vvEPA
United States
Environmental Protection
Agency
600R05161
Field Study of the Fate of
Arsenic, Lead, and Zinc at the
Grou nd-Water/Su rf ace-Water
Interface
Reducing
Iron
Oxidation
Iron
Reduction
Sulfate
Reduction
11860 11870 11880 11890 11900
Energy, eV
Redox Controls on Contaminant Speciation and Mobility
Surface Water
Discharge
(Dissolved &
Suspended Solids)
©
Internal Recycling Process
Chemocline
(aq) 2(.q)
©
As-HFO(s) + HzS(aq) (or other reductants)
As-FeS,,,
(or other Refill-bearing minerals
and organic matter)
I ,L. *
SoHds Dissolution
indJor R* tuspunslon
-------
EPA/600/R-05/161
December 2005
Field Study of the Fate of Arsenic, Lead,
and Zinc at the Ground-Water/
Surf ace-Water Interface
Robert G. Ford, Richard T. Wilkin,
Cynthia J. Paul, Frank Beck, Jr., and Tony Lee
U.S. Environmental Protection Agency
Office of Research and Development
National Risk Management Research Laboratory
Ground Water and Ecosystems Restoration Division
Ada, Oklahoma 74820
Kirk G. Scheckel
U. S. Environmental Protection Agency
Office of Research and Development
National Risk Management Research Laboratory
Land Remediation and Pollution Control Division
Cincinnati, Ohio 45268
Patrick Clark
U. S. Environmental Protection Agency
Office of Research and Development
National Risk Management Research Laboratory
Technology Transfer and Support Division
Cincinnati, Ohio 45268
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
Notice
The U. S. Environmental Protection Agency through its Office of Research and
Development funded the research described herein. This report has been subjected
to the Agency's peer and administrative review and has been approved for publication
as an EPA document. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use. All research projects making
conclusions or recommendations based on environmental data and funded by the
U.S. Environmental Protection Agency are required to participate in the Agency
Quality Assurance Program. This project was conducted under an approved Quality
Assurance Project Plan. The procedures specified in this plan were used without
exception. Information on the plan and documentation of the quality assurance
activities and results are available from the Principal Investigator.
-------
Foreword
The U.S. Environmental Protection Agency is charged by Congress with protecting the Nation's land, air,
and water resources. Under a mandate of national environmental laws, the Agency strives to formulate
and implement actions leading to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research program is providing data
and technical support for solving environmental problems today and building a science knowledge base
necessary to manage our ecological resources wisely, understand how pollutants affect our health, and
prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for investigation of technologi-
cal and management approaches for preventing and reducing risks from pollution that threatens human
health and the environment. The focus of the Laboratory's research program is on methods and their
cost-effectiveness for prevention and control of pollution to air, land, water, and subsurface resources;
protection of water quality in public water systems; remediation of contaminated sites, sediments and
ground water; prevention and control of indoor air pollution; and restoration of ecosystems. NRMRL
collaborates with both public and private sector partners to foster technologies that reduce the cost of
compliance and to anticipate emerging problems. NRMRL's research provides solutions to environmental
problems by: developing and promoting technologies that protect and improve the environment; advanc-
ing scientific and engineering information to support regulatory and policy decisions; and providing the
technical support and information transfer to ensure implementation of environmental regulations and
strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan. It is
published and made available by EPA's Office of Research and Development to assist the user com-
munity and to link researchers with their clients. Characterization of contaminant transport across the
ground-water/surface-water transition zone is an important component of risk characterization for sites
that are a source of contaminants within an urban watershed. Defining the chemical and biological fac-
tors that control the fate of ground-water contaminants that are discharged into surface water is critical to
the design of effective risk management strategies. This report summarizes findings from an extensive
field investigation conducted to determine the fate of inorganic contaminants within the headwaters of an
urban watershed and to support the design and evaluation of remedial strategies to mitigate exposure to
contaminants in surface water and sediments.
ien G. Schmelling, Director
Ground Water and EcosysterrtS Restoration Division
National Risk Management Research Laboratory
-------
Contents
Foreword iii
Figures vii
Tables ix
Acknowledgments xi
Executive Summary xiii
Chapter 1
Purpose 1
Study Scope 3
Chapter 2
Site Background 5
Study Approach 5
Site Characterization Methods 9
Field Measurements 9
Ground Water 9
Surface Water 9
Laboratory Measurements 10
Water Chemistry 10
Sediments 10
Sediment Chemistry and Mineralogy 10
Mineralogical Characterization 11
X-ray Absorption Spectroscopy 11
Arsenic XAS 11
Pb and Zn XAS 11
Sediment Extraction Procedures 12
Surficial Sediments 12
Buried Sediments 12
Scanning Electron Microscopy-Energy Dispersive Spectroscopy 12
Chapter 3
Site Hydrology 15
Media-Specific Contaminant Distributions 24
Ground Water 24
Surface Water 24
Sediments 27
Chemical Speciation of Metals in Sediments 32
Sediment Extractions 32
Surficial Sediments 32
Buried Sediments 33
Element Speciation by X-ray Absorption Spectroscopy 38
Scanning Electron Microscopy 43
Chapter 4
Stability of Contaminants in HBHA Pond Sediments 47
Monitoring Long-Term Behavior (or Performance) of HBHA Pond 49
Relevance to Other Sites 49
References 51
Appendix A 55
Appendix B 57
Appendix C 63
-------
Figures
Figure 1 Conceptual model of geochemical zones that play a dominant role in governing solid-liquid
partitioning of arsenic (lead and zinc) across the GW/SW transition zone at the Industri-Plex
Superfund Site 2
Figure 2 Prominent land features within the Industri-Plex Superfund Site and adjacent to the HBHA
Pond 6
Figure 3 (A) Arsenic (As), benzene (Bz), and toluene (Tl) concentrations (ppb) from temporary
ground-water sampling locations as a function of depth below ground surface (feet) for the
study site 7
Figure 4 (A) Close-up view of permanent ground-water monitoring network; arrows show approximate
centerline of arsenic and BTEX plumes discharging into HBHA Pond 8
Figure 5 (A) Patterns in surface water and ground-water flow budgets within the HBHA Pond as
documented by Aurilio et al. (1994) and Wick et al. (2000) 16
Figure 6 Aerial distribution and time-dependent variability for total dissolved arsenic concentrations
(ppb) detected in tubing wells established adjacent to the HBHA Pond 17
Figure 7 Aerial distribution and time-dependent variability for total dissolved zinc concentrations (ppb)
detected in tubing wells established adjacent to the HBHA Pond 18
Figure 8 Aerial distribution and time-dependent variability for benzene concentrations (ppb) detected
in tubing wells established adjacent to the HBHA Pond 19
Figure 9 Vertical patterns in water quality within the northern part of the HBHA Pond on sampling
dates in April 2000, August 2000, and April 2001 20
Figure 10 Vertical patterns in water quality within the central part of the HBHA Pond on sampling
dates in April 2000, August 2000, and April 2001 21
Figure 11 Vertical patterns in water quality within the southern part of the HBHA Pond on sampling
dates in April 2000, August 2000, and April 2001 22
Figure 12 Evidence of water column mixing within the HBHA Pond as a result of a large surface
water flow event 23
Figure 13 The concentration distribution of (A) alkalinity, (B) sulfate, (C) ammonia-nitrogen, and
(D) total organic carbon (TOG) and benzene in ground water 25
Figure 14 The distribution of dissolved As, Zn, Fe, and total S in shallow ground water adjacent to
the northern and eastern margins of the HBHA Pond 26
Figure 15 Temporal trends in water chemistry for the NML sampling station and adjacent ground-water
monitoring locations (TW07 andTW02) 27
Figure 16 The distribution of arsenic (As) in sediments collected from the HBHA Pond 28
Figure 17 The distribution of lead (Pb) and zinc (Zn) in sediments collected from the HBHA Pond 29
Figure 18 The distribution of chromium (Cr) and iron (Fe) in sediments collected from the HBHA
Pond 30
Figure 19 The distribution of sulfur (S) and organic carbon (OC) in sediments collected from the
HBHA Pond 31
VII
-------
Figure 20 The ratio of As, Fe, Pb, and Zn extracted by HCI for unoxidized and oxidized sediments
collected from suboxic zones within the HBHA Pond 37
Figure 21 (A-D) Comparison of the amount of As and Fe extracted by an ascorbate solution as a
function of total Fe and S content in sediments
(oxic = open symbols, suboxic = filled symbols) 37
Figure 22 X-ray diffraction data showing the relative proportion of hematite and ferrihydrite in the
<2 urn size fraction isolated from surficial sediment samples WI01, WI01-NEP, and
WI02-NEP collected near the north-northwestern margin of the HBHA Pond 38
Figure 23 Locations of a selection of sediments collected from oxic and suboxic zones within the
HBHA Pond 40
Figure 24 XANES data for (A) a series of reference compounds and (B) a selection of sediments
collected within the HBHA Pond 41
Figure 25 XANES data for (A) a series of reference compounds and (B) a selection of sediments
collected within the HBHA Pond 41
Figure 26 Aerial distribution of contaminant sediment speciation within the HBHA Pond based on
LCF-XANES analysis; contour lines delineate depth to sediment (meters) 42
Figure 27 Characterization of the mineralogy of the clay-sized sediment fraction and zinc speciation
in the silt-sized fraction of surficial sediment sample WI02-NEP before and after incubation
for a period of 2.5 years 43
Figure 28 Representative compositional spectra for samples imaged using SEM-EDS;
UP = unidentified peak 45
Figure 29 Image of a pyrite framboid observed in a surficial (oxic) sediment
(WI02-NER clay-sized fraction) 45
Figure 30 Images of iron (hydr)oxide precipitates collected near the chemocline within the water
column of the HBHA Pond (PS-3) 46
Figure 31 Illustration of the apparent geochemical processes controlling solid-solution partitioning
of arsenic, lead, and zinc within the HBHA Pond 48
Figure C.1 Location of XANES and EXAFS regions of an XAS spectrum 64
Figure C.2 Aerial photo and architectural diagram of the Advanced Photon Source at Argonne
National Laboratory, Chicago, IL 65
Figure C.3 Experimental configuration for transmission data collection 65
Figure C.4 Experimental configuration for fluorescence data collection 65
Figure C.5 Standard raw XAFS spectra illustrating the three regions: (A) pre-edge, (B) edge step,
and (C) EXAFS 66
Figure C.6 Solid sample sandwiched between pieces of Kapton tape 67
Figure C.7 Sample backfilled into opening of a Teflon block and secured with tape 67
Figure C.8 Autosampler template for multi-sample analysis 67
Figure C.9 Fluorescence data collection of metal hyperaccumulation in plant leaves 68
Figure C.10 Raw fluorescence data Pb sorption on ferrihydrite 68
Figure C.11 Background corrected spectra of Figure C.10 69
Figure C.12k-space conversion of spectra in Figure C.11 69
Figure C.13The k3-weighted -function of Figure C.12 70
Figure C.14 Radial distribution (or structure) function of Fourier transformed k-space data for
Figure C.13 70
Figure C.15 ATOMS input file for magnetoplumbite listing crystallographic information 71
viii
-------
Figure C.16FEFF file used to determine fitting paths for EXAFS modeling showing the interaction of
a central Pb atom (Pb1) with two different oxygen atoms (O5 and O3) and three different
iron atoms (Fe2, Fe5, and Fe4) 71
Figure C. 17 Structural data derived from ab-initio calculated fitting paths for Pb sorption on
ferrihydrite 72
Figure C.18XANES spectra (thin line) and derivative of XANES spectra (thick line) for aqueous Zn2+
(blue) and ZnS (red) 73
Figure C.19XANES spectra for arsenite [As(lll) - blue] and arsenate [As(V) - red] 73
Figure C.20 Linear combination fitting of X-ray absorption near edge spectroscopy data (LCF-XANES)
for a sediment (Sedimentl) sample with multiple Zn species 74
Tables
Table 1 Percent of Total Element Released by Selected Wet Chemical Extraction Tests 35
Table 2 Percent of Total Element Released by Selected Wet Chemical Extraction Tests 36
Table 3 Atomic Percentages of Elements in 5 Samples Determined by SEM-EDS 44
Table A.1 Analytical Methods, Detection Limits, Precision, and Accuracy for Measurement of
Aqueous Chemistry 55
Table A.2 Analytical Methods, Detection Limits, Precision, and Accuracy for Measurement of Solid
Phase Chemistry 56
Table B.1 Concentrations of Selected Elements in Halls Brook Holding Area Pond Core NC01
57
Table B.2 Concentrations of Selected Elements in Halls Brook Holding Area Pond Core CC02
58
Table B.3 Concentrations of Selected Elements in Halls Brook Holding Area Pond Core SC02
58
Table B.4 Concentrations of Selected Elements in Halls Brook Holding Area Pond Grab Sediment
Samples 59
Table B.5 Concentrations of Selected Elements in Halls Brook Holding Area Pond Cores 60
Table B.6 Results from LCF-XANES Fits of the Pb XANES Data Collected for Sediments from
Suboxic and Oxic Zones within the HBHA Pond 61
Table B.7 Results from LCF-XANES Fits of the Zn XANES Data Collected for Sediments from
Suboxic and Oxic Zones within the HBHA Pond 62
IX
-------
Acknowledgments
Joseph LeMay (U.S. EPA-Region 1, Boston, MA), Thomas Holdsworth (U.S. EPA/ORD-Cincinnati), and
Tim Bridges (U.S. EPA-Region 1 Laboratory, Chelmsford, MA) provided valuable assistance and guidance
during field sampling. Ning Xu, Sandra Saye, and Jihua Hong provided analytical support for determi-
nation of metals and arsenic speciation; Lynda Pennington, Kelly Bates, and Brad Scroggins provided
analytical support for determination of aqueous carbon, nitrogen species and major anions (Contracts
#68-0-98-138 and #68-0-03-097). Becky Butler (U.S. EPA/ORD-Ada) along with Martha Williams and
Trina Perry (Contract #68-W-01-032) assisted with editing and formatting for publication. The authors
greatly appreciate support provided by the staff of DND-CAT and PNC-CAT. DND-CAT is supported by
the E.I. DuPont de Nemours & Co., The Dow Chemical Company, the U.S. National Science Foundation
through Grant DMR-9304725, and the State of Illinois through the Department of Commerce and the
Board of Higher Education Grant IBHE HECA NWU 96. PNC-CAT is supported by the U.S. Department
of Energy, Basic Energy Sciences, under Contract DE-FG03-97ER45628, the University of Washington,
and grants from the Natural Sciences and Engineering Research Council of Canada. Use of the Ad-
vanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office
of Energy Research under Contract W-31-102-Eng-38. This document benefited significantly from criti-
cal and constructive reviews from Christopher Impellitteri (U.S. EPA), Janet Hering (California Institute of
Technology), Madeline Schreiber (Virginia Polytechnic Institute & State University), and Sonia Nagorski
(University of Alaska Southeast).
xi
-------
Executive Summary
Physical and chemical interactions between adjacent ground-water and surface-water bodies are impor-
tant factors impacting water budget and contaminant transport within a watershed. These interactions
are also of importance for hazardous waste site cleanup within the United States, since about 75% of
sites regulated under the Resource Conservation and Recovery Act (RCRA) and the Comprehensive
Environmental Response, Compensation, and Liability Act (CERCLA, or Superfund) are located within a
half mile of a surface water body. The boundary between adjacent ground-water and surface-water bod-
ies is referred to as the ground-water/surface-water (GW/SW) transition zone. The transition zone plays
a critical role in governing contaminant exchange and transformation during water exchange between the
two water bodies. The transition zone is host to a wide diversity of aquatic organisms, and it also can
serve as a sink for contaminants transported in surface water or ground water. Ultimately, the potential
for human exposure within a watershed and the health of the ecosystem inhabiting the transition zone will
depend on the bioavailability of accumulated contaminants. The extent of contaminant bioavailability will,
in part, be dictated by partitioning reactions that control the distribution and speciation of contaminants
within water and sediments in the GW/SW transition zone.
The purpose of this document is to illustrate some of the chemical processes that govern contaminant
transport and speciation during water exchange across the GW/SW transition zone. The focus of this
document is the assessment of metal speciation transformations in contaminated sediments. Results
from a field investigation of the fate of arsenic, lead, and zinc transported across the GW/SW transition
zone at a contaminated site are presented in order to illustrate the importance of using a site conceptual
model and to provide an example of approaches that may be used to characterize the spatial and tem-
poral distributions of inorganic contaminant speciation. The field site described in this report is located
immediately downgradient from the Industri-Plex Superfund Site in Woburn, MA and is characterized
by a ground-water contaminant plume that discharges into the Halls Brook Holding Area (HBHA) Pond
resulting in contamination of surface water and sediments.
The results from this field investigation provide insight into the source of inorganic contaminants within the
HBHA Pond and present a conceptual framework relative to the design of strategies to mitigate human
exposure to site-derived contaminants. Spatial and temporal trends in ground-water data indicate that
arsenic and zinc observed within surface water and sediments of the HBHA Pond are primarily derived from
continuing ground-water discharge. In contrast, lead observed in sediments appears either to be derived
from historical discharges or is currently derived from sediment transport or soil erosion from upgradient
source areas. In addition, the vertical distribution and temporal patterns in arsenic concentrations within
the water column of the HBHA Pond indicate that sediment dissolution/desorption processes contribute
to the overall dissolved concentration of this contaminant. The fate of these inorganic contaminants is
coupled to the fate of iron and sulfate derived from ground-water discharge. Iron (hydr)oxides are actively
produced in oxic portions of the HBHA Pond, while iron sulfides are produced in suboxic/anoxic portions
of the HBHA Pond. The generation of iron (hydr)oxides is a result of oxidation and precipitation of ferrous
iron upon contact with oxygen within oxic portions of the HBHA Pond. The generation of iron sulfides
is tied to microbial sulfate reduction coupled with degradation of anthropogenic and naturally occurring
dissolved organic compounds in discharging ground water within suboxic/anoxic portions of the HBHA
Pond. These newly formed precipitate phases possess significant sorption capacity for arsenic, lead,
and zinc. The retention of these solids within deeper portions of the HBHA Pond water column and/or
sediments helps to mitigate downgradient transport of these freshly deposited contaminants through the
watershed. Analytical data are presented to define the spatial distribution and chemical speciation of
arsenic, lead, and zinc within HBHA Pond sediments relative to the spatial distribution of predominant
redox processes throughout the pond.
XIII
-------
While the observed distributions of arsenic, lead, and zinc in ground water, sediments, and surface water
described in this report are ultimately unique to site-specific characteristics, some observed patterns in
contaminant geochemistry are likely relevant to other sites of contamination. Specifically, the results of
this study may have application to sites where an anoxic iron-rich ground-water plume encounters an
oxygenated environment (e.g., anoxic landfill leachates or organic contaminant plumes). The results of
this intensive field investigation provide useful information on technical approaches to characterize inor-
ganic contaminant transport through subsurface redox gradients that are frequently established at sites
with co-occurring organic and inorganic contaminant plumes. As demonstrated for this particular site, the
ultimate fate of inorganic contaminants will depend on both site-specific characteristics (e.g., hydrology
and soil/sediment mineralogy) and the chemical properties of the contaminant in question.
XIV
-------
Chapter 1
Purpose
Physical and chemical interactions between adjacent ground-water and surface-water bodies are important factors im-
pacting water budget and nutrient/contaminant transport within a watershed (Winter et al., 1998). These interactions are
also of importance for hazardous waste site cleanup within the United States, since about 75% of sites regulated under
the Resource Conservation and Recovery Act (RCRA) and the Comprehensive Environmental Response, Compensa-
tion, and Liability Act (CERCLA, or Superfund) are located within a half mile of a surface water body (Tomassoni, 2000;
Biksey and Gross, 2001). The boundary between adjacent ground-water and surface-water bodies is referred to as the
ground-water/surface-water (GW/SW) transition zone. The transition zone plays a critical role in governing contaminant
exchange and transformation during water exchange between the two water bodies. The transition zone is host to a
wide diversity of aquatic organisms, and it also can serve as a sink for contaminants transported in surface water or
ground water. Ultimately, the health of the ecosystem inhabiting the transition zone will depend on the bioavailability of
accumulated contaminants. The extent of contaminant bioavailability will, in part, be dictated by partitioning reactions
that control the distribution and speciation of contaminants within water and sediments in the GW/SW transition zone.
The purpose of this document is to illustrate some of the chemical processes that govern contaminant transport and
speciation during water exchange across the GW/SW transition zone. The focus of this document is the assessment
of metal speciation transformations in contaminated sediments. Results from a field investigation of the fate of arsenic,
lead, and zinc transported across the GW/SW transition zone at a contaminated site are presented in order to illus-
trate the importance of using a site conceptual model and to provide an example of approaches that may be used to
characterize the spatial and temporal distributions of inorganic contaminant speciation. The field site described in this
report is characterized by a ground-water contaminant plume that discharges into a pond resulting in contamination of
surface water and sediments. While the observed distributions of ars.enic, lead, and zinc in ground water, sediments,
and surface water described in this report are ultimately unique to site-specific characteristics, some observed patterns
in contaminant geochemistry are likely relevant to other sites of contamination. Specifically, the results of this study
may have application to sites where an anoxic iron-rich ground-water plume encounters an oxygenated environment
(e.g., anoxic landfill leachates or contaminant plumes).
Assessment of the factors controlling contaminant transport and distribution between interacting ground-water and
surface-water bodies will be guided by the conceptual model that delineates the relevant hydrologic and biogeochemical
processes. Typically, a conceptual model is developed based on site-specific data, which is subsequently revised in an
iterative fashion with continued accumulation of data that document relevant processes active on site. The complexity
of the conceptual model and the extent of required refinement or revision will be dictated by site heterogeneity and
process variability in time and space.
In a companion publication, a site conceptual model was developed for the GW/SW transition zone at the Industri-Plex
Superfund Site largely based on temporal measurements of surface water chemistry adjacent to the region of con-
taminated ground-water discharge within the Halls Brook Holding Area (HBHA) Pond (see Figures 6 and 11 in Ford,
2005; http://www.epa.gov/ahaazvuc/pubs/rschbrief.html}. In general, three geochemical zones that played an important
role in governing the partitioning of arsenic between liquid and solids across the GW/SW transition zone were identi-
fied within the proposed site conceptual model as depicted below in Figure 1. The emphasis of the initial report was
directed towards illustration of the importance of surface water hydrologic dynamics on observed temporal variations in
contaminant concentrations in a setting where contaminated ground water was suspected as the primary contaminant
source. The spatial and temporal dynamics observed for arsenic concentrations within the HBHA Pond water column
were best explained by considering joint contributions from both contaminated ground-water discharge and sediment
dissolution/desorption reactions. However, the characteristics of arsenic partitioning to sediments within the HBHA
Pond were not specifically explored in that document. Additional data are provided in this report to provide greater
understanding of the characteristics of sediments deposited within the HBHA Pond and the processes that control the
solid-phase partitioning of arsenic, lead, and zinc.
