EPA/600/R-94/018
February 1994
ABIOTIC TRANSFORMATION OF CARBON TETRACHLORIDE
AT MINERAL SURFACES
by
Michelle Kriegman-King and Martin Reinhard
Department of Civil Engineering
Stanford University
Stanford, California 94305-4020
CR-816776
Project Officer
Stephen R. Hutchins
Processes and Systems Research Division
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ADA, OKLAHOMA 74820
-------
TECHNICAL REPORT DATA
(Pieaxe read Instructions on the reverse before complel
1. REPORT NO.
2.
•4. TITLE AND SUBTITLE
ABIOTIC TRANSFORMATION OF CARBON TETRACHLORIDE
AT MINERAL SURFACES '
PB94-144698
5. REPORT DATE
Fohrnarv 1 QQA
6. PERFORMING ORGANIZATION COOE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
Michelle Kriegman-King and Martin Reinhard
9. PERFORMING ORGANIZATION NAME ANO AOORESS
Department of Civil Engineering
Stanford University
Stanford, CA 94301-4020
10. PROGRAM ELEMENT NO.
AC4C1A
11. CONTRACT/GRANT NO.
CR-816776
12. SPONSORING AGENCY NAME ANO AOORESS
Robert S. Kerr Environmental Research Lab. - Ada, OK
U.S. EPA
P.O. Box 1198
Ada OK 74820
13. TYPE OF REPORT ANO PERIOD COVERED
Final Report 09/90 - 09/93
14. SPONSORING AGENCY CODE
EPA/600/15
15. SUPPLEMENTARY NOTES
Project Officer Stephen R. Hutchins 405/436-8563
16. ABSTRACT
Transformation of carbon tetrachloride (CCIJ by biotite, vermiculite, and pyrite in the presence of hydrogen sulfide (HS")
was studied under different environmental conditions. In systems containing biotite and vermiculite, the rate of CCI4
transformation was dependent on the temperature, HS" concentration, surface concentration, and Fe(ll) content in.the
minerals. At 25°C, .the half-life of CCI4 with 1 mM HS' was calculated to be 2600, 160, and 50 days for the homogeneous,
vermiculite (114 m2/L) and biotite (55.8 m2/L) systems, respectively. The transformation rate with biotite and vermiculite
was nearly independent of pH in the range 6 - 10 at constant HS" concentration. The rate dependence on Fe(ll) content of
the sheet silicates suggested that the transformation occurs at surface sites where HS" is associated with Fe(ll).
CCI4 reacted relatively rapidly in 1.2-1.4 m2/L pyrite with >90% of the CCI4 transformed within 12-36 days at 25°C. The
observed rate law supports a heterogeneous reaction mechanism. The reactivity of CCI4 with pyrite increased in the order
air-exposed pyrite/aerabic, air-exposed pyrrte/HS", air-exposed pyrite/anaerobic, and acid-treated pyrite/anaerobic but
overall varied only by a factor of 2.5 The CCI4 transformation products were observed to vary under different reaction
conditions. Approximately 80-85% of the CCI4 was transformed to CS2 which hydrolyzed to C02. Only 5-15% of the CCI4
was reduced to CHCI3. In the pyrite systems, CO2.was the major transformation product formed under aerobic conditions
whereas CHCIS was largely formed under anaerobic conditions. Formation of some CS2 was observed in all pyrite
systems.
17.
KEY WORDS ANO DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
COS ATI Field. Group
ORGANICS ANAEROBIC
AQUIFER KINETICS
ABIOTIC
TRANSFORMATION
MINERAL
CARBON TETRACHLORIDE
HYDROGEN SULFIDE
BIOTITE
VERMICULITE
PYRITE
PATHWAY
18. DISTRI3UTION STATEMENT
RELEASE TO THE PUBLIC
19. SECURITY CLASS
-------
DISCLAIMER
The information in this document has been funded wholly or hi part by the United
States Environmental Protection Agency under cooperative agreement CR-816776 to Stanford
University. The 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
environmentally related measurements and funded by the Environmental Protection Agency
are required to participate in the Agency Quality Assurance Project Plan. The procedures
specified hi 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.
11
-------
FOREWORD
EPA is charged by Congress to protect the Nation's land, air, and water systems.
Under a mandate of national environmental laws focused on air and water quality, solid
waste management and the control of toxic substances, pesticides, noise and radiation, the
Agency strives to formulate and implement actions which lead to a compatible balance
between human activities and the ability of natural systems to support and nurture life.
The Robert S. Kerr Environmental Research Laboratory is the Agency's center of
expertise for investigation of the soil and subsurface environment. Personnel at the
Laboratory are responsible for management of research programs to: (a) determine the fate,
transport, and transformation rates of pollutants in the soil, the unsaturated and the saturated
zones of the subsurface environment; (b) define the processes to be used in characterizing the
soil and subsurface environment as a receptor of pollutants; (c) develop techniques for
predicting the effect of pollutants on ground water, soil and indigenous organisms; and (d)
define and demonstrate the applicability and limitations of using natural processes, indigenous
to the soil and subsurface environment, for the protection of this resource.
Chlorinated solvents are among the most pervasive of ground-water contaminants and
represent a significant fraction of total contamination incidents. This research was conducted
to define rates of abiotic (chemical) transformation and fate mechanisms for carbon
tetrachloride in the presence of natural mineral surfaces in aquifer systems. The data show
that natural mineral surfaces can enhance the rate of degradation of carbon tetrachloride
several orders of magnitude beyond that observed in homogeneous systems. This research
demonstrates the importance of abiotic reactions in reducing concentrations of chlorinated
solvents, especially in the presence of mineral surfaces typically found in aquifers. These
types of reactions need to be considered and addressed in the use of models for predicting
transport and fate of haloaliphatic compounds in the subsurface.
Clinton W. Hall
Director
Robert S. Kerr Environmental
Research Laboratory
111
-------
ABSTRACT
This report addresses the ability of natural mineral surfaces to abiotically transform
halogenated organic compounds in subsurface environments. The research focuses on carbon
tetrachloride as the halogenated organic and biotite, vermiculite, and pyrite as the mineral
surfaces. The CCU transformation rates and products were quantified under different
environmental conditions.
The disappearance of CCU was significantly faster in the presence of mineral surfaces than
in homogeneous solution. In systems containing the sheet silicates and HS', the rate of reaction
was dependent on the temperature, HS' concentration, surface concentration, and Fe(II) content
in the minerals. At 25 °C, the half-life of CCU with 1 mM HS- was calculated to be 2600, 160,
and 50 days for the homogeneous, vermiculite (114 m2/L) and biotite (55.8 m2/L) systems,
respectively. The transformation rate with biotite and vermiculite was independent of pH over
the range of 6 to 10 at constant [HS*]. The heterogeneous transformation of CCU with HS' and
biotite was first order with respect to the biotite surface concentration up to 280 m2/L. Above
0.5 mM HS' and 55.8 m2/L biotite, the CCU transformation rate was independent of [HS']
suggesting that the reaction is heterogeneous. In addition, the reactivity of CCU with Aerosil
200 (amorphous silica)/HS' and the rate dependence on Fe(H) content of the sheet silicates
suggest that the transformation occurs at sites where HS' is associated with Fe(n).
was observed to be relatively reactive with 1.2-1.4 m2/L pyrite under all reaction
conditions studied with >90% of the CCU transformed within 12-36 days at 25 °C. A zeroth-
order rate dependence on the CCU concentration in the presence of 1.2 m2/L pyrite supports a
heterogeneous reaction mechanism. The reactivity of CCU with pyrite increased in the order:
air-exposed pyrite/aerobic, air-exposed pyrite/HS', air-exposed pyrite/anaerobic, and acid-treated
pyrite/anaerobic. The transformation rates only varied by a factor of 2.5 for all the conditions
studied. An iron oxide coating, identified to be FeOOH using x-ray photoelectron spectroscopy,
was detected on the pyrite that had been reacted aerobically. It was likely that this coating is
responsible for the slower transformation rates observed under aerobic conditions.
The CCU transformation products were observed to vary under different reaction
conditions. In the sheet silicate/HS- systems, a new pathway to form CO2 was identified.
Approximately 80-85% of the CCU was transformed to C$2 which hydrolyzed to CO2. Only 5-
15% of the CCU was reduced to CHCls. In the pyrite systems, CO2 was the major
transformation product formed under aerobic conditions whereas CHCls was largely formed
under anaerobic conditions. Formation of some CSz was observed in all pyrite systems,
suggesting that CCU or a reaction intermediate must react directly with pyrite-S.
This report was submitted in fulfillment of the grant EPA CR-8 16776 by the researchers in
the Environmental Engineering and Science program, Department of Civil Engineering, Stanford
University, under the sponsorship of the U.S. Environmental Protection Agency. This report
covers the period September 7, 1990 to September 6, 1993, and work was completed as of
September 6.
IV
-------
CONTENTS
SECTION 1 1
Research Objectives and Hypotheses 1
SECTION 2 , 3
SECTIONS 5
SECTION 4 6
Research Approach 6
Reference Solids, Pretreatments, and Reagents 6
Filling and Sealing of Ampules 7
Experiments with Sheet Silicates 7
Synthetic Solids Experiments 7
Experiments with Pyrite 8
Chemical Analyses 8
CCU and CHCls Analysis 8
CS2 Analysis 9
CO Analysis 9
Formate Analysis 9
Radiolabeled Products Analysis 9
Adsorbed Formate and CO2 10
Aqueous Pyrite Oxidation Products 10
Surface Pyrite Oxidation Products 10
Mineral Characterization 11
Sheet Silicate Characterization 11
Synthetic Solids Characterization 13
Pyrite Characterization 14
Kinetic Evaluation of Data 14
Adsorption 15
SECTION 5
Transformation of CCLt by Sheet Silicates 16
Effect of Biotite and Vermiculite on CCU Transformation 16
-------
CCU Transformation Products 17
Temperature Dependence 22
pH Dependence 23
Effect of Solids Concentration 24
Effect of Sulfide Concentration 25
Effect of Ferrous Iron Content 26
Summary of Sheet Silicate Results 27
Transformation of CCU by Synthetic Solids 28
Transformation of CCU by Pyrite 29
Effect of Pyrite Pretreatment on CCU Transformation Rate 31
CCU Transformation Products 34
Pyrite Oxidation Products 39
Proposed Mechanism at Pyrite Surface 43
Summary of Pyrite Experiments 45
REFERENCES 46
VI
-------
LIST OF FIGURES
Figure 1: XPS spectra of the Fe 2p peak for biotite reacted for 6 months in anaerobic
Milli-Q water and 0.1 M HS". No differences are observed in the
oxidation state of iron in the two samples. Peaks are charge shifted by
approximately 7.0 eV [[[ . ................... 12
Figure 2: SEM micrograph of biotite cleavage sheet after reaction in anaerobic
Milli-Q water for 6 months at 25 °C.r [[[ 13
Figure 3: SEM micrograph of biotite cleavage sheet after reaction in 0.1 M HS' for
6 months at 25 °C [[[ 13
Figure 4: First-order plots of CCU transformation with biotite (SCbi0tite=55-8 m2/L)
and vermiculite (SCvennic=l 14 m2/L), [HS-]=1 mM, pH=8.6 at 50 °C ..... . ....... 17
Figure 5: CCU transformation products from reaction with [HS"]=lmM, SCbiotite=
55.8 m2/L, pH=8.8 at 50 °C. Lines represent best fit using eqs. 8, 12, and
13 [[[ : .............................. . ................................... is
Figure 6: Carbon disulfide (€82) was identified as an intermediate of CCU
transformation by the agreement between the CS2 fraction measured by
gas chromatography and the "unidentified" volatile fraction measured
from 14C experiments [[[ 19
Figure 7: Proposed CCU transformation pathways in HS- solution containing
biotite. Shadowed boxes are products and intermediates detected in this
study. Compounds in brackets are intermediates proposed from the
literature (as cited in Criddle and McCarty, 1991) and this report.
Reaction pathways with dashed arrows have been observed in other
studies (as cited in Criddle and McCarty, 1991) ................................................. 20
Figure 8: CCU transformation rate at 1 mM HS~, 50 °C, and 55.8 m2/L biotite or
114 m2/L vermiculite in the pH range of 6-10. Error bars are 95%
confidence intervals around k'homo or k'hetero ................................... • ............... ^4
Figure 9: The effect of biotite surface concentration (SCbiotite) on the CCU
transformation rate (k'heteto) at 50 °C. The best-fit line for the data at 1 1.2
and 55.8 m2/L passes through the origin. Error bars are 95% confidence
intervals... [[[ . .............................. 25
Figure 10: The effect of HS- concentration on the CCU transformation rate (k'tetero)
and the fit to the proposed rate law (eq. 3) with SCbiotite=55.8 m2/L at
50 °C. Above [HS~] - 0.5 mM, the rate shows a zeroth-order dependence
on [HS~] (b2=0). When [HS']<0.5 mM, b2 is estimated to be 1.34 or
-------
Figure 12: Zeroth-order dependence of the disappearance of CCU in the presence of
air-exposed pyrite reacted under aerobic and anaerobic conditions at
25 °C. The control represents the disappearance of CCU in a 1 mM NaCl
anaerobic solution 32
Figure 13: The disappearance of 100 mM CCU with 1.2 m2/L pyrite reacted
anaerobically at 25 °C. Appearance of formic acid was measured with an
ion chromatograph 35
Figure 14: Disappearance of CCU in the presence of pyrite under aerobic conditions
at 25 °C. Appearance of the products, CS2 and CO2, with model results
assuming the only path to form CO2 is CCU -> CS2 -> CO2 37
Figure 15: Disappearance of CC14 in the presence of pyrite under aerobic conditions
at 25 °C. Appearance of the products, CS2 and CO2, with model results
assuming the only path to form CO2 from CC14 is: CCU -> Intermed. ->
CS2 -> CO2 ...38
Figure 16: Disappearance of CCU in the presence of acid-treated pyrite under
anaerobic conditions at 25 °C. Appearance of the products, CS2 and CC>2,
with model results assuming the only path to form CC«2 is: CCU -> CS2 ->
CO2 39
Figure 17: XPS spectra comparing S 2p peak for pyrite reacted under anaerobic
conditions in the presence and absence of 1 mM CCU • 40
Figure 18: XPS spectra of Fe 2p peak of air-exposed pyrite reacted aerobically. The
Fe(in) peaks are indicative of an iron oxide coating on the pyrite 40
Figure 19: XPS spectra of O Is peak of air-exposed pyrite reacted aerobically. The
peaks at 530.1 and 531.2 eV coincide with the O Is peak for FeOOH
(Moulder etal., 1992) 41
Figure 20: XPS spectra of the S 2p peak comparing fresh-ground pyrite reacted
anaerobically with air-exposed pyrite reacted aerobically. There is no
significant 'difference in the S 2p peak shapes and binding energies under
these reaction conditions 42
Figure 21: Proposed CCU transformation pathways with pyrite. Compounds in
shadowed boxes are measured intermediates or products. Compounds in
brackets are proposed intermediates 44
Vlll
-------
LIST OF TABLES
Table 1: Specific surface area determined by BET and iron content for minerals and
synthetic solids 11
Table 2: Pseudo-first-order rate constants obtained for CCU transformation
pathways using eqs. 8,12, and 13 23
Table 3: Arrhenius parameters for CCU transformation with 1 mM HS" at 50°C.
Ea and InA were calculated using k'i,i,et for biotite and vermiculite and k'i
was used for the homogeneous systems 23
Table 4: Results of CCU transformation in systems containing HS', gibbsite and
Aerosil 200 at 50°C 29
Table 5: Zeroth-order rate constants for CCU transformation with pyrite reacted
under aerobic and anaerobic conditions at 25 °C 32
Table 6: Comparison of pyrite oxidation rate by CCU with literature rates of
oxidation by O2> Fe3"1", and H2O2 at room temperature 33
Table 7: CCU product distribution from reaction with pyrite under aerobic and
anaerobic conditions at 25 °C 34
Table 8: Rate constants for the disappearance of CCU and appearance of
intermediates and products from reaction with pyrite under aerobic and
anaerobic conditions at 25 °C 36
Table 9: Effect of environmental conditions on S:Fe(II) in the near-surface of pyrite
where S:Fe(II) was determined using XPS 43
IX
-------
ACKNOWLEDGMENTS
We thank Art F. White for suggesting that we study oxygenated systems in the pyrite
studies, David King for helping us solve differential equations and Stephen R. Hutchins for
serving as our Project Officer.
-------
SECTION 1
INTRODUCTION
Halogenated organic compounds are prevalent contaminants in ground water and are often
components of landfill leachate and hazardous waste. Recent reports suggest a link between
chlorinated hydrocarbons and increased incidence of many illnesses ranging from breast cancer
to endocrine disorders (Hileman, 1993). Although there are proposals to ultimately eliminate use
of many of these anthropogenic compounds in our society, their current widespread use
predicates the need to understand the fate of these compounds in subsurface environments. Until
recently, most studies of organic contaminants were conducted at the "system" level in which the
disappearance of these compounds in sediment slurries was investigated without an attempt to
understand the processes controlling the disappearance. In 1986, Macalady et al. promulgated
the need to understand the chemical, physical, and biological processes that potentially control
the fate of these compounds, ultimately leading to a predictive capability.
