United States
Environmental Protection
Agency
Municipal Environmental Research EPA-600/2-80-1 34
Laboratory ; August 1980
Cincinnati OH 45268
Research and Development
Physical Chemistry of
Virus Adsorption and
Degradation on
Inorganic Surfaces
Its Relation to
Wastewater Treatment
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields,
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution-sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
PHYSICAL CHEMISTRY OF
VIRUS ADSORPTION AND DEGRADATION ON INORGANIC SURFACES
Its Relation to Wastewater Treatment
by
James P. Murray
Stanford University
Stanford, California 94305
Grant No. R-805016
Project Officer
Albert D. Venosa
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental iProtection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement or recommenda-
tion for use.
11
-------
FOREWORD
The U.S. Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health and
welfare of the American people. Noxious air, foul water, and spoiled land are
tragic testimonies to the deterioration of our natural environment. The com-
plexity of that environment and the interplay of its components require a
concentrated and integrated attack on the problem. !
Research and development is that necessary first step in problem solution;
it involves defining the problem, measuring its impact, and searching for solu-
tions. The Municipal Environmental Research Laboratory develops new and
improved technology and systems to prevent, treat, and manage wastewater and
solid and hazardous waste pollutant discharges from municipal and community
sources, to preserve and treat public drinking water supplies, and to minimize
the adverse economic, social, health, and aesthetic effects of pollution. This
publication is one of the products of that research and provides a most vital
communications link between the researcher and the user community.
This report presents a detailed fundamental investigation of the mechanisms
that control adsorption of viruses to mineral surfaces in the environment. The
findings have important implications regarding which natural soil and aquatic
systems are conducive to viral dissemination and those that are not. In addi-
tion, application of the optimum adsorbent for wastewater purification and
water recycling was investigated in detail. Promising results were obtained
that could eventually lead to the development of a new nontoxicogenic water
reclamation process.
Francis T. Mayo
Director
Municipal Environmental
Laboratory
111
Research
-------
ABSTRACT
The DLVO-Lifshitz theory of colloid stability was applied to adsorption
of poliovirus on oxide surfaces common in soil and aquatic environments.
Excellent agreement was found between colloid stability theory and adsorption
free energies calculated from mass-action principles. Colloid stability
theory now provides an organized frame of reference with which to understand
virus adsorption in the environment. On some surfaces, notably 3-Mn09, CuO,
and Al metal, kinetic analysis of data from multiple extractions and sedi-
mentation analysis reveals that virus was actually degraded by adsorption.
A column experiment also indicated that Al metal also effectively degraded
virus in the presence of secondary wastewater effluent. Potential applica-
tions to wastewater treatment are suggested.
This report was submitted in fulfillment of Grant No. R-805016 by
Stanford University under the sponsorship of the U.S. Environmental Protection
Agency. This report covers the period September 1, 1976 to March 1, 1978.
IV
-------
CONTENTS
Foreword * . ill
Abstract iv
Figures ......... vi
Tables vii
Abbreviations and Symbols ix
Acknowledgements xii
1. Introduction . . 1
Proj ect Obj ectives . 1
Importance of Problem and Review of Previous Work. ... 1
2. Conclusions 4
3. Recommendations. 5.
4. Materials and Methods 6
Virus and Assay 6
Statistics of Assay Data 7
Buffers 8
Adsorbents 8
ChemicalSj Isotopes, and Analytical Methodology. ..... 9
5. Mechanism of Poliovirus Adsorption ...... 11
Thermodynamic Approach '..... 11
Determination of AGacjs 11
Application of DLVO-Lifshitz Theory 21
Electrostatic Components of AGads 22
Electrodynamic Componentsof AGa
-------
FIGURES ;
Number : Page
1 Adsorption and degradation apparatus 14
2 Adsorption-desorption isotherms of poliovirus on oxides at
0.02 ionic strength and pH 7 18
3 Adsorption-desorption isotherms of poliovirus on oxides at
0.305 ionic strength and pH 7 19
4 Adsorption-desorption isotherms of poliovirus on A1203 and
aluminum metal at 0.02 ionic strength and pH 7 • 20
5 Diagrammatic representation of double layers about virus and
solid „ . . 23
6 Variation of poliovirus adsorption with pH and a-Si02 at 0.02
ionic strength as predicted by DLVO theory ..... 34
7 Kinetic plot for a-Fe203 Sample C. 43
8 Kinetic plot for 3-Mn02 Sample C 45
9 Kinetic plot for CuO Sample C 46
10 Kinetic plot for a-Al203 Sample C . ...... 49
11 Kinetic plot for Al Metal Sample C ......... 50
12 Sedimentation profiles of virus degraded by Al metal 51
13 Schematic diagram of bench-scale process experiment 54
14 Photograph of bench-scale process experiment 55
15 Results from process experiment 57
VI
-------
TABLES
Number | Page
1 Surface Areas of Adsorbents !..... 16
2 Free Energies of Poliovirus Adsorption on Oxide Surfaces as
Determined by Mass-Action Equilibria ; 21
I
3 Normalized Concentration of Poliovirus Infectivity in
Equilibrium with Various Solids 22
4 Zeta Potentials of Oxides and Virus in pH 7 NaHC03-NaCl
Solutions 26
5 Comparison of Theoretical Isoelectric Points with Values
Reported for Other Enteroviruses . . . 27
6 Double Layer Interaction Potentials between Virus arid Solid
Phases According to the LSA Expression ........... 27
i
7 Comparison of Differences in AGads and Udi at 0.02 1 and
0.305 I I. . . . . ' 28
8 Individual Self-Interaction Hamaker Coefficients Taken from
Visser (65) and Calculated Mixed Coefficients ... |...... 31
9 Comparison of Lifshitz-Van der Waals Potentials to Residual
Free Energy Components after Correcting for Double-Layer
Repulsion j ..... 31
10 Comparison of Virus Initially Adsorbed (Adsi) with Virus
Recovered (Adsr) for a-Si02 j 41
11 Comparison of Virus Initially Adsorbed (Ads^) on a-Fe^O,
with Virus Recovered by Kinetic Analysis of Data from
Four Extractions CAdso) i 42
12 Comparison of Virus Initially Adsorbed (Ads^) on g-Mn02 with
Virus Recovered by Kinetic Analysis of Data from Four
Extractions CAdsQ). 44
13 Comparison of Virus Initially Adsorbed (Ads^) on CuO with
Virus Recovered by Kinetic Analysis of Data from Four
Extractions (Adso) • 47
VII
-------
Number Page
14 Comparison of Virus Initially Adsorbed (Adsj.) on a-Al203 with
Virus Recovered by Kinetic Analysis of Data from Four
Extractions (Adso) .... 48
15 Comparison of Virus Initially Adsorbed (Ads^) on Al Metal
with Virus Recovered by Kinetic Analysis of Data from Four
Extractions CAdso) . .' 52
16 Composition of Composite Secondary Effluent (Unchlorinated)
Obtained from Palo Alto Sewage Treatment Plant 56
Vlll
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
ACS
BET
CPM
COD
DLVO
ddd H20
g
Hz
I
kJ/mol
LSA
M
MEM
ND
PFU
SD
s
THM
U
SYMBOLS
All,22,33
A-132,131
Adsi
Adsn n_i
Adso
Ads
*v
^*6Q
CH<70
colloid stability
n
American Chemical Society
Brunauer, Emmett, and Teller isotherm for surface area measurement
counts per minute
chemical oxygen demand
Derjaguin, Landau, Verwey and Overbeek theory of <
double distilled deionized water
gram, gravitational unit (acceleration)
Hertz, cycles per second
ionic strength
kilojpules per mole
linear superposition constant charge assumption
molar
minimal essential medium
not determined
plaquerfprming unit
standard deviation
second
trihalpmethane
internalional unit, e.g. for concentration of penicillin
individual self-interaction Hamaker coefficients for Van der
Waals potential calculation
complex Hamaker coefficients '
virus initially adsorbed :•
virus adsorbed after extraction sequence n, n^l
virus recoverable with an infinite number.of extractions
virus recovered after elution from Si02
virus adsorbed after adsorption (1), desorption (2) in mechanism
experiments i
virion radius, poliovirus = 27 nm ,
background counts in scintillation counting
fitting coefficient ~0.32 x 10-16/s, for Bargema.n and Van
Voorst Vader theory j
virus concentration after elution from SiC>2
equilibrium concentration of virus in liquid phase
water concentration in solution, mol/kg
concentration entering column :
virus concentration in elution medium after extraction sequence n
IX
-------
c«n
Gout
c
X
es
AG
cov_ion
AGres
AGydW
AS0
At
e
el>2,3
F
Fads
F132
I1'2
k
*B
K
m
N
Ma
n
n_;n+
u>
WR
W132
PI
PFU/mllj2
PPR
PHPZC
R
s
a
T
Cn minus virus left over from previous extraction
concentration leaving column
virus concentration not .adsorbed after adsorption step for
degradation studies with SiC>2
velocity of electromagnetic radiation in vacuum
integration variable for Lifshitz function
characteristic equilibrium separation for electrodynamic forces
characteristic equilibrium separation distance for electro-
static interaction
free energy of adsorption-Gibbs function
covalent-ionic bond contribution to AGads
double layer contribution to AGads
residual energy after subtracting U^ from AGads
Van der Waals contribution to AGads :
entropy considerations in AGads not accounted for in other terms
time intervals between extraction sequences
electronic charge
zero frequency dielectric constant relative to vacuum
dielectric permittivity of free space ' ;
complex frequency dependent dielectric susceptibility function
of materials 1,2,3
real coefficient of ei,2,3
imaginary coefficient of e^ 2 3
Faraday's constant
fraction of virus adsorbed
electrodynamic induction force (Van der Waals) between materials
1 and 2, immersed in medium 3
tanh (yi32/4) / • '
imaginary modulus (-1)1/2
rate constant
Boltzmann's constant
Debye inverse length = (2e2nz2/eoevkBT)l/2
number of pipetting operations
number of counts or plaques in statistical evaluation-
Avogadro's number
extraction sequence number in degradation studies
anion or cation number (perm^) in double layer formulations
complex frequency WR + ic
real part of to
value of complex Lifshitz frequency integral
packing index virions per m2
1 refers to titer in adsorption supernatant, 2 to desorption
supernantant ;
virus particle to PFU ratio
pH where particle (virus) has no net charge
parts per billion
surface potential
Stern potential of particle 132
gas constant
specific surface area of adsorbent
theoretical standard deviation
temperature
x
-------
Tot
t
e
uVdW
Va
Vd
Ve
Vn
Vr
wa
wr
x
X
total virus in system
time
fraction of surface covered
double layer interaction potential energy
Van der Waals interaction potential energy
volume during adsorption step
volume during desorption step
volume during elution with Si02
volume during extraction sequence n
residual solution volume left over from adsorption step or
previous extraction
weight of adsorbent
weight of adsorbent remaining after initial
tion studies on Si02
water molecules desorbed per virion adsorbed
sequence number for final extraction
sampling in degrada-
in mechanism.studies
charge of anion or cation
zeta potential of substances 1,2
XI
-------
ACKNOWLEDGEMENTS
The first section of this report, which defines the mechanisms by which
poliovirus adsorbs to various materials, was the subject of the Ph.D. dis-
sertation of the author. In this regard, financial assistance by Mr. John L.
Murray, Dr. John R. Murray, Dr. H.H. Huang, and Sigma-Xi is acknowledged.
The U.S. Environmental Protection Agency supported the concluding phas^e of
the dissertation, degradation studies, and bench scale process experiments.
The advice and laboratory space provided by Drs. G.A. Parks, C.E.
Schwerdt, and P.L. McCarty are much appreciated. Dr. R.O. James, currently
at the C.S.I.R.O. Textile Physics Div., Ryde, N.S.W., Australia, assisted
with fundamental concepts of the DLVO-Lifshitz theory. Dr. B.D. Korarit and
K. Lonberg-Holm are thanked for permission to cite unpublished data. 1S.
Enriquez, S. Laband, M. Sanchez, E. Slack, and E. Terehoff are thanked for
technical assistance. Especial thanks belongs to the Division of Applied
Sciences, Ms. Ellen Campbell and Ms. Dot Franzosa of Harvard University
for assistance with final report preparation.
XI1
-------
SECTION 1
INTRODUCTION
PROJECT OBJECTIVES
The objectives of this investigation were to determine how potentially
pathogenic enteric viruses adsorb to mineral surfaces in
the environment, to
investigate and quantify actual degradation of virus by adsorption to various
surfaces, and to apply information obtained from the first two studies to-
wards development of a virus removal process applicable to water purification
and wastewater reclamation. i
,{
IMPORTANCE OF PROBLEM AND REVIEW OF PREVIOUS WORK j
!
A recent survey conducted by Dr. Martin Goldfield (1:) of the New Jersey
Department of Health indicates that viral disease, particularly viral gastro-
enteritis and infectious hepatitis, continues to be transmitted through con-
taminated water even in the United States. Virus transmission problems are
an active environmental concern. The massive epidemics of waterborne hepati-
tis in Delhi, India, in 1955, involving 35,000 cases (2) and of waterborne
Salmonella typhimurium in Riverside, Calif., in 1965, involving 16,000 cases
(3) point to potential hazards associated with waterborne, disease.
Problems associated with the problem of controlling viruses in water
have been reviewed by Dr. Gerald Berg of the U.S. Environmental Protection
Agency (4) . , i
Enteric viruses have been isolated from raw and treated domestic efflu-
ents and in relatively uncontaminated water (5) . Even low levels may be
considered significant because exceptionally low titers of virus, e.g. 2
plaque forming units (PFU) with poliovirus type 1, strain SM (6), have been
shown to infect human volunteers. The question of what levels of removal are
required to prevent transmission of disease by drinking water has yet to- be
defined . i
In soil and aquatic environments, adsorption of enteroviruses to various
mineral surfaces appears to be a major control on virus dispersal in the en-
vironment. Pioneering work on this topic was done at Stanford by Drewry and
Ryan (7,8). They showed that coliphages Tl, T2, and T5 were rapidly taken up
by soil and sand columns, and that soil columns, especially those made with
soils containing abundant clay minerals, favored virus removal when compared
to sand columns. They concluded that adsorption to soil mineral surfaces was
responsible for the removal, and observed that filter aid|s, high ionic
-------
strength (I), and coagulants enhanced removal.
Cookson and North (9,10) investigated adsorption of bacteriophage on
activated carbon by kinetic and surface esterification experiments. Their
work demonstrates the importance of electrostatic interactions in adsorp-
tion. Previous investigators (11) have suggested that Van der Waals inter-
actions might be involved in virus adsorption. This investigation has shown
that these interactions play a dominant role, vary markedly between different
substances, and can be understood from a physical perspective.
