Estimating the Oral Bioavailability of Arsenic in Soil

1
Estimating the Oral Bioavailability of Arsenic in
Soil
Mark O. Barnett, Ph.D. Jae K. Yang ,
Ph.D. Jonathan L. Subacz Department of Civil
Engineering Auburn University, AL
Philip M. Jardine, Ph.D. Environmental Sciences
Division Oak Ridge National Laboratory
Scott E. Fendorf, Ph.D. Stanford University
2
Background

• Project Scientist for remediation of Lower East
  Fork Poplar Creek (LEFPC), a Hg-contaminated
  CERCLA site in early 1990s.
• In vitro bioaccessibility and speciation studies
  led to the adoption of a site-specific relative
  Hg bioavailability of 10 (30 originally
  proposed).
• Barnett, M. O. and R. R. Turner (2001).
  Bioaccessibility of mercury in soils. Soil and
  Sediment Contamination 10(3) 301-316.
• Barnett, M. O., L. A. Harris, R. R. Turner, R. J.
  Stevenson, T. J. Henson, R. C. Melton and D. P.
  Hoffman (1997). Formation of mercuric sulfide in
  soil. Environmental Science and Technology
  31(11) 3037-3043.
• Barnett, M. O., L. A. Harris, R. R. Turner, T. J.
  Henson, R. E. Melton and R. J. Stevenson (1995).
  Characterization of mercury species in
  contaminated floodplain soils. Water, Air, and
  Soil Pollution 80(1) 1105-1108.

3
Background

• Funded by the Strategic Environmental Research
  and Development Program (SERDP) since 2000 to
  look at the processes governing the
  bioaccessibility/bioavailability of metals (As)
  in soil, with an emphasis on
• 1) developing a predictive capability and
• 2) correlating macroscopic bioaccessibility/bioava
  ilability with microscopic speciation.
• Views expressed herein are my own and do not
  necessarily reflect the views of past or present
  sponsors, co-investigators, or collaborators.

4
Purpose of Presentation

• Present some of our recent results for arsenic.
• Discuss these results in the context of this
  afternoons panel topic Validation and Quality
  Assurance of Predictive Models of Metals
  Bioavailability.

5
Methods

• 1. Collected and characterized thirty-six
  uncontaminated soils from seven major soil orders
  in the U.S.
• 2. Spiked soils with soluble As(V) (and As(III))
  and measured solid-phase bioaccessibility over
  time using the stream-lined PBET of Drexler, et
  al.
• 3. Developed multivariable linear regression
  models to predict steady-state bioavailability
  from soil properties.
• 4. Validated models using actual contaminated
  soils sites.
• 5. Examined solid speciation with
  synchrotron-generated XAS.

6
Why add soluble As?

• Although the generally low bioavailability of As
  in soils in mining areas has been well
  documented, other forms of these metals may be
  more bioavailable.
• Soluble metals added to soils represent a worst
  case scenario from a bioavailability standpoint.
• Any reduction in bioaccessibility/bioavailability
  of soluble metals added to soils is due to the
  properties of the soil and not some pre-existing
  metal speciation, which is difficult to measure
  and subject to change over time.
• Major soil properties are arguably the most
  stable aspect of soils.

7
Aging reduces As(V) bioaccessibility

• Seventeen of thirty-six soils (47.2) exhibit a
  significant reduction in bioaccessibility over
  six months.
• However, only four of thirty-six soils (11.1)
  exhibit significant reduction in bioaccessibility
  from three to six months.
• pH is the major factor controlling aging, with
  pHlt6 promoting aging.

8
Fe oxides and pH affect steady-state As(V)
bioaccessibility
Soil Fe-oxides and pH control As(V)
bioaccessibility on 36 DoD relevant soils. Model
accounts for 80 of variability in As(V)
bioaccessibility. Model based on common soil
properties is statistically robust at the 99
confidence limit.
9
EXAFS speciation of As(V)
XANES spectra of a As(V)-contaminated soil and
the same soil spiked with additional As(V).The
mechanism of As(V) sequestration is similar for
both soil conditions, that being the formation
of strong inner sphere complexes with Fe-oxides.
As contaminated soil As spike As contaminated
soil
As
10
In vitro validation
Model was able to predict the in vitro
bioaccessibility of five As(V)-spiked soils
(triangles) within a root mean square error of
10.
11
In vivo validation
Model was able to predict the in vivo
bioavailability of As(V) in nine contaminated
soils (squares) within a root mean square error
of lt10.
Soils from in vivo study courtesy of Nick Basta
and Stan Casteel.
12
Comparison to DoD contaminated soils

• The measured As bioaccessibility was very low for
  most soils. The model estimated conservatively
  for all soils

13
Panel Questions

• What are the intended uses of the results of in
  vitro bioavailability models and what confidence
  is needed to support their application?
• How should the validity of a predictive in vitro
  model for bioavailability be assessed?
• What would constitute an adequate validation of
  an assay for site specific bioavailability of a
  metal?
• In assessing the bioavailability of metals at a
  site where there are some in vitro and some in
  vivo split sample comparisons, what information
  will be needed to extrapolate from areas beyond
  those with split samples (if such comparisons can
  be made at all)?

14
At what stage in the process is the model being
applied?

• Bioavailability data can be incorporated into
  risk assessments at the screening level (Tier IB)
  as well as in the baseline risk assessment (Tier
  II). The results of the Tier IB assessment can
  be used to remove sites from further
  consideration or for early identification as to
  whether or not a bioavailability adjustment is
  potentially useful in the baseline risk
  assessment.
• Guide for Incorporating Bioavailability
  Adjustments into Human Health and Ecological Risk
  Assessments at U.S. Navy and Marine Corps
  Facilities. Part 1 Overview of Metals
  Bioavailability, Naval Facilities Engineering
  Service Center, 2000.

15
At what stage in the process is the model being
applied?

• According to Teresa Bernhard, U.S. Navy
  Facilities Engineering Command, we need the
  ability to answer questions such as
• When do we consider doing a bioavailability
  study?
• What are the benefits of considering a
  bioavailability study?
• Do site data indicate potential for lower
  site-specific bioavailability?

16
What are the results going to be used for?

• According to Jim Ryan, U.S. EPA, to use
  bioavailability adjustments at metal contaminated
  sites, we need to
• Measure bioavailability in more than one way
• Understand the reasons for the reduced
  bioavailability
• Understand something about the long-term
  stability.

17
Results for As(V)

• Labile As(V) added to soil (i.e., the worst case
  exposure scenario) is rapidly sequestered in most
  soils, significantly reducing its
  bioaccessibility below 100.
• A model based on soil properties (pH and Fe
  oxides) was able to describe As(V)
  bioaccessibility in a wide range of soils. This
  model was able to predict the bioavailability of
  As in field soils (as measured by swine dosing
  studies) within 10.
• Mechanistic understanding of reduced
  bioaccessibility.
• Since these results are due to generic soil-metal
  interactions rather than metal-specific
  speciation, they are valid as long as the major
  soil properties dont change.

18
Acknowledgements

• We gratefully acknowledge the U. S. Department of
  Defense Strategic Environmental Research and
  Development Program and Dr. Andrea Leeson for
  supporting this work.