An example of a progression from the "system" level to the process level is the
transformation of hexachloroethane (HCA) by sediment samples from the Canadian Forces Base
in Borden, Ontario. Criddle et al. (1986) found that HCA is transformed to tetrachloroethylene
(PCE) in sterilized Borden sediment slurries. Subsequently Curtis (1991) identified the reactive
component of the Borden sand to be organic matter associated with the sediment fraction
containing quartz, feldspars, and carbonate minerals. Curtis (1991; Curtis and Reinhard, 1993)
further investigated the reactivity of humic substances and model compounds with HCA and
other haloaliphatic organics at the process level. Although this work does not allow prediction of
the fate of HCA under diverse environmental conditions, it did provide a deeper understanding of
the processes that may be controlling reactivity in some sediment systems.
In his studies to identify the reactive component in Borden sand, Curtis (1991) suggested
that ferrous iron bearing minerals may play a role in the transformation of haloaliphatic
compounds; Although they were not the major reductant in the case of Borden sand, ferrous iron
bearing minerals can play a major role in other environments. Anderson et al. (1992; 1994)
proposed that such minerals were responsible for the reduction of Cr(VI) to Cr(IH) in a sand and
gravel aquifer at Otis Air Force Base, Cape Cod, MA. Furthermore, at the same site, Barber et
al. (1992) found that chlorinated benzenes were adsorbed to organic matter associated with the
iron bearing minerals. If the haloaliphatics are associated with organic matter at iron bearing
phases, ferrous iron bearing minerals may be able to reduce haloaliphatic compounds at
environmentally significant rates.
Research Objectives and Hypotheses
The objectives of this research were to test the ability of ferrous iron bearing minerals to
abiotically transform haloaliphatic compounds in sulfidic environments and to identify the
organic's transformation products. The work focused largely on carbon tetrachloride as the
halogenated organic and biotite, vermiculite, and pyrite as the ferrous-iron bearing minerals.
CCU was chosen as the compound of interest because it is a contaminant frequently found in
ground water, it is a suspected human carcinogen (Sax and Lewis, 1987), it causes ozone
depletion, and it is a greenhouse gas (Rowland, 1991). Additionally, CCLt is known to react
relatively rapidly in the presence of reductants (for example, Castro and Kray, 1963; 1966).
Biotite and vermiculite are sheet silicates that are commonly found as detrital materials in
-------
sedimentary rocks (Deer et aL, 1982) and sulfidogenic conditions are often observed in plumes of
hazardous waste and landfill leachate (Barbash and Reinhard, 1989a). To obtain additional
insight into the reaction mechanism of CCU with the sheet silicates and sulfide, experiments
were conducted with muscovite, pyrite, and the synthetic solids, Aerosil 200 (amorphous silica)
and gibbsite. In-depth studies of pyrite reactivity were conducted because pyrite, an iron sulfide
mineral, is ubiquitous in sulfidic environments (Howarth and Teal, 1979; Luther et al., 1982;
Lord and Church, 1983) and is potentially the most significant environmental player of the
systems studied.
It was hypothesized that:
(1) CCU is transformed faster in the presence of mineral surfaces than in homogeneous
solution;
(2) the rate of transformation is a function of the reaction conditions such as the type of
mineral, iron content, pH, temperature, sulfide concentration, presence of oxygen,
and solids concentration;
(3) the CCU transformation products will vary under different reaction conditions.
The research approach used to address these hypotheses and an overview of the research project
are delineated below. Additional hypotheses such as the influence of natural organic matter,
competing oxidants, and co-solvents on the CCU transformation rates and products in
heterogeneous systems were not addressed in this project
-------
SECTION!
SUMMARY AND CONCLUSIONS
This research demonstrates, for the first time, that mineral surfaces may play a significant
role in the fate and transport of haloaliphatic compounds in subsurface environments. The
results of this work provide insight into the rates of CCU transformation under different
environmental conditions, the CCU transformation products, and the mechanism of CCU
transformation at mineral surfaces. The specific conclusions that can be drawn from this work
are as follows:
(1) The disappearance ofCCfy in sulfidic systems was significantly faster in the presence
of mineral surfaces (biotite, vermiculite, pyrite, and Aerosil 200) than in homogeneous solution.
In sulfide solutions containing biotite, vermiculite, or Aerosil 200, CCU was transformed an
order of magnitude faster than in homogeneous systems containing only HS*. At 25 °C, the half-
life for the disappearance of CCU in the presence of 1 mM HS' was 2600, 160 and 50 days for
the homogeneous, vermiculite (114 m2/L), and biotite (55.8 m2/L) systems, respectively. In
addition, the reaction of CCU with pyrite at 25°C was orders of magnitude faster than reactions
with 1 mM HS- or 0.1 mM Fe(E) in solution at 50°C. In the pyrite systems 50% of the CCU
reacted within 6-20 days at 25 °C. These results along with additional evidence suggest that the
with CCU is heterogeneous.
(2) The disappearance of CCU in the sheet silicate and Aerosil 200 systems followed first
order kinetics with respect to the CCU concentration. In the presence of pyrite, however, the
disappearance followed zeroth-order kinetics (MDL to 1 (iM) with respect to the CCU
concentration, indicating that reactive surface sites were saturated with CCU> and that the rate of
reaction was not dependent on the solution concentration. The possible reaction mechanisms of
CCU degradation at the Aerosil 200 interface by HS" remain to be investigated.
(3) The rate of transformation ofCCLf with the sheet silicates and sulfide depended on the
following reaction conditions: temperature, surface concentration, sulfide concentration, and
ferrous iron content in the minerals. The CCU transformation rate was investigated over the
range of 6-10 at constant [HS'Jshowed and showed a very shallow minimum at near-neutral pH.
The role of the proton and the hydroxide ion in the rate-limiting step could not be elucidated.
The Arrhenius parameters were calculated from the temperature dependence studies. The
activation energies were 122, 91, and 60 kJ/mol for the homogeneous, vermiculite and biotite
systems, respectively. The disappearance of CCU was first order with respect to the surface
concentration up to a surface concentration of 280 m2/L biotite. The dependence on sulfide
concentration was studied over a range of 0.02-4 mM HS' at a fixed surface concentration.
Below 0.5 mM HS", the CCU transformation rate was roughly first order with respect to [HS"],
whereas above 0.5 mM HS~, the transformation rate was independent of [HS-]. These results
suggest that the biotite surface sites are saturated with HS~ above 0.5 mM [HS-]. Lastly, a
relationship between the ferrous iron content in muscovite, vermiculite, and biotite, and the CCU
transformation rate intimate that ferrous iron plays a role in the reaction with CCU-
(4) The rate of transformation of CCLf with pyrite varied with reaction conditions in the
following order: air-exposed pyrite/aerobic < air-exposed pyrite/sulfide < air-exposed
pyrite/anaerobic < acid-treated pyrite/anaerobic. The rate constants varied by only a factor of
-------
2.5 for all the conditions studied. HS* inhibited the CCU transformation rate by pyrite relative to
systems reacted anaerobically in the absence of HS~. The fresh-ground/anaerobic data could not
be included in this evaluation because of the large scatter in the data, probably caused by
heterogeneity introduced by grinding.
(5) The CCl4 transformation products varied greatly as a function of the reaction
conditions. In the sheet silicate/sulfide systems CS2 was identified as a major intermediate that
hydrolyzed to CO2, accounting for >85% of the CCU transformed. The hydrolysis of CS2
establishes a new pathway to form CO2 anaerobically. In the pyrite systems, CS2 was detected
under all reaction conditions, suggesting that CCU or an intermediate must react directly with the
pyrite surface. Under aerobic conditions, CO2 was the major transformation product (80%), but
kinetic modeling could not resolve the pathway to form CO2- At this point it is not known if
CO2 is formed via reaction of the trichloromethyl radical with 02, or the €82 hydrolysis
pathway. In the fresh-ground pyrite systems, roughly 50% of the CCU was transformed to
In all systems studied, formate and carbon monoxide were minor products.
(6) Under aerobic conditions, an iron oxide coating developed on the pyrite surface. Using
x-ray photoelectrpn spectroscopy (XPS), this coating was identified to be FeOOH. It is likely
that this coating is responsible for the slower reaction rates and different product distribution
measured under aerobic conditions.
(7) Reaction of CCU with the model solids, Aerosil 200 and gibbsite, provided insight into
the transformation mechanism of CCU by the sheet silicates in the presence of sulfide. In
systems containing HS% CCU was reactive with Aerosil 200 and not with gibbsite (y-Al(OH)3),
demonstrating that HS~ may be associated with silica in the sheet silicate experiments. However,
the reactivity with Aerosil was not as great as the sheet silicates, suggesting that ferrous iron was
playing a role in the transformation of CCU-
(8) The rate of transformation of CCl4 with the sheet silicates was dependent on both the
ferrous iron content and the sulfur concentration, indicating that the reaction occurs at sites
where sulfide is associated with structural ferrous iron. It is unlikely that CCU is reacting with
secondary iron sulfide phases because (a) the CCU product distribution formed in the fresh-
ground pyrite system was very different from that formed in the sheet silicate system; and (b) the
CCU transformation rate ..was independent of sulfide concentration above a certain level. If
secondary iron sulfides are controlling the CCU transformation rate, higher sulfide
concentrations should precipitate more iron sulfides through enhanced weathering, thereby
causing an increase in the CCU transformation rate.
(9) In the pyrite system, the near-surface was depleted of iron after reaction in water, while
the oxidation state of the pyrite-S appeared to remain the same. The high sulfur concentration at
the near-surface makes it likely that pyrite-S is the reductant of CC\4, rather than pyrite-Fe. At
circumneutral pH, these surface sulfur sites are negatively charged. Under anaerobic conditions
it is proposed that an electron is transferred from FeSS~ to CCU and that the intermediate
FeSSCCls is formed. Both the surface and solution conditions dictate the decomposition
pathway of this intermediate. Under aerobic conditions the site of electron transfer is not
determined as it may occur across the FeOOH coating or at unoxidized sulfur sites.
-------
SECTION 3
RECOMMENDATIONS
The results of this study show that mineral surfaces may play a significant role in the fate of
halogenated organics in the environment Although they can be quantitatively applied to natural
systems only with some difficulty, these data show that rate constants measured in deionized
water will greatly underpredict the actual transformation rates. In pyrite- or sulfide-rich
environments, abiotic transformation pathways may be significant on the timescale of ground
water transport. Predictive capabilities are complicated at this point due to the confounding
effects of natural organic matter, co-solvents, competing oxidants, and microbial activity. To
address these issues, continued research to further understand the surface chemistry of pyrite, the
CC14 transformation pathway at the pyrite surface, and the reactivity of CCU and other
polyhalogenated aliphatics under field conditions is necessary. This work will ultimately lead to
predictive capabilities.
The pyrite surface was relatively reactive with CCU even under aerobic conditions;
therefore, it is conceivable that a pyrite-based treatment system could be engineered. Further
studies would have to be conducted in order to (1) identify the rate determining step of the
reaction, (2) test the efficiency of the method in column and batch reactors, (3) control the CCU
product distribution, and (4) measure the pyrite oxidation products and ensure that they are
harmless. In addition, the engineered system would have to be tested with other haloaliphatics to
see if they could also be transformed and if they inhibited or effected the transformation of CCU-
-------
SECTION 4
MATERIALS AND METHODS
Research Approach
The primary goal of this research was to test the ability of mineral surfaces to transform
CCU under different environmental conditions. The base case reactions conditions to measure
the disappearance rate of 1 uM CCU in the biotite, vermiculite, and synthetic solid systems were
pH=7.5-8.5, temperature=50 °C, and [HS*]=1 mM. The solids concentrations for the base case
were 55.8 m2 biotite/L, 114 m2 vermiculite/L, 700 m2 Aerosil/L, and 100 m2 gibbsite/L.
Controls were established by reacting CCU with HS' at the same temperature and pH, but in the
absence of the solids. To study the effects of environmental conditions on the reaction rate,
experiments with biotite were conducted over a pH range of 6-10, temperature range of 37.5-
62.7 °C, solids concentration of 11.2-280 m2/L, and [HS']=0.02-4 mM.
For all pyrite/CCU transformation rate and product studies, 1 uM CCU was reacted in
aqueous systems containing 1.2-1.4 m2/L pretreated pyrite at pH 6.5 and 25 °C, except the
experiments conducted with sulfide which were at pH 7.75. Since a pH buffer was not used
because of the potential confounding effects on CCU reactivity, experiments were conducted in a
1 mM NaCI ionic medium. Controls were established by reacting CCU under the same
conditions in the absence of pyrite in addition to reacting CCU in homogeneous solutions of
Fe2+aqOrHS-.
For the studies of the pyrite oxidation products, 0.1-1 mM CCU was reacted with large
particles of pretreated pyrite (0.2 g) to maximize oxidation on a small pyrite surface area.
Controls were established by reacting pyrite under the same conditions, but in the absence of
CCU. These experiments were designed to measure the oxidation due to both pyrite
pretreatments and reaction with CCU-
Reference Solids, Pretreatments, and Reagents
Biotite, vermiculite, and muscovite were obtained from Ward's Scientific Establishment,
Inc. (Rochester, NY). The minerals were wet ground with an Osterizer blender (using
deoxygenated MillirQ water, Millipore Corp., Bedford, MA) in an anaerobic glove box (Coy
Products, Ann Arbor, MI) which had a 90% N2/10% H2 atmosphere. After grinding, the
minerals were freeze-dried, dry-sieved, and stored in the anaerobic glove box. The 200-50 mesh
U.S. Standard (75-300 Jim) size fraction was used for the experiments. Aerosil 200 (>99.8%
SiO2, Degussa, Ridgefield Park, NJ) and gibbsite (ALCOA, Alcoa Center, PA) were rinsed twice
with Milli-Q water before using.
Pyrite (Zacatecas, Mexico) was obtained from Ward's Scientific Establishment, Inc.
(Rochester, NY). All grinding and pretreatment of the pyrite was conducted in a glovebox
containing a 90% N2/10% H2 atmosphere (Coy Products, Ann Arbor, MI). Pyrite was ground
with a ceramic mortar and pestle, sieved to 75-300 um, sonicated in deoxygenated Milli-Q water,
and "air-dried" in the glovebox. Acid-treatment was conducted in the glovebox using 20 g
batches of pyrite. Sonicated pyrite was washed twice for 2 minutes with 20 mL of 1 N HC1,
rinsed 10 times with 40-mL aliquots of deoxygenated Milli-Q water, and "air-dried" in the
glovebox. The sonicated pyrite to be exposed to air was placed in an amber-glass screw-top vial.
The vial was removed from the glovebox and opened to the atmosphere for 3 days.
6
-------
The following compounds were obtained from commercial sources and used as received:
carbon tetrachloride (>99%), trichloroethylene (99%), and chloroform (99%). Radiolabeled
(14C) CCU was obtained from Sigma Chemical Co. (St. Louis, MO), diluted in Milli-Q water
and stored in flame-sealed glass ampules at 4 °C in the dark. Sodium sulfide (Na2S-9H20,98%)
and carbon disulfide (>99%) were stored in an Ar-filled glove bag. Individual crystals of sodium
sulfide were rinsed with deoxygenated Milli-Q water to remove oxidation products and wiped
dry with cellulose tissue before use. Anhydrous sulfate (Na2SO4, >99%) and sulfite (Na2SOs,
>99.5%) were stored in a desiccator. A 0.1 N thiosulfate standard solution (Fluka Chemical Co.,
Buchs, Switzerland) was stored in the anaerobic glovebox. Anydrous pentane (>99%) was
stored under a nitrogen atmosphere to prevent contamination by halogenated organics.
Filling and Sealing of Ampules
Experiments with Sheet Silicates
All transformation and adsorption studies were conducted in flame-sealed glass ampules
because the reaction times were on the order of weeks to months at elevated temperature (50°C ).
The ampule method was based on and adapted from Barbash and Reinhard (1989b). Ampules
were washed with 10% nitric acid, rinsed with deionized water and oven dried overnight at 110
°C. Ampules were weighed and filled with a known amount of sheet silicate. To minimize
Fe(n) oxidation in the autoclave, mineral-filled ampules were autoclaved under a nitrogen
atmosphere. The nitrogen atmosphere was achieved by placing the ampules in a pressure cooker
that contained 100 mL of water. The pressure cooker was evacuated three times to 75 mm Hg,
backfilled with 99.99% nitrogen, and autoclaved at 121 °C for 20 min (Curtis, 1991). After
autoclaving, the ampules were placed in the anaerobic glove box to allow oxygen to exsolve for a
minimum of 2 days. Stock solutions of sulfide were made fresh daily by transferring pre-
weighed and washed sodium sulfide crystals into 50 mL of deoxygenated water in the glovebox.
The amount of HC1 or sulfide stock used to make the sulfide solution was determined by the
parameters of a given experiment This sulfide solution will herein be referred to as the "buffer".
A pH buffer or background electrolyte was not used to eliminate potential confounding effects on
the CCU transformation rate.