Other investigations on virus^ adsorption have been extensively reviewed
by Gerba .et jil. (11) . In general, low pH (3.0 to 6.0) favors adsorption
compared to high pH (8.0 to 10.0), and divalent cations favor adsorption com-
pared to monovalent cations. Organic material may compete with viruses
for adsorption sites, but in some cases, virus adsorption is still quite
strong even in the presence of high concentrations of organics. For example,
Y-F6203 effectively adsorbs respiratory syncytial virus from infected egg
allantoic fluids (12). Mitchell and Jannasch (13) and Smith et.al. (14) show
that adsorption of virus to suspended solids in seawater inhibits the rapid
decrease in titer found in controls not containing solids, possibly protecr-
ting the virus from spontaneous decay. In addition, Shaub and Sagik (15)
show that virions of mengo virus retain infectivity when adsorbed to bento-
nite. Our investigation has first demonstrated that adsorption of virus to
some inorganic surfaces actually degrades the virus into small fragments.
Problems associated with potential cancer risks to man by chlorinated
drinking water have only recently been recognized (16). Over 700 organic com-
pounds have so far been detected in water (17). Chlorination of these in
water treatment plants causes formation of trihalomethanes (THMs) and other
halogenated substances. A significant fraction of these compounds are muta-
genic (18), carcinogenic, teratogenic, and fetotoxic (17). Page et al. (17)
have recently estimated potential gastrointestinal and urinary tract cancer
mortalities for THMs with linear accumulated dose - animal testing models,
and also with epidemiologic associations. Estimates of 20 to 300 cancer
deaths annually per million population corresponded to a drinking water THM
concentration of 250 ppb. This level of contamination is higher than that
found in many municipal drinking water supplies, but is certainly within the
range reported by EPA. EPA's report to Congress (16) also shows that THM3
are far higher in chlorinated drinking water than in the same raw water
supplies, indicating that chloririation is responsible for most of the THMs
present. Risks imposed by other compounds produced by chlorination are not
included in animal dose extrapolations, but are probably included in esti-
mates based on epidemiologic associations.
The total adverse impact on human health and possibly the potential for
environmental quality degeneration caused by chlorination of water is clearly
significant. 1
An alternative disinfectant, ozone, has recently been shown to produce
mutagens in domestic wastewater effluent (19) . Adverse health effects could
be implicated here as well, although this cannot be stated with certainty
from Ames test results alone.
-------
It is clear that new disinfection processes must be designed which do
not increase the cancer risk of the water consumer. Discharge of chlorinated
wastewater to the environment might have other long-term deleterious effects
as well. It was in response to these needs that this investigation was ini-
tiated.
-------
SECTION 2
CONCLUSIONS
Using poliovirus type 1 as a model enterovirus, it was established with •
a wide range of materials, that adsorption is controlled by, the balance be-
tween electrostatic double layer interactions and electrodynamic Van der
Waals interactions. This control was demonstrated by showing that adsorption
free energies calculated from mass action arguments are in good agreement
with adsorption potentials calculated from the Derjaguin-Landau-Verwey-
Overbeek-Lifshitz (DLVO-Lifshitz) theory of colloid stability (20-22). Since
this virus is quite characteristic of enteroviruses in terms of size-:-27 nm
diameter, and geometry—icosahedral-spheroidal (23), electrokinetic proper-
ties of many types, and survival in the .environment (24-), these
adsorption studies should be representative of enteroviruses in general.
The DLVO-Lifshitz theory provides us a fundamental, organized frame of
reference in which to understand virus adsorption, and allows prediction of
the types of materials and solution conditions conducive to virus adsorption
and those that are not.
Some solids, e.g. Mn02, CuO,, and most notably, Al metal, degrade virus
during adsorption. Techniques for recovering all viruses from solids that
did not degrade them, e.g. Fe203,, A1203, may in the future, be applicable
for evaluating all virus adsorbed to other solids.
When information gained in mechanism and degradation studies was applied
to a bench scale system with a 1 cm bed of Al and 0.7 min residence time,
it was found that Al metal removes >99% of the input virus titer from secon-
dary effluent at steady-state. Al metal is substantially more effective than
granular activated carbon for the purpose of virus removal. It has good
potential for eventual use in advanced water and wastewater treatment systems.
-------
SECTION 3
RECOMMENDATIONS
It is recommended that future investigations of virus adsorption be
centered around the following objectives: ;
1. DLVO-Lifshitz theory should be the framework from which future
investigations of virus adsorption are conducted.
2. Investigation should continue in regards to role of conformational
states on virus adsorption. ''
3. Competition for adsorption sites by naturally occurring organics
deserves additional investigation, particularly !with aluminum metal
as the adsorbent. '
4. Differences in adsorption characteristics of different viruses needs
additional study. Adsorption and degradation of infectious hepati-
tis is especially important, because it is the most: significant
waterborne viral disease. This, of course, depends on the develop-
ment of a practical assay system for hepatitis, Iwhich is currently
not available. :
Additional investigation should be made in regards to the use of alumi-
num metal as an adsorbent in an adsorption-degradation unit process. An
engineering design and economic optimization study which determines how the
metal might most effectively be used to provide complete virus removal at
minimum cost is required prior to application in treatment plants
-------
SECTION 4
MATERIALS AND METHODS
VIRUS AND ASSAY
Plaque-purified poliovirus type 1, strain LSc2ab, obtained from Dr. F.
L. Schaffer, Naval Biomedical Laboratories, Oakland, California, was propa-
gated and assayed in rhinovirus adapted HeLa cells obtained from Dr. V.V.
Hamparian, Medical Microbiology, Ohio State University, Columbus, Ohio.
17
The virus was labeled by growing it in media containing H-uridine
[5,6 3H ] - 0.5 to 54 uci/ml, and 14C-amino acid mixture 0.4 to 15 vtci/ml,
obtained from New England Nuclear, in the presence of 0.5 to 1 ug/ml actino-
mycin D, gift of Merck, Sharp and Dohme Research Labs. For mechanism and
degradation experiments, the virus was purified of unincorporated label and
cellular debris by differential ultracentrifugation, extraction with C2C13F3,
sucrose gradient rate zonal sedimentation, isopycnic banding in a CsCl grad-
ient, and dialysis against saline-bicarbonate 0,02 I buffers.,In degrada-
tion experiments, preparations were filtered through 0.45- urn-membrane filters
(Millipore Corp.} after dialysis. For process experiments, virus was labeled
with mixed 14c-amino acids alone and required only partial purification by
C.,C1,F_ extraction and differential ultracentrifugation.
^* o o
Sucrose gradients of purified preparations showed that viruses were not
clumped (25) but existed as a suspension of discrete virions. Fresh prepara-
tions were used for all experiments. When a significant,decrease in the
infectivity of the stock was observed, the preparation was discarded and a
new one prepared. Fourteen batches were prepared for this study.
Ratios in absorbance of purified preparations at 260 to 280 nm varied
from 0.91 to 1.65, slightly less than the ideal ratio (26), 1.70 ± p.02.-
This indicates that if the preparations-are essentially free of contaminating
nucleic acids, a minor amount of protein impurity is present even after rigo-
rous purification.
Assays were done by the monolayer method of Dulbecco and Vogt as adapted
by McLain and Schwerdt (27). Cells were maintained in Eagle's minimal essen-
tial medium (MEM) with Earle's or Hanks' salts containing 400 U/ml penicillin
G and 100 Ug/ml streptomycin sulfate plus 10% fetal bovine serum. Media and
serum were obtained from Grand Island Biol. Co. and Microbiol. Assoc. For
degradation and process experiments, the penicillin concentration was reduced
to 100 U/ml.
-------
Where contamination was a severe problem, as in virus assay in the pres-
ence of secondary effluent, additional antibiotics were added to the assay
overlay medium. Concentrations were; penicillin G - 1000 U/ml, strepto-
mycin sulfate - 100 ug/ml, gentamycin sulfate - 40 yg/ml,; and amphotericin B
- 2.5 yg/ml. No effect of these high concentrations on titer was seen in
controls. ;
Serial dilutions were prepared in Hanks' balanced salt solution, from
Grand Island Biol. Co., plus 3.6% MEM and 0.4% fetal bovine serum. This mix-
ture contains albumin levels adequate to block virus adsorption to dilution
tubes (28,29),
Radioactivity was measured by liquid scintillation counting in a Tri -
Carb model 3003 instrument with Insta Gel, both from Packard Instruments.
Virus particle to PFU ratios were required for measurement of surface
area coverage. Ratios were determined from plaque assays and total virus
particle counts obtained from spectroscopic information,!using an absorbance
constant of 1.32 x 10^2 particles per absorbance unit/cm,, This constant was
obtained from extinction coefficients given by Charney el: a_l. (30), - a 60
subunit (isomolar with respect to structural polypeptides) model of the vir-
ion, and the polypeptide molecular weight determinations made by Maizel and
Summers (31). Joklik and Darnell (32) found a constant of 1.3 x 1012 par-
ticles per absorbance-unit/cm, using a similar approach. r
STATISTICS OF ASSAY DATA
Fractional standard deviations (SD) of individual plaque
r (titer) and sample radioactivity can be estimated with the
SD = (ma2
N2
assay data
expression
(1)
where m = total number of pipetting steps,
cr = mean standard deviation of pipetting operations ,
N = number of plaques or counts taken,
b = background counts.
This expression, derived from statistical analysis presented, by Wang et al.
(33), assumes that the formation of plaques or counts are random events, and
that pipetting variation, measured at 0.75% to 3.25%, is normally distributed.
A value of 0.04 was used for 0, twice the measured mean pipetting variation,
to account for increased variance attributed to non-random selection of
pipettes used in preparing tubes for serial dilutions. The non-randomness
is present because one pipette was commonly used to prepare as many as ten
dilution tubes in sequence. Comparison of standard deviations predicted
by Eq. 1. with values obtained from replicate titrations and liquid scintilla-
tion counts gave excellent agreement (34) . ^
-------
BUFFERS
-2
Adsorption experiments were conducted in 1.093 x 10 ^ M NaHCOs plus 9.156
x lO-3 M NaCl* (0.02 I buffer), and in 1.516 x 10~2 M NaHC03 plus 2.941 x 10'1
M NaCl (0.305 I buffer)"!". Buffers were prepared in double-distilled deionized
water (ddd H20) and equilibrated with 5% C02 in air to pH 7 at 15°C.
Degradation experiments were carried out in 0.02 I buffer. Sequential
extractions were done with "elution media" of the following compositions:
Normal lOx Concentrate
NaHCC-3
Na2 ethylene diamine tetraacetate
Sodium dodecyl sulfate
Glycine
Bacto tryptone (Difco)
Fetal bovine serum
PH
The lOx concentrate produced a medium almost identical to normal,elution
medium when diluted 1:10 with 0.02 I buffer. All buffers except those used in
process experiments were sterilized by 0.22 ym membrane filtration.
Process experiments also used composite unchlorinated secondary effluent
from the Palo Alto wastewater treatment plant. ""-'•".
0.01 M
0.001 M
0.2%
0.01 M
0.1 mg/ml
1%
8.8
0.01 M
2%
0.1 M
1 mg/ml
10%
12.0 (before
dilution)
9.8 (after
dilution)
ADSORBENTS .... .
Adsorbents selected for study were non-porous a-Si02, a-Fe203, 3-Mn02,
CuO, a-Al203, Si metal and Al metal. This group of materials has a wide range
of dielectric, electrokinetic, and chemical properties for assistance in re-
solving the adsorption mechanism. Because many of these,materials are common
in nature, adsorption information on these solids should be relevant in an
environmental context.
The Si02, Min-U-Sil 5 powder, gift of Pennsylvania Glass Sand Corp., was
cleaned by repeated overnight refluxing in aqua regia, 16 M WQ^, and| then by
*Some buffers were accidentally prepared with 1% less NaCl in adsorption
mechanism studies.
tBuffer compositions were determined by chemical titrations and equilibrium
calculations including activity corrections and complexation (34).
-------
rinsing in ddd 1^0 between and after refluxing. Refluxing in acid was con-
tinued >_ 5 times until analysis of the supernatant showed that additional
refluxing would remove little of the remaining iron impurity. The Si02 used
in process experiments was Extra Pine Jasper, 0.03 to 0.08 mm in diameter,
likewise a gift of Pennsylvania Glass Sand Corp. It was wiashed extensively
in deionized water to remove particles < 0.03 mm in. diameter and cleaned as
described above. It was rinsed in dilute NaOH prior to acid refluxing to
help disperse adhering particles. . '
Mn02, and CuO powders were ACS reagent grade materials aged for
14 days at 90-95° C and washed extensively with ddd H20 to remove finer par-
ticulates. A1203, Linde A alumina powder, manufactured by Union Carbide, New
York, N.Y., was obtained from Dr. R.L. Bassett, U.S. Geological Survey, Menlo
Park, California. It was aged overnight in ddd H20 at 25° C and rinsed once
in ddd H20 prior to use. Silicon metal semiconductor wafers, 30 ohm-cm N and
0.3 ohm-cm P types, were manufactured by Monsanto Chemical Company, St. Louis,
Missouri. Aluminum metal was ACS reagent powder or metal foil, obtained from
Reynolds Metals Corp. Aluminum metal was cleaned by rinsing in dilute NaOH
and HC1, and ddd H20. I
Oxide structural forms were confirmed by x-ray diffraction. Greek let-
ters a, 3, and j, refer to structural forms. The oxides and;Al powder were
sterilized by heating to 200° C for _> 2 hr, and were stored dry. Si metal was
sterilized by overnight exposure to high intensity UV (~104 erg/ cm2 sec).
The solids CuO, Al metal, A1203, and Si metal, are thermodynamically
unstable for the given reaction conditions in adsorption-desorption experi-
ments. The stable phases are malachite, Q^COH^COg (35) , IT -A^Og (36,37),
and a-Si02 (37), respectively. The surfaces of these materials may well be
altered to the stable phase during experiments, except for Si metal, where the
oxidation is known to be kinetically limited, the surface phase altering in-
stead to amorphous silica (38) . i
i
Clay minerals were not selected because they have at least three types
of sites, plate, edge, and interlayer, which would substantially complicate
analysis of adsorption information. ;
CHEMICALS, ISOTOPES, AND ANALYTICAL METHODOLOGY J
Chemicals used during this study were ACS reagent or higher quality mate-
rial unless specified otherwise. 3H20 solutions for use in process experi-
ments were purchased from New England Nuclear. j
In plaque assays, scintillation counting, and coliforra assays, standards
were always analyzed concurrently with samples. Data from'these samples were
compared only to standards run in parallel. Occasionally samples required
retitration by plaque assay. If standards run in parallel during both assays
were comparable, data obtained by retitration were accepted. If not, the data
point was discarded or the experiment repeated.
-------
COD, suspended solids, and total coliform (membrane filter) determina-
tions conformed to procedures listed in Standard Methods for the Examination
of Water and Wastewater (39).
-------
SECTION 5
MECHANISM OF POLIOVIRUS ADSORPTION
THERMODYNAMIC APPROACH ;
I
The purpose of this section is to evaluate the mechanisms by which ente-
ric viruses adsorb to representative inorganic surfaces, using poliovirus
type 1, strain LSc2ab as the model virus. ;
The overall free energy of an adsorption process CAGfUjs) can be broken
down into various components, c~.
