The Development of Varying Methodologies to Speciate and Monitor the Interactions of Selenium and Environmental Contaminants in Plants

Home , Mustard plant

The development of varying methodologies to speciate and monitor the interactions of selenium and environmental contaminants in plants

A dissertation submitted to the

Division of Research and Advanced Studies of the University of Cincinnati

In partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

In the Department of Chemistry of the College of Arts and Sciences

2008

Scott Ellington Afton

B.S., Chemistry, Andrews University, 2004

Committee Chair: Dr. Joseph A. Caruso

Abstract of dissertation

There is a multitude of contaminated waste sites worldwide due to anthropogenic activities. When assessing the potential toxicological effects of environmental contaminants, prior concern has dealt with simply quantifying the total concentration of the particular contaminating element. With the increasing awareness of the often significant differences in toxicity between the varying environmental contaminants, elemental speciation and percent distribution must be determined. Conventional remediation efforts have been effective in contaminant removal, but are generally very costly. As a result, phytoremediation, which utilizes plants, has recently gained popularity for removing contaminants from soil. It is also known that a toxic concentration of selenium and arsenic/mercury, if administered simultaneously, produce a nontoxic metabolite in a mammal. The studies in this dissertation utilize novel methodologies to speciate and monitor the interactions of selenium and environmental contaminants, arsenic and mercury, in plants.

Eight predominant selenium and arsenic species were simultaneously separated using ion-pairing reversed phase liquid chromatography (IPRP) coupled with inductively coupled plasma mass spectrometry (ICPMS) and electrospray ionization ion trap mass spectrometry within 18 minutes and applied to river water, plant extract and urine matrices. The differences metabolic pathways, including location and identity, of selenium and arsenic species were elucidated after single and simultaneous supplementation in the Chlorophytum comosum, spider plant via size exclusion chromatography (SEC) and IPRP coupled to ICPMS. Although total elemental analysis demonstrates a selenium and arsenic antagonism, a compound containing selenium and arsenic was not found in the general aqueous extract of the plant.

iii

Increasing the difference between the voltage on the extraction lens and the octopole, plus using a positive voltage between the octopole and quadrupole provided the lowest sulfur detection limits using xenon as the collision reaction cell (CRC) gas in ICPMS. Energy

32 + 32 + discrimination is the predominant mechanism for removing the O2 interference of S .

Optimization parameters were also suggested for a standard solution comprised of 52 elements, including sulfur. Similar detection limits were acquired when comparing Xe to He or H2 CRC gas with a general trend comparable to detection levels for He.

Capillary reversed phase chromatography (capRPLC) and SEC were coupled to ICPMS to investigate the metabolic fate of mercury in conjunction with or exclusion of selenium in the

Allium fistulosum. The data suggests a possible selenium-mercury association in a proteinaceous macromolecule which is not readily translocated to the aerial plant regions. Data from x-ray fluorescence mapping of a freshly excised root and capRPLC-ICPMS of homogenized root extract suggest the formation of a mercury-selenium species and a similarly structured mercury- sulfur species predominantly residing in the cell wall of the epidermal root tissue. Utilizing x- ray absorption near edge structure analysis, the local environment of mercury and selenium qualitatively coincided with the mercury-selenium species formed in a mammal via a Hg-Se-

S(GSH) moiety. The local environment of mercury also coincided with (GS)2Hg.

Acknowledgements

This dissertation is dedicated to my parents, Rick and Anne Afton. Besides being kind enough to bring me into the world, they have continuously supported me throughout my many years of education financially, spiritually and educationally. I hope the accomplishment of this dissertation represents a small token of my appreciation. I also would like to thank my parents for the multitude of important life lessons that have been instilled in me such as the seven P’s: prior proper planning prevents pitiful poor performance. I would also like to acknowledge my sister, Danielle, for conversations we have had throughout my graduate career. I could not have come this far without my family’s constant love and support.