Ampules were filled with approximately 13.5 mL of buffer that was filtered through a
sterilized 0.2 ^im nylon filter (Nalgene Corp., Rochester, NY). Each ampule was covered with a
polyvinylidene chloride (Saran™) sheet, secured at the neck with a 3/16" inner-diameter ring of
latex tubing, removed from the glove box, spiked with an appropriated volume of an aqueous
solution saturated with CCU» and flame-sealed under a stream of nitrogen. After sealing,
ampules were weighed to obtain the exact amount of buffer added, and placed in the dark in a
constant temperature water bath (±0.1 °C) for the duration of the experiment Weight loss upon
sealing was less than 0.005% and was not considered significant Daily to weekly, ampules were
removed from the bath, shaken by hand, and returned to the bath. At each sampling time, two
ampules for each experimental condition were removed from the constant-temperature bath and
preserved at 4 °C in the dark until they could be extracted. Before extraction, ampules were
centrifuged at 4 °C and 1400 g for 20 min.
Synthetic Solids Experiments
Transformation studies in the presence of Aerosil 200 and gibbsite were also conducted in
10-mL flame-sealed glass ampules. However, due to the small particle size, the solids were
added to the ampules as a slurry. The slurry was made outside the glove box and autoclaved for
20 minutes at 120°C. The slurry was deoxygenated by bubbling the solution with nitrogen which
-------
passed through a sterile 0.2 um nylon filter for 1 hour. The deoxygenated slurry was placed in
the glove box and bubbled with the glove box atmosphere for 30 minutes. Sodium sulfide was
added to the slurry at this time. To fill the ampules, a sterilized syringe with a sterilized spinal
needle was fitted to the slurry with sterilized PTFE tubing and three-way valve. Once the
ampules were filled, they were covered with a Saran™ sheet and secured at the neck with a 3/16-
in. inner-diameter ring of latex tubing. All other procedures in the synthetic solids experiments
were identical to those described above, except the ampules were centrifuged at 2800 g for 30
min.
Experiments with Pyrite
Transformation studies in the presence of pyrite were conducted in 10-mL flame-sealed
glass ampules that were acid washed, oven dried, and placed in the anaerobic glovebox to outgas
for 2 days (Kriegman-King and Reinhard, 1992). Air-exposed pyrite was placed back into the
glovebox. For aU reaction conditions, 0.2 g (± 0.003 g) pyrite was weighed into the ampules.
For all conditions, except the aerobic system, a 1 mM NaCl solution was deoxygenated
with NI for 1 hour and placed in the glovebox. The NaCl solution was then sparged with
glovebox air for a minimum of 30 min. Sulfide solutions were made as described above.
Ampules were filled with approximately 13.5 mL of either the NaCl or sulfide solution that was
filtered through a sterile 0.2 um nylon filter (Nalgene Corp., Rochester, NY). Once the ampules
were filled, they were spiked and sealed while taking precautions to maintain anaerobic
conditions (Barbash and Reinhard, 1989b; Kriegman-King and Reinhard, 1992). The XPS and
pyrite-sulfur oxidation-experiments were conducted at 0.1-1 mM CCU. CCU was added as a
methanolic spike rather than an aqueous spike. Sealed ampules were weighed and placed in the
dark in a 24.7 °C constant-temperature water bath (± 0.1 °C) for the duration of the experiment
For ampules to be reacted aerobically, the 1 mM NaCl solution was sparged with
laboratory air for 45 min to strip any residual CHCls from the Milli-Q water. Pyrite-filled
ampules were removed from the glovebox and filled with approximately 13.5 mL of a 1 mM
NaCl solution that was also filtered. Ampules were spiked with CCU and immediately flame-
sealed.
For all conditions, ampules were shaken manually once daily for the first 10 days of
reaction and then every other day until the experiment was complete. At each sampling time,
two ampules for each experimental condition were removed from the constant-temperature water
bath, centrifuged at 4 °C and 2800 g for 20 min, and stored at 4 °C for a maximum of 3 hours
until preparation for analysis.
Chemical Analyses
CCk and CHGh Analysis.
Reaction solutions were extracted in 2.9-mL glass vials with PTFE/silicone septum-lined
screw top caps. Ampules were cracked open, 2 mL of aqueous solution were transferred to an
extraction vial containing 0.5 mL of pentane, and the vial was shaken for 15 sec. Extraction vials
were stored inverted at -5 °C in the dark for up to 24 h before GC analysis. Extraction vials were
removed from the freezer, defrosted to room temperature, and spiked with an appropriate amount
of internal standard (trichloroethylene). Vials were mixed on a shaker table for 30 min at 350
rpm. The extractant was analyzed for CCU and CHCls using a Hewlett-Packard 5890 gas
chromatograph equipped with a 63Ni electron capture detector and a DB-1 column (J&W
Scientific, Rancho Cordova, CA), 15 m x 0.535 mm, with a film thickness of 1.5 um. Helium
-------
was used as the carrier gas and Ar/CH4 as the make-up gas. The gas chromatograph was
calibrated daily with a minimum of 5 calibration standards and duplicate measurements were
made for each sample or standard. If the measurements did not agree within 10%, a third
injection was made.
CS2 Analysis
A 2.9-mL aliquot of reaction solution from the ampules was transferred to an extraction
vial (described above) to fill the vial with no headspace. The vials were stored at 4 °C in the
dark for up to 48 h before GC analysis. Aqueous samples (2-2.4 mL) were analyzed for €82
using a Tekmar Model ALS Purge and Trap with a Hewlett-Packard 5890 gas chromatograph
equipped with a 10.0 eV Tracor photoionization detector and a Quadrex 007-502 column
(Quadrex Corp., New Haven, CT), 75 m x 0.535 mm, with a film thickness of 2.5 urn. Helium
was both the carrier gas and the make-up gas. The GC was calibrated daily using external
standards at a minimum of four calibration levels.
CO Analysis
CO was determined by transferring a 7.1 mL aliquot of aqueous solution from an ampule
to an 8.7 mL vial with a PTFE/silicone septum-lined screw top cap. The vial was shaken at 350
rpm for 10 min to equilibrate the CO partitioning between the headspace and the aqueous phase.
A 0.4 mL headspace sample was analyzed on a Reduction Gas Detector (Trace Analytical,
Menlo Park, CA) using 30 mL/min air as the carrier gas. The gas phase CO concentration was
calculated by comparing the peak height to CO gas standards (Scott Specialty Gases, San
Bernadino, CA). The total amount of CO was calculated using a Henry's constant of 0.96
amvm3/mol at 20 °C (Tchobanoglous and Schroeder, 1985).
Formate Analysis
An ion chromatograph (Dionex Series 4000i) equipped with a conductivity detector was
used to measure the formate concentration from reaction of CC14 with pyrite. A Dionex AS-4A
separator column was used with 5 mM borate eluent at a flow rate of 2.0 mL/min. The detection
limit for formate was approximately 1 uM. Although the aqueous non-volatile fraction was
identified to be formate by matching retention times of an unknown peak formed in pyrite
systems with formate standards on the ion chromatograph, formate was not directly quantified in
most of these systems. Herein, the non-volatile fraction is assumed to be formate.
Radiolabeled Products Analysis
To determine the product distribution in experiments with radiolabeled substrate, three 1-
mL aliquots of reaction solution were taken from each ampule. Aliquot 1 was acidified with 0.2
mL of 1 N H2SO4 and purged with N2 for 10-15 min. This procedure stripped the volatiles and
CO2 from solution, leaving the non-volatiles behind. After purging, 10 mL of Universe! (ICN
Biomedicals, Inc., Irvine, CA) liquid scintillation cocktail were added to the sample. Aliquot 2
was treated with 0.2 mL of 1 N NaOH and purged for 10-15 min, thereby stripping only the non-
CO2 volatiles. Again, 10 mL of scintillation cocktail were added to the sample after purging.
Aliquot 3 was immediately added to 10 mL of scintillation cocktail containing 0.2 mL 1 N
NaOH (not purged) to determine the total radioactivity. The CO2 fraction was then calculated by
subtracting the counts in aliquot 1 from those in aliquot 2; and the volatile fraction was
calculated as the difference between the counts in aliquot 3 and those in aliquot 2. The efficiency
of this method to strip CO2 under acidic conditions and to retain CO2 under basic conditions was
tested by adding 0.13-0.56 mL of 1 N H2SO4 or 1 N NaOH, respectively, to the 1 mL aliquots.
There was no significant difference in the results with increasing amounts of acid or base added.
-------
All samples were then counted twice on a Packard Tricarb Model 4530 liquid scintillation
counter for 10 min. Measured counts-per-minute were converted to disintegrations-per-minute
using the external standard method.
Adsorbed Formate and CO2
Adsorbed formate and CO2 were obtained using a presumptive method. The supernatant
was removed from the ampules using a syringe, 0.4 mL of 1 M H2SO4 or 1 M NaOH was added
to the ampules, and ampules were allowed to sit for 10 min. One-mL dilution water was added
to the ampules and mixed with the slurry. After letting the pyrite settle for 10 min, a 1-mL
aliquot of the aqueous phase was placed into a scintillation vial, purged with N2 for 15 min, and
counted on a Packard Tricarb Model 4530 liquid scintillation counter. The extreme acidic and
basic conditions presumably desorbed both the CO2 and formate. The pHpzc of pyrite has been
reported to be 1.2 (Fornasiero et al., 1992). By purging the system with N2, the CO2 was
stripped from solution at acid pH leaving the "adsorbed" formate in solution, whereas the basic
solution retained both the "adsorbed" CO2 and formate. The loss of "adsorbed" CG>2 via
precipitation of carbonate minerals at high pH was assumed insignificant because the supernatant
containing metal ions, such as Fe2+, was removed. This method did not allow measurement of
the adsorbed volatile fraction.
Aqueous Pyrite Oxidation Products
For analysis of potential aqueous pyrite oxidation products SOs2', SC>42% S2Os2',
centrifuged ampules were cracked open in the anaerobic glovebox to prevent oxidation of SOs2'
and S2O32'. A 0.5-mL aliquot of supernatant was placed in a 10-mL volumetric flask and diluted
to 10 mL with deoxygenated Milli-Q water. A portion of diluted sample was then transferred to
a 2.9-mL glass vial to fill the vial with no headspace and sealed with a PTFE/silicone septa-lined
cap. The sample vials were stored in the glovebox up to 6 hours and were removed from the
glovebox immediately before analysis. Standards at three calibration levels were made in the
glovebox with deoxygenated Milli-Q water and treated identically to the unknowns. Samples
were withdrawn from the vials with a 2-mL gas-tight syringe that had been filled with N2 and
analyzed on a Dionex Series 40001 ion chromatograph equipped with a conductivity detector. A
Dionex AS-5 separator column with 4.5 mM Na2COs/2.0 mM NaOH eluent at 1.5 mL/min was
used to separate SOs2', SO42-, and
Surface Pyrite Oxidation Products
Ampules were cracked open in the glovebox, and the supernatant was removed. Using
tweezers and a microbiological loop, a large particle of pyrite was removed from the ampule,
dipped in deoxygenated Milli-Q water, and adhered to an XPS sample holder using double-stick
tape or silver paint. Samples were allowed to dry overnight in a desiccator in the glovebox.
Within the glovebox, the desiccator was transferred to a glovebag and sealed with low-oxygen
permeable tape (Coy Products, Ann Arbor, MI). The sample holder was transferred into the XPS
instrument by sealing the glovebag opening around the transfer chamber of the instrument and
flushing the glovebag with nitrogen. The desiccator was opened and the sample holder was
placed in the transfer chamber. XPS analysis was conducted using a Surface Science S-Probe
equipped with a monochromatic Al-Ka x-ray source. A spot-size of 150 x 800 urn was used
with a pass energy of 150 eV for broad scans and 50 eV for narrow scans. Normalized
concentrations were calculated for the peak of interest by dividing the peak area by a sensitivity
factor comprising the Scofield cross-section (Scofield, 1976) and the relative kinetic energy of
the photoelectrons (Hyland and Bancroft, 1990).
10
-------
Mineral Characterization
Sheet Silicate Characterization
Specific surface area and iron content of the sheet silicates are shown in Table 1. Surface
area was measured on an Acusorb 2100E surface area analyzer by the BET method using
krypton adsorption. Ferrous iron and total iron in the sheet silicates were measured using a
modification of several 1,10-phenanthroline methods (Fadrus and Maly, 1975; Begheijn, 1979;
Fritz and Popp, 1985). A 10-20 mg sample of mineral was added to a pre-weighed 4 oz
polypropylene bottle. Approximately 20 mg of 1,10-phenanthroline and 50 mg of nitnlotriacetic
acid (NTA) were added to the bottles. Since ferric iron-NTA complexes are stronger than ferric
iron-phenanthroline complexes, NTA was used to complex with ferric iron, allowing only ferrous
iron to form a complex with phenanthroline. To dissolve the minerals, 3 mL of 10% H2SO4 and
0.5 mL of 48% HF were then added to the bottles. The bottles were placed in a boiling water
bath for 30 min. Once removed from the bath, 20 mL of 10% citric acid and 5 mL of 4% boric
acid were added. Samples were then diluted to 100-135 mL with Milli-Q water and weighed.
The samples were shaken by hand and then divided in half for the total and ferrous iron
determinations. Approximately 20 mg of hydroquinone was then added to the samples
designated for total iron analysis. All sample bottles were shaken at 350 rpm for 30 minutes, and
analyzed on an Hewlett-Packard Model 8451A UV/visible spectrophotometer at 510 nm.
Concentrations were determined using ferrous ammonium sulfate standards at 5 calibration
levels. This method of iron determination in minerals was tested on a rock standard. The Fe(II)
and total Fe contents agreed with the rock standard within 10%.
Wet chemical analysis of the sheet silicates was not conducted for trace metals other than
iron. X-ray photoelectron spectroscopy (XPS) conducted on cleavage sheets of the biotite and
vermiculite did not show the presence of any redpx sensitive trace metals besides iron. Based on
a semi-quantitative analysis, the biotite contained <2 wt % Ti. The amount of total iron
measured by this semi-quantitative analysis agreed with the iron content obtained from wet
chemical analysis for both biotite and vermiculite (Table 1).
Table 1: BET surface area determined by BET and iron content for solids.
Mineral
Biotite
Vermiculite
Muscovite
Aerosil 200
Gibbsite
Fresh-ground Pyrite
Acid-treated Pyrite
Air-exposed Pyrite
Specific Surface
Area (n^/g)
1.45
2.94
0.75
200
11.4
0.10
0.11
0.083
Fe(H)
Wt % (g/g)
3.1
1.3
0.6
n.m.b
d
n.m.
n.m.
n.m.
Fe(ffl)
Vft%(K/&
3.1
0.8
1.1
<0.002c
d
n.m.
n.m.
n.m.
a Calculated by difference between total iron and Fe(H).
b n.m.=not measured.
c Measured by Degussa, Ridgefield Park, NJ. Speciation not indicated.
d No iron detected using XPS with detection limit approx. 10% of monolayer (C.J. Papelis,
University of Michigan, personal communication).
11
-------
To study the effect of sulfide on biotite, cleavage sheets of biotite were placed in amber
bottles containing deoxygenated Milli-Q water or 0.1 M sulfide. The bottles were sealed with
PTFE-lined caps, wrapped with anaerobic tape, and stored in the glovebox. After 6 months, the
cleavage sheets were analyzed using XPS to look for changes in the iron speciation and for
association of sulfide with the near surface. XPS analysis was conducted on a VG ESCA
instrument with a non-monochromatic Al-Ka x-ray source and a pass energy of 20 eV. As
shown in Figure 1, there was no significant difference in the Fe 2p line shape with biotite treated
in an aqueous solution or in a sulfide solution. When adjusted for charge shifting, the Fe 2p peak
in Figure 1 corresponds to Fe3* indicating that all of the iron at the near surface of the basal
plane is oxidized. A small amount of sulfur was detected at the near surface using XPS (data not
shown), but the amount and speciation was not quantifiable. However, the biotite reacted with
sulfide had a greyish, metallic luster that was not present on the other biotite sample. SEM
photographs (Figures 2 and 3) depict a difference in the two samples: the biotite reacted with
sulfide (Figure 3) appears to be more weathered than the biotite reacted in water (Figure 2).
fiiorite in
0.1 M HS~
Biotite in
Milli-0 Water
Figure 1:
700 705 710 715 720 725 730 735 740 745 750
Binding Energy / eV
XPS spectra of the Fe 2p peak for biotite reacted for 6 months in anaerobic Milli-
Q water and 0.1 M HS~. No differences are observed in the oxidation state of iron
in the two samples. Peaks are charge shifted by approximately 7.0 eV.
12
-------
Figure 2: SEM micrograph of biotite cleavage sheet after reaction in anaerobic Milli-Q
water for 6 months at 25 °C (white bar equals 100 ^im).
Figure 3: SEM micrograph of biotite cleavage sheet after reaction in 0.1 M HS- for 6
months at 25 °C (white bar equals 100 pm).