AGads = AGdl
AGVdW+
AG
hyd
(2)
each of which arises from a different type of interaction. The subscripts
dl, VdW, hyd, and cov-ion refer to energy contributions that come from the
interaction of double layers, Van der Waals potentials, surface hydration,
and covalent-ionic interactions respectively. T is the absolute temperature,
and AS0 .represents entropy changes not accounted for in other terms.
and then to compare
theory, where
examined by crit-
importance in
are
The approach has been to evaluate the overall AGads
these values to values of the components calculated from
possible. Terms not possible to evaluate 'analytically
ical hypothesis-and-test experiment or by comparison to their
analogous systems.
DETERMINATION OF AG , .
ads
Mass Action Equation of State
The overall free energy of adsorption can be determined with the mass
action equation of state derived by Dhar et al.
x(l-6)-
Jeq
"HoO
exp
RT
where Cec, = molality of adsorbate, solution phase
Cit c= molality of H90
2 z
11
(3)
-------
R = gas constant
T = absolute temperature in degrees Kelvin :
9 = fraction of surface covered by adsorbate
x = relative size factor; molecules of solute desorbed per molecule
adsorbate adsorbed.
Evaluation of valid free energies from this equation requires determina-
tion of the variables Ceq, CH?O, e> and x> and establishing that the system
is reversible and at equilibrium. The variables Ceq and 9 are determined
from adsorption-desorption isotherms, i.e., Ceq versus 9 plots, measured with
plaque assay data in conjunction with virus particle to PFU ratios, space
filling models of the virus structure, and measured surface areas of the ad-
sorbents. The relative size factar x, estimated to be ~50, is"obtained
from space filling models as well, assuming ~3 percent of a virion face will
approach a surface closely enough to displace one water molecule.
A temperature of 15 ° C was chosen for free energy determinations because
it is characteristic of the average natural water temperatures in temperate
climates.
If adsorption and desorption data points for a given isotherm coincide
within the normal range of data scatter, then .conditions required for deter-
mination of AGads of the adsorption reaction, that is equilibrium and reversi-
bility, are met. This approach is far better than simply following the
course of adsorption experiment with time, because it evaluates both equili-
brium and reversibility criteria simultaneously.
Adsorption-Desorption Experiments
Adsorption-desorption experiments on oxides and Al powder were done in
sterile 15 ml screw-capped borosilicate glass tubes, covered with flamed
alumimun foil to prevent leakage. Procedures for the experiments are listed
below in sequence.
1. Addition of the solid phase, 0.02 to 3 g, depending on the solid, to the
reaction vessel, weighing before and after addition.
2. Rinsing the solid twice with the same type of buffer used in adsorption
to eliminate possible shifts in buffer composition due to double layer
formation on the solid. For experiments with CuO, Mn02, and Fe203,
buffers were pre-equilibrated _> 4 hr with the respective solid phases.
Centrifugation at lOOOg for 10 min in an IEC PR-2 refrigerated centrifuge
was used for solid-liquid separation.
3. Addition of 10 ml of buffer to the reaction and resuspending solids with
a vortex mixer.
C R
4. Addition of virus 10 to 10 PFU with micropipette.
5. Adsorption of virus during slow rotation (0.5 or 1 RPM) for 2 hr in
a rotator mounted in a 15°C incubator. In addition to slow rotation,
12
-------
samples were shaken by hand at 30 min intervals to Insure good mixing.
6. Centrifugation of tubes as described in step 2 and sampling for residual
infectivity and radioactivity. The buffer was then aspirated and the
reaction vessels weighed to determine the quantity of residual buffer
left around the solid.
I
7. Addition of 10 ml of buffer, not containing virus, to; the reaction vessel
and replacement of flamed, foil cap. Solids were then resuspended gently
with a vortex mixer.
8. Desorption of virus into "clean" buffer during slow rotation for 2 hr,
shaking at 30 min intervals as described in step 5.
9.
Centrifugation, as in steps 2 and 6, followed by sampling the supernatant
for determination of the equilibrium concentrations of virus for the
desorption step. ' , •
In adsorption-desorption experiments with Si02 an additional step of
Centrifugation for 10 min at 15,000g in the 40.2 rotor ofja Spinco Model L
ultracentrifuge from Beckman Instruments was required for complete solid-
liquid separation on aliquots withdrawn from the reaction,vessel for assay.
Adsorption experiments on aluminum foil were done in T-25 tissue culture
flasks from Corning Glass Works. Experiments were done in an identical manner
as those described above, except that they were placed on a sample oscillator
which gently rocked the samples at 1.5 Hz. : '
The reaction times for both the adsorption and desorption steps were
increased to 10 hr, and Centrifugation was not required for solid-liquid
separation. ; . .
Adsorption-desorption experiments on Si metal were done in Corning 60 x
15 mm polystyrene, tissue-culture petri dishes to which th'e Si chips had been
attached with polystyrene cement. Experiments were done as described with
Al foil, except that the reaction time was decreased to 4jhr for both the
adsorption and desorption steps and petri dishes were placed on the oscilla-
tor in a sealed container with an atmosphere of 5% C02 in jair to keep the
pH constant. The apparatus is shown in Figure 1. !
Controls demonstrated that rotation and oscillation during the experi-
ments ;and dissolution products of the adsorbents had no significant effect on '
virus titers with 5 hr incubation times, and that adsorption to reaction ves-
sel surfaces was statistically insignificant. ,'
Radioactivity adsorption data corresponded to infectivity adsorption
data for Si02 and Si metal, the weakest adsorbents. On other materials,
however, more radioactivity was present in the equilibrium supernatants
than infectivity, even though both were reduced to exceptionally low levels. '
Because purified preparations contain minor amounts of impurity and degrada-
tion products, which cannot be avoided, this discrepancy may be due to
13
-------
TfilUVXV
Figure 1. Adsorption (a) and degradation (b) apparatus. Buffers
for experiments with transition metal oxides were pre-
equilibrated with respective solid phases (c).
14
-------
preferential adsorption of virus compared to impurity.
point on Si02 at 0.02 I, infectivity data and particle to
to construct adsorption isotherms in all cases.
for one data
PFU ratios were used
Except
Surface Area Determinations
Specific surface area determinations were required for the calculation
of 6, the fraction of available surface covered by virus particles at equi-
librium. BET theory (41) and N2 adsorption at 77.5°K to 77.6°K were used
where feasible, to determine surface area, assuming a molecular site occu-
pancy of 16.2 A2. Adsorption data were obtained with an Orr_ Model 2100
instrument manufactured by Micromeritics Instruments Corp,; "Results are
presented in Table 1. Accuracy of ±20% is suggested by Jura (42). The BET
value we obtained for Min-U-Sil 5_ is quite consistent with results obtained
by other investigators, (e.g., 43,44). :
These BET values are compared to surface areas obtained from Kozeny
air permeability methods (45) as a check on validity. A 0.39.m2/g (BET-Kr)
standard, obtained from Dr. G.A. Parks, Applied Earth Sciences, Stanford
University, was used to calibrate the Kozeny instrument. Kozeny values are
also presented in Table 1. !
According to the discussion of Jura (42) and Schultz (46) these two
methods of surface area determination are not directly comparable. BET
surface areas are considered most accurate (42) and these have been selected
for calculating 6 where available. Kozeny methods (46) measure only surface
in a packed column that is adjacent to air flow paths, and which could be
expected to give smaller areas than BET methods. j
i
The BET surface area of Al metal powder was estimated from the Kozeny
value times the average BET:Kozeny ratio (3.5 ± 1.6) for other samples. On
Al foil and Si (monocrystalline) semiconductor chips, the surface areas were
estimated by direct measurement times a roughness factor of two, estimated
from geometric considerations of etched metallic surfaces.
Adsorption Isotherms and AG ,
Having obtained adsorption-desorption data and specific surface areas,
data were assembled as isotherms.
Adsorption-desorption isotherm data were calculated with the following
equations:
PFU/ml x PPR x 10 g/kg
1,2
eq
Na x p
15
(4)
-------
TABLE 1. SURFACE AREAS OF ADSORBENTS
Solid
a-Si02*
3-Mn02
CuO
a-A!203
Al Powder
Al Foil*
Si Metal*
BET-m2/g
3.95 + 0.79
2.33 ± 0.47
0.38 ± 0.08
0.29 ± 0.06
14.6 ± 2.9t
(0.13)*
0.0012 m2/g
0.0039 m2
Kozeny - m2/g
(duplicates)
1.43, 1.44
0.61, 0.79
0.11, 0.10
0.16, 0.16
2.38, 2^34
0.037, 0,038
ND§.
ND \
*BET surface area of Si02 used in degradation experiments was 5.49 ±
1.10 m2/g. .:. r- ,
fDetermination by Dr. R. L. Bassett, U.S. Geol. Survey. ;.-.,.
^Estimated from average BET:Kozeny ratio found with other samples.
§Not determined. -
#Areas obtained by direct measurement times a roughness factor of
two. - -.'-.-"•:
, Ads1 Ox PPR '•-'• • •-
and Q _ •!•» ^ _-•-•-
u —
w x S x PI
a a
where Ads1 = Tot - PFU/m^ x Va
Ads2 = Ads.,^ + PFU/ml1 x Vr - PFU/ml2 x Vd
; &
(6)
(7)
and PFU/ml- 2 = plaque forming units per ml. Subscript 1 refers to
' adsorption supernatants, subscript 2 refers to desorp-
tion supernatants.
PPR = particle to PFU ratio,
PI = packing index., 1.58 :x 1015 virion sites per m2 C34),
p = buffer density,
16
-------
N
a
w
a
S
a
Ads
Tot
V
a.
V
r
V,
1,2
= Avogadro's number,
= weight of adsorbent,
= specific surface area,
= virus adsorbed to solid,
= total virus initially added,
= volume of buffer during adsorption,
= volume of buffer remaining, after aspiration,
= volume during desorption.
The subscripts 1 and 2.in these expressions refer to adsorption and desorption
steps, respectively. j
Isotherms determined on Si02, Fe_0 ,.™iu ,
and 0.305 ionic strength, are presented3in Figures 2 and 3, respectively.
•MnO~, and CuO, determined at 0.02
Isotherms determined on A120 and Al metal, at an ionic strength of 0.02, are
presented in Figure 4. Isotherms were fit by linear regression analysis of
log:log plots with 1:1 slopes appropriate for low values of 0. A log Ceq
value of -17 corresponds to approximately 5 PFU/ml, a value of -11 corresponds
to approximately 5 x 1Q6 PFU/ml. Approximate nature of correspondence is
related to variation in virus particle to PFU ratios among preparations.
i
On Si metal, we could estimate only a limiting value, because the amount
which appeared to be adsorbed (Ads.. _) was within the statistical precision
of the individual assays. ' i
j
Where adsorption and desorption data points coincide, within the normal
range of data scatter, isotherms appear to be reversible and to approach
equilibrium. This was found to be the case on all materials except on Al
metal and on CuO at high ionic strength, where the reactions were clearly
irreversible. For these two cases, on the desorption step, in all samples
except one - where we counted one plaque, data points were below detection
limits of routine plaque assay procedures. !
i
At low surface coverage, where (1-9) •*- 1, the reversible reactions are
linear, within the normal range :of data scatter. The appairent linearity and
reversibility of these reactions suggest that the interaction energy between
virus and adsorbent does not change substantially as a function of the loca-
tion on the surface where adsorption takes place. !
Adsorption free energies, calculated from Eq. 3, are listed in Table 2.
The^values were obtained by linear regressions with 1:1 slopes at a 0 of
10~ . Standard deviations are based on data scatter about, the regressions.
17
-------
r
-2 -\
-3
-4
log 0
-6
-8
'y-x
= 10
0.73
yx
= 10
0.74
10
0.30
0.44
Adsorption A, o,
Qesprpfion D,+,
-17
-16
-15
-14
-13
-12
-11
log Ceq/mol kg
Figure 2. Adsorption-desorption isotherms of poliovirus on oxides ;at
0.02 ionic strength and pH 7. 0 is the fraction of surface
covered, Ceq is the equilibrium virus concentration (mol/kg)
in solution. Syx is the standard deviation (n-2 degrees of
statistical freedom) of log 9 about regression lines. The .error
bar represents theoretical precision of plaque assay data.
18
-------
-2
-3
-4
Irreversible
a
.v=10
0.34
-5 -
log 0
-6 A
Sy.x=10
0.57
-8
Adsorption A, o, •
Desoiption D, +, x
-17 -16 -15 -14 -13 -12 -11
Ceq/mol kg"
Figure 3. Adsorption-desorption isotherms of poliovirus on oxides at
0.305 ionic strength and pH 7. The error bar represents theo-
retical assay precision. < designations refer to points
below detection limits of the assay system, corresponding
to <8 PFU/ml.
19
-------
-3
-4
-5
log 0
-6
-7
-1
-2
-3
log 0
-4
-5
*-Al203
c 0-?7
byx = 10
Adsorption •
Desorption x
-16
-15
-14
log Ceq/mol kg
-13
-1
-12
-------
TABLE 2.
Solid
a-Si09
&*
Ct~*F 6 0
g-Mn02*
CuO*
a-Al203
Al Metal
i
Si Metal
i
FREE ENERGIES OF POLIOVIRUS ADSORPTION OIJI OXIDE SURFACES
AS DETERMINED BY MASS-ACTION
AGa(js - kJ/mol
0.02 I
-32.4 + 2.4
-52.9 ± 4.0
-58.9 ± 1.7 •
-65.8 ±4.0
-52.3 ± 1.5
Very Strongt
> - 39.5
EQUILIBRIA
AGacj<; - kJ/mol
|0.:505 I
;-50.0 ± 3.1
1-59.9 ± 1.2
1-65.3 ± 2.1
ij
1 Strongt
]- ND*
i ND
1 ND
*These calculated values may be 4 to 9 kJ/mol too negative on the basis of
our degradation^ studies. !
tlrreversible reactions. \
*Not determined. i
Actually, as presented in the next major section of ;this report, virus
is degraded by adsorption on some surfaces, notably Mn02* CuO, and Al metal.
In these cases, we estimate that our free energies as listed may be 4 to 9
kJ/mol too negative. This degree of error is only twice1the normal range of
data scatter about the isotherms. It does not change the free energy
values enough to alter the purpose to which these free energies are put, that
is, evaluation of the adsorption mechanisms. •;
Because it is rather difficult to visualize the differences in effective-
ness of these solids as adsorbents from free energy data, expected equilibrium
titers on these materials were calculated assuming a PPRiof 10^, 25°C, a 9 of
10-5, and have been listed in Table 3; This normalizes all system variables
except the chemical nature of the solid phase. In terms of equilibrium titer,
there is at least 5 orders of magnitude difference between the effectiveness
of the strongest and weakest adsorbents. i
' ' !