I would like to thank my advisor Dr. Joseph Caruso “Doc” for his continued support over the past few years. Of the many benefits gained by joining the Caruso group, Dr. Caruso has granted me almost complete independence to decide the direction and projects of my research.

This has allowed for multiple failures and successes that have made me into the scientist I am today. In addition, he was there to provide support and guidance when needed and entrusted responsibilities to me that allowed for external collaborations that otherwise would not have been possible. He has been a wonderful advisor and friend. I would also like to thank Judy Caruso for her kindness and support. She has always made me feel as though I was part of her extended family.

In addition to my advisor, I would like to thank my committee members Dr. William

Heineman, Dr. Bruce Ault and a previous member Dr. Theresa Reineke. They each have offered invaluable advice and support that aided in reaching my research goals and I greatly appreciate the time and effort they extended. I would also like to thank the rest of the analytical faculty: Dr.

Patrick Limbach, Dr. Thomas Ridgway and Dr. Apryll Stalcup for their support and insightful

vi questions. In addition, I would like to thank Dr. Peter Padolik and Dr. John Breiner for all the good times during my teaching assistantship.

I would like to thank the many collaborators that I have had the privilege of working with that has resulted in multiple publications and still continues to date: Dr. David McNear from the

University of Kentucky; Dr. Steve Sutton and Dr. Matt Newville from Argonne National

Laboratories; Dr. Julio Figueroa, Dr. Katarzyna Wrobel and Dr. Kazimierz Wrobel from the

Universidad de Guanajuato; Dr. Jeff Lehman from Otterbein College; Dr. Mary Beth Genter, Dr.

Zhenyu Qin, Dr. Michael Craig, Elizabeth LaPensee, Dr. Nira Ben-Jonathan, Dr. Erin Haynes,

Dr. Bin Wang, Scott Schneider and Jed Thorn from the University of Cincinnati. I would also like to thank Agilent Technologies and CEM Corporation for continued support through instrumentation, which has enabled the Caruso group the capability to investigate and answer interesting research questions.

I would also like to thank the past and present members of the Caruso group for their support, advice and willingness to incorporate a family atmosphere in the office and laboratory:

Dr. Oktay Cankur, Dr. Baki Sadi, Dr. Juris Meja, Dr. Katie DeNicola, Dr. Monica Shah, Dr.

Santha Yathavakilla, Dr. Sarath Jayasinghe, Dr. Kevin Kubachka, Dr. Douglas Richardson,

Allison Krentz, Heather Trenary, Dr. Jenny Ellis, Dr. Kirk Lokits, Qilin Chan, Yaofang Zhang,

Karolin Kroening, Jen Siverling, Chris Tompson, Dean Stuart, Brittany Catron and Renee Easter.

It would not be possible to list everyone outside the Caruso group that has made my graduate experience enjoyable; however, I would specifically like to thank a few good men, which include Phillip Durham, Michael Haven and Kevin Parker. I thank you for all of the good times, off-color jokes and the plethora of awkward moments that have culminated into my graduate experience, which will be greatly missed.

vii

Last but certainly not least, I would like to thank my best friend and my wife, Dr. Lisa

Afton. I will try not to hold a grudge against you for finishing your doctorate in audiology before me for too long. My wife was a major influence on me when I considered the possibility of attending graduate school. During our simultaneous graduate careers, we have shared in good times and hardships that have brought us closer together and strengthened our marriage. My graduate journey would have been nearly as fruitful without your love and support. I look forward to the many adventures and experiences ahead. I love you sooooooooooooooooooooo much.

