was applied to the removal of T2 bacteriophage, MS2 bacteriophage, and poliovirus type 1 (Sabin) from water and wastewater. Since the effectiveness of this process is governed by the adsorption of the viruses onto the magnetic iron oxide seed, magnetite, those factors influencing adsorption were investigated. It was found that the adsorption of the three viruses studied to magnetite at a concentration of 500 ppm increased in the pre-sence of cations, was not affected when the pH was varied from 5 to 9, and could be described by the Freundlich adsorption isotherm. Viruses were suspended in two wastewater effluents and subsequently removed by magnetic filtration. It was shown that organic color interfered with the adsorption process. The interference was reduced by the addition of CaC12, by the dilution of the wastewater with distilled water or by removing the organic color with activated carbon. Competition for adsorption was also exerted by 100 ppm of egg albumin, casein and dextran resulting in a significant reduction in the removal of poliovirus. Additionally, the infectivity of the magnetite-bound viruses was studied. Poliovirus, adsorbed to magnetite, was 100% infective whereas T2 phage displayed a lower and somewhat variable infectivity. Attempts were made to desorb poliovirus from magnetite by resuspension of the magnetite pellet in a variety of possible eluents. The best eluent was an isotonic, 10% solution of fetal calf serum buffered at the pH of 9 with Tris buffer. Preliminary results of the concentration of poliovirus by magnetic filtration are also presented and discussed. The experimental data obtained in this work show that magnetite is a good adsorbent towards viruses and that magnetic filtration can be effectively used for the removal of viruses from water and wastewater. vi

PAGE 10

PUBLICATIONS 1. G. Bitton, O. Pancorbo and G. E. Gifford. 1976. Adsorption of Poliovirus Type 1 to Magnetite and Subsequent Removal by Magnetic Filtration. Abstract, Amer. Soc. for Microbiology Ann. Meeting, Atlantic City, N.J. 2. G. Bitton, J. L. Fox and H. G. Strickland. 1975. Algae from Florida Lakes by Magnetic Filtration. 30:905-908. Removal of Microbiol. 3. G. Bitton, O. Pancorbo and G. E. 1976. Factors Affecting the Adsorption of Poliovirus to Magnetite in Water and Waste wate'r". Water Res." Vol. !Q. (in press) 4. G. Bitton, J. L. Fox, G. E. Gifford and O. Pancorbo. 1976. Utilisation d'electroaimants dans l'elimination des virus et des algues presents dans les eaux de surface et les eaux usees. Annales de l' ACM'" 43: 133. Pancorbo, O. C. 1976. wastewater by magnetic Dept. of Environmental Florida, Gainesville. THESIS The removal of viruses from water and filtration. 115 pp. M.S. thesis, Engineering Sciences, University of vii

PAGE 11

INTRODUCTION Viruses are obligatory intracellular parasites ranging in size from 100 A to 4000 A in diameter (Pollard, 1953). They contain a single type of nucleic acid, either DNA or RNA, surrounded by a protein coat. The ionization of certain protein coat groups results in viruses acquiring a negative charge at neutral pH. The interaction of viruses with their environment is determined by the nature of their protein coat and is influenced by their ability to adsorb to surfaces. Bitton (1975) reviewed the adsorption of viruses to surfaces and its importance in the removal of viruses by water treatment processes and in the concentration and recovery of viruses. Sproul et al. (1969) proposed that probably the most important mechanism for virus "inactivation" in waste water treatment plants is adsorption. Among the various adsorbents studied, iron oxides';have been shown,to display excellent sorptive capacity towards viruses (Bitton and Mitchell, 1974; Pearson and Metcalf, 1974; Rao et al., 1968; Warren et al., 1966). Using magnetite as an adsorbent-,-Bitton and Mi tchell(f974) removed the bacteriophage T7 by a "magnetic fil tration technique.'" This process consists of adsorbing the virus onto the magnetic iron oxide seed, magnetite, and subsequently removing the virus-seed mixture from the water by pouring through a filter placed in a background magnetic field. The purpose of this study was to delineate further the tiveness of magnetic filtration towards the removal of viruses, mainly those of animal origin, from water. Factors known to influence the adsorption of viruses onto surfaces and that could, therefore, affect the adsorption of viruses onto magnetite were investigated. These included pH, electrolyte valency and concentration, virus type and concentration, time allowed for adsorption, and concentration and type of organic substances in the suspending medium. It was felt that the knowledge gained from this study could be applicable in the removal of pathogenic enteric viruses present in water and wastewater. REVIEW OF LITERATURE AdsorpfionofViruses to Surfaces Viruses are able to adsorb to biological (Tolmach, 1957) and nonbiological surfaces (Bitton, 1975). The adsorption process is usually influenced by such factors as the type of virus, nature of the surface involved, pH, and electrolyte, and aTganic matter content of the suspending medium. Due to their widespread occurrence in soil and aquatic environments and to their large surface area, clay minerals have been studied for their sorptive capacity towards viruses. The adsorption of entero-

PAGE 12

-2-viruses to such clays as kaolinite, illite, montmorillonite, and bentonite has been shown to be significant and to increase in the presence of cations with trivalent cations being more effective than divalent or monovalent cations (Carlson et al., 1968; Schaub and Sagik, 1975; Schaub et al., 1974). Virus-Sorption onto clays has also been found to be-independent of pH over the range of 3 to 9 (Bartell et al., 1960; Schaub et al., 1974) and to conform to the Freundlich:rsotherm (Schaub et al., 1974). Organic materials such as egg and bovine albumin, calf serum were also shown to interfere significantly with the adsorption of viruses to clays (Carlson et al., 1968; Schaub and Sagik, 1975). Variability in the adsorptive capacity of different clays towards viruses was proposed by Carlson et al. (1968) to result from differences in clay surface exchange-capacity which is determined by the surface charge density and clay particle geometry. Activated carbon is extensively used in water and wastewater treatment and consequently, its ability to adsorb viruses has been investigated. Cookson (1969) studied the adsorption of T4 bacteriophage on activated carbon and determined that the adsorption rate was influenced by the pH and the ionic strength of the medium. It was proposed that the adsorption process involved an electrostatic interaction between amino groups on the virus and carboxyl groups on the activated carbon. At very high ionic strengths as well as at low pH's, the tail fibers were unavailable for adsorption (Lauffer and Bendet, 1962) and this decreased the rate of reaction. The adsorption was described as a diffusion-limited process (Cookson, 1967), conformed to the Langmuir isotherm and did not inactivate the virus (Cookson and North, 1967). Gerba et al. (1975) found that the adsorption of poliovirus type 1 (straiU-LSc) to activated carbon obeyed also the Freundlich isotherm and was reduced in the presence of wastewater effluents. DesQrption of viruses from an activated charcoal column occurred when the wastewater pH was increased or when glycine buffer adjusted to pH 11.5 was used to elute the virus. The adsorption of viruses onto such surfaces as glass, celite, nitrocellulose, aluminum oxide and gold has been reported and was found also to be influenced by the presence of cations and organic materials (Bache and Quilligan, 1966; Shepard and Woodend, 1951; Valentine and Allison" 1959). Implication of Sorptive Phenomena in the Removal of Viruses Pathogenic enteric viruses mainly from human and animal origin are constantly present in our water systems and constitute a serious public health hazard. Consequently, a great deal of research has been devo1ted to finding efficient water treatments for their elimination. The removal of viruses by water (Berg, 1973a; Berg, 1973b; Berg, 1975; Sproul, 1972) and wastewater (Berg, 1973a; Berg, 1973b;

PAGE 13

-3-Berg, 1975; Grabow, 1968) treatment processes usually involves storage, biological and tertiary treatment followed by disinfection with chlorine, iodine, bromine or ozone. Among the various secondary (biological) wastewater treatment processes, the activated sludge system has been the most intensively studied for the removal of viruses. Clarke et al. (1961) found that the adsorption of coxsackie A9 virus and poliovirus type 1 (Mahoney) to activated sludge conformed to the Freundlich isotherm and resulted in the removal of about 90% of the viruses. The virus-sludge complex was shown to be very stable since only a small percentage of adsorbed viruses could be recovered. However, the removal or inactivation of viruses in stabilization ponds has received little attention. Data from laboratory experiments did not show evidence of active removal of poliovirus and reovirus by algae, but there was a significant inactivation of viruses by sunlight at the surface of sewage maturation ponds (Malherbe and Strickland-Cholmley, 1967). Sobsey and Cooper (1973) observed that the reduction of poliovirus in stabilization pond water was due to adsorption to solids. Tertiary wastewater treatment processes are designed to further polish the quality of secondary effluents with regards to suspended solids and nutrients containing nitrogen and phosphorus. Among the various tertiary wastewater treatment processes, activated carbon has been shown to remove viruses very poorly. The factors influencing the adsorption of viruses to activated carbon have already been discussed. Sproul et al. (1969) reported that virus removal from wastewater by activated carbon is not dependable. Coagulation and flocculation using cationic polyelectrolytes as prime coagulants or as coagulant aids in the presence of alum: has also been studied and was found to be effective in the removal of viruses (Amirhor and Engelbrecht, 1974; Chaudhuri and Engelbrecht, 1972; Thorup et al., 1970). This process was shown to be salt dependent (Thorup-et-al., 1970) and unfortunately, organic matter interfered with moval (Amirhor and Engelbrecht, 1974; Chaudhuri and Engelbrecht, 1972). Even though sand filtration plays a role as a tertiary treatment process, its standard and more important use in potable water treatment is the primary reason for the intensive investigation of its role in virus removal. Dieterich (1953) found that adsorption was the primary mechanism responsible for the removal of a bacteriophage during sand filtration and that egg albumin significantly interfered with virus removal. Cations, on the other hand, were reported to enhance the retention of viruses by sand columns (Lefler and Kott, 1974). Although sand is a poor adsorbent due to its small surface area (Dieterich, 1953), significant virus removal can be achieved by combining sand filtration with alum flocculation (Gilcreas and Kelly, 1955; Robeck 1962). The removal of viruses from water by diatomaceous-earth filtration (Brown et al., 1974a; Brown et al., 1974b) and by the use of coal assorbent (Oza and Chaudhuri-,-1975; Oza et al., 1973) was also investigated, and the results were encou-

PAGE 14

-4-raging. However, more research on the effectiveness of these two treatments is in order. Use of Adsorbents in the Concentration of Viruses Viruses occur in natural water bodies in very low concentrations and therefore, their detection is dependent upon their adequate concentration. Various virus adsorbents have been investigated for their potential application in the concentration of viruses from dilute suspensions (Hill et al., 1971). The membrane-adsorption technique has been effectively concentrate viruses. Wallis and Melnick (1967a) showed that the adsorption of viruses on membrane filters was enhanced in the presence of cations and inhibited in the presence of proteinaceous matter such as cell extracts or serum. The optimal pH for adsorption was found to be 5. This method was further used in the concentration of viruses from water (Hill et al., 1974) and.:wastewater (Wallis and Melnick, 1967b). Syntheticinsoluble polyelectrolytes (Wallis et al., 1970, 1971), cationic exchange resins (Muller and Rose, 1952;-Fuck and Sagik, 1953) and anionic exchange resins (LoGrippo, 1950; Puck and Sagik, 1953) have also been found to be effective virus adsorbents and consequently, to possess the ability to concentrate and purify viruses. Finally, precipitates of aluminum hydroxide, aluminum phosphate, and calcium phosphate have been used effectively to concentrate viruses (Wallis and Melnick, 1967c, 1967d). Viruses may also be concentrated by a variety of techniques which do not involve adsorption onto surfaces (Hill et al., 1971; Shuval et al., 1967) but these methods will not be discussed in this review-.---Among the various adsorbents used for virus concentration and removal, iron oxides have received little attention. Warren et al. (1966) reported for the first time the adsorption of influenza-virus to an iron oxide, hematite. Concentration and purification of this virus were achieved by adsorption onto hematite and subsequent elution with a 10% sodium phosphate solution at a pH between 7.5 and 8.5 resulting in 10-fold concentrates. Rao et al. (1968) later used an iron oxide (M.O. 2530) column to adsorb andconcentrate a strain of coxsackie virus A9. It was found that concentrations as high as 2 x 10 6 pfu/ml of this virus could be removed, without detecting any virus in the filtrate, by 25 grams of the iron oxide. Concentration of the coxsackie virus A9 by adsorption to the iron oxide and subsequent elution with foetal calf serum resulted in 5-fold concentrates with recoveries ranging from 55% to 95%, depending on the initial virus concentration. Pearson and Metcalf (l974) studied the adsorption of several enteroviruses to the same iron oxide, M.O. 2530, and found virus recovery most effective when elutions were made from thin layers of iron oxide under alkaline pH and in the presence of 3% isotonic beef extract. Unfortunately, the use of columns or thin layers of iron oxide may ultimately result in the clogging of the