These results show that sulfide does interact with the biotite surface, but the type and extent of
effect sulfide has on the surface is unknown. The only studies on the effect of sulfide on
weathering or dissolution of minerals has been conducted on iron oxides (Pyzik and Sommer,
1981; dos Santos Afonso and Stumm, 1992; Peiffer et al., 1992). These studies show that sulfide
promotes reductive dissolution of iron oxides.
Synthetic Solids Characterization
The only chemical data that we have for Aerosil 200 is from the manufacturer. They do not
report the presence of trace metals, except for Ti, which is present at <0.02 wt %. The gibbsite
13
-------
surface was chemically pure as determined by XPS (CJ. Papelis, University of Michigan,
personal communication). The BET surface areas and iron contents for Aerosil 200 and gibbsite
are presented in Table 1.
Pyrite Characterization
Single point BET surface area of pyrite was determined using a Micrometrics Flowsorb n
2300 system with 30% N2/70% He gas mixture. The specific surface areas are summarized in
Table 1. The chemical composition of the near-surface of fresh-cleaved pyrite was measured
using XPS. The S:Fe ratio was determined to be 2.1. One in ten analyses of reacted pyrite
samples showed the presence of 1.1 atomic % Cu.
Kinetic Evaluation of Data
Observed pseudo-first-order rate constants (k'pbs) for the disappearance of CCU in the sheet
silicate and synthetic solids systems were calculated from regressions of m([CCU]t/[CCU]o) vs.
time where [CCUlo and [CCUlt were the CCU concentrations at time=0 and time=t, respectively.
In the pyrite systems zeroth-order rate constants QH°CQA) were calculated from linear regressions
of [CCUlt/fCCUlo vs. time wherein k°ccu= -(slopeXfCCUlo)- At least nine data points were
used to determine each rate constant in this study. The standard error of the slope for the linear
regression was used to calculate the 95% confidence intervals for the rate constant. In most of
the heterogeneous experiments, the 95% confidence intervals were less than 25% of the rate
constant. The error was much higher (up to 100%) in the homogeneous experiments because
there was relatively little transformation of CCU over the course of the experiment. For
experiments with k'0bs<0.001 day1, the transformation was considered negligible over the time-
period studied and k'obs was assumed to equal zero.
The rate law for the disappearance of CCU in the sheet silicates systems was hypothesized
as follows:
(i)
(2)
(3)
where a, pi, P2, yl, y2, and 8 represent the reaction order with respect to reduction in solution
and at the mineral interface, k'homo and k'hetero are pseudo-first-order rate constants, and ki^o.
kns- and khetero are intrinsic rate constants. Eqs. 1, 2 and 3 assume a priori that the
heterogeneous and homogeneous rate constants are first order with respect to [CCU] (e.g. a=l).
The mineral surface area concentration (SC) was calculated from the product of the solids
loading (g/L) and the specific surface area (m2/g) of the mineral. The pKa for the first
dissociation of H2S at the reaction temperature (MUlero, 1986) was used to calculate the HS~
concentration based on the pH and the amount of total sulfide added to the system. For
homogeneous systems, k'0bs = k'homo» where k'homo accounts for the reactions with H20 and HS'
in solution. In the heterogeneous systems, k'hetero = k'obs -
Reaction orders for the different reactants (i.e. HS', pH, CCU, SC) were determined by the
method of isolation under pseudo-first order conditions. When calculating reaction orders or
Arrhenius parameters, individual rate constants (k'O were calculated for each data point, i, using
the integrated form of a first-order rate law (eq. 4),
14
-------
in order to account for the scatter in the original data (Barbash and Reinhard, 1989b). For the
heterogeneous systems, k'h0mo was subtracted from k'j, yielding a heterogeneous individual rate
constant, denoted as k'i,het.
Adsorption .
Adsorption of CCU onto biotite and vermiculite was determined by control experiments
using radiolabeled CCU at 25 °C. Comparison of the aqueous CCU concentration in the
homogeneous and heterogeneous systems over four weeks showed less than 3% adsorption.
Because the adsorption of CCU was so small, CCU measurements in the transformation studies
were not corrected for adsorption.
15
-------
SECTION 5
RESULTS AND DISCUSSION
Transformation of CCU by Sheet Silicates
Laboratory studies were conducted to identify and quantify the environmental parameters
that govern the transformation rate of CCU in a heterogeneous aqueous environment containing
sulfide and sheet silicates (biotite and vermiculite). The parameters studied were temperature,
pH, mineral surface area, and sulfide concentration. The reaction rate was hypothesized to be a
function of surface area because the reactions have been shown to be faster in the presence of
mineral surfaces (Kriegman-King and Reinhard, 1991). Sulfide was also expected to play a role
in the kinetics because it could act as either an electron donor or a nucleophile. Data are
presented in terms of [HS~] instead of total sulfide because HS~ is the dominant species at near-
neutral pH, HS" is a much stronger nucleophile than H2S (Haag and Mill, 1988; Barbash and
Reinhard, 1989a), and HS" is a stronger reductant than H2S (Luther, 1990). Lastly, pH
dependence was studied because it affects the surface charge of the minerals, the aqueous
speciation of sulfide, and probably the surface speciation of sulfide.
Sheet silicates and sulfide were hypothesized to influence the mechanism of CCU
transformation in the following ways (Kriegman-King and Reinhard, 1991):
(1) CCU can undergo electron transfer with ferrous iron in the sheet silicate and the
oxidized iron can then be re-reduced by sulfide;
(2) CCU can react with sulfide that is adsorbed to the sheet silicate; and
(3) sulfide can react with dissolved iron (released to solution from mineral dissolution) to
form a secondary mineral, an iron sulfide, which can then react with CCU-
Structural ferrous iron in biotite has a lower reduction potential than aqueous ferrous iron
(White and Yee, 1985). In hypotheses 2 and 3, the adsorbed sulfide or iron sulfide-sulfide can
act as an electron donor or as a nucleophile. As far as we know, the reduction potential and/or
nucleophilicity of adsorbed sulfide on any solid has not been measured or compared with the
reduction potential and/or nucleophilicity of aqueous sulfide. Polysulfides (SX2~, where x is the
number of sulfur atoms) are as strong or stronger nucleophiles than HS~ (Haag and Mill, 1988;
Vairavamurthy and Mopper, 1989; Luther, 1990). In these systems, SX2~ may be forming from
reaction of HS~ with ferric iron in the sheet silicates. In the pH range of 7-9, $42' and Ss2' are
the polysulfides that are predicted to be present in solution (Luther, 1990). As a fourth
hypothesis, SX2~ may be reacting with CCU via electron transfer or nucleophilic substitution. If
secondary mineral formation of pyrite (FeS2) occurs, the $22~ group on pyrite may exhibit a
reactivity similar to that of the lower polysulfides (n=2 or 3), which are slightly less reactive
electron donors or nucleophiles than aqueous HS' (Luther, 1990). However, the iron sulfides,
pyrite and marcasite, have been shown to increase the rate of CCU transformation compared with
homogeneous rates in systems containing aqueous ferrous iron or HS~ (Kriegman-King and
Reinhard, 1991).
Effect of Biotite and Vermiculite on CCl^ Transformation
Figure 4 shows the effect of biotite (SC=55.8 m2/L) and vermiculite (SC=114 m2/L) on the
transformation rate of 1 \iM CCU in the presence of 1 mM HS~ at pH 8.6, 50°C. The average
observed rate constants (k'0bs) for similar systems are 0.020±0.011,0.08110.005, and 0.1210.012
16
-------
day1 for the homogeneous, vermiculite, and biotite systems, respectively. The results show that
(1) disappearance of CCU is first-order (o=l); (2) the minerals increase the transformation rate
of CCU over the rate that occurs in homogeneous solution; and (3) biotite is more reactive than
o
U
u
u
N_X
e
Homogeneous
Vermiculite
Biotite
10 15
Time (days)
Figure 4: First-order plots of CCU transformation with biotite (SCbi0tite=55.8 m2/L) and
vermiculite (SCvennic=114 m2/L), [HS']=1 mM, pH=8.6 at 50 °C.
vermiculite. Jeffers et al. (1989) studied the hydrolysis of CCU in the temperature range of 96-
170 °C and also found that the disappearance of CCU was first-order with respect to CCU.
Using their Arrhenius parameters (lnA=36.3 day1; Ea=114.5 kJ/mol), the rate of hydrolysis
(kH2o) of CCU was calculated to be 0.002 day1 at 50°C. Although the rate reported by Jeffers
et af. (1989) may be too high because the experiments were conducted in stainless steel tubes that
could have promoted the transformation of CCU, it is the best hydrolysis data available for CCU-
In homogeneous systems, the CCU transformation rate in the presence of sulfide is at least an
order of magnitude greater than k^O when [HS']>0.5 mM. In heterogeneous systems, the CCU
transformation rate is faster than kp^O with [HS']>0.05 mM and SCbi0tite=55.8 m2/L. Even
though H2O is nearly five orders of magnitude more concentrated than HS% HS" is more reactive
with CCU than is H^O. This reactivity is further enhanced in the presence of minerals. These
results are not surprising because HS' is a much stronger nucleophile than HaO or OH" (Swain
and Scott, 1953; Haag and Mill, 1988; Barbash and Reinhard, 1989b; Schwarzenbach and
Gschwend, 1990) and it may act as a reductant
CCU Transformation Products
Figure 5 depicts the disappearance of CCU and the appearance of products in a
heterogeneous system containing biotite (SC=55.8 m2/L) and 1 mM HS'. Carbon disulfide (CS2)
was identified as a major intermediate by matching retention times with standard CS2 solutions
on one column using a gas chromatograph equipped with a photoionization detector. As shown
17
-------
in Figure 6, €82 concentrations measured by gas chromatography agree with the "unidentified"
volatile fraction calculated by difference between the total volatile fraction determined from 14C
data and the known volatile fractions (CCU and CHCls) measured by gas chromatography.
About 65% of the CCU was transformed to CO2 after 60 days. At this time, approximately 20%
of the CS2 was remaining. Chloroform, formed via reductive dehalogenation of CCU, reached a
maximum of 10%. CHCls was shown to be persistent in these systems by reacting 5 uM CHQs
under the same conditions as the CCU experiments. The half-life of CHCls in the presence of
55.8 m2/L biotite and 1 mM HS~ at 50°C was measured to be 172 days, whereas the half-life for
hydrolysis of CHCla at pH 7.75 and 50°C is 5000 days (Jeffers et al., 1989). Carbon monoxide
and a non-volatile component were measured as products in very small quantities (<5%
combined) in the CCU systems. The non-volatile product, detected by 14C fractionation
measurements, has not been identified, although Criddle and McCarty (1991) observed the
appearance of formate as a major transformation product of CCU in an electrolytic cell.
o
U
U
u
u
^^
o
'•S
a
b
*»
I
o
U
CC14
CHC13
CS2
Non-volatile
C02
Mass Balance
Figure 5:
Time (days)
CCU transformation products from reaction with [HS']=lmM, SCbk>tite= 55.8
m2/L, pH=8.8 at 50 °C. Lines represent best fit using eqs. 8,12, and 13.
18
-------
u
u
D
O
s
a
1
0.4-
0.2-
o.o
o
o
§
-0.2
O "Unidentified" Volatile
D Measured CS2
10
20
30
40
50
Time (days)
Figure 6: Carbon disulfide (CS2) was identified as an intermediate of CCU transformation
by the agreement between the CS2 fraction measured by gas chromatography and
the "unidentified" volatile fraction measured from 14C experiments.
The only other known report of CS2 as a CCU transformation product was observed in
systems with fumarate respiring and fermenting Escherichia coli (Griddle et al., 1990), although
the pathway to form CS2 from CCU was not known. In the fumarate respiring and fermenting £.
coli systems, CS2 was a minor product or intermediate (4.3% and 1.6% of the added CCU.
respectively). Studies on the hydrolysis and oxidation of CS2 show that CS2 is hydrolyzed to
CO2 by hydroxide ion (Hovenkamp, 1963; Adewuyi and Carmichael, 1987). Adewuyi and
Carmichael (1987) proposed the following steps for CS2 hydrolysis:
CS2 + OH~ —> CS2OH~ (slow)
CS2OH~ + OH~ —+ CS02H~ + HS~
where the hydrolysis of €82 to dithiocarbonate (CS2OH') is the rate-limiting step. Assuming
€82 is stoicniometrically converted to CO2, about 85% of the CCU is ultimately transformed to
CO2 in these systems.
The proposed chemical transformation pathways for CCU under anaerobic conditions are
summarized in Figure 7. The products and intermediates in the shadowed boxes were detected in
our experiments. As summarized by Griddle and McCarty (1991), the first step for the
transformation of CCU has been proposed to be a one-electron reduction to form a
trichloromethyl radical and CK This radical can follow several different pathways such as
additional electron transfer to form a dichlorocarbene and Cl", dimerization to form
hexachloroethane, or electron transfer and protonation to produce CHCls.
19
-------
cl Cl
\ <*
cf
2H2O 4Hd
Hydrolysis
Carbon tetrachloride
r°i
2x
cf
SM*
-S.
*C1
Hexachloroethane Trichlorpmethyl
(not detected) radical
*^^b
^" Trichloromethanethiolate
e'-t-H*
V.
Dichlorotbionomethane
Chloroform
(5-15%)
Dichlorocarbene
2H2O
c
.
Carbon
monoxide
(1-2%)
HCOOH
Formic acid
(Non-volatile product
detected (3-6%), formic acid?)
Methylene
chloride
(not detected)
disulfide
(intermediate)
dioxide
(81-86%)
Figure 7: Proposed CCU transformation pathways in HS' solution containing biotite.
Shadowed boxes are products and intermediates detected in this study.
Compounds in brackets are intermediates proposed from the literature (as cited in
Criddle and McCarty, 1991) and this report. Reaction pathways with dashed
arrows have been observed in other studies (as cited in Criddle and McCarty,
1991).
20
-------
The only pathway previously suggested to form CO2 under anaerobic conditions is direct
hydrolysis (Criddle and McCarty, 1991). There has been no evidence that CCU can undergo
direct hydrolysis to CO2, although it is chemically feasible. Jeffers et al. (1989) did not propose
a hydrolysis pathway. In our systems, €82 appears to be a major intermediate which is
transformed to CO2- The hypothesized pathway to form CS2 is discussed below.
CCU most likely reacts via an electron transfer reaction to form CCLf which decomposes
to a trichloromethyl radical (Shaik, 1983; Pross, 1985a, b). The trichloromethyl radical can react
with sulfide or its oxidation products in one of the following ways: (1) it can react with HS" and
release H ; (2) it can first accept another electron to form the trichloromethyl anion which then
reacts with Sr2' or S2Q32' to release S(x-i)2" or SOs2', respectively; or (3) it can first react with
Sx2" or 82032- producing S(x-i)2~ or SOs2', respectively, and a trichlorothiomethyl radical which
then accepts an electron. These three proposed pathways all result in the formation of trichloro-
methanethiolate (CClaS*). Sx2' and 8203^ are likely to be present in these systems by reaction
of HS" and ferric iron in the minerals. The proposed intermediate, CClsS', should decompose to
form dichlprothionomethane (Cl2C=S) which reacts with HS' to form CS2- As discussed above,
CS2 is ultimately hydrolyzed to CO2 by OH'. Rather than reacting with HS~ to form CS2,
Cl2C=S could react with water to form CO2 via a carbonyl sulfide (COS) intermediate.
However, in these systems CS2, not COS, was formed since the analytical procedure to detect
CS2 could distinguish between COS and CS2- It should be noted that CS2 could not be measured
in low-sulfide systems (<0.5 mM HS") because the CS2 hydrolysis rate was faster than the CCU
transformation rate.
Rate constants for the different CCU transformation pathways can be evaluated by
considering the following relationship:
*•' — t' +t' 4-fr' ( si
* obs ~ * CS2 ^K C#CJ, Tlt NV v-v
where k'cS2» k'CHCls and k'Nv are pseudo-first-order rate constants that describe the formation of
CS2, CHC13, and the non- volatile product, respectively. CO was not measured regularly and was
detected in such small quantities that it was not considered in this analysis. To calculate a rate
constant for the appearance of CS2 and its subsequent hydrolysis, the following equation was
used:
(6)
[CS2] (7)
where the pseudo-first-order rate constant, k'co2' describes the hydrolysis of CS2- This
differential equation can be solved by dividing eq. 7 by eq. 1 and using an integrating factor
(Kreyszig, 1983), F, where F=[CCU]'k'c°2/k'obs. The solution is:
(8)
V*0fa «-C02
The rate of appearance of CO2 can be described by eq. 9
d[C02] = J,
dt co* 2
which can be solved for [CO2] by substitution with eq. 8 and integration.
21
(9)
-------
(10)
Rate constants for the formation of CHQs and the non-volatile product were obtained from eq.
11,
jrr'ur'i i
and
These equations were solved for [CHCla] and [Non-vol] by substituting CGU with the integrated
form of eq. 1 and integrating to obtain
and (12)
[Non - vof] = ". _ expKter) (13)
The rate constants, k'cs2 k'CQ2' ^'CHC13» an^ k'nv were estimated by fitting the data to
eqs. 8, 12, and 13 using both the nonlinear curve-fitting statistical package, SYSTAT (SYSTAT,
Inc., Evanston, IL), and the best visual fit. The curve-fitting results for one experiment are
shown in Figure 5, and estimated rate constants are contained in Table 2. The curve-fitting
results show good agreement across experiments in terms of the fraction of CCLj reacting via the
three different pathways. Applying eq. 5 to the curve-fitting results, a mass balance of 95-100%
was obtained in these experiments.
