United States
                  Environmental Protection
                  Agency	
Robert S. Kerr Environmental
Research Laboratory
Ada OK 74820
                  Research and Development
EPA/600/S6-89/002 Sept. 1989
AEPA         Project Summary
                   Assessment of  the Potential  for
                   Transport of Dioxins and
                   Codisposed  Materials  to
                   Groundwater
                   Richard W. Walters, Zohreh Yousefi, Amy L. Tarleton, Stanley A. Ostazeski,
                   and David C. Barry
                    Parameters relevant to the sorptive
                  transport of polychlorodibenzo-p-
                  dioxins (PCDDs) through soils were
                  evaluated In  laboratory  experiments
                  involving batch snake testing and sat-
                  urated-flow soil column techniques.
                  The experiments  were conducted
                  using  water/methanol mixtures and
                  four uncontaminated sorbents (two
                  surface soils from Times Beach, MO
                  and aquifer materials from Traverse
                  City, Ml and Lula, OK). Five "re-
                  labeled PCOO cogeners, including
                  2,3,7,8-tetra (T4COO), 1,2,3,4,7-penta,
                  1,2,3,4,7,8-hexa, 1,2,3,4,6,7,8-hepta,
                  and octa (O8CDD), and  three  codls-
                  posed materials pentachlorophenol
                  (POP), and chloroberusene (CB) were
                  used.  The  partition coefficient (KD)
                  for sorptien of T«CDD from water was
                  found  to be in good agreement with
                  the water-phase K0 predicted by log-
                  linear  extrapolation according  to the
                  cosolvent theory by using  KD data
                  generated) for water/methanol mix-
                  tures.  This observation validates the
                  use of'log-linear extrapolation to esti-
                  mate  water-phase KD  values for
                  PCDDs using cosolvent data. KD
                  values for sorption of PCDDs by sur-
                  face soils at volume fraction solvent
                  (fs) of 0.5-0.9 were reduced  by a
                  factor  of up to 2.5 when PCP or CB
                  were present Reductions in KD for
                  PCDDs in the presence of PCP and
                  CB  increased with decreasing fs to
                  enable a better understanding of the
                  influence of codisposed materials
                  such as PCP and CB on the mobility
of PCDDs under environmental con-
ditions typified  by low f, (i.e., f,
< < 0.5). Sorption KD values for the
aquifer materials were normalized on
organic carbon content (f^) to yield
values  of KQC which were in general
agreement with Koc values  deter-
mined for the surface soils. This ob-
servation suggests that sorption by
the organic matter content or the
aquifer materials was sufficient to
mask sorption to mineral surfaces.
Desorption of  PCOOs from the  sur-
face soils appeared to be reversible
but was  limited  by kinetics, with
roughly 50-90% of reversible-desorp-
tion equilibrium being attained within
a contact period of 30-50 days.
  This  Project Summary was  de-
veloped lay EPA's  Robert S. Kerr En-
vironmental Research Laboratory, Ada,
OK, to  announce key findings of the
research prefect that is fully docu-
mented in a separate report  of the
same title (see Protect Report ord-
ering information at back).
Introduction
  Polychlorodibenzo-p-dioxins as en-
vironmental  contaminants have perhaps
received most widespread  national at-
tention because they were  present  in
waste oils that were used for dust control
in Missouri.  PCDD-contamination of soils
and sediments  in  various  locations
throughout the United States and in other
countries  have  been  documented.

-------
Concentrations  of  PCDDs  in  soil and
sediment samples as high as 1600 ppb
to 100 ppm have been observed.
  Until recently it was  believed  that
PCDDs were relatively immobile in soils
because  of their low water  solubilities,
and that  there  was little potential for
PCDDs to be leached from contaminated
soils. Field observations have shown that
PCDDs move faster through soils  than
would be expected based on their water
solubilities. Numerous theories have been
proposed to  explain  the  "facilitated
transport" of  PCDDs and other highly
hydrophobic organic  contaminants, in-
cluding a cosolvent which accounts for
enhanced  mobility in  terms  of the
reduction in soil sorption in the presence
of cosolvents.
  The research  presented in the full re-
port involved  three basic objectives;  1)
determine  the  relative  effect  of the
presence of codisposed materials on the
sorptive  transport  of  PCDDs   through
surface soils, 2) evaluate the equilibrium
sorption of PCDDs by aquifer materials,
and 3) evaluate the kinetics of desorption
of PCDDs from surface soils.
Procedure