PAGE 15

-5-filter (Rao et al., 1968). Due to the open matrix of the filter used in Magnetic Separation (HGMS), the problem of clogging is greatly minimized. The Process of High Gradient Magnetic Separation (HGMS) Convential'magnetic separators have been used for many years to remove strongly magnetic iron-bearing particles larger than 100 microns in size from non-magnetic media. These devices are used to remove magnetic impurities in a variety of feeds, and to concentrate magnetic materials such as iron ores for their beneficiation (Ober teuffer, 1973). Recently, a great deal of research has been devoted to the development of "High Gradient Magnetic Separation." This process, unlike convential magnetic separation, is able to separate weakly paramagnetic materials of micron size by maximizing the magnetic force on such particles. The magnetic force on a particle is given by the equation, dH F = VM-x dx (1) Where V is the volume of the particle, M is the magnetic moment of the particle in the field H, and is the field gradient in the x direction (Oberteuffer, 1973). In the case of paramagnetic materials, and M = XH dH Fx = V XH dx (2) (3) where X is the magnetic susceptibility. From equation 3, it can bedH seen that both a large magnetic field H and a large field gradient will yield a maximum magnetic force Fx. A large magnetic force is x required in order to overcome competing forces (e.g., gravitational) that oppose the magnetic separation of particles (Oberteuffer, 1973). A high gradient magnetic separator consists of a filter containing a matrix (usually stainless steel wool) placed in a strong magnetic field. Such a matrix provides a very large number of trapping sites for susceptible particles and enables high flow rates (50-150 gal. min.-l ft.-2) due to its loosely packed structure. With the magnetic field on, the magnetic particles in the feed slurry are trapped at the points and edges of the matrix fibers while nonmagnetic constituents of the slurry pass through the filter easily. Due to the low residual magnetization of the matrix fibers, magnetic particles trapped in the matrix during the feed. mode can be easily washed out by turning the magnetic field off and backflushing (Mitchell et al., 1975a, 1975b). Several industrial applications have been found for high gradient magnetic separation. In the clay industry, for example, this process is used to clean kaolin from micron size iron stained titaniferous

PAGE 16

-6-materials (Oder, 1973). Such impurities can seriously reduce the use of kaolin as a paper coating material. High gradient magnetic separation has also been used in the beneficiation of semitaconites which are made up of very small particles of weakly magnetic iron oxides intermixed with gangue material such as slate and chert. Kelland (1973) has used successfully the high gradient magnetic separation technique to separate the iron minerals from the gangue and this process resulted in an increase in the iron grade (iron content) of the semitaconites. This technique has also been applied to the desulfurization of coal. Trindade and Kolm (1973) obtained a 50 to 75% reduction in the sulfur content of Brazilian coal. High gradient magnetic separation has been shown to be effective in the removal of various pollutants from water and wastewater (DeLatour and Kolm, 1976; Mitchell et al., 1975a). Removal may be achieved in one of two ways depending upon the nature of the pollutant. For magnetic contaminants, such as the fine iron oxide particles found in steel mill effluents and boiler feed waters, high gradient magnetic separation may be used alone to obtain their removal. However, for the more common non-magnetic pollutants found in natural waters, a magnetic seeding technique is required. This technique involves the adsorption of non-magnetic contaminants onto a magnetic seed, magnetite, and subsequent passage of the mixture through a high gradient magnetic separator. The magnetite and adsorbed pollutants are trapped on the matrix of the separator and can be easily washed out when the magnetic field is turned off. This process has been called "magnetic filtration" and it has been shown to reduce effectively coliforms, color, turbidity (DeLatour, 1973; Mitchell et al., 1975a, 1975b), phosphate (Bitton et al., 1974; DeLatour, 1973)-,-algae in lake water (Bitton et al., 1975) and the phage T7 (Bitton and Mitchell, 1974). (1974) found that the removal of the bacteriophage T7 by this process was dependent on the presence of cations and independent of virus concentration in the range of 30 pfu/ml to 14 x 10 3 pfu/ml. Although having many similar properties, bacteriophages and animal viruses are, nevertheless, inherently different. Therefore, an animal virus study should be included when assessing the virus removing effectiveness of any treatment. Unfortunately, no research has previously been done on the removal of animal viruses by magnetic filtration. Additionally, little is known about the effect of such factors as pH and the presence of organic materials on the removal of viruses by this process. The research reported in this:report attempts to fill in these gaps in the knowledge of virus removal by magnetic filtration. Finally, it can be said that high gradient magnetic filtration is an efficient process capable of removing a wide variety of pollutants from wastewater. Consequently, its use either as a replacement for small sewage treatment facilities or in large facilities as an advanced treatment process is not far off (Mitchell et al., 1975a, 1975b). --

PAGE 17

-7-MATERIALS AND METHODS All 'virological work was performed aseptically and in the case of poliovirus, all work was done in a room previously sterilized by ultraviolet irradiation. All glassware which came into @vernight in Haemo-Sol (Scientific in tap water and distilled water. claving. Viruses Used and Their Assays contact with the virus was soaked Products) and subsequently rinsed Glassware was sterilized by auto-The viruses used in this study were the T2 bacteriophage, T4R(II) bacteriophage, MS2 bacteriophage and poliovirus type 1 (Sabin strain). Table 1 shows some important properties of these viruses. Bacteriophages were used in this study to obtain preliminary data on a particular phase of research and to resolve any difficulties in the experimental procedure prior to initiating experiments with poliovirus. Representatives from two different groups of bacteriophages were used: MS2 phage representing the group of small, RNA phages and T2 and T4R(II) phages representing the large, DNA phages possessing a contractile tail. These two types of bacteriophages were selected in an attempt to compare their behavior during magnetic filtration with that of poliovirus and to determine which is a better model of enterovirus removal by magnetic filtration. The bacteriophage MS2 should be a better model of poliovirus behavior than T2 and T 4R(II) bacteriophages due to its greater similarity to poliovirus in size, particle mass and nucleic acid type (Table 1). Preliminary investigations showed that T2 and T4R(II) bacterio-phages behaved similarly during magnetic filtration, and therefore, subsequent experimentation was only carried out with T2 phage. Polio virus type 1 (Sabin) was used as a representative human enteric virus. It is generally agreed that the use of an animal virus in a viral adsorption study increases the reliability of the results obtained. The T2 and T4R(II) bacteriophage stocks and their host Escherichia coli B were obtained from Dr. Donna H. Duckworth, Department of Immunology and Medical Microbiology, University of Florida. The phages were refrigerated at in NW-1) medium (Adams, 1959) and the host, E. coli B, was maintained at room temperature on 2% nutrient agar (Difco) slants. A complete list of all media and solutions used in this study including their composition and is presented in the Appendix. These T-even bacteriophages were assayed by the plague technique as described by Adams (1959) using the double layer plating procedure. The MS2 bacteriophage stock was obtained from Dr. Anthony Pfister, Department of Immunology and Medical Microbiology, University of Florida. The host Escherichia coli C3000 (ATCC #15597) was

PAGE 18

Table 1. Properties of viruses used in this study Virus Nuc1eic a Isoe1ectricb Sizec (mll) Dry Partic1ed pH Stabili tye Acid Point Mass Range (x 10-16 g) T2 phage DNN) 4.2 head -65 x 95 3.3 5.0 9.0' tail -25 x 100 T4 phage DNAb head 65 x 95 3.3 5.0 9.0 tail -25 x 100 MS2 phage RNAb 3.9 25 0.06 Poliovirus RNAt ) 4.5 and 7.0 27 0.14 3.6 -8.4 a From Davis et a1. 1973. b From: Matide1 1971; Overby et al. 1966; Sharp et a1. 1946. c From Overby et a1. 1966; Schwerdt and Schaffer, 1955; Williams and Fraser, 1953. d e From Kellenberger, 1962; Overby et a1. 1966; and Schaffer, 1955; Taylor et a1. 1955. Frmm Bachrach and Schwerdt, 1952; Putnam, 1953; Sharp et a1. 1946. I 00 I

PAGE 19

-9-obtained from Miles Laboratories, Kankakee, Illinois. The phage was refrigerated at 4C in Tris buffer (see Appendix) and the host, E. coli C3000, was maintained at room temperature on 2% nutrient agar (Difco) slants. Assay of this virus was by the plaque technique using the double layer procedure (Adams, 1959). The Poliovirus type 1 (Sabin strain) stock suspension used in this study was prepared by infecting a monolayer culture of AV3 cells in a 32 ounce bottle. After a 1 hour adsorption period with tilting at 15 minute intervals, 40 m1 of Eagle's Minimal Essential Medium (MEM) plus 10% fetal calf serum (see Appendix) were added. After two days of incubation at 37fiC, the overlay medium containing poliovirus was decanted, centrifuged for 15 minutes at 1500 rpm to remove debris and then distributed to 1 m1 ampules which were immediately frozen at -70C. The poliovirus was assayed by the plaque assay technique on AV3 (human amnion) cell mono1ayers. This assay was performed by inoculating a drained AV3 monolayer with 0.2 m1 of the virus suspension which had been diluted in Eagle's MEM + 5% calf serum + 0.03 M Hepes buffer (see Appendix) to yield 100-300 plaques per bottle. Following inoculation, a @ne hour adsorption period with tilting at 15 minute intervals was allowed. The infected cells were then overlayed with 4 m1 of methyl cellulose overlay (see Appendix). After incubation at 37C for 48 hours, the methyl cellulose overlay was decanted and the monolayer was stained with crystal violet (see Appendix). PPlaques were subsequently counted using an Omega Enlarger B22 (Simmon Bros., Woodside, N.Y.). Magnetic Filtration of Adsorbed Viruses The salts used in this study were NaC1 (Fisher Scientific, Fairlawn, N.J.), CaC12 (Fisher) and A12(S04)3 18 H20 (Ma11inckrodt, St. Louis, Mo.). The magnetite was supplied by Fisher Scientific (catalog no. 1119) and its particle size distribution was determined using a Coulter Counter Industrial Model B (Coulter Electronics, Illinois). Figure 1 shows that the particle size varied from 3 to 30 with a maximum count between 3 and 5 The adsorption of viruses to magnetite and the subsequent removal of the magnetite-virus complex from the water suspension by filtration through a magnetic separator was undertaken as described below. Appropriate volumes from concentrated stock solutions of magnetite, salt, and added successively in batch experiments to samples of water resulting in a final volume of 100 m1. The pH of each solution was measured using a Beckman Expandomatic SS-2 pH meter and the resulting pH's were usually between 6 and 7. If required, pH adjustments were made using 0.1

PAGE 20

5 4 .jJ t: :::l ci U 0 3 r-O'l l I 0 ....J 0 I 2 1 o 10 30 Particle Figufe 1. Particle size distribution of magnetite (Fisher) as measured with the Coulter Counter Industrial Model B (Coulter Electronics, Illinois)

PAGE 21

-11-Figure 2. High gradient magnetic filter used to separate the virus-magnetite complex from the water suspension This filter consists of a stainless steel wool matrix (compaction equal to lag/50 cm3 ) placed in a background magnetic field of 2000 gauss.

PAGE 23

-13-N NaOH or HCl. These mixtures were then shaken at room temperature at approximately 170 oscillations/min and poured through a filter placed in a background magnetic field of 2000 gauss (see Figure 2). The filter was made of a matrix of stainless steel wool (supplied by Dr. E. Maxwell, Francis Bitter National Magnet Laboratory, M.l.T., Cambridge, Mass.) having the compaction of 10 g/50 cm3 and magnetized a background magnetic field. Each filtration experiment was carried out in triplicate and one control without magnetite was included in each experimental condition. Samples were removed and assayed for virus immediately after the shaking period (i.e., just prior to filtration) and after filtration through the magnetic separator. The controls without magnetite were included in order to account for loss of infectivity due to adsorption to containers or inactivation. The work described above was all done aseptically. Wastewater Effluents Used in Adsorption Experiments The wastewater effluents used in the adsorption experiments were "dome water" and an activated sludge effluent. "Dome water" is a secondarily treated effluent sampled at the edge of a cypress dome, located north of Gainesville, Florida. This effluent originated from a package treatment plant and became highly colored after standing in the dome site. When required, the removal of these coloring materials was undertaken by filtration through an activated carbon column, 11 cm in length and 3 em in diameter. The column contained 10 cm of Filtrasorb 400 (Calgon Co.) and 1 cm of fine particles of Hydrodarco B (Atlas Powder Co., N.Y.). The flow rate through the activated carbon column was 0.37 ml/min. The activated sludge effluent was sampled, after secondary settling and prior to chlorination, at the University of Florida campus sewage treatment plant. The effluents were immediately brought back to the laboratory and subsequently clarified by filtration through a Whatman filter no. 41, sterilized by autoclaving and stored at 4C until used. Total organic carbon, turbidity, color, pH and conductivity were determined in the laboratory using a Beckman 915 total organic carbon analyzer, a Hach model 2l00A turbidimeter, a Bausch and Lomb. Spectronic 88, a Beckman Expandomatic SS-2 pH meter and a conductivity bridge Model RC l6B2, respectively. The values for the various characteristics of the effluents are shown in Table 2.