APPLICATION OF DLVO-LIFSHITZ THEORY - j
i
:l
The approach in evaluating the contributions from double layer inter-
actions and Van der Waals interactions according to the DLVO-Lifshitz theory,
was to compare differences in AGads with differences in potentials (U)
evaluated theoretically. In evaluating the potentials, as detailed later, a
highly complex interaction of aspheroidal-icosahedral, deformable. virus with a
real surface has been approximated with a sphere-plate model. To adjust for
errors introduced by this approximation, characteristic sepsi.ration distances in
the models between virus and solid were taken to be adjustable parameters. Because
21 ' • . !•
-------
TABLE 3. NORMALIZED CONCENTRATION OF POLIOVIRUS INFECTIVITY IN EQUILIBRIUM
WITH VARIOUS SOLIDS
Solid Phase
a-Si02
CuO*
a-A!203
Al Metal*
Si Metal
Equilibrium Virus Titer log PFU/ml
0.02 I
7.15 + 0.43
3.56 + 0.71
2.50 + 0.29
1.30 + 0.71
3.66 ± 0.26
•-> —00
> 5.90
0.305 I
4.07 ± 0.55
2.34 + 0.21
1.39 + 0.36
-> -00
ND
ND
Includes significant amount of degradation as well as adsorption.
assumptions in double layer models and Van der Waals models are different,
values for characteristic separations are not expected to be the same.
The test of the comparison of differences in AGa(js and U are then two-
fold: (1) do potentials match free energies for all solid tested, (2) are
values selected for characteristic separation distances realistic regarding
current understanding of the molecular structure of solid-liquid interfaces,
i.e., 0.1 to 1 nm. If these two considerations are met, we have good evidence
that DLVO-Lifshitz'theory explains virus adsorption on solids tested.
ELECTROSTATIC COMPONENTS OF AGads
A major contribution to the free energy of adsorption comes from the
interaction of double layers that form about the virus and solid phases.
Double layers exhibit localised concentration excesses of either negative or
positive ions adjacent to charged surfaces in solution.
Viruses and oxides are thought to develop charged surfaces in solution
by ionization of prototropic groups on their surfaces (47). Current treatment
of double layers show at least two principle regions, the inner or Stern
layer and the outer or Gouy layer. These are shown diagrammatically in
Figure 5.
When double layers from different particles interact, they develop
repulsive or attractive forces depending on whether they are of like or oppo-
site electrical charge. Integration of these forces over distance from large
distances of separation until the particles collide will give potential energy
values which can evaluate double layer contributions to the AGad's-
22
-------
Stern Layer
Ionizing
Residues
1
1
1
\ 3 1
v 1
>> V
\ -N
/ 1 n_
Bulk
olution
Surface
Hydroxyls
|
1
1
1
A
v 1
\ i
\J
1
j HO
Figure 5. Diagrammatic representation of double layers about virus and
solid. The presentation shows ionizing residues and surface
hydroxyl groups thought responsible for charge development;
the Stern layer of bound cations, anions, and water molecules;
and the Gouy layer with cation (n+) excesses and anion (n_)
deficits, e^, £3, and £2 are tne complex dielectric suscept-
ibility functions important in controlling the differences
in magnitudes of Van der Waals interactions with different
materials. i
23
-------
We have used the constant charge LSA* expression derived by Gregory
(48) for sphere-plate boundary conditions
Udl = Na 128
Pa + O.Snm) irnknT
'V D
exp(-K
(8)
where U,, = double layer interaction potential per mole,
ay = virion radius - 13.5 run; 0.5 ran is added to account for Stern
layer thickness according to Smith (49),
3 '
n = anion or cation density (per m ) in solution,
/
N = Avogadro's number,
cL
kn = Boltzmann's constant,
o
22 %
K = (2e nz / e e kRT) , the Debye inverse length,
e = relative dielectric constant of immersion medium at zero
frequency,
e = dielectric permittivity of free space,
tanh (y1 7/4)
•*->*•
*1,2 = Ze*l,2 /kBT>
z = charges per ion,
e = electronic charge,
ij;1 9 = Gouy Chapman surface potential,
J., £•
interaction separation distance for electrostatic potential.
.
J.,
This expression predicts potentials which are intermediate between those pre-
dicted by exact constant surface potential and exact constant charge expres-
sions. A similar intermediacy is observed with experimental data pertaining
to the adhesion force of polystyrene latex particles on cellophane (50.) .
The only experimental data required to compute values of U^i for a given
system for sphere-plate boundary conditions are measured values of surface
potentials i|>i 2 f°r the interacting particles under a given set of experimen-
tal conditions, and the characteristic distance between them in the adsorbed
* Linear superposition assumption.
24
-------
state. On the basis of Smith's work on oxide electrokinetics, it was assumed
that ^i 2 were approximately equal to zeta potentials (g]_; j) measured by
electrophoresis experiments. A characteristic separation distance des here
was assumed to be zero at coincidence of the Stern surfaces. This corresponds
to a value of 1 .nm between effective virus and substrate surfaces (49)•
j
Zeta Potential Determinations for Oxides ' |
Zeta potentials of the oxides were determined by conventional micro-
el ectrophoresis with a Zeta-Meter, manufactured by Zeta-Meter, Inc.
Electrophoretic mobilities were measured in 0.02 I buffers under 'condi-
tions closely paralleling virus adsorption-desorption experiments. 'Zeta
potentials were then calculated with the Smoluchowski equation (51). In 0.305
I buffers, the ionic strength was too high to permit accurate measurement.
Zeta potentials were estimated from values measured at 0.02 I using Hunter
and Wright's correlation (52)*. Values are given in Table 4.
i
I
Zeta Potential of Poliovirus j
The virus zeta potentials were measured by zonal electrophoresis in a
RNAse-free sucrose gradient, and an extension of the retarded, relaxed Smolu-
chowski equation (34,53). This extension allows correction for variation of
viscosity with distance in the gradient. Again, the value of ? at 0.305 I
is estimated by applying Hunter and Wright's correlation.) Values are included
in Table 4. :
Correspondence of Theoretical Isoelectric Points with Reported Values
As a check on the validity of the virus zeta potential measurements and
to determine whether or not this strain of virus is electrokinetically simi-
lar to other enterovirus types, isoelectric points, of the virus were estimated
from zeta potential information and compared to reported values obtained
by isoelectric focusing.
Estimates were obtained by first extrapolating the c'ounterion concen-
tration in Hunter and Wright's correlation to 10~6 M for a value of the sur-
face potential OP0). Estimates using this procedure are quite close to
values obtained by Hunter and Wright (52) using more invojlved methods. Then
the pH at which the virus has zero charge (pHp^c) was calculated with the
Nernst equation (47) j •
RT
= 2.303 ^- (PHpzc - PH)
(9)
* Hunter and Wright (52) found far oxide surfaces that the measured zeta po-
tentials plot approximately linearly as a function of the logarithm of count-
erion concentration and the C=0 intercept corresponds to a cbunterion concen-
tration of ~1 M. |
25 j . .'
-------
TABLE 4. ZETA POTENTIALS (mv) OF OXIDES AND VIRUS IN pH 7 NaHC03-NaCl SOLUTIONS
0.02 I
0.305 I*
a-Si02
a-Fe203
3-Mn02
CuO
Poliovirus 1 LSc2ab
-44.5 +'
-28.8 ±
-20.7 ±
-17.6 +
-5.9 ±
8.5t
2.5
5.9
6.1
0.9
-13.4
-8.7
-6.2
-5.3
-1.8
±2.6
± 0.8
± 1.8
+ 1.8
± 0.3
*Estimated with Hunter and Wright's (52) correlation,
"i"Values are standard deviations.
where F is Faraday's constant and pH is the pH of the solution in which elec-
trophoresis was done. Data required for more accurate evaluation procedures,
e.g. the Yates et_al. model (54), are not yet available for viruses. The
pHpzc should be approximately equal to an isoelectric point in the absence'of
strong binding of ions in solution.
Comparison of estimates with reported values is presented in Table 5.
Values appear to be in good agreement with those of other types (55-59).
Mandel (56) found that infectious poliovirus exists in two conformational
states, A and B, each of which has a different isoelectric point,, Virus that
had been inactivated appeared to have an isoelectric point similar to what was
measured for B state virus. There was a small amount of inactivated material
(Table 5) in one of our experiments that agrees with Mandel's findings.
Mandel's conformational bimodalism has been found with many different
types of enteroviruses, and should play a major role in governing adsorption
of viruses to materials in the environment.
Double-Layer Interaction Potential Evaluation
Having obtained estimates of the zeta potentials of oxides and virus for
the adsorption reactions, Eq. (8) can be used to evaluate double-layer inter-
action potentials. These are presented in Table 6 and are repulsive in all
cases.
To evaluate the degree to which the adsorption-free energy is controlled
by double-layer interactions, differences in the AGa(js and Udl at two ionic
strengths were compared. This comparison, shown in Table 7, gives excellent
agreement.
26
-------
TABLE 5. COMPARISON OF THEORETICAL ISOELECTRIC POINTS WITH VALUES REPORTED
FOR OTHER ENTEROVIRUSES {
Virus
Poliovirus 1
LSc2ab
Poliovirus 1
Brunenders
Poliovirus It
Brunhilde
Poliovirus 1
CHAT
Poliovirus 1
Mahoney
Poliovirus 2
Sab in T2
Coxsackievirus
A21
Echovirus 1
Farouk
State Isoelectric Point Reference
A 6.6
non- infect.* mat. < 5.7
A 7.4
B 3.'8
A 7.0
B 4.5
A 7.5
B 4.5
A 8.5 + 0.1
B ?
A 6.5
B .4.5
A 6.1
B 4.8
A? 5.6
B 5.1
_
This study
(55)
(56)
(57)
(58)
4
*
*r
(59)
*The Nernst method of estimating pHjgp should give an overestimate if the
electrophoresis pH is far from the isoelectric point (54).
tMandel found essentially identical results with zonal electrophoresis as a
function of pH as with isoelectric focusing. i
*B.D. Korant and K. Lonberg-Holm, personal communication (1977),
TABLE 6. DOUBLE-LAYER INTERACTION POTENTIALS BETWEEN VIRUS AND SOLID PHASES
ACCORDING TO THE LSA EXPRESSION \
Solid
Udl(0.02 I) - kJ/mol
Udl (0.305 I) - kJ/mol
a-Fe203
CuO
18.9 ± 4.8*
12.7 ± 2.2
9.2 ± 3.0
7.9 ± 3.0
1.9 + 0.5
1.2 ± 0.2
0.9 ±0.3
0.7 + 0.3
* ± values are standard deviations.
27
-------
TABLE 7. COMPARISON OF DIFFERENCES IN AG , and U,, at 0.02 I and 0.305 I
ads dl
Solid
AG
ads
' °2
- 0.305 I) - kJ/mol Udl (0.02 1-0.305 I) -kJ/mol
a-SiO
a-Fe0
2
3-MnO
CuO
2
0_
3
2
17
7
6
.6 ±
.0 +
.3 ±
t
4
4
2
.0
.3
.6
17
11
8
7
.0 ±
.5 +
.4 ±
.2 ±
4
p.
3
3
1 *
.8
.2
.0
.0
* ± values are standard deviations.
t Cannot be determined because the isotherm at 0.305 I is irreversible.
ELECTRODYNAMIC COMPONENTS OF AG ,
ads
It is apparent that the virus is quite strongly adsorbed in spite of the
repulsive nature of all mineral-virus double layer interaction potentials
studied so far. The electrostatic interactions are weak where adsorption is
strong and are strong where adsorption is weak. This suggests that other
strong attractive forces must be involved in adsorption.
Colloid stability and polymer adsorption theory (60) suggest that Van
der Waals interactions should be of major importance in systems involving
interaction of particles of virus size.
Evaluation of Van der Waals interactions was made with a combined
Lifshitz-Hamaker approach. Van der Waals interactions are a result of elec-
tromagnetic fields generated from dipole oscillators in one material, i.e.
electronic oscillations about nuclei of constituent atoms and vibrations or
rotations of molecular structures containing separated electrical charges,
interacting with dipole oscillators in an adjacent material. If there is an
intervening material, e.g. an aqueous solution, interacting electromagnetic
fields are screened by dipoles in the immersion medium.
Dzyaloshinski, Lifshitz, and Pitaevski (22) developed a general theory of
Van der Waals interactions (the Lifshitz theory) from continuum quantum field
electrodynamic considerations. At small separations, < 5 run, and at zero
temperature (°K) the electrodynamic induction force (F]^) between two mate-
rials, designated by subscripts 1 and 2, immersed in a third medium, desig-
nated by subscript 3, according to Dzyaloshinski e_t al. (22), is
28
-------
132
167T
ed
,
/ '/
o o
- el
£3J
X
- 1 1
J
1
dX (.10)
where
n
d
X
e
ed
= Planck's constant h/2ir,
= interaction separation distance,
= integration variable (22),
.. - „ = complex dielectric susceptibility functions: e; = e' (to) + ie"(w),
' ' as a function of the complex frequency w = aL + i E,.
Evaluating the integral and describing the dielectric, susceptibility as
a function of the imaginary frequency component (22) gives
h" -
to.
132 " 87T2d 3 132
ed
(11)
where
0)
132
= f
J
iQ -e,(iQ ] [e (iQ - e (i
(12)
Describing the dielectric susceptibility as a function of the imaginary
component of the complex frequency allows ready interpolation between e
values measured at different frequencies by capacitance bridge measurements
or spectroscopic information (61). j
The Lifshitz function is of the same form as older pairvrise summation equa-
tions derived by Hamaker for plate-plate geometries j
132
132 &rd
ed
(13)
The complex Hamaker coefficient &••,-??> is related (22) to, the Lifshitz fre-
quency integral by . j
A =
132
4ir W132
(14)
for short-distance, Munretarded" interactions. Short-distance formulations
are appropriate for this system, because nearly all the interaction energy is
developed at very short separation distances compared to distances where
retardation (62) becomes important. 'i •
29
-------
Evaluation of Hamaker coefficients by Lifshitz theory will give Infor-
mation which can be used in the sphere-plate* Hamaker expressions (63)
U - M 152
UVdW ~ "Na 6
2a (a + d
vv v
ed
2a
- In
(d
ed
2a
ed
(15)
appropriate for our system to calculate Van der Waals potentials (U, ), se-
lecting 0.1 nm for ded- Vdw
It is mathematically convenient to evaluate individual self-interaction
coefficients, and evaluate complex coefficients by more recent developments.
We apply mixing laws derived by Bargeman and Van Voorst Vader (64)
132
and
3h
16ir2b.
CAj/2 - Aj/2) 0
.1/2 f .1/2
f*-ryry L-">T T
66 11
vl/2 _ 1/2.
A V2,
- A22 ^
16TT bi
1 3h
16ir2b.
1 1
1 3h
A__A__)
£i£* *_I*J
rA A ~\ '
LA11A33J
(16)
131
^U ~
16fr2b.
1 _ 1 1
A1/2)2
A33 J
ra A
,1/2
(17)
3h
-16,
where b. is a fitting parameter equal to ~0.32 x 10~10/s.