-------
GW/SW Transition Zone
HBHA Pond
Chemodine
Upgradient
Aquifer
Iron Oxidizing Zone
Oxidation-precipitation of ferrous iron
leading to the production of iron
(hydr)oxides that sequester arsenic from
solution
Iron Reducing Zone
Reductive dissolution of settling iron
(hydr)oxides coupled with release of
arsenic to solution
Sulfate Reducing Zone
Precipitation of iron sulfides that
sequester a fraction of arsenic derived
from ground-water discharge and
dissolution of settling iron (hydr)oxides
Figure 1 Conceptual model of geochemical zones that play a dominant role in governing solid-liquid partitioning of
arsenic (lead and zinc) across the GW/SW transition zone at the Industri-Plex Superfund Site.
The issue of contaminated sediments within the HBHA Pond is important relative to the desire to reduce contaminant
exposure and potential health risks. While elimination of the ground-water plume would benefit site restoration, this
action alone would not eliminate all risks associated with historical contamination due to the recalcitrance of inorganic
contaminants within the HBHA Pond. Based on the simplified conceptual model of the geochemical processes control-
ling the fate of arsenic, lead, and zinc shown in Figure 1, three key questions were posed relative to the information
needed to characterize the site:
1) What is the chemical composition of the ground water discharging into the HBHA Pond and how does this
vary spatially and temporally?
2) What are the specific processes that result in the partitioning of arsenic, lead, and zinc to solids that are
deposited within the HBHA Pond?
3) What is the stability of these solids (and their associated contaminants) and what physicochemical processes
can disturb the functionality of the HBHA Pond relative to the downgradient migration of contaminants?
Site characterization data on the temporal and spatial distributions of dissolved arsenic in the northern portion of the
HBHA Pond and the upgradient aquifer were reported by Ford (2005). This information provided partial answers to
the three questions posed above. Specifically, water chemistry data indicated that: 1) nearly continuous stratification
(i.e., development of a stable chemocline) within the HBHA Pond caused by discharge of contaminated ground water
with elevated dissolved solids content lead to the development of an internal redox process that controlled arsenic and
iron precipitation-dissolution reactions, 2) the concentration of dissolved arsenic within the water column of the HBHA
Pond was controlled by the balance between ferrous iron oxidation coupled with precipitation of iron (hydr)oxides near
the chemocline and the reductive dissolution of the newly formed iron (hydr)oxides as they settled through the anoxic
hypolimnion, and 3) the apparently stable stratification within the HBHA Pond could be disrupted by an exceptionally
high surface water influx, although the chemocline was re-established through time and the arsenic-iron redox cycle
restored. However, the results documented by Ford (2005) provided limited detail of the chemical composition of ground
water, and they provided no specific insight into the extent and characteristics of arsenic (lead and zinc) partitioning
to sediments for dissolved and particulate contaminants accumulated within the hypolimnion. Accordingly, the objec-
tives of this report are to elaborate on the physicochemical processes controlling arsenic, lead, and zinc partitioning to
sediments within the HBHA Pond and to evaluate the stability of sediment-associated contaminants along with overall
performance of the HBHA Pond in mitigating contaminant migration.
-------
Study Scope
As will be discussed later, the fate of arsenic, lead, and zinc is tied to the interaction between ground water and surface
water within the HBHA Pond. Sediments within the HBHA Pond play a critical role in the biogeochemical linkage of
these two water sources. These sediments originate from two sources: 1) mineral precipitates derived from discharging
ground water, e.g., due to ferrous iron oxidation and precipitation of iron (hydr)oxides, and 2) deposition of particulates
derived from terrestrial sources, e.g., biotic sources within the water column and along the margin of the HBHA Pond
and eroded soils. Several studies have been conducted to examine the fate of arsenic and other metals across the
sediment-water interface within the water column of lakes and reservoirs (e.g., Aggett and O'Brien, 1985; Aggett and
Kriegman, 1988; Seyler and Martin, 1989; Balistrieri et al., 1994; Spliethoff et al., 1995; Sohrin et al., 1997; Kneebone
et al., 2002; Martin and Pederson, 2002; Senn and Hemond, 2002). In many of these studies, it has been shown that
the fate of arsenic is tied to the chemical cycling of iron. Iron may be continuously recycled within the water body for
systems in which anoxia develops within the hypolimnion or underlying sediments. Reduced iron generated in the
lower regions of a lake can diffuse upward within the water column and re-precipitate upon contact with oxygen or other
oxidants (e.g., Davison et al., 1982; Sholkovitz and Copland, 1982). Dissolved arsenic within the water column can
partition to the newly-formed iron (hydr)oxides either through coprecipitation at the time of formation or adsorption to
settling iron (hydr)oxide particles (Seyler and Martin, 1989; Balistrieri et al., 1994; Spliethoff et al., 1995; Sohrin et al.,
1997; Senn and Hemond, 2002). Arsenic partitioned to the settling iron (hydr)oxides can subsequently be remobilized
during reductive dissolution within the hypolimnion or following sedimentation depending on the spatial distribution of
reducing conditions within the lake system. This 'natural' biogeochemical cycle may overlap with hydrologically-driven
contaminant fluxes across the GW/SW transition zone, which can cause misleading interpretations of the environmental
risk that is directly attributable to discharge of contaminated ground water (See 'Summary and Conclusions' in Sholkovitz
and Copland, 1982.).
The purpose of this study was to evaluate the fate of arsenic, lead, and zinc within the HBHA Pond system immediately
downgradient from the Industri-Plex Superfund Site, which hosts contaminated soils and ground water resulting from
industrial activities dating back to the mid-1800s (Davis et al., 1994; Aurilio et al., 1995; Wick and Gschwend, 1998).
The fate of arsenic and other inorganic contaminants in this system is coupled to the chemical cycling of iron, sulfur,
and carbon due to the characteristics of the ground-water plume discharging into the surface water system. The results
of field measurements to ascertain the spatial and temporal variations of the geochemistry within the GW/SW transition
zone and the chemical speciation of arsenic, lead, and zinc across redox gradients within HBHA Pond sediments are
examined below to identify the factors controlling the mobility of these contaminants.
-------
Chapter 2
Site Background
The Aberjona Watershed is an industrialized 65-km2 watershed located in northeastern Massachusetts. The watershed
boundary lies within the towns of Reading, Woburn, Winchester, and Medford. The Aberjona Watershed extends to the
Mystic Lakes, which drain into the Boston Harbor via the Mystic River. Historic industrial activities have resulted in
contamination within the northern extent of the watershed. Specifically, activities at the Industri-Plex Superfund Site,
a 245-acre industrial park located in Woburn, Massachusetts, resulted in the deposition of inorganic and organic con-
taminants in soils, sediments, and ground water. In response to public health concerns, a Record of Decision (ROD)
for the Industri-Plex Superfund Site was signed in 1986, addressing on-site soil, sediment, and hot spot ground water
contamination.
Contamination at the Industri-Plex Superfund Site was a result of historical industrial activity. From the mid-1800s to
the 1930s, leather tanning was the dominant industry in Woburn, and the Aberjona River and its tributaries served
as the main conduits for tannery wastewater (Durant et al., 1990). During the same period, starting in the late 19th
century, sulfuric acid and arsenical pesticide manufacturing took place near the headwaters of the watershed in an
area now designated as the Industri-Plex Superfund Site. Estimates suggest that 270 metric tons of arsenic may still
exist within the site boundaries (Aurilio, 1995). By-products such as unused portions of animal hides from tanning and
glue-manufacturing operations from 1934-1969 were disposed on site. It is hypothesized that leached degradation
products from these materials have contributed to the mobilization of arsenic via the ground water from source areas
to a wetland that serves as a source of surface water to the Aberjona River (Davis et al., 1994). Ground water that,
in part, originates from within the boundaries of the Industri-Plex Superfund Site discharges into a pond known as the
Halls Brook Holding Area (HBHA) Pond. This surface water feature was constructed to serve as a hydraulic retention
basin to mitigate flooding during periods of peak surface water discharge. Sources of surface water to the HBHA Pond
include Halls Brook, a perennial stream located on the western edge of the pond, and an intermittent stream (Atlantic
Avenue Drainway) that conveys water from an upgradient wetland and stormwater runoff from Atlantic Avenue and
nearby parking facilities. The HBHA Pond is located in the northern-most portion of a wetland area referred to as the
Halls Brook Holding Area. An aerial view of the study location, including portions of the Industri-Plex Superfund Site
('hide piles') is provided in Figure 2.
The ground water beneath the Industri-Plex Superfund Site and the study area in North Woburn has been designated as
a non-drinking water source area by the state of Massachusetts. This designation may influence selection of remedial
alternatives to address the contaminated soils, ground water, surface water, and sediments. It is possible that most, if
not all, of contaminated ground water discharges to the HBHA Pond down gradient from the Industri-Plex Superfund
Site. If true, this pond may sequester or retard down gradient transport of arsenic, lead, and zinc. However, there is
uncertainty as to the speciation of these contaminants in the various environmental media, interactions between the
surface water and ground water, and stability of contaminated sediments within the HBHA Pond. A critical aspect to this
site investigation was the resolution of these uncertainties prior to considering remedial action alternatives for contami-
nation at the Industri-Plex Superfund Site. This report expands on the description of the contaminated ground-water
plume originating from the Industri-Plex Superfund Site and the processes that control the fate of site contaminants as
they are transported across the GW/SW transition zone (Ford, 2005).
Study Approach
The field sampling activities and laboratory analyses performed under this study provide a current assessment of ar-
senic, lead, and zinc distribution in ground water, surface water, and sediments at the Industri-Plex Superfund Site and
within the HBHA Pond. Site characterization was directed towards identifying the predominant chemical processes
controlling arsenic, lead, and zinc migration to and sequestration within the HBHA Pond. Field-based sampling was
carried out over a period of 35 months to assess time-dependent trends in contaminant mobility within the Industri-Plex
Superfund Site and the HBHA Pond. The data derived from this effort provide a means for 1) assessing the long-term
assimilative capacity within the HBHA Pond, and 2) the potential for future mobilization of arsenic, lead, and zinc par-
titioned to sediments.
-------
200ft
:,>••* V.
* Atlantic AvenueDrainway \
'
Industri-Plex Ground Water '•
.. " Treatment Building V*
Regional
Ground-Water
' Ffow
Hafte Brook
Figure 2
Prominent land features within the Industri-Plex Superfund Site and adjacent to the HBHA Pond. Land
features labeled as 'hide piles' are suspected source terms for metal contaminants within the Industri-Plex
Superfund Site. Image was derived from May 1995 aerial photograph obtained from MassGIS.
Results from ground water, surface water, and sediment sampling and characterization are summarized and evalu-
ated in the sections and appendices that follow. Collection of ground water, surface water, and sediment samples was
completed during the following dates: October 13-21, 1999; November 30 - December 3, 1999; March 27 - April 6,
2000; May 16-18, 2000; August 22-30, 2000; March 26 - April 14, 2001; May 11-17, 2001; September 10-21, 2001; and
September 20-23, 2004. The study was initiated by defining the extent of the plumes of arsenic and BTEX compounds
within the aquifer upgradient to the HBHA Pond. This was accomplished by collecting ground-water samples at tem-
porary wells installed at discrete depths using direct-push equipment (Figure 3A). Patterns in water chemistry data
collected from these sampling locations were then used to guide the installation of permanent wells that were used to
monitor temporal trends in water chemistry adjacent to and within the HBHA Pond (Figure 3B). An effort was made to
capture the vertical heterogeneity anticipated within the GW/SW transition zone by the installation of discrete sampling
points at various locations (illustrated in Figure 4).
-------
Figure 3 (A) Arsenic (As), benzene (Bz), and toluene (Tl) concentrations (ppb) from temporary ground-water sampling locations as a function of
depth below ground surface (feet) for the study site. NO = not detected, BLQ = below limit of quantitation. Image was derived from May
1995 aerial photograph obtained from MassGIS. (B) Permanent ground-water monitoring locations established adjacent to the HBHA Pond.
Images were derived from April 2001 aerial photograph obtained from MassGIS.
-------
' to. •
TW10
Arsenic
TW06
TEX
200ft
TW02
TW04
TW01
NlVfb
• TWOS
TWOS
B
NML
i
TW02
i
manhole cover
TW07
1
Sondeju
Multi-
Sampling
Station
Single
Completion
Well
Multiple ,
Completion •
Well
Diffusion
Sampler
Figure 4 A) Close-up view of permanent ground-water monitoring network; arrows show approximate centerline of
arsenic and BTEX plumes discharging into HBHA Pond. B) Illustration of the approach used to provide
depth-resolved spatial coverage of the GW/SW transition zone adjacent to the HBHA Pond. Sampling in-
tervals were completed with 6-inch stainless steel screens connected to 3/8-inch Teflon-lined polyethylene
tubing, except for nylon membranes used for cells in diffusion sampler.
-------
Site Characterization Methods
Water physicochemical parameters were measured in the field along with the collection of water and sediment samples
during the sampling dates previously listed. The objective of this effort was to define the chemical structure of the
system with particular attention being given to delineation of the GW/SW transition zone. Historical data had been col-
lected as part of previous site characterization efforts to demonstrate a net decrease in arsenic (and other contaminant)
concentrations in transitioning from the upgradient aquifer to the outlet of the HBHA Pond. However, limited detailed
liquid and solid phase characterization data had been collected to aid in establishing the specific process(es) that
controlled the transport of inorganic contaminants through the system (Aurilio et al., 1994; Davis et al., 1996; Ahmann
et al., 1997). In particular, insufficient information was available to assess the influence of the seasonal dynamics of
water flow and microbial processes on the geochemistry of inorganic contaminants within the HBHA Pond. Wick and
others (Wick and Gschwend, 1998a; Wick and Gschwend, 1998b; Wick et al., 2000) had characterized the HBHA Pond
system with respect to the fate of synthetic organic compounds that entered the pond via ground-water discharge, and
this provided a solid basis from which to design a site characterization program. A range of geochemical data was
collected to refine and support the general site conceptual model that had been developed to describe the function of
the HBHA Pond relative to arsenic migration, i.e., the sequestration of arsenic, lead, and zinc discharged into the pond
via partitioning to Fe-bearing minerals that were continuously deposited onto the sediment layer. Details of sample
collection and characterization protocols are described below.
Field Measurements
Ground Water
The major and trace element chemistry and redox characteristics of ground water were assessed through field- and
laboratory-based measurements on water samples collected from the aquifer immediately upgradient and below the
HBHA Pond. In addition, the concentration distribution of dissolved contaminants was assessed to determine if there
were interactions between inorganic and organic contaminants that may influence biogeochemical processes active
within and adjacent to the GW/SW transition zone. Sufficient data were collected to define the overall redox chemistry
within ground water and how this influenced the chemical speciation of arsenic prior to its discharge into the HBHA
Pond. In general, this required measurements to determine the speciation of soluble carbon, iron, and sulfur including
the spatial and temporal variability in these constituents.
Ground-water samples were extracted from Teflon-lined tubing wells equipped with 6-inch stainless steel screens that
had been driven to depth using direct-push equipment. Samples of ground water were collected using a peristaltic
pump through a closed flow cell following stabilization of pH, specific conductance, turbidity, oxidation-reduction potential
(ORP), and dissolved oxygen (Puls and Paul, 1995). Filtered (0.45-um, cellulose acetate membrane) ground-water
samples were collected for laboratory analyses. The concentration of ferrous iron was determined by spectrophotomet-
ric detection of its complex with 1,10-phenanthroline at 510 nm (Hach, 1992a). The concentration of ferrous iron plus
dithionite-reducible ferric iron (total iron) was also assessed by spectrophotometric detection of the ferrous iron complex
with 1,10-phenanthroline at 510 nm following reduction with dithionite (Hach, 1992b). The alkalinity of water samples
was determined by titration with standardized sulfuric acid to a colorimetric endpoint near pH 4.5.
Surface Water
Depth profiling within the northern, central, and southern regions of the HBHA Pond was carried out at various times of
the year to assess the chemical structure of the water column and to delineate the processes controlling contaminant
partitioning between solution and solids that subsequently deposit onto the underlying sediments. General water chem-
istry measurements were conducted as a function of depth either by lowering an instrumented sonde down through the
water column or by pumping water to the surface through an enclosed and instrumented flow cell (Figure 4B). The sonde
or flow cell was instrumented to collect readings of temperature, dissolved oxygen, pH, specific conductance, and ORP
(platinum electrode). Water samples were collected as a function of depth within the water column by pumping water
to the surface for field determinations (ferrous iron and alkalinity) or laboratory measurements to determine the major
and trace element chemistry. A permanent multi-screened sampling station (NML) was installed within the northern
end of the HBHA Pond to facilitate multiple sampling rounds during or following storm events. As shown below and by
the work of Wick and others (Wick and Gschwend, 1998a; Wick et al., 2000), depth-resolved measurements within the
HBHA Pond water column were critical towards developing a clear understanding of the physicochemical processes
that controlled the fate of contaminants discharged from the upgradient aquifer.
Depth-resolved surface water chemistry within the HBHA Pond was assessed using aYSI 6820 Multi-Parameter Water
Quality Monitor to measure temperature, pH, specific conductance, ORP, and dissolved oxygen. Water samples were
collected at each depth for determination of turbidity and dissolved/particulate concentrations of arsenic, iron, and
organic carbon. Suspended solids were collected by in-line pressure filtration using a peristaltic pump from various
depths within the HBHA Pond water column for mineralogical characterization.
-------
Surveys of the distribution of specific conductance throughout the HBHA Pond were conducted prior to and following
a storm event to understand the dynamics of physicochemical stratification throughout the water column. Continuous
measurements of specific conductance as a function of depth were collected using a Geoprobe Direct Image® Electrical
Conductivity System that had been adapted for deployment from a pontoon boat. Transects of specific conductance
were developed by contouring data from multiple vertical profiles collected within northern and southern portions of the
HBHA Pond. The locations of vertical profiles were recorded using a Trimble Navigation GeoExplorer GPS Receiver
with post-processing correction of locational data. Data collection events occurred on April 5, 2001 and May 11, 2001
following a major surface water flow event at the end of March 2001 (Ford, 2005).
Laboratory Measurements
Water Chemistry
Dissolved metals in ground water were determined by ICP-OES or ICP-MS for filtered samples acidified to pH<2 with
concentrated nitric acid (Fisher Optima). The concentration of benzene was determined by purge-and-trap gas chroma-
tography for filtered samples acidified to pH<2 with concentrated sulfuric acid (Fisher Optima). Dissolved sulfate was
determined by capillary electrophoresis with indirect UV detection (Standard Methods, 1999). Nitrate and ammonium
were determined by flow injection analysis with a Lachat Quikchem 800 Analyzer (U.S. EPA, 1983; Methods 353.2
and 350.1, respectively). Dissolved organic carbon was determined by UV catalyzed persulfate digestion with infrared
detection of evolved CO2 (U.S. EPA, 1983; Method 415.2).
Dissolved (0.2-um filter, nylon membrane) and total concentrations of metals in surface water samples were determined
by ICP-OES or ICP-MS for samples acidified to pH<2 with concentrated nitric acid (Fisher Optima). Samples for particu-
late metals were microwave-digested in 10% v/v nitric acid prior to analysis. Dissolved arsenic speciation (0.2-um filter,
nylon membrane) was determined by ion chromatography-hydride generation-atomic fluorescence spectrophotometry
(IC-HG-AFS) for samples acidified to pH<2 with concentrated hydrochloric acid (Optima) in amber polyethylene contain-
ers and stored at 4°C for a period of less than five days (McCleskey et al., 2004). Separate holding time studies using
a representative range of site-specific water chemistries demonstrated that changes in inorganic arsenic speciation
were less than the analytical error for a holding time of seven days or less (data not shown). The speciation method
was optimized for detection of As(lll), As(V), MMA(V), and DMA(V) using gradient elution with a phosphate mobile
phase (Gomez-Ariza et al., 2000).
Sediments
The major element and contaminant chemistry was defined in sediments collected from various locations within the
HBHA Pond during different times of the year. Locations that were deemed important included regions of obvious
ground-water discharge, i.e., ground-water seeps evidenced by the precipitation of iron (hydr)oxides, as well as regions
that represented varying stages along the redox gradient from shallow to deeper water levels within the pond. Shallow
sediment grab samples were collected throughout the HBHA Pond to help define the chemical processes that controlled
the distribution of contaminants across the sediment-water interface. In addition, sediment cores were collected to
greater depth in order to define the depositional history and diagenetic processes that exert an influence on inorganic
contaminants accumulated within the system. The predominant sediment mineralogy at various locations across the
system redox gradient was assessed along with characterization of the chemical speciation of arsenic, lead, and zinc
using chemical extraction tests as well as X-ray absorption spectroscopy on preserved sediment samples. Previous
work had shown that iron monosulfides were a significant component in sediments deposited/formed within suboxic
regions of the HBHA Pond (Wilkin and Ford, 2002). Results documented below demonstrated the predominance of iron
(hydr)oxides in the reactive fraction of sediments deposited/formed within oxic regions of the HBHA Pond.
Sediment samples were collected in December 1999, April 2001, and September 2001 from the HBHA Pond. The great-
est density of sampling points was along the center axis and along the northeast shore of the pond, where contaminated
ground water discharges into the HBHA Pond. Sediment samples were retrieved from water depths ranging from 0.5 to
4.5 m. Cores were collected using a 5-cm diameter piston-coring device. The cores were capped immediately after their
recovery and kept upright before and during freezing. Surface sediments were collected into N2-purged plastic bags by
pumping highly fluidized sediments to the surface using a peristaltic pump. Surface sediments and cores were frozen
within 1 h of collection from the pond bottom. While still frozen, sediment cores were sub-sampled by cutting off 2- to
5-cm thick sections, which were immediately bagged and kept frozen for subsequent solid-phase analyses.
Sediment Chemistry and Mineralogy
Surficial (oxic) sediments were dried in air to constant mass prior to chemical measurements or mineralogical determi-
nations. The <2 mm size fraction was isolated from these sediments by sieving prior to further analysis. Buried sedi-
ments collected from suboxic zones were thawed and dried at room temperature in an anaerobic glove box (96:4 v/v
N2-H2 gas mixture). Excess pore water was squeezed out by hand and decanted from the sediments prior to drying.
10
-------
Dried samples were homogenized with an agate mortar and pestle and retained in the anaerobic glove box. Sample
splits were removed from the glove box and allowed to oxidize in an oven at 60°C for 72 hours.
Methods used for determination of the elemental composition of sediments are listed in Appendix A along with method
detection limits, analytical accuracy, and precision. All solid-phase concentrations are reported on a dry weight basis. Total
carbon and carbonate carbon were determined on oven-dried samples using a UIC (Model CM5014) carbon coulometer
system (Huffmann, 1977). Organic carbon content was derived from the difference between total carbon and inorganic
carbon. For total carbon analysis, a 20-300 mg aliquot of powdered sediment was combusted at 950 °C. Inorganic
carbon was analyzed as carbon dioxide released from a sample after reaction with hot 5% perchloric acid. Total sulfur
was determined by mixing oven-dried sediment samples with vanadium pentoxide and combusting the mixture in the
presence of high-purity oxygen-gas at 1050 °C (Atkin and Somerfield, 1994). Sulfur dioxide produced is quantitatively
titrated using a sulfur coulometer (UIC Model CM5014S). Concentrations of metals and metalloids in the sediments
were determined by microwave assisted digestion in 10% HNO3 (modified EPA Method 3051), followed by inductively
coupled plasma optical emission spectrometry (ICP-OES, Perkin Elmer Optima 3300DV). Microwave digestion was
carried out using a CEM Mars 5 system at a total pressure of 120 psi (8.5 bar) and maximum temperature of 170 °C.