PAGE 24

Table 2. Characteristics of the secondary effluents used in the magnetic filtration of viruses Parameter pH Color (mg/1 as Pt) Turb idi ty (JTU) TOC (mg/1) Conductivity at 25C) Dome Watera 6.20 -7.65 350.0 750.0 3.0 4.0 26.5 37.0 227.3 385.0 Dome Water #lb 7.10 575.0 4.0 31.0 385.0 Dome Water #1 After Activated Carbon Treatment 8.25 (adjusted to 7.10 with 0.1 N HC1) 8.0 4.0 7.0 320.9 Activated Sludge EffluentC 7;40 -7.73 30.0 35.0 0.6 -2.0 4.0 12.0 385; 0 472.7 a Various batches of dome water were collected and the values given for each parameter represent the high and low measurements obtained. b This is the first batch of dome water collected. c Various batches of activated sludge effluent were collected and the values given for each parameter represent the high and low measurements obtained. I i-' ..,. I

PAGE 25

-15-Organics Used in Adsorption Experiments The interference by organics with the adsorption of poliovirus to magnetite and consequently, with the removal of the virus by magnetic filtration was investigated. The organic substances used were egg albumin (cat. no. B2SS), casein (cat. no. B337), both supplied by Difco Laboratories (Detroit, Mich.) and dextran-S x 105 MW (cat. no. 05251) supplied by Sigma Chemical Co. (St. Louis, Mo.). Stock solutions (0.1%) of these substances were prepared in distilled water and were then sterilized by autoclaving and stored at 4C. Infectivity of Virus Adsorbed to Magnetite In order to determine the infectivity of virus adsorbed to magnetite, an experiment was undertaken as described below. Viruses were shaken 170 oscillations/min.) for 20 min. in the presence of 500 ppm of magnetite and 1610 ppm of CaC12' Low virus concentrations were used so that samples might be plated directly with-out requiring dilution. Samples were removed and directly plated immediately after addition of the virus and after 20 min. adsorp-tion period. Following the adsorption period, the magnetite was pelleted by placing the mixtures next to a magnet and the supernatants were subsequently assayed for virus. Controls without magnetite were studied in order to account for loss of infectivity due to adsorption to containers or inactivation. Elution of Adsorbed Poliovirus from Magnetite Attempts were made to elute the adsorbed poliovirus from magnetite. The procedure employed is described below. Poliovirus was shaken 200 oscillations/min.) for 20 min.; lhn the presence of 500 ppm of magnetite and 1610 ppm of CaC12' Ten ml aliquots were then removed and distributed into sterile test tubes. The magnetite in each test tube was then pelleted by placing the mixture next to a magnet. Supernatants were removed for assay and then discarded. The magnetite pellets were resuspended in 10 ml of the eluents shown in Table 20 (the composition and source of these eluents appear in the Appendix). After a 1 hour elution period with periodic at 4C (except for glycine buffer which was allowed only 1 min of contact time), the magnetite was pelleted again. The supernatants were then assayed for desorbed viruses. Concentration of Poliovirus by Magnetic Filtration Recovery of poliovirus from 4 liters of distilled water containing initial virus concentrations ranging from 1.0 x 103 to 1.3 x 10 3 pfu/ml was undertaken as described below. Poliovirus was shaken 200 oscillations/min.) for 20 min. in the presence of 1000 ppm of magnetite and 1610 ppm of CaC12' The mixtures were subsequently

PAGE 26

-16-poured through the magnetic filter in order to separate the polio virusffiillagnetite complex from the water suspension. Poliovirus was then eluted from the magnetite surface with a small volume of 10%, isotonic fetal calf serum, Tris buffered, pH = 9. The procedure used for the recovery of poliovirus varied with each experiment performed. In a first experiment, 100 ml of the eluent was passed twice through the filter with the filter in the magnetic field and once (the last passage) with the filter outside the magnetic field. The eluauee (separated from any magnetite) was assayed for virus after each passage through the filter. The Sala magnet pictured in Figure 3 was used in this experiment (and only this experiment). In a second experiment, 50 ml of the eluent was introduced into the filter and then the filter was shaken manually for 1 min and then shaken mechanically overnight at 4C. After each shaking period, the eluate was separated from magnetite and then assayed for viruses. In the third experiment, the matrix of the filter was removed (along with all the magnetite) and placed in a sterile beaker containing 100 ml of the eluent. The matrix was then sonicated in an ice bath for 10 min. using the standard probe of a Branson Sonifier (Danbury, Conn.) S-75 which has a power output of 75 watts and an ultrasonic frequency of 20 kc/sec. The eluate was then assayed for viruses. Statistical Treatment of Data The two statistical procedures used to treat data were: 1. the small-sample, t test for comparing two means, and 2. linear regression analysis to determine the least squares regression line for a set of points and the coefficient of determination r2 which is a measure of the strength of the relationship represented by the regression line. The Hewlett-Packard Calculator Model 98l0A and Statistics were used to perform the statistical analysis.

PAGE 27

-17-Figure 3. High gradient magnetic filter supplied by Sa1a Magnetics, Cambridge, Mass. This filter was used in the concentration of Poliovirus, experiment 1.

PAGE 28

-18-

PAGE 29

-19RESULTS AND DISCUSSION The purpose of this study was to determine the effectiveness of magnetic filtration in removing viruses from water. Consequently, factors that could influence the adsorption of viruses onto magnetite were investigated. These included cation valency and concentration, pH, time allowed for adsorption, magnetite concentration, virus type and concentration and organic substances in the suspending medium. Effect of Cation Valency and Concentration The effect of monovalent (Na+, K+), divalent (Ca++, Mg++) and trivalent (Al+++) cations on virus adsorption to magnetite and on the subsequent virus removal by magnetic filtration was studied (Tables 3-12). It was observed that, in the presence of any salt under consideration, the removal of poliovirus and MS2 phage approached or exceeded the 99% level when the salt concentration was at or above 20 ppm. The percent removal of T2 phage was below that of poliovirus or MS2 phage for any cation used and at any concentration. The highest removal of T2 phage (95.9%) was achieved in the presence of 16100 ppm of CaC12 (see Table 5). The removal obtained for the T4r (II) phage in the presence of 1610 ppm of CaC12 or 1380 ppm of MgC12 (Table 6) was similar to that found for the T2 phage (Table 5). In the presence of 100 ppm of alum (see Tables 9 and 10), however, the removal of the T4r (II) phage (99.2%) greatly exceeded that displayed by the T2 phage (85.4%). There was no significant difference in the removal of the phage T2 in the presence of K+ OT Na+ cations (Table 3). Calcium cations (Ca++) were found more effective than Mg++ cations at ionic strengths equal to 4.35 x 10-2 and 4.35 x 10-1 (see Tables 5 and 6) It has been shown by other workers that the electrolyte content of the suspending medium is important for the adsorption of viruses to particles such as iron oxides, clays, activated carbon, polyelectrolytes, sand and soil (Bitton andMi tchell, 1974; Carlson et al., 1968; Cookson, 1969; Thorup et al., 1970; Lefler and Kott-,-1974; Drewry and Eliassen, 1968). Furthermore, it is known that, on a concentration or molar basis, trivalent cations are more efficient than divalent cations which in turn are more efficient than monovalent ones. Viruses are colloidal particles having a net negative surface charge at neutral pH. Most of the adsorbents studied also carry a net negative surface charge at neutral pH. An increase in the ionic strength of the suspending medium leads to a reduction of the thickness of the double-layer around the particles which are then bound by attractive forces, namely, London-Van der Waals forces (Clark et al., 1971; Osipow, 1962). The adsorption of viruses to magnetite-,-as seen in Tables 3 through 12, follows the trends discussed above with the exception

PAGE 30

-20Table 3. Effect of monovalent cations (Na+ and K+) on the removal of bacteriophage T2 by magnetic filtration Ionic Strength (]l) Concentration (ppm) % Removal of T2a by Magnetic Fi1trationb of the Salt Used of the Salt Used NaCL or KC1 NaC1 NaC1 KC1 4.35 4.35 4.35 4.35 b x 10-4 25.4 3Z244 62.1 46.8 x 10-3 254 56.1 55.9 x 10-2 ''3240 72.9 79.5 x 10-1 25Zl!00 65.2 61.0 The initial bacteriophage T2 concentration was 2000 pfu/m1. The bacteriophage T2 was shaken oscillations/min) for 20 min in the presence of 300 ppm of magnetite and various concentrations of NaC1 or KC1. These mixtures were subsequently poured through the magnetic filter.

PAGE 31

-21-Table 4. Effect of a monovalent cation (Na+) on the removal of Poliovirus Type I (Sabin) by magnetic filtration Ionic Strength Concentration % Removal of Poliovirusa of NaCl ( ]1) of NaCl (ppm) by Magnetic Filtrationb 4.35 x 10-4 25.4 93.0 4.35 x 10-3 254 98.7 4.35 x 10-2 2540 99.1 4.35 x 10-1 25400 98.9 a The initial Poliovirus concentration was 15 x 10 3 pfu/ml. b The Poliovirus was shaken 200 oscillations/min) for 20 min in the presence of 500 ppm of magnetite and various concentrations of NaCl. These mixtures were subsequently poured through the magnetic filter.

PAGE 32

-22-Table 5. Effect of divalent cations (Ca++ and Mg++) on the removal of bacteriophage T2 by magnetic filtration Ionic Strength Concentration (ppm) % Removal of T 2 a by of the Salt Used of the Sal t Used Magnetic Filtrationb CaC12 or MgC12 CaC12 MgC12 CaC12 MgC12 4. t;Jg X l(,D-4 16.1 13.8 26.1 35.4 4.35 x 10-3 161 138 82.1 80.7 4.35 x 10-2 1610 1380 93.1 81.2 4.35 x 10-1 16100 13800 95.9 82.3 a The initial bacteriophage T2 concentration was 2000 pfu/ml. b The bacteriophage T2 was shaken 130 oscillations/min) for 20 min in the presence of 300 of magnetite and various concentrations of CaC12 or MgC12 These mixtures were subsequently poured the filter.

PAGE 33

-23-Table 6. Effect of divalent cations (Ca++ and Mg++) on the removal of bacteriophage T 4R(II) by magnetic filtration Ionic Strength Concentration (ppm) % Removal of T4R(II)a by Magnetic Fi1trationb of the Salt Used of the Salt Used CaC12 or MgC12 CaC12 MgC12 CaC12 Mg0l;2 4.35 4.35 4.35 4.35 a b x 10-4 16.1 13.8 62.6 63.3 x 10-3 161 138 60.1 69.8 x 10-2 1610 1380 85.1 71.2 x 10-1 16100 13800 92.5 76.4 The inttia1 bacteriophage T4R(II) concentration was 2000 pfu/m1. The bacteriophage T4R(II) was shaken 130 oscillations/min) for 20 min in the presence of 300 ppm of magnetite and various concentrations of CaC12 or MgC12 IIhae mixtures were subse poured through the magnetic filter.

PAGE 34

-Z4Table 7. Effect of divalent cations (Ca++ and Mg++) on the removal of bacteriophage MSZ by magnetic filtration Ionic Strength of the Salt Used Concentration (ppm) of the Salt Used % Removal of MSZa by Magnetic Filtration b CaC1Z or MgC1Z CaC1Z MgClz CaC1Z MgClz 4.11 x 10-Z 1610 1380 99.6 99.6 b ID[b initial bacteriophage MSZ concentration was 7000 pfu/ml. bacteriophage MSZ was shaken 130 oscillations/min) for ZO min in the presence of 500 ppm of magnetite and the concentration of CaC1Z or MgC1Z above. These mixtures were subsequently poured through the magnetic filter.

PAGE 35

-25-Table 8. Effect of a divalent cation (Ca++) on the removal of Poliovirus Type I (Sabin) by magnetic filtration Ionic Strength Concentration % Removal of Poliovirusa of CaC12 (11) of CaC12 (ppm) by Magnetic Filtrationb 4.35 x 10-4 16.1 16.0 4.35 x 10-3 161 99.8 4.35 x 10-2 1610 99.6 4.35 x 10-1 16100 99.1 a The initial Poliovirus concentration was 15 x 10 3 pfu/ml. b The Poliovirus was shaken (rv 200 oscillations/min) for 20 min in the presence of 500 ppm of magnetite and various concentrations of CaC12. These mixtures were subsequently poured through the magnetic filter.

PAGE 36

-26-Table 9. Effect of a trivalent cation (Al+++) on the removal of bacteriophage T2 by magnetic filtration Ionic Strength of A12(S04)3 l8H20 Concentration of A12 (S04)3 l8H20 (ppm) % Removal of T2a by Magnetic Filtrationb 4.35 x 10-4 13.05 x 10-4 21.75 x 10-4 20 60 100 64.8 43.6 85.4 a b The initial bacteriophage T2 concentration was 5600 pfu/ml. The bacteriophage T2 was shaken 130 oscillations/min) for 20 min in the presence of 300 ppm of magnetite, various concentrations o-f A12(S04)3 l8H 2 0 and pH's adjusted to 6.0 0.1 with 1.ON NaHC03' These mixtures were subsequently poured through the magnetic filter.