Self- interaction coefficients (A^i, &-22> A53) are evaluated by Visser
(65) from Lifshitz theory or Ninham and Parsegian's ("66) macroscopic approx-
imations. Appropriate values are presented in Table 8, along with mixed
coefficients (A,-, and A.. _„) calculated with equations 16 and 17.
To. determine to what extent Van der Waals potentials account for the
residual free energy controlling adsorption (AGres) , double layer contri-
butions were subtracted from the overall free energy and the AGres ob-
tained was compared to Uydw evaluated by Lifshitz-Hamaker theory. The com-
parison, presented in Table 9, shows that the Van der Waals potentials can
account for all of the remaining energy. This means that the balance 'of
double layer interaction and Van der Waals interaction can account for the
principal contribution to virus adsorption free energies with these solids.
A value for de,j (0.1 nm) was selected that is shorter than that used
by other investigators, e.g. 0.2 rim (67) and 0.4 to 1 nm (65). This probably
is in response to the larger contact area expected between virus and pxide
surface than given by the sphere-plate interaction model, and also for pos-
sible effects of solution damping (61) .
*Refer to Figure 5 for diagrammatic presentation.
30
-------
TABLE 8. INDIVIDUAL SELF-INTERACTION HAMAKER COEFFICIENTS TAKEN FROM
VISSER (65) AND CALCULATED MIXED COEFFICIENTS
Material
Water
Polystyrene (virus)
"Oxides, etc."
Metals
Quartz
Material
Quartz-Water-Virus
"Oxides"-Water-Virus
"Metal"-Water-Virus
Polystyrene-Water-Virus
TABLE 9. COMPARISON
-13
Aj- Values - erg x 10
4.38
6.15-6.60
10.6-15.5
16.2-45.5
8.0-8.8
-13
' A 2 Values - erg x 10
0.33-0.49
0.53-1.06,
0.91-2.97
0.17-0.26
OF LIFSHITZ^VAN DER WAALS POTEN1
FREE ENERGY COMPONENTS AFTER CORRECTING FOR DOUBLE
Solid
a-Si02
a-Fe203
3-Mn02
CuO
Al Metal
Polystyrene
uvdw - kj/mo1
-43.0 to -63.9
-69.1 to -138
M
it
-118 to -387
-22.2 to -33.9
*Values are standard deviations .
•{•Irreversible reactions, A Gacjs could not be determined.
31
Theory
Lif shitz
"
it
Macroscopic
Approximation
Theory
Bargeman
and
Van Voorst
Vader
IALS TO RESIDUAL
LAYER REPULSION
AG - kJ/mol
res
-51.9 ± 3.2*
-61.0 + 1.2
.'-•• -66.1 ± 2.1
• -73.7 ± 5.0
' . Very Strongt
Weak - No Ads .
-------
Lifshitz theory predicts that we should have stronger adsorption on Si
metal than on Si02- This effect is not observed in our adsorption results.
It is suspected that kinetic difficulties in approach to the Si metal surface
or the thickness of the oxide coating* on the metal may prevent strong
adsorption. ;
On Al metal, however, even though ellipsometryt suggests an oxide coating
we see exceptionally strong adsorption, characteristic of metal predictions
by Lifshitz theory. This may involve local thin spots in the coating allowing
dipole-dipole contributions from the metal proper to interact with adsorbing
virus.
In any case, with"the Lifshitz theory and also with double layer inter-
action theory, the conditions for accepting the models — ,viz, correspondence
of potentials with free energies and reasonable values for characteristic
separation distances, have been met.
OTHER POSSIBLE COMPONENTS
Even though DLVO-Lifshitz theory explains virus adsorption with surpris-
ing accuracy, it should be recognized by the critical investigator that the
fit could be merely coincidental.,
We are unable to develop models substantiating significant involvement
of covalent-ionic interactions,hydrophobic interactions, and electrostatic
induced image interactions being important in virus adsorption to materials
studied here. This topic is discussed at length in (34) and (70). Minor con-
tributions firom AGa(js nonetheless f may come from these other types of inter-
action in some cases.
The estimate of ~50 interfacial water molecules displaced by adsorption
of one virus particle suggests that energy not accounted for previously is
required for this transfer. LeNeveu et al's.(71) measurements of the chemical
potentials of hydration associated water molecules are 0.2 kJ/mol to 0.01
kJ/mol at 0.4 to 1.0 nm phospholipid bilayer separations. For the virus system
this gives an estimated 0.5 to 10 kJ/mol repulsive energy due to transfer of
hydration water to the bulk liquid. The precision to which DLVO-Lifshitz
theory can be fit to free energy data could easily accommodate this level
contribution to AG , .
ads
Ion-exchange concepts are inapplicable to our problem because viruses are
orders of magnitude larger than ions. Ion exchange occurs as a result of
charge development on surfaces of materials (35) •. In our system, these devel-
opments are expressed in the interaction of double layers.
*3 to 6 nm thick by ellipsometry for Si exposed to air for several months (68).
tSmith (69) found by ellipsometry that the oxide coat on Al metal is ~5 nm
thick after electropolishing in HC104-ethanol mixtures and ~20 nm thick
after acid-dichromate etching.
32
-------
Good fit of the DLVO-LifsMtz theory suggests that additional entropy
contributions may be small. :
PREDICTIONS FROM DLVO-LIFSHITZ THEORY •
I
The DLVO-Lif shitz theory is capable of predicting adsorption characteris-
tics of viruses in aquatic systems.
For spherical particles, keeping all other parameters; constant, if Eq. 8
is combined with Eq. 15, and the combined .double layer and Van der Waals
potential is varied with particle radius, large particles are predicted to
adsorb more strongly than small ones. In general, Wallis ;and Melnick's
empirical studies (72) of the adsorption of various virus 'types on aluminum
hydroxide and phosphate precipitates exhibit this trend, but there is con-
siderable scatter in the correlation. Adenoviruses and reoviruses,
which are larger than enterbviruses, adsorb less strongly \ than enteroviruses
to these precipitates. Differences in electrokinetic properties and the
configuration of-the interaction with the surfaces due to different viral
architectures are thought to be the cause of this apparent discrepancy.
Tailed bacteriophages are morphologically very different, from entero-
viruses. In these phages, in contrast to enteroviruses, different types of
protein are found in different locations on the virion surfaces and their
electrical properties are expected to be considerably different. Adsorption
of these phages should be unrepresentative of enterovirus adsorption.
Icosahedral phages such as f2 may correspond to enteroviruses in terms of
the configuration of the interaction, but if their electrical properties are
much different, adsorption characteristics may be different as well.
The combined DLVO-Lifshitz expression predicts that If the virus and
solid have similar charge, increasing the ionic strength of the supporting
solution will increase the tendency of virus to adsorb. Conversely, if they
are of opposite charge, an increase in ionic strength will decrease the ten-
dency to adsorb. ;
Using Si02 £ measurements, and for Mandel's (70) poliovirus - estimat-
ing the zeta potentials by Hunter and Wright's correlation and the Nernst .
equation, Eq. 8 can be used to estimate the effect of|pH on adsorption.
The Van der Waals component -51.9 kJ/mol for Si02 was taken from Table 9.
An ionic strength of 0.02 with monovalent cations and anions was selected.
Our prediction is presented in Figure 6. The pH at which!the isoelectric
point shifts from the A to the B conformation was taken from Mandel's (56)
data. It may, actually, shift when adsorbed to a surface at a different pH.
At high pH (8 to 9) the adsorbed fraction of total virus present in the
system (Facis) is very small. As the A state isoelectric point is approached-
about pH 7, adsorption is strong. It decreases on lowering the pH between 5
and 6, however, until transition "to the B state occurs and we have strong
adsorption again around pH 4.5. Of course, differences in electrical pro-
perties with different adsorbents will change predicted adsorption on these ,
materials to be noticeably different from Si02. Some investigations such as
those of Berg (4) who found an adsorption maximum at pH 7'on membrane filters
33
-------
log Fac|s
Figure 6. Estimated variation of poliovirus adsorption with pH on
one m2 of a-Si02 at 0.02 ionic strength as predicted by
DLVO theory. Transition from the A to B state is taken
from Mandel. Arrows represent pH regions of relatively
strong predicted adsorption.
34
-------
with poliovirus and those of Netty Buras (personal communication, 1976) who
found adsorption maxima at about pH 7 and pH 4.5 on magnetite (FegC^) but
decreased adsorption at intermediate pHs, agree with DLVO-Lifshitz predictions
in Figure 6. Other studies such as those of Gerba et al. !(73) on poliovirus
adsorption to activated carbon, however, do not show the two maxima expected.
Lifshitz theory predicts that materials with high dielectric susceptibil-
ities such as metals and metal sulfides, should develop higher Van der Waals
potentials, and therefore be better virus adsorbents than 'materials such as
silicate minerals, clay minerals, and humic substances, i.e.,
Metals > Sulfides > Transition Metal > SiO,
(strong) Oxides
Most
etc.> Organics
(weak)
The surface area of a given material must also be taken into account.
Large amounts of surface area favor more adsorption. Carei must be taken when
extending this generalization to porous materials such as activated carbon
and clay minerals, however, because the surface area included in pores may
not be accessible to adsorbing virus. Additional discussion of environmental
implications has been presented earlier (34, 70).
Aluminum metal appears to be particularly good in regard to adsorbing
virus. This material will be examined in greater detail in following sections
concerning virus degradation and removal in column experiments.
35
-------
SECTION 6
DEGRADATION OF POLIOVIRUS BY ADSORPTION
KINETIC APPROACH
In the last section, the principal mechanisms by which virus particles
adsorbed to various solids were resolved. Another critical question that
deserved attention was whether or not virus retained or lost infectivity
during or after adsorption. This question was investigated by study of the
infectivity and labeled viral protein and nucleic acid concentrations re-
covered in desorption experiments.
Only on Si02 could all 3H-RNA and 14C-protein be recovered in one ex-
traction. All other solids allowed only a small fraction of the radioactivity
initially adsorbed to be eluted. Lack of recovery of virus from various
adsorbents is a widely recognized phenomenon (74). The question of whether
this lack of recovery represents strong adsorption or inactivation will now
be addressed. Kinetic theory applied to infectivity and radioactivity data
from multiple extractions provides an answer to this question. It allows
calculation of total virus (both infectivity and radioactivity) recoverable
with an infinite number of extractions. Comparisons between stock prepara-
tions and desorbed virus can then be made for estimates of specific infecti-
vity loss.
Derivation of Kinetic Expressions
For multiple extraction data, the. mass balance expression can be written
Adsn-l - Adsn = C'nVn ' -<")
where Adsn is the total virus adsorbed for extraction n, C'n is the concen-
tration due to desorption during step n, Vn is the volume, and n-1 refers to
the previous extraction.*
A first-order rate of virus desorption from the surface
d Ads
dt
= -k Ads (19)
Variables Ads, C1 and others pertaining to amounts of virus can refer to
either infectivity or radioactivity.
36
-------
where k is the rate constant and t is elapsed time was applied because
surface coverage («1%) is low, which minimizes the probability of
particle-particle interactions. Other systems, that have high surface
coverage, could have significant adsorbate-adsorbate interactions on the
surface which could, cause desorption reactions to have orders different
from one. No back reaction occurs during elution because controls demon-
strated that virus does not adsorb to any of our solids in the presence of
our elution medium. We also assume that dk/dn -» 0 for our systems, and this
assumption is evaluated during discussion of these results.
Substitution and integration of Equations 18 and 19 gives
Since
V C' =
n n
^tl - exp (-kAt)].
(20)
for any n,
where
Ads
n
Cf = D Ads
n = exp (kAt) - 1
V
(21)
(22)
and substituting into Eq. 18 we have
C' - C1 , = CT DV .
n n-1 n n
(23)
Keeping V and At invariant with respect to n, and assuming that our
extractions are a continuous process, we have «
dC'
Integrating this expression gives
In C1
r = -DV dn
= -DV n
(24)
(25)
where x represents the sequence number of the final extraction. Plotting
In C1 versus n will give a straight line with slope -DV and intercept
In C' . !
37
-------
The amount recoverable in an infinite number of extractions is then
' Adso - 2. , (26)
by Eq. 22.
DEGRADATION EXPERIMENTS
Degradation studies were performed on all solids used for the adsorption
mechanism studies except Si metal. A different batch of Si02 with a slightly
greater surface area (Table 1) was used in these experiments.
Experiments were done in ethyl ene oxide or UV sterilized polyethylene
scintillation vials. Virus was adsorbed to the solids for periods of two hr
to 24 hr in 10 ml of 0.02 I NaCl + NaHC03 pH 7 buffer at 15 ± 2°C prior to
elution. The following weights of solids were used in the experiments;
values after + represent standard deviations: H
a-Si02 4.02 ± 0.01 g (initial), 2.64 ± 0.06 g (residual, w )
a-Fe203, 3-Mn02, CuO, Al metal 1.00 ± 0.01 g
a-Al90 0.0803 ± 0.0002 g
£• o
Experiments with Si02
On Si02, after virus was adsorbed to the solids for the incubation time
periods mentioned above, the mixture was sampled and lOx concentrated elution
medium was added in the appropriate amount. The samples were then placed
on the sample rotator and eluted at 25 ± 2°C for one hr for extraction.
Afterwards, they were centrifuged in an IEC model PR-2 refrigerated centri-
fuge for 5 min at 250g and sampled.
Both aliquot sampled after adsorption and that sampled after elution
required additional clarification at 15,000g for 10 min in a Spinco model L
ultracentrifuge with a 40.2 rotor.
The liquid volumes were monitored at every step during the experiment by
recording the pipetted input volumes or weighing. The amount of silica with-
drawn from the reaction vessels was determined by rinsing the solid portion
of the withdrawn aliquot after sampling with ddd H^O and weighing after
drying at 200°C.
The amount of virus initially adsorbed (Ads.) is
Ads. = ^ (Tot - V Cu) (27)
38
-------
where
Tot = total virus initially added,
w = amount of solid remaining in reaction vessel after sampling,
w = total solid initially added,
a
V = solution volume in reaction vessel during adsorption,
3.
C = concentration of virus not initially adsorbed.
The recovered virus (Adsr) is
Ads = C V -
r e e
C V
u r
(28)
where Ce is the virus concentration after elution, Ve is the volume of eluting
solution in the reaction vessel, and Vr is the residual volume of solution
left after withdrawing the aliquot to determine unadsorbed'virus.
Experiments with Other Solids j
To obtain degradation information on other solids it, was necessary to
perform multiple extractions and apply the kinetic analysis derived earlier.
j
i
Virus was adsorbed to solids under the same conditions used for SiO~ for
periods of two hr and 24 hr. j
The samples were then centrifuged for 5 min at 250g in the PR-2 re-.
frigerated centrifuge and sampled for residual virus. The supernatants were
then aspirated and 10 ml of elution medium (Ix) were added;to the reaction
vessels. The samples were placed on the rotator and eluted for 15 min at
25 •+ 2°C followed by 5 min centrifugation at 250g. in the PR-2, sampling
for eluted virus, and aspiration of the remaining medium. This procedure
was repeated four times in sequence. All volumes were monitored during the
experiment by pipetting or weighing.