-SEA

viii

Table of contents

Abstract of dissertation iii

Acknowledgments vi

Figures 4

Tables 7

Chapter 1 - Methodologies used for biological contaminant remediation and speciation

1.1 Contaminant remediation 1.1.1 Ex situ methodologies 1.1.2 In situ methodologies 1.1.3 Phytoremediation 1.1.4 Selenium, arsenic and mercury remediation 1.2 Inductively coupled plasma mass spectrometry (ICPMS) 1.2.1 Instrument setup and theory 1.2.2 Theory of interference removal with the collision/reaction cell 1.3 High performance liquid chromatography (HPLC) 1.3.1 Size exclusion chromatography (SEC) 1.3.2 Ion-pairing reversed phase chromatography (IPRP) 1.4 X-ray absorption fine structure (XAFS) 1.4.1 Instrument setup and operation 1.4.2 X-ray absorption fine structure theory 1.5 References

Chapter 2 – Simultaneous characterization of selenium and arsenic analytes via ion-pairing reversed phase chromatography with inductively coupled plasma and electrospray ionization ion trap mass spectrometry for detection; applications of river water, plant extract and urine matrices

2.1 Abstract 2.2 Introduction 2.3 Experimental 2.3.1 Instrumentation 2.3.2 Reagents and standards 2.3.3 Sample acquisition 2.3.4 Chromatographic optimization 2.4 Results and discussion 2.4.1 Chromatographic performance

2.4.2 Electrospray ionization studies 2.4.3 Varying matrix applications 2.5 Conclusion 2.6 Acknowledgements 2.7 References

Chapter 3 – The Characterization of Arsenic and Selenium Metabolites in Chlorophytum Comosum, Spider Plant, via IPRP-ICPMS and SEC-ICPMS

3.1 Abstract 3.2 Introduction 3.3 Experimental 3.3.1 Instrumentation 3.3.2 Reagents and standards 3.3.3 Plant growth and supplementation 3.3.4 Total selenium and arsenic determination 3.3.5 Extraction procedures of plant tissues 3.4 Results and discussion 3.4.1 Total element accumulation 3.4.2 Root extract characterization of selenium and arsenic species 3.4.3 Leaf extract characterization of selenium and arsenic species 3.5 Acknowledgements 3.6 References

Chapter 4 – Investigating the mechanisms and feasibility of xenon as a cell gas in inductively coupled plasma mass spectrometry for multi-element detection including sulfur with application to bottled water

4.1 Abstract 4.2 Introduction 4.3 Experimental 4.3.1 Instrumentation 4.3.2 Reagents, standards and materials 4.3.3 Analytical protocol 4.4 Results and discussion + 4.4.1 Elucidating the O2 removal mechanism by Xe 4.4.2 Sulfur optimization via ICPMS 4.4.3 Multi-element detection via sulfur optimization 4.4.4 Evaluation of Xe versus He or H2 as a CRC gas 4.4.5 Quantification of bottled water impurities 4.5 Acknowledgements 4.6 References

Chapter 5 – Comparing the molecular makeup of mercury in conjunction or exclusion of selenium in the Allium fistulosum, green onion

5.1 Abstract 5.2 Introduction 5.3 Experimental 5.3.1 Instrumentation 5.3.2 Reagents and standards 5.3.3 Plant cultivation and supplementation 5.3.4 Total selenium and mercury determination 5.3.5 Extraction procedures of plant tissues 5.3.6 A. fistulosum root extract digestion 5.4 Results and discussion 5.4.1 A. fistulosum total element accumulation 5.4.2 A. fistulosum extract analysis via SEC-ICPMS 5.4.3 Enzymatic proteolysis of the A. fistulosum root extract 5.4.4 Acid hydrolysis of the A. fistulosum root extract 5.4.5 A. fistulosum extract analysis via IPRP/capRPLC-ICPMS 5.5 Acknowledgements 5.6 References

Chapter 6 – Exploring the structural basis mercury/selenium antagonism in Allium fistulosum - green onion

6.1 Abstract 6.2 Introduction 6.3 Experimental 6.3.1 Plant cultivation and supplementation 6.3.2 capRPLC-ICPMS analysis 6.3.3 XAFS analysis 6.4 Results and discussion 6.4.1 Analysis of A. fistulosum root extract via capRPLC-ICPMS 6.4.2 Mercury and selenium location in the A. fistulosum root 6.4.3 Mercury and selenium speciation in the A. fistulosum root 6.5 Acknowledgements 6.6 References

Chapter 7 – Conclusions and future directions

Figures

Figure 1.1 Potential metabolic pathway utilized during contaminant remediation in a plant.