PAGE 37

-27-Table 10. Effect of a trivalent cation (Al+++) on the removal of bacteriophage T4R(II) by magnetic filtration Ionic Strength of Concentration of % Removal of T 4R(II)a by Magnetic Filtrationb A12 (S04)3 l8H20 A12 (S04)3 l8H20 (ppm) x L@-4 20 41.7 13.05 21.75 a b x 10-4 60 99.5 x 104 100 99.2 The initial bacteriophage T4R(II) concentration was 3000 pfu/ml. The bacteriophage T4R(II) was shaken 130 oscillations/min) for 20 min in the presence of 300 ppm of magnetite, various concentrations of A12 ( S04)3 l8H20 and pH's adjusted to 6.0 0.1 with 1.ON NaHC03. These mixtures were subsequently poured through the magnetic filter.

PAGE 38

-2STable 11. Effect of a trivalent cation (Al+++) on the removal of bacteriophage MS2 by magnetic filtration Ionic Strength of Concentration of % Removal of MS2a bb A12 (S04)3 lSH2 0 A12(S04)3 lSH 2 0 Magnetic Filtration (ppm) 4.35 x 10-4 20 99.6 13.05 21. 75 a b x 10-4 60 100.0 x 10-4 100 100.0 The initial bacteriophage MS2 concentration was 3700 pfu/ml. The bacteriophage MS2 was shaken ('V 130 oscillations/min) for 20 min in the presence of 500 ppm of magnetite, various concentrations of A12(S04)3 lSH20 and pH's adjusted to 6.0 + 0.1 with O.lN NaOH. These mixtures were subsequently poured through the magnetic filter.

PAGE 39

-29-Table 12. Effect of a trivalent cation (Al+++) on the removal of Poliovirus Type I (Sabin) by magnetic filtration T@1lI.9:c Strength of Concentration of pHa of % Removal of POliovirusb by A12(S04)3 l8H 2 O 3 tppm) l8H20 Solution (ll) Magnetic Filtrationc 4.35 x 10-4 20 4.64 81.6 13.05 x 10-4 60 4.40 98.5 21.75 x 10-4 100 4.29 90.4 4.35 x 10-3 200 4.14 91.7 a b c The pH's reported represent the actual measured pH's of the solutions; no pH made. The initial Poliovirus concentration was 15 x 10 3 pfu/ml. The Poliovirus was shaken 200 oscillations/min) for 20 min in the presence of 500 ppm of magnetite and various concentrations of A12(S04)3 l8H20. These mixtures were subsequently poured through the magnetic filter.

PAGE 40

-30of the 93% removal of poliovirus and 6Z.l% removal of TZ phage obtained in the presence of only Z5.4 ppm of NaCl (Tables 3 and 4). The adsorption of poliovirus was also found to decrease when the alum concentration was increased to above 60 ppm (see Table lZ). As no pH adjustments were made when adding the alum, this decrease in sorption was probably due to the lowering of the pH by 100 and ZOO ppm of alum to pH values of 4.Z9 and 4.14, respectively. These values are below the isoelectric point of 4.5 reported by Mandel (1971) for poliovirus type 1. We also noted that in the pH experiment described in the next section, the adsorption of poliovirus was also weak below pH 5. One may aiso add that, among the three bacteriophages studied, MSZ is the only one which has an adsorption pattern similar to that of poliovirus. Effect of pH The pH of the suspending medium has been reported to be significant in the adsorption of viruses to such surfaces as soil, membrane filters and synthetic insoluble polyelectrolytes (Reece, 1967; Wallis and Melnick, 1967a; Wallis et al., 1971). In order to simulate the pH of most natural adsorption of TZ phage, phage and poliovirus to magnetite was studied in the pH range of 4 to 9. Figure 4 shows that, above pH 5, the removal of poliovirus was i:limthe 98-99% level and did not vary significantly. However, below pH 5, a decrease in removal was observed for this virus. For the phages TZ and MSZ, the removal (98-99% level) did not vary significantly in the pH range of 4 to 9. Bacteriophages TZ and MSZ, and poliovirus type 1 have isoelectric points of 4.Z, 3.9 and 4.5, respectively (see Table 1). Consequently, the studied pH 4 was above the isoelectric point of MSZ, only slightly below the isoelectric point of TZ and significantly below the isoelectric point of poliovirus. It is postulated that the adsorption of poliovirus onto magnetite was inhibited by lowering the pH of the suspending medium sufficiently below the isoelectric point of the virus. Unfortu nately, no data was collected on the electrophoretic mobility of magnetite particles in order to draw more definite conclusions. Kinetics of the Adsorption Process The kinetics of the adsorption process was investigated and it was found that 10 to ZO minutes of shaking were sufficient to allow an optimum removal of poliovirus (see Table 13). For the phage MSZ, equilibrium was reached,wi;thl8.s little,as5to 10 minutes of shaking (Table 13). On the other hand, the removal of TZ phage increased with time in the range of shaking times studied and did not reach an equilibrium value (see Table 13). From the above results, i ttmrssdecided to allow ZO minutes shaking time for the adsorption experiments. This period of shaking was suf-

PAGE 41

Figure 4. Effect of pH on the removal of bacteriophage T2, bacteriophage MS2 and Poliovirus by magnetic filtration Bacteriophage T2 (5 x 10 3 pfu/ml), bacteriophage MS2 (5 x 103 pfu/ml) and Poliovirus Type I (15 x 103 pfu/ml) were each shaken 170 oscillations/ min) for 20 min in the presence of 500 ppm of magnetite, 1610 ppm of CaC12 and varying pH (adjusted with O.lN NaOH and O.lN HCl). These mixtures were subsequently poured through the magnetic filter. I Vol ..... I

PAGE 42

o o -32-T I I I OJ en rO ..c: 0... N ls a sn.A A J 0 LEA 0 ill a OJ en rO ..c: 0... N V) 0;:-. """ til ::::l So-'r> 'r-0 c.. co :c 0... L() o

PAGE 43

-33-Table 13. Effect of time of shaking on the removal of bacteriophage T 2 bacteriophage MS2 and Poliovirus Type I (Sabin) by magnetic filtration Time a b c d e of Shakinga % Removal by Magnetic Fil trationb (min) T2c Msit PoUoviruse 1 56.3 90.4 66.8 5 80.9 99.2 94.7 10 83.2 100.0 100.0 20 95.3 98.5 99.5 30 97.1 99.8 98.9 The flasks shaken at 170 oscillations/min. The:mmggBtic filtration was undertaken in the presence of 500 ppm of magnetite and 1610 ppm of CaC12. The initial bacteriophage T2 concentration was 7 x 103 pfu/ml. The initial bacteriophage MS2 concentration was 4 x 10 3 pfu/ml. The initial concentration was 3 x 10 3 pfu/ml.

PAGE 44

-34-ficient for the optimum removal of poliovirus and MS2 phage but apparently not for T2 phage. Nevertheless, a longer period of shaking was not allowed for the phage T2 in order to avoid the large inactivation of the virus which wOl!l,ld occur when shaken for periods longer than 20 minutes. Effect of Magnetite Concentration The concentration of magnetite, used as a seed material for magnetic filtration, is an important factor to consider in light of the economic feasibility of the process. Figyre 5 shows that 300 ppm of magnetite wasrsufficient for the optimum removal of the phage MS2 (6 x 103 and poliovirus (8 x 103 pfu/ml) and 500 ppm is sufficient'for the phage T2 (4.5 x 103 pfu/ml). Consequently, adsorption experiments were carried out in the presence of 500 ppm of magnetite in order to insure optimal conditions for adsorption and subsequent removal by magnetic fiI tration. Effect of Initial Virus Concentration The Freundlich adsorption isotherm (Fair et al., 1968) has often been used to show that removal of a solute (e.g., virus) from an aqueous solution by solid media was an adsorption process. This empirical relation is expressed as y/m = kc l / n (4) where, for the specific case in which the adsorbate is a virus, y/m is the quantity of virus removed per unit weight of adsorbent (e.g., magnetite) and c is the concentration of virus remaining in solution at equilibrium. The constant k has been described as a measure of the surface area of the solid phase and the constant n as an indicator of the intensity of adsorption (Reece, 1967). The Freundlich equation is usually used in the logarithmic form, log (y/m) = log It + l/n log c, (5) which indicates a linear variation of log (y/m) with log c. Con sequently, a double-logarithmic plot of data conforming to the Freundlich isotherm should give a line with slope l/n and log k as the / y-intercept. The interaction of viruses with such surfaces as activated carbon, activated sludge, stabilization pond solids and soil (Gerba et al., 1975; Clarke et al., 1961; Sobsey and Cooper, 1973; Drewry andJ6liassen, 1968) has been shown to obey the Freundlich isotherm. This general tendency of virus-surface systems to conform to the Freundlich isotherm shows the importance of adsorption

PAGE 45

Figure 5. Effect of magnetite concentration on the removal of bacteriophage T 2 bacteriophage MS2 and Poliovirus Type I (Sabin) by magnetic filtration Bacteriophage T2 (4.5 x 10 3 pfu/m1), bacteriophage MS2 (6 x 103 pfu/m1) and Poliovirus (8 x 10 3 pfu/m1) were each shaken 170 oscillations/min) for 20 min in the presence of 1610 ppm of CaC12 and various concentrations of magnetite. These mixtures were subsequently poured through the magnetic filter. I Vl Ul I

PAGE 46

U1 (J) U1 :::s s... > 40 0 E (J) 0::: 1001 K: me :a: _______ 80 60 T 2 phage .. -----e MS2 phageO 0 PoliovirusrJ () 200 400 600 800 1000 Magnetite concentration (ppm) I V-l Q'\ I

PAGE 47

-37-Figure 6. Freundlich isotherms for the adsorption of bacteriophage T 2 bacteriophage MS2 and Poliovirus Type I (Sabin) to magnetite: variable virus concentration Ba6te:tH:(\1]ilhllgegf r bacteriophage MS2 and Poliovirus were each shaken 170 oscillations/min) for 20 min in the presence of 500 ppm of magnetite and 1610 ppm of CaC12. These mixtures were subsequently poured through the magnetic filter. Linear regression analysis yielded the least squares regression lines drawn: Slope y-Intercept r2 T2 phage 1.035 1.354 0.99 MS2 phage 1.166 2.314 0.98 Poliovirus 0.736 2.743 0.99

PAGE 48

OJ +-> ',-+-> OJ s::: en to E 4o en E OJ Cl. "0 OJ > o E OJ ',> o 7 6 5 4 3 2 o ...;.382 T2 phage .. --.... MS2 3 4 Loglo virus concentration in effluent (pfu/ml) 5

PAGE 49

-39-in removing viruses from the water suspension. In order to determine the nature of the removal process by magnetite, an experiment was performed in which the T2 phage, MS2 phage and poliovirus concentration was varied but the magnetite concentration was held constant. Figure 6 shows that the results obtained for the three viruses conform to the Freundlich isotherm. The Freundlich isotherm for the T2 phage was found to be below the isotherms for poliovirus and the MS2 phage (Figure 6). This could be explained by the larger size of the T2 phage in comparison to poliovirus and the phage MS2 (see Table 1). Due to its larger size, less T2 phage can be adsorbed per mg of magnetite when compared to a smaller virus such as poliovirus or the MS2 phage. Also, notice that the y-intercept ("log k" in the Freundlich equation) is less for the T2 phage 0.354) than for poliovirus (2.743) or the phage MS2 (2.314). Since the constant k in the Freundlich equation has been described as a measure of the surface area of the adsorbent, then a decrease in its value, as seen for the phage T2, may indicate a reduction of the surface area available to viruses for adsorption onto magnetite. In this case, the surface area is not decreased but the size of the virus is increased. The difference in the adsorption pattern may also be due to the fact that T2 phage has a tail whereas MS2 phage and poliovirus are tailless. Effect of Wastewater Effluents on the Adsorption of Viruses to Magnetite Wastewater effluents contain organic materials which may compete with viruses for adsorption onto solids (Amirhor and Engelbrecht, 1974; Carlson et al., 1968; Dieterich, 1953; Gerba et al., 1975) and may lead to a-decreased removal of the infective particles. Consequently, experiments were undertaken to determine if the adsorption of viruses to magnetite was influenced by organics present in two wastewater effluents, "dome water" and University of Florida campus activated sludge effluent (see Table 2 for characteristics of these effluents). Table 14 shows that, in the absence of CaC12 dome water strongly interferes with the removal of the phage T2 (0.0% removal) and that this interference could be reduced (but not completely eliminated) with the addition of CaC12' It can be seen in Table 15 that the removal obtained in the presence of dome water and 1610 ppm of CaC12 for the phage T2 can be increased by diluting the wastewater effluent (and the organics present) in distilled water. Removal of the organic color present in dome water with an activated carbon treatment greatly reduced the interference exerted by this effluent (94.9% removal of T2 --see Table 15). However, the removal of T2 phage found in distilled water (98.3%) could not be obtained in dome water regardless of previous treatment. Apparently, the activated carbon did not remove all the interfering substances from the water. Table 16 shows clearly that dome

PAGE 50

-40Table 14. Effect of a divalent cation (Ca++) on the removal of bacteriophage TZ by magnetic filtration in the presence of dome water Ionic Strength of CaClZ (]l) 4.35 x 10-4 4.35 x 10-3 4.35 x 10-Z 4.35 x 10-1 Concentration of CaCl Z (ppm) 0 16.1 161 1610 16100 % Removal of TZa by Magnetic Filtrationb in the presence of Dome WaterC 0.0 16.4 44.8 76.0 86.9 a The initial bacteriophage TZ concentration was 5.6 x 10 3 pfu/ml. b The bacteriophage TZ was shaken 130 oscillations/min) for ZO min in the presence of dome water, 500 ppm of magnetite and various concentrations of CaClZ. These mixtures were subsequently poured through the magnetic filter. c The dome water was sampled at a cypress dome located north of Gainesville, Florida.