Temperature Dependence
Since experiments like those shown in Figures 4 and 5 were on the order of weeks to
months at 50°C, the temperature dependence of CCLt transformation was studied to allow
extrapolation of rates to environmentally relevant temperatures. From the temperature data
collected at 37.5, 50.0 and 62.7°C, we calculated values for the Arrhenius activation energy (Ea)
and pre-exponential (A) for the homogeneous, vermiculite, and biotite systems. For the
heterogeneous systems, the Arrhenius parameters were calculated using the relationship,
k'ijiet=kIi - k'homo- Calculated Ea and A values are listed in Table 3. Lower Ea values in the
heterogeneous systems indicate that these reactions will dominate the homogeneous reactions to
an even greater extent at environmentally relevant temperatures. For example, at 50°C, CCLj in
the presence of biotite and 1 mM HS' reacts 8 times faster than CCU in the absence of biotite,
whereas at 15°C, the biotite system reacts 125 times faster than the homogeneous system.
22
-------
Table 2: Pseudo-first-order rate constants obtained for CCU transformation pathways using
eqs. 8,12, and 13.
Experiment8
Biotitel
Biotite2
Vermiculite
k'obs
0.185 d-1
0.129
0.102
k'cs2
O.lSd-1
0.11
0.088
k'CQ2
0.06 d-1
0.03
0.035
k'cHCia
0.022 d-1
0.006
0.015
k'NV
0.003 d-l
0.008
0.001
Sumb
0.175 d-1
0.124
0.104
a Experiments were conducted with 1 mM HS~, 55.8 m2/L biotite or 114 m2/L vermiculite at
50°C. The pH was 8.6, 8.8, and 8.3 for the biotitel, biotite2 and vermiculite experiments,
respectively.
b Sum= k'cs2 + k'CHCls + k'NV
Table 3: Arrhenius parameters for CCU transformation with 1 mM HS~. Ea and InA were
calculated using k'ithet for biotite and vermicuh'te and k'i was used for the
homogeneous systemsa.
System
Homogeneous
Vermiculite
Biotite
Ea [kj/mol]
122±32b
91.3±8.4
59.9±13.3
ln(A) [InCdar1)]
41.0±2.1a
31.4±0.5
19.9±0.9
a Data collected at pH 7.5 and in the temperature range 37.5 - 62.7°C
b 95% confidence intervals
pH Dependence
The effect of pH on the CCU transformation rate was studied over the pH range of 6-10 at
constant HS* concentration (not constant total sulfide concentration). Since several data points of
the pH6 and pHIO data were collected at 62.7°C, these rate constants were extrapolated to 50°C
assuming that the Ea values tabulated above were independent of pH. The extrapolated data
were then combined with other rate constants obtained at 50°C. As shown in Figure 8, in the pH
range of 6-10, k'homo and k'hetero did not show a first order pH dependence. In the heterogeneous
systems, the rate appears to go through a shallow minimum between pH 7-9. Although die rates
appear to increase toward pH 6 and 10, the reaction order of H+ does not exceed 0.15. The
reason for the higher rates at pH 10 is unknown, but may be due to increased mineral dissolution
at this pH. The high experimental error at pH 10 make data interpretation difficult, however.
The amount of CHCls produced from CCU also did not vary significantly as a function of pH
(data not shown).
23
-------
^
^N
I
(A
•a
•**
V
is
o
e
E
e
!*
U.3'
0.4-
0.3-
;
0.2-
o.i-
'
o.o-
O Homogeneous
A Vermiculite
D Biotite
i
1
4
1 i
: e
k fi - 5
L T a TT J ^
^ T *W U (
O S
o V T^ o
J
i
I
5.67 8 9 10
pH
Figure 8: CCLt transformation rate at 1 mM HS% 50 °C, and 55.8 m2/L biotite or 114 m2/L
vermiculite in the pH range of 6-10. Error bars are 95% confidence intervals
around k'homo or k'hetero-
Effect of Solids Concentration
If the reaction of CCU with HS~ is heterogeneous, k'hetero should be dependent on SC. All
of the above experiments were conducted at a biotite surface area concentration (SCbiotite) of
55.8 m2/L. The effect tite=55.8 m2/L, the ratio of HS" to SCbiotite decreases by a factor of 5. At
this lower HSVSCbiotite ratio, the rate of CCU transformation should be approximately 2.5 times
slower (see Effect of Sulfide Concentration and Figure 10). If k'hetero measured for
SCbiotite=280 m2/L is adjusted for the lower HSVSCbiotite ratio, k'hetero,adjusted is approximately
0.5 day1. First-order behavior with respect to SCbiotite is therefore observed up to 280 m2/L
(5=1).
24
-------
0.25
~ 0.20'
VH
I
<
& 0.151
« o.io -
0.05-
0.00
§
0
100 200 300
Surface Area Cone. [mA2/L]
Figure 9: The effect of the biotite surface concentration (SCbiotite) °n
transformation rate (k'hetero defined eq. 1) at 50 °C. The best-fit line for the data
at 11.2 and 55.8 m2/L passes through the origin. Error bars are 95% confidence
intervals.
Effect of Sulfide Concentration
The effect of HS' concentration was studied over a range of 0.02-4 mM HS' at constant
(55.8 m2/L) and 50°C. The data indicate that above 0.5 mM HS', the CCU
transformation rate is independent of [HS'] (Figure 10). In this range with pH=7.5-8.8,
k'hetero=0-10±0-01 day1. Below 0.5 mM HS% CCU transformation rate appears to be dependent
on HS" concentration. The CCLt systems with 0.02 and 0.05 mM HS' did not undergo
substantial transformation and therefore have a large error associated with them. The reaction
order was evaluated using a logarithmic plot of both k'hetero versus [HS-] and k'i^et versus [HS'].
The data at 0.02, 0.05 and 0.09 mM HS" could not be corrected for k'homo because the
homogeneous rate constants were too small to be measured. The reaction order ((32) was
estimated to be 1.34 and 1.16i0.21 for the ^.'hetero and k'yiet methods, respectively. At this point,
we assume that the heterogeneous CCU transformation rate is dependent on the HS'
concentration below 0.5 mM, although this dependence is not confirmed. The reaction order
(pl) with respect to HS' in the homogeneous systems could not be measured under our reaction
conditions due to the slow transformation rates in the low sulfide concentration systems. The
data conceptually adhere to a Langmuir model sorption/reaction in which rapid sorption at the
reactive surface sites proceeds the rate-limiting reaction. In this model, the rate increases up to
surface site saturation and then becomes independent on the reactant concentration. The data
indicated in Figure 10 indicates saturation above 0.5 mM HS'. However, fitting the data to this
model yielded unrealistic Langmuir parameters suggesting that the model assumptions (i.e.
uniform surface sites, non interaction of adsorbed molecules) are not appropriate.
25
-------
0
Of)
O
Slopes: k'i,het=U6±0.21
k'hetero=1.34
O klietero
— Fit: klietero
— Fit:k'i,het
-5
-4 -3
log(HS- [M])
-2
Figure 10:
The effect of HS' concentration on the CCU transformation rate (k'ketero) and the
fit to the proposed rate law (Equation 3) with SCbiotite=55.8 rn^/L at 50 °C.
Above [HS~] = 0.5 mM, the rate shows a zeroth-order dependence on [HS']
(P2=0). When [HS']<0.5 mM, p2 is estimated to be 1.34 or 1.1610.21 using
k'hetero and k'jjiet data, respectively. Error bars are 95% confidence intervals.
Effect of Ferrous Iron Content
The effect of ferrous iron on the CCU transformation rate was studied by comparing three
structurally similar sheet silicates, biotite, vermiculite, and muscovite. Based on an increasing
bulk Fe(II) content, the transformation rate increases in the order, muscovite, vermiculite, and
biotite (Figure 11). Although ferrous iron appears to play a role in these reactions, the
significance of this relationship is not known because (1) iron was measured as a bulk quantity,
not as a surface concentration; and (2) the sheet silicates, although being all 2:1 sheet silicates
have quite different chemical compositions. Based on rates of potassium release, biotite (White
and Yee, 1985) undergoes dissolution significantly faster than muscovite (Knauss and Wolery,
1989). Differences in dissolution rates may explain the observed differences in CCU
transformation rates compared with the bulk ferrous iron content
26
-------
0.2
<
>»
«
2
V
0.1-
0.0
Biotite
Venniculite
Muscovite
A
A
O
O
O
O
Figure 11:
0 1 2
Ferrous Iron Content [g Fe(n)/ampule]
The effect of bulk ferrous iron content in the minerals, muscovite, vermiculite,
and biotite on k'hetero at 50 °C.
In the introduction, it was also suggested that S42' and Ss2' may be formed by reaction with
ferric iron in the sheet silicates. Although these polysulfides should be more reactive with CCU
than HS' or pyrite, there is no relationship between the ferric iron content in the sheet silicates
and the CCU transformation rate, suggesting that reaction of CCU with Sx2' is not controlling the
transformation rate. In light of the ferrous iron content results and studies of CCU
transformation with pyrite and marcasite (results presented in Kriegman-King and Reinhard,
1991 and below), it seems feasible that CCU may be undergoing transformation via reaction with
HS* associated with ferrous iron sites and/or with a secondary iron sulfide phase.
Summary of Sheet Silicate Results
This section shows that HS' with either biotite or vermiculite increases the CCU
transformation rate over rates that occur in homogeneous solution. For the homogeneous
reaction at 25°C, the half-life of CCU with 1 mM HS' was calculated to be 2600 days. In the
presence of 1 mM HS' and vermiculite (114 m2/L) or biotite (55.8 m2/L) 25°C, CCU removal
was first order with half-lives of 160 and 50 days, respectively (calculated using the Arrhenius
parameters). On a surface area normalized basis, CCU transformation due to the presence of
biotite is approximately six times greater than vermiculite.
The results of product studies suggest that reaction the major transformation pathway of
CCU with HS- is via CS2 to CO2. It is proposed that CCU undergoes reduction to form a
trichloromethyl radical which then reacts with HS', Sx2', or S2O32' to form CS2- The
intermediate, CS2, hydrolyzes to CO2 at rates that may be appreciable relative to ground water
residence times. At 50 °C, the rate constant for the disappearance of CS2 ranged from 0.03 to
0.06 day1. Although Arrhenius constants were not measured for the hydrolysis of CS2 to CO2
under reaction conditions herein, literature values for Ea (43.5 kJ/mol~Philipp, 1955 as cited in
Adewuyi and Carmichael, 1987); 50.0 kJ/mol~Adewuyi and Carmichael (1987)) result in a
hydrolysis rate of 0.006-0.015 day1 at 25 °C (half-life of 45-110 days) which can be significant
27
-------
on the scale of ground water transport About 85% of the CCU is ultimately transformed to
in these systems. Reductive dehalogenation of CCU to CHCls was observed, although it
contributes to only 5-15% of CCU transformation.
Data were presented that showed the effect of temperature, pH, mineral surface area, and
sulfide concentration on the CCU transformation rate. At 50 °C, pH 6-9, SCt>iotite=0-280 m2/L
and low HS' concentrations ([HS-]<0.5 mM), the reaction orders from eq. 3 were determined to
be: cc=l, (32=1.2, and 6=1. The pH dependence in the environmentally relevant pH range was
too low to be determined reliably and both yl and y2 may be assumed to be zero. At high HS-
concentrations ([HS']=0.5-4 mM) and SQ>iotite<55.8 m2/L, the rate of disappearance of CCU in
heterogeneous systems was independent of HS- concentration ((32=0).
Finally, ferrous iron in the sheet silicates appears to be playing a role in the transformation
of CCU with HS'. It is most likely that the reaction is occurring at sites where HS* is associated
with ferrous iron. CCU may also be reacting with iron sulfides formed by iron dissolution and
subsequent precipitation with sulfide (Kriegman-King and Reinhard, 1991), but there is no direct
evidence from SEM or XPS for iron sulfide formation. Although the mechanism of
heterogeneous CCU transformation is unknown, this work clearly shows the significance of
mineral surfaces on the transformation rates of a halogenated organic compound compared with
homogeneous rates when HS- is present.
Transformation of CCU by Synthetic Solids
In the previous section, both biotite and vermiculite were shown to react with CCU in the
presence of HS-. It was unclear how the minerals and HS- increased the transformation rates.
The sheet silicates, biotite and vermiculite, are 2:1 sheet silicates that are comprised of an
aluminum hydroxide layer (gibbsite) sandwiched between two silicon dioxide (silica) layers.
The point of zero charge (pHpzc) of gibbsite is between 8.5 and 10 (Davis and Hem, 1989;
Sposito, 1989), and the pHpzc of amorphous silica is approximately 2-3 (Sposito, 1989). Katser
and Ropot (1986) studied the adsorption of HS' on natural sheet silicates and showed that 0.1 g/L
slurries of bentonite, hydromica, and palygorskite adsorbed more than 90% of 0.6 mM HS"
within 4 hours. However, no information is available in the literature on the redox properties of
sulfide sorbed by mica surfaces. Experiments were conducted with Aerosil 200 (amorphous
SiO2) or gibbsite (cc-Al(OH)3) and HS- to examine the effect of sheet silicate components mat do
not contain significant amounts of iron on the CCU transformation rate. .
The data presented in Table 4 show that k'0bs measured in the presence of gibbsite (5-100
m2/L) and HS* did not increase over k'0bs measured in homogeneous systems. The Aerosil
200/HS- system significantly increased k'0bs over homogeneous and gibbsite rates, even when
k'obs is normalized for surface area. These data suggest that silica/HS- system is capable of
promoting the transformation of CCU, rather than the gibbsite/HS- system. These results are
surprising because one would not expect Aerosil 200, with a pHpzc of 2-3 (Sposito, 1989), to
adsorb HS\ However, Aerosil 200 is a fumed silica with isolated silanol groups that tend not to
adsorb water (Burneau and Banes, 1990).
28
-------
Table 4: Results of CCU transformation in systems containing HS~, gibbsite and Aerosil
200at50°C.
System Solids Cone
(m*fL)
Homogen.
it
n
11
«d
Gibbsite
tt
it
Aerosil 200
—
5
50
100
700
pH
7.2
7.5
7.6
8.0
7.7
8.0
8.4
9.2
7.9
[HS-]»
(mM)
3.4
1.1
0.88
0.95
0.91
0.95
0.98
1.0
4.5
k'obsb Number of
(day1) Data Points
0.031±0.022
0.021±0.007
0.019±0.010C
0.006±0.002
0.0052±0.0047
0.006±0.004
0.003±0.002
0.007±0.004
0.2610.06
13
11
45C
17 .
21
17
17
17
13
a[HS-] calculated using the pKa=6.73 at 50°C (Millero, 1986).
b95% confidence interval around k'0bs.
cAverage of three experiments with 15 data points each.
dMeasured in the presence of 1 mM Tris buffer.
These isolated groups may be able to adsorb HS" at these sites, even at pH 7-8. The adsorption
of HS' at the isolated silanol groups is postulated to be of the form
)SiOH+~SH > QSiOHSH)-
(14)
where sulfide undergoes hydrogen bonding with the silanol surface. Although Aerosil 200 may
not be an adequate model of silica surfaces in sheet silicate minerals because of its unusual
surface, these results still show that "non-ferrous iron bearing solids" can act as catalysts to
promote CCU transformation. It is also conceivable that trace amounts of iron (<0.002 wt %) in
the Aerosil 200, rather than the silica surface, were responsible for the transformation of CCU in
the presence of HS~. A comparison of the reactivity of Aerosil to biotite and vermiculite on a
surface area basis shows that Aerosil is not as reactive as the natural sheet silicates.
Transformation of CCU by Pyrite
Iron sulfide minerals are ubiquitous in sulfate reducing environments (e.g. Howarth and
Teal, 1979; Luther et al., 1982, Lord and Church, 1983; Swider and Mackin, 1989; Huerta-Diaz
and Morse, 1992). The iron sulfide minerals pyrite and marcasite were shown to reduce CCU to
CHCls, but most of the CCU was not accounted for and the reaction mechanism was unknown
(Kriegman-King and Reinhard, 1991). Sulfur has been the proposed electron donor in the
oxidation of pyrite by different oxidants (Singer and Stumm, 1970; Umana, 1979; Goldhaber,
1983; Moses et al., 1987; Nicholson et al., 1988; Hyland and Bancroft, 1989, 1990; Mycroft et
al., 1990). CC\4 is proposed to accept an electron from pyrite-sulfur TC* (anti-bonding) orbital.
The electron is transferred from the TC* orbital on the pyrite-S to a c* (anti-bonding) orbital in
CCU which is symmetry allowed according to frontier molecular orbital theory (Luther, 1990).
