United States
Environmental Protection
Agency
Water Engineering
Research Laboratory
Cincinnati OH 45268
 Research and Development
EPA/600/S2-87/002 May 1987
Project Summary
Factors  Affecting the
Bioavailability of  Cadmium

R. B. Corey, P. A. Helmke, D. R. Keeney, and G. C. Gerloff
  Results are presented  from two
studies designed to identify factors af-
fecting the bioavailability  of Cd and
other trace metals in soils amended
with municipal sewage sludges and to
develop analytical methods for measur-
ing those factors. The research included
field, greenhouse, growth-chamber, and
laboratory studies. The  field  studies
were conducted on the agricultural re-
search stations  at  Arlington  and
Hancock, Wisconsin. Growth-chamber
studies were done at the University
Biotron, and the greenhouse and lab-
oratory studies  were conducted in
Department of Soil Science facilities at
the University of Wisconsin-Madison.
  The first study was entitled  "Heavy
Metal Bioavailability in Sludge-amended
Soils" and included  (1) field  studies
with two sludges containing about 200
mg Cd/kg to determine the effects of
application rate, sludge properties, soil
texture, and liming on uptake of Cd and
other trace metals by corn, (2) green-
house studies of the same factors but
with three sludges and ten soils, and (3)
laboratory studies to develop methods
for measuring parameters that  have
been shown to affect bioavailability of
the trace metals. The second study was
entitled "Bioavailability of Cadmium"
and was concerned mainly with further
developing and  applying methods  for
predicting bioavailability of Cd and
other trace metals in soils amended
with sewage sludge.
  This Project Summary was developed
by EPA's Water Engineering Research
Laboratory, Cincinnati, OH, to announce
key findings of the research project that
Is fully documented In a separate report
of the  same title (see Project  Report
ordering Information at back).
 Introduction
  Municipal sewage sludges have been
 shown to be effective sources of plant
 nutrients, particularly N and P, when
 applied to agricultural  land.  However,
 some sludges  contain relatively high
 levels of trace metals, which under some
 circumstances  can produce  toxic re-
 actions in plants and/or  animals. The
 trace metal Cd is of  particular concern
 because it can accumulate in sufficient
 quantities in plant tissues to be toxic to
 animals or humans ingesting those tis-
 sues over an extended period. The pur-
 pose of the investigations reported here
 was to identify  sludge and soil char-
 acteristics that affect the bioavailability
 of Cd and other trace metals  and to
 develop analytical methods for measuring
 those factors. As our knowledge of the
 sludge-soil-plant system increased during
 these studies, it became apparent that a
 particularly effective research approach
 would be to devise methods that would
 provide  inputs to a computer  model
 describing solute transport and plant
 uptake.
  Transport of solutes in a soil  system
can occur by mass flow (moving with the
 labile soil water) or by diffusion (moving
 in response to a concentration gradient).
Transport of metal downward in the soil
profile occurs dominantly by mass flow,
whereas transport to plant roots is mainly
a diffusion process (even for some con-
servative solutes such as nitrate), except
when the species being absorbed are at
very high concentrations. The soil and
plant factors that affect the rate at which
a plant  absorbs  a  given metal from
solution in a diffusion-limited system are
shown in Equation 1. This equation is a
modification of an uptake equation that
describes the diffusive radial flux of solute

-------
(in our case a metal) from an isotropic
medium (soil) to a cylindrical sink (plant
root),  in time, t assuming depletion of a
cylindrical volume of soil  surrounding
each segment of root.
U = CMb(1
     - exp
                 -27raAir0L»t
                                  (D
                 D,0f
1.65 rn
Soil factors:  C,, = initial concentration of
                metal in soil solution
             b = buffer  power,  the
                change  in  concentra-
                tion of total  labile forms
                (adsorbed + dissolved)
                per unit of change in
                concentration of dis-
                solved metal
            A| = fractional area of water
                contacting root
             0 - volumetric  water  con-
                tent
            D, = diffusion coefficient for
                the metal in soil solution
             f = conductivity factor re-
                lated  to length of dif-
                fusion path
Plant factors: U = uptake per unit volume
                of soil in time, t
             a = root-absorbing  power
                (uptake  flux  density)/
                (cone, at root surface)
             r0 = root radius
             rh = half-distance  between
                roots
            LV = root density
  We have developed analytical methods
that determine the proportions of metals
present as dissolved free ions,  dissolved
complexes, and reactive adsorbed (labile
forms). These methods allow us to cal-
culate the dissolved  metal concentration,
CN,  and buffer power, b,  in  Equation 1.
We also developed methods  for control-
ling metal-ion activities at realistic soil-
solution  levels  in solution  cultures to
study the effects of speciation of dissolved
metals on the root-absorbing power, a,
and on competition  among metal ions in
the uptake process.
Development of Analytical
Methods