PAGE 51

-41-Table 15. Removal of bacteriophage T2 by magnetic filtration in the presence of dome water Description Dome WaterC Dome Water I1il1lltibed 1:2 in Distilled Water Dome Water kiiill.1lJ[1ten 1: 4 in Distilled Water Activated Carbon Filtered Dome Water Distilled Water % Removal of T 2 a by Magnetic Filtrationb 74.3 75.7 88.6 94.9 98.3 a The initial bacteriophage T2 concentration was 4 x 10 3 pfu/ml. b c The bacteriophage T2 was shaken 130 oscillations/min) for 20 min in the presence of 500 ppm of magnetite, 1610 ppm of CaC12 and the various solutions described above. These mixtures were subsequently poured through the magnetic filter. The dome water was sampled at a cypress dome located north of Gainesville, Florida.

PAGE 52

-42-Table 16. Removal of Poliovirus Type I (Sabin) by magnetic filtration in the presence of dome water Description Dome WaterC Dome Water + 1610 ppm CaC12 Distilled Water d Distilled Water + 1610 ppm CaC12 Activated Carbon Filtered Dome Water Activated Carbon Filtered Dome Water + 1610 ppm CaC12 % Removal of Poliovirusa by Magnetic Filtrationb 0.0 96.3 99.4 99.6 99.2 99.1 a The initial Poliovirus concentration was 15 x 103 pfu/ml. b c d The Poliovirus was shaken 200 oscillations/min) for 20 min in the presence of 500 ppm of magnetite and the various solutions described above. These mixtures were subsequently poured through the magnetic filter. The dome water was sampled at a cypress dome located north of Gainesville, Florida. The distilled water was adjusted to the same .conductivity as the dome water (conductivity equal to 385.0 at 250C) with CaC12

PAGE 53

-43-water also interferes with the removal of poliovirus. This interference was significantly reduced by adding 1610 ppm of CaC12 (96.3% removal) or by removing the organic color with activated carbon treatment (99.2% removal). The interference of dome water with the adsorption of viruses on magnetite is probably due to the ful vic acid fraction released in the water by the decomposition of cypress needles and other leaves within the cypress dome. It is generally known that fulvic acids are soluble in water whereas humic acids are not although some low molecular fractions may be (Flaig, 1960; Prakash and Rashid, 1968). It has been found that dome water also interferes with the adsorption of poliovirus on soil and may be efficient for the elution of soil-adsorbed viruses (Bitton, et a1., 1976). The other effluent used in this study was the University of Florida campus activated sludge effluent. The interference exerted by the organics present in this effluent was slight. When this effluent was used as the suspending medium for poliovirus, the removal was 96.7% in the absence of any salt and was enhanced to 99.4% with the addition of 1610 ppm of CaC12 (Table 17). This removal was similar to the one observed in distilled water adjusted to the conductivity of the activated sludge effluent (385 at 250C) or containing 1610 ppm of CaC12 (Table 17). It is possible to the two wastewater effluents, dome water and campus activated sludge effluent, for their interference with the adsorption of poliovirus onto magnetite (Tables 16 and 17). The interference was mostly apparent with the dome water and was completely eliminated by filtering the dome water through an activated carbon column. In the presence of CaC12, the campus activated sludge effluent displayed a better virus removal (99.4%) than the dome water (96.3%). The main difference between the two types of effluents (st:fe Table 2) was the color content. The dome water had a color ranging from 350.0 to 750.0 units, whereas the campus activated sludge effluent had a color ranging from 30.0 to 35.0 units. Therefore, it could be postulated that the organic color was responsible for the observed interference. It should be added that dome water is not a typical wastewater effluent due to its high organic color. Consequently, the results of the campus activated sludge effluent should be considered when assessing the virus removing potential of magnetic filtration in the presence of wastewater effluents. Additional experiments were carried out in which the virus concentration was held constant and the magnetite concentration was varied. The data in Figures 7, 8 and 9 shows that, in the presence or the absence of secondary wastewater the adsorption of bacteriophage T2, bacteriophage MS2 and poliovirus

PAGE 54

-44-Table 17. Removal of Poliovirus Type I (Sabin) by magnetic filtration in the presence of activated sludge effluent Description % Removal of Po1iovirusa by Magnetic Fi1trationb Activated Sludge Eff1uentC 96.7 Activated Sludge Effluent + 1610 ppm CaC12 99.4 Distilled Water d 99.4 Distilled Water + 1610 ppm CaC12 99.6 a b c d The initial Poliovirus concentration was 8 x 103 pfu/m1. The Poliovirus was shaken 200 oscillations/min) for 20 min in the presence of 500 ppm of magnetite and the various solutions described above. These mixtures were subsequently poured through the magnetic filter. The activated sludge effluent was sampled at the University of Florida campus sewage treatment plant. The distilled water was adjusted to the same conductivity as the activated sludge effluent (conductivity equal to 385.0 at 2S0C) with CaC12.

PAGE 55

Figure 7. Freundlich isotherms for the adsorption of bacteriophage T2 to magnetite in distilled water and dome water: variable magnetite concentration Bacteriophage T2 (4.5 x 10 3 pfu/ml) was shaken 130 oscillations/min) for 20 min in the presence of 1610 ppm of CaC12, various concentrations of magnetite (ranging from 100 ppm to 1000 ppm) and distilled .. Water of dome water. These mixtures were subsequently poured through the magnetic filter. Linear regression analysis yielded the least squares regression lines drawn: Distilled Water Dome Water slope 0.595 0.718 y-intercept 2.889 1.167 r2 0.86 0.99 I .j:>. c.n I

PAGE 56

-46-T ..--.. I ..... I E I .......... ::l 4-0-S-.f-I a; c: +> a) to ::l :3: M ..... S-4--0 a; 4-a; .f-I a; ..... to r-:3: c: \ ..... ..... .f-I a; III E c: ..... 0 0 Cl Cl ..... +> to s... .f-I c: a; U c: 0 N U '. N I-\ a; O"l ,. to oJ:: 0- 0 .... S-a; .f-I U to ..a 0 r-r-. O"l 0 -I o (6wjnJd) cl:j. lau6ew JO 6w .,lad paAOWa.A G 1 a6eljdo 0 LOOl

PAGE 57

Figure 8. Freundlich isotherms for the adsorption of bacteriophage MS2 to magnetite in distilled water and activated sludge effluent: variable magnetite concentration Bacteriophage MS2 (6.3 x 10 3 pfu/ml) was shaken 130 oscillations/min) for 20 min in the presence of 1610 ppm of CaC12' various concentrations of magnetite (ranging from 100 ppm to 1000 ppm) and distilled water or activated sludge effluent. These mixtures were subsequently poured through the magnetic filter. Linear regression analysis yielded the least squares regression lines drawn: Distilled Water Activated Sludge Effluent slope 0.769 1.021 y-intercept 3.082 1.272 r2 0.99 0.95 I ..,. ---.] I

PAGE 58

0) E -;5 4-0.. Q) ...... Q) s:: 0) ro E 4o 0) E 4 -0 Q) > o E Q) N Vl :E: Q) 0) ro ..s:: 0.. o '" ",e / / / '" W' [eJ '" .,/ / '" / ..I' [oJ 2 3 I Distilled Water .. i L Activated Sludge Effluent 0 0 o III 2 3 4 Loglo bacteriophage MS2 concentration in effluent (pfu/ml) I .j::> 00 I

PAGE 59

Figure 9. Freundlich isotherms for the adsorption of Poliovirus Type I (Sabin) to magnetite in distilled water and wastewater effluents: variable magnetite concentration Poliovirus (8 x 103 pfu/ml) was shaken 200 oscillations/min) for 20 min in the presence of 1610 ppm of CaC12, various concentrations of magnetite (ranging from 100 ppm to 1000 ppm) and distilled water, dome water or activated sludge effluent. These mixtures were subsequently poured through the magnetic filter. Linear regression analysis yielded the least regression lines drawn: Distilled Water Activated Sludge Effluent Dome Water slope 0.715 1.395 2.590 y-intercept 2.897 1.011 -3.313 r2 0.98 0.94 0.99 I ..j:>. \D I

PAGE 60

5 01 [.] E '-(J ::::s 4a. '-'" Q) .jJ .... ". +-l Q) / c 01 tt1 ", E ,.4 4-4 e''''' 0 01 E [.yo [OJ SQ) 0.. I (J1 -0 0 Q) I > [0] 0 E Q) S-III ::::s S-.... > Distilled Water .. --... .... 3 Activated Sludge Effluent 0 0 o Dome Water 0 0 r-01 .0 -l I I I o I 1 2 3 Il LoglO Poliovirus concentration in effluent (pfu/ml)

PAGE 61

-51-onto magnetite conformed to the Freundlich isotherm. However, in the presence of wastewater effluents, the Freundlich isotherms had a lower y-intercept ("log kIf in the Freundlich isotherm equation) than that found in distilled water. For example, in Figure 7, for the phage T2, the y-intercept of the Freundlich isotherm was as high as 2.889 in the absence of organic materials and decreased to 1.167 in the presence of dome water. For the phage MS2 (see Figure 8), the y-intercept was 3.082 in distilled water and dropped to 1.272 in the presence of activated sludge effluent. Figure 9 shows that, for poliovirus, the y-intercept was as high as 2.897 in distilled water and decreased to 1.011 in the presence of campus activated sludge effluent and -3.313 in the presence of dome water. This consistent drop in the value of the y-intercept ('I'ihQg kIf) when the viruses are suspended in wastewater effluents aan be explained in the following manner. The constant k in the Freundlich isotherm equation has been described as a measure of the surface area of the solid phase (Reece, 1967) and, therefore, a decrease in its value may indicate a reduction in surface area of magnetite available to viruses for adsorption. This reduction is due to the occupation of adsorption sites on the magnetite by competing organics present in wastewater. Effect of Organics on the Adsorption of Poliovirus to Magnetite A variety of organic substances such as bovine albumin, egg :m.lbuniin. meat infusion broth and cell extracts have been shown to compete with viruses for adsorption onto solids (Carlson et al., 1968; Shepard and Woodend, 1951; NRililis and Melnick, 119)67a). In order to determine if organics would also interfere with the adsorption of poliovirus to magnetite, an experiment was undertaken using egg and dextran. Table 18 shows that 1000ppm of these materials significantly :iiutterfered with the removal of poliovirus. Casein is known to be a highly efficient eluent of virus from cellulose membranes (Wallis and Melnick, 1967b) and therefore, its complete interference with the removal of poliovirus (0.0%) is not surprising (Table 18). Moreover, it should be noted that 100 ppm of these organic materials is an extremely high concentration. For example, Carlson et al. (1968) showed that as little as 2ppm of egg albumin reduced the adsorption of the phage T2 to the clay KaoIilTite 4 from 93% to 27%. It can be concluded that a reduction in the removal of viruses by magnetic filtration in the presence of high concentrations of an interfering substance is to be Infectivity of Viruses Adsorbed to Magnetite The previous sections have dealt with the factors ing the adsorption of viruses to magnetite. It was learned that

PAGE 62

-52-Table 18. by egg albumin, casein and dextran with the removal of Poliovirus Type I (Sabin) by magnetic filtration Description 100 ppm Egg Albumin 100 ppm Casein 100 ppm Dextran (5 x 105 MW) % Removal of Poliovirusa by Magnetic Filtrationb 42.4 0.0 7.1 99.6 a The initial Poliovirus concentration was 1.0 x 10 4 pfu/ml. b The Poliovirus was shaken 200 oscillations/min) for 20 min in the presence of 500 ppm of magnetite, 1610 ppm of CaC12 and the various solutions described above. These mixtures were subsequently poured through the magnetic filter.