Solutes
  All PCDDs used in the experiments
were radiolabeled with C-14. Individual
stock  solutions  of  the radiolabeled
PCDDs were  prepared by  dissolving
them in methanol.
  Other radiolabeled compounds used in
the experiments included pentachloro-
phenol (PCP), methanol and water. All
radiolabeled compounds were used as
received from the supplier.
So//s and Aquifer Materials
  The soils obtained  from Times Beach,
MO were the same as those used in pre-
vious investigations of the  sorption  of
PCDDs to  soils.  These soils were air
dried, sieved through 0.3  mm standard
sieves, and characterized for pH, cation
exchange  capacity, organic matter  con-
tent, organic carbon, organic nitrogen and
texture. The soils are referred to here as
soil 91 (low-organic carbon soil) and soil
96 (high-organic carbon soil).
  Lula, OK and Traverse City, Ml aquifer
materials  were  provided  by  the  EPA
Project Officer and were used as
received.
Solvents
  Methanol  and  methylene chloride sol-
vents used in the experiments described
were pesticide grade  and were used as
received  from the supplier.  Water was
treated by reverse osmosis,  activated
carbon bed and  a pair  of  mixed bed
deionizers. Further, the water was dosed
with  0.01% NaN3, as a biocide  and  its
ionic strength was adjusted to 0.01 M
using CaCI2.  The pH of  the  water was
adjusted to 7.0 using NaOH.
Scintillation Cocktail
  An insta-gel liquid  scintillation  cocktail
obtained  from United Technologies was
used for all radioisotope analyses.
Analyses

Liquid Scintillation Counting
  Analytical determinations were made
by  liquid scintillation counting (LSC)
using a Model 1219 Rackbeta counter.
  Liquid-phase  solute concentrations  (C
or Ce) were  determined by  sampling
three aliquots of 1-3 ml of liquid phase
and adding the sample to 10 ml of Insta-
gel cocktail in 20-ml glass counting vials.
  Soil-phase  solute concentrations (S of
Se) were  determined primarily by the dif-
ference of the total solute minus liquid-
phase  solute. However, to confirm mass
balances  direct determination of  soil-
phase solute concentrations was made to
determine mass balances.
Data Manipulation
  Isotherm data were evaluated by linear
regression to determine  best-fit  param-
eters for sorption  constants.  Best-fit
estimates of  KD,  the coefficient of  de-
termination (r2), and the standard  error of
the  estimate  (s) were calculated using
standard  equations.
  Values of Kobs  were determined  from
the  effluent concentration  profiles  ob-
tained  from saturated-flow  soil  column
experiments.  Kobs is the value of  KD
calculated from  the proper  equation
under the assumption of local equilibrium
and  a  value  of Vr equal  to  the  ratio of
sorbing solute to nonsorbing solute re-
tention time as determined by the center-
of-mass of the respective effluent  peaks.
  Least-squares regression analysis was
used to determine the best-fit estimates
of the slope  (m) and intercept  (b)  for
proposed  cosolvent  equations.  The
values of m and b and the regression sta-
tistics  were determined  by using
standard techniques.
  Desorption  data were  regressed ac-
cording  to empirical  exponential and
power-curve models.
Solubility Determinations
  Batch techniques already described to
evaluate  T4CDD  solubility  in water/
methanol mixtures were used to evaluate
the solubilities of the other PCDDs.  E>
periments were conducted in 1-ml glas
microvials.  PCDD  crystals were pre
coated  onto the inner  wail  near  th
bottom  of vial  by delivering a  workin
solution containing  PCDD  and the
evaporating  the solvent with nitrogei
One ml of liquid phase was then added t
the precoated microvial, the  vials wet
capped  and covered with aluminum  fo
and then placed on  a shaking table. Th
vials were shaken daily for 15 minutes ft
contact periods up to 28  days.