Total Labile Metals
  The accepted  standard method for
determining labile elements  is isotopic
exchange if suitable isotopes are avail-
able.  Thus, 109Cd and 65Zn were equili-
brated with samples of sand and silt loam
soils with and without addition  of  102
Mg sewage sludge/ha. Isotopically ex-
changeable Cd and Zn were determined
after 16 hr, and the results were com-
pared with amounts of Cd  and  Zn ex-
tracted  by various diethylenetriamine-
pentaacetic acid (DTPA) and ethylenedia-
minetetracetic acid (EDTA) solutions.
Metal  extraction  by 0.005M EDTA in
0.01M CaCI2 at a 1:2.5 soihsolution ratio
gave the best agreement with isotopically
exchangeable  values. DTPA extraction
consistently gave low results.

Metal Speciation In Soil
Suspensions
  The method developed for determining
metal  speciation  in  soil  suspensions
involves equilibration of separate  soil
samples with each of four solutions: (1)
0.01 M Ca(N03)2,  (2) 0.01 M CaC12, (3)
0.01 M Ca(N03)2 +  10;45M EDTA  (or hy-
droxyethylenediaminetriacetic  acid
(HEDTA), and (4) 0.01 M Cad 2 + 0.005M
EDTA. If Cd is not detectable in the 0.01 M
Ca(N03)2 solution,  Cd(N03)2  is added to
all solutions to give a final Cd concentra-
tion of 5 mg/L. Equilibration with Solution
1 gives total  dissolved metals. The in-
crease in Cd concentration in Solution 2
over Solution 1 results from the formation
of cadmium chlorine complexes. Because
the stability constants for these complexes
and the CI" activity are known, the activity
of Cd2* can be calculated. The increases
in metal concentrations  in Solution  3
over Solution 1  result from the formation
of metal-EDTA complexes. The activity of
any  divalent  metal ion,  (M2t),  can be
calculated from Equation 2 once the Cd2t
activity and relevant competitive stability
constant (K) are known:
                                                         tion  1. An overall competitive stability
                                                         constant (KN) for the natural complexes
                                                         can be calculated from Equation 3.
       (M2+) = K(Cd2+)
                     [M-EDTA]
                                                   (2)
                     [Cd-EDTA]
If concentrations of any of the metals are
below detection limits in Solution 1 but
are detectable in  Solution 2, activities
can  be calculated, but information on
dissolved complexes cannot be obtained.
The  Solution 4 extract is a  measure of
labile metals. From the ratio of metal
activities in solution to labile metal con-
centrations, distribution coefficients can
be calculated for the partitioning between
solution and solid phases. Concentrations
of these metal ions can be determined
from  the calculated activities and ionic
strength.  Concentrations of  dissolved
metal complexes (M-ch), are determined
by subtracting the  calculated metal ionic
concentration from the respective total
metal concentration determined in Solu-
                                                                        (M2lCd-ch]

                                                                        (Cd+2HM-ch]
                                                                                                                    (3)
  The method used for calculating free
metal ion  activities  assumes that  the
addition of CaCI2 IN Solution 2 or EDTA in
Solution 3 does not alter the equilibrium
significantly; that is, the amount extracted
into solution is not a  significant part of
the labile fraction. If this assumption does
not hold, the calculated activities will be
lower than  those in a natural system.