Figure 1.2 A brief summary of the proposed selenium metabolic pathway in a plant, adapted from Terry et al.

Figure 1.3 A block diagram depicting the general schematic of an inductively coupled plasma mass spectrometer.

Figure 1.4 A block diagram comparing a triple quadrupole mass filter used in conjunction with electrospray ionization and a single octopole and quadrupole mass filter used in conjunction with inductively coupled plasma ionization.

32 Figure 1.5 A block diagram representing the three potential mechanisms for O2 discrimination allowing for 32S monitoring.

Figure 1.6 An illustration depicting the general mechanism in size exclusion chromatography resulting in the separation of analyte mixtures.

Figure 1.7 An illustration depicting the general mechanisms in ion-pairing reversed phase chromatography.

Figure 1.8 An illustration depicting the general mechanisms in ion-pairing reversed phase chromatography after an altercation in eluent buffer.

Figure 1.9 A block diagram depicting the general schematic of x-ray absorption fine structure analysis.

Figure 1.10 A block diagram depicting the absorption x-rays (hν) resulting in photo-electron propagation into the continuum and the subsequent x-ray fluorescence from the decay of an excited state electron.

Figure 1.11 A diagram depicting an x-ray absorption pattern through the photoelectric process.

Figure 2.1 An IPRP-ICPMS chromatographic separation of arsenic standards at (A) 250 ng ml-1 and (C) 1 ng ml-1, selenium standards at (B) 250 ng ml-1 and (D) 1 ng ml-1 with SeMet at 5 ng ml-1 and chloride at (A) 1000 µg ml-1.

Figure 2.2 IPRP-ESI-ITMS extracted ion chromatograms of the arsenic standards (top): m/z 107, 137, 139 and 141 and the selenium standards (bottom): m/z 129, 145, 196 and 335 at 10 µg ml-1.

Figure 2.3 IPRP-ICPMS chromatograms of river water, (A) 75As profile and (B) 78Se profile, and ~50 ng ml-1 spiked river water (C) 75As profile and (D) 78Se profile.

Figure 2.4 IPRP-ICPMS chromatograms of Allium Fistulosum root extract, (A) 75As profile and (B) 78Se profile, and ~50 ng ml-1 spiked Allium Fistulosum root extract (C) 75As profile and (D) 78Se profile.

Figure 2.5 IPRP-ICPMS chromatograms of urine, (A) 75As profile and (B) 78Se profile, and ~50 ng ml-1 spiked urine (C) 75As profile and (D) 78Se profile.

Figure 3.1 The C. comosum accumulation of arsenic and selenium for the varying supplementation types administered during the cultivation process.

Figure 3.2 78Se and 75As SEC-ICPMS chromatograms of the root extract from the C. comosum after varying supplementation combinations.

Figure 3.3 78Se and 75As IPRP-ICPMS chromatograms of the root extract from the C. comosum after varying supplementation combinations.

Figure 3.4 78Se and 75As SEC-ICPMS chromatograms of the leaf extract from the C. comosum after varying supplementation combinations.

Figure 3.5 A summary of the metabolic pathway for the water soluble selenium and arsenic species after varying supplementation types in soil, rhizosphere, roots and leaves of the C. comosum (HMW= high molecular weight compounds, Seorg = organic selenium species).

-1 Figure 4.1 Graphs depicting the m/z 32 response after the addition of a 250 ng ml Na2SO4 standard or DDW as a blank with Xe (top) or He (bottom) CRC gas using a -16 V quadrupole bias and varying octopole bias voltages.