PAGE 63

-53-under appropriate conditions, viruses can be efficiently adsorbed to magnetite and subsequently removed by magnetic filtration. Experiments were then performed to determine the infectivity of the adsorbed virus. We used low concentrations of viruses so that samples might be directly plated and thereby, avoid dilutions which could lead to the desorption of viruses. Table 19 shows that T2 phage (97.5%), MS2 phage and poliovirus (98.9%) were effectively adsorbed to magnetite. The infectivity of adsorbed poliovirus and MS2 phage was >100.0%. The elevated titer (>100.0%) of the adsorbed virus could have resulted from virus disaggregation when shaken for 20 minutes. Schaub and Sagik (1975) have also observed that enteric viruses adsorbed to either Montmorillonite clay or other naturally occurrmng solids displayed an elevated titer. These authors proposed that the clay particles may function in allowing the virus to establish better proximity to the cells. Table 19 also shows that the infectivity of adsorbed T2 phage was lower than that displayed by and much more variable (57.8% and 93.5% infective). Puck and Sagik (1953) have shown that T2 phage, after its attachment to a cationic resin, was split into its DNA and protein fractions and therefore, could not be recovered in active form. It appears that upon adsorbing to this surface, the T2 phage ejaculates its DNA much as it does ordinarily at the surface of its host cell. It is postulated that the reduced infectivity of T2 phage may have resulted from the ejaculation of its upon adsorbing to magnetite. Furthermore, the variability found in infectivity may be due to the way in which the phage is adsorbed. If the phage is adsorbed by the tail to the magnetite, then ejaculation and inactivation may occur. However, if adsorption occurs on the phage head, then inactivation does not occur. Finally, it can be said that viruses adsorbed to magnetite ate infective. Moore et al. (1975) have studied the association of poliovirus and phages-r2 T7 and f2 with clays and suspended solids. They found that all the viruses tested, with the exception of f2 phage, were infective in the adsorbed state. In our study, we found that the tailless RNA phage, a similar behavior as poliovirus type 1 and these two viruses were 100% infective in the adsorbed state. In view of these findings, it should be clear that the safe treatment and disposal of virusladen magnetite is critical in controlling the spread of disease. Elution of Poliovirus from Magnetite Prior to initiating experiments on the concentration of poliovirus by magnetic filtration, it was important to determine which solution would be best for the elution of poliovirus from magnetite. Table 20 shows the eluents used and the percent

PAGE 64

Table 19. Infectivity of bacteriophage T2, bacteriophage MS2 and Poliovirus Type I (Sabin) adsorbed to magnetite Virus a Direct Plate Direct Plate @ T = 0 min @ T = 20 min (pfu/ml) (pfu/ml) T2 phage 1190 700 790 740 MS2 phage 717 800 % of Adsorbed Virus 57.8 93.5 >100.0 Supernatant (pfu/ml) 30 20 o % Virus Adsorbed c 97.5 97.5 100.0 Poliovirus 1200 1335 >100.0 13 98.9 a b c The bacteriophage T2, bacteriophage MS2 and Poliovirus were each shaken 170 oscillations/ min) for 20 min in the presence of 500 ppm of magnetite and 1610 ppm of CaC12 Samples were removed and assayed by direct plating (no dilutions made) immediately after addition of the virus (T = 0 min) and after the adsorption period (T = 20 min). Following the adsorption period, the magnetite was pelleted by placing the mixtures next to a magnet and then the supernatants were assayed for virus. Controls without magnetite were also studied and they showed no significant inactivation of the three viruses during the 20 minutes of shaking. The % infectivity of adsorbed virus was determined using the following equation: ttiectivit' = Direct Plate @ T = 20 (Efu/ml) SUEernatant (Efu/ml) X 100 Y Direct Plate @ T = 0 (pfu/ml) -Supernatant (pfu/ml) The % virus adsorbed is based on the percentage of the direct plate at T = 0 min. I CJ1 -1>0 I

PAGE 65

-55-Table 20. Elution of Poliovirus Type I (Sabin) from the magnetite surface Eluentb Used Distilled Water 1610 ppm CaC12 Tris Buffer, pH = 9 3%, isotonic Beef Extract, isotonic buffered, pH = 9 10%, isotonic Fetal Calf Serum, isotonic buffered, pH = 9 Eagles Minimal Essential Medium (MEM) + 5% Calf Serum + 0.03 M Hepes Buffer, pH = 7.0 Glycine Buffer, 0.05 M, pH = 11.3 Dome Water,d Tris buffered, pH = 9 % RecoveryC of Adsorbed Poliovirusa 7.1 l.8 9.9 3.2 59.lB 39.5C 8.8 a The Poliovirus was shaken (rv 200 oscillations/min) for 20 min in the presence of 500 ppm of magnetite and 1610 ppm of CaC12. Aliquots were then removed and the magnetite was pelleted by cpiLacing the mixture next to a magnet. Supernatants were assayed and showed that 99% of the initial Poliovirus concentration (1.0 x 10 4 pfu/ml) was adsorbed to magnetite. b c d Elution was undertaken as follows: the magnetite pellets were resuspended in the indicated eluents for 1 hour at 40C (except for glycine buffer which was allowed only 1 min of contact time), were then pelleted again and finally, the supernatants were assayed for viruses. The % recovery values displaying different superscript capital letters are significantly different at the 0.01 level for a twotailed test. If no superscript capital letter appears, then a test of significance was not performed. The dome water was sampled at a cypress dome located north of Gainesville, Florida.

PAGE 66

-56-recoveries obtained. The composition and source of the eluents used appear in the Appendix. The 1610 ppm CaC12 solution was the medium in which adsorption initially took place and. as expected, the re.c(i)weTyof adsorbed poliovirus using this solution as eluent was very low (108%). The 'best eiluent was 10% solution of fetal calf serum, Tris buffered, pH = 9 and it enabled the desorption of 77.9% of poliovirus from the magnetite surface (Table 20). These results are not surprising since Rao et al. (1968) also used a similar solution to effectively elute viruses from an iron oxide. However, the low recovery obtained with an isotonic, 3% solution of beef extract, Tris buffered at pH 9 (3.2% recovery) was not expected. This solution has been shown to be an efficient eluent when desorbing viruses from an iron oxide;surface (Pearson and Metcalf, 1974; Rao et al., 1968). Rao et al. (1968) has stated that not all lots of beef-extract virus eluting capacity. Thus,it appears that the lot of beef extract used in this study probably had a low virus eluting capacity. Generally, the best eluents are proteinaceous materials adjusted to an alkaline pH (8-9). In many cases, good eluents also effectively interfere with the initial adsorption of viruses to surfaces. Dome water, which had previously been shown to interfere with the adsorption of poliovirus to magnetite, was studied to determine its efficiency in eluting poliovirus from magnetite. Table 20 shows that dome water, Tris buffered, pH = 9 could only elute 8.8% of adsorbed poliovirus. Therefore, in this case, a good interfering substance (dome water) was shown not to be a good eluent. Experiments with glycine buffer, 0.05M, pH = 11.3 were performed allowing only 1 minute of contact time (see Table 20). This was done in order to avoid the inactivation of the virus due to the high pH. In fact, when 1 hour of contact time was allowed, no infectivity could be detected in the eluate as compared to 39.5% recovery of adsorbed poliovirus when the contact time was 1 minute. From the results presented, it was decided to use the 10%, isotonic fetal calf serum, Tris buffered, pH = 9 solution as the eluent in the concentration experiment described in the next section. Concentration of Poliovirus by Magnetic Filtration Viruses occur in natural water bodies in very low concentrations and, therefore, their detection is dependent upon their adequate concentration. Various virus adsorbents have been used to effectively concentrate viruses and these have been discussed in the literature review section. Since magnetite can under conditions, over 99% of poliovirus, we have carried out a preliminary experiment to investigate the application of magnetic filtration to the concentration of viruses. Table 21 shows the results obtained when poliovirus was conctmtTated using this technique. We studied the recovery of viruses under various conditions and it

PAGE 67

Table 21. Concentration of Poliovirus Type I (Sabin) by magnetic filtration Experiment No. Initiala Virus Conc. (pfu/ml) Volume of Sample Fil trate Virus Conc. (pfu/ml) % RecoveryC Concentration Factor 1 2 3 a b c 1.0 x 103 4 liters o 1.3 x 103 4 ilib:rs 17 1.2 x 103 4 liters o Hi t tpflsS g;:): 0.6 2nd pass: 4.S 3rdpass: 6.5 1 min: 11.7 5vernight: 50.7 >100.0 40 so 40 Recovery of Poliovirus from 4 liters of distilled water containing the initial virus concentrations shown was undertaken as described below. Poliovirus was shaken 200 oscillations/min) for 20 min in the presence of 1000 ppm of magnetite and 1610 ppm of CaC12' Tile.: mixtures were subsequently poured through the magnetic filter in order to separate the from the water suspension. Poliovirus was then eluted from surface with 100 ml (or 50 ml) of 10% isotonic fetal calf serum, Tris buffered, pH = 9. The procedure for eluting the Poliovirus varied with each experiment and appears in the Materials and Methods section. The headings 1st pass, 2nd pass, 3rd pass, 1 min and overnight refer to when the eluate was assayed for viruses. In experiment 1, the eluate was assayed after the 1st, 2nd, and 3rd pass through the filter. In experiment 2, the eluate was assayed after shaking manually for 1 min and after shaking mechanically overnight at 4 oC. The % recovery values were obtained by using the following equation: (7) 9,: R .. -Eluate Virus Conc. (pfu/ml) 1 x 100 o ecovery [Initial Virus Conc, (pfu/mI9 .'" Filtrate Vbms onc. (pfu/ml)] x I U1 -....] I

PAGE 68

-58-was found that the number of recovered virus particles from 4 li ters of water increased with the degree of contact between the eluent used and the magnetite-virus complex. For example, in experiment 1 (see Table 21), recovery increased with the number of passes of the eluent through the filter. Since the flow rate through the filter was very high, there was not enough time for effective contact between eluent and virus. Consequently, the highest recovery achieved was 6.5%. In experiment 2 ('iti'alble 21), by adding the eluent into the filter and shaking, the contact was increased that of experiment 1 and the re achieved increased accordingly to 11.7% and 50.7% for 1 minute shaking and overnight shaking, respectively. Sonica-tion has been used effectively to desorb microorganisms, including viruses, from surfaces (Puleo et al., 1967; Tschider et al., 1974). In experiment 3 (Table of the matrix-afforded enough contact between eluent and virus such that a recovery of >100.0% was obtained. The elevated titer (>100.0%) cOllll41 have resulted from disaggregation of the virus as explained before or simply from plating error. The concentration factors in the experiments described above were 40 and 80 and the initial virus concentrations ranged from 1.0 x 103 to 1.3 x 103 pfu/ml. In order to fully assess the effectiveness of magnetic filtration in concentrating viruses, the concentration of lower initial virus titers (e.g., <5 pfu/ml) in natural waters should be attempted. Finally, it can be said that magnetic filtration appears promising as a concentrating technique. It is an easy procedure and since it does not require a long time to perform, inactivation of virus does not occur. CONCLUSIONS Based on the findings of this investigation, the following conclusions may be drawn: 1. The adsorption of bacteriophage T2, bacteriophage MS2 and poliovirus type 1 (Sabin) to magnetite, and the subsequent removal by magnetic filtration increased in the presence of cations. Furthermore, on a concentration or molar basis, trivalent cations were more efficient than divalent cations which in turn were more efilio:l:ebttltihJ:JawlQIWln(1)\llaJllmtT1$l.es. 2. The adsorption of the three viruses to magnetite remained constant in the pH range of 5 to 9. 3. Removal by magnetic filtration in the 98-99% level was achieved with as little as 300 ppm of magnetite for poliovirus and MS2 phage, and 500 ppm of magnetite for T2 phage. 4. The adsorption of the three viruses to magnetite conformed to the Freundlich adsorption isotherm. However, the isotherm for

PAGE 69

-59-the T2 phage was below the isotherms for poliovirus and MS2 phage. This indicates that less T2 phage is adsorbed per mg of magnetite due to the larger size of this virus in comparison to poliovirus or MS2 phage. 5, Dhteraction between magnetite and virus was hindered in a highly colored wastewater effluent. However, this interference was reduced by the addition of CaC12, by the dilution of the wastewater with distilled water or by removing the organic color with activated carbon. 6. The interference by an activated sludge effluent on the removal of poliovirus by magnetic filtration was slight. This interference was completely eliminated by the addition of 1610 ppm of CaC12 (99.4% removal). This effluent is more typical than the highly colored effluent described above if one the effect of wastewater effluents upon the removal of viruses by magnetic filtration. 7. Competition for adsorption was also exerted by 100 ppm of egg albumin, casein and dextran resulting in a significant reduction in the removal of poliovirus. This finding is not unexpected since 100 ppm of these organic materials is an extremely high concentration. 8. Poliovirus and MS2 phage, adsorbed to magnetite, were 100% infective. The infectivity of adsorbed T2 was lower than that of poliovirus or MS2 phage and much more variable. The reduced infectivity of T2 phage may result from the ejaculation of its DNA upon adsorbing to magnetite. 9. From the results of the infectivity of adsorbed virus, the Freundlich isotherms and other experiments, it is clear that MS2 phage is a better model of poliovirus by magnetic filtration than T2 phage. Consequently, it is recom mended that future research in this area, involving only a bacteriophage, should be performed with an RNA phage such as MS2. 10. Poliovirus was effectively eluted from the magnetite surface with an isotonic, 10% solution of fetal calf serum, Tris buffered, pH = 9 (77.9% recovery of adsorbed virus). 11. A 40-fold concentration of poliovirus was achieved using magnetic filtration. Recovery of the virus (initial concentration equal to 1.2 x 103 pfu/ml) from 4 liters of distilled water was >100.0% when sonication of the matrix, in the presence of the eluent described above, was undertaken. More research is needed on the concentration of lower virus titers in natural waters by this technique.