Speciation of Dissolved Metals
  Dissolved metals are partitioned  be-
tween hydrated free-ions,  ion pairs, and
organic  and  inorganic  complexes. The
concentrations of total dissolved species
are usually within the detection limits of
sensitive analytical  methods such  as
flameless  atomic absorption spectro-
photometry (FAAS), but present methods
that measure activities of free metal ions
are either not sensitive enough to mea-
sure the low activities present in most
soil solutions, or  they are subject to  in-
terference by components of the solution.
The method for determining metal specia-
tion in soil suspensions estimates free-
ion activities, but it requires contact with
a  solid  phase  to maintain  a constant
metal activity. In order to overcome these
limitations  a sensitive method utilizing a
Donnan* equilibrium system was devel-
oped. In this method a cation-exchange
membrane separating  a donor compart-
ment (through which the test solution is
pumped) from a much smaller compart-
ment containing an  acceptor solution.
The latter is similar in composition to the
test solution with respect to major cations.
Free  metal ions  equilibrate  across the
membrane in about 1 hr, but transfer of
complexed species is greatly retarded.
Thus, analysis  of the  acceptor solution
after 1 hr by FAAS gives the concentration
of the free ion, from which an activity
may be calculated. The detection limit is
determined by the detection limit of the
FAAS analysis.


Characterization of Soluble
Complexlng Agents
  Chelating resins which  provide a  re-
servoir  for metal ions and  maintain
                                                         •Mention of trade names or commercial products
                                                          does not constitute endorsement or recommenda-
                                                          tion for use

-------
constant metal ionic activity are used to
control  metal-ion  activities  so that
competitive selectivity coefficients can be
determined for  dissolved ligands.  The
chelating resin is placed in a dialysis bag
and equilibrated with the solution which
originally contains the dissolved ligand or
ligands but not the metal ion. The amount
of metal complex formed is the difference
between the  total metal concentration
and the metal ion concentration supported
by the resin.  Competitive stability con-
stant (K) is calculated from the following
equation where M and N are  different
metals, L is the dissolved ligand and ML
and NL and the concentration  of metal
ligands:
                [ML]  (N2+)
             K = -
                                   (4)
                [NL]
  A method was developed for estimating
the Cu-complexing capacity of dissolved
ligands. In this method, samples of solu-
tion containing the dissolved ligand re-
ceived varying  amounts of added Cu2*.
The difference between the total Cu con-
centration determined by FAAS and Cu2+
concentration (determined either with the
Donnan-equilibri urn/cat ion-exchange
membrane system or with an ion-specific
electrode when Cu2+ activities are above
10~6M) gave the Cu-complexing capacity
at the measured Cu2+ activity.

Cadmium Adsorption  Isotherms
  A method was developed for adsorbing
Cd onto soil surfaces under conditions of
constant pH and Cd2+ activity. To produce
solutions with known and constant Cd2+
activity  varying  ratios of  Cd-saturated
chelating resin to Ca-saturated chelating
resin were mixed and added to a dialysis
bag. A pH-adjusted, weak-acid resin was
also added for pH control. The resin mix-
ture in  a  dialysis bag was then equili-
brated  with  a soil  sample  in  0.01 M.
Ca(NO3)2, and the Cd adsorbed on the soil
was extracted with acid and determined.
The Cd adsorbed on  the  soil must  be
small compared with  that contained in
the chelating resin so that a nearly con-
stant, known Cd2+ activity  is maintained.
Competition for adsorption sites on the
soil  is limited to Cd2* and Ca2+ as the
resin adsorbs other metals  that were
originally adsorbed on the  soil. Competi-
tive adsorption at known metal activities
could be studied by including other metals
in the chelating resin.

Control of Solute Concentrations
In Nutrient Solutions
  If nutrient solutions are  to be used for
determining root-absorbing powers for
trace metals, concentrations of those
metals  at  the root surfaces must  be
known. However, the nutrient solutions
do not buffer metal concentrations, and
even rapidly flowing solutions  do not
maintain constant metal concentrations
at root surface when concentrations are
in normal soil solution  ranges. For this
reason, a resin-buffered nutrient solution
system  was developed. In this system, a
strong-acid  resin controls the ratios of
Ca2+, Mg2+,  K+, and Mn2+;  a weak-acid
resin buffers pH and assists in controlling
ratios  of-  the previously  mentioned
cations; a  chelating resin  controls the
ratios of trace metals,  including  Cd2t,
Zn2+, and  Cu2+; and a strong-acid  resin
partially  saturated  with  (AI(OH)2+)n
polynuclear complexes  controls  P con-
centrations  and  assists in  controlling
ratios of the major cations. Anions such
as NO3  and S04= must be maintained by
periodic additions. For most of the cations,
the ratios are controlled by the resins; but
the absolute concentrations are a function
of ionic strength. Thus the  total solute
concentration as indicated by electrical
conductivity must be controlled within
rather narrow limits. This system has
been used successfully with the resins
enclosed in filter membrane bags, floating
free  in the pot, or contained  in a column
through which the nutrient  solution is
recirculated.