The reliability of this digestion method was continuously checked using international certified reference materials (NIST
2710, NIST 2780, and CCRMP LKSD-1). Arsenic recovery from these reference materials, for example, ranged from
88% to 97% of the certified values.
Mineralogical Characterization
Precipitate mineralogy was determined by powder X-ray diffraction (XRD) using a Rigaku MiniFlex diffractometer. XRD
data were collected using Fe Ka radiation via continuous scan in 0.02 °20 steps with a count time of 12 sec/step. The
<2 um size fraction of surficial sediment samples was isolated via centrifugation of aqueous suspensions. Air-dried
samples were prepared as smear mounts with methanol on zero background quartz slides (typically 10-15 mg of sample).
Mounted samples were allowed to air dry prior to data collection. External calibration using a smear mount of the NIST
SRM 640b (silicon) was carried out on a regular basis to confirm goniometer angular position accuracy and to estimate
peak position error due to sample displacement. Peak identification for crystalline phases was achieved by reference
to the ICDD Powder Diffraction File database (JCPDS, 1993). Identification of 6-line and 2-line ferrihydrite was carried
out by reference to data reported by Stanjek and Weidler (1992).
X-ray Absorption Spectroscopy
Arsenic XAS
The oxidation state and bonding environment of arsenic associated with iron oxidation precipitates was examined us-
ing X-ray absorption near edge structure (XANES) spectroscopy. Arsenic K-edge spectra were collected at sectors
20-BM (DuPont-Northwestern-Dow Collaborative Access Team; DND-CAT) and 5-BM (Pacific Northwest Consortium
Collaborative Access Team; PNC-CAT) at the Advanced Photon Source, Argonne National Laboratory (Argonne, IL).
Dried sediment samples were loaded into 1-mm-thick plastic sample holders and sealed with strips of Kapton tape.
Absorption spectra were collected at the As K-edge (11,867 eV) in fluorescence mode using a 13-element solid-state
Ge-detector. The synchrotron was operated at 7.0 GeV and at a nominal 100 mA fill current. The energy of a Si(220)
double-crystal monochromator was calibrated using arsenic or gold foil. The monochromator step size was 0.25 eV
per step in the XANES region (11845-11900 eV). Multiple scans were collected and summed for each sample (3 to 9).
The XANES fluorescence data were normalized to the edge-jump height, and the K-edge inflection point was deter-
mined as the energy at the maximum in the first derivative of the normalized spectra. Mineral reference compounds
were synthesized in the laboratory, including arsenate adsorbed to hematite (Ford, 2002), arsenite coprecipitated with
siderite, and arsenite coprecipitated with mackinawite (Wilkin and Ford, 2002). All fluorescence spectra were collected
at room temperature. For surficial (oxic) sediments and the arsenate-hematite reference sample, no attempt was made
to exclude sample exposure to air. However, buried (suboxic) sediments and the reference samples prepared using
arsenite were protected from oxygen exposure in an inert atmosphere prior to and during collection of fluorescence
data in order to prevent sample oxidation. The relative percentages of As(V)-O, As(lll)-O, and As-S species in the study
samples were assessed by linear combination fitting (LCF) of XANES data employing WinXAS Version 3.1 (Ressler,
1998) using reference spectra for As(V)-hematite, As(lll)-siderite, and As(lll) coprecipitated with FeS. The LCF-XANES
analysis demonstrated that an orpiment-like phase (As2S3) was an insignificant component for all examined samples.
Pb and Zn XAS
The sediment samples were analyzed for Pb and Zn by XANES spectroscopy to determine metal speciation parameters.
XANES spectroscopic data were collected at beamline 20-BM (Pacific Northwest Consortium - Collaborative Access
Team) at the Advanced Photon Source at Argonne National Laboratory, Argonne, IL. The electron storage ring operated
at 7 GeV with a top-up fill mode. Three to five scans were collected at ambient temperature in fluorescence mode with a
13-element Ge detector. A 0.5 mm premonochromator slit width and a Si(111) double crystal monochromator detuned
11
-------
by 20% to reject higher-order harmonics were employed. For XANES analyses, the sediment samples were loaded into
Teflon sample holders and sealed with Kapton tape within an oxygen-free glovebox. The samples were transferred to
the experimental hutch in a plastic bag which was purged with Ar during transfer and data collection to avoid oxidation
from atmospheric oxygen. No beam-induced oxidation was evident. No background Pb or Zn was detected in the purge
bag material or the Kapton tape. Energy was calibrated to the first inflection of ZnK (9.659 keV) or PbLII (15.200 keV)
metal foil standards and was collected simultaneously with the spectrum of each sample for the respective metals. Due
to the relative low concentration of Pb and high concentration of As in the samples, the PbLII edge energy was utilized
for Pb data collection rather than the usual PbLIII edge energy of 13.055 keV. The PbLIII La1 fluorescence energy of X-ray
emission is 10.552 keV which shouldered on the overwhelming AsK Ka1 emission energy peak at 10.544 keV. The col-
lected scans for a particular sample were merged, the data were then normalized, and the background was removed
by spline fitting using WinXAS 3.1 (Rehr and Ankudinov, 2001, 2003). The XANES spectra of Pb and Zn in sediment
samples and references were then assembled in a format suitable for linear combination fitting (Beauchemin et al., 2002;
Roberts et al., 2002; Scheinost et al., 2002; Scheckel and Ryan, 2004) of the X-ray absorption near edge spectroscopy
(LCF-XANES) data to identify the major Pb and Zn species present in the sediment samples.
Sediment Extraction Procedures
Surficial Sediments
A series of single-step chemical extractions were employed to assess the relative stability of metals associated with oxic
sediments exposed at the surface near visible ground-water seeps along the north-northwestern margin of the HBHA
Pond. The extraction steps employed in this procedure are documented below. All stock solutions were prepared using
reagent grade chemicals and deionized water (Millipore® Milli-Q3RO/Milli-Q2 system). All glassware had been previ-
ously soaked in dilute HNO3 and rinsed thoroughly with deionized water.
i) Weakly Adsorbed: A 20 ml aliquot of 1 M MgCI2 (Baker®) at pH 7.0 was added to 45 ml polyethylene cen-
trifuge tubes containing either 0.1 grn solid for WI01-NEP, WI02, WI02-NEP, and WI04 or 1.0 gm for WI01.
Samples were prepared in duplicate. The tubes were continuously shaken for 1 hr (Tessier et al., 1979)
and centrifuged for 10 min at 15,000 RPM. The supernatant was carefully pipetted off and filtered through
0.2 um Nuclepore® filter.
ii) Strongly Adsorbed: The residue from (i) was extracted with 19 ml 0.005 M NaH2PO4 (Sigma®) at both pH 4
(Bartlett and James, 1988; Welch and Lico, 1998). Samples were shaken for 24 hours, and supernatants
were collected in the same manner as described in step (i).
Hi) Amorphous Iron Oxides: The ascorbate solution for Step 3 was prepared following the method of Ferdel-
man (1988). Sodium citrate (0.2 M C6H5Na3O7-2H2O) and sodium bicarbonate (0.6 M NaHCO3) were added
to deionized water. The solution was then deaerated with nitrogen and ascorbic acid (0.4 M C6H8O6) was
added with a resulting pH of 8.0. Sediment samples were extracted for 24 hours with 19 ml of the ascorbate
solution, and the supernatants were collected in the same manner as described in step (i).
iv) Iron Sulfides: The final step in the sequential extraction scheme involved adding 19 ml 0.5 M HCI (Aldrich®)
to the sediment samples and shaking for 1 hour (Kostka and Luther, 1994). Supernatants were collected
in the same manner as described in step (i).
Buried Sediments
A series of single-step chemical extractions were performed to better constrain the solid-phase partitioning of arsenic,
lead, and zinc in buried sediments collected from the bottom of the HBHA Pond. All solutions were deoxygenated prior
to use, and extractions were carried out using unoxidized sediments in an anaerobic glove box. The following solutions
were used: 0.5 M MgCI2, 0.1 M Na2CO3, 0.02 M ascorbic acid (buffered to pH 7 in 0.6 M NaHCO3), and 1 M HCI. Dilute
hydrochloric acid extractions were conducted on both unoxidized and oxidized sediment samples. Supernatant solu-
tions were filtered through 0.22-um syringe filters and acidified with concentrated nitric acid. The analysis of metals in
the solutions obtained following chemical extraction was carried out using ICP-OES.
All liquid extractant samples were acidified to pH < 2 using ultra-pure HNO3 and analyzed for the elements of interest
(arsenic, iron, lead, zinc) by ICP-OES. Samples with low arsenic concentration were also analyzed using GFAAS. Qual-
ity assurance measures performed on these analyses included analytical duplicates, known analytical quality controls
(AQCs), check standards, and blanks. Non-detect results were observed for all blank samples, and the results for all
other QC samples met the required data quality requirements (Appendix A).
Scanning Electron Microscopy-Energy Dispersive Spectroscopy
Scanning Electron Microscopy was used to evaluate the morphology and spatial relationships among mineral precipi-
tates for two surficial sediments, suspended solids collected adjacent to the chemocline within the HBHA Pond, and
12
-------
one buried sediment sample. In addition, energy dispersive X-ray spectroscopy (EDS) was conducted on polished and
unpolished samples to determine the composition of sediments on a semi-quantitative basis. Samples for SEM/EDS
analysis were stored in an anaerobic glove box and then embedded in an epoxy resin. The sample mounts (1" diameter
round mounts) were ground and polished using diamond abrasives and coated with a thin layer of gold prior to being
placed within the SEM sample chamber. The sample of suspended solids was the only unpolished sample analyzed for
these studies. The grain mount procedure used for this sample is discussed below.
Secondary electron and back-scattered electron images were obtained using a JEOL 6360 SEM. The instrument was
operated using an 18kV and 20 kV accelerating potential and a beam current of about 8.5 nA. Micrographs were obtained
at a range of magnifications from 1500X to 10000X. Copper grids obtained from SPI Supplies (West Chester, PA) were
used to verify quantitative lengths. EDS spectra were acquired using an OXFORD Instruments Model INCA300/GEM
EDS Unit. Elemental concentrations were calculated using INCA software and cobalt metal as a standard reference
material to insure semi-quantitative accuracy.
Polished thin-sections were made from sediments that had previously been imbedded in resin. Thin-section samples
were sonicated in methanol for 30 minutes prior to mounting, rinsed with RO water, and dried with an air-stream (canned
air) before proceeding any further. After the drying process, each prepared thin-section was mounted to an aluminum
stub using double-sided carbon tape, and then a strip of silver paint was applied from the aluminum stub to the top of
the sample thin-section for conductivity purposes. The silver paint was dried thoroughly with a stream of air before the
coating procedures were attempted. A 10-second gold sputter coat was applied to these samples in preparation for
SEM analysis. One of the sediment thin-sections was treated differently in the sample preparation because a pyrite
framboid was identified under an optical light microscope. The framboid could not be located with the SEM after the first
sample preparation. The thin-section was stripped of the first 10-second gold coat with a methanol/sonication process
for 30 minutes, rinsed with RO water, and then followed by a 2% potassium iodide/sonication process for 30 minutes.
The thin-section sample was then rinsed with RO water, dried with a stream of air (canned air), mounted on an alumi-
num stub with double-sided carbon tape, two strips of silver paint were applied on opposite sides of the thin-section for
conductivity reasons, and finally, a fresh 10-second gold coating was applied. By stripping the first gold coat, we were
able to mark the site of interest so that the framboid could be more easily located in the SEM system.
The suspended solid sample collected from the southern portion of the HBHA Pond was stored in an anaerobic glove-
box until ready for SEM/EDS sample preparation. A small sample (<1 mg) was suspended in 1 mL of methanol, soni-
cated for 10 minutes, agitated, and then sonicated a second time for 10 minutes. A carbon planchet was mounted on
an aluminum stub with double-sided carbon tape, and then one drop of the suspension was placed on the top surface
of the planchet. The sample was immediately placed into a dessicator for 2 days of vacuum drying. The sample was
then removed from the dessicator, two strips of silver paint were applied (opposite sides) and dried immediately with an
air stream (canned air) before the gold coating could be applied. The sample was then coated with gold for 10 seconds
to achieve good conductivity for the SEM/EDS analyses.
13
-------
Chapter 3
Site Hydrology
A portion of the ground-water plume discharges into the HBHA Pond that also receives surface water inputs from a
perennial stream (Halls Brook) on its western edge and an intermittent runoff channel at its northwestern end (Figure 2).
The HBHA Pond was built in the early 1970s for flood control. The HBHA Pond has a maximum depth of approximately
5 meters (15 feet), and it discharges into a heavily vegetated wetland area. The HBHA Pond discharge reconstitutes
Halls Brook, which meanders through the wetland until discharge into the Aberjona River. Independent estimates
of water flux into the HBHA Pond indicate that approximately 7-60% of the water input is derived from ground water
throughout the year (Aurilio et al., 1994; Wick et al. 2000). Trends in ground water and surface water inputs into the
HBHA Pond documented by these authors are illustrated in Figure 5. Data shown in Figure 5A and Figure 5B indicate
that the influx of ground water remains relatively constant, although the fraction of ground-water flow relative to the
total system flow varies significantly as a function of surface water inputs. For example, during late January 1998, the
relative fraction of ground-water flow varied from an average of approximately 60% down to 10% following an increase
in surface-water flow (Figure 5B, see history of water flow in Aberjona River at USGS 01102500). Water level measure-
ments in the HBHA Pond (this study) illustrate rapid fluctuations in water budget in this system in response to changes
in surface-water influx (Figure 5C).
Based on salt balance calculations, Wick et al. (2000) have estimated that a contaminated ground-water plume (denoted
'Qhigh' by the authors) contributes approximately 1 - 9% of the water input into the HBHA Pond throughout the year (3-27%
of ground-water input). Based on specific conductance measurements throughout the HBHA Pond, these authors
concluded that the primary location of contaminated ground-water discharge was within the northern end of the lake.
Independent measurements of ground water and lake water chemistry (including dissolved arsenic) from the current
study support this finding and show that contaminants are primarily discharged within the north-northeastern portion of
the pond (Figures 6-8; Pb below detection in ground water). However, arsenic contamination is derived from a larger
portion of the adjacent ground-water aquifer than the hydrocarbon contaminants studied by Wick et al. (1998, 2000).
Thus, their estimates of contaminated ground-water flux may be low with respect to the discharging arsenic plume. The
distribution of arsenic and benzene as a function of depth for three ground-water monitoring locations proximate to the
north-northeastern shoreline of the HBHA Pond (TW06, TW07, TWOS) indicates that the hydrocarbon plume occurs
over a smaller region than the arsenic plume. While the highest arsenic concentrations were encountered at location
TW07, there were still substantial concentrations of arsenic in the vicinity of the hydrocarbon plume (TWOS). Since
additional hydrologic data were not collected during this study, there is insufficient information to properly constrain an
estimate of the flux of arsenic derived from ground water. However, published data of changes in the vertical distribu-
tion of dissolved arsenic in the HBHA Pond water column at location NML demonstrate that observed concentrations
of arsenic within this pond are derived from a combination of ground-water inputs and sediment dissolution processes
(Ford, 2005).
An important component of the conceptual model developed thus far is the maintenance of a stable chemocline within
the HBHA Pond that serves to regulate the redox cycling of Fe-bearing solids and associated contaminants. Thus, it
was important to assess the relative stability of this physical aspect of the HBHA Pond water column during the period
of study. A series of water quality measurements are shown in Figures 9-11 as a function of depth at north, central,
and south sampling locations for three sampling dates. In general, the vertical trends in specific conductance, dis-
solved oxygen, and ORP document relatively uniform stratification (i.e., formation of chemocline) throughout the water
column. Dissolved oxygen concentrations are low (limit of detection), and measured ORP is negative near the sedi-
ment-water interface. In some instances, a turbidity maximum was observed near the chemocline, which was attributed
to iron (hydr)oxides precipitating as a result of ferrous iron oxidation. The chemocline was disrupted in the central and
southern portions of the HBHA Pond during the April 2001 sampling date as indicated by a uniform profile for specific
conductance as a function of depth (Figures 10 and 11). This was due to a significant increase in surface-water influx
15
-------
30000
a-
T3
CD
"E
»- HBHA Pond Oufflow
o Ground-Water Inflow
Surface-Water Inflow
Percent GW
l '—' '—' '
USGS 01102500 Aberjona River at Winchester, MA
100
- 80
-60
-40
-20
- 0
"0
CD
O
3
3
CL
!
Date
*
E
,_- 1.0-
«> 08-
i u-u
1 0.6 :
§
-a 0.4-
c
o
CL 0.2-
m o.o-
/ / x
<« w- "V
1
- -_NM^J
v^
_,
nJ4v-ju___
1
] 11
1 1
C
JUL
j^v__y •
1 //
.._/*
3000
- 2000
- 1000
0
^> J* J»-
C
CO
O —i.
O o
Date
Figure 5 (A) Patterns in surface water and ground-water flow budgets within the HBHA Pond as documented by
Aurilio et al. (1994) and Wick et al. (2000). (B) Concurrent flow variations within the Aberjona River are also
shown for the USGS 01102500 gaging station located approximately 2.5 miles south of the HBHA Pond
(Winchester, MA). (C) Water level variations (blue columns) recorded within the HBHA Pond (this study)
relative to concurrent flow variations observed at the USGS 01102500 gaging station.
(Figure 5C), a fraction of which was contributed by Halls Brook. The chemocline was depressed'to greater depth on
this sampling date in the northern portion of the HBHA Pond (Figure 9). Maintenance of stratification in this portion
of the pond is likely due to its proximity to the primary point of discharge of contaminated ground water with high dis-
solved solids. However, as shown in Figure 12 based on contours of specific conductance within the water column in
the northern and southern portions of the HBHA Pond, the chemocline is re-established throughout the pond to es-
sentially pre-storm conditions within a period of less than a month. Analysis of historical surface water flow records at
the USGS Winchester Monitoring Station indicates that flow events of sufficient magnitude to disrupt the chemocline
may occur at a frequency of about every four years (data not shown). Water quality measurements conducted as part
of this study indicate that any disruption to the chemocline is relatively short-lived, which is consistent with historical
records for the HBHA Pond (Wick et al., 2000).
16
-------
Depth (ft bgs)
10.58 15.58
5/00
8/00
4/01
9/01
1079
1063
398
862
2091
2052
1020
1881
Depth (ft bgs)
12.25 15.00
4/00
5/00
8/00
3/01
9/01
729
1389
1249
1114
1334
TWOS Depth (ft bws)
Date 13.00
Depth (ft bgs)
12.25 17.50
3/00 ND
5/00 ND
8/00 ND
TW02 Depth (ft bws)
Date 14.00
4/00
5/00
8/00
3/01
9/01
TW04 Depth (ft bws)
Date 14.00
TWOS Depth (ft bgs)
Date 8.00 13.00 17.50
3/00
5/00
8/00
3/01
9/01
Depth (ft bws)
9.00
TWOS Depth (ft bws
Date 11.00
Depth (ft bgs)
3.50 8.50
3/00 395
5/00 601
8/00 912
4/01 712
9/01 1214
5/00 ND
8/00 ND
9/01 33
TW11 Depth (ft bgs)
Date 3.50
Figure 6 Aerial distribution and time-dependent variability for total dissolved arsenic concentrations (ppb) detected
in tubing wells established adjacent to the HBHA Pond. Depths are shown in feet below ground surface
(ft bgs) and feet below water surface (ft bws) for tubing wells installed on land and within the HBHA Pond,
respectively. Depths for tubing wells TW01, TW02, TW03, TW04, and TWOS were determined at the time
of installation. NM = not measured, ND = not detected (<33 ppb). Image derived from May 1995 aerial
photograph obtained from MassGIS.
17
-------
TW10 Depth (ft bgs)
TW07 Depth (ft bgs)
Date 12.25 15.00 20.00
TWOS Depth (ft bws)
Date 13.00
TW06 Depth (ft bgs)
Date 12.25 17.50 22.50
3/00 2398
5/00 2211
8/00 1423
4/00 1713 4479 6577
5/00 1372 3388 10340
8/00 1427 4295 16337
TW02 Depth (ft bws)
Date 14.00
4/00
5/00
8/00
3/01
9/01
TWOS Depth (ft bgs)
Date 8.00 1300 17.50
4//00 ND
5/00 ND
8/00 ND
4/01 ND
9/01 32
TWOS Depth (ft bws)
Date 11.00
TW12 Depth (ft bgs)
Date 3.50 8.50
5/00 328
8/00 353
9/01 287
TW11 Depth (ft bgs)
Date 3.50
TW13 Depth (ft bgs)
Date 3.50
Figure 7 Aerial distribution and time-dependent variability for total dissolved zinc concentrations (ppb) detected in
tubing wells established adjacent to the HBHA Pond. Depths are shown in feet below ground surface (ft
bgs) and feet below water surface (ft bws) for tubing wells installed on land and within the HBHA Pond,
respectively. Depths for tubing wells TW01, TW02, TW03, TW04, and TWOS were determined at the time
of installation. NM = not measured, ND = not detected (<14 ppb). Image derived from May 1995 aerial
photograph obtained from MassGIS.
18
-------
TW07 Depth (ft bgs
TWOS Depth (ft bws)
Date 13.00
Depth (ft bgs)
12.25 17.50
TW02 Depth (ft bws)
Date 14.00
4/00
5/00
8/00
3/01
TW04 Depth (ft bws)
Date 14.00
TWOS Depth (ft bgs)
3/00
5/00 1
8/00 NM
3/01 2
Depth (ft bws)
9.00
TWOS Depth (ft bws)
Date 11.00
3/00
5/00
8/00
4/01
4070
1760
NM
2590
Figure 8 Aerial distribution and time-dependent variability for benzene concentrations (ppb) detected in tubing wells
established adjacent to the HBHA Pond. Depths are shown in feet below ground surface (ft bgs) and feet
below water surface (ftbws) for tubing wells installed on land and within the HBHA Pond, respectively. Depths
for tubing wells TW01, TW02, TW03, TW04, and TWOS were determined at the time of installation. NM =
not measured, ND - not detected (<1 ppb). Image derived from May 1995 aerial photograph obtained from
MassGIS.
19
-------
-«-T, °C
0 5 10152025
NORTH
A—Sp. Cnd.nS/cm
0 4000 8000
0 _|
50-
100-
150-
200-
250-
300-
350
I I I
t
H I
(, I
llJs
r^' \
A • D
I ' I ' I ' 'I
02468
-c^pH
-A- DO, mg/L
0 5 1015202
0 ! '
50 -
100-
150-
200-
250-
300-
f T?