The total virus initially adsorbed is
Ads. = Tot - V C ,
i a u
(29)
For calculation of the amount of material recoverable with an infinite number
of extractions (Adso), concentrations in the elution supernatants which
desorb during the given extraction (C'...) are-required. They are
obtained from analyzed concentrations (Cn) by 1
n
= C - 'C ' frr=-
n
n-r
(30)
39
-------
where Vr is the volume of residual buffer or elution medium left in the
reaction vessel from the previous operation.
The value of Adso can then be obtained from the expression derived
earlier,
Adso = -£ (26)
where values of C and D are obtained by linear regression analysis of C1
data. °
Controls for these experiments showed that there was no significant
loss of virus or infectivity in the reaction vessels when incubated with
0.02 I buffers pre-equilbrated (> 1 week) with solid phases. Elution
media had no significant effect on the virus titer for times equal to those
used for extraction experiments, and no significant effect on the HeLa cell
assay system when diluted 1:30 or greater with media used to prepare serial
dilutions. Theoretical minimum estimates of standard deviations of initially
added virus and recovered virus were made from analysis of data scatter
about regression lines and theoretical counting plus pipetting variations.
Other sources of variation, such as differences in site-specific interaction
energy over the surface,- cannot be quantified at present. 'Duplicate determi-
nations indicate slightly greater variation than theoretically predicted.
RESULTS
If all inf ectivity, %-RNA, and ^C-pTotein initially adsorbed within
expected variation of the system is recovered, virus is not significantly
inactivated by the presence of the surface. If all 3H-RNA and 14c_protein
is recovered but significant losses of infectivity are observed, we can
conclude that virus is inactivated and not just merely adsorbed. These
hypotheses are valid even if preparations contain minor levels of radioactive
impurities, provided that we have no additional,inactivation due to the
combined effects of adsorbent and elution medium.
Recovery information for Si02 is presented in Table 10. All the infec-
tivity, 3H-RNA, and 14c-protein initially adsorbed is recovered and no
significant loss of infectivity is observed for the two hr and 24 hr samples.
For other solids, application of the kinetic theory derived earlier is •
required to determine whether or not virus is being inactivated by adsorption
on the surface. The kinetic plot mentioned earlier gives us additional infor-
mation concerning degradation. If regression lines of normalized (C'/Adsi)
concentration data are coincident, virus infectivity is not altered> provided
that degradation products desorb in the elution medium. If regression lines
are linear, the assumption regarding the invariance of desorption rate with
extraction sequence number is indicated to be appropriate. Recovery informa-
tion is presented in Tables 11 to 15 and sample kinetic plots (sample C for
all cases) are presented in Figures 7 to 11. .
40
-------
TABLE 10. COMPARISON OF VIRUS INITIALLY ADSORBED (Ads.) WITH VIRUS
RECOVERED (Ads )
2-hr Incubation —
Measurement Sample
Infectivity (log PFU) A
B
3H-RNA (log CPM) A
B
C-Protein (log CPM) A
B
24-hr Incubation —
Measurement Sample
Infectivity (log PFU) C
D
SH-RNA (log QPM) c
D
14C-Protein (log CPM) C
D
PFU = plaque forming units,
*Values are standard deviation estimates.
for a-Si02 !
Duplicates A,B :
Adsi !
8.46 + 0.08* 1
< 8.41 ± 0.09 1
3.22+0.02 i
3.17 ± 0.03
I
3.20 ± 0.02 :
3.15 + 0.03 '
i
Duplicates C,D
Ads . !
8.46 ± 0.10 !
8.49 ± 0.10 ,
3.41 ± 0.03 \
3.49 ± 0.02 ;
2.29 ± 0.02 |
2.99 ± 0.02 :
CPM = counts per
<
Ads
r
8.71 ± 0.10
8.63 ± 0.11
3.17 ±0.03
3.12 ± 0.04
3.13 + 0.03
3.13 ± 0.03
Ads
r
8.14 + 0.10
8.30 ± 0.09
3.38 ± 0.04
3.56 ± 0.03
2.86 ± 0.03
2.96 ± 0.03
minute .
On Fe203 as with Al203 there is a possibility that inactivation of virus
occurs on the surface, but it does not appear very significant. On Mn02 we
see demonstrable amounts' of inactivation, both by examining recovery informa-
tion and by observing that the regression lines of normalized infectivity
data plot distinctly below radioactivity data. On CuO, inactivation is
quite marked by the criteria mentioned above. In addition, the amount of
virus inactivated'increases with the length of time virus was incubated with
the adsorbents prior to extraction. .i
On aluminum metal, however, inactivation is quite dramatic. For two-hr
samples, recovered infectivity is decreased approximately three orders of
magnitude compared to that initially adsorbed. For 24-hr samples, greater
than four orders of magnitude decrease was observed. Only with the first
extraction was any infectivity (86 PFU/ml) detectable. A minimum value for
41. ' ! .
-------
TABLE 11. COMPARISON OF VIRUS INITIALLY ADSORBED (Ads^ on a-Fe203
WITH VIRUS RECOVERED BY KINETIC ANALYSIS OF DATA F.ROM FOUR EXTRACTIONS(AdsQ)
2-hr Incubation —Duplicates A,B
Measurement
Infectivity (log PFU)
3H-RNA (log CPM)
14C-Protein Clog CPM)
Sample
A
B
A
B
A
B
Ads.
8.61 ± 0.06*
8.61 ± 0-06
3.58 ± 0.02
3.58 ± 0.02
3.41 ± 0.02
. 3.41 + 0.02
AdsQ
8.33 ± 0.20
7.56 ± 0.10
3.20 ± 0.21
3.19 ± 0.23
3.31 ± 0.31
3.27 +.0.34
24-hr Incubation —Duplicates C,D
Measurement
Infectivity (log PFU)
3H-RNA (log CPM)
14C-Protein Clog CPM)
Sample
C
D
C
D
C
D
Ads.
8.14 ± 0.07
8.14 ± 0.07
3.74 ± 0.02
3.74 ± 0.02
3.18 + 0.02
3.19 + 0.02
Ads
o
7.74 + 0.25
7.75 ± 0.32
3.53 ± 0.16
3.50 ± 0.11
3.02 + 0.21
2.87 + 0.08
PFU = plaque forming units, CPM-= counts per minute,
*Values are standard deviation estimates.
the slope of -0.2 was estimated here, less than half that found with the
two-hr samples. This gives a conservative estimate for recovery of infectivi-
ty. The less than 104 PFU detection limit is imposed by the requirement of
diluting elution medium 1:30 or more for assay, and by constraints defined
by the mathematics of the system. A lower bound on the amount inactivated
can be predicted from PFU assays and extraction volumes. Both duplicates
contained 1 PFU per inoculated extraction media volume. This gives
approximately 1Q3 PFU recovered for the total volumes used during the first
extractions of the aluminum metal recovery experiment.
The logarithms of extraction concentration data approach linearity when
plotted as a function of n, but exhibit a slightly concave upwards trend_in
some cases. This indicates that materials desorb from a given surface with
similar but not quite identical rates.
42
-------
- 3
- 7
In
-15
-19
-23
PFU
n
Fe203
Figure 7.
Kinetic plot fora-Fe203, Sample C. C'/Adsi is the concentration
of virus that desorbs during each sequential extraction, n=l to 4;
divided by the total virus initially adsorbed to the solid.
Plotted curves are least square fit linear regressions. The error
bar represents theoretical assay standard deviation.
43
-------
TABLE 12. COMPARISON OF VIRUS INITIALLY ADSORBED (Ads^ on 3-Mn02 WITH VIRUS
RECOVERED BY KINETIC ANALYSIS OF DATA FROM FOUR EXTRACTIONS (Adso)
2-hr Incubation — Duplicates
Measurement
Infectivity (log PFU)
3H-RNA (log CPM)
14C-Protein (log CPM)
Sample
A
B
A
B
A
B
Ads^
8.57 ± 0.09*
8.57 ± 0.09
2.93 ± 0.02
2.93 ± 0.02
3.14 + 0.02
3.14 ± 0.02
AdS0
7.53 + 0.36
7.86 ± 0.11
3.09 ± 0.15
3.09 ± 0.20
3.05 + 0.09
2.97 ± 0.19
24-hr Incubation — Duplicates
Measurement
Infectivity (log PFU)
3H-RNA (log CPM)
14C-Protein (log CPM)
Sample
C
D
C
D
C
D
MSi
8.37 ± 0.13
8.37 ± 0.13
3.73 ± 0.02
3.74 ± 0.02
'3.15 ± 0.02
3.16 ± 0.02
AdsQ
7.13 ± 0.24
7.43 ± 0.31
3.04 + 0.24
3.02 ± 0.23
2.70 i 0.19
2.68 ± 0.18
PFU = Plaque forming units, CPM = counts per minute.
*Values are standard deviation estimates.
This is also seen in the slight decrease in recovery of H-RNA ai)d
^c-protein seen in some samples. The decrease is barely measurable if
at all, however, which suggests that the invariant rate constant assumption
used here introduces little error in the recovery calculation.
DEGRADATION BY ALUMINUM METAL
In some cases, especially CuO, 3H-RNA data points and regression lines
are widely separated from l4C-protein information. This indicates that 3H-RNA
and 14C-protein are desorbing differently and suggests that the virus is
being degraded by the surface.
To investigate this problem further, the experiment with Al metal where
inactivation is most pronounced was repeated, but this time extracting once
into a small volume using lOx elution medium as described for Si02. The
eluted material was then layered on top of 15 to 30% RNAse-free sucrose
gradients prepared in 0.02 I buffer at pH 8.4, equilibrated with atmospheric_
C02. Gradients containing virus from a parallel control which did not contain
44
-------
- 3
- 7 -
In
-15
-19 -
-23
^C-Protein
PFU
n
Figure 8. Kinetic plot for £>-Mn02, Sample C. C'/Adsi is ithe concentration
of virus that desorbs during each sequential extraction,
n = 1 to 4; divided by the total virus initially adsorbed to the
solid. Plotted curves are least-square fit linear regressions.
The error bar represents theoretical assay standard deviation.
45
-------
- 3
- 7
In
-15
•19
-23
14-
C- Protein
3H_
H-RNA
D
PFU
n
CuO
Figure 9. Kinetic plot for CuO, Sample C, C'/Ads.^ is the concentration of
virus that desorbs during each sequential extraction, n = 1 to 4,
divided by the total virus initially adsorbed to the solid.
Plotted curves are least-square fit linear regressions. The error
bar represents theoretical assay standard deviation.
46
-------
TABLE 13. COMPARISON OF VIRUS INITIALLY ADSORBED (AdSi) On CuO WITH VIRUS
RECOVERED BY KINETIC ANALYSIS OF DATA FROM FOUR EXTRACTIONS (Adso)
Measurement
2-hr Incubation —
Sample
Infectivity (log PFU) A
B
3H-RNA (log CPM)
14
C-Protein (log CPM)
Measurement
A
B
A
B
24 -hr Incubation —
Sample
Infectivity (log PFU) C
D
3H-RNA (log CPM)
14C-Protein (log CPM)
PFU =
* Values are standard
aluminum metal, and rtii
in the Spinco model L
Fractions were colled
C
D
C
D
plaque forming units,
deviation estimates .
Duplicates A,B
Ads..^
8.48 ± 0.09*
8.48 ± 0.09
2.91 ± 0.02
2.92 + 0.02
3.09 ± 0.02
3.09 ± 0.02
Duplicates C,E
Ads..^
8.65 ± 0.08
8.65 ± 0.08
2.91 ± 0.02
2.92 ± 0.02
3.11 ± 0.02
3.12 ± 0.02
CPM = counts p
iterial eluted from Al metal surfaces
ultracentrifuge for conditions prese
:ed for 3H-RNA and 14C-protein analys
Ads
o
-7.52 ± 0.32
6.76 ± 0.28
1.97 ± 0.31
1.97 ± 0.43
2.70 ± 0.13
2.60 ± 0.16
Ads
o
6.24 ± 0.17
6.51 ± 0.18
1.62 ± 0.39
1.76 + 0.22
2.40 ± 0.16
2.40 ± 0.23
er minute.
were centrif uged
nted in Figure 12.
is.
Virus from the stock suspension sedimented at the expected 155S. A
small amount of 3H-RNA in the control separated from the virus and sedimented
at 29S. Counts indicated in Figure 12 were normalized to (represent the total
counts in the elution medium for both the control and Al-eluted material.
Sedimentation coefficients were calculated using McEwen's (75) method
and a virus density of 1.34 g/cm3 (CsCl gradient) and RNA density of 1.66
g/cm3 (76). Infectious RNA is reported to sediment at 32S to 35S (77).
47
-------
TABLE 14. COMPARISON OF VIRUS INITIALLY ADSORBED (Adsi) on o-Al203 WITH VIRUS
RECOVERED BY KINETIC ANALYSIS OF DATA FROM FOUR EXTRACTIONS'
Measurement
Infectivity Clog PFU)
3H-RNA Clog CPM)
14C-Protein Clog CPM)
2-hr Incubation —
.Sample
A
B
A
B
A
B
Duplicates A,B
Ads..
8.60 ± 0.07*
8.60 ± 0.07
3.49 ± 0.02
3.49 ± 0.02
3.39 + 0.02
3.39 ± 0.02
AdsQ
9.22 ± 0.41
8.60 ± 0.43
3.43 -t 0.41
3.46 + 0.38
3.46 + 0.33
3.54 ± 0.36
24-hr Incubation — Duplicates C,D
Measurement
Infectivity Clog PFU)
3H-RNA Clog CPM)
14C-Protein Clog CPM)
Sample
C
D-
C
D
C
D
Adsi
8.73 ± 0.07
8.73 ± 0.07
3.51 ± 0.02
3.51 ± 0.02
3.46 ± 0..02
3.46 ± 0.02
Adsb
8.52 ± 0.85
7.86 ± 0,09
3.24 ± 0.28
3.24 ± 0.31
3.28 ± 0.28
3.26 ± 0:33
PFU = plaque forming units, CPM = counts per minute.
*Values are standard deviation estimates.
Virus eluted from aluminum metal was degraded into small fragments that
remained at the top of the gradient. The degraded material at the top of this
gradient contains far more labeled material than low molecular weight impuri-
tiespresent in the original preparation as seen in the control. This rules out
the possibility that the material representing degraded virus could be due only
to impurities in the original preparation.
DISCUSSION
Virus is degraded on aluminum metal and apparently not at all on A1203-
Lifshitz calculations suggest that the metal should generate substantially
stronger Van der Waals forces than the oxide. Earlier, it was suggested
that partial penetration of the oxide coating known to be present on the
48
-------
- 3
- 7
In
-15.
-19 .