To determine if the reaction is energetically favorable, the energy of the lowest unoccupied
molecular orbital (ELUMO) °f the oxidant has to be less than or within 6 eV of the energy of the
highest occupied molecular orbital (EHOMO) of the electron donor (Luther, 1990). For the case
29
-------
of CCU and pyrite, ELUMO= -0.28 eV (Shaik, 1983) and $K)MO= -3.9 eV (Luther, 1990)
indicating that die reaction is energetically favorable.
The purpose of this study was to assess the reactivity of CCU with pyrite (FeSi),
specifically to: (1) study the ability of CCU to oxidize pyrite in the presence and absence of O2
and HS", (2) measure the effect of pretreating the FeS2 surface with Ofc and acid, (3) determine
the CCU transformation products, (4) investigate the effect of reaction conditions on the product
distribution, and (5) monitor the aqueous and surface oxidation products of pyrite. Since C>2 is
known to oxidize S22* groups (Moses et al, 1987; Luther, 1990) and to react with the
trichloromethyl radical intermediate (Asmus et al., 1985), both the rates and products of CCU
transformation are expected to be influenced by the presence of C«2.
The pyrite treatments and reaction conditions studied were as follows: air-exposed pyrite
reacted aerobically, air-exposed pyrite reacted anaerobically, air-exposed pyrite reacted in the
presence of sulfide, fresh-ground pyrite reacted anaerobically, and acid-treated pyrite reacted
anaerobically. These conditions (except the acid treatment) were chosen to simulate different
geochemical scenarios. Although pyrite is formed in anaerobic environments, pyrite is often
exposed to aerobic conditions upon weathering (e.g., White et al., 1991). Under oxic conditions,
an iron oxide coating will develop on the pyrite surface and will inhibit the reactivity with O2
(Nicholson et al., 1990) and presumably other oxidants, such as CCU. Additionally, O2 may
compete directly with CCU for reaction sites. The extent of competition or inhibition by O2 and
the effect on the CCU product distribution were investigated. Air-exposed pyrite may be re-
introduced into an anaerobic or sulfide-rich environment since sulfide is often present in plumes
from hazardous waste sites and landfills (Barbash and Reinhard, 1989a). Sulfide may be able to
regenerate the pyrite surface through reductive dissolution of the iron oxide coating (Pyzik and
Sommer, 1981; dos Santos Afonso and Stumm, 1992; Peiffer et al., 1992). Although fresh-
ground pyrite does not necessarily mimic natural unoxidized pyrite because of increases in
surface energy induced by grinding (Papirer et al., 1993), it is the best approximation for natural
pyrite that has not been exposed to oxygen. Acid-treated pyrite was studied because acid
treatment is commonly used in pyrite dissolution and oxidation research to remove high strain
areas and to obtain a reproducible surface (Goldhaber, 1983; Taylor et al., 1984; McKibben and
Barnes, 1986; Moses et al., 1987; Nicholson et al., 1988). However, the relationship between
acid-treated pyrite and in situ natural pyrite is also uncertain.
Oxidation rates of pyrite by oxidants such as ferric iron, oxygen, and hydrogen peroxide
have been measured under different reaction conditions (Wiersma and Rimstidt, 1984;
McKibben and Barnes, 1986; Moses and Herman, 1991). In these studies, the rates of pyrite
oxidation were measured either by the disappearance of the oxidant or appearance of sulfate.
This indirect method yields accurate results if the stoichiometry of the reaction under
consideration is known. Assuming that pyrite-S is oxidized to SO42' while the iron redox state is
unchanged, the oxidation of pyrite by 62 and Fe3+ can be described by the overall reactions
(Moses and Herman, 1991)
(15)
and
FeS2(s) + l4Fe^} + 8#20(0 -» ISFfa + 25$^ + 16J3J, . (16)
If the appearance of sulfate or the disappearance of ferric iron are used to monitor the rate of
pyrite oxidation, then the pyrite oxidation rate equals
30
-------
Rate= _
dt 2 dt 14 dt
When using SC>42' to monitor the pyrite oxidation rate, it is assumed that relatively stable
intermediates are not formed. The only aqueous sulfur intermediate detected in the oxidation of
pyrite at circumneutral pH was S2(>32', but its accumulation was negligible (Moses et aL, 1987).
The stoichiometry for the oxidation of pyrite by CCU is unknown. Assuming that the
oxidation with O2 and CCU are analogous, we may hypothesize
FeS2(s) +14CC/4(- + 8ff20(/) -» Fe +2S(% + 14qJC-(-rt +I4C£,> +16^,, .(18)
2(s) 4(-rt
where 14 moles of CCU react with 1 mole of pyrite to form the intermediate ClsC- and sulfate.
Under anaerobic conditions with CCU as the only oxidant, pyrite-S may not be fully oxidized to
SC>42-; partially oxidized sulfur compounds such as polysulfides or thiosulfate may be relatively
stable. This equation also assumes that CCU is reduced only to ClsG; whereas for formation of
CHCls, CO, or HCOOH, two electrons are required resulting in a stoichiometric coefficient of 7
rather than 14, thereby doubling the pyrite oxidation rate. By monitoring the disappearance of
CCU, we can use a stoichiometric coefficient of 14 to provide a low estimate of the pyrite
oxidation rate, thus enabling us to compare the reactivity of pyrite with CCU and other oxidants.
Effect of Pyrite Pretreatment on CCl^ Transformation Rate
In order to establish the reaction kinetics for the disappearance of CC14 with 1.2 m2 pyrite,
the data were fit to both a first- and a zeroth-order reaction model. Rate constans were obtained
for the first- and the second-order models were obtained from the slopes of semi-logarithmic
plots of [CCU]/[CCU]0 versus time and linear plots of [CCU]/[CCU]0 versus time, respectively.
An example of the results of the kinetic model fits for the O2-Exp/Aer and Acid/An experiments
are depicted in Figure 12. The data showed a much better adherence to zeroth-order model than
to the first-order model. A poor fit was found for the fresh-ground system (R2 = 0.65)
presumably due to heterogenous nature of the freshly cleaved surfaces.
A zeroth-order dependence on oxidant concentration is expected when a heterogeneous
reaction is chemically controlled, rather than diffusion controlled (Goldhaber, 1983). Rate
constants for the disappearance of CCU (k°CCu) are shown in Table 5. The rate constants were
normalized by the pyrite surface concentration, assuming the reaction was first-order with
respect to the surface concentration, SC (Lasaga, 1981; Moses and Herman, 1991), according to
the equation
(19)
where SC is the product of the specific surface area of the solid [m2/g] and the solids loading
[g/L].
31
-------
Figure 12:
n O2-Exp/AerData
- - • First Order Fit: O2-Exp/Aer
Zeroth-OrderFit O2-Exp/Aer
A Acid/An Data
First-Order Fit Acid/An
Zeroth-OrderFit Acid/An
0
Time (day)
Disappearance of CCU in the presence of air-exposed pyrite reacted under aerobic
conditions and acid-treated pyrite reacted under anaerobic conditions at 25 °C.
Data are fit with both first- and zeroth-order kinetic models.
Table 5: Zeroth-order rate constants for CCU transformation with pyrite (1.2-1.4 m^/L)
reacted under aerobic and anaerobic conditions at 25 °C.
Pyrite Conditions
Air-exposed, Aerobic
Air-exposed, HS-
Air-exposed, Anaerobic
Fresh-Ground
Acid-Pretreated
Slope (d-l)
0.025
0.031
0.057
0.056
0.082
R2
0.93
0.87
0.85
0.65
0.96
k°CCl4
(mol/m2- d)
0.021
0.026
0.047
0.039
0.053
95% Confidence
Interval
0.017-0.026
0.020-0.032
0.035-0.049
0.022-0.056
0.046-0.060
In the acid-treated pyrite system, >90% of 1 uM CCU was transformed within 12 days at
25 °C, whereas half-lives in homogeneous solution are 1400 days for 1 mM US' at 25 °C and
105 days for 0.1 mM Fe2+aq at 50 °C (Kriegman-King, 1993). Assuming an activation energy of
60-120 kJ/mol, the half life of CCU with 0.1 mM Fe2+aq at 25 °C ranges from 700-4500 days.
The aqueous Fe2+ and HS" concentrations in suspensions of pyrite in deoxygenated Milli-Q
water have recently been measured to be approximately 0.1 mM, in a 1:2 ratio, respectively (L.
Ronngren and S. Sjoberg, University of Umea, Sweden, personal communication, 1993),
suggesting that the solubility product of pyrite is higher than reported previously. The CCU
32
-------
transformation rates that were measured in the Fe2+ and HS' aqueous systems were at
concentrations that would be expected in the pyrite systems. The rate data therefore support a
surface-controlled reaction mechanism for the transformation of CCU by pyrite because (1)
zeroth-order kinetics were observed and (2) the reaction with pyrite was much faster than in
homogeneous solutions. Over longer time periods, however, the reaction may become diffusion
controlled when an iron oxide coating forms on the pyrite and the oxidant and products have to
diffuse through the oxide coating (Goldhaber, 1983; Nicholson et al., 1990).
In Table 5, the data show that CCU reacts the fastest with the acid-treated pyrite, although
the air-exposed pyrite reacted anaerobically is not statistically slower. As expected, the slowest
transformation rate was observed when CCU was reacted with pyrite under aerobic conditions.
However, the rate constant was only 2.5 times slower than the acid-treated system. The large
error associated with the fresh-ground pyrite system precludes comparison with other rate
constants. The stresses induced by grinding (Papirer et al., 1993), however, may still be
responsible for the scatter in the fresh-ground/anaerobic system. This heterogeneity was likely
removed during pretreatment by O2 or acid.
The rate constants are compared to rates of pyrite oxidation by O2, Fe3*, and H2O2 in
Table 6. In order to compare pyrite oxidation rates by CCU to rates by other oxidants, it is
assumed that CC14 oxidizes pyrite-S to SO42". Using eq. 18, the molar pyrite oxidation rate is
1/14 the CCU disappearance rate. When the pyrite oxidation rate is calculated from the
disappearance rate of CCU. competing reactions by other oxidants such as O2 are not accounted
for. Although solution conditions and pretreatments differ, it is clear that CCU reacts with pyrite
as fast or faster than 02 and Fe3+ (Table 6).
Table 6: Comparison of pyrite oxidation rate by CCU with literature rates of oxidation by
62, Fe3+, and H2O2 at room temperature.
This work
Moses and
Herman (1991)
McKibben and
Barnes (1986)
Wiersma and
Rimstidt (1984)
Pretreatment
Air-exposure
Air-exposure
Mild HC1 Wash
Boiling HC1 Wash
(t
u
Mild HNOs Wash
41
((
None
pH
6.5
6.5
6.5
6.0
6.0
6.0
2.0
2.0
2.0
2.0
Oxidant
CCU, 02
ecu
ccu
Fe(in)
02
O2, Fe(m)
Fe(DI)c
O2C
H202c
Fe(ffl),02
Oxidation Rate
[nmol/m2-s]
15a,b
45a
70s
10
1
0.5
5
500
2X106
10
a Assumes a stoichiometry of 14 moles of CCU reduced to C\3,C- for 1 mole of pyrite oxidized to
so42-.
b Represents pyrite oxidation rate due to CCU, not the total rate due to both O2 and CCU-
c Did not observe zeroth-order dependence for oxidants. Used 0.3 mM oxidant concentration to
calculate rate.
In environmental situations where CCU will likely be present with O2 and Fe3+, pyrite
oxidation by CCU will not be necessarily inhibited or out-competed by O2 and Fe3+. Hydrogen
33
-------
peroxide reacts with pyrite orders of magnitude faster than CCU, O2, or Fe3*. However, an
oxide coating will develop, and with time, the reactivity of the pyrite surface toward CCU, or any
oxidant, will eventually become diffusion controlled (Nicholson et al., 1990).
Reaction of air-exposed pyrite in the presence of sulfide shows that treatment of an
oxidized pyrite surface with HS* does not restore the reactivity of pyrite. Rather, sulfide appears
to inhibit die transformation of CCU by pyrite, even relative to the air-exposed/anaerobic pyrite
system. At pH 7.75, 85% of the sulfide is present as HS*; and more than 100 pM is present as
H2S. Because of the observed zeroth-order dependence on CCU, reaction sites on pyrite are
inferred to be saturated with CCU when [CCU]=luM and [H2S] is 100 times more concentrated
than CCU, it is conceivable that H2S blocks CCU reaction sites. Characterization of the pyrite
surface chemistry is necessary to understand the interaction of sulfide species with the pyrite
surface.
CCU Transformation Products
As shown in Table 7, the CCU product distribution varies greatly depending on the reaction
conditions even though k°cci4 only varies by a factor of 2.5.
Table 7: CCU product distribution from reaction with pyrite under aerobic and anaerobic
conditions at 25 °C.
Condition
(Time)8
Air-exposed;
Aerobic
(42 d)
Air-exposed,
HS- (31 d)
Air-exposed,
Anaerobic
(20d)
Fresh-Ground
(13d)
Acid-
Pretreated
(13d)
CC14
0-1%
0-10%
0-1%
1%
6-10%
CHCb
5-6%
21-22%
28-30%
48%
20-21%
CS2
11-15%
NMe
0-3%
2%
19-20%
CO2
52-59%
NM
26-30%
10%
17%
Formateb
2%
NM
7-9%
5%
4%
Adsorbed'
10%
(2%NV
8%C02)
NM
12%
(7%NV
5%CO2)
12%f
9%
(2%NV
7%CO2)
Mass
Balance*1
84-87%
NM
82-84%
78%
78%
a Reaction time in days is in the parentheses.
b Formate was not directly measured in these experiments. The formate concentration was
assumed to equal the non-volatile concentration measured using 14C analysis.
c Adsorbed amount does not account for volatile compounds adsorbed. NV=non-volatiles.
d Mass balance of aqueous volatile compounds was obtained in all cases. Total radioactivity in
solution + adsorbed non-volatile and CO2 fractions are equal to the mass balance within 5%.
Missing fraction likely to be adsorbed volatiles.
e NM=not measured.
f Breakdown of adsorbed products not measured.
34
-------
Under aerobic conditions, the major product was CO2 (60-70%). Including the hydrolysis
of CS2 to CO2 (Adewuyi and Carmichael, 1987; Kriegman-King and Reinhard, 1992), CO2
accounts for 70-80% of the CCU transformed. In contrast, the fresh-ground pyrite system forms
approximately 50% CHCls and ultimately only 10-20% CO2- The total CO2 amount in the
fresh-ground system is a rough estimate because the speciation of the adsorbed fraction was not
measured. Interestingly, some CS? was formed in all systems suggesting that the CCU or
reactive intermediates react with 82*" sites on pyrite, even in the presence of 02. The volatile
fraction measured from scintillation counting agreed with the sum of the CCU> CHCls, and €82
fractions, that CO and C^Cltf are not formed in these systems. A loss in total radioactivity from
solution as a function of time suggests that a fraction of the CCU or its transformation products is
adsorbed to pyrite. Assuming the method to measure adsorbed CO2 and HCOOH is valid, it is
likely that the missing fraction is a compound that is volatile when desorbed and is not detected
with our experimental methods.
In an experiment conducted anaerobically with 100 jiM CCU and 1.2 m2/L pyrite, the
appearance of formate was monitored. As shown in Figure 13, after disappearance of 95% of the
CCU, approximately 4-6 fiM formate was detected, accounting for 4-6% of the initial CCU
concentration. For comparison, in l^C-labeled CCU experiments, approximately 5% of the
CCU is transformed to a non-volatile product when reacted with pyrite anaerobically (Table 7).
The direct measurement of formate by ion chromatography and its agreement with the fraction of
non-volatile products measured by scintillation counting agree with the above assumption that
the non-volatile compound is formate.
u
•^1
u
u
^^
e
4->
e
V
w
e
0
l.2~
1f\ J
.O1
•
0.8-
•
0.6-
0.4-
0.2-
]^
o
D
B
D D
O
D
D
• : : x *
D CC14
• HCOOH
1
0
10
Time (day)
15
20
Figure 13: The disappearance of 100 uM CCU with 1.2 m^/L pyrite reacted anaerobically at
25 °C. Appearance of formic acid was measured with an ion chromatograph.
35
-------
Air-Exposed Pyrite/Aerobic-
Using the product data as a function of time, the rate constants for the formation of the
products can be estimated. Because the disappearance of CCU was zeroth order, the appearance
of the products is also assumed to be zeroth order except when secondary products are formed
that require more complex rate laws. The rate constants for the appearance of CHQs and
HCOOH (k'cHCls an^ k°NV> respectively) were calculated assuming a zeroth-order rate law
(Table 8). Under aerobic conditions, CO2 can form via hydrolysis of CS2 (Kriegman-King and
Reinhard, 1992) or via reaction of O2 with the trichloromethyl radical (Asmus et al., 1985). If
any CO2 is formed via hydrolysis of CS2, then the appearance of CO2 cannot be modeled by a
zeroth-order equation. Assuming COi is formed only by the CS2 pathway, eqs. 20 (a, b) can be
used to solve for [€82] and [CO2] as a function of time.
and
(20 a, b)
where k^ is a zeroth-order rate constant for the appearance of €82 and JfcCOl is a first-order rate
constant for the formation of CO2- Using SYSTAT (SYSTAT, Inc, Evanston, IL), the €82 data
were fit with the equation
(21)
to solve for the rate constants,
Table 8:
and kCOi (Table 8). These constants were then substituted
Rate constants for the disappearance of CCLt and appearance of intermediates and
products from reaction with pyrite under aerobic and anaerobic conditions at
25 °C.