Figure 4.2 A graph overlying the 4He and 129Xe responses after percent stepwise addition of He (20 ml min-1) and Xe (1 ml min-1) CRC gas, respectively.

Figure 4.3 Xe gas flow rate optimization plots displaying sulfur signal (250 µg ml-1), blank (DDW) and BEC for (A) 32S, (B) 33S and (C) 34S with (D) 34S inset; shown using a quadrupole bias, -16 V and an octopole bias, -50 V for a net voltage difference of +34 V.

Figure 4.4 A contour plot of detection limits for 32S as a function of the alteration in quadrupole and octopole bias.

Figure 4.5 A contour plot showing the percentage of 52 elements exhibiting a detection limit within 3x of the lowest detection limit obtained per element as a function of the alteration in the octopole and quadrupole bias.

Figure 4.6 A bar graph depicting the 7 of 52 elements which possessed a difference in detection limits greater than 2.6x the lowest detection limit obtained per element when using Xe versus He or H2 as the CRC gas.

Figure 4.7 A bar graph displaying the elemental profile concentration found in bottled water from four major manufacturers represented as A-D (lowest to highest priced).

Figure 5.1 The A. fistulosum accumulation of selenium and mercury for the varying supplementation types administered during the cultivation process.

Figure 5.2 80Se and 202Hg SEC-ICPMS chromatograms (0.1 kDa to 7 kDa) of the leaf and root extract from the A. fistulosum after varying supplementation combinations.

Figure 5.3 80Se and 202Hg SEC-ICPMS chromatograms (10 kDa to 600 kDa) of the root extract from the A. fistulosum after varying supplementation combinations.

Figure 5.4 80Se and 202Hg SEC-ICPMS chromatograms (0.1 kDa to 7 kDa) of the root extract following enzymatic digestion from the A. fistulosum after varying supplementation combinations.

Figure 5.5 80Se and 202Hg SEC-ICPMS chromatograms (0.1 kDa to 7 kDa) of the root extract following acid digestion from the A. fistulosum after varying supplementation combinations.

Figure 5.6 78Se and 202Hg capRPLC-ICPMS chromatograms of the root extract from the A. fistulosum after varying supplementation combinations.

Figure 6.1 34S, 78Se and 202Hg capRPLC-ICPMS chromatograms of the root extract from the A. fistulosum after varying supplementation with HgII or SeIV & HgII.

Figure 6.2 X-ray fluorescence maps of a center section from an excised SeIV & HgII supplemented A. fistulosum root. This depicts the elemental distribution of selenium and mercury: the highlighted circles indicate certain areas of element co-localization.

Figure 6.3 X-ray fluorescence maps of the edge section from an excised SeIV & HgII supplemented A. fistulosum root depicting the elemental distribution of mercury and selenium.

Figure 6.4 A diagram representing the cross section of a monocot root.

Figure 6.5 X-ray absorption near edge structure spectra depicting the Se K-edge absorbance of selected spots of selenium and mercury co-localization from the SeIV & HgII supplemented A. fistulosum excised root and standards adapted from Gailer et al.

Figure 6.6 X-ray absorption near edge structure spectra depicting the Hg LIII-edge absorbance of selected spots (A) and (B) of selenium and mercury co-localization from the SeIV & HgII supplemented A. fistulosum excised root and standards adapted from Gailer et al.

Figure 7.1 The proposed metabolic pathways leading to the formation of the mercury-selenium and mercury-sulfur entities in the A. fistulosum root

Tables

Table 1.1 Contaminant remediation methodologies

Table 1.2 Summary of commonly used cell gases for ICPMS adapted from Koppenaal et al.