/ n
r /I
i i i ' ' i
02468
0 5 1015202
0 ' '
50 -
100-
150-
200-
250-
300-
350
T A
o /A
" / T
(I & U
o I i
y j
1 1 1 1 1 1
02468
n
50-
100-
150-
200-
250-
300-
5
0 -,
50-
100-
150-
200-
250-
300-
5
0
50-
100-
150-
200-
250-
300-
350
I
/
t
/"
^
I i | , | i ,
8 8 ° 8 |
-*- ORP, mV
50-
100-
150-
200-
250-
300-
5 o
J o
, I , I
0
L
)
1
I
1
>
^
0
, 1 , 1 ,
50 100 150
Turbidity, NTU
0 4000 8000
n ,1,1
I
<
1 ' 1 ' 1 ' '
| 8 ° 8 S
50-
100-
150-
200-
250-
300-
*
' )
V -••-
V ...--o
1 o •
} 0 50 100 150
c
n
T '
i r
/
/
I
i ' i ' i ' i '
8 8 ° 8 i
50-
100-
150-
200-
250-
300-
350
) 4000 8000
, 1 ,
A
^
3 i
>,
iV
^
0 50 100 1
50
April 2000
August 2000
April 2001
Figure 9 Vertical patterns in water quality within the northern part of the HBHA Pond on sampling dates in April 2000,
August2000, and April 2001. Symbols: T= temperature, DO = dissolved oxygen, ORP = oxidation-reduction
potential, and Sp. Cnd. = specific conductance.
20
-------
CENTRAL
(
o
50-
100-
150-
200-
250-
300-
350
3 5 1015202
T t
i > |A
< » A
{T r
n(
i ' i ' i ' i ' i
02468
-o-pH
-A- DO, mg/L
0 5 1015202
0 I I I I
50-
100-
150-
200-
250-
300-
350
(
0 _|
50-
100-
150-
200-
250-
300-
350
t T
\ n
\i\
f 4 \
1 i b
i ' i ' ' i ' i
02468
) 5 1015202
i , , 1 i 1 ,
T ^?
<> 11
0 II
" I \
0 I fa
i I I
1 ' 1 ' 1 ' 1 ' 1
02468
O
o
50-
100-
150-
200-
250-
300-
350
5
0 -,
50-
100-
150-
200-
250-
300-
350
5
50-
100-
150-
200-
250-
300-
-r
i •
ir
ir
^
1 ' 1 ' 1 ' 1 '
| 8 ° 8 I
-tt- ORP, mV
c
o J
50-
) 4000 8000
, I ,
r
100 -3)
150 4
200-
250-
300-
350
» :
¥
, i i i .
) 50 100 150
> Turbidity, NTU
0 4000 8000
n ,1,1
1 ' 1 ' 1 ' 1 '
| 8 ° 8 \
50-
100-
150-
200-
250-
300-
350
*
4
V-
\ ""•••' °
*
0 50 100 150
si
c
n
\
f
/
I
I
1 ' 1 ' 1 ' 1 '
8 8 ° 8 £
50-
100-
150-
200-
250-
300-
350
s c
) 4000 8000
, I ,
+
jk
j.
i-
^
^
1 I ' I '
) 50 100 1
50
April 2000
August 2000
April 2001
Figure 10 Vertical patterns in water quality within the central part of the HBHA Pond on sampling dates in April 2000,
August2000, and April 2001. Symbols:J= temperature, DO = dissolvedoxygen, ORP = oxidation-reduction
potential, and Sp. Cnd. = specific conductance.
21
-------
-•— T, °C —^~ Sp. Cnd-pS/cm
0 5 10152025 OUUin Q 4QOO 800Q
o
50-
100-
150-
200-
250-
300-
350
1 1 1
K
0
— D-
— A-
0 "
50 -
100-
150-
200-
250-
300-
T^n
i
>
i
/
50-
100-
150-
200-
250-
300-
•iKn
i 1 i
*
»
i
<»
\
\
0 A
2468 ooooo 0 50 100 150
- DO, mg/L
-*- ORP, mV
0 5 10152025 0 4000 8000
Q I i i I i I i I i n n ilil
50-
100-
150-
200-
250-
300-
350
n
0
50 -
100-
150-
200-
250-
300-
•^e;n
*
/
i
f
^
50 -
100-
150-
200-
250-
300-
icin
f
4
%.
r°
ol
I , | , | , | — | , | , | , | , — i | i | i
2468 oooS2 050 100 150
0 5 10152025 0 4000 8000
Qiililili n n ili
50 -
100-
150-
200-
250-
300^
350
(
i
i
i
i
i
1
0
t An
i A n
' 1 T
I A D
' T T
i A n
50-
100-
150-
200-
250-
300-
i«;n
r
i
t
y
I i | i | i I — - | i | , i | i
2468 ^^^^^
50-
100-
150-
200-
250-
300-
-350
C
A
'
>
L
4 L
A
1 1 ' 1 '
) 50 100 1
50
April 2000
August 2000
April 2001
Figure 11 Vertical patterns in water quality within the southern part of the HBHA Pond on sampling dates in April 2000,
August2000, and April 2001. Symbols:T= temperature, DO = dissolved oxygen, ORP = oxidation-reduction
potential, and Sp. Cnd. = specific conductance.
22
-------
ro
w
kTWOT
North V^
Conductivity \ ,„ , ,xx
Transect \ \ \2 v
South
Conductivity vls\\>r
Transect
D North Conductivity Transect
April 2001
May 2001
April 2001
May 2001
South Conductivity Transect
Figure 12 Evidence of water column mixing within the HBHA Fondas a result of a large surface water flow event. (A) Location of conductivity transects
collected within the north and south portions of the HBHA Pond. Contour lines indicate depth to sediment (meters). (B) Specific conduc-
tance (\iS crrr1) distribution immediately following (April 2001) and approximately one month (May 2001) after the large surface water flow
event at the end of March 2001 (See Winchester flow data in Figure 5C.).
-------
Media-Specific Contaminant Distributions
Ground Water
The general distributions of major ion and contaminant metal concentrations in ground water are depicted in Figures 13
and 14. Patterns in the concentrations of alkalinity, sulfate, and ammonia-nitrogen between well pairs that span the
GW/SW transition zone point to the stimulation of microbial activity within this region of the aquifer (specifically well
pairs TW07-TW02 and TW08-TW01). There is a systematic decrease in the concentration of sulfate within ground
water in moving from a location immediately upgradient to underneath the HBHA Pond (Figure 13B). This decrease
is accompanied by concurrent increases in alkalinity and ammonia-nitrogen for these well pairs (Figure 13A and 13C).
These trends appear to be confined to a region along the northeastern portion of the HBHA Pond. This region of the
pond also corresponds to the location of the BTEX plume (Figure 13D).
Patterns in inorganic contaminant concentrations within the shallow aquifer appear more complex. The highest concen-
trations of zinc observed within the upgradient aquifer occur along the northern margin of the HBHA Pond (Figure 14).
For arsenic, hot spots within ground water occur along the northern margin and also the north-eastern side of the HBHA
Pond in a similar region as the BTEX plume. The co-occurrence of elevated arsenic, zinc, and sulfate along the northern
portion of the HBHA Pond points to upgradient oxidative weathering of sulfides as a possible source. This is consistent
with historical records that document the operation of a sulfuric acid production facility within the Industri-Plex Superfund
Site boundary (Aurilio et al., 1995). A common process used in the production of sulfuric acid involved burning pyrites
to generate sulfur dioxide gas. The waste products from this process were disposed on site and pose a long-term
source of iron, sulfate, and associated inorganic contaminants in shallow ground water (Spanish pyrites with up to 1 %
arsenic by weight were often utilized). The region of elevated arsenic along the north-eastern side of the HBHA Pond
may be due to an additional buried source or the development of chemical conditions conducive to maintaining arsenic
in a dissolved state (e.g., iron reduction). There is an apparent source of sulfate at depth at location TW08-3 that may
contribute to the elevated concentration of this dissolved component at location TW01. However, sulfate concentrations
are significantly lower at shallow depths within this part of the aquifer (See Site 5, TW08-1, and Site 6.).
In general, the highest concentrations of inorganic contaminants discharging from ground water appear to be confined
to the north-northeastern shoreline of the HBHA Pond. While inorganic contaminant concentrations appear to be lower
in the region of the BTEX plume, it is clear that this portion of the aquifer is exerting an influence on the chemistry within
the northern region of the HBHA Pond. The extent of this influence will be examined in greater detail in the following
section.
Surface Water
Patterns in surface water chemistry have been reviewed by Ford (2005) for the HBHA Pond. Briefly, the chemistry
within the pond is controlled by the relative proportions of surface water and ground-water flows. Due to the physical
configuration of the HBHA Pond and the differences in dissolved solids concentrations between the main surface water
inflow (Halls Brook) and the ground-water plume, the pond generally remains stratified. This stratification, or chemo-
cline, acts to limit vertical diffusion of dissolved constituents entering the system via ground-water discharge at depth.
The barrier to vertical transport within the water column plays an important role in maintaining the redox cycling of iron
and co-precipitated arsenic (and metals). In general, ferrous iron entering the system via ground-water discharge is
precipitated within the oxic (shallow) portion of the water column, leading to sequestration of dissolved arsenic. The
iron (hydr)oxides (poorly crystalline ferrihydrite, Ford et al., 2005) settle to the sediment layer and are subsequently dis-
solved in the reducing conditions that are established by the microbial degradation of natural and contaminant organic
compounds that accumulate within the HBHA Pond. The mineralogical characteristics of the iron (hydr)oxide precipitates
and their rate of formation (data not shown) are comparable to those of natural, poorly crystalline ferrihydrite formed
in terrestrial systems (Schwertmann, and Fischer, 1973; Davison and Seed, 1983; Schwertmann and Murad, 1988;
Perret et al., 2000)
However, this near-permanent stratification can be interrupted by a sufficient surface water flow event. Such an event
occurred during the study period (end of March 2001) resulting in complete mixing within the water column at the central
and southern portions of the HBHA Pond (i.e., below Halls Brook discharge). The northern portion of the pond remained
partially stratified (chemocline suppressed in depth), since this is the primary location of discharge for the contaminant
plume. As shown by Ford (2005), the stratification is re-established along with the steady-state redox chemistry that
controls iron and arsenic cycling. Based on the ground-water characterization program during this study, it became
apparent that two plumes discharge into the north-northeastern portion of the HBHA Pond. The northern plume is the
primary source of arsenic and iron into the system. In contrast, the primary input of BTEX compounds is located along
the northeastern margin of the pond. The input of sulfate is more widely distributed across these two areas of plume
discharge (Figures 11 and 12). Assessment of the temporal variations of the HBHA Pond water chemistry as a func-
tion of depth confirms that two separate plumes intersect to discharge into the north-northeastern portion of the HBHA
24
-------
B
I
4000-6600
2000-4000
800-2000
0-800
Alkalinity
(mg L'1 CaCO3)
TWOS
Benzene
-------
Site 3
1090nS/cm
440 ug/L As
2040 ug/L Zn
10 mg/L Fe
91 mg/L S
Site 4
410 uS/cm
3 ug/L As
1970 ug/L Zn
0 2 mg/L Fe
21 mg,LS
TW02
7160uS/cm
100 ug/L As
20 ugiL Zn
0 2 mg/L Fe
24 mg/L S
Sitel
2630 uSicm
60 ug/L As
56 ug/L Zn
0 4 mg/L Fe
90 mg/L S
1290uS/cm
1360 ug/LAs
69 ug/LZn
28 mg/L Fe
117 mg/LS
TVTO7-1
1245 uS/cm
860 ug/LAs
660 ug/L Zn
24 mg/L Fe
1 30 mg/L S
TW07-2
1016uS/cm
1300 ug/L As
104 ug/LZn
23 mg/L Fe
88 mg/L S
TW07-3
2180uS/cm
169 ug/L As
76 M9/L Zn
68 mg/L Fe
325 mg/L S
TW01
14590uS/cm
700 ug/L As
17 ug/L Zn
6 mg/L Fe
280 mgiL S
Sites
810 uS/cm
650 ug/L As
49 ug/L Zn
28 mg/L Fe
8 mg/L S
TW08-1
7190uS/cm
500 ^g/L As
80 ug/L Zn
2 mg/L Fe
TW08-3
12730uS
-------
4/5/01
4/10/01
5/14/01
9/14/01
As, mg/L
0246
0-
100
200-
I
5 300-
Q.
0>
Q
400
500-
0
'Ml
1 2
(x103)
Alk, mg CaCO3/L
0369
I . I
Zn, mg/L
0.0 0.5 1.0
NH3-N, mg/L
(x103)
I ' I ' I
0 100200
Fe, mg/L
0 1 2
SO4> mg/L
(x103)
0 600 1200
Na, mg/L
Figure 15 Temporal trends in water chemistry for the NML sampling station and adjacent ground-water monitoring
locations (TWO? and TW02). NML data are shown as open black symbols connected with lines for multiple
sampling dates during the period of April 5, 2001, to September 14, 2001; adjacent ground-water data for
locations TW07-1, TW07-2, TW07-3, and TW02 are shown as open red symbols for multiple sampling dates
during the period of April 6, 2000, to September 11, 2001. Arsenic data for the NML sampling station are
from Ford (2005).
Sediments
Spatially-resolved concentration distributions for arsenic, lead, zinc, chromium, iron, sulfur, and organic carbon in the
sediments of the HBHA Pond are shown in Figures 16-19. Raw concentration data are also tabulated in Appendix B.
The highest concentrations of arsenic, chromium, and zinc were observed near the location of plume discharge and
further downgradient within the HBHA Pond. In contrast, the highest lead concentrations are proximate to the location
of the Atlantic Avenue Drainway discharge (Figure 2), which may be attributed to erosion or runoff from areas within
the Industri-Plex Superfund Site. Because sediment concentration maxima for arsenic, chromium, and zinc coincide
with regions where contaminated ground water discharges into the pond, it is reasonable to suspect that ground water
transport coupled with metal deposition near the ground water-surface water interface is a primary mechanism con-
trolling metal concentrations in the pond sediments. This pattern is particularly evident for sulfur, arsenic, and zinc
where the highest concentrations of these elements appear to be associated with the area of plume discharge in the
north-northeastern region of the HBHA Pond. Analysis of sediment cores showed that the highest concentrations of
arsenic, lead, and zinc were associated with sediments that have been deposited following construction of the HBHA
Pond (Appendix B). The underlying sandy sediments possess concentrations of carbon, iron, and sulfur that are lower
by an order-of-magnitude or more.
A comparatively more even concentration distribution is observed for organic carbon (Figure 19). This pattern is likely the
result of multiple sources of carbon to the pond, including ground water discharge, carbon input from Halls Brook, and
seasonal deposition of plant materials and woody debris. High concentrations of organic carbon appear to be clustered
in areas of low relief at the bottom of the pond suggesting that sedimentary deposition processes play an important role
in regulating concentrations of organic carbon. With the exception of the locations of visible ground-water seeps along
the north-northeastern margin of the HBHA Pond, iron is also distributed uniformly in shallow sediments throughout
the pond (Figure 18, typically 8-12 wt%). The lack of iron concentration gradients in the pond sediments may be tied
27
-------
to bottom-water redox conditions that are characteristically iron-reducing. That is, iron is highly soluble and mobile in
deep bottom-waters of the HBHA Pond; consequently, there are no focused loci of iron accumulation. The highest sedi-
ment concentrations of iron are found in red-colored, shallow sediments (<1 m) over which waters are typically oxidizing
(>2 mg/L dissolved O2). Under oxidizing conditions, iron is highly insoluble, and iron would be expected to accumulate
in the surface sediments, especially in areas that receive discharge of reducing, ferrous iron-laden water.
^r^.,
• s «t&.
\0«^\
^V, \s , \
~"~<^. \ '*> 9 \
Xx \ M \ 50m
\-=^ 3« \ "^~
N
\ 9 \
\x » \
As, M9 9 1 ^ \
3 0-500 \
• 500-1000 V ^0 )
• 1000-1500 X^X
• >1500 x\^
\s
30-
25-
&20-
I1"
10-
5-
---
0 500 1000 1500 2000 2500
As.pg g1
Figure 16 The distribution of arsenic (As) in sediments collected from the HBHA Pond. The top panel shows the
location of sediment samples for which compositional data are reported. The aerial distribution of arsenic
is based on grab samples and the top 2-cm interval of core samples; contour lines delineate depth to sedi-
ment (meters). The frequency plot includes all grab samples and all sediment core depth intervals; data are
compiled in Appendix B.
28
-------
ng g
° 0-200
• 200-400
• 400-600
• >600
25
20
15-
10-
5-
200 400 600 800 1000 1200 1400 1600
M9
0-2000
2000-4000
4000-10000
>1 wt.%
2000 4000 6000 8000 1000012000140001600018000
Figure 17 The distribution of lead (Pb) and zinc (Zn) in sediments collected from the HBHA Pond. The aerial distribu-
tions of lead and zinc are based on grab samples and the top 2-cm interval of core samples; contour lines
delineate depth to sediment (meters). The frequency plot includes all grab samples and all sediment core
depth intervals; data are compiled in Appendix B.
29
-------
Cr, MQ g
° 0-200
• 200-400
• 400-600
• >600
25
20-
15-
10-
5-
200 400
600
800 1000
30
25-
20-
15-
10-
5-
-i
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
Fe, Wt.%
Figure 18 The distribution of chromium (Cr) and iron (Fe) in sediments collected from the HBHA Pond. The aerial
distributions of chromium and iron are based on grab samples and the top 2-cm interval of core samples;
contour lines delineate depth to sediment (meters). The frequency plot includes all grab samples and all
sediment core depth intervals; data are compiled in Appendix B.
30
-------
60-
50-
40-
30-
20-
10-
0 2 4 6 8 10 12 14 16
S, Wt %
30
25
20-
15-
10-
5-
10
OC, wt.%
20
Figure 19 The distribution of sulfur (S) and organic carbon (OC) in sediments collected from the HBHA Pond. The
aerial distributions of sulfur and organic carbon are based on grab samples and the top 2-cm interval of core
samples; contour lines delineate depth to sediment (meters). The frequency plot includes all grab samples
and all sediment core depth intervals; data are compiled in Appendix B.
31
-------
Chemical Speciation of Metals in Sediments
Sediment Extractions
The objective of the chemical extraction tests was to develop a sense of the degree to which arsenic, lead, and zinc
were partitioned to sediments in a relatively labile form. It was assumed that the fraction of contaminants released
by mild extractant solutions (i.e., MgCI2, phosphate, or Na2CO3) or extractant solutions designed to solubilized poorly
crystalline iron oxides or sulfides would represent the most labile pool. From this perspective, MgCI2 was chosen to
extract lead and zinc present in a readily exchangeable fraction (Tessier et al., 1979); whereas, solutions of phosphate
or Na2CO3 were chosen to extract arsenic adsorbed on mineral surfaces (Jackson and Miller, 2000). Solutions of
ascorbate or hydrochloric acid were chosen to dissolve poorly crystalline iron oxides and sulfides (Kostka and Luther,
1994; Ford et al., 1999). Extraction with HCI was also studied due to its use for the acid volatile sulfide-simultaneously
extracted metals (AVS-SEM) procedure commonly employed to determine metal speciation and potential bioavailability
in sediments (Wilkin and Ford, 2002).
An implicit assumption in the following analysis was the existence of a direct association between extracted iron oxides
and sulfides and co-extracted arsenic, lead, and zinc. The reliability of this assumption was tested via independent
assessment of the solid phase speciation of these elements using X-ray absorption spectroscopy (discussed in a later
section). It should also be noted that the various extraction solutions were applied in parallel. Thus, contaminants
solubilized in either the ascorbate or HCI extractions could potentially have been derived from a variety of partitioning
environments such as ion exchange sites and/or more stable environments such as the internal structures of coprecipi-
tates (e.g., zinc incorporated in the structure of an iron oxide) or pure phase precipitates (e.g., zinc sulfide). The slight
differences in pH or the concentration of the primary component in the extraction solutions employed for surficial (oxic)
versus buried (suboxic) sediments was due to different personnel handling the two sediment classes. It was assumed
that these differences were sufficiently insignificant to prevent comparison of extraction trends for the two sediment
classes. It should also be noted that selectivity and completeness of the HCI and ascorbate extractions were not cali-
brated against sediments spiked with reference minerals. Rather, reliance was placed on published data where these
procedures have been employed.
Surficial Sediments
Sediment samples were collected from the north-northwestern margin of the HBHA Pond in the vicinity of visible ground-
water seeps. The sediments collected from this location were reddish-orange and consistent with the precipitation of
poorly crystalline iron oxides due to the rapid oxidation and precipitation of ferrous iron derived from ground-water dis-
charge. A series of chemical extraction tests were carried out on 5 sediment samples from this location (<2 mm size
fraction). The solutions used to extract these sediments varied in their aggressiveness for the extraction of contaminants
and dissolution of iron oxides or iron sulfides. Results of the chemical extraction tests are presented in Table 1 as
percent recoveries of the total amount of arsenic, lead, zinc, and iron in the sediment samples. Total concentrations of
these metals in the test sediments are presented in Appendix B. The chemical extraction tests included a spectrum of
reagents from relatively mild 1 M magnesium chloride to more aggressive 0.5 M hydrochloric acid.
Test results indicated that arsenic, lead, and zinc were not readily solubilized from the sediments using a 1 M solution of
MgCI2. Typically, <1% of the total amount of sediment arsenic, lead, or zinc was leached using the magnesium chloride
solution (Table 1); although, a greater fraction of zinc was extracted by this reagent (mean 15.0 %, max. 25.9%). This
easily extracted fraction of zinc may be attributed to magnesium competition with zinc for ion exchange sites. Solutions
containing phosphate were also examined to assess the degree to which arsenic could be desorbed from the sediment
without significant solid phase dissolution. The fraction of extractable arsenic increased to as much as 15.9% in the
presence of phosphate at pH 4, but less arsenic was extracted in the presence of phosphate at pH 7. The increase in
extractable arsenic relative to the MgCI2 solution can be attributed to the strong competition from phosphate for adsorption
sites on sediment minerals (Jackson and Miller, 2000). A significant fraction of Zn (up to 27.4%) was also extracted in
the presence of the pH 4 phosphate solution. This result may be attributed to competition with sodium for ion exchange
sites and/or desorption of zinc from surface complexation sites due to the low pH (McBride, 1994). An insignificant
fraction of lead was extracted from the oxic sediments in the presence of the MgCI2 or phosphate solutions. In general,
the highest fractions of arsenic, lead, and zinc were extracted in parallel with significant dissolution of iron-bearing
minerals in these sediments. Mineralogical characterization of the clay-sized fraction from these sediments indicated
that ferrihydrite was a dominant iron oxide phase. Ferrihydrite is readily solubilized via reductive- or acid-dissolution
processes in ascorbate or HCI solutions (Ford et al., 1999; Larsen and Postma, 2001). The lower fraction of extracted
lead for the ascorbate versus the HCI extraction (and zinc to a lesser extent) may be due, in part, to precipitation of a
carbonate phase during the procedure.