PAGE 70

-60-The experimental data obtained in this work show that magnetite is a good adsorbent towards viruses and that magnetic filtration can be effectively used for the removal of viruses from water and wastewater. Magnetic filtration also appears promising as a concentrating technique.

PAGE 71

-61ACKNOWLEDGMENTS The authors wish to thank Dr. William H. Morgan, Director of the Florida Water Resources Research Center, for his help and patience during the course of this investigation. We are also indebted to Michael Duke for his valuable help in this work.

PAGE 72

-62-REFERENCES Adams, M. H. 1959. Bacteriophages. Interscience Publishers, Inc., New York, N.Y. Amirhor, P. and Engelbrecht, R. S. 1974. Interaction of virus and polyelectrolyte in wastewater effluent. 7th Int. Conf. Water Poll. Res., Paris. Bachrach, H. L. and Schwerdt, C. E. Purification studies on Lansing poliomyelitis virus: pH stability, CNS extraction and butanol purification experiments. J. Immun., 69:551-561. Bartell, P., Pierzchala, W. and Tint, H. 1960. The adsorption of enterioviruses by activated attapulgite. J. Am. Pharm. Assoc. Sci. 49:1-4. Berg, G. 1973a. Removal of viruses from sewage, effluents and waters. 1. A review. Bull. WId. Hlth. Org., 49:451-460. Berg, G. 1973b. Removal of viruses from sewage, effluents and waters. 2. Present and future trends. Bull. WId. Hlth. Org., 49:461-469. Berg, G. 1975. Microbiology-detection, occurrence, and removal of viruses. J. Wat. Pollute Control Fed., 47:1587-1595. Bitton, G 1975. Adsorption of viruses onto surfaces in soil and water. Water Res., 9:473-484. Bitton, G., Fox, J. L. and Strickland, H. C. 1975. Removal of algae from Florida lakes by magnetic filtration. Appl. Microbiol., 30:905-908. Bitton, G., Masterson, N. and Gifford, G. E. 1976. Effect of a secondary treated effluent on the movement of viruses through a cypress dome soil. J. Env. Qual. (in press). Bitton, G. and Mitchell, R. 1974. The removal of Escherichia coli bacteriophage T7 by magnetic filtration. Water Res., Bitton, G., Mitchell, R., De Latour, C. and Maxwell, E. 1974. Phosphate removal by magnetic filtration. Water Res., Boche, R. D. and Quilligan, J. J., Jr. 1966. Adsorption to glass and specific antibody inhibition of l25I-labeled influenza virus. J. Immun., 97:942-950.

PAGE 73

-63-Brown, T. S., Malina, J. F., Jr. and Moore, B. D. 1974a. Virus removal by diammaceous-earth filtration -part 1. J. Am. Wat. Wks. Assoc., 66:98-102. Brown, T. S., Malina, J. F., Jr. and Moore, B. D. 1974b. Virus removal by diatomaceous-earth filtration -part 2. J. Am. Wat. Wks. Assoc., 66:735-738. Carlson, G. F., Jr., Woodard, F. E., Wentworth, D. F. and Sproul, O. J. 1968. Virus inactivation on clay particles in natural waters. J. Wat. Pollut. Control Fed., 40:R-89-l06. Chaudhuri, M. and Engelbrecht, R. S. 1972. Removal of viruses from water by chemical coagulation and flocculation. J. Am. Wat. Wks. Assoc., Clark, J. W,i;l Viessman, W., Jr. and Hammer, M. J. 1971. Water Supply and Pollution Control, 2nd ed., International Textbook Company, N.Y. Clarke, N. A., Stevenson, R. E., Chang, S. L. and Kabler, P. W. 1961. Removal of enteric viruses from sewage by activated sludge treatment. Am. J. Publ. Hlth., Cookson, J. T., Jr. 1967. Adsorption of viruses on activated carbon. Adsorption of Escherichia coli bacteriophage T4 on activated carbon as a diffusion-limited process. Environ. Sci. Tech., 157-160. Cookson, J. T., Jr. 1969. Mechanism of virus adsorption on activated carbon. J. Am. Wat. Wks. Assoc., Cookson, J. T., Jr. and North, W. J. 1967. Adsorption of viruses on activated carbon. Equilibria and kinetics of the attachment of Escherichia coli bacteriophage T4 on activated carbon. Environ. Sci. Tech., 1 :46-52. Davis, B. D., Dulbecco, R., Eisen, H. N., Ginsberg, H. S. and Wood, W. B. 1973. Microbiology. Harper and Row, Hagerstown, Maryland. De Latour, C. 1973. Magnetic separation in water pollution control. IEEE Transactions on Magnetics, MAG-9:3l4-3l6. De Latour, C. and Kolm, H. 1976. High gradient magnetic separation: A water treatment alternative. J. Am. Wat. Wks. Assoc., Dieterich, B. H. 1953. The behavior of bacteria virus in contact with ordinary and uniform filter sand. Master's thesis, Harvard Univ., Cambridge, Mass.

PAGE 74

-64-Drewry, W. A. and Eliassen, R. 1968. Virus movement in groundwater. J. Wat. Pollut. Control Fed., 40:R257-27l. Fair, G. M., Geyer, J. C. and Okun, D. A. 1968. Water and Wastewater Engineering, Vol. 2. Water purification and wastewater treatment and disposal. John Wiley and Sons, Inc. New York. Flaig, W. 1960. Chemie der humusstoffe. Acta Chem. Fenn. A, 33:229-251. Gerba, C. P., Sobsey, M. D., Wallis, C. and Melnick, J. L. 1975. Adsorption of poliovirus onto activated carbon in wastewater. Environ. Sci. Tech., Gilcreas, F. W. and Kelly, S. M. 1955. Relation of coliformorganism test to enteric virus pollution. J. Am. Wat. Wks. Assoc., 47:683-694. Grabow, W. O. K. 1968. The virology of wastewater treatment. Water Res., l:675-70l. Hill, W. F., Jr., Akin, E. W. and Benton, W. H. 1971. Detection of viruses in water: A review of methods and application. Water Res., Hill, W. F., Jr., Akin, E. W., Benton, W. H., Mayhew, C. J. and Jakubowski, W. 1974. Apparatus for conditioning unlimited quantities of finished waters for enteric virus detection. Appl. Microbiol., 27:1177-1178. Kelland, D. R. 1973. High gradient magnetic separation applied to mineral beneficiation. IEEE Transactions on Magnetics, MAG-9: 307-310. Kellenberger, E. 1962. Vegetative bacteriophage and the maturation of the virus particles. Adv. Virus Res., Lauffer, M. A. and Bendet, I. J. 1962. Comments on "Bio-physical properties of bacteriophage T2'" Biochim. Biophys. Acta, 55:211-214. Lefler, E. and Kott, Y. 1974. Virus retention and survival in sand. In Virus Survival in Water and Wastewater Systems, edited by Malina, J. F., Jr. and Sagik, B. P., Center for Research in Water Resources, University of Texas, Austin. LoGrippo, G. A. 1950. anion exchange resin. Partial purification of viruses with an Proc. Soc. Exp. BioI. Med., 74:208-211.

PAGE 75

-65-Malherbe, H. H. and Strickland-Cholmley, M. 1967. Survival of viruses in the presence of algae. In Transmission of Viruses by the Water Route, edited by Berg, G., Interscience, New York. Mandel, B. 1971. Characterization of type 1 poliovirus by electrophoretic analysis. Virology, 44:554-568. Mitchell, R., Bitton, De Latour, C. and Maxwell, E. 1974. Magnetic separation: A new approach to water and waste treatment. 7th Int. Conf. Water Poll. Res., Paris. Mitchell, R., Bitton, G. and Oberteuffer, J. A. 1975a. High gradient magnetic filtration of magnetic and non-magnetic contaminants from water. Separation and Purification Methods, Mitchell, R., Chet, I., Bitton, G., Oberteuffer, J. and Harland, J. R. 1975b. High gradient magnetic separation of microorganisms and other non-magnetic suspended solids from waste water. 48th Annual Conf. Water Poll. Cont. Fed.; Miami Beach, Florida. Moore, B. E., Sagik, B. P. and Malina, J. F., Jr. 1975. Viral association with suspended solids. Water Res., Muller, R. H. and Rose, H. M. 1952. Concentration of influenza virus (PR8 strain) by a cation exchange resin. Proc. Soc. Exp. BioI. Med., 27-29. Oberteuffer, J. A. 1973. High gradient magnetic separation. IEEE Transactions on Magnetics, MAG-9:303-306. Oder, R. R. 1973. In High Gradient Magnetic Separation Symposium, edited by Oberteuffer, J. A. and Kelland, D. Cambridge, Mass. Osipow, L. I. 1962. Surface ahemistry. Robert E. Krieger Publish ing Company, Huntington, N. Y. Overby, L. R., Barlow, G. H., Doi, R. H., Jacob, M. and Spiegelman, S. 1966. Comparison of two serologically distinct ribonucleic acid bacteriophages. J. Bacteriol., 91:442-448. Oza, P. P. and Chaudhuri, M. 1975. Removal of viruses from water by sorption on coal. Water Res., Oza, P. P., Sriramulu, N. and Chaudhuri, M. 1973. Preliminary investigation on the use of coal for removing viruses from water. Proc. Symp. on Environmental Pollution, edited by Saraf, R. K., CPHERI, Nagpur, India.

PAGE 76

-66Pearson, F. and Metcalf, T. G. 1974. The use of magnetic iron oxide for recovery of virus from water. Completion report, Project No. A-014-NH, under Grant No. Water Resources Research Center, Univ. of New Hampshire, Durham. Pollard, E. C. 1953. The Physics of Viruses. Academic Press, New York. Prakash, A. and Rashid, M. A. 1968. Influence of humic substances on the growth of marine phytoplankton: Dinoflagellates. Limnol. Oceanog.,1:1..:598-606. Puck, T. T. and Sagik, B. 1953. Virus and cell interaction with ion exchangers. J. Exp. Med., Puleo, J. R., Favero, M. S. and Tritz, G. J. 1967. Feasibility of using ultrasonics for removing viable microorganisms from surfaces. Contamination Control Journal, Putnam, F. W. 1953. Bacteriophages: nature and reproduction. Adv. Protein Chem., 8:175-284. Rao, V. C., Sullivan, R., Read, R. B. and Clarke, N. A. 1968. A simple method for concentrating and detecting viruses in water. J. Am. Wat. Wks. Assoc;. 60:1288:-1294. Reece, R. E. 1967. Virus sorption on natural soils. Master's thesis, University of Arkansas, Fayetteville. Robeck, G. G., Clarke, N. A. and Dostal, K. A. 1962. Effectiveness of water treatment processes in virus removal. J. Am. Wat. Wks. Assoc., 54:1275-1290. Schaub, S. A. and Sagik, B. P. 1975. Association of enterioviruses with natural and artificially introduced colloidal solids in water and infectivity of solids-associated virions. Appl. Microbiol., 30:212-222. Schaub, S. A., Sorber, C. A. and Taylor, G. W. 1974. The association of enteric viruses with natural turbidity in the aquatic environment. In Virus Survival in Water and Wastewater Systems, edited by Malina, J. F., Jr. and Sagik, B. P., Center for Research in Water Resources, University of Texas, Austin. Schwerdt, C. E. and Schaffer, F. L. properties of purified poliomyelitis N.Y. Acad. Sci., 1955. Some physical and chemical virus preparations. Annals Sharp, D. G., Hook, A. E., Taylor, A. R., Beard, D. and Beard, J. W. 1946. Sedimentation characters and pH stability of the T2 bacteriophage of Escherichia coli. J. BioI. Chem., 165:259-270.