Table 2.1 Instrumental conditions

Table 2.2 The studied species in this work with their pKa values

Table 2.3 Chromatographic figures of merit

Table 3.1 Instrumental conditions in this study

Table 4.1 Pertinent ionization potentials and isotopic distributions

Chapter 1

Methodologies used for biological contaminant remediation and speciation

1.1 Contaminant remediation

Through varying sources including natural causes, i.e. volcanic eruption, global industrialization, agricultural practices, faulty waste disposal and purification methods, and a general increase in anthropogenic activities, the level of elemental contaminants in the environment has increased to toxic concentrations for living organisms in a multitude of worldwide locations. The growing recognition of the significant ecological hazards, including the impact on human health from these contaminants, led Congress to pass the Comprehensive

Environmental Response, Compensation, and Liability Act (CERCLA) also known as Superfund on December 11, 1980 with an initial trust fund of 1.6 million dollars.

When reevaluated in 1986, only six sites of the thousands of listed hazardous waste sites were removed from the national priority list and the significance of contaminant removal was realized. To combat the ineffectiveness of CERCLA through needed improvements, Congress passed the Superfund Amendments and reauthorization Act (SARA) in 1986. In summary, the limit of toxin removal was authorized to be consistent with long-term remedial action or long- term cleanup. Required cleanup actions were to meet state and federal laws to the extent where practical. Federal facilities were made applicable to CERCLA. The EPA was required to consider alternatives to sediment disposal and to treat wastes to the extent practicable. Specific deadlines were set for negotiating and settling remediation details with responsible parties.1

Since its inception in 1980, the Environmental Protection Agency (EPA) has placed a total of 47,281 sites in the Superfund data system. After expending billions and billions of dollars, 35,053 sites have been removed or archived leaving 12,228 active sites in need of remediation.1 Due to the variety in location and contamination type, varying methods for

9 remediation have been developed and implemented. Several of the prominent methodologies are listed in Table 1.1.

Table 1.1 Contaminant remediation methodologies

Ex situ In situ

Thermal treatment Sediment capping

Soil washing Electrokinetic separation Physical separation Chelation Phytoremediation

1.1.1 Ex situ methodologies

Ex situ methodologies provide remediation of environmental contaminants by transporting the polluted soil/sediment off-site, where treatment is administered, and returned once the hazard level is considered acceptable. Advantages include: fast remediation which is nearly independent of the site geology; parameters monitoring process optimization, efficiency and final results can be easily controlled; and extracted materials can be recycled (especially in the case of metals). Disadvantages include: a potentially high risk of contaminant exposure for workers and the environment during evacuation operations; expensive machinery is required to transport contaminated soil off-site to a fixed location; and remediation is often difficult to implement given various infrastructures (i.e. in an urban setting).2

A brief summary of the common ex situ methodologies is listed below. Thermal treatment of contaminated soils has been recently cited in the literature.3, 4 Thermal treatment techniques are generally classified as desorption or destruction according to their operational

10 temperatures. For desorption, the contaminated soil is heated from 150 °C to 500 °C. Pollutants are physically separated via a gas stream upon volatilization. For destruction, the contaminated soil is heated from 600 °C to 900 °C (occasionally greater temperatures are used). Pollutants are neutralized after undergoing a chemical modification.4 Microwave induction is shown to improve the conventional thermal treatment process.3

Soil washing involves the separation of contaminants from the soil matrix by solubilizing the contaminants in a washing solution, and has recently been successfully used for the treatment of soils.5, 6 Overall, the treatment process has been shown to be effective in remediating a wide range of organic and inorganic contaminants via chemical extraction and separation in addition to traditional physical processes. A major area for remediation optimization in soil washing utilizes the varying sized particles, mineral and organic sediments, comprising soil. In general, contaminants will reside and be concentrated in the fine particle fraction (smaller than a silt size of 74 µm) which is physically separated from the sand and oversize fractions.6 Additional extraction is accomplished through use of varying chelators depending on the contaminant species. Recent advances have attempted to combine soil washing with additional remediation methods in situ.5

1.1.2 In situ methodologies

In situ methodologies provide remediation of environmental contaminants by administering treatment on-site until the hazard level is considered acceptable. Advantages include: no soil/sediment evacuation is necessary, fixed facilities are not required, there is minimal exposure to workers, heavy industrial equipment is not required, the process is usually more cost effective, applications are generally appropriate for large sites, and contaminated sites