Batch  Shake  Testing
  The  batch shake testing  techniqu
which was  used to  generate  data pn
sented in this report was based on pul
lished procedures to evaluate sorptic
and desorption of PCDDs to/from soils.

Experiments Involving
Water/Methanol  Mixtures
  Batch  experiments involving  wate
methanol mixtures were  conducted in 1
ml conical glass centrifuge tubes.
  Tubes were dosed by  direct addition
soil or aquifer material (0.2-1.0 gm), liqu
phase and solute (delivered  by dire
dosing of stock solutions). A liquid-pha:
volume of 12-ml was added to the tub<
except for experiments involving  PCP <
solute, in which case the tubes were filk
completely with liquid phase.
  Experiments conducted to evaluate tl
sorption of  PCDDs  in the  presence
PCP and  CB  were  performed at co
centrations from  1-10% of solubility.
  After  dosing, the  tubes were  place
horizontally on a shaking table and we
shaken  for 15  minutes every  hoi
Contact periods ranged from 2 hr to 30
and varied  according to the type of  e
periment (kinetic/equilibrium) and  tl
soil-solute-liquid phase system und
study. Following contact, tubes were r
moved  from  the  shaking  table ai
centrifuged at 2,000 x g for 10 minute
to achieve  adequate separation of liqi
and solid phases.

Water-Phase Experiments
  Two batch procedures were utilized
generate isotherm data for the sorption
T4CDD  from water by soil 91. Both s<
of experiments were conducted  using
50-ml  round bottom glass  centrifu
tubes fitted with  Teflon-lined screw caj
Soil and water doses to the tubes we
approximately 50 mg and 40 ml.
PCP Screening  Experiments
  Preliminary  experiments  were cc
ducted to validate mass  balances and I

-------
centrifugation procedure used to evaluate
sorption of PCP.
  Using general batch  shake  test pro-
cedures mass balance experiments were
conducted with and without soil to assess
recovery of PCP.


Kinetic Evaluations
  Time  series experiments  involving
sorption of PCP from water by soils 91
and  96 and  sorption  of  P5CDD  and
OgCDD by the two aquifer materials were
conducted to determine the contact time
necessary to  attain sorption equilibrium
and  to  qualitatively  evaluate  sorption
kinetics.
  Experiments  involving PCP were con-
ducted by using 0.3 g of soil and a PCP
dose such that initial concentration in the
liquid phase  was 0.035  ng/ml. Liquid
volume in these experiments was 15-17
ml as  necessary  to  completely fill the
centrifuge tube in order  to  minimize
volatile losses  of  PCP into the  head-
space. Tubes were contacted for various
times and prepared in triplicate for each
contact periods.
  Experiments  involving aquifer material
utilized contact periods of 2-72 hr. These
experiments  were  performed at a fs  of
0.65  for  08CDD.  P5CDD and 08CDD
were  chosen  on  the basis of known
kinetic data  for  surface soils  which
indicated that  these  PCDDs bound the
kinetic behavior of  all PCDDs studied.
Saturated-Flow Soil Column
Testing

  Saturated-flow soil column testing was
performed to validate  the use  of  the
column techniques for estimating  the KD
of the PCDDs. These column techniques
were then applied to study the sorption of
PCDDs in the presence of PCP and CB.
  These soil  column experiments used
either a glass column (2.5 cm x 25 cm)
fitted with  an adjustable plunger, to  ac-
commodate variable soil masses in  the
column, or a stainless steel column (0.48
cm x 10 cm).

Column Packing
  The glass column  was slurry packed
by adding a known mass of soil (15-25 g)
to roughly  20 ml of liquid phase and
slowly pouring portions  of the  slurry into
the  top  of the column. Solvent  was al-
lowed to flow from the column by gravity,
and  small amounts of slurry  or fresh
solvent were  continuously  added to  the
top  of  the  column to maintain  a liquid
level above the soil at all times.
  The steel column was dry packed with
a known mass of soil (2-3g) with  the aid
of a vacuum pump.