-23
n
Al 2 03
Figure 10.
Kinetic plot for a-Al203, Sample C. C'/Ads.^ is
of virus that desorbs during each sequential
n = 1 to 4, divided by the total virus initially
to the solid. Plotted curves are least-square
regressions. The error bar represents theoretical
standard deviation.
49
14C-Prbtem'0
the concentration
extraction,
adsorbed
fit linear
assay
-------
- 3
- 7
In
-15
-19
-23
14C-Protein
. ^
——O—««.™«ion limi,
n
AI Metal
Figure 11. Kinetic plot for Al metal, Sample C. C'/Ads^ is the concentra-
tion of virus that desorbs during each sequential extraction,
n = 1 to 4, divided by the total virus initially adsorbed to the
solid. Plotted lines are least-square fit linear regressions.
50
-------
E
jo
4) .
M t
CO ^/
in *
J2 to
Z
o
u
m
d.
ft!
^
^~
CO
s
O
<
Z
at
CN
o
o
o
CN
§
00
o
o
o
O
O
O
00
o
o
o
51
- -O
"O
00
0)
CO
-------
TABLE 15. COMPARISON OF VIRUS INITIALLY ADSORBED (Adsi) on Al METAL WITH
VIRUS RECOVERED BY KINETIC ANALYSIS OF.DATA FROM FOUR EXTRACTIONS (Adso)
Measurement
2-hr Incubation — Duplicates A,B
Sample
Adsf
Infectivity (log PFU)
3H-RNA (log CPM)
14C-Protein Clog CPM)
A
B
A
B
A
B
8.59 ± 0.07*
8.59 ± 0.07
3.50 + 0.02
3.50 ± 0.02
3.44 + 0.02
3.44 ± 0.02
5.81 ± 0.46
5.96 ± 0.84
3.40 ± 0.18
3.34 ± 0.19
3.43 + 0.14
3.38 + 0.16
24-hr Incubation — Duplicates C,D
Measurement
Infectivity (log PFU)
Sample
C
D
Ads..^
8.27 ± 0.11
8.27 ± 0.11
Adso
< 4
< 4
H-RNA (log CPM)
14
C-Protein Clog CPM)
C
D
C
D
3.12 ± 0.05
3.11 ± 0.05
3.32 ± 0.02
3.31 ± 0.02
2.73 ± 0.10
2.81 ± 0.05
2.76 + 0.09
2.69 ± 0.13
PFU = plaque forming units, CPM = counts,per minute.
*Values are standard deviation estimates.
surface by these forces generated in the metal was responsible for the strong
adsorbent properties of aluminum. These strong Van der Waals forces are
suspected also to be involved in degradation, perhaps by stressing the
adsorbed virion irreversibly.
The metal is demonstrated to be an effective substance for virus ad-
sorption as well as degradation. This indicates that effective filters
could be designed with aluminum metal to remove virus from aerosols and
contaminated water and wastewater, if all pathogenic viruses react similarly.
Additional work must be done with viruses such as influenza, to confirm
suspected broad-spectrum application to virus aerosol filters., and if pos-
sible, with hepatitis A for water and wastewater filters. In the next sectipn,
experimental application to water and wastewater systems will be presented.
52
-------
SECTION 7 :|
APPLICATION OF ALUMINUM METAL TO WASTEWATER DECONTAMINATION
INTRODUCTION . - j .
The final obj ective of the proj ect is to determine whe:ther or not mate-
rials selected by the first two studies could be effectively a.pplied to decon-
tamination of wastewater containing enteric viruses. i
i . •"
Studies of adsorption and degradation suggest that aluminum metal may be
effectively applied to decontamination processes. In adsorption-filtration
operations, we predict that it should be far better for virus removal than
diatomaceous earth (78) or activated carbon (73). In addition, it is inexpen-
sive and its very low levels of dissolution products, Al^+ and associated
hydroxo complexes (-SIS), are ubiquitous in natural waters and are considered
non-toxic. <
A potential problem with its application, however, may; arise if organics
present in water or wastewater block virus adsorption. (
BENCH-SCALE PROCESS EXPERIMENT . . . <
A bench-scale process experiment was constructed to determine whether or
not this blockage would occur with wastewater. The experiment was designed to
be very conducive to rapid virus breakthrough, using a 1-cm thick bed of coarse
aluminum metal powder and a residence time of only0.7min. ; With this design,
if effective virus removals are obtained, there is good evidence that the
material could be applicable to treatment plant conditions.
The apparatus is shown in Figures 13 and 14. Buffer 00.02 I - pH 7) was
continuously pumped at 3.00 + 0.01 ml/min with a Sage model 220 syringe pump
into a magnetically operated flash mixing device at 15 ± 2°C. Shortly after
the start of the experiment, l^c-protein labeled virus suspended in 0.02 I
buffer containing 1% fetal bovine serum to prevent loss of virus to the
syringe needle, plus 1.32 (± 0.05) x 10s CPM/ml tritiated water (3H-H20),
was injected into the flash mixer. A Sage model 355 syringe pump was used
with a flow rate of 3.29(± 0.07) x 10~3 ml/min for this injection. The input
virus concentration was 1.14 (± 0.18) x 106 PFU/ml. i
Midway during the experiment (~3hr), the reservoir delivering 0.02 I
buffer was changed to one containing secondary effluent obtained from the Palo
Alto treatment plant (Table 16) with an added 148 ± 6 CPM/ml 3H-H20.* The
°*'. i
The effluent was allowed to settle for 2 days at 4°C prior to this experiment
to remove coarse suspended material which might stop flow in the tubing used
with the syringe pumps. Total coliforms'= 5.90 (+ 0.87) x 10^ per 100 ml.
53 i
-------
Labeled
Virus
Flash Mix
Fractions
Buffer
Secondary Effluent
5% CO-
in air
Figure 13. Schematic diagram of bench-scale process experiment,
54
-------
Figure 14. Photograph of Bench-Scale Process Experiment
55
-------
-------
'••
' '
'
1
c
I
\
\
\
\
\
\
)
A . I
\
\ »- !
\ "^_ * ^
\ \x /
V •- \
^ ® \ -s
'I u\ = i
*^^ *•!•• I *f ^^J 0
O nZZ \ ^J \ | | TJ
r^ s \^ §
\ M
\ \ t £
AH v <
\ \ ' -2
Nk N 'i
] — ' ^"-> Is*
1 H ' *-*f _
X.--~~ '
/ • • E j
O --~""~ 3
ZC ^s * ^
(VJ ^^ I .y
^^ \ o
>r d
1 ' -
I I
^^^ >
<^_ ^
•7 CN CO -r}- to
1 1 'I 1 i
^lu
Uo> 1
^0 ; |
57 ' \
1
•
O
• O
in
•
4-J
C
o • 1
' ° -O S
Or bd
•*- w
^ CD
us 's
o
0
fl> PM
.§ £ «
•" 33
™^r w
O i
> 1
-— 0)
O PO
o i i
- o ^ £
CN ^
CD
CO
o "^
. O "H
"~
-------
that fluid flow from the virus and buffer reservoirs remains constant during
the experiment.
We find greater relative amounts of ^C-protein ^n t^ie column effluent
than infectious virus. This represents a decreased tendency of virus degrada-
tion products or l^c-iabeled impurities to adsorb compared to infectious virus.
Coliform breakthrough begins at the same time as 3H-H20 breakthrough.
3H-H20 breakthrough indicates that secondary effluent replaced 0.02 jl buffer
in the column. A steady-state output coliform concentration reached! that of
the input, indicating that the bacteria are. resistant to aluminum metal
adsorption and degradation.
The concentrations of infectious virus, total coliforms, and other con-
stituents in the input were determined by sampling the stock input fluids.
Infectivity of the virus preparation was shown to be stable for at least
24 hr in Al equilibrated 0.02 I buffer^ Incubation of virus with our second-
ary effluent for ~6 hr showed no significant loss in titer.
Total coliforms were detenmined immediately after sampling collected
fractions. Since the stock and samples were assayed at the same time, the
respective titers can be compared for recovery. If the coliform numbers in ..
the stock effluent decreased in viability in the effluent during the course of
the experiment, a similar decrease in viability in the collected fractions
would be expected, because the solution conditions after filtration were
essentially the same as in the stock. ,
The suspended solids concentration in input effluent was 1.0 mg/1 and
< 0.4 mg/1 in pooled fractions. COD concentrations are listed below:
Input Secondary Effluent (plus 1:1000 virus stock) 7.6 mg/1
316-321 Void Volumes Eluted 78 mg/1
Output 418-423 " " " 54 mg/1
546-551 " " " 52 mg/1
Only minor amounts of COD appear to be removed by the Al metal column.
Similar experiments with freshly precipitated manganese oxyhydroxides
in the upper region bf a 10 cm column' of coarse Si02 did not give effective
virus removals even with 0.02 I buffer conditions.
DISCUSSION
In Figure 15 we show that the virus infectivity in the Al metal output
was dramatically reduced, and that secondary effluent did not inhibit adsorp-
tion, even after 300 void volumes had been eluted. The column was only 1 cm
thick and the column had a residence time of only 0.7 min. If this degree
of removal is extrapolated to a. similar column 4 cm thick, made of the same
aluminum powder - with identical flow rates, approximately 12 orders of
magnitude reduction in virus titer is estimated by a first order removal rate.
58
-------
Gerba et al. (73) studied poliovirus removal from filtered secondary
effluent in 46 cm, 20 to 50 mesh, granular activated carbon columns with
3 min residence times. Breakthrough of PFU occurred at ~10 bed volumes and
the steady-state concentration of virus was only reduced to!85 ± 20 percent
of the input. Aluminum metal appears to be orders of magnitude more effective.
After ~300 bed volumes, the organics in the secondary effluent plus
those from the 1% fetal bovine serum used to stabilize the input virus inocu-
lum failed to block steady-state virus removal by Al metal. Operation of an
adsorption-filtration unit process may eventually be hindered, however, by
long-term buildup of organic films. Before making a realistic estimate of
the cost-effectiveness of Al metal for virus removal, answers to the potential
surface coating problem must be obtained. Al metal effervesces on reaction
with both acid and base in aqueous systems. This might well be incorporated
into backwashing operations, to strip minor amounts of the metal plus
organic films from the surface, greatly extending the useful operation of the
process metal. j ' ' '.
This investigation has demonstrated that Al metal is effective in
removing poliovirus type 1, strain LSc2ab from wastewater. jFurther investiga-
tion is clearly needed to demonstrate effectiveness with other viruses,
especially hepatitis type A. Finally, engineering optimization studies are
needed to determine whether or not unit processes are cost-effective compared
to other potentially safe virus removal operations..
59
-------
REFERENCES
1. Goldfield, M. Epidemiological Indicators for Transmission of Viruses by
Water. In: Viruses in Water, G. Berg, H. L. Bodily, E. H. Lennette,
J. L. Melnick, and T. G. Metcalf, eds. Amer. Pub. Health Assoc., Wash-
ington, D.C., 1976, pp. 70-85.
2. Melnick, J. L. Detection of Virus Spread by the Water Route. In:
Viruses and Water Quality: Occurrence and Control, V. Snoeyink and V.
Griffin, eds. Univ. 111. Bull. Vol. 69, No. 1, Urbana, 111., 1971.
pp. 114-125.
3. Stone, E. M., et al_. A Waterborne Epidemic of Salmonellosis in River-
side, California, 1965. Amer. Jour. Epidem. 93 (1): 33-54, 1971.
4. Berg, G. Integrated Approach to Problem of Viruses in Water. J. Sanit.
Engr. Div., Amer. Soc. Civil Engr. 79 CSA6): 867-882, 1971.
*
5. Denis, F. Incidences Epidemiologiques de la Contamination Virale des
Aux. Rev: Epidem., MSd. Soc. et Sante1 Publ. 21 (4): 273-302, 1973.
6. Plotkin, S. A., and M. Katz. Minimal Infective Doses of Viruses for
Man by the Oral Route. In: Transmission of Viruses by the Water Route,
G. Berg, ed. Interscience, New York, 1967. pp. 151-166.
7. Brewry, W. A. An Experimental and Theoretical Study of the Movement of
Viruses through Groundwater. Ph.D. Thesis, Stanford University, Stan-
ford, Calif., 1968. 150 pp.
8. Ryan, W. Studies of the Movement of Viruses through Groundwater. Engr.
Thesis, Stanford University, Stanford, Calif., 1967. 137 pp.
9. Cookson, J. T., and W. J. North. Adsorption of Viruses on Activated
Carbon. Environ. Sci. Tech. 1 (1): 46-52, 1967.
10. Cookson, J. T. Adsorption of Viruses on Activated Carbon. Environ. Sci.
Tech. 1 (2): 157-160, 1967.
11. Gerba, C. P., C. Wallis, and J. L. Melnick. Fate of Wastewater Bacteria
and Viruses in Soil. J. Irrig. Div. Amer. Soc. Civil Engr., 101 (IR3):
157-174, 1975.
12. Warren, J., M. Kende, and K. Takano. The Adjuvant Effect of Powdered
Ferric Oxide? Enhancement of Response-of Mycoplasma. Prieumoniae and Res-
piratory Syncytial Virus Vaccines. J. Immun. 102 (5): 1300-1308, 1969.
60
-------
13. Mitchell, R., and H. W. Jannasch. Processes Controlling Virus Inactiva-
tion in Seawater. Environ. Sci. Tech. 3 (10): 941-943, 1969.
14. Smith, E. M., C. P. Gerba, and J. L. Melnick. Role of Sediment in the
Persistence of Enteroviruses in the Estuarine Environment. Appl. Envi-
ron. Microbiol. 35 (4): 685-693, 1978.
15. Schaub, S. A., and B. P. Sagik. Association of Enteroviruses with Natu-
ral and Artificially Introduced Colloidal Solids in Water, and Infecti-
vity of Solids Associated Virions. Appl. Microbiol. 30 (2): 212-222,
1975. ;
16. Environmental Protection Agency. Preliminary Assessment of Suspected
Carcinogens in Drinking Water - Report to Congress, December, 1975.
17. Page, T., R. Harris, and J. Bruser. Removal of Carcinogens from Drink-
ing Water: A Cost-Benefit Analysis. California Inst. Technology Social
Science Working Paper 230, 1979, 82 pp. 1
18. Simmon, V. F., K. Kauhanen, and R. Tardiff. Mutagenic, Activity of Chemi-
cals Identified in Drinking Water. In: Progress in Genetic Toxicology,
D. Scott, B. A. Bridges, and F. Sobels, eds. Elsevier- North Holland,
Biomedical Press, New York. pp. 249-258, 1977. !
19. Gruener, N. Mutagenicity of Ozonated, Recycled Water. Bull. Environm.
Contam. Toxicol. 20: 522-526, 1978.
I
20. Derjaguin, B., and L. D. Landau. Theory of the Stability of Strongly
Charged Lyophobic Sols and of the Adhesion of Strongly Charged Particles
in Solutions of Electrolytes. Acta Physicochim. U.R.S;.S. 14: 633, 1941.