Rate Constant8
keen [mol/m2-d]
k° [mol/nf-d]
k°CSl [mol/m2-d]
kcsi [L/m2-d]
kc02 [L/m2-d]
kCHCl, [mol/m2.d]
k°m [mol/m2-d]
Air-exposed Pyrite
Reacted Aerobically
No Interm.b
R2adi=0-78d
0.021
—
0.0078
—
0.040
0.00092
0.00040
Intermediate0
R2adi=0.85
0.021
0.020
~
0.012
0.012
0.00092
0.00040
Acid-treated Pyrite
Reacted Anaerobically
No interm.b
R2adl-=0.85
0.053
~.
0.022
—
0.12
0.012
0.0027
Intermediate0
R2adi=0-85
0.053
0.023
--
7.7
0.12
0.012
0.0027
a k°= zeroth-order rate constant; k - first-order rate constant Symbols are defined in text
c CCU -> Int -> CS2 -> CO2
d R2adj accounts for the number of fitting parameters.
36
-------
than was measured, suggesting that CO2 is formed via the reaction of CCly with O2. However,
the curve for the appearance of €82 also does not fit the data well (R2adjusted=0-78). There is a
time lag before [€82] starts to increase, suggesting formation of a relatively stable intermediate
in the path to form CS2-
0
10
20 30
Time (day)
40
Figure 14:
Disappearance of CCU in the presence of pyrite under aerobic conditions at 25
°C. Appearance of the products, CS2 and CO2, with model results assuming the
only path to form CO2 is CCU -> CS2 -> CO2.1
If it is hypothesized that a stable intermediate is formed, then the appearance of CS2 can be
modeled using the equations,
d[Intermed.]
Jt
= k° - kcs^ [Intermed.]
(22)
(23)
where k° is the zeroth-order rate constant for the formation of the intermediate. In eqs. 22 and
23, the rate constant for the appearance of CS2 (^C52) ** now assumed to be first order. Using
the solution to eq. 22 which is of the same form as eq. 21, eq. 23 can be solved to give the CS2
concentration,
[C&] =
(24)
37
-------
The curve-fitting results obtained from eq. 24 are shown in Table 8 and Figure 15. There is an
improvement in the €82 fit (R2adjusted=0-85), and the CO2 data are also fit quite well. The curve
predicts formation of 5-10% more CO2 than was measured in solution which might correspond
to the 8% CO2 adsorbed. The appearance of the hypothetical intermediate is also included in
Figure 15. At the end of the experiment, the model predicts that the intermediate attains a
steady-state concentration of approximately 15% which agrees with the missing mass balance
(Table 7).
0 CC14
CS2
"»' C02
Unknown Int.
Figure 15:
20 30
Time (day)
Disappearance of CC14 in the presence of pyrite under aerobic conditions at 25
°C. Appearance of the products, €82 and CO2, with model results assuming the
only path to form CO2 from CCU is: CCLj -> Intermed. -> €82 -> CO2-
Acid-Treated Pyrite/Anaerobic--
A similar fitting analysis was conducted on the results from the acid-treated pyrite system.
In this case, no significant difference in the CS2 fit was observed if the appearance of an
unknown intermediate was included. As shown in Table 8, the rate constant for the
disappearance of the unknown intermediate I (k°) is relatively large, indicating that the
intermediate I is very short-lived. Therefore, k° is thus approximately equal to the rate constant
for the appearance of €82 (fc^,)- In Figure 16 the predicted CO2 concentrations do not
accurately describe the measured CO2 within the analytical precision of ±5%. The CO2 data do
not have a lag; CO2 appears to form without a stable intermediate. Although the over-predicted
CO2 concentrations again likely correspond to the adsorbed CO2 (Table 7), the measured Cp2 do
not have a lag; CO2 appears to form independently of €82 hydrolysis through a yet unidentified
pathway.
38
-------
5 10
Time (day)
15
Figure 16: Disappearance of CCU in the presence of acid-treated pyrite under anaerobic
conditions at 25 °C. Appearance of the products, C$2 and CO2, with model
results assuming the only path to form CO2 is: CCU -> C$2 -> CO2.
Pyrite Oxidation Products
Aqueous Sulfur Oxidation Products—
The presence and absence of CCU did not affect the aqueous sulfate concentrations (data
not shown). In an experiment in which pyrite was reacted anaerobically with 100 ^iM CCU,
consistent formation of 2-4 uM S2O32' was observed. The control, reacted in the absence of
CCU, contained only a sporadic appearance of SOs2" at <3 uM. These data suggest that S2O32'
is an intermediate or product formed during oxidation of pyrite by CCU- S2Oa2" presumably
reacts with CCU or CCU transformation intermediates until it is fufiy oxidized to SO42'. SOs2'
was not detected in any of these systems.
Surface Oxidation Products—
To measure the surface pyrite oxidation products, experiments were conducted with large
pyrite particles both in the presence and absence of 1 mM CCU- After 5 weeks of reaction at 25
°C, no measurable amount of CCU had reacted, but after 5 more weeks of reaction at 50 °C,
approximately 1 uM CCU had reacted.
As shown in Figure 17, XPS analysis of the sulfur 2p peak does not show any effect on the
near-surface of pyrite from reaction with 1 uM CCU- The concentration of pyrite surface sites
that had reacted with CCU is estimated to be 10^- Ifr5 mole/m2 which is too low to be detected
using XPS. Using scanning tunneling microscopy (STM), early stages of oxidation of pyrite in
air have been observed to occur at isolated spots or patches (C. Eggleston, EAWAG, Diibendorf,
Switzerland, personal communication, 1993). Techniques such as in situ STM or atomic force
39
-------
microscopy (AFM) may be more helpful to study surface oxidation of pyrite by CCL* in the
future.
No CC14
f V^^^AVA^A^V
•_ » I
169
165 161
Binding Energy (eV)
157
Figure 17: XPS spectra comparing S 2p peak for pyrite reacted under anaerobic conditions in
the presence and absence of 1 uM CCL*.
0
738 731
724 717 710
Binding Energy (eV)
703
Figure 18: XPS spectra of Fe 2p peak of air-exposed pyrite reacted aerobically. The Fe(IQ)
peaks are indicative of an iron oxide coating on the pyrite.
40
-------
Oxidation of the near-surface of pyrite due to reaction under aerobic conditions was
observed using XPS. The pyrite reacted aerobically for 5 and 10 weeks showed significant
formation of an iron oxide at the surface.
en
V
o
540
—i " 1 ' 1
536 532 528
Binding Energy (eV)
524
520
Figure 19:
XPS spectra of O Is peak of air-exposed pyrite reacted aerobically. The peaks at
530.1 and 531.2 eV coincide with the O Is peak for FeOOH (Moulder et al.,
1992).
The iron oxide is assumed to be FeOOH, as evidenced by the shape of the Fe 2p line
(Figure 18) and the O Is line (Figure 19) (Moulder et al., 1992). The peaks at 530.35 eV and
531.77 eV are the same size and width, within 5%. According to Moulder et al. (1992), the
FeQOH and FeOQH peaks occur at 530.1 and 531.2 eV, respectively. Additionally, the ratio of
FeOOH-Fe to FeOOH-O is 0.45 which is close to the expected ratio of 0.5. Goethite (oc-FeOOH)
and lepidicrocite (y-FeOOH) cannot be distinguished using XPS. The O Is peak in other iron
hydroxides, such as Fe(OH)3, would appear as one peak because the oxygens are in an identical
environment The O Is peak for water is at 533.2 eV (Moulder et al., 1992), which likely
corresponds to the broader peak at high binding energy (532.85 eV). The peak energies in Figure
19 do not correspond exactly to those in the literature, probably due to the electronic difference
between pure FeOOH and an FeOOH coating on pyrite. The mechanism of formation of iron
hydroxides on pyrite is proposed to occur via surface complexation of Fe(OH)3_nIH' species on
the pyrite surface that ultimately recrystallize to goethite (Fornasiero et al., 1992). Therefore it is
assumed that the iron oxide coating formed on pyrite is goethite. There was not a significant
difference in the sulfur peak (2p) from the air-exposed/aerobic samples compared with the fresh-
ground pyrite reacted anaerobically (Figure 20). The sulfur peaks were mostly attributed to
pyrite-sulfur (>90%) (Hyland and Bancroft, 1990; Szargen et al., 1992). Small shoulders
observed on the low and high binding energy sides of the sulfur 2p peak can be attributed to FeS
(Moulder et al., 1992) and partially oxidized sulfur, such as polysulfides and elemental sulfur
(Hyland and Bancroft, 1990; Szargen et al., 1992), respectively. Small differences in these
shoulders observed under distinct reaction conditions were too small to be quantified.
41
-------
Under all reaction conditions, the near-surface of the pyrite was depleted in ferrous iron.
The S:Fe(II) ratio was greater than 4 in all cases (Table 9), whereas it was 2.1 for fresh-cleaved
pyrite. The depth of analysis using XPS decreases by the factor (1 - e'n) where n is the number
of the attenuation lengths of the photoelectron, (A,). Therefore, 63% of the signal is within 1 X.
Since X=18A for the S 2p peak (Baltrus and Proctor, 1990), the depth of this leached layer is
probably a several unit cells thick (5.4 A per unit cell). With more than one in two Fe2+ leached
from the near surface of the pyrite lattice, a shift in the S 2p peak is expected. This shift would
be evidenced as peak broadening of the S 2p due to contributions of both FeS2-S and leached
82-8. However, no significant differences were observed in the S 2p peak shape for fresh-
cleaved pyrite and pyrite reacted in water for 10 weeks (data not shown). It appears that iron is
being leached from the pyrite surface while the electronic structure of the sulfur is apparently
unchanged.
Fresh-ground/
Anaerobic
Air-exposed/
Aerobic
169
165 161
Binding Energy (eV)
157
Figure 20:
XPS spectra of the S 2p peak comparing fresh-ground pyrite reacted anaerobically
with air-exposed pyrite reacted aerobically. There is no significant difference in
the S 2p peak shapes and binding energies under these reaction conditions.
In experiments with sphalerite (ZnS) at pH 5-7, Ronngren et al. (1991) observed leaching
of Zn2+ to solution, without a significant release of S. The appearance of Zn2+ in solution
corresponded with a loss of two times the number protons from solution, suggesting a hydrolysis
or ion exchange reaction of the form,
2H++
}SH2
.2+
(25)
(Ronngren et al., 1991). Similarly, ion exchange of 2H+ for Pb2+ was observed for galena (Sun
et al., 1991). Although the solution chemistry was not quantified for pyrite, the XPS data
intimate that a similar leaching reaction is occurring in these systems. The binding energy for
S 2p in a species such as S2H2 is not known, however it is expected to be at higher binding
energy because the nuclei of H22+ are less shielded than Fe2+. Nonetheless, it is conceivable that
the binding energy shift of S in H2S2 is less than the resolution of the instrument (0.8 eV), thus
42
-------
explaining the lack of peak broadening in the S 2p peak. The effect of this leaching on the
reactivity of the pyrite toward oxidation is unknown.
Table 9: Effect of environmental conditions on S:Fe(II) in the near-surface of pyrite where
S:Fe(II) was determined using XPS.
Reaction Conditions
Fresh-cleaved pyrite
Air-exposed pyrite, Aerobic*
Air-exposed pyrite, Sulfide3
Air-exposed pyrite, Anaerobic3
Fresh-ground pyrite, Anaerobic3
Acid-treated pyrite, Anaerobic3
S:Fe(n)
. 2.1
4.6
6.7
6.0
4.7
4.8
3 After reaction in water for 10 weeks.
Proposed Mechanism at Pyrite Surface
Sulfur is the proposed electron transfer site in reactions of CCU with pyrite because the
surface was depleted in iron and €82 was detected under all reaction conditions. Since the pyrite
surface was negatively charged under the reaction conditions in this study, the surface sites are
proposed to be predominantly of the form, >FeSS*. It is assumed that the amphoteric nature of
the leached pyrite surface is similar to that of pyrite-S. In the absence of oxidation.of the pyrite
surface, the reaction with CC14 is proposed to occur via the reaction,
)FeS - 5f + CC/4 -» [ }FeS -S-+ CC/4
)FeS -S- CO,
(26)
It is unlikely that a concerted Stf2 reaction occurs between pyrite-S and CCU. Rather, a radical
reaction is proposed above because of the resonance stability of the three-electron C-C1 bond in
CCU-" (Shaik, 1983; Pross; 1985 a, b). An electron would be transferred from the n* orbital on
the pyrite-sulfur to a c* orbital in CCU. Since CCU or an intermediate must react directly with
the pyrite surface, formation of the proposed FeSSCCls intermediate is plausible. Subsequent
reaction of FeSSCCla is uncertain, but speculation is summarized in Figure 21 and discussed
below.
In the fresh-ground pyrite system, CHCls is the major product The reaction probably
proceeds with an additional electron to form a trichloromethyl anion (:CCl3') which can react
with a proton in solution to form CHCls. The pyrite-sulfur will be hydroxylated and
subsequently oxidized until a sulfoxy species forms that is stable in solution. No sulfoxy species
were detected on the pyrite surface using XPS, either because the concentrations were too low or
the hydroxylation reactions are fast. As predicted in the above pathway, the sulfoxy species,
', was detected in systems with extensive reaction by CCU-
43
-------
tl
^
cf
Carbon tetrachloride
->FeSS"
Carbon disulfide
Carbon dioxide
Figure 21: Proposed CCU transformation pathways with pyrite. Compounds in shadowed
boxes are measured intermediates or products. Compounds in brackets are
proposed intermediates.
44
-------
The path to form CS2 will depend on the strength of the S-S bond relative to the S-
bond. As indicated in Figure 21, FeSSCCls may react with OH' to form FeSOH and "SCCls
(trichloromethylthiolate) or FeOH and 'SSCCla (trichloromethyldithiolate). Both the
trichloromethyldithiolate and the trichloromethylthiolate may decompose in separate pathways to
form thiophosgene (dichlorothionomethane) which may react again with another sulfide group at
the pyrite surface to form CSi- If a one-electron transfer is assumed as the intial step (Equation
26), the carbon in CCU (+IV) is first reduced to form the transition state {Cl-CCly (+HI)}
followed by oxidation to form €82 (+IV). In this reaction scheme the disulfide group on pyrite
undergoes a disproportionation reaction wherein one sulfur of the disulfide group is reduced to
the -Q state (€82) and one sulfur of the disulfide is oxidized to the 0 state (>FeSOH).
The path to form CO2 can occur via hydrolysis of CS2 or reaction of the trichloromethyl
radical with O2- The former pathway is assumed to prevail under anaerobic conditions.
However, under aerobic conditions, CO2 is the major product in the pyrite system and both
pathways are possible. Although the modeling results were inconclusive, they were consistent
with a pathway in which all the CO2 is formed by hydrolysis of CS2- The formation of €82
under these conditions was proposed to occur via a stable intermediate that is adsorbed to the
surface. XPS data show significant formation of FeOOH on the pyrite surface. Goethite can
influence the reaction of CCU in several ways. Electron transfer can occur across the iron oxide
coating because FeOOH is a semi-conductor. CCU may be accepting an electron at the
FeOOH/water interface to form Clj,C-. The radical can then react with O2 to form CO2 or with
pyrite-S to form CS2- It seems unlikely that Cl^C- would be formed at the oxide surface and then
react with pyrite-S because the reaction rate of ClsC- with O2 is close to the diffusion limit
(Asmus et al., 1985). Lastly, CCU may be reacting with pyrite-S to form an intermediate such
as, •SSCCls, that is stabilized at the FeOOH/water interface since the pHpzc of goethite is 7.3-
8.3 (Kingston et al., 1972; Sigg and Stumm, 1981). Although the data preclude determination of
the pathway to form CO2, it is clear that the presence of O2 significantly affects the CCU
transformation pathway.
Summary of Pyrite Experiments
Pyrite was shown to be very reactive toward CCU- Under all environmental conditions
studied, >90 % of 1 uM CCU reacted within 12-36 days at 25 °C in the presence of 1.2-1.4 m2/L
pyrite. Zeroth-order dependence on the CCU concentration was observed, suggesting a surface-
controlled reaction. The rates of transformation of CCU under all conditions are fast enough
such that oxidation of pyrite by CCU should not be curtailed in the presence of competing
oxidants, such as Fe^*" and O2- The CCU product distribution was very dependent on reaction
conditions. Under aerobic conditions, CC«2 was the major transformation product; whereas in the
fresh-ground pyrite system, CHCls was the major product
Although these laboratory rates of CCU transformation by pyrite cannot yet be extrapolated
to field conditions because of the confounding effects of natural organic matter, co-solvents, and
competing oxidants, the reactivity under different environmental conditions can be compared
qualitatively. In a sulfate reducing environment, the transformation rate of CCU may be retarded
by the presence of sulfide, whereas in the absence of aqueous sulfide; the fastest transformation
rates would be expected. Under aerobic conditions, oxide coatings on pyrite will likely retard the
transformation of CCU; but the transformation products are environmentally desirable.