32
-------
Buried Sediments
A series of selected chemical extraction tests were carried out on 14 buried sediment samples collected from the bot-
tom of the HBHA Pond. These tests were conducted to provide insight into contaminant speciation and the conditions
that could favor metal release from the sediments. Results of the chemical extraction tests are presented in Table 2 as
percent recoveries of the total amount of arsenic, lead, zinc, and iron in the sediment samples. Total concentrations
of these metals in the test sediments are presented in Appendix B. The chemical extraction tests included a spectrum
of reagents from relatively mild 0.5 M magnesium chloride to more aggressive 1 M hydrochloric acid. Test results in-
dicated that arsenic and other selected metals analyzed were not readily solubilized from the sediments using 0.5 M
magnesium chloride; typically <6% of the total amount of sediment arsenic or metal was leached using the magnesium
chloride solution (Table 2). These results indicate that arsenic, lead, zinc, and iron are tightly bound to the solid matrix
of the sediments. As expected, alkaline extraction solutions (0.1 M sodium carbonate) were ineffective in solubilizing
lead, zinc, and iron; whereas, arsenic was partially recovered using the sodium carbonate solution (Table 2). Similar to
the surficial sediments, significant fractions of arsenic and iron were extracted by the ascorbate solution, while extracted
fractions of lead and zinc were below detection in this solution. Significant fractions of arsenic, lead, zinc, and iron were
extracted in HCI for "unoxidized" and "oxidized" buried sediments. These results suggest the importance of contaminant
partitioning to poorly crystalline iron (hydr)oxides and/or acid volatile sulfides. Mineralogical characterization of surficial
sediments (discussed below) and a representative sample of buried sediments from the HBHA Pond (Wilkin and Ford,
2002) confirm that both mineral types could potentially be present in sediments deposited within the HBHA Pond.
Apparent differences for the extracted fractions of arsenic and lead between "oxidized" and "unoxidized" sediments
treated with HCI suggest a potential artifact associated with the predominant mineralogy in the treated sediment. As
shown by Wilkin and Ford (2002), acid volatile sulfides (in particular, iron monosulfides) may be present as a significant
component of buried sediments. Typically, one would expect formation of poorly crystalline ferrihydrite as the endproduct
of rapid oxidation of iron monosulfides. The high extractable fractions of arsenic, lead, zinc, and iron in these "oxidized"
sediments are consistent with an association with either a 'native' poorly crystalline iron (hydr)oxide as well as a similar
phase produced upon oxidation of iron monosulfide. Thus, the observed artifact can be attributed to some chemical
process that limits the solubility of arsenic and lead in HCI for "unoxidized" sediments.
The ratios of extracted arsenic, lead, zinc, and iron for "unoxidized" and "oxidized" buried sediments are plotted in Figure 20
as a function of total sulfur content. In the absence of analytical artifacts, one would anticipate a value close to unity
for this ratio in sediments dominated by acid volatile sulfides ("unoxidized") and/or poorly crystalline iron (hydr)oxides
("oxidized"), since both mineral phases are extracted in hydrochloric acid. Results plotted for arsenic indicate that this
metal is less extractable for the "unoxidized" sediment, i.e., extracted ratio « unity. However, as shown by Wilkin and
Ford (2002) this behavior can be attributed to an analytical artifact of the acid extraction commonly employed in the
acid-volatile sulfide-simultaneously extracted metals (AVS-SEM) procedure rather than to any inherent difference in
extractability. Specifically, hydrogen sulfide that is released during acid extraction reacts with co-extracted arsenic to
form an orpiment-like phase at acidic pH, which is confirmed through calculation of orpiment solubility at the extraction
pH (Wilkin and Ford, 2002). It appears there may be a related concern for Pb, which again displayed a decreased
extractability for the "unoxidized" buried sediment with an increase in sulfide content. The extractability ratios clustered
around unity for Fe and Zn in these sediment samples are consistent with the high solubility of iron and zinc mono-
sulfides at acid pH. Extract ratios that exceed unity value at low sulfur concentrations may result from the conditions
used to oxidize sediments. Exposure to air in a drying oven at 60°C was employed to oxidize sulfides. The elevated
drying temperature may have resulted in the production of an iron oxide phase(s) that was resistant to extraction using
hydrochloric acid (e.g., Stanjek and Weidler, 1992). This would result in extraction ratios greater than unity, since iron
and zinc associated with an acid-resistant iron oxide phase would be less extractable. This process would be less likely
for sediments with exceptionally high sulfide concentrations, since the formation of a more extractable ferric sulfate
phase(s) would be favored.
Due to the anticipated importance of poorly crystalline iron (hydr)oxides and iron sulfides in HBHA Pond sediments,
ascorbate and hydrochloric acid extraction solutions were employed to assist in differentiating the relative proportion of
these minerals. Peltier et al. (2005) have shown that extraction tests designed to target easily reducible iron (hydr)oxides
via reductive dissolution at acidic pH also extract a significant fraction of acid volatile sulfides. However, this potential
artifact was avoided by employing reductive dissolution with ascorbate buffered at a near-neutral pH (Kostka and Luther,
1994). As a point of reference, the results of ascorbate extractions of both surficial and "unoxidized" buried sediments
for arsenic and iron are shown in Figure 21. Except for sample WI01, arsenic was completely extracted in the pres-
ence of ascorbate for all surficial sediments, which corresponded with an extractable iron fraction of more than 80%.
The lower amount of extractable arsenic and iron in sediment WI01 may be attributed to the crystallinity of the domi-
nant iron oxide in this sample. As shown in Figure 22, hematite is the dominant iron oxide in the <2 urn size fraction
from this sediment. X-ray diffraction data for the <2 urn size fraction of sediment samples WI01-NEP and WI02-NEP
show an increased relative proportion of poorly crystalline 6-line ferrihydrite, consistent with the observed change in
33
-------
the fraction of ascorbate-extractable iron. As shown by Kostka and Luther (1994), hematite is dissolved to a lesser
extent than ferrihydrite by ascorbate under the conditions employed. The increase in extractable iron for samples WI01,
WI01-NEP, and WI02-NEP is consistent with the increased relative proportion of a poorly crystalline 6-line ferrihydrite
in this series of samples. The low extractability of arsenic concurrent with the significant fraction of hematite in sample
WI01 suggests the incorporation of arsenic in hematite during nucleation and growth of this mineral as suggested by
experimental studies by Ford (2002). However, this observation remains uncertain without a specific knowledge of the
diagenetic history leading to hematite formation in this sediment.
The fraction of ascorbate-extractable arsenic varied significantly for the set of buried sediments that was analyzed.
Based on the results shown in Figure 21 A, there were two distinct sediment groups in which arsenic was either >90%
extractable (6 samples) or <40% extractable (6 samples) with a set of transition sediments between these two limits
(5 samples). Examination of the fraction of ascorbate-extractable arsenic as a function of the total sulfur content of
buried sediments indicates a general correlation between these two parameters (Figure 21C). It should be noted that
a similar correlation (although weaker) apparently exists between ascorbate-extractable iron and total sulfur content for
buried sediments (Figure 21D). Insignificant extraction of iron sulfides is anticipated for ascorbate extractions (Kostka
and Luther, 1994), suggesting that arsenic may be partitioned to iron sulfides for buried sediments with low ascorbate-
extractable arsenic. Sediments in which the iron mineralogy is dominated by iron oxides are anticipated to have similar
extractable iron fractions in the presence of ascorbate and HCI. This hypothesis is supported by results presented in
Figure 21E. All surficial sediments have an ascorbaterHCI extractable ratio for iron >1 compared to buried sediments
with ratios of <0.5. Ratios greater than unity for surficial sediments can be attributed to extraction, in the presence of
ascorbate, of iron oxides that are more crystalline than ferrihydrite (Dos Santos Afonso et al., 1990; Larsen and Postma,
2001). These results point toward the presence of acid volatile iron sulfides in all of the buried sediments. In general,
it appears that the iron mineralogy for buried sediments is increasingly dominated by acid volatile iron sulfides with
increasing sulfur content.
Comparison to observed trends in apparent lead and zinc partitioning in surficial and buried sediments suggests dif-
ferences in the partitioning environment of these contaminants relative to arsenic (Tables 1 and 2). The ascorbate-
extractable fractions of lead and zinc were below analytical detection for "unoxidized" buried sediments (Table 2). The
observation that a majority of lead and zinc (more than 80% and 90% of lead and zinc, respectively) were extracted
in "oxidized" buried sediments suggests that a significant fraction of these contaminants are partitioned to acid volatile
sulfides. As for the surficial sediments, the presence of a high concentration of bicarbonate may cause precipitation
of lead/zinc carbonates during the ascorbate extraction. In addition, re-adsorption of lead/zinc to the surface of un-
extracted minerals may also have resulted due to the absence of citrate in the procedure applied for buried sediments.
Citrate would act as a complexant to help maintain extracted metals in solution. Overall, the results from sediment
extractions employing ascorbate and HCI were less definitive for delineation of lead and zinc speciation. This limitation
was addressed for a limited set of sediment samples using X-ray absorption spectroscopy to directly determine the
in-situ speciation of arsenic, lead, and zinc.
34
-------
Table 1 Percent of Total Element Released by Selected Wet Chemical Extraction Tests. Range and Mean Value
are Reported for Replicate Analyses of 5 Sediment Samples Collected from the North-Northeastern
Margin of the HBHA Pond
Solution
1 M MgCI2
0.005 M NaH2PO4
0.005 M NaH2PO4
Ascorbate
0.2 M Na-citrate
0.6 M NaHCO3
0.4 M Ascorbic acid
0.5 M HCI
As
<1%
mean
0.2%
8.3 to
15.9%
mean
1 1 .6%
2.2 to
5.1%
mean
3.3%
39.2 to
127.0%
mean
101.8%
16.7 to
66.9%
mean
32.0%
Pb
0.2 to 1 .0%
mean
0.5%
mean
mean
t
4.5 to
33.2%
mean
17.6%
17.2 to
95.7%
mean
61 .0%
Zn
7 7 to
25.9%
mean
15.0%
5.8 to
27.4%
mean
16.2%
mean
0.2%
20.0 to
80.3%
mean
57.6%
27.1 to
95.3%
mean
71 .4%
Fe
<1%
mean
0.01%
mean
0.3%
mean
0.04%
27.6 to
115.9%
mean
88.0%
16.4 to
97.0%
mean
70.2%
Association
Water-soluble; labile; easily de-
sorbed (pH 7)
Water-soluble; strongly adsorbed
(PH4)
Water-soluble; strongly adsorbed
(pHT)
Soluble with a mild reductant
(pH8)
Acid-soluble
(pH<2)
Notes: Samples used in extraction tests: I/WO 7,
in the extraction solution.
WI01-NEP, WI02, WI02-NEP, WI04. t indicates element was undetected or below quantitative limits
35
-------
Table 2 Percent of Total Element Released by Selected Wet Chemical Extraction Tests. Range and Mean Value
are Reported for 14 Sediment Samples Collected from the HBHA Pond
Solution
0.5 M MgCI2
0.1 M Na2CO3
0.02 M Ascorbic acid 0.6
M NaHCO
1 M HCI
"oxidized" sediment
1 MHCI
"unoxidized" sediment
As
1 to
26.5%
mean
5.8%
10.9 to
73.5%
mean
25.9%
16.1 to
108%
mean
65.7%
71 to
106%
mean
91 .7%
<1 to
20.9%
mean
13.2%
Pb
<1%
mean
t
<1%
mean
t
<1%
mean
t
80.0 to
113.1%
mean
99.1%
3.6 to
105.8%
mean
63.5%
Zn
0.1 to
1 .9%
mean
0.6%
<1%
mean
t
<1%
mean
t
90.3 to
109.4%
mean
100.8%
62.2 to
105%
mean
98.2%
Fe
0.3 to
1 1 .0%
mean
4.3%
<1%
mean
0.3%
6.1 to
51 .0%
mean
23.2%
56.8 to
107%
mean
86.7%
83.0 to
107%
mean
98.5%
Association
Water-soluble; labile; easily de-
sorbed (pH 6.8)
Base-soluble; complexed with
organic carbon
(pH 10.9)
Soluble with a mild reductant (pH
7.0)
Acid-soluble; leaching character-
istic of air-exposed sediment at
pH<2
Acid-soluble; leaching characteris-
tic of unaltered sediment at pH<2
Notes: Samples used in extraction tests: SC0401-1, SC0401-2, SC0401-3, SC0401-5, SC0401-6, SC0401-7, NC0901-1, NC0901-2, NC0901-3,
NC0901-4a, NC0901-4b, NC0901-4c, NC0401-1 (0-2cm), NC0401-1 (2-4cm), NC0401-6(0-2cm), NC0401-6(2-4cm). "Oxidized"sediments extracted
in HCI were exposed to air to oxidize sulfide to sulfate prior to extraction, while "unoxidized" sediments were processed in an inert atmosphere to
prevent oxidation of sulfides. t indicates element was undetected or below quantitative limits in the extraction solution.
36
-------
o ^
II
If
££
m^
UJ to
"§=0
o 'x
2 g
X 3
UJ
1.5-
-
1 n
-
0.5-
0.0-
1.5-
1 n
0.5-
0.0-
(
m " (A)
?v?^ Q
^* QD
O
O
i i i I i i | i | i | i |
(B)
**\\
^& *£? A A
i i i i i i | i | i i |
) 2 4 6 8 10 12 14
• Fe
O Zn
A Pb
* As
Total S, wt%
Figure 20 The ratio of As, Fe, Pb, and Zn extracted by HCI for unoxidized and oxidized sediments collected from suboxic
zones within the HBHA Pond.
WI01-NEP ,~,WI02-NEP
S
S
120H ® !'xJ'' «£> w
100-|®i A* WI04
80^
60-
40-
20-
0-
120 J
100-
80-
60-
40-
20-
0 -
C0901-1 *
WI01 **
>1) A
^-"^^ NC0901-4C
© * A
As
1 i ' i ' i ' i ' i ' i
Fe I J I, (B)
I t
,
T 4
•
•• ••
i A (°)
A
A A
2t A
A
4 * A
I ' I ' I ' I ' I ' I ' I ' I
^
•m'm
1 • . -
• • • »
- 120
- 100
- 80
- 60
- 40
-20
- 0
p- 120
- 100
- 80
- 60
- 40
- 20
0
0 5 10 15 20 25 30 0 2
Total Fe, wt%
4 6 8 10 12 14
Total S, wt%
o
l£
§1
f !
LL
£..\i —
1.5-
1.0-
0.5-
n n
_ a Oxic (E)
ill • Suboxic
1
1
'•!.
"» . '. * -
°-° i ' i ' i ' i '
0.0
0.5 10 15
S Fe, mole/mole
2.0
Figure 21 (A-D) Comparison of the amount of As and Fe extracted by an ascorbate solution as a function of total Fe
and S content in sediments (oxic = open symbols, suboxic = filled symbols). (E) Changes in the ratio of
ascorbate-to-HCI extractable Fe as a function of the molar ratio of S and Fe in sediments. X-ray diffraction
data are shown below for sediment samples WI01, WI01-NEP, and WI02-NEP (Figure 20). X-ray absorp-
tion spectroscopic data are shown below for arsenic in sediment samples WI01-NEP, WI04, NC0901-1, and
NC0901-4C (Figure 23).
37
-------
sn
CD
•
i
b
**
WI01 28.3%
WI01-NEP83.6%
WI02-NEP107.7%
10 20 30 40 50 60 70 80 90
20, Fe Kcc
Figure 22 X-ray diffraction data showing the relative proportion of hematite and ferrihydrite in the <2 um size fraction
isolated from surficial sediment samples WI01, WI01-NEP, and WI02-NEP collected near the north-north-
western margin of the HBHA Pond. The starred peaks in the pattern for sample WI01 correspond with the
dominant peaks in reference hematite (PDF 33-0664). Mean percent iron extracted by ascorbate in duplicate
samples is shown to the right of the sample labels. The diffraction pattern for sample WI02-NEP is dominated
by a poorly crystalline 6-line ferrihydrite (see peaks marked with arrows).
Element Speciation by X-ray Absorption Spectroscopy
To provide a constraint on the interpretation of the extraction results, the in-situ speciation of arsenic, lead, and zinc
were determined for a representative set of samples using X-ray absorption spectroscopy. The results of contaminant
speciation analyses employing X-ray absorption spectroscopy are documented in Figures 23-25. Specifically, linear
combination fitting (LCF) of the X-ray absorption near edge (XANES) region of the spectra was used to delineate ele-
ment speciation in sediments. Reference spectra collected for model solid phase compounds were used in the fitting
procedure in order to estimate the relative fraction of representative contaminant-mineral associations in each analyzed
sediment sample. In aggregate, these results attest to the importance of the redox gradients within the HBHA Pond
and attest to the apparent degree of disequilibrium that exists within the sediments.
The results from a selection of samples characterized for arsenic speciation are shown in Figure 23, and the aerial
distribution of arsenic phase associations within the HBHA Pond are shown in Figure 26. Analysis of these samples
indicates that the fate of arsenic is directly tied to the iron minerals that dominate within oxic and suboxic portions of
the HBHA Pond. Arsenic speciation is dominated by adsorption/coprecipitation of As(lll) or As(V) with iron (hydr)oxides
that precipitate in shallow ground-water seeps near the margin of the HBHA Pond (WI01-NEP, WI04). The speciation
of arsenic transitions to As(lll) partitioned to iron monosulfides within the deepest buried sediments (NC0901-4C) with
a fraction of As(lll) remaining bound to iron (hydr)oxides or iron carbonate at shallower depths within the HBHA Pond
(NC0901-1). It should be noted that arsenic associated with orpiment (As2S3) represented an insignificant fraction
(<10%) within buried sediments. The position of the arsenic absorption edge in all sediment samples that were evalu-
ated was significantly higher than that observed for mineral specimens and laboratory-prepared As2S3 (data not shown).
The arsenic speciation results for these samples are broadly consistent with the extraction results shown in Figure 21 A.
These results are consistent with the extraction data previously discussed where the bulk of the sediment-associated
arsenic was extracted using reagents that targeted poorly crystalline iron (hydr)oxides and iron sulfides (Tables 1 and 2).
The relative fraction of arsenic bound to oxygen increases for sediments with higher iron (hydr)oxide content (FeAsc/FeHCI
ratio near unity), while the fraction of arsenic bound to sulfur increases with higher iron sulfide content (low FeAsc/FeHCI
ratio). Therefore, the stability of solid phase arsenic in these sediments will partly be dictated by the stability of these
metastable minerals.
38
-------
The results from a selection of samples characterized for lead speciation are shown in Figure 24 (all data fits in Appen-
dix B), and the aerial distribution of lead phase associations within the HBHA Pond are shown in Figure 26. Analysis of
these samples indicates that lead speciation is dominated by partitioning to iron (hydr)oxide minerals. This pattern holds
even in sediments collected from the bottom of the HBHA Pond where there is evidence for iron- and sulfate-reduction
(Figures 13 and 15) and formation of iron sulfides (Figure 21E; low FeAsc/FeHCI ratios). A fraction of lead is partitioned
to a sulfidic phase within buried sediments, but in most cases, this represents a minor fraction of the species of lead
identified based on the LCF-XANES approach. This observation in combination with FeAsc/FeHCI extraction ratios >25%
for half of the buried sediment samples shown in Figure 21E clearly indicates that iron (hydr)oxides formed within oxic
portions of the HBHA Pond survive the reducing conditions established at the bottom of the pond. The persistence of
iron (hydr)oxides in zones of iron- and sulfate-reduction in sediments has been observed by others (Kostka and Luther,
1994). Since both poorly crystalline iron (hydr)oxides and iron sulfides are extracted by HCI, this suggests potential
limitations to the use of tests such as the AVS-SEM procedure to define element speciation.
Finally, the results from a selection of samples characterized for zinc speciation are shown in Figure 25 (all data fits in
Appendix B) and the aerial distribution of zinc phase associations within the HBHA Pond is shown in Figure 26. The
results of this analysis contrast with those for lead in that zinc sulfides play a more dominant role in the speciation of
this metal in buried sediments. Zinc partitioned to an iron (hydr)oxide phase is still observed as a significant compo-
nent within all sediments indicating that the survival of this mineral fraction still plays a role in zinc speciation within
the suboxic zone of the HBHA Pond. The decrease in the abundance of the sediment species for zinc relative to lead
is consistent with the greater stability of the lead sorption complex on iron (hydr)oxide surfaces (Stumm, 1992). In
general, the importance of sulfide minerals to element speciation within buried sediments appears to follow the order
As > Zn > Pb.
A sediment incubation study was conducted to observe the changes in metal speciation that may occur as iron
(hydr)oxides formed in the oxic zone are deposited in the suboxic zone within the HBHA Pond. In this study, surficial
sediments from the northern margin of the HBHA Pond were incubated with their native pore water in closed vessels
within the laboratory at room temperature. No attempt was made to control the biogeochemical conditions within the
vessel during incubation for a period of 2.5 years. The silt (<53 urn) and clay (<2 urn) fractions of the fresh and in-
cubated sediments were isolated by sieving and centrifugation to facilitate mineralogical characterization of the most
reactive mineral fraction. Significant alterations of the iron mineralogy were observed as a result of incubation. X-ray
diffraction data collected on the clay fraction of a representative sediment sample showed that a poorly-crystalline 6-
line ferrihydrite was dominant in the fresh sediment (Figure 27A). However, this mineral was transformed to a mixture
of goethite and siderite during the incubation period. Zinc speciation data were collected for the silt-sized fraction of
the fresh and incubated sample to determine the concurrent changes in contaminant speciation that accompanied the
observed transformations in bulk mineralogy (Figure 27B; arsenic and lead data not collected). Analysis of these data
using the LCF-XANES approach demonstrated significant changes in zinc speciation during sediment incubation. The
speciation of zinc was dominated by a combination of a hydroxide-like precipitate and adsorption/coprecitation with
an iron (hydr)oxide phase in the fresh sample. During incubation, a zinc sulfide phase formed at the expense of the
zinc-iron (hydr)oxide phase. The results of the XRD and LCF-XANES analyses demonstrate that iron-reducing and
sulfate-reducing conditions developed during the period of incubation, potentially driven by the growth of native microbial
communities in the closed environment. Unfortunately, data on the chemical conditions during aging and evidence for
microbial reactions were not collected during this experiment. However, the results of this experiment are consistent
with the observed change in zinc speciation in transitioning from surficial (oxic) to buried (suboxic) sediments within the
HBHA Pond. Specifically, iron- and sulfate-reducing processes play a critical role in determining contaminant speciation
in sediments deposited within the HBHA Pond.
39
-------
Oxic
Suboxic-
Anoxic
NC0901-4c
MB^^^^^j
As-FeS
Iron
Oxidation
Iron
Reduction
Sulfate
Reduction
11860 11870 11880 11890 11900
Energy, eV
Figure 23 Locations of a selection of sediments collected from oxic and suboxic zones within the HBHA Pond. XANES
data are shown for sediments that span the observed redox gradient within the system relative to representa-
tive reference compounds.
40
-------
8
I
Q
T3
9)
N
• Galena (PbS)
- Massicote (PbO)
-Anglesite (PbSO4)
Hydrocerrusite (Pb2(OH)2CO3)
- Pb-Ferrihydrite
15.15 15.20 15.25 15.30
Energy, keV
15.35
-------
0)
c
!c
o
Z
CD
J3
CD
Pb-S
Zn-S
As-S
Pb-O
'
Zn-O
As-0
Relative
Abundance
Relative Easting
Figure 26 Aerial distribution of contaminant sediment speciation within the HBHA Pond based on LCF-XANES analysis;
contour lines delineate depth to sediment (meters). FeAs/FeHQI is a measure of the relative fraction of iron
(hydr)oxide minerals. An 'x' denotes a location for which data is not available.