Factors Affecting Cadmium
Bioavailability

Sludge Characteristics
  Data from laboratory, greenhouse, and
field  all  show that sludge characteristics
are extremely important, if not dominant,
in controlling Cd (and Zn) availability to
plants. From 18% to 63% of the total Zn,
and from 12% to 63% of the total Cd in
the sewage sludges studied was isotopi-
cally exchangeable.  The greatest  per-
centage was for Chicago sludge, whereas
the lowest was for a sludge that received
soluble Fe during the treatment process.
When corrected for the effects of pH, the
fractions of isotopically exchangeable Zn
and Cd extracted by 0.01 M Ca(NO3)2  were
about the same for all sludges and soils
studied. This  result suggests that the
mechanism  by which solubility is  con-
trolled is coprecipitation (solid solution)
of Zn and Cd with phosphates of Fe, Al, or
Ca, with hydrous oxides of Fe or Al, or
with  CaC03 during the treatment process.
  According to solid solution theory, the
solubility of a trace constituent should be
proportional to its relative concentration
on the surface of the precipitate. Thus for
equal  concentrations of Cd in  different
sludges, the solubility should decrease as
the  amount  of the precipitate's major
component (FePO4, AIP04, etc.) increases
with  respect to the  Cd concentration.
This effect is apparent when comparing
results for the Chicago sludge (low Fe, Al,
and P) with those for the Wisconsin
Rapids sludge (high Fe and P). Although
total concentrations of Cd were nearly
the same for both sludges, the isotopically
exchangeable Cd, solution Cd, and Cd in
barley leaves in the field were all much
higher for the Chicago sludge.
   Similarly, in sludges with similar levels
of the major precipitant, the solubility of
Cd should increase with the  total  Cd
concentration. This result is borne out
(although not as  clearly  as would  be
desired because of  the low additions in-
volved) in  the  greenhouse  study com-
paring plant uptake of Cd at equal  Cd
additions from the Oshkosh (9 mg Cd/kg)
and Wisconsin Rapids (180  mg Cd/kg)
sludges.
   If these observations on a few sludges
were found to hold  as a general rule, Cd
solubility might be predicted on the basis
of sludge  composition and verified by
analyzing for the capacity factor by iso-
topic dilution or ligand extraction and for
the intensity factor  by equilibration with
Sr(N03)2 or Ca(NO3)2.

Sludge Application Rate
  At high sludge application rates, the
properties of the sludge tend to determine
Cd and Zn availability, with the effects of
soil properties being limited primarily to
pH control. At low sludge  rates,  the
adsorptive properties of the soil appear to
exert some control.
  Increasing  additions of the  high-Cd
Wisconsin Rapids sludge to soils in the
greenhouse resulted in proportionate
increases in DTPA-extractable Cd. Similar
increases  in the levels of soluble DTPA-
extractable and isotopically exchangeable
Zn and Cd were observed in the field as
the rate of sludge  application  was  in-
creased. However,  these  proportionate
increases  in labile metal levels were not
consistently reflected in crop uptake. The
Cd concentration  in  corn  tissue, par-
ticularly in the third  greenhouse crop,
tended to plateau at about 80  Mg/ha
(18.3 kg Cd/ha) in the Plainfield sand
and  160 Mg/ha (36.6 kg Cd/ha) in the
Piano silt  loam. Tissue Cd levels in the
field also  tended to plateau at higher
sludge rates, but the  effect  was not as
apparent as in the greenhouse  because

-------
of the lower rates used in the field. This
relationship was somewhat confounded
in the field by lower pH induced by the
high sludge rates.
  In the field, the concentrations of Cd in
corn ear leaves tended to reach a plateau
between 1 and 1.5 mg/kg. Very little Cd
was found  in  the corn grain.  In  the
greenhouse, the Cd concentration in plant
tissue from the third corn crop reached a
plateau at about 7 mg/kg. This level is
much higher than in the field, even con-
sidering that the greenhouse tissue was
less mature. This result suggests that the
greater transpiration ratio of the green-
house crop may have resulted in greater
Cd uptake  because of  the  increased
amount of Cd arriving at the root surface
by mass flow. Alternatively, Cd uptake in
the field may have been reduced by root
penetration  into deeper soil regions not
affected by sludge addition.