45
-------
REFERENCES
Adewuyi, Y.G. and G.R. Carmichael (1987) Kinetics of hydrolysis and oxidation of carbon
disulfide by hydrogen peroxide in alkaline medium and application to carbonyl sulfide,
Environ. Sci. Technol., 21: 170-177.
Anderson, L.D., D.B. Kent, and J.A. Davis (1992) Batch experiments characterizing the
reduction of Cr(VI) using suboxic material from a mildly reducing sand and gravel aquifer,
Abst Am. Chem. Soc. Mtg., April 5-10,1992, San Francisco, CA.
Anderson, L.D., D.B. Kent, and J.A. Davis (1994) Batch experiments characterizing the
reduction of Cr(VI) using suboxic material from a mildly reducing sand and gravel aquifer
Environ. Sci. Technol., in press.
Asmus, K.D., D. Bahnemann, K. Krischer, M. Lai, and J. Monig (1985) One-electron induced
degradation of halogenated methanes and ethanes in oxygenated and anoxic aqueous
solutions, Life Chem. Rep., 3: 1-15.
Baltrus, J.P. and A. Proctor (1990) Composition of overlayers on oxidized pyrite surfaces, Appl.
Surf. ScL, 44: 147-150.
Barbash, I.E. and M. Reinhard (1989a) The Reactivity of Sulfur Nucleophiles Toward
Halogenated Organic Compounds in Natural Waters, in Biogenic Sulfur in the
Environment, E.S. Saltzman and WJ. Cooper, eds., ACS Symposium Series 393, American
Chemical Society: Washington, D.C., pp 101-138.
Barbash, I.E. and M. Reinhard (1989b) Abiotic dehalogenation of 1,2-dichloroethane and 1,2-
dibromoethane in aqueous solution containing hydrogen sulfide, Environ. Sci. Technol., 23:
1349-1358.
Barber, L.B., n, E. M. Thurman, and D.D. Runnells (1992) Geochemical heterogeneity in a sand
and gravel aquifer: Effect of sediment mineralogy and particle size on the sorption of
chlorobenzenes, /. Contain. Hydrol., 9: 35-54.
Begheijn, L.Th. (1979) Determination of iron(H) in rock, soil, and clay, Analyst, 104: 1055-
1061.
Burneau, A. and O. Barres (1990) Comparative study of the suface hydroxyl groups of fumed
and precipitated silicas. 2. Characterization by infrared spectroscopy of the interactions
with water, Langmuir, 6: 1364-1372.
Castro, C.E. and W.C. Kray, Jr. (1963) The cleavage of bonds by low valent transition metal
ions. The homogeneous reduction of alkyl halides by chromous sulfate., /. Am. Chem.
Soc., 85: 2768-2773.
Castro, C.E. and W. C. Kray, Jr. (1966) Carbenoid intermediates from polyhalomethanes-and
chromium(JJ). The homogeneous reduction of geminal halides by chromous sulfate., J.
Am. Chem. Soc., 88: 4447-4455.
46
-------
Criddle, C.S. and P.L. McCarty (1991) Electrolytic model system for reductive dehalogenation in
aqueous environments, Environ, Sci. TechnoL, 25: 973-978.
Criddle, C.S., J.T. DeWitt, and P.L. McCarty (1990) Reductive dehalogenation of carbon
tetrachloride by Escherichia coli K-12, AppL Environ. Microbiol., 56: 3247-3254.
Criddle, C.S., P.L. McCarty, M.C. Elliott, and J.F. Barker (1986) Reduction of hexachloroethane
to tetrachloroethylene in ground waterwater, J. Contam. Hydrol., 1: 133-142.
Curtis, G.P. (1991) Reductive dehalogenation of hexachloroethane and carbon tetrachloride by
aquifer sand and humic acid., Ph.D. Dissertation, Department of .Civil Engineering,
Stanford University, Stanford, CA.
Curtis, G.P. and M. Reinhard (1993) Reductive dehalogenation of hexachloroethane and carbon
tetrachloride and bromoform by anthrahydroquinone disulfonate and humic acid (submitted
Environ. Sci. TechnoL, March, 1993).
Davis, J. A. and J.D. Hem (1989) The surface chemistry of aluminum oxides and hydroxides, In
The Environmental Chemistry of Aluminum., (G. Sposito, ed.), Chapter 7, CRC Press, Inc.,
Boca Raton, FL.
Deer, W.A., R.A. Howie, and J. Zussman (1982) An Introduction to the Rock Forming Minerals,
Longman Group Limited, Essex, England.
dos Santos Afonso, M. and W. Stumm (1992) The reductive dissolution of iron(IQ) (hydr)oxides
by hydrogen sulfide, Langmuir, 8: 1671-1675.
Fadrus, H. and J. Maly (1975) Suppression of iron(in) interference in the determination of
iron(n) in water by the 1,10-phenanthroline method, Analyst, 100: 549-554.
Fornasiero, D., V. Eijt, and J. Ralston (1992) An electrokinetic study of pyrite oxidation,
Colloids Surf., 62: 63-73.
Fritz, S.F. and R.K. Popp (1985) A single-dissolution technique for determining FeO and Fe2O3
in rock and mineral samples, Amer. Mineral., 70: 961-968.
Goldhaber, M.B. (1983) Experimental study of metastable sulfur oxyanion formation during
pyrite oxidation at pH 6-9 and 30 °C, Amer. J. Sci., 283: 193-217.
Haag, W. and T. Mill (1988) Some reactions of naturally occurring nucleophiles with
haloalkanes in water, Environ. Toxicol. Chem., 7: 917-924.
Hileman, B. (1993) Concerns broaden over chlorine and chlorinated hydrocarbons, Chem. Eng.
News, April 19,1993, pp 11-20.
Kingston, F.J., A.M. Posner, and J.P. Quirk (1972) Anion adsorption by goethite and gibbsite. L
The role of the proton in determining adsorption envelopes, J. Soil Sci., 23: 177-192.
Hovenkamp, S.G. (1963) Sodium dithiocarbonate as a by-product in xanthating reactions. A
contribution to the chemistry of viscose, J. Polymer Sci., Pan C., No. 2,341-355.
47
-------
Howarth, R.W. and J.M. Teal (1979) Sulfate reduction in a New England salt marsh, Limnol.
Oceanogr.,24: 999-1013.
Huerta-Diaz, M.A. and J.W. Morse (1992) Pyritization of trace metals in anoxic marine
sediments, Geochim. Cosmochim. Acta, 56: 2681-2702.
Hyland, M.M. and G.M. Bancroft (1989) An XPS study of gold deposition at low temperatures
on sulphide minerals: reducing agents, Geochim. Cosmochim. Acta, 52: 385-394.
Hyland, M.M. and G.M. Bancroft (1990) Palladium sorption and reduction on sulphide mineral
surfaces: An XPS and AES study, Geochim. Cosmochim. Acta, 54: 117-130.
Jeffers, P.M., L.M. Ward, L.M. Woytowitch and N.L. Wolfe (1989) Homogeneous hydrolysis
rate constants for selected chlorinated methanes, ethanes, ethenes, and propanes, Environ.
Sci. Technol., 23: 965-969.
Katser, R.P. and V.M. Ropot (1986) A study of hydrosulfide ion sorption on natural sorbents,
Khim. Teckhnol. Vody, 8: 81-82.
Knauss, K.G. and T.J. Wolery (1989) Muscovite dissolution kinetics as a function of pH and
time at 70 °C, Geochim. Cosmochim. Acta, 53: 1493-1501.
Kreyszig, E. (1983) Advanced Engineering Mathematics, 5th ed, J. Wiley & Sons, New York.
Kriegman-King, M.R. (1993) Abiotic transformation of carbon tetrachloride at mineral surfaces,
Ph.D. Dissertation, Department of Civil Engineering, Stanford University, Stanford, CA.
Kriegman-King, M.R. and M. Reinhard (1991) Reduction of hexachloroethane and carbon
tetrachloride at surfaces of biotite, vermiculite, pyrite and marcasite, In Organic Substances
and Sediments in Water (R.A. Baker, ed.) Vol n, Lewis Publishers, Chelsea, MI, pp 349-
364.
Kriegman-King, M.R. and M. Reinhard (1992) Transformation of carbon tetrachloride in the
presence of sulfide, biotite, and vermiculite, Environ. Sci. Technol., 26: 2198-2206.
Lasaga, A.C. (1981) Rate laws of chemical reactions, In Kinetics of Geochemical Processes,
(A.C. Lasaga, ed.), Reviews in Mineralogy, Vol 8, Mineralogical Society of America,
Washington, D.C., pp 1-68.
Lasaga, A.C. (1984) Chemical kinetics of water-rock interactions, J. Geophys. Res., 89: 4009-
4025.
Lord, C.J., IE and T.M. Church (1983) The geochemistry of salt marshes: Sedimentary ion
diffusion, sulfate reduction, and pyritization, Geochim. Cosmochim. Acta, 47: 1381-1391.
Luther, G.W., III (1990) The frontier-molecular-orbital theory approach in geochemical
processes, In Aquatic Chemical Kinetics, (W. Stumm, ed.), Wiley Interscience, New York,
pp 173-198.
Luther, G.W., m, A. Giblin, R.W. Howarth, and R.A. Ryans (1982) Pyrite and oxidized iron
mineral phases formed from pyrite oxidation in salt marsh and estuarine sediments,
Geochim. Cosmochim. Acta, 46: 2665-2669.
48
-------
Macalady, D.L., P.O. Tratnyek, and T.J. Grundl (1986) Abiotic reduction reactions of
anthropogenic organic chemicals in anaerobic systems: a critical review, J. Contam.
Hydrol,\: 1-28.
McKibben, M.A. and H.L. Barnes (1986) Oxidation of pyrite in low temperature acidic
solutions: rate laws and surf ace textures, Geochim. Cosmochim. Acta, 50: 1509-1520.
Millero, F.J. (1986) The thermodynamics and kinetics of the hydrogen sulfide system in natural
waters, Mar. Chem., 18: 121-147.
Moses, C.O., D.K. Nordstrom, J.S. Herman, and A.L. Mills (1987) Aqueous pyrite oxidation by
dissolved oxygen and by ferric iron, Geochim. Cosmochim. Acta, 51: 1561-1571.
Moses, C.O. and J.S. Herman (1991) Pyrite oxidation at circumneutral pH, Geochim.
Cosmochim. Acta, 55: 471-482.
Moulder, J.F., W.F. Stickle, P.E. Sobol, and K.D. Bomben (1992) Handbook of X-ray
Photoelectron Spectroscopy (J. Chastain, ed.), Perkin-Elmer Corporation, Eden Prairie,
MN.
Mycroft, J.R., G.M. Bancroft, N.S. Mclntyre, J.W. Lorimer, and I.R. Hill (1990) Detection of
sulphur and polysulfides on electrochemically oxidized surfaces by X-ray photoelectron
spectroscopy and Raman spectroscopy, J. Electroanalyt. Chem. and Interfacial.
Electrochem., 292: 139-152.
Nicholson, R.V., R.W. Gillham, and E.J. Reardon (1988) Pyrite oxidation in carbonate-buffered
solution: 1. Experimental kinetics, Geochim. Cosmochim. Acta, 52: 1077-1085.
Nicholson, R.V., R.W. Gillham, and EJ. Reardon (1990) Pyrite oxidation in carbonate-buffered
solution: 2. Rate control by oxide coatings, Geochim. Cosmochim. Acta, 54: 395-402.
Papirer, E., J.M. Perrin, B. Siffert, G. Philipponneau, and J.M. Lamerant (1993) The influence of
grinding on the surface properties of a-aluminas, J. Coll. Int. Sci. 156: 104-108.
Peiffer, S., M. dos Santos Afonso, B. Wehrli, and R. Gachter (1992) Kinetics and mechanism of
the reaction of H2S with lepidicrocite, Environ. Sci. TechnoL, 26: 2408-2413.
Pross, A. (1985a) A general approach to organic reactivity: The configuration mixing model, In
Advances in Physical Organic Chemistry, (V. Gold and D. Bethell, eds.), Academic Press,
San Diego, pp 99-196.
Pross, A. (1985b) The single electron shift as a fundamental process in organic chemistry: The
relationship between polar and electron transfer pathways, Ace. Chem. Res., 18: 212-219.
Pyzik, A.J. and S.E. Sommer (1981) Sedimentary iron monosulfides: Kinetics and mechanism
of formation, Geochim. Cosmochim. Acta, 45: 687-698.
Ronngren, L., S. Sjoberg, Z. Sun, W. Forsling, and P.W. Schindler (1991) Surface reactions in
aqueous metal sulfide systems. 2. Ion exchange and acid/base reactions at the ZnS-H2O
interface, J. Coll. Int. Sci., 145: 396-404.
49
-------
Rowland, F.S. (1991) Stratospheric ozone in the 21st century: The chlorofluorocarbon problem,
Environ. Sci. TechnoL, 25: 622-628.
Sax, N.L. and R.J. Lewis, Sr. (1987) Hazardous Chemicals Desk Reference, Van Nostrand
ReinholdCo., New York, pp 299-300.
Schwarzenbach, R.P. and P.M. Gschwend (1990) Chemical transformations of organic pollutants
in the aquatic environment, In Aquatic Chemical Kinetics, (W. Stumm, ed.), J. Wiley &
Sons, New York, 199-233.
Scofield, J.H. (1976) Hartree-Slater subshell photoionization cross-sections at 1254 and 1487 eV,
J.Elec.Spectr.ReLPhen.,*: 129-137.
Shaik, S.S. (1983) a- and p-carbon substituent effect on Si«j2 reactivity. A valence-bond
approach, J. Am. Chem, Soc., 105: 4359-4367.
Sigg, L. and W. Stumm (1981) The interaction of anions and weak acids with the hydrous
goethite (a-FeOOH) surface, Colloids and Surfaces, 2: 101-117.
Singer, P.C. and W. Stumm (1970) Acid mine drainage~The rate limiting step, Science, 167:
1121-1123.
Sposito, G. (1989) The Chemistry of Soils, Oxford University Press, New York.
Stone, A. and JJ. Morgan (1987) Reductive dissolution of metal oxides, In Aquatic Surface
Chemistry, (W. Stumm, ed.), J. Wiley & Sons, New York, 221-254.
Stumm, W. and JJ. Morgan (1981) Aquatic Chemistry, J. Wiley and Sons, New York, p 456.
Sun, Z., W. Forsling, L. Ronngren, S. Sjoberg and P.W. Schindler (1991) Surface reactions in
aqueous metal sulfide systems. 3. Ion exchange and acid/base properties of hydrous lead
sulfide, Colloids. Surf., 59: 243-254.
Swain, C.G. and C.B. Scott (1953) Quantitative correlation of relative rates. Comparison of
hydroxide ion with other nucleophilic reagents toward alkyl halides, esters, epoxides, and
acyl halides., J. Am. Chem. Soc., 75: 141-147.
Swider, K.T. and I.E. Mackin (1989) Transformations of sulfur compounds in marsh-flat
sediments, Geochim. Cosmochim. Acta, 53: 2311-2323.
Szargan, R., S. Karthe, and E. Suoninen (1992) XPS studies of xanthate adsorption on pyrite,
Appl. Surf. Sci., 55: 227-232.
Taylor, B.E., M.C. Wheeler, and D.K. Nordstrom (1984) Stable isotope geochemistry of acid
mine drainage: experimental oxidation of pyrite, Geochim. Cosmochim. Acta, 48: 2669-
2678.
Tchobanoglous, G. and E.D. Schroeder (1985) Water Quality, Addison-Wesley Publishing Co.,
Menlo Park, CA.
Umana, A.F. (1979) Kinetics of oxidative dissolution of pyrite by aqueous chlorine species,
Ph.D. Dissertation, Department of Civil Engineering, Stanford University, Stanford, CA.
50
-------
Vairavamurthy, A. and K. Moppet (1989) Mechanistic studies of organosulfur (thiol) formation
in coastal marine sediments, In Bio genie Sulfur in the Environment, (E.S. Saltzman and
WJ. Cooper, eds.), ACS Symposium Series 393, American Chemical Society, Washington,
D.C., pp 231-242.
White, A.F. and A. Yee (1985) Aqueous oxidation-reduction kinetics associated with coupled
electron-cation transfer from iron-containing silicates at 25 °C, Geochim. Cosmochim.
Acta, 49: 1263-1275.
White, G.N., J.B. Dixon, R.M. Weaver, and A.C. Kunkle (1991) Genesis and morphology of
iron sulfides in gray kaolins, Clays Clay Min., 39: 70-76.
Wieland, E., B. Wehrli, and W. Stumm (1988) The coordination chemistry of weathering: HI. A
generalization on the dissolution rates of minerals, Geochim. Cosmochim. Acta, 52: 1969-
1981.
Wiersma, C.L. and J.D. Rimstidt (1984) Rates of reaction of pyrite and marcasite with ferric iron
at pH 2, Geochim. Cosmochim. Acta, 48: 85-92.
51
-------
t S
------- |