-------
2000
Fresh
800-
2400-
2000-
1600
1200
800
1 I '
Fernhydnle
Quartz
Incubated
Sidente
Quartz
I
I—r
10 20 30
CD
100 of
50 _
I
^?
s?
8
CD
•e
$
j^
T3
-------
Table 3 Atomic Percentages of Elements in 5 Samples Determined by SEM-EDS. Sample PS-3 is Composed
of 2-line Ferrihydrite; Poorly Crystalline 6-line Ferrihydrite is the Dominant Iron (Hydr)oxide in Samples
WI02-NEP and WI01-NEP. Atomic Percentages for Fe and O are Shown in Parentheses under the PS-3
Column for the Ferrihydrite Structure as Proposed by Towe and Bradley (1967)
Element
(emission
shell)
S(K)
Fe(K)
Si(K)
0(K)
Ca(K)
K(K)
Zn(K)
AI(K)
As(L)
Cu(K)
Na(K)
magnifica-
tion
WI02-NEPA
<2 urn
"oxic
sediment"
ND
11.16%±
0.52%
1 .84% ±
0.20%
37.94% ±
1.31%
ND
*0.36% ±
0.12%
ND
* 0.49% ±
0.17%
ND
ND
ND
2000x
WI02-NEPA
<2 urn
"pyrite
framboid"
49.63% ±
1 .08%
50.37% ±
1 .08%
ND
ND
ND
ND
ND
ND
ND
ND
ND
4300x
WI02-NEPB
2 urn < d <
53 urn
"oxic
sediment"
ND
19.63%±
0.80%
7.78% ±
0.46%
25.75% ±
1 .54%
0.56 ±0.1 7%
ND
ND
0.92% ±
0.19%
ND
ND
*0.47% ±
0.21%
SOOOx
PS-3
"water column
sample"
ND
61.92%±
0.64%
(64.58%)
3.26% ± 0.29%
30.25% ±
0.57%
(33.33%)
1.83% ±0.1 3%
ND
ND
0.89% ±0.1 6%
0.85% ± 0.26%
ND
ND
10000x
PS-3
Compositional
data from acid
digestion
2.54 wt%
34.28 wt%
0.87 wt%
~
1.21 wt%
0.16wt%
0.11 wt%
0.21 wt%
0.44 wt%
-
~
-
SC0401-3
"anoxic
sediment"
7.98% ±
0.32%
12.08% ±
0.48%
1 .48% ±
0.16%
20.44% ±
1.13%
*0.44% ±
0.11%
ND
1 .38% ±
0.36%
1 .52% ±
0.17%
ND
ND
ND
2000x
otes: ND = Not Detected
Below Limit of Quantitation (BLQ)
44
-------
WI02-NEPA
<2|am
(framboid)
12345
Energy (KeV)
Figure 28 Representative compositional spectra for samples imaged using SEM-EDS; UP = unidentified peak.
Figure 29 Image of a pyrite framboid observed in a surficial (oxic) sediment (WI02-NEP, clay-sized fraction).
45
-------
Figure 30 Images of iron (hydr)oxide precipitates collected near the chemocline within the water column of the HBHA
Pond (PS-3).
46
-------
Chapter 4
Stability of Contaminants in HBHA Pond Sediments
Assessment of ground-water data collected during this study and historic data collected within the study site indicate
that the primary point of discharge for contaminated ground water is into the HBHA Pond. Long-term monitoring at
the HBHA Pond surface water discharge indicates that dissolved arsenic is typically below a concentration of 100 ppb.
This suggests that arsenic and metals discharged via site-derived ground water are partially sequestered within the
HBHA Pond. The apparent geochemical processes controlling the solid-solution partitioning of arsenic, lead, and zinc
are illustrated in Figure 31. In general, arsenic solid-solution partitioning is linked to precipitation-dissolution cycles
that control the distribution of iron between dissolved and oxidic/sulfidic mineral phases. Arsenic remains in a relatively
labile chemical state, although the physicochemical characteristics of the HBHA Pond result in significant sequestra-
tion within the pond boundary. The cycling of lead and zinc within the HBHA Pond water column is less clear due to
the lack of clear chemical patterns in dissolved concentrations for these contaminants. However, the solid-solution
partitioning of lead and zinc in sediments also appears to be linked to the production of iron (hydr)oxides and sulfides.
These contaminants are also sequestered within the boundaries of the HBHA Pond, primarily in association with sedi-
ments. As noted previously (Figure 14 and 17), zinc derived from ground-water discharge is efficiently sequestered in
sediments. The observed release of arsenic from shallow sediments confirms the partial instability of iron (hydr)oxides
deposited on the bottom of the HBHA Pond (Ford, 2005). Dissolved concentrations of arsenic are a maximum within
the hypolimion near the sediment-water interface where iron (hydr)oxide dissolution prevails with a limited capacity for
production of iron sulfides (until sediment burial and more intense sulfate reduction). This observation is consistent with
projections made by Hounslow (1980), where regions of iron reduction coupled with insufficient sulfate reduction are
anticipated to result in the most labile/mobile forms of arsenic. The circumneutral pH near the sediment-water interface
also contributes to selective sequestration of zinc versus arsenic in sulfate-reducing zones due to the higher solubility
of iron and arsenic sulfides (Wilkin and Ford, 2002). Ultimately, although the HBHA Pond retains a significant fraction
of arsenic and metals derived from ground-water discharge, it could still supply a continual flux of arsenic and metals to
downgradient wetlands and the Aberjona River via discharge of dissolved contaminant species or suspended solids.
The critical factors maintaining arsenic, lead, and zinc at depth within the HBHA Pond hypolimnion are 1) the presence of
a fairly stable chemocline and 2) the high capacity for metal sorption to iron (hydr)oxides and/or iron sulfides formed at
the oxic-anoxic transition zone (Taillefert et al., 2000; Ford, 2002; Taillefert and Gaillard, 2002). The capacity for arsenic
and metal sequestration from the HBHA Pond water column is dependent on a continual supply of ferrous iron into the
system, while the continual production of iron sulfides is dependent on maintenance of sulfate reduction processes
within the hypolimnion. The arsenic (and potentially lead and zinc) sequestration capacity will decline concurrent with
a decline in ferrous iron from the upgradient aquifer. Since performance of the HBHA Pond for arsenic removal is also
dependent on continued stratification of the water column, assessment of the long-term sedimentation capacity is also
an important factor. Water column monitoring data from this study suggest that the chemocline is typically located
approximately 100-200 cm above the sediment-water interface depending on the season. While there are no data
to assess sediment accumulation rates for the HBHA Pond, it is not anticipated that sediment in-filling would cause
disruption of the chemocline. Regardless, projections of potential sediment erosion or turnover within the HBHA Pond
will depend on knowledge of modifications to the pond geometry over time. Buried sediments within the HBHA Pond
have accumulated significant concentrations of metals such as lead and zinc. Mineralogical characterization and
contaminant speciation indicate that both iron (hydr)oxides and iron sulfides play a significant role in this process within
the hypolimnion and buried sediments. The relative distribution of these two mineral classes will depend on the extent
of sulfate-reduction and the rates of dissolution of the iron (hydr)oxides deposited onto buried sediments (Wersin et
al., 1991; Poulton et al., 2004). The stability of these sediments will depend on the maintenance of iron- and sulfate-
reducing conditions at depth within the HBHA Pond. Current and historical ground-water data suggest that these
processes will continue into the future, but consideration should be given to the impact of water column turnover and
the potential long-term depletion of the reducing capacity within the aquifer as the BTEX plume diminishes over time
(e.g., Saulnier and Mucci, 2000).
47
-------
Since the location and mass of arsenic (and iron) within contaminated soils upgradient from the HBHA Pond are poorly
defined, it is difficult to assess the long-term capacity of the HBHA Pond. Given this level of uncertainty, it is recom-
mended that a long-term monitoring strategy be implemented that continuously tracks arsenic and redox chemistry within
the aquifer and the HBHA Pond. This appears to be less of a concern for lead and zinc, since elevated concentrations
of these metals were rarely observed within the HBHA Pond water column. A proposed strategy is outlined below.
Recommendations for site monitoring and potential remediation apply only to discharge of contaminated ground water
into the HBHA Pond and are not intended to address potential contaminant migration in surface water and ground water
downgradient from the HBHA Pond.
'2(g)
Chemocline
As-HFO(s) (Pb/Zn?)
:Fe(")(aq)
» !AS
Sediment Speciation
Zn
distal & near discharge
distal
near discharge
distal
near discharge
Figure 31 Illustration of the apparent geochemical processes controlling solid-solution partitioning of arsenic, lead,
and zinc within the HBHA Pond: 1) Fe(ll) oxidation near chemocline upon encountering dissolved oxygen
and coprecipitation of As with iron (hydr)oxides, 2) settling of precipitated iron (hydroxides, 3A) reductive
dissolution of a fraction of settling iron (hydr)oxides generating dissolved Fe(ll) and arsenic, 3B) deposition
and diagenisis of a fraction of precipitated iron (hydr)oxides - partial conversion to iron sulfides, and 4)
vertical transport of dissolved Fe(ll) and As up to the chemocline. The speciation of arsenic, lead, and zinc
in sediments located near and distal to the primary zone of contaminated ground-water discharge is shown
(red = oxide association; black = sulfide association).
48
-------
Monitoring Long-Term Behavior (or Performance) of HBHA Pond
It is evident that the HBHA Pond currently serves to remove a fraction of the metal burden from ground water originating
from the Industri-Plex Superfund Site. However, the ability of the HBHA Pond to function in this capacity over the long
term will be controlled by 1) continued maintenance of the chemical stratification within the water column and 2) con-
tinued supply of iron and sulfate from site ground water for the production of iron (hydr)oxides and sulfides.
As shown via the HBHA Pond water column sampling during April 2001, high surface water flow events can perturb the
chemical stratification (Ford, 2005). While the chemocline was re-established within the HBHA Pond, there are insuf-
ficient monitoring data to track the impact of these transient flow events on the long-term performance of the HBHA
Pond for metal sequestration. Forecasting the impact of surface flow events is critical due to site development, which
can modify patterns in surface runoff into the HBHA Pond. It is recommended that a long-term monitoring program
incorporate input and output flow measurements for the HBHA Pond in order to establish whether the major flow event
occurring at the end of March 2001 was anomalous or is likely to occur on a more frequent basis. Note that continued
discharge from Halls Brook should be maintained, since this continuous source of fresh water helps maintain chemical
stratification within the HBHA Pond. As demonstrated by depth-resolved measurements, the highest concentration of
dissolved As occurs at the bottom of the HBHA Pond and is relatively invariant throughout the year. This mass of dis-
solved As will be distributed throughout the water column during turnover, leading to increased mass flux of As at the
shallow outlet. This is a critical factor, since the HBHA Pond is a potential long-term source of arsenic transport to the
Aberjona River. Based on the observed perturbation to the performance of the HBHA Pond during the March 2001
surface flow event, it is recommended that control structures be installed to better regulate surface water inputs into
the HBHA Pond. Structures to divert excess flows during precipitation/snow melt events of similar magnitude to that
observed during March - April 2001 will facilitate maintenance of the chemocline within the HBHA Pond, which appears
critical to the precipitation and settling characteristics of the iron cycle within the water column.
It is recommended that a more comprehensive site monitoring strategy be implemented in the near-term to better track
performance of the HBHA Pond system and detect possible failure. The current focus of monitoring efforts outside of
this study has entailed extensive evaluation of surface water inputs and discharge out of the HBHA Pond. In order to
properly assess performance of the HBHA Pond to sequester arsenic and metal contaminants, it is recommended that
a permanent ground-water monitoring network be established within the aquifer upgradient and downgradient from
the HBHA Pond. The network should be sufficient to document the extent of the arsenic contaminant plume and its
elimination down gradient to the HBHA Pond. Contaminants of concern and the general redox chemistry of ground
water should be monitored concurrent with surface water sampling events. In general, quarterly sampling will provide
sufficient frequency to monitor the dynamics of the ground-water plume and the HBHA system, but it is recommended
that additional efforts be made to assess system response to storm events. Assessment of seasonal patterns in con-
taminant mobility and sequestration within the HBHA Pond is required to establish the long-term viability of the metal
sequestration reactions active within the HBHA Pond to mitigate mobilization of arsenic and metals.
There is evidence that the chemical stratification within the HBHA Pond can be interrupted during large flow events.
Ford (2005) has documented the loss of chemical stratification in the central and southern portions of the HBHA Pond
following a large flow event during March 2001. The flow-induced mixing resulted in distribution of the entire budget of
hypolimnetic arsenic throughout the water column in these portions of the pond and ultimately resulted in a temporary
increase in arsenic export from the HBHA Pond. Evaluation of contemporary and historical water flow records at a
hydrologic monitoring station approximately 2.5 miles south of the HBHA Pond (USGS 01102005) indicates that an
event of this magnitude could potentially occur on a 3-4 year cycle. These results suggest that efforts to control high-
flow events into the HBHA Pond will be important to limiting down gradient transport of arsenic and metals derived from
discharge of contaminated ground water. However, the ultimate performance of the HBHA Pond system will be limited
by net sedimentation, which will slowly consume the storage capacity for contaminated water and sediments within the
deeper portions of the pond.
Relevance to Other Sites
The observations of the dynamics of iron mineralogical cycling and the chemical speciation of arsenic, lead, and zinc
from this study site may be relevant to other sites. The differential influence of iron- and sulfate-reduction processes
on the solid-solution partitioning of inorganic contaminants observed in this study is important relative to the design of
the characterization plan for other sites with complex metal and metalloid contaminant mixtures. The overall stability of
sediment-associated contaminants as well as the technologies that may be employed for their remediation will depend
on the types of mineral associations controlling contaminant solid-phase speciation. The results of the work described
in this document indicate that, of the three inorganic contaminants that were studied, arsenic posed the greatest concern
due to its higher potential mobility. This may apply to other sites where inorganic contaminants are a primary concern.
The extent to which this could be defined for the HBHA Pond was dependent on development of a comprehensive
knowledge of the influence of iron- and sulfate-reduction on the solid-phase partitioning of arsenic within the system
49
-------
boundaries. This study demonstrated a general correspondence between observed patterns in the aqueous distribution
of arsenic, lead, and zinc relative to the predominant solid-phase speciation patterns for these contaminants based on
chemical extractions and spectroscopic analyses. In general, this work supports the applicability of using chemical ex-
tractions to help define contaminant speciation. However, results for the specific contaminants indicate that care should
be taken during the design of an extraction protocol relative to observed site conditions. Ultimately, the reliability and
utility of extraction protocols to define inorganic contaminant speciation will be predicated on concurrently developed
knowledge of the mineralogical composition of the tested solids.
50
-------
References
Aggett, J., and M.R. Kriegman. The extent of formation of arsenic(lll) in sediment interstitial waters and its release to
hypolimnetic waters in Lake Ohakuri Water Resources 22: 407-411 (1998).
Aggett, J., and G.A. O'Brien. Detailed model for the mobility of arsenic in lacustrine sediments based on measurements
in Lake Ohakuri. Environmental Science and Technology 19: 321-238(1985).
Ahmann, D., L.R. Krumholz, H.F. Hemond, D.R. Lovely, and F.M.M. Morel. Microbial mobilization of arsenic from sedi-
ments of the Aberjona watershed. Environmental Science and Technology 31: 2923-2930 (1997).
Atkin, B.R and C. Somerfield. The determination of total sulfur in geological materials by coulometric titration. Chemi-
cal Geology 111: 131 -134 (1994).
Aurilio, A.C., R.R Mason, and H.F. Hemond. Speciation and fate of arsenic in three lakes of the Aberjona watershed.
Environmental Science and Technology 28: 577-585 (1994).
Aurilio, A.C., J.L. Durant, H.F. Hemond, and M.L. Knox. Sources and distribution of arsenic in the Aberjona watershed,
eastern Massachusetts. Water Air Soil Pollution 78:1-18 (1995).
Balistrieri, L.S., J.W. Murray, and B. Paul. The geochemical cycling of trace elements in a biogenic meromictic lake.
Geochimica et Cosmochimica Acta 58: 3993-4008 (1994).
Bartlett, R.J. and R.R. James, Chromium in the Natural and Human Environment, New York: John Wiley & Sons,
1988.
Beauchemin, S., D. Hesterberg, and M. Beauchemin. Principal component analysis approach for modeling sulfur K-
XANES spectra of humic acids. American Journal of Soil Science Society 66: 83-91 (2002).
Biksey, T.M., and E.D. Gross. The hyporheic zone: linking groundwater and surface water - understanding the paradigm.
Remediation 12: 55-62 (2001).
Davis, A., J.H. Kempton, A. Nicholson, and B.Yare. Groundwater transport of arsenic and chromium at a historical tan-
nery, Woburn, Massachusetts, U.S.A. Applied Geochemistry 9: 569-582 (1994).
Davis, A., C. Sellstone, S. Clough, R. Barrick, and B.Yare. Bioaccumulation of arsenic, chromium and lead in fish: con-
straints imposed by sediment geochemistry. Applied Geochemistry 11: 409-423 (1996).
Davison, W. and G. Seed. The kinetics of the oxidation of ferrous iron in synthetic and natural waters. Geochimica et
Cosmochimica Acta 47: 67-79 (1983).
Davison, W., C. Woof, and E. Rigg. The dynamics of iron and manganese in a seasonally anoxic lake; direct measure-
ment of fluxes using sediment traps. Limnol. Oceanography 27: 987-1003 (1982).
Dos Santos Afonso, M., P.J. Morando, M.A. Blesa, S. Banwart, and W. Stumm. The reductive dissolution of iron oxides
by ascorbate: The role of carboxylate anions in accelerating reductive dissolution. Journal of Colloid and Interface
Sc/ence138: 74-82(1990).
Durant, J.L., J.J. Zemach, and H.F. Hemond. The history of leather industry waste contamination in the Aberjona wa-
tershed: A mass balance approach. Civil Engineering Practice 5: 41-66 (1990).
Ferdelman, T.G. "The distribution of sulfur, iron, manganese, copper, and uranium in a salt marsh sediment core as
determined by sequential extraction methods." Masters Thesis, University of Delaware, 1988.
Ford, R.G. Rates of hydrous ferric oxide crystallization and the influence on coprecipitated arsenate. Environmental
Science and Technology 36: 2459-2463 (2002).
Ford, R.G. The impact of ground water-surface water interactions on contaminant transport with application to an arsenic
contaminated site, EPA/600/S-05/002. Cincinnati, OH: U.S. Environmental Protection Agency, 2005.
Ford, R.G., K.M. Kemner, and P.M. Bertsch. Influence of sorbate-sorbent interactions on the crystallization kinetics of
nickel- and lead-ferrihydrite coprecipitates. Geochimica et Cosmochimica Acta 63: 39-48 (1999).
51
-------
Ford, R.G., R.T. Wilkin, and G. Hernandez. Arsenic cycling within the water column of a small lake receiving contami-
nated ground-water discharge. Chemical Geology (accepted) (2005).
Gomez-Ariza, J.L.D. Sanchez-Rodas, I. Giraldez, and E. Morales. A comparison between ICP-MS and AFS detection
for arsenic speciation in environmental samples. Talanta 51: 257-268 (2000).
Hach. "Iron, ferrous." In Water Analysis Handbook, Loveland, CO: Hach Co., 1992a, 343-346.
Hach. "Iron, total." In Water Analysis Handbook, Loveland, CO: Hach Co., 1992b, 353-357.
Hounslow, A.W. Ground-water geochemistry: arsenic in landfills. Ground Water 18: 331-333 (1980).
Huffman, E.W.D. Performance of a new carbon dioxide coulometer. Microchemical Journal 22: 567-573 (1977).
Jackson, B.R, and W.R Miller. Effectiveness of phosphate and hydroxide for desorption of arsenic and selenium species
from iron oxides. Soil Science Society of America Journal 64:1616-1622 (2000).
JCPDS. Mineral Powder Diffraction File Databook; Sets 1-42, International Centre for Diffraction Data: Swarthmore,
PA, 1993.
Kneebone, RE., PA. O'Day, N. Jones, and J.G. Hering. Deposition and fate of arsenic in iron- and arsenic-enriched
reservoir sediments. Environmental Science and Technology 36: 381 -386 (2002).
Kostka, J.E., and G.W. Luther III. Partitioning and speciation of solid phase iron in saltmarsh sediments. Geochimica
et Cosmochima Ada 56: 1701 -1710 (1994).
Larsen, O., and D. Postma. Kinetics of reductive bulk dissolution of lepidocrocite, ferrihydrite, and goethite. Geochimica
et Cosmochimica Ada 65: 1367-1379 (2001).
Martin, A.J., andT.F. Pederson. Seasonal and interannual mobility of arsenic in a lake impacted by metal mining. Envi-
ronmental Science and Technology 36: 1516-1523 (2002).
McBride, M.B. Environmental Chemistry of Soils, New York: Oxford University Press, 1994.
McCleskey, R.B., O.K. Nordstrom, and A.S. Maest. Preservation of water samples for arsenic(lll/V) determinations: an
evaluation of the literature and new analytical results. Applied Geochemistry 19: 995-1009 (2004).
Peltier, E., A.L. Dahl, and J.-F. Gaillard. Metal speciation in anoxic sediments: When sulfides can be construed as
oxides. Environmental Science and Technology 39: 311 -316 (2005).
Perret, D., J.-F. Gaillard, J. Dominik, and O. Atteia. The diversity of natural hydrous iron oxides. Environmental Science
and Technology 34: 3540-3546 (2000).
Poulton, S.W., M.D. Krom, and R. Raiswell. A revised scheme for the reactivity of iron (oxyhydr)oxide minerals towards
dissolved sulfide. Geochimica et Cosmochimica Ada 68: 3703-3715 (2004).
Puls, R.W., and C.J. Paul. Low-flow purging and sampling of ground-water monitoring wells with dedicated systems.
Ground Water Monitoring and Remediation 15: 116-123 (1995).
Rehr, J.J., and A.L. Ankudinov. Progress and challenges in the theory and interpretation of X-ray spectra. Journal of
Synchrotron Radiation 8: 61-65(2001).
Rehr, J.J., and A.L. Ankudinov. New developments in the theory and interpretation of X-ray spectra based on fast paral-
lel calculations. Journal of Synchrotron Radiation 10: 43-45 (2003).
Ressler, T. WinXAS: A program for X-ray absorption spectroscopy data analysis under MS-Windows. Journal of Syn-
chrotron Radiation 5: 118-122 (1998).
Roberts, D.R., A.C. Scheinost, and D.L. Sparks. Zinc speciation in a smelter-contaminated soil profile using bulk and
microspectroscopic techniques. Environmental Science and Technology 36: 1742-1750 (2002).
Saulnier, I., and A. Mucci. Trace metal remobilization following the resuspension of estuarine sediments: Saguenay
Fjord, Canada. Applied Geochemistry 15:191-210 (2000).
Scheckel, K.G., and J.A. Ryan. Spectroscopic speciation and quantification of Pb in phosphate amended soils. Journal
of Environmental Quality 33: 1288-1295 (2004).