Pore Volume Determination
  Carbon  14 labeled methanol  and 3H
labeled water were used as inert  tracers
for determination of the hydraulic deten-
tion volume (HDV) of the soil column.
Solute Retention Experiments
  Pulsed input of the radiolabeled solute
similar to that used to evaluate column
HDV was used in column retardation ex-
periments. A plot of effluent sample con-
centration versus  cumulative effluent
volume was prepared, and the solute ret-
ention volume was determined from the
center-of-mass of the  effluent  solute
peak.
Results and Discussion

Solubility Determinations
  The  results of experiments to deter-
mine the solubility  of P5CDD,  H6CDD,
H7CDD and OaCDD in methanol are sum-
marized m Table 1.
  Values of  os determined  from  the
solubility data for each of the PCDDs and
for PCP are also listed in Table 1. These
values  represent the slope of the  log-
linear relationship between the log of the
mole fraction solubility and the volume
fraction cosolvent.

Sorption to Surface Soils
  The  time series  experiments  suggest
that equilibrium for  sorption of PCP to
soil 91  was achieved in less than one day
of contact, while a 30-day contact period
was necessary for equilibrium when soil
96 was used.
              Table 1.    Summary of Results for Solubility Determinations for PCDDs and PCP in WatertMethanol Mixtures

                                             Measured Solubility, mg/L            as
f*
T4CDD
Mean
St. Dev.
n
P5CDD
Mean
St. Dev.
n
H6CDD
Mean
St. Dev.
n
H7CDD
Mean
St. Dev.
n
OgCDD
Mean
St. Dev.
n
PCP
Mean
St. Dev.
n
0.50 1.0

10.6
0.5
33

0.10 52
0.2 2
18 6

20
2
15

24
2
12

4.0
0.6
56

3. 1x103 180xl03
0.6
9


6.25



6.09



7.01



7.35



7.35



3.8



-------
              Table 2.
Summary of Isotherm Parameters Determined for Sorption of PCP by Soils 91 and 96

                             Volume Fraction Methanol, fa
                                        0.0
                           0.25«
0.25«
                                                                                0.5
                                                                    0.75
Soil 91
KD, mUg
r*
n
pH*>
Km,oc, mollg
log (Km.0c)
Soil 96
KD, mUg
'2
n
pH«>
«m.oc, mol/g
log (Km,oc)
Molar Volume, mUmol

18
0.999
12
7.0
150
2.18

180
0.96
14
6.6
130
2.11
18.0

4.5
0.93
13
6.5
33
1.52

96
0.98
15
6.3
61
1.78
20.5

5.0
0.93
14
6.9
37
1.57







20.5

1.1
0.69
7
6.5
6.9
0.84

12
0.98
15
6.2
6.5
0.87
24.7








2.0
0.60
8
6.0
0.87
-0.06
29.8
              aReplicate isotherms for soil 91 at fs of 0.25 were determined for contact periods of 2 days and 4 days,
              respectively.
              bpH is the value of the pH in solution at the end of the contract period.

              Table 3.     Summary of Kobs Values Determined for Sorption of P5CDD, H6CDD, H7CDD and O8CDD by Soil
              91 from Water/Methanol Mixtures

                                                    Volume Fraction Methanol, fs
                                               0.75a
                                  0.9"
Molar Volume, mUmol
P5CDD
Kobs, mUg
Km.oc. mollg
log (Km,oc)
H6CDD
Kobs, mUg
Km oc, mollg
log (Km,oc)
Kobs, mUg
Km,oc, mollg
log (Xmoc)
OgCDD
Kobs, mUg
Km oc, mollg
log (KmtOC)
29.8