PAGE 77

-67-Shepard, C. C. and Woodend, W. G. 1951. The adsorptive behavior of bacteriophage T2 with certain inorganic materials. J. Immun., 66:385-393. Shuval, H. I., Cymbalista, S., Fatal, B. and Goldblum, N. 1967. Concentration of enteric viruses in water by hydro-extraction and two-phase separation. In Transmission of Viruses by the Water Route, edited by Berg, G., Interscience, New York. Sobsey, M. D. and Cooper, R. C. 1973. Enteric virus survival in algal-bacterial wastewater treatment systems-I-Laboratory studies. Water Res., 1:669-685. Sproul, O. 19WihnNirus inactivation by water treatment. J. Am. Wat. Wks. Assoc., 64:31-35. Sproul, O. J., Warner, M., LaRochelle, L. R. and Brunner, D. R. 1969. Virus removal by adsorption in wastewater treatment processes. Adv. Water Pollut. Res., 14th Int. Conf. Water Pollut. Res., Prague. Taylor, N. W., Epstein, H. T. and Lauffer, M. A. 1955. The particle weight, hydration and shape of the T2 bacteriophage of Escherichia coli. J. Am. Chem. Soc., 77:1270-1273. Thorup, R. T., Nixon, F. P., Wentworth, D. F. and Sproul, O. J. 1970. Virus removal by coagulation with polyelectrolytes. J. Am. Wat. Wks. Assoc., 62:97-101. Tolmach, L. J. 1957. Attachment and penetration of cells by viruses. Adv. Virus Res., !:63l-ll0. Trindade, S. C. and Kolm, H. H. 1973. Magnetic desulfurization of coal. IEEE Transactions on Magnetics, MAG-9:3l0-3l3. Tschider, S. R., Berryhill, D. L. and Schipper, I. A. 1974. Membrane concentration of infectious bovine rhinotracheitis virus from water. Appl. Microbiol., Valentine, R. C. and Allison, A. C. 1959. Virus particle adsorption. I. Theory of adsorption and experiments on the attachment of particles to non-biological surfaces. Biochim. Biophys. Acta, 34:10-23. Wallis, C. and Melnick, J. L. 1967a. Concentration of enteroviruses on membrane filters. J. Virology, 1:472-477. Wallis, C. and Melnick, J. L. 1967b. Concentration of viruses from sewage by adsorption on Millipore membranes. Bull. WId. Hlth. Org., 36:219-225.

PAGE 78

-68-Wallis, C. and Melnick, J. L. 1967c. Concentration of viruses on aluminum and calcium salts. Amer. J. Epidemio1., Wallis, C. and Melnick, J. L. 1967d. Concentration of viruses on aluminum phosphate and aluminum hydroxide precipitates. In Transmission of Viruses by the Water Route, edited by Berg, G., Interscience, New York. Wallis, C., Melnick, J. L. and Fields, J. E. 1970. Detection of viruses in large volumes of natural waters by concentration on insoluble po1ye1ectro1ytes. Water Res., .!: 787-796. Wallis, C., Melnick, J. L. and Fields, J. E. 1971. Concentration and purification of viruses by adsorption to and elution from insoluble polye1ectro1ytes. App1. Microbio1., 703-709. Warren, J., Neal, A. and Rennels, D. 1966. Adsorption of myxoviruses on magnetic iron oxides. Proc. Soc. Exp. Bio1. Med., 12: 1250-1253. Williams, R. C. and Fraser, D. 1953. Morphology of the seven T-bacteriophages. J. Bacterio1., 66:458-464.

PAGE 79

-69-APPENDIX COMPOSITION OF MEDIA AND SOLUTIONS USED Media Used in T2 and T4r(II) Phage Work All of these media were described by Adams (1959). 1. Basal medium for phage plating: 8 grams nutrient broth, dehydrated (Difco) 5 grams NaCl (Fisher) Make up to 1 liter with distilled water. 2. Nutrient agar for plates: Solution 1 above plus 15 grams per liter of Bacto-Agar (Difco). Between 25 and 30 ml are poured per plate and the plates are dried in an incubator at 370C overnight before use. 3. Soft agar overlay: Solution 1 above plus 7 grams per liter of Bacto-Agar (Difco). This solution is distributed in 2.5 ml amounts to test tubes. Prior to assaying, the soft agar is melted in a boiling water bath and then held in a 460C water bath. 4. M-9 medium for storing phages: Solution I: 3 grams KH2P04 (Fisher) 6 grams Na2HP04' anhydrous (Fisher) 1 gram NH4Cl (Fisher) Make up to 900 ml with distilled water. Solution II: 4 grams glucose (Fisher as dextrose) 100 ml of distilled water. Solution III. 1.3 grams MgS04 (Fisher) 100 ml of distilled water. Autoclave each solution separately. Prior to use, add sterilely 9 parts of solution I to one part of Solution II and 0.1 part of Solution III. 5. Agar slants for maintaining the host Escherichia coli B: 2 grams nutrient agar (Difco) 100 ml distilled water. Distribute 5 ml into 20 test tubes and autoclave. Allow to harden slanted after autoclaving. All media described above were kept refrigerated at 40C and usually used within 2 days of preparing.

PAGE 80

-70Media Used in MS2 Phage Work The L medium, L agar and L soft agar were described by Overby et al. (1966). 1. L medium: 2. 10 grams tryptone (Difco) 1 gram glucose (Fisher as dextrose) 5 grams yeast extract (Difco) 10 grams NaCl (Fisher) 0.22 grams CaC12 (Fisher) Make up to 1 liter with distilled water. L agar for plates: L medium above plus Bacto-Agar (Difco). poured per plate and an incubator at 370C 15 grams per liter of Between 25 and 30 ml are the plates are dried in overnight before use. 3. L soft agar overlay: L medium above plus 10 grams per liter of Bacto-Agar (Difco). This solution is distributed in 2.5 ml amounts to test tubes. Prior to assaying, the soft agar is melted in a boiling water bath and then held in a 460C water bath. 4. Tris buffer (0.05 M Tris, 0.1 M NaCl, pH 7.6) for suspending (storing) phages as recommended by Miles Laboratories: 1.39 grams Trizma base (Sigma) 6.06 grams Trizma HCl (Sigma) 5.85 grams NaCl (Fisher) Make up to 1 liter with distilled water. The resulting pH is 7.6 at 250C. The host Escherichia coli C3000 was maintained on 2% nutrient agar (Difco) slants as described for E. coli B. All media described above were kept refrigerated at 40C and usually used within 2 days of preparing. Media Used in Poliovirus Type I (Sabin) Work 1. Gey's Balanced (BSS) is the common diluent for cell culture: Gey's A (lOx): 70 grams NaCl 3.7 grams KCl 3.01 grams Na2HP04 l2H20 0.237 grams KH2P04 lQQ ml 0.1% phenol red 10 grams glucose 900 ml glass distilled water 5 ml chloroform as a preservative

PAGE 81

-71-This stock solution of Gey's A is stored at room temperature unautoclaved and is diluted 1:10 and autoclaved when needed. Gey's B (20x): 0.42 grams MgC12 6H20 0.14 grams MgS04 7H20 0.34 grams CaC12 100 ml glass distilled water Gey'sC (20x): 2.25 grams NaHC03 100 ml glass distilled water Bubble C02 into Gey's C until pH is less than 7.6. Dispense and tightly cap. Gey's Band Care autoclaved without further dilution. To make the complete Gey's Balanced Salt Solution (BSS) add: 90 parts Gey's A (Ix) 5 parts Gey's B (20x) 5 parts Gey's C (20x) 2. Hepes buffer (1 M) stock solution: 47.7 grams Hepes 190 ml Gey's A (Ix) 10 ml Gey's B (20x) 16 ml 2 M NaOH (8g/l00 ml) Dispense and autoclave. 3. Streptomycin-penicillin (lOOOx) stock solution: Solution I: 1.0 gram streptomycin 8 ml Gey's A (Ix) Solution II: 106 units of penitillin 4 ml of Solution I. Solution II contains 125 mg of streptomycin and 2.5 x 105 uni ts of penicillin per ml which is 1000x of what is required. Therefore, it must be diluted 1:1000 in the final solution. 4. Eagle's Minimal Essential Medium (MEM) using Gey's BSS + 10% fetal calf serum: 300 ml Gey's A (Ix) 20f)ml Gey's B (20x) 20 ml Gey's C (20x) 8 ml MEM essential amino acids (SOx) (International Scientific, Cary, Ill.)

PAGE 82

-72-4 ml vitamins (lOOx) (International Sci.) 4 ml glutamine (lOOx) (International Sci.) 0.4 ml strep.-pen. stock (lOOOx) 40 ml fetal calf serum (International Sci.) 5. Eagle's MEM using Gey's BSS + 5% calf serum + 0.03 M Hepes buffer at pH = 7: This solution was used to make most virus dilutions. The solution is made by substituting, in Solution 4 above, 20 ml of calf serum (International Scientific) + 12 ml of 1 M Hepes buffer stock solution + 8 ml of Gey's A (Ix) for 40 ml of fetal calf serum. 6. Solutions required for the removal of AV3 cells from glass (trypsinization): Solution I: Pre-trypsin wash: This solution removes all traces of serum (which contains trypsin inhibitors) as well as Ca++ and Mg++ions. 300 ml Gey's A (Ix) 5 ml Gey's C (20x) Dispense and autoclave. Solution II: 1% versene (EDTA) stock in Gey's A: 2.0 grams ethylene diamine tetraacetate (EDTA) 10 ml 2 M NaOH (8g/l00 ml) 20 ml Gey's A (lOx) 170 ml glass distilled water Dispense and autoclave. Solution III: 2.5% trypsin stock: 1.0 gram, trypsin (Difco, 1:250) 100 ml glass distilled water Sterilize by cold filtration. Dispense in 5 ml aliquots to screw-capped test tubes and freeze for storage. Solution IV: Standard versene-trypsin solution: This solution is used to remove the AV3 cells from the 32 ounce bottles in which they have been growing prior to their distribution to plaque bottles.

PAGE 83

-73100 ml Gey's A (Ix) 4 ml Gey's C (20x) 4 ml stock trypsin (2.5%) 4 ml stock versene (1% in Gey's A) This solution is good for only one day. 7. Methyl cellulose overlay for AV3 cells (1% methyl cellulose + 5% fetal calf serum) : Solution I: 300 ml glass distilled water 6 grams methyl cellulose (1500 centipoise) Autoclave and then allow to cool to room temperature, shaking vigorously every hour to avoid layering. Refrigerate. Solmtion II: 2x Eagle's MEM: 350 ml glass distilled water 120 ml Eagle's MEM (lOx) with Hanks' salts (International Scientific) 50 ml Gey's C (20x) 60 ml fetal calf serum (International Sci.) 25 ml Hepes buffer (1M) stock solution 12 ml glutamine (lOOx) (International Sci.) 1.2 ml strep.-pen. stock (lOOOx) Combine equal amounts of Solution I and II to make the methyl cellulose overlay. 8. Crystal violet: This stain is used to make the plaques on the cell monolayer visible to the naked eye. Solution I: 20 grams crystal violet 200 ml absolute ethanol Allow this solution to sit overnight. Solution II: 8 grams ammonium oxalate 800 ml distilled water Mix I and II and dilute 1:10 with tap water. Solutions Used for the Elution of Poliovirus from Magnetite The eluents used are listed below along with their composition. 1. Distilled water (Carlson et al., 1968; Puck and Sagik, 1953)

PAGE 84

-74-2. 1610 ppm CaC12 3. Tris buffer, pH = 9 0.076 g Trizma HCl (Sigma, St. Louis, Mo.) a .547 g Trizma base (Sigma) 100 ml distilled water 4. 3%, isotonic beef extract, Tris buffered, pH = 9 (Pearson and Metcalf, 1974; Rao et al., 1968) 0.076 g Trizma HCl (Sigma) 0.547 g Trizma base (Sigma) 0.87 g NaCl (Fisher) 3.0 g beef extract (Difco) 100 ml distilled water 5. Eagle's Minimal Essential Medium (MEM) + 5% calf serum + 0.03 M Hepes buffer, pH = 7.0 (Pearson and Metcalf, 1974) The composition of this solution can be found in this appendix in the section dealing with the media used in poliovirus work. 6. 10%, isotonic fetal calf serum, Tris buffered, pH = 9 (Rao et al., 1968; Wallis and Melnick, 1967a, 1967b; WalliS-et al., 1971; Wellings et 1974) 0.076 g Trizma HCl (Sigma) 0.547 g Trizma base (Sigma) 0.87 g NaCl (Fisher) 10 ml fetal calf serum (International Scientific, Inc.) 90 ml distilled water 7. Glycine buffer, 0.05 M, pH = 11.3 (Hill et al., 1974) Solution I: 0.375 g glycine (Fisher) 100 ml distilled water Solution II: 10 N NaOH prepared: 40 g NaOH (Fisher) 100 ml distilled water Within 2 hours of using, add 0.54 ml of Solution II to 100 ml of Solution I. 8. Dome water, Tris buffered, pH = 9

PAGE 85

" -750.076 g Trizma Hel (Sigma) 0.547 g Trizma base (Sigma) 100 ml dome water (see composition in Table 2) All the solution above were sterilized by autoclaving. Serum, when required, was added stl'1dlely after autoclaving.