Scheinost, A.C., R. Kretzschmar, and S. Pfister. Combining selective sequential extractions, X-ray absorption spec-
troscopy, and principal component analysis for quantitative zinc speciation in soil. Environmental Science and
Technology 36: 5021-5028(2002).
Schwertmann, U., and W.R. Fischer. Natural "amorphous" ferric hydroxide. Geoderma 10: 237-247 (1973).
Schwertmann, U., and E. Murad. The nature of an iron oxide-organic iron association in a peaty environment. Clay
Minerals 23: 291-299 (1988).
52
-------
Senn, D.A., and H.F. Hemond. Nitrate controls on iron and arsenic in an urban lake. Science 296: 2373-2376 (2002).
Seyler, P., and J.-M. Martin. Biogeochemical processes affecting arsenic species distribution in a permanently stratified
lake. Environmental Science and Technology 23: 1258-1263 (1989).
Sholkovitz, E.R., and D. Copland. The chemistry of suspended matter in Esthwaite Water, a biologically productive lake
with seasonally anoxic hypolimnion. Geochimica et Cosmochimica Acta 46: 393-410 (1982).
Sohrin, Y., M. Matsui, M. Kawashima, M. Hojo, and H. Hasegawa. Arsenic biogeochemistry affected by eutrophication
in Lake Biwa, Japan. Environmental Science and Technology 31: 2712-2720 (1997).
Spliethoff, H.M., R.R Mason, and H.F. Hemond. Interannual variability in the speciation and mobility of arsenic in a
dimictic lake. Environmental Science and Technology 29: 2157-2161 (1995).
Standard Methods for the Examination of Water and Wastewater. Section 4140 B Capillary Ion Electrophoresis with
Indirect UV Detection, 20th Edition, pp. 1-4 to 1-13 (1999).
Stanjek, H., and RG. Weidler. The effect of dry heating on the chemistry, surface area, and oxalate solubility of synthetic
2-line and 6-line ferrihydrites. Clay Miner 27: 397-412 (1992).
Stumm, W. Chemistry of the Solid-Water Interface, New York: John Wiley & Sons, Inc., 1992.
Taillefert, M., C.-R Lienemann, J.-F. Gaillard, and D. Perret. Speciation, reactivity, and cycling of Fe and Pb in a meromictic
lake. Geochimica et Cosmochimica Acta 64: 169-183 (2000).
Taillefert, M., and J.-F. Gaillard. Reactive transport modeling of trace elements in the water column of a stratified lake:
iron cycling and metal scavenging. Journal of Hydrology 256: 16-34 (2002).
Tessier, A., P.G.C. Campbell, and M. Bison. Sequential extraction procedure for the speciation of particulate trace met-
als. Analytical Chemistry 51: 844-851 (1979).
Tomassoni, G. "A federal statutory/regulatory/policy perspective on remedial decision-making with respect to ground-
water/surface-water interaction." In Proceedings of the Ground-Water/Surface-Water Interactions Workshop,
Denver, CO, 1999.
Towe, K.M., and W.F. Bradley. Mineralogical constitution of colloidal "hydrous ferric oxides." Journal of Colloid and Inter-
face Science 24: 384-392 (1967).
U.S. Environmental Protection Agency. Methods for chemical analysis of water and wastes. EPA/600/4-79/020. Wash-
ington, DC, 1983.
Welch, A.H., and M.S. Lico. Factors controlling As and U in shallow ground water, southern Carson Desert, Nevada.
Applied Geochemistry 13: 521 -539 (1998).
Wersin, P., P. Hohener, R. Giovanoli, and W. Stumm. Early diagenetic influences on iron transformations in a fresh-water
lake sediment. Chemical Geology 90: 233-252 (1991).
Wick, L.Y., and P.M. Gschwend. Source and chemodynamic behavior of diphenyl sulfone and ortho- and para-hydroxybi-
phenyl in a small lake receiving discharges from an adjacent superfund site. Environmental Science and Technology
32: 1319-1328(1998a).
Wick, L.Y., and P.M. Gschwend. By-products of a former phenol manufacturing site in a small lake adjacent to a Super-
fund site in the Aberjona watershed. Environmental Health Perspectives 106:1069-1074 (1998b).
Wick, L.Y., K. McNeill, M. Rojo, E. Medilanski, and P.M. Gschwend. Fate of benzene in a stratified lake receiving con-
taminated groundwater discharges from a superfund site. Environmental Science and Technology 34: 4354-4362
(2000).
Wilkin, R.T., and R.G. Ford. Use of hydrochloric acid for determining solid-phase arsenic partitioning in sulfidic sedi-
ments. Environmental Science and Technology 36: 4921-4927 (2002).
Winter, T.C., J.W. Harvey, O.L. Franke, and W.M. Alley. Ground water and surface water: a single resource. U.S. Geologi-
cal Survey Circular 1139: 77 pp (1998).
53
-------
54
-------
Appendix A
Analytical Performance for Laboratory Methods
Table A.1. Analytical Methods, Detection Limits, Precision, and Accuracy for Measurement of Aqueous Chemis-
try
Parameter
Benzene, Toluene
TOC, DOC
Fe2+
Arsenic, Iron, So-
dium, Lead, Sulfur,
Zinc
NO3-N
NH3-N
so4
Method
Purge-and-Trap Gas
Chromatography
Wet Oxidation/Infrared
Detection
EPA Method 41 5.2
Colorimetric, 1,10-phen-
anthroline
ICP-OES
EPA Method 353.2
EPA Method 350.1
Capillary electropho-
resis
Detection Limit
0.29 ug/L, benzene
0.22 ug/L, toluene
0.02 mg C/L
0.01 mg/L
0.033 mg/L, As
0.035 mg/L, Fe
0.479 mg/L, Na
0.01 5 mg/L, Pb
0. 137 mg/L, S
0.01 4 mg/L, Zn
0.004 mg N/L
0.03 mg N/L
0.1 mg/L
Precision
±15% based on
duplicate analyses
of unknowns
±10% based on
duplicate analyses
of unknowns
±10% based on
duplicate analyses
of unknowns
±10% based on
duplicate analyses
of unknowns
±10% based on
duplicate analyses
of unknowns
±10% based on
duplicate analyses
of unknowns
±10% based on
duplicate analyses
of unknowns
Accuracy
±20% based on
spike recovery for un-
knowns
±20% based on
spike recovery for un-
knowns
±20% based on
spike recovery for un-
knowns
±20% based on
spike recovery for un-
knowns
±20% based on
spike recovery for un-
knowns
±20% based on
spike recovery for un-
knowns
±20% based on
spike recovery for un-
knowns
n.d. not determined
55
-------
Table A.2. Analytical Methods, Detection Limits, Precision, and Accuracy for Measurement of Solid Phase Chem-
istry
Parameter
Total Carbon
Total Sulfur
Arsenic, Chro-
mium, Iron,
Lead, Zinc
Method
Coulometry/
combustion
Coulometry/
combustion
ICP-OES/microwave
assisted digestion/
chemical extraction
Detection Limit
0.01 wt% C for a
100-mg sample
0.005 wt% S for a
1 00-mg sample
1 to 1 0 ppm for 1 00-mg
sample (depending on
element) and 20 mL
extraction volume
Precision
±3.2% based on 32
analyses of CaCO3
±5.1% based on 35
analyses of NIST
1646a
±5 to 15% based on
duplicate or tripli-
cate analyses of un-
knowns
Accuracy
±2.8% using CaCO3
(12.0 wt%C)
±2.5% using
NIST1646a
(0.35 wt% S)
Variable, depend-
ing on element and
reference material
As (±3 to 12%)
n.d. not determined; standard reference materials used: NIST 1646a Estuary sediment, NIST 2710 Montana soil, NIST2780 Hard rock mine waste,
CCRMP Lake sediment-1
56
-------
Appendix B
Sediment Composition Data
Table B.1. Concentrations of Selected Elements in Halls Brook Holding Area Pond Core NC01
Core Sample
NC01-1
NC01-2
NC01-3
NC01-4
NC01-5
NC01-6
NC01-7
NC01-8
NC01-9
NC01-10
Depth cm
1.4
2.8
4.2
5.6
7.0
8.4
10.0
18.3
26.6
34.9
Fe wt%
9.91
9.65
9.54
9.38
9.38
10.20
10.10
2.21
1.51
0.37
S
wt%
2.20
1.80
2.20
2.11
2.10
1.77
0.05
0.04
0.04
0.03
TOC
wt%
15.24
15.58
14.27
14.61
13.32
14.52
5.95
0.27
0.17
0.03
As
ug/gm
836
857
828
839
834
952
639
150
101
10
Pb
ug/gm
593
610
613
628
607
668
606
507
106
34
Cr
ug/gm
839
970
1060
815
856
940
691
501
87.4
13.8
Zn
ug/gm
5920
5560
5710
5130
5290
6060
3990
2850
609
71
Notes: Core collected 4/03/2000. Fe, As, Pb, Cr, and Zn determined by ICP-OES after microwave-assisted extraction
in 10% nitric acid. TOC=total carbon minus inorganic carbon, determined using an UIC, Inc. carbon coulometer. S
determined by coulometery (UIC, Inc.).
57
-------
Table B.2. Concentrations of Selected Elements in Halls Brook Holding Area Pond Core CC02
Core Sample
CC02-1
CC02-2
CC02-3
CC02-4
CC02-5
CC02-6
CC02-7
CC02-8
CC02-9
CC02-10
CC02-11
CC02-12
CC02-13
CC02-14
Depth cm
2.1
4.1
6.3
8.4
10.5
12.6
14.7
16.8
21.1
25.4
31.9
38.4
45.9
53.4
Fe wt%
9.89
11.01
11.40
11.80
12.36
14.55
10.10
7.21
1.51
0.37
0.74
0.61
0.62
0.75
S
wt%
1.36
1.38
1.34
1.43
1.73
1.31
1.41
1.00
0.20
0.04
0.23
0.05
0.05
0.04
TOC wt%
15.37
17.06
14.58
14.64
14.26
13.09
11.71
9.81
1.91
0.33
0.33
0.29
0.12
0.05
Aspg/
gm
1120
1101
1020
1050
914
945
639
455
101
10.3
18.5
6.49
2.00
6.22
PbM9/
gm
429
425
514
528
521
633
606
507
106
34.0
49.5
9.8
4.7
3.0
Cr |jg/gm
593
573
830
901
826
830
691
501
87.4
13.8
31.5
15.5
8.9
7.4
Zn (jg/gm
3500
3499
4170
4380
4420
4884
3990
2850
609
71
373
121
45
24
Wotes: Core collected 4/03/2000.
carbon minus inorganic carbon,
Fe, As, Pb, Cr, and Zn determined by
determined using an UIC, Inc. carbon
ICP-OES after microwave-assisted extraction in 10% nitric acid. TOC=total
coulometer. S determined by coulometery (UIC, Inc.).
Table B.3. Concentrations of Selected Elements in Halls Brook Holding Area Pond Core SC02
Core
Sample
SC02-1
SC02-2
SC02-3
SC02-4
SC02-5
SC02-6
SC02-7
Depth cm
3.0
6.0
9.0
12.0
15.0
18.0
21.0
Fe wt%
10.20
11.94
10.30
3.40
2.72
1.28
0.07
S
wt%
2.54
2.25
2.18
1.37
0.57
0.17
0.03
TOC
wt%
10.33
11.32
9.51
5.04
6.46
1.87
0.13
As
M9/gm
1570
1439
1510
232
200
66
10
Pb
ng/gm
403
455
412
315
452
226
202
Cr
H9/gm
529
594
557
328
462
219
187
Zn
Mg/gm
2780
3612
3060
1790
1860
357
255
Notes: Core collected 4/03/2000. Fe, As, Pb, Cr, and Zn determined by ICP-OES after microwave-assisted extraction in 10% nitric acid. TOC=total
carbon minus inorganic carbon, determined using an UIC, Inc. carbon coulometer. S determined by coulometery (UIC, Inc.).
58
-------
Table B.4. Concentrations of Selected Elements in Halls Brook Holding Area Pond Grab Sediment Samples
Sediment
Sample
WI01
WI01-NEP
WI02
WI02-NEP
WI04
SC0401-1
SC0401-2
SC0401-3
SC0401-4
SC0401-5
SC0401-6
SC0401-7
NC0901-1
NC0901-2
NC0901-3
NC0901-4a
NC0901-4D
NC0901-4C
NC0901-4d
Depth
cm
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<20
<20
<20
<20
Fe
wt%
3.80
7.70
24.60
18.50
26.30
7.46
8.94
11.90
5.03
7.59
8.56
8.65
1.90
3.37
1.65
7.00
7.04
8.23
9.66
S
wt%
0.20
0.18
0.47
0.65
0.35
1.14
2.44
10.95
0.37
1.59
2.85
2.32
0.83
2.24
0.51
6.71
5.68
6.89
9.91
TOC
wt%
3.10
3.00
4.75
5.24
5.26
5.24
10.26
6.69
4.53
14.18
18.61
17.34
1.50
2.29
0.25
13.32
12.98
12.86
11.09
As
Mg/gm
494
830
715
630
840
682
1150
1680
539
676
1070
973
207
356
117
1430
1490
1690
2050
Pb
|jg/gm
794
572
203
262
115
274
392
398
494
526
481
360
92
144
20
464
461
649
548
Cr
(jg/gm
197
122
106
162
49
511
698
250
148
593
694
549
65
107
23.6
431
437
478
606
Zn
Mg/gm
583
767
3034
4589
1517
2550
3790
17500
3920
3440
3190
2590
836
1580
682
4300
4780
7480
9810
Notes: 'Wl'sediments collected on 8/24/2000; 'SC040T sediments collected on 4/03/2001; and 'NC0901' samples collected in September 2001. Fe,
As, Pb, Cr, and Zn determined by ICP-OES after microwave-assisted extraction in 10% nitric acid. TOC=total carbon minus inorganic carbon,
determined using an UIC, Inc. carbon coulometer. S determined by coulometery (UIC, Inc.).
59
-------
Table B.5. Concentrations of Selected Elements in Halls Brook Holding Area Pond Cores
Core Sample
NC0401-1
NC0401-1
NC0401-1
NC0401-1
NC0401-2
NC0401-2
NC0401-3
NC0401-3
NC0401-3
NC0401-3
NC0401-3
NC0401-4
NC0401-4
NC0401-4
NC0401-4
NC0401-4
NC0401-5
NC0401 -5
NC0401-5
NC0401 -5
NC0401-5
NC0401 -6
NC0401-6
NC0401-6
NC0401-6
NC0401-6
NC0401-6
NC0401-7
NC0401-7
NC0401-8
Depth
cm
1.0
3.0
5.0
7.0
1.5
4.5
1.0
3.0
6.0
10.0
14.0
1.0
3.0
6.0
9.5
12.5
1.0
3.0
5.0
7.5
11.5
1.0
3.0
5.0
7.0
11.0
13.5
3.0
9.0
4.0
Fe
wt%
8.87
6.47
1.84
0.57
0.35
0.91
8.78
6.61
2.34
1.71
0.68
5.94
7.05
0.82
0.44
0.40
7.85
12.00
3.85
5.70
1.25
7.87
15.60
3.85
2.12
1.20
0.32
1.63
1.01
0.57
S
wt%
1.80
1.87
0.31
0.10
0.09
0.41
1.61
1.65
0.18
0.10
0.05
2.89
2.17
0.16
0.06
0.03
5.42
4.54
0.49
0.43
0.04
4.86
13.86
1.60
0.49
0.12
0.02
0.21
0.09
<0.01
TOC wt%
11.92
12.28
3.92
1.39
1.01
2.84
8.82
7.91
2.42
1.56
0.92
13.27
13.32
0.46
0.02
0.03
11.62
10.21
0.69
2.57
0.05
12.66
7.02
0.14
0.26
0.23
0.06
3.18
3.08
0.07
As
Mg/gm
1020
850
183
68
19
154
284
330
126
88
6
694
1020
20
2
2
1060
1960
342
405
21
1550
2570
65
78
32
2
158
107
<2
Pb
Lig/gm
763
755
1550
837
1070
350
923
957
1430
734
38
369
506
18
2
0.5
459
557
244
681
27
452
404
27
62
50
6
1210
784
4
Cr
Mg/gm
549
893
421
31
11
140
640
500
300
221
17
389
811
17
6
6
508
632
120
149
22
617
359
22
37
27
5
278
368
8
Zn
|jg/gm
4270
6320
1240
77
67
1120
7170
4390
505
334
46
2270
4090
104
24
18
3080
10300
1900
1210
185
5520
15500
273
301
162
29
2100
275
15
Notes: NC0401 cores collected in April 2001. Fe, As, Pb, Cr, and Zn determined by ICP-OES after microwave-assisted extraction in 10% nitric acid.
TOC=total carbon minus inorganic carbon, determined using an UIC, Inc. carbon coulometer. S determined by coulometery (UIC, Inc.).
60
-------
Table B.6. Results from LCF-XANES Fits of the Pb XANES Data Collected for Sediments from Suboxic and Oxic
Zones within the HBHA Pond
Weighted Percent Contribution
NC01-1
NC01-3
NC01-6
NC01-7
NC0401-1 0-2
NC0401-1 2-4
NC0401-50-2
NC0401-52-4
NC0401-60-2
NC0901-4B
NC0901-4C
NTW4
SC0401-3
SC0401-6
Galena
19.4
12.7
5.0
37.8
11.6
4.1
16.1
3.2
4.9
3.6
5.3
20.3
4.2
4.1
s
1.11
1.09
0.21
0.32
1.24
0.28
0.86
0.25
0.21
0.26
0.11
1.74
0.32
0.15
Pb-Sorbed
Ferrihydrite
80.6
87.3
95.0
62.3
88.4
95.9
83.9
96.8
95.2
96.4
94.7
79.7
95.8
95.9
s
1.10
1.09
0.21
0.32
1.23
0.28
0.86
0.26
0.21
0.26
0.11
1.73
0.32
0.15
F-Test
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Residual
0.774
0.807
0.875
0.622
0.851
0.543
0.856
0.822
0.823
0.744
0.784
0.747
0.872
0.808
s = standard deviation
61
-------
Table B.7. Results from LCF-XANES Fits of the Zn XANES Data Collected for Sediments from Suboxic and Oxic
Zones within the HBHA Pond
Weighted Percent Contribution
NC01-1
NC01-3
NC01-6
NC01-7
NC0401-1 0-2
NC0401-1 2-4
NC0401-50-2
NC0401-52-4
NC0401-60-2
NC0401-62-4
NC0901-1
NC0901-4B
NC0901-4C
NTW4
SC0401-1
SC0401-3
SC0401-6
SC0401-7
WI02-N1-53D2
WI02-N2-53D2
WI02-N2B 2um
WI02-NINE
Sphalerite
73.8
70.5
75.7
49.5
55.0
73.6
67.2
73.4
79.6
72.9
65.3
80.6
78.3
75.1
68.5
84.6
70.7
69.1
--
21.7
~
--
s
0.09
0.09
0.10
0.11
0.11
0.09
0.09
0.11
0.11
0.09
0.09
0.10
0.09
0.09
0.09
0.09
0.09
0.09
--
0.40
~
~
Zn-Sorbed
Ferrihydrite
26.2
29.5
24.3
13.6
45.0
26.4
32.8
26.6
20.4
27.2
34.7
19.5
21.7
24.9
31.5
15.4
29.3
30.9
42.8
30.8
24.9
75.0
s
0.09
0.09
0.09
0.17
0.10
0.09
0.09
0.10
0.09
0.09
0.09
0.03
0.09
0.09
0.09
0.09
0.09
0.09
0.40
0.60
0.39
0.50
Smith-
son ite
--
-
-
--
-
~
-
~
~
-
-
~
-
-
-
-
-
-
--
-
12.4
9.0
s
-
~
-
-
-
-
-
-
-
-
-
-
-
--
-
-
--
-
--
-
0.29
0.50
Zn(OH)2
~
--
-
36.9
--
-
-
-
-
-
-
-
~
-
-
-
--
-
57.2
47.4
62.8
15.9
S
~
"
-
"
-
~
"
-
~
--
-
~
~
"
~
~
~
0.40
0.50
0.49
0.6
F-
Test
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Residual
1.82
1.42
1.65
2.64
2.22
1.50
1.81
1.72
1.28
1.45
1.51
1.41
1.58
1.71
1.88
1.29
2.26
1.89
1.59
1.81
2.38
1.68
s = standard deviation
62
-------
Appendix C
Description of Methods for Data Collection and Analysis to Determine Element Speciation
Employing X-ray Absorption Spectroscopy
Introduction
X-ray Absorption Spectroscopy (XAS) refers to a technique that employs analysis of X-ray Absorption Fine-Structure
(XAFS) to determine structural and chemical speciation of elements in various matrices. At characteristic energies
for a given element, absorption of an x-ray results in ejection of a core level electron. The ejected photo-electron may
subsequently be scattered from neighboring atoms and interfere with itself resulting in modulation of the energy of the
photo-electron and, therefore, the absorption event (http://cars9.uchicago.edu/xafs/xas_fun/xas_fundamentalapdf). In
practice, XAFS involves probing and analyzing the modulation of the x-ray absorption probability of an atom due to the
chemical and physical state of the atom. XAFS spectra are especially sensitive to the formal oxidation state, coordination
chemistry, and the interatomic distances, coordination number, and species of the atoms in the surrounding proximity
of the selected element of interest. As a result, XAFS provides a practical and straight-forward way to determine the
chemical state and local atomic structure for a selected atomic species. XAFS can be used in a wide variety of systems
and bulk physical environments.
Current practice for collection of XAFS data usually involves the use of a synchrotron as the source of x-rays. The flux
of x-rays from a synchrotron facility is sufficiently high to prevent the need to conduct experiments in a vacuum or the
removal of water. Thus, an important aspect from an environmental perspective is that XAFS can be used as an in-situ
Spectroscopy allowing for the investigation of samples in their natural state. Another unique aspect of this technique
is that it provides a means to collect element-specific data. Since each element absorbs x-rays at unique and discrete
energies, it is possible to selectively collect chemical and structural data for a specific element. In addition, XAFS probes
the short-range structure of a substance and is, therefore, not limited to analysis of materials possessing long-range
structural order, e.g., as required for x-ray diffraction analysis. Thus, XAFS measurements can be used to probe the
chemical and structural speciation of noncrystalline material, disordered compounds, and solutions feasible; matrices
that are most relevant to environmental systems.
Though XAFS measurements can be operationally simple, interpretation of XAFS data involves a mixture of modern
physics and chemistry, and a complete mastery of the data analysis can be somewhat challenging. Though the basic
phenomena are well understood, an accurate theoretical treatment is fairly involved, and in some respects, still an area
of active research.
X-Ray Absorption Fine Structure Spectroscopy
X-ray Absorption Fine-Structure (XAFS) is the modulation of the x-ray absorption coefficient at energies near and above
an x-ray absorption edge. XAFS is also referred to as X-ray Absorption Spectroscopy (XAS) and is subdivided into
2 regimes (Figure C.1):
XANES - X-ray Absorption Near-Edge Spectroscopy
EXAFS - Extended X-ray Absorption Fine-Structure
that contain related, but slightly different, information about the local coordination and chemical state of an element.