1.7
8.6
0.93

3.6
18
1.3
4.8
24
1.38




37.9

0.20
0.79
-0.10

0.23
0.91
-0.04
0.40
1.6
0.20

0.82
3.2
0.51
40.4

0.072,0.057
0.27, 0.27
-0.57, -0.68

0.70, 0.092
0.38, 0.35
-0.42, -0.46
0.72, 0.77
0.45, 0.47
-0.35, -0.39

0.27, 0.79
0.79, 0.77
-0.70,-G 75
              Experiments were conducted at a column flow rate of 8 mL/hr (pore velocity of 73.5 m/day), with ps of
              1.4 g/mL and e of 0.79.
              ^Experiments were conducted at a column flow rate of 8 mUhr (pore velocity of 0.77 m/day), with pB of
              1.8 g/mL and £ of 0.51.
              ^Experiments were conducted at column flow rates of 12 mLlhr and 20 mLJhr (pore velocities of 1.78 and 2.96
              m/day), respectively, with pB of 1.75 g/mL and e of 0.33.
  Equilibrium isotherm data for the sorp-
tion of PCP by soils  were generated by
batch techniques at fs of 0.0, 0.25 and 0.5
and for soil 96 at fs of 0.75. The results of
regression  analysis of the isotherm data
obtained for all fs studied are summarized
in Table 2.
  The Kobs values for sorption of PCP by
soil 91 were also estimated from column
experiments.
                  The log  (Kmoc) values for  sorption of
                PCP by soils 91 and 96 plotted against fs
                agree with the  results  obtained from
                batch  and column techniques over  the
                entire range in fs studied.

                Sorption  of T4CDD from  Water

                  The value of log (K^.) determined from
                the experiment involving unwashed  soil
                  and a contact  period  of 48  hours w£
                  6.44,  while  the  value of  log  (K0
                  determined from the experiment involvin
                  prewashed  soils and a contact period
                  ten days was 6.66.
                     The value of  log (Koc) of 6.67 dete
                  mined by  the  second  procedure  co
                  responds to a value of log (Km oc) of 5.4
                  The experimental value of the  wate
                  phase partition coefficient  is  in  goc

-------
agreement with the value predicted by
log-linear  extrapolation of the  water/
methanol data,  namely 5.30.  Based on
this  observation, it  appears that  for
T4CDD the cosolvent theory applies over
the entire  range of fs, 0.0-1.0 for water/
methanol mixtures. This provides strong
support  for the  validity of this theory to
PCDDs and other highly hydrophobic or-
ganic contaminants, and  also provides
support  for the use of  log-linear extrapo-
lation to estimate water-phase  K^. values
for PCDDs by using data generated from
water/methanol mixtures.
Sorption of PCDDs from  Mixtures
of Water and Methanol
  Linear sorption  isotherms  were ob-
served for  all PCDDs  studied. The  KD
values for the four PCDDs are generally
comparable  (ranging  from 1.8-3.8 ml/g)
and  do  not show the  expected trend of
increasing with increasing  hydrophobicity
of the PCDD.
  The results of column studies used in
sorption experiments  are  presented  in
Table 3.
  The values of KD and Kobs determined
for  the  PCDDs  by batch and  column
techniques, respectively were normalized
on  foc  converted to  molar  units  by
dividing by  liquid-phase molar volume.
The results of the batch experiments  de-
termined in  the present study  are  in
general  agreement with the log-linear re-
lationship  based on the  data of other
investigators. The  results obtained  by
column  techniques were consistently
lower than expected from  the log-linear
relationship of the batch data only.
  The relatively  low  values of  Kobs  in
comparison to KD  were believed to be
attributed  to non-attainment of local
equilibrium within the soil  column for the
flow rates utilized.
Sorption of PCDDs in the
Presence  of PGP and CB
  The effect of  the presence of PCP on
the sorption  of PCDDs was evaluated by
batch and column techniques.
  Batch experiments were conducted to
generate  sorption  isotherm  data  for
08CDD  with both soils  at PCP con-
centrations of 25 and  200 ppm. Contact
periods of 2 and 36 days were  utilized for
soils 91 and 96,  respectively. For both
soils and both  PCP concentrations,  the
sorption of 08CDD in  the  presence  of
PCP was roughly 30% of the sorption of
08CDD in the absence of PCP.
  Column  experiments were  conducted
for P5CDD, H6CDD and H7CDD using soil
)1, with  three liquid phases; 1) fs of 0.75,
2) fs of 0.75  and a PCP concentration of
1000 ppm, and 3) fs of 0.75  and a CB
concentration of 10 ml/L. For each PCDD
studied,  the  lowest Kobs values were
observed for the solvent/CB system. The
K(,bs for both solvent/CB and solvent/PCP
systems were consistently below Kobs for
the solvent  only.  These normalized
values  show  that  PCDD  sorption
generally decreased by  25%  in the
presence of  PCP  and  50%   in the
presence of CB.
  Column experiments were  also  con-
ducted for all  PCDDs at fs of 0.9 with and
without PCP present. The  results for  all
PCDDs indicate that there  was generally
no  significant change  in  Kobs  in the
presence of PCP relative to values in the
absence of PCP at a fs value of 0.9.
  The reduced sorption of  PCDDs in the
presence of cosolutes may be explained
by  a  number of  factors,  including  the
effect of the cosolute on the solubility of
PCDDs and/or the effect of the cosolute
on PCDD-soil organic matter interactions.
However, it is suspected that the reduced
sorption of PCDDs in the presence of the
cosolutes  may be explained in terms of
either the effect of soil-phase cosolute  on
the soil-phase activity coefficients of the
PCDDs  or  the  results  due  to the
competition  between  the sorbed
cosolutes  and PCDDs. This speculation
would support the observation that the
effects become  more  pronounced  at
lower fs  because  greater soil-phase
concentrations of the cosolutes would  be
expected via sorption as fs decreases.