21. Verwey, E. J. W., and Overbeek, J. Th. G. Theory of the Stability of
Lyophobic Colloids. Elsevier, New York, 1948. 205 ppL
22. Dzyaloshinski, I. E., E. M. Lifshitz, and L. P. Pitaevski. The General
Theory of Van der Waals Forces. Adv. Phys. 10 (37): 165-209, 1961.
23. Fenner, F., B. R. McAuslan, C. A. Mims, J. Sambrook, and D. 0. White.
The Biology of Animal Viruses, 2nd ed. Academic Press, New York, 1974.
834 pp.
• I
24. England, B., R. E. Leach, B. Adame, and R. Shinosaki. ' Virologic Assess-
ment of Sewage Treatment at Santee, Calif. In: Transmission of Viruses
by the Water Route, G. Berg, ed. Interscience, New York, 1967; pp. 401-
417. ;
25. Chang, S.' L. Statistics of the Infectious Units of Animal Viruses.
In: Transmission of Viruses by the Water Route, G. Berg, ed. Inter-
science, New York, 1967. pp. 219-234. '
61
-------
26. Schwerdt, C. E., and F. L. Schaffer. Purification of Poliomyelitis
Viruses Propagated in Tissue Culture. Virol. 2 (5): 665-678, 1956.
27. McClain, M. E., and C. E. Schwerdt. Use of HeLa Cells for Plaque Assay
of Poliomyelitis Virus. Fed. Proc. Amer. Soc. Exptl. Biol. 13 (1):
505, 1954.
28. White, A., P. H. Handler, and E. L. Smith. Principles of Biochemistry,
5th ed. McGraw-Hill, New York, 1973. pp. 806,
29. Charney, J., R. A. Machlowitz, and D. S. Spicer. The Adsorption of
Poliovirus Antigen to Glass Surfaces. Virol. 18 (3): 495-497, 1962.
30. Charney, J., R. Machlowitz, A. A. Tytell, J. F. Sagin, and D.;S. Spicer.
The Concentration and Purification of Poliomyelitis Virus by Use of
Nucleic Acid Precipitation. Virol.,15 (3):. 269-280, 1961.
31. Maizel, J. V., and D. Summers. Evidence for Differences in Size and
Composition of the Poliovirus Specific Polypeptides in Infected HeLa
Cells. Virol. 36 (1): 48-54, 1968.
32. Joklik, W. K., and J. E. Darnell. The Adsorption and Early Fate of
Purified Poliovirus in HeLa Cells. Virol. 13 (4): 439-447, 1961.
33. Wang, C. H., D. L. Willis, and W. D. Loveland. Radiotracer Methodology
in the Biological, Environmental, and Physical Sciences. Prentiss-Hall,
Englewood Cliffs, N.J., 1975. 480 pp,
34. Murray, J. P. Thermodynamics of Poliovirus Adsorption. Ph.D. Thesis,
Stanford University, Stanford, Calif., 1978. 196pp.
35. Stumm, W. and J. J. Morgan. Aquatic Chemistry. Interscience, New York,
1970. 583 pp. :
36. Wiese, G. R., and T. W. Healy. Solubility Effects in Al^O and TiO
Colloidal Dispersions. J. Colloid Interface Sci. 52 (3): 452-457, 1975.
37. Wagman, D. D., W. H. Evans, V. B. Parker, I. Halow, S. M. Bailey, and
R. H. Schumm. Selected Values of Chemical Thermodynamic Properties.
N.B.S. Tech. Note 270-3. U.S. Govt. Print. Office, Washington, D.C.,
1968. 264 pp.
38. DiStephano, T. H., and D. E. Eastman. .Photoemission Measurements of the
Valence Levels of Amorphous Si02 . Phys. Rev. Letters 27 (23): 1560-
1562. 1971. I
39. Taras, M. J., A. E. Greenberg, R. D. Hoak, and M. C. Rand. Standard
Methods for the Examination of Water and Wastewater, 13th ed. Amer. Pub-
lic Health Assoc., Washington, D.C., 1971. 874 pp.
62
-------
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
"53.
Dhar, H. P., B. E. Conway, and K. M. Joshi. On the Form of Adsorption
Isotherms for Substitutional Adsorption of Molecules of Different Sizes.
Electrochim. Acta. 18 (11): 789-798, 1973. !
Brunauer, S., P. Emmett, and E. Teller. Adsorption ojf Gases in Multi-
molecular Layers. J. Amer. Chem. Soc. 60 (2): 309-319, 1938.
I
Jura, G. The Determination of the Area of the Surfaces of Solids.
In: Physical Methods in Chemical Analysis, Vol. 2., W. G. Berl, ed.
Academic Press, New York. 1951. pp. 255-303. I
Vuceta, J. Adsorption of Pb(II) and Cu(II) on a-Quartz from Aqueous
Solutions: Influence of pH, Ionic Strength, and Complexing Ligands.
Ph.D. Thesis, Calif. Inst. of Tech., Pasadena, Calif.! 1976. 206 pp.
O'Melia, C. R., and W. Stumm. Aggregation of Silica Dispersions by
Iron(III). J. Colloid Interface Sci. 23 (3): 437-447, 1967.
American Association for Testing Materials. Standard]Methods for Test
of Fineness for Portland Cement by Air Permeability Apparatus, C-204-73
Vol. 13, 1973, pp. 182-188. j
Schultz, N. F. Measurement of Surface Areas by Permeametry. Internat.
J. Min. Proc. 1: 65-79, 1974. •-
I
Parks, G. A. Adsorption in the Marine Environment, in: Chemical
Oceanography, Vol. 1, G. Riley and G. Skirrow, eds. Academic Press,
New York, 1976. pp. 241-308. j
:j
Gregory, J. Interaction of Unequal Double Layers at Constant Charge.
J. Colloid Interface Sci, 51 (1): 44-51, 1975. j
]
Smith, A. L. Electrokinetics of the Oxide Solution Interface. J. Col-
loid Interface Sci. 55 (3): 525-530, 1976. !
Visser, J. The Adhesion of Colloidal Polystyrene Partlicles to Cello-
phane as a Function of pH and Ionic Strength. J. Colloid Interface Sci
55 (3): 664-677, 1976. j
i
Wiersema, P. H., A. L. Loeb, and J. Th. G. Overbeek. 'calculation of
the Electrophoretic Mobility of a Spherical Colloid Particle: J.
Colloid Interface Sci. 22 (1): 78-99, 1966. j
j
Hunter, R. J., and H. J. L. Wright. The Dependence of Electrokinetic
Potential on Concentration of Electrolyte. J. Colloid Interface Sci
37 C3): 564-580, 1971. j
Murray, J. P. and C. E. Schwerdt. Calculation of Zeta: Potentials of
Suboptical Particles by Electrophoresis in Sucrose Gradients. Unpub.
Res. 1978. !
63
-------
54. Yates, D. E., S. Levine, and T. Healy. Site-Binding Model of the Elec-
trical Double Layer at the Oxide/Water Interface. J. Chem. Soc. Faraday
Trans. I. 70: 1807-1818, 1974.
55. LaColla, P., M. Marcialis,, G. P. Mereu, and B. Loddo. Behavior of a
Guanidine-Dependent Strain of Poliovirus 1 in Sucrose Density and pH
Gradients. Experientia 28 (9): 1115-1117, 1972.
56. Mandel, B. Characterization of Type 1 Poliovirus by Electrophoretic
Analysis. Virol. 44 (3): 554-568, 1971.
57. Ward, R. L. Mechanism of Poliovirus Inactivation by Ammonia. J. Virol.
26 (2): 299-305, 1978.
58. Floyd, R., and D. G. Sharp. Viral Aggregation: Effects of Salts on the
Aggregation of Poliovirus and Reovirus at Low pH. Appl. Environ. Micro-
biol. 35 (6): 1084-1094,, 1978. . .
59. Young, D. C., and D. G. Sharp. Presented at ASM National Meeting,
Honolulu, 1979.
60. Lipatov, Yu. S., and L. M,. Sergeeva. Adsorption of Polymers. Halsted
Press, New York, 1974. 177 pp. ;
61. Parsegian, V. A. Long Range Van der Waals Forces. In: Physical Chem-
istry: Enriching Topics from Colloid and Surface Science, H. van Olphen
and K. Mysels, eds., I.U.P.A.C. 1.6, Theorex, La Jolla, Calif.1, pp. 27-
72, 1975.
62. Richmond, P. The Theory and Calculation of Van der Waals Forces. In:
Colloid Science, Vol. 2, D. Everett, ed. The Chem. Soc., London, 1974.
pp. 130-172.
63. Hamaker, H. C. The London-Van der Waals Attraction between Spherical
Particles. Physica IV (10): 1058-1072, 1937.
64. Bargeman, D. and F. Van Voorst Vader. Van der Waals Forces between
Immersed Particles. J. Electroanal. Chem. Interfacial Electrochem. 37
(1): 42-52, 1972.
65. Visser, J. Adhesion of Colloidal Particles. In: Surface and Colloid
Sci., Vol. 8, Chap. 1, E. Matijevic, ed. Interscience, New York, 1976.
pp. 3-84.
66. Ninham, B., and V. A. Parsegian. Van der Waals Forces, Special Charac-
teristics in Lipid-Water Systems and a General Method of Calculation
Based on Lifshitz Theory. Biophys. J. 10: 646-674, 1970.
67. van Oss, C. J., et al. Comparison between Attachment and Detachment
Approaches to the Quantitative Study of Cell Adhesion to Low-Energy
Solids. J. Colloid Interface Sci., 59 (3): 505-515, 1977.
64
-------
68. Luke's, F. Oxidation of Si'and GaAs in Air at Room Temperature. Surface
Sci. 30 (1): 91-100, 1972. «
|
69. Smith, J1. Auger Electron Spectroscopy and Ion Sputter Profiles of Oxides
on Aluminum. Surface Sci. 55 (2): 601-624, 1976. i
70. Murray, J. P., and G. A. Parks. Poliovirus Adsorption of Oxide Surfaces
- Correspondence with the DLVO-Lifshitz Theory of Colloid Stability.
A.C.S. Adv. in Chem. Series, Particulates in Water, M. Kavanaugh and J.
Leckie, eds. In Press, 1980. , j
i
71. LeNeveu, D. M., R. Rand, V. Parsegian, and D. Gingell.; Measurement and
Modification of Forces between Lethicin Bilayers. Bidphys. J. 18: 209-
230, 1977. i
72. Wallis, C., and J. L. Melnick. Concentration of Viruses on Aluminum
Phosphate and Aluminum Hydroxide Precipitates. In: transmission of .
Viruses by the Water Route, G. Berg, ed. Interscience, New York, 1967.
pp. 129-138. |
i
73. Gerba, C. P., M. Sobsey, C. Wallis, and J. L. Melnick.! Adsorption of
Poliovirus onto Activated Carbon in Wastewater. Env. jSci. Tech. 9 (8):
727-731, 1975. !
74. Hill, W. F., E. W. Akin, and W. H. Benton. Detection of Viruses in
Water: A Review of Methods and Application, In: Virus, and Water
Quality: Occurrence and Control, V. Snoeyink and V. Griffin, eds.,
Univ. 111. Bull. Vol. 69, No. 1, Urbana, 111., 1971. pp. 17-46.
75. McEwen, C. R. Tables for Estimating Sedimentation through Linear
Concentration Gradients of Sucrose Solution. Anal. Bijochem. 20 (1) :
114-149, 1967. !
76. Jordan, L., and H. D. Mayor. Studies on the Degradation of Poliovirus
by Heat. Microbios 9 (33-34): 51-60, 1971. j
i
i
77. Stone, M. R. Studies on the Reconstitution and Structure of Three
Strains of Poliovirus Type I. Ph.D. Thesis, Stanford lUniversity,
Stanford, Calif., 1977. 175 pp. j ..
78. Brown, T. S., J. F. Malina, B. D. Moore, and B. P. Sagjik. Virus Removal
by Diatomaceous Earth Filtration. In: Virus Survival in Water and
Wastewater Systems, J. Malina and B. Sagik, eds. Univ. Texas Press,
Austin, 1974. pp. 129-144.
65
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO,
EPA-600/2-80-134
3. RECIPIENT'S ACCESSIOWNO.
4. TITLE AND SUBTITLE
PHYSICAL CHEMISTRY OF VIRUS ADSORPTION AND
DEGRADATION ON INORGANIC SURFACES
Its Relation to Wastewater Treatment
5. REPORT.DATE
August 1980 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
James P. Murray
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Applied Earth Sciences
Stanford University
Stanford, California 94305
10. PROGRAM ELEMENT NO.
P.E.35B1C A/14
11. CONTRACT/GRANT NO.
R-805016
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory--Cin.,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final Sept. 1976-March 1978
14. SPONSORING AGENCY CODE
EPA/600/14
IB. SUPPLEMENTARY NOTES
Project Officer: Albert D. Venosa (513) 684-7668
16. ABSTRACT
The DLVO-Lifshitz theory of colloid stability is applied to adsorption
of poliovirus on oxide surfaces common in soil and aquatic environments.
Excellent agreement was found between colloid stability theory and adsorption
free energies calculated from mass-action principles. Colloid stability
theory now provides an organized frame of reference with which to understand
virus adsorption in the environment. On some surfaces, notably S-Mn02, CuO,
and Al metal, kinetic analysis of data from multiple extractions and sqdi-
mentation analysis reveals that virus was actually degraded by adsorption.
A column experiment also indicated that Al metal also effectively degraded
virus in the presence of secondary wastewater effluent. Potential applica-
tions to wastewater treatment are suggested. '
This report was submitted in fulfillment of Grant No. R-805016 by:
Stanford University under the sponsorship of the U.S. Environmental Projection
Agency. This report covers the period September 1, 1976 to March 1, 1978.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Physical chemistry, Adsorption, Adsorbents.
Desorption, Viruses, Disinfection, Waste
treatment, Aluminum inorganic compounds,
Copper inorganic compounds, iron inorganic
compounds, silicon inorganic compounds
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Colloid stability
theory, Van der Waals
forces, Double layer
interaction, Virus
inactivation, Virus
adsorption
07D
OGM
13B
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)'
Unclassified
21. NO. OF PAGES
78
20. SECURITY CLASS (Thtspage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
66
U.S. GOVERNMENT PRINTING OFFICE: 1980--657-165/0125
-------
-------
-------
-------
CD
00
o
o
s*? %!!
Q) Q)
TOW)
T
ON
O
O
CX5
Lo
-C-
5 o c
5o °"
531
I? I
ill
1 8 3
Q.O 3
3? £
3-3
(D O
01 I
as
3 ^
SS
2S
f.q
cn
•a
CD
o
00 c
o ^J
o 3-
7T i
n
CD
33
D)
> I
(Q
(D
O
5'
111
Q) r»i
-'
o
I
a>
oo
m > TJ m
-0(0 3 3
>§!!•
co o o o
en 6' q
IPI
------- |