X-rays (light with wavelength 0.03 12 A or energy 1 E 500 keV) are absorbed by all matter through the pho-
toelectric effect: An x-ray is absorbed by an atom, promoting a core-level electron (K, L, or M shell) out of the atom
and into the continuum. The atom is left in an excited state with an empty electronic level (a core hole). The electron
ejected from the atom is called the photoelectron. When X-rays are absorbed by the photoelectric effect, the excited
core-hole will relax back to a "ground state" of the atom. A higher level core electron drops into the core hole, and a
fluorescent X-ray or Auger electron is emitted.
X-ray Fluorescence: An x-ray with energy equal to the difference of the core levels is emitted.
Auger Effect An electron is promoted into the continuum from another core level.
X-ray fluorescence occurs at discrete energies that are characteristic of the absorbing atom and can be used to identify
the absorbing atom.
63
-------
e-
o
(A
.Q
XANES
EXAFS
12.8 13 13.2 13.4 13.6
Energy (keV)
13.8
14
Figure C.1 Location of XANES and EXAFS regions of an XAS spectrum. The characteristic energy required to excite
core level electrons is unique to each element and is known as the absorption edge or edge step.
The intensity of an X-ray beam as it passes through a material of thickness, t, is given by the absorption coefficient,
M'-
I = I0e-jjt
where I0 is the X-ray intensity hitting the material, and I is the intensity transmitted through the material. The absorption
coefficient depends strongly on X-ray energy, E, and atomic number, Z, and on the density, , and atomic mass, A:
u ( Z4)/(AE3)
In addition, u has sharp absorption edges (Figure C.1) corresponding to the characteristic core level energies of the
atom. The energies of the K-edge absorption edges go roughly as EK ~ Z2. All elements with Z>18 have either a K- or
L-edge energies between 3 and 35 keV, which can be accessed at many synchrotron sources.
X-ray Absorption Spectroscopy Data Collection
While room-sized accelerators exist for conducting XAS studies, the intensity pales in comparison to accelerators
found at synchrotron radiation facilities such as Department of Energy National Laboratories (Figure C.2). Figure C.2
shows the Advanced Photon Source (APS) synchrotron radiation research facility at Argonne National Laboratory in
the southeastern suburbs of Chicago, IL. Applying components of Figure C.2B to explain structures in Figure C.2A, the
operation of the APS synchrotron facility entails (A) production of electrons in a linear accelerator which are deposited
into (B) the booster/injector ring to bring the electron packets near the speed of light. The electron beam is then sent to
the storage ring (C) from which beamlines (D) as either insertion devices, ID, or bending magnets, BM, are constructed
for experimental research (E). In-line with the research beamlines are monochromaters that tune the electron beam to
selected energies via Bragg diffraction and must be capable of energy resolutions of ~ 1 eV at 10 keV.
XAS data collection can be divided into two realms: 1) configuration of beamline equipment and 2) sample preparation.
XAS measures the energy dependence of the x-ray absorption coefficient, u(E), at and above the absorption edge of
a selected element. u(E) can be measured in two experimental configuration setups:
Transmission: The absorption is measured directly by measuring what is transmitted through the sample
(Figure C.3):
u(E)t -
Fluorescence: The re-filling of the deep core hole is detected. Typically the fluorescent x-ray is measured
(Figure C.4).
64
-------
B
Figure C.2 Aerial photo and architectural diagram of the Advanced Photon Source at Argonne National Laboratory,
Chicago, IL.
Q
X-ray
Source
L Detector
I Detector
Monochromator
Sample
Figure C.3 Experimental configuration for transmission data collection.
monochromator
slits
Ion
Chambers
Figure C.4 Experimental configuration for fluorescence data collection.
Sample
65
-------
Upon commencing an XAS scan to collect data, one has the option to use step-scan, continuous scan, or dispersive
XAFS modes. Typically, step-scan is used so that regions can be defined, and specific time can be allocated at step
to ensure adequate counts for statistical purposes.
Typical scan parameters for spectral regions for environmental samples include (Figure C.5):
A. Sample pre-edge to get background trend
(Range ~ -100 eV to -20 eV, 5 eV sampling)
C. Sample edge region
(-20 eV to +40 eV, 1 eV)
C. EXAFS region
Uniform in k-space (to > 12 A\ sampling 0.05 A1)
Prefer increased integration time per point at high k
Sample preparation can be as crucial as data collection setup when attempting to collect quality data. Although much
imagination can be employed for sample preparation, many considerations must be examined to ensure that the integrity
of the sample is maintained. For example, with redox sensitive materials such as arsenic, one must be mindful of pos-
sible oxidation of arsenite either from oxygen in the air, as a result of materials in the sample holder, and over exposure
of the electron beam. Another important issue to evaluate for sample preparation is the adequacy of the beamline one
chooses to conduct the experiments. This can be overcome by first working out the absorption lengths of the mate-
rial at the relevant energies. One should check for beamlines with the needed energy range and focal properties for
the intended samples. If the X-rays can get through the sample with only a few absorption lengths of attenuation, i.e.
small, homogeneous particle sizes, one can consider transmission data collection which can be superior to fluorescence
measurements. However, the edge step must be large enough for a transmission measurement which is hampered by
concentration of the element of interest. Thus, if the sample is dilute or inhomogeneous, fluorescence data collection
is better and most often employed.
If the energy is too low, absorption from air and windows can be a problem (a general rule is absorption energy de-
creases as atomic weight decreases, so lighter elements such as chromium can have artifacts which heavier elements
such as arsenic can avoid) but can be overcome by enclosing the sample and detection chambers in a non-ionizing gas
environment. Once an accurate understanding of detector limitations is accomplished for sample preparation, sample
mounting should be designed with simplicity in mind.
O
t»
•"\_**
12.8
13
13.2 13.4
Energy (keV)
13.6
13.8
14
Figure C.5 Standard raw XAFS spectra illustrating the three regions: (A) pre-edge, (B) edge step, and (C) EXAFS.
The XANES portion of the spectra contains all of edge step and small portions of the pre-edge and EXAFS
regions (Figure C. 1).
66
-------
Placement of the sample within an experimental hutch at a synchrotron facility is outlined in Figures C.3 and C.4, for
transmission and fluorescence data collection, respectively. In both setups, the sample is placed after the I0 detector
which measures the intensity of the beam before interaction with the sample. The only difference between sample
mounting for transmission and fluorescence data collection is that the sample must be situated at a 45° angle to the
incident electron beam for fluorescence mode (Figure C.4) to allow maximum fluorescence detection to occur and mini-
mize elastic scattering. As such, a fluorescence detector for fluorescence data collection must be placed perpendicular
to the electron beam and in line with the sample (Figure C.4). While transmission data can be collected from samples
offset of the penetrating beam, data quality is theoretically better by allowing the beam to pass directly through the
sample placed perpendicular to the electron beam (Figure C.3) with a suitable detector aligned with the electron beam
(Figure C.3) behind the sample.
Sample holders can vary significantly but follow a basic design. Sample holders can be as simple as smearing a solid
material on the tacky side of tape and folding the tape back onto itself to secure the sample (Figure C.6). Another com-
mon sample holder is a small block of non-reactive material with a depression or hole lathed into the holder to which
the sample is added with tape securing the opening (Figure C.7). This type of sample holder allows analysis of solids,
slurries, and solutions. The thickness and size of openings can be customized. An adaptation of this sample holder
in conjunction with a movable sample stages and automation software yields a basic autosampler (Figure C.8) which
permits all types of samples for analysis.
Figure C.6 Solid sample sandwiched between pieces Figure C.7
of Kapton tape.
Sample backfilled into opening of a Teflon
block and secured with tape.
Figure C.8 Autosampler template for multi-sample analysis.
67
-------
Traditional spectroscopic sample holders, such as thin films and thin sections, are often employed as well for special-
ized projects, but the above mentioned methods are more suitable for typical environmental analyses. However, some
of the traditional holders are being examined for the purpose of biological specimens and modified approaches can be
used to deal with samples with unusual physical characteristics (Figure C.9).
Figure C.9 Fluorescence data collection of metal hyperaccumulation in plant leaves.
Extended X-Ray Absorption Fine-Structure Data Analysis
Whether transmission or fluorescence data are collected, the data reduction and analysis are essentially the same. The
steps to data analysis are 1) reduce the raw spectra (Figure C.10) to k-space (conversion of energy [eV] to wavelength
[inverse distance]), 2) apply a Fourier transform to convert the k-space data into R-space (conversion of wavelength to
actual distance), and 3) XAFS data modeling.
12.8 13
13.2 13.4 13.6
Energy (keV)
13.8
14
Figure C.10 Raw fluorescence data Pb sorption on ferrihydrite
68
-------
Background correction of a raw spectrum involves assigning a baseline value of zero to the pre-edge region and a
normalized unity value of one to the EXAFS region (Figure C.11).
o
1
II 1 1 1 1
12.8 13 13.2 13.4 13.6 13.8 14
Energy (keV)
Figure C.11 Background corrected spectra of Figure C. 10.
The conversion of energy (E) to k-space (Figure C.12) involves first the identification of the threshold energy, Eo, which
is the energy maximum of the edge step. Thereafter, one isolates the EXAFS region in terms of the wave behavior of
the photoelectron (k) created in the absorption process by the equation:
k =
_ J2m(E-E0)
h2
where E is energy, Eo is the absorption edge energy, is Planck's constant, and m is the electron mass.
B-
o
-7
-2
13
Figure C.12 k-space conversion of spectra in Figure C.11.
69
-------
By selecting and emphasizing (k-weighting) only positive k-space, one generates oscillations as a function of the pho-
toelectron wave number also known as the (chi) (k)-function (Figure C.13). The oscillations in Figure C.13 correspond
to different near-neighbor coordination shells which, after Fourier transformation (Figure C.14), can be described by
the EXAFS equation:
tf,/,(*X"v'.
*«;
Where f(k) and (k) are scattering properties of the atoms neighboring the excited atom, N is the number of neighboring
atoms, R is the distance to the neighboring atom, and 2 is the disorder in the neighbor distance. Though complex in
appearance, the EXAFS equation allows one to determine N, R, and z if the scattering amplitude, f(k), and phase-shift,
(k), are known. Since f(k) and (k) depend on the Z of the neighboring atoms, EXAFS is also sensitive to the atomic
species of the neighboring atom.
10
Figure C.13 The ^-weighted -function of Figure C. 12. The oscillations describe the photoelectron wave number as
the photoelectron constructively and destructively interacts with neighboring atoms around PC.
Now that the energy spectrum in Figure C.11 is converted to k-space (Figure C.13), one can Fourier transform the
k-space data in R-space (distance). The Fourier transformation is critical to XAFS analysis and is often the area of
confusion for novice data analyzers. Since the photoelectron effect causes backscattering in XAFS data collection, a
phase shift causes real values of distance to be offset at least 0.5 A when determining interatomic bond distances in
the data modeling techniques. For example, the protruding peak located at approximately 1.8 A in Figure C.14 for the
Fourier transformed k-space data is actually determined to be 2.53 A in the modeling procedure to follow.
Figure C.14 Radial distribution (or structure) function of Fourier transformed k-space data for Figure C. 13.
70
-------
The final step in EXAFS data analysis is modeling which is often the most complex and time consuming aspect of the
procedure. The most common method of modeling is to develop a library of fitting paths generated crystallographic pa-
rameters. Figure C.15 shows the crystalline unit cell parameters for magnetoplumbite, an iron oxide with small amounts
of Pb in the structure. A commercially available software program, ATOMS, is available for this purpose to record what
and where the information was derived (title and notes sections), space group identity and unit cell angles (abc) of the
phase, the ab-initio calculation sphere in angstroms (rmax), which element to centralize the calculation around (core),
and the 3-dimensional positions (xyz) of each element with corresponding tags (Pb1, Fe2, etc).
This ATOMS input file is used to generate possible coordinations of elements within magnetoplumbite as a FEFF
file (commercially available software) (Figure C.16). (NOTE: The XAS research user community is rather small, and
many software packages are available free of charge from the international experts.) The FEFF program then uses
this information to generate, via ab-initio calculations, fitting paths to describe the interatomic bond distances (R) and
coordination numbers (N) with neighboring elements.
title name: magnetoplumbite
title formula: PbFe_12O_19
title sites. Pb1,Fe1-5, O1-5
title refer!: Moore etal.(1989) Am. Mm. 74,1186-1194
title refer?:
title schoen.
title notesl Pb2+, Fe3+ oxidation state
title notes2: modified from Moore et al's structure'
title notesS: see notes below
space p 63/m m c
a =5.873 c =23007
rmax = 6.00
core = Pb1
atom
Pb 0.66667 0.33333 025000 Pb1
Fe 000000 0.00000 0.00000 Fe1
Fe 000000 0.00000 0.25000 Fe2
Fe 033333 0.66667 0.02730 Fe3
Fe 0.33333 0.66667 0.19000 Fe4
Fe 0.16900 -0.16900 -0.10900 Fe5
O 0.00000 000000 015100 O1
O 0.33333 0.66667 -0.05500 O2
0 0.18400 -0.18400 0.25000 O3
O 0 15500 -0 15500 0.05200 O4
O 0.50400 -0.50400 0.15000 O5
% notes'
% Moore et al put the Pb at (0.72,0.384,1/4), a site with 12j symmetry,
% and have an occupation of 1/6 for this site. This is a large distortion
% of the Pb environment from the "normal" (2/3, 1/3, 1/4) position (with 2d
% symmetry) used here.
% Since fractional occupation is poorly defined in the context of local
% structure, it isn't supported in atoms. So the point of high symmetry is
% needed (otherwise unphysically short Pb-Pb bonds are made).
% similarly, the Fe2 position has been moved from (0,0,0.256) with an
% occupation of 1/2 to (0,0,1/4) with full occupation.
% the Pb-O and Fe-O near-neighbor distances for all 5 Fe sites agree
% fairly well with Moore et al.
TITLEname magnetoplumbite TITLEfomnula PbFe 12O 19 TITLE sites Pb1 Fe1-5 O1-5
TITLErefer! Moore etal (1989) Am Mm 74 1186-1194 TITLE refer2 TITLE schoen TITLE notesl
Pb2+. Fe3» oxidation state TITLE notes2 modified from Moore et al's structure TITLE notesS see
notes below
HOLE 4 10 ' Pb L3 edge (13035 0 eV), second number is S0*2
" mphase mpath mfeff mchi
CONTROL 1111
PRINT 1113
RMAX 60
"CRITERIA curved plane
"DEBYE temp debye-temp
"NLEG 8
POTENTIALS
" ipot Z element
0 82 Pb
1 82 Pb
2 26 Fe
3 8 O
ATOMS
* x y
0 00000
1 65468
-0 82737
1 65468
-0 82737
-0 82737
-0 82737
263123
-0 17634
263123
-2 45494
-2 45494
-0 17634
1 69537
-3 39080
1 69537
0 83581
0 83581
-1 67167
0 83581
-1 67167
0 83581
3 39074
-t 69542
3 39074
-1 69542
-1 69542
-1 69542
"this list contains 84 atoms
0 00000
0 00003
-1 43298
0 00003
-1 43298
1 43304
1 43304
-1 31552
-2 93647
1 31558
-1 62092
1 62098
2 93653
-2 93647
0 00003
2 93653
-1 44767
-1 44767
0 00003
1 44772
0 00003
1 44772
0 00006
-2 93644
0 00006
-2 93644
2 93656
2 93656
ipot tag
0 00000
2 30070
2 30070
-2 30070
-2 30070
2 30070
-2 30070
0 00000
0 00000
0 00000
0 00000
0 00000
0 00000
0 00000
0 00000
0 00000
3 24399
-3 24399
3 24399
3 24399
-3 24399
-3 24399
1 38042
1 38042
-1 38042
-1 38042
1 38042
-1 38042
0
3
3
3
3
3
3
3
3
3
3
3
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
distance
Pb1
O5 1
05" 1
OS 1
05~1
05 ~2
05~2
O3 1
O3 1
O3 2
O3 2
O3 3
03 3
Fell
Fe2 2
Fe2 2
FeS 1
FeS 1
FeS "2
FeS 2
Fe5 2
FeS 2
Fe4 1
Fe4~1
Fe4 1
Fe4"1
Fe4 2
Fe4 2
0 00000
2 83394
2 83394
2 83394
2 83394
2 83397
2 83397
294176
294176
294179
2 941 79
294182
294182
3 39074
3 39080
3 39080
3 64935
3 64935
3 64937
3 64937
3 64937
3 64937
3 66097
3 66097
3 66097
3 66097
366106
366106
Figure C.15 ATOMS input file for magnetoplumbite
listing crystallographic information.
Figure C.16 FEFF file used to determine fitting paths for
EXAFS modeling showing the interaction of
a central Pb atom (Pb1) with two different
oxygen atoms (O5andO3) and three different
iron atoms (Fe2, FeS, and Fe4).
Once theoretical fitting paths are generated, one can return to the radial distribution function (Figure C.14) to conduct
the actual fitting protocol. Again, using one of several software packages for this procedure allows one to evaluate real
data collected at a synchrotron facility relative to fitting paths from model data. An overlay of the resulting model fit to
the actual data shows the goodness of the model parameter fitting and determines the relevant information to under-
stand the overall system in terms of the coordination environment (R and N) (Figure C.17). Figure C.17 shows that
Pb is octahedrally coordinated (N=6) with a bond distance of 2.53 A to oxygen (Pb-O shell: oxygen is the first nearest
neighbor) followed by Pb-Fe shell indicating a coordination number of approximately 2 and an interatomic bond distance
of 3.67 A. The Pb-Fe data suggest a bidentate (2 bonds) innersphere (oxygen between the Pb and Fe atoms) sorption
complex for Pb sorption on ferrihydrite.
71
-------
Pb-O Shell, N=6.1,R=2.53A
o>
•o
a
'E
o>
ra
ra
o
Pb-Fe Shell, N=2.3, R=3.67A
4
R[A]
Figure C.17 Structural data derivedfrom ab-initio calculated fitting paths for Pbsorption on ferrihydrite. The red curve is
the sample data reduced from a raw spectrum, and the blue dots represent the fit from the modeled pa ths.
The fitting data can then determine coordination numbers (N) and interatomic bond distances (R).
X-Ray Absorption Near Edge Spectroscopy Data Analysis
The XANES portion of the spectrum (Figure C.1) is a much larger signal than EXAFS which allows data collection for
samples with lower concentrations and less than ideal sample conditions. The XANES portion is typically limited to
within the range of -100 eV to 300 eV. These values vary depending on the element of interest and shape of the spec-
trum. The interpretation of XANES is complicated by the fact that there is not a simple chemical or physical description
of the spectrum since XANES is uniquely different for each element; a fingerprint as an analogy. However, there is
significant chemical data in the XANES region, notably the formal valence and coordination environment. The valence
is identifiable by the position of the maximum edge energy by taking the derivative of the XANES region (Figure C.18).
Typically, the energy difference between oxidation (valence) states of a given element can range significantly. For
example, identification of arsenite (As3+) and arsenate (As5+) can be easily distinguished (Figure C.19). In the case of
aqueous Zn and ZnS in Figure 18, Zn is in the divalent (2+) valence; however, there is a slight shift (1 eV) of the ZnS
spectrum to a lower energy versus the Zn-O coordination of aqueous Zn. One can clearly notice distinctly different
shapes of the Zn XANES spectra which allow for understanding of the influence of ligand type (first nearest neighbor)
(Zn-O and Zn-S) on the coordination environment. For the As species in Figure C.19, oxygen is the nearest neighbor
for both, resulting in similar spectra but offset in E due to different oxidation states. Note that the higher oxidation state
of arsenate is at a higher E value than arsenite. This is intuitively correct when considering an oxidized atom will exert
more effort to hang on to remaining electrons.
While analysis of XANES data can provide information on the oxidation state and the first coordination shell, the struc-
tural information is much more limited than what is accessible via EXAFS analysis. However, not all cases require one
to know the level of information provided by EXAFS. XANES data analysis is significantly less labor intensive and in
some cases can provide the level of information necessary to answer research questions.
Data Analysis of Complex, Heterogeneous Samples
In many instances for environmental samples, the speciation of metals can result in multiple phases. This can make data
analysis difficult. For example, EXAFS data analysis is providing information on the average coordination numbers and
bond distances for a given shell. When multiple species are present, these parameters are organized into one value that
does not represent the complexity of the metal species. The same problem arises when interpreting the coordination
environment of metals for XANES data analysis. To overcome this issue, one can apply a statistical fitting procedure
that seeks to strip the multiple components of a sample spectrum into individual parts through the assistance of known
reference spectra. The two most common methods are linear combination fitting (LCF) and principle component analy-
sis (PCA). LCF analysis of XANES and XAFS spectra (LCF-XANES and LCF-XAFS) is simple to apply to normalized
XANES spectrum or the k2or3-weighted -function from EXAFS data reduction. The goal in this procedure is to accu-
72
-------
Figure C.18 XANES spectra (thin line) and derivative of XANES spectra (thick line) for aqueous Zrf+ (blue) and ZnS
(red).
c
o
".5
e-
o
(A
JO
Arsenate (AsV)
Arsenite (Aslll)
11.85
11.87
11.89 11.91
E (keV)
11.93
11.95
Figure C.19 XANES spectra for arsenite [As(lll) - blue] and arsenate [As(V) - red].
mulate enough relevant reference spectra that can explain and represent the unknown environmental sample. Through
the use of available software, one selects an unknown spectrum to evaluate and multiple known reference spectra to fit
against the unknown. By repeating the procedure and removing nonessential reference spectra, one can gain a semi-
quantitative analysis of the major metal species present in the unknown sample (Figure C.20). Detailed information as
collected in EXAFS analysis is not possible, but identity of multiple species in the sample is accomplished.
For typical environmental samples, the use of finger-printing methods such as LCF-XANES and LCF-XAFS can be a
very powerful tool to determine metal speciation when multiple phases are present. This approach proves very effec-
tive in monitoring contaminated sites to evaluate changes in metal speciation either through in-situ amendments or
monitored natural attenuation.
73
-------
00 —
»8
'\
«30
— LEAST -SQUARES FIT ( scute
— FILE Sedmestl tut
Iterations
(CHI) '2
Residual
Nc F-Test i
1541-
I*. par'ial r
*; EC shit'
J ^11935
-0 10«04<>5i<>
partial c
*4 EC shift
1C xaaes - F
*S partial i
*t> EC shut
ibscr;.'-; en r^rnsc'xon fox 1C X&HSS Fx*
~pecic *l tZni ^ ~ 1 *)?"!*?
Spec le <» [ fexr xisvdri teJ^Hn } c 1 > * Jj
Spec:* «? [Fr«n>linit*; c -3 * 3 0^04
Figure C.20 Linear combination fitting of X-ray absorption near edge spectroscopy data (LCF-XANES) for a sediment
(Sedimentl) sample with multiple Zn species. The red curve is the sample data, and the blue curve is the
fitted result. The LCF-XANES results indicate Zn speciation as 87% ZnS, 4% Zn sorbed to ferrihydrite,
and 9% ZnFe2O4 (franklinite).
------- |