Sorption by Aquifer Materials
  Linear isotherms were observed for  all
PCDDs for fs values of 0.25 and 0.65 with
the exception of T4CDD and O8CDD.
  Comparison of  the  experimental  data
points for the  aquifer materials with the
data for the surface soils,  and their cor-
responding  regression  equations, in-
dicates that values of log (Km oc) for both
sorbents  are well  described by a single
curve for  T4CDD,  H6CDD and H7CDD.
For P5CDD,  the experimental values  of
lOQ (Km,oc) f°r  the aquifer materials  at
each  fs'  are  significantly  greater  than
those  determined  from surface soil. For
O8CDD, the  experimental  values of log
(Km.oc) foe the aquifer materials are
slightly below those predicted  for the
surface soils.
  The general agreement between values
of log (Km oc) for the aquifer materials and
the surface soils for PCDDs suggests that
foc alone is  the  dominant   factor
determining the extent  of  sorption for
highly hydrophobic organic contaminants
such as PCDDs, presumably because the
organic sorption  is sufficiently  strong to
dominate the weaker sorption to minerals.


Desorption From Surface Soils
  From limited data, it appears that the
sorption of PCDDs from the surface soils
into  water/methanol mixtures  may  be
reversible, but that desorption  is char-
acterized  by very low rates.
  Preliminary experiments conducted to
assess batch and column techniques for
the study of sorptive transport of PCDDs
through soils  in  the  presence  of diesel
fuel, were found to be infeasible, owing to
the formation of multiple  phases. These
experiments suggested  that  relatively
stable water/methanol/diesel fuel emul-
sions  were readily  transported through
the soil column. Further, although the soil
column had been  apparently flushed of
diesel fuel at a high relative velocity, it
was  observed  that the column had a
residual  capacity  to retain significant
amounts  of  diesel  fuel  which  was
released  from  the column  as  a stable
emulsion  at low liquid-phase flow rates.