11 are often deemed useable for revegetation. Disadvantages include: the process efficiency is generally difficult to verify; site-specific conditions are influenced by the type, chemistry and structure of soils; contaminant remediation depth is limited; long term maintenance is often required; remediation is limited with low soil permeability; efficiency is influenced by concentration, fractionation and speciation of the contaminants; and the addition or injection of chemical reagents is often required, which may provide potential contamination to the aquifer.2

A brief summary of the common in situ methodologies is listed below. Sediment capping, additionally termed in situ capping, utilizes the placement of specified remedial material at the soil-water interface to prevent interaction between the benthic community (the ecological region at the lowest level of a body of water such as an ocean or a lake) and the surface water. The presence of sand, the traditional material employed for capping, removes the source of organic substrates, where biotransformations generally occur, by shifting the deposition of organic matter due to the inserted cap-water interface. Additionally, the natural flow of groundwater seepage has been shown in previous studies to be disrupted after implementation of some synthetic capping materials causing changes to the localized discharge zones.7

Electrokinetic remediation of contaminated soils has been recently cited in the literature.8-

10 Environmental contaminants are remediated through electrokinetic treatment after the insertion of two electrodes directly into the soil, or after the addition of an electrolyte solution that permeates the soil, and a low direct current is imposed.11 The combination of the soil, groundwater and electrodes are theoretically considered an electrochemical cell. In addition to the chemical alteration caused by redox reactions, the applied potential difference between the inserted electrodes causes metal electrodeposition at the cathode and the oxidation of organic

12 contaminants producing carbon dioxide and water at the anode. Electrokinetic remediation was shown in prior studies to be particularly useful over alternate techniques for contaminated soil with low permeability.9 Recent studies have investigated mechanisms during electrokinetic remediation with the addition of chelators such as ethylenediaminetetraacetic acid (EDTA).10

1.1.3 Phytoremediation

Phytoremediation is an in situ method which utilizes plants to clean up contaminated sites through a variety of mechanisms. It is generally regarded as a cost-effective and environmentally benign method, which is comparatively aesthetically pleasing relative to other remediation methods. A distinct disadvantage of using phytoremediation is the relatively long period of time that must be allotted for adequate soil remediation compared with some of the ex situ techniques previously mentioned. Effective plants possess the ability to accumulate a vast amount of contaminant and produce an abundant amount of biomass in a short period of time.

Previous studies have focused on natural hyperaccumulating plants, which can accumulate an abundant amount of contaminant in their aerial compartments such as leaves or shoots during normal growth and reproduction, for effective remediation. Additional studies have shown that hyperaccumulating plants often are slow growing, possess shallow root systems and have limited bioproductivity.12 Consequently, genetic modification is currently being investigated in a variety of plants in order to further improve the general effectiveness and utility of phytoremediation in the environment. Several recent reviews briefly summarizing current efforts utilizing gene alteration in plants for increased phytoremediation of metal and organic contaminants are given.13, 14

An improved general understanding of the mechanisms utilized by plants for the extraction, accumulation, transport, modification and volatilization of soil contaminants is currently needed to effectively target areas for genetic modification, thereby, increasing the potential applications for phytoremediation. The basic metabolic pathway utilized by plants for contaminant remediation is illustrated in Figure 1.1.

Figure 1.1 Potential metabolic pathways utilized during contaminant remediation in a plant.

The rhizosphere is defined as the dynamic microenvironment immediately surrounding the plant roots, where conditions may vary significantly compared to the surrounding bulk soil.15

In rhizosphere degradation, microbial enzymatic activity is utilized for the degradation of soil contaminants. The breakdown products can be volatilized or incorporated in the rhizosphere matrix for plant extraction. In addition, plants may release stabilizing or chelating compounds in

14 the rhizosphere aiding in contaminant accumulation.16 As distinguishing between microbial and plant contributions has been historically difficult, they generally comprise the soil-root interface.