Time Following Sludge
Application
  In both  the greenhouse  and  field
studies, DTPA-extractable Cd tended to
increase with time,  but  the  effect ex-
tended over a longer period in the field,
probably because  the  sludge particles
were coarser and took longer to interact
completely with the soil. Mixing sewage
sludge  with soil  did not  change  the
isotopic exchangeability of either com-
ponent immediately (i.e.,  the concentra-
tion of  isotopically exchangeable Zn in
the mixture remained the weighted aver-
age of the values for the soil and sludge).
However, mixing  did increase the con-
centration of  isotopically exchangeable
Cd by 30% to 40% in the mixture after 1
year.
  The concentration of Cd in tissue tended
to increase from crop-to-crop, both in the
greenhouse and the field. This result is
contrary to  those  of most field studies
reported in  the literature. However, in-
terpreting year-to-year differences in crop
Cd concentrations as due to changes in
lability of sludge Cd alone is questionable
because Cd  uptake  from a given  soil
might  also  be affected by changes in
environmental conditions such as tem-
perature, moisture, pH, soluble salts, and
evapotranspiration.

SoilpH
  The combination of liming and sludge
application resulted in a  wide pH range
— greater than 2 pH units in the Piano
field plots. Isotopic exchange studies on
soil samples from these plots indicated
that the concentration of isotopically ex-
changeable Zn in  the  sludge-amended
soil was independent of pH. However,
isotopically exchangeable Cd at pH 6.2
after 1 year was only 80% of that mea-
sured at pH 5.4. Similarly, DTPA-extract-
able Cd on the limed plots in 1979 (after
3 years of equilibration) averaged 67% of
that of the unlimed plots. Graphs of the
negative  logarithms of solution concen-
tration versus pH for control and sludge-
treated plots gave  a slope of 1.0 for Zn
and 0.7 for Cd over a pH range of 4.6 to
6.8. An increase in 1 pH unit therefore
decreased the intensity factor of bioavail-
ability (concentration in solution) tenfold
for Zn but only fivefold for Cd. The values
of the slopes were independent of the
sludge application rate.
  This marked decrease in  soluble Cd
with increasing  pH is  not consistently
reflected in a corresponding  decrease in
Cd uptake in the field, however. Although
the  limed and unlimed plots on the Piano
soil differed by approximately 1 pH  unit,
no significant  effects of pH  on Cd  con-
centration in the leaves of corn or barley
were  observed.  Only in the corn  leaf
tissue from Hancock at the high rates in
1978 and at all rates in 1979 was there a
markedly greater Cd concentration in corn
ear  leaves from unlimed plots. Why the
increased solution Cd at lower pH did not
result in  consistently greater Cd uptake
from the Piano soil  is not apparent.
  In contrast to the field studies, where
there was little effect of pH on Cd uptake,
addition of excess  CaC03 to 10 soils in
the greenhouse caused marked decreases
in Cd  uptake. This result  is consistent
with others  in the  literature, indicating
relatively low Cd availability in calcareous
soils.  The pH  range in the greenhouse
study was higher than those used in the
field study, so the results are not directly
comparable.

Soluble Salts
  Equilibration  of Wisconsin  Rapids
sludge with solutions of varying Ca con-
centrations resulted in solution Cd levels
that were directly related to  Ca concen-
tration. This  result  is in accord with Cd-
Ca reactions on the chelating resin. Zinc,
Ni, and Fe also increased with increasing
Ca concentration, but only about half as
much as Cd. Similar results were obtained
with soils from the field  plots when
equilibrated with varying concentrations
of Ca or Sr.
  Tissue Cd  levels  in the second green-
house corn crop tended to be lower for
high-sludge  treatments that had been
leached than for lower-sludge treatments
that had not been leached. This suggests
that Cd levels in the soil solution may
have been increased at higher salt levels.
Such a response implies that either the
salts in the sludge were present in the Ca
form or that  the other cations induced
greater displacement of Ca from exchange
sites, which then competed with Cd for
more selective adsorption sites. This rela-
tionship is by no means clearcut, but  it
does point out the need to control soluble
salt concentration 'following sludge ap-
plication,  especially in  greenhouse
studies.