Conclusions
  The sorption and desorption of PCDDs
by surface  soils  and the  sorption of
PCDDs by aquifer materials has been in-
vestigated by using batch shake testing
and  saturated-flow soil  column tech-
niques.  Experiments were conducted
using  water and water/methanol mixtures
and  two  model codisposed  materials
(pentachlorophenol (PCP)  and chloroben-
zene (CB)).
  Sorption isotherm data were generated
by batch  techniques to evaluate the sorp-
tion of T4CDD from water by soil 91. A
contact period  of 10 days,  and  a water
prewash procedure which was used to re-
move nonseparable suspended  particles
(NSP) from the water phase, appeared to
be adequate for evaluating an equilibrium
KD value which was not  biased by the
inadvertent sampling of NSP during liquid
phase analysis. A linear sorption isotherm
was obtained which indicated that KD was
30,600 (log (KD) = 4.49). A value of Km oc
of 2.58x105 (log (Kmoc) = 5.41) was de-
termined  by dividing this KD value by f^
and by the molar volume of water. This
corresponds to a  value of log  (Km oc) of
5.30 predicted by log-linear extrapolation
of KD data  generated  by other  in-
vestigators for  T4CDD and  soils 91  and
96 and water/ methanol  mixtures.  This
observation provides support for the ap-
plicability of the  cosolvent theory to the
sorption   of  PCDDs  to  soils from
water/methanol mixtures over the entire

-------
range in fs  from 0.0-1.0,  and  provides
justification  for  log-linear extrapolation
using data generated for water/methanol
mixtures to  predict  equilibrium  water-
phase  KD values for PCDDs and other
highly hydrophobic organic contaminants.
  The effect of the presence of PCP on
the sorptive  transport of PCDDs through
soils was studied  by using  batch and
column  techniques,  in which the  liquid
phase consisted of water/methanol mix-
tures ranging in fs from 0.5-0.9 and con-
taining  PCP at  concentrations ranging
from 1-10% of solubility. These experi-
ments  indicated that the sorptive reten-
tion  of PCDDs by soils was reduced by
about  a factor  of  3 when  PCP was
present in the  liquid  phase to  that ob-
served in the absence of PCP.
  The effect of the presence of CB  on
the sorptive  transport of PCDDs through
soils was studied by using column tech-
niques,  in which the liquid  phase con-
sisted of water/methanol mixtures ranging
in fs from of 0.75-0.9 and a  CB concen-
tration of 10  mL/L. These experiments in-
dicated  that the sorptive  retention  of
PCDDs  by soils was  reduced by a factor
of 2  in the presence of CB at an fs of 0.75
relative  to that observed in the  absence
of CB.  Experiments to evaluate the
sorption of  PCDDs  by aquifer materials
included  an assessment  of  sorption
kinetics and  equilibrium KDs. The  results
of these experiments suggested  that a
two-day contact period was  sufficient  to
achieve sorption  equilibrium  for  all
PCDDs  studied for these low carbon soil
materials. Sorption  isotherm  data were
generally  linear, and equilibrium KDs,
when converted to Km  (partition  coef-
ficient),  were log-linearly related to fs  in
accordance  with the  cosolvent theory.
When Km values were normalized on f^
of the aquifer material, the resulting KmiOC
values were in general  agreement with
Km.oc values determined for surface soils
from the  Times Beach, MO area. It  is
suspected that the agreement  between
aquifer material and surface  soil sorption
for PCDDs indicates that sorption to  or-
ganic  carbon  is  sufficient to  mask
sorption to mineral surfaces.
   Experiments to evaluate the desorption
of PCDDs from surface soil were con-
ducted  by  batch techniques, in  which
PCDDs were previously sorbed by soil
for thirty days prior to the initiation of the
desorption experiments. The results  of
these experiments  suggested  that de-
sorption  appeared to be  generally  re-
versible but  was limited  by kinetics, with
roughly 50-90% of reversible-desorption
equilibrium being attained within a 30-day
contact period.
Recommendations