As the type of plant and growth environment influence the amount, diversity and activity of microbial populations,17 targeting specific genes for modification in microbial populations is not trivial.

Phytoextraction involves the removal or induction of contaminants from the soil through the root epidermal tissues. For additional translocation to the aerial regions of the plant, the contaminants enter the xylem through a membrane pump or channel after passing the endodermal cells.17 Phytoextraction can be generally classified as continuous or induced.

Continuous phytoextraction utilizes plants that accumulate high concentrations of environmental contaminants throughout their life cycle (i.e. hyperaccumulators). Induced phytoextraction utilizes a single time point for enhanced contaminant accumulation via the addition of accelerants or chelators to the soil (EDTA is commonly used for increased metal uptake).18

Additional increase in effective phytoextraction of contaminants from the soil may be accomplished by proper pH adjustment and frequent monitoring.12

Phytoaccumulation involves the ability of the plant to stabilize and concentrate soil contaminants upon extraction. In the case of metals, transporters and binding proteins play an essential role in sequestration. Metallothioneins and phytochelatins are two classes of cysteine- rich peptides, which are capable of binding metals through a sulfur bond (R-SH). Plants contain metallothionein genes for encoding peptides that are generally composed of 60 to 80 amino acids and contain 9 to 16 cysteine residues.19 Metallothioneins have been shown in previous studies to function in the accumulation, transportation and sequestration of metals during normal metabolism.16 As the metallothionein-metal association can be glutathionated, it may be

15 suggested that long-term storage in the vacuole of a plant cell is probable.20 Phytochelatins are composed of the amino acids glutamate (Glu), cysteine (Cys) and glycine (Gly) in a polymer that can be described as (γ-Glu-Cys)n-Gly (n = 2 to 11). As a natural defense after metal exposure to plants, phytochelatins are synthesized via an enzymatic reaction catalyzed by phytochelatin synthase. The complexed metal can be translocated to the plant cell vacuole for storage.21 In addition to metallothioneins and phytochelatins, small organic acids have also been shown to play a role in metal detoxification in a plant.22 As the accumulated contaminant may enter the food chain through ingestion by animals causing increased toxicity, precautions must be taken when utilizing extensive genetic modification resulting in increased phytoaccumulation for remediation purposes.

In phytovolatilization, translocated contaminants to the plant leaves are volatilized to the atmosphere through the stomata. Extensive studies have provided increased contaminant removal through genetic modification. Other parameters, including water flow intake based on plant species and environmental conditions, also depict the efficiency of contaminant removal.17

In addition, monitoring contaminant relocation upon volatilization must be investigated to assess environmental hazards in the surrounding sites. Additional information where genetically engineered plants are utilized for improved phytovolatilization of metal contaminated sediments is sited.23

1.1.4 Selenium, arsenic and mercury remediation

While generally considered an essential element for animals, selenium has not been

2- proven essential for plants. The oxidized forms of selenium, selenate (SeO4 ) and selenite

2- (SeO3 ), commonly found in the environment are highly soluble and readily accumulated by

16 plants. Selenate is accumulated in a plant root through the sulfate transporter Sultr 1;2.24 The metabolism of selenium in a plant is a complex process and is briefly summarized in Figure 1.2.

Extensive efforts for phytoremediation have focused on genetic modification for increased accumulation and volatilization of selenium. The overexpression of ATP sulfurylase in Brassica juncea (Indian mustard) resulted in an increase of selenium accumulation.25 The increase in selenium volatilization has been shown after the overexpression of SeCys methyltransferase in B. juncea and Arabidopsis thaliana resulting in the formation of dimethyldiselenide (DMDSe).26, 27

In addition, complementary overexpressions of ATP sulfurylase and SeCys methyltransferase have been shown effective in B. juncea.28

Figure 1.2 