Speciation of Dissolved Cd
  Speciation  studies indicated that most
of the Cd and Zn in the soil solution were
in the free-ion form, whereas Cu was
more than 90%  complexed. In a Biotron
study using the resin-controlled nutrient
solution system, the Cd2+, Zn2+, and Cu2+
activities were controlled at levels char-
acteristic of a sludge-amended soil, but
total metal concentrations were changed
two to four orders  of  magnitude  by
complexing with EDTA.  Concentrations
of Cd in the plant shoots were not affected
by the presence of the soluble complex,
and concentrations of Zn and Cu were
increased less than 30% even though the
total Cu concentration in solution in-
creased  10,000 times. Metal  uptake
therefore appears to be a function of the
free-ion concentration at the root surface.
  Speciation  results  with Cd and Cu  in
soil solutions from  control and  sludge-
amended soils by the Donnan membrane
equilibrium  technique were compared
with those calculated by the GEOCHEM
program. The nine-ligand Mixture Model
of GEOCHEM generally tended to predict
higher proportions of free Cd2+ and Zn2*
than those experimentally determined.

Butter Power
  The Cd buffer  power is the change in
total labile Cd per unit change in dissolved
Cd. If  curves relating adsorbed Cd to
solution Cd were linear, the distribution
coefficient,  Cdadsorbed/Cdsolutlon, would
equal the  buffer power. A  survey of
literature values showed that distribution
coefficients  decreased  dramatically  as
Cd/Ca ratios increased.  The reason  is
probably that sites with a very high affinity
for Cd are available at low Cd concentra-
tions; but at  higher  Cd concentrations,
these sites are filled, and Cd has to
compete with Ca for sites of  lower Cd
specificity. Under these  conditions, Cd
buffer  powers determined at relatively
high Cd/Ca ratios drastically underesti-

-------
mate buffer powers at Cd/Ca ratios oc-
curring  in normal soil solution. In fact,
the literature survey indicates that most
distribution coefficients have been deter-
mined at unrealistically  high Cd values
and that buffer powers derived from those
values may be as much as three orders of
magnitude too low for normal soils. The
relationship of  buffer power to uptake
shown  in Equation 1 indicates that a
difference of this magnitude would have
a very great effect on predicted uptake.


Cadmium-Zinc Interactions
  The addition of Zn(N03>2to unamended
Plainfield sand resulted in increased up-
take of Cd by plants grown in the green-
house, most likely as a result of increased
Cd  in solution caused by displacement
from the relatively small number of ad-
sorption sites in this soil. Similar  Zn
additions to Piano soil and to both soils
containing sludge resulted in little addi-
tional Cd uptake, probably because the
greater quantity of sorption sites furnished
by the soil and the sludge could accom-
modate the Zn with little Cd displacement.
These assumptions were supported  by
the  Cd concentrations  determined  in
0.01 M Ca(N03)2 extracts from the above
soil treatments. Zn addition to the Plain-
field soil greatly increased relative solu-
tion Cd  levels, whereas the increases in
solution Cd levels in the Piano soil and
the  sludge-amended  soils were much
smaller.
  Effects of Cd/Zn ratios in solution  on
Cd and Zn uptake by tomatoes grown in
the resin-controlled nutrient system were
studied  in a Biotron experiment. Ratios of
Cd/Zn were varied over four orders of
magnitude, with Cd varying from 10~93 to
1 (T 5M and Zn from 10'7 9 to 10'5 2M.  Cu
was maintained at 10"113M. A linear  re-
lationship (r2=0.94)  was found between
log 
-------
     R. B. Corey, P. A. Helmke. D. R. Keeney. and G. C. Gerloffare with the University
       of Wisconsin, Madison, Wl 53706.
     J. A. Ryan is the EPA Project Officer (see below).
     The complete report, entitled "Factors Affecting the Bioavailability of Cadmium,"
       (Order No. PB  87-147 435/AS; Cost: $30.95, subject to change) will be
       available only from:
             National Technical Information Service
             5285 Port Royal Road
             Springfield, VA22161
             Telephone: 703-487-4650
     The EPA Project Officer can be contacted at:
             Water Engineering Research Laboratory
             U.S. Environmental Protection Agency
             Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
     BULK RATE
POSTAGE & FEES P/
       EPA
  PERMIT No. G-35
Official Business
Penalty for Private Use $300

EPA/600/S2-87/002
           0000329   PS
          CH?rfrnEAR80RN  STREET
          CHICAGO               IL    606()4

-------