  Additional research  in six areas of the
transport  of PCDDs  andrelated con-
taminants through soils is recommended:
  1. Sorption of PCDFs. No  data have
been reported on the sorptive partitioning
of  polychlorodibenzofurans  (PCDFs).
High-purity standards of these contami-
nants are now  available in radiolabeled
form,  and  it is  suggested that  issues
relevant to  sorptive partitioning be inves-
tigated using several PCDF congeners.
  2. Sorption from Water/Methanol Mix-
tures. The sorption of PCDDs as a func-
tion of volume fraction cosolvent  (fs) for
water-miscible  cosolvents  is  generally
well understood. However, further  re-
search is necessary to better understand
solubility and sorption  behavior at low fs
(e.g., fs below 0.05-0.1). This work should
be  conducted using PCDDs and PCDFs
as  well as slightly   less hydrophobic
contaminants,  such  as polycyclic
aromatic hydrocarbons (PAHs) and poly-
chlorinated biphenyls  (PCBs).  Research
should also be  initiated to evaluate  the
cosolvent  effects attributed to  ethylene
glycol, a water-miscible solvent which is
relevant to  the manufacturing  process of
PCP and  thus  relevant to waste sites
contaminated by PCDDs.
  3. Sorption in  the Presence  of  Codis-
posed Materials. Additional  research is
needed to better understand the sorption
of  PCDDs  and  related contaminants in
the presence of codisposed  materials.
Three groups of codisposed materials are
important. The  first  group consists of
slightly soluble  solvents.  Much is known
regarding  the sorption of  hydrophobic
contaminants in the presence of miscible
cosolvents  such  as methanol  and ace-
tone,  but  little  is known regarding  the
effect  of  immiscible  solvent such  as
methylene chloride, benzene, and aniline.
The second  group consists  of  slightly
soluble solids. For example, the effect of
the presence of 2,4,5-trichlorophenol  and
terpenes, as well as PCP, at low fs should
be  evaluated. The third group consists of
oil/carrier  liquids. There  is a  significant
requirement to  further understand  the
movement  of oil through  soils, and  the
manner in which oil movement affects the
sorptive  transport  of  PCDDs  and
codisposed materials. Thus, research
should  be conducted to evaluate  the
sorptive transport of oils, oils  and PCP,
and oils, PCP and terpenes,  as well as
the sorptive transport of PCDDs in  the
presence  of these materials.  This  re-
search  should focus  on evaluating  the
relative contributions to transport attribu-
ted to dissolved oil constituents versus oil
droplets and oil/water  emulsions,  ir
eluding water-phase sorption-desorptioi
dislodging  of entrapped droplets,  an
potential for emulsion formation.
  4. Sorption of PCDDs in the present
of colloids.  Much is  known regarding tti
sorptive transport  of moderately hydrc
phobic organic contaminants in the  pre:
ence of colloidal organic matter, and th
work suggests that colloids  can  substar
tially  increase  the mobility of organi
contaminants  through soils.  Addition
work is necessary to extend this researc
to include  highly  hydrophobic  organic
such as PCDDs.
  5. Sorption of PCDDs by aquifer mi
terials. Additional research is needed  I
evaluate the relative mineral contributior
to  sorptive transport  of PCDDs  an
similar organics through  low  f^ aquif<
materials.  This research should  focus c
organic contaminants that  are very hydrc
phobic, such as  PAHs and PCBs, as we
as PCDDs. The research should include
variety of  sorbents  to  enable  a  bett<
fundamental understanding of the relativ
effects of  KOW, mineral surface area, ar
foe on sorption to low-foe sorbents.
  6. Desorption of PCDDs. Desorption i
PCDDs from soils at low fs in the  pre:
ence  of codisposed materials,  such <
oil, should be investigated. This researc
should include development of  a mod
which incorporates desorption kinetics
enable better prediction of the moveme
of PCDDs through soils under real-wor
conditions.

-------
  R. W. Walters, Zohreh Yousefi, Amy L. Tarleton,  Stanley A. Ostazeski, and David
        C. Barry are with University of Maryland, College Park, MD 20742
  Carl G. En field is the EPA Project Officer (see below).
  The complete report, entitled  "Assessment  of the Potential for Transport  of
        Dioxins and  Codisposed Materials to Groundwater," (Order No. PB 89-
        166 6071 AS; Cost: $21.95, subject to change) will be available only from:
           National Technical Information Service
           5285 Port Royal Road
           Springfield, VA 22161
           Telephone: 703-487-4650
  The EPA Project Officer can be contacted at:
           Robert S. Kerr Environmental Research Laboratory
           U.S. Environmental Protection Agency
           Ada, OK 74820
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300

EPA/600/S6-89/002
          0000*5833
                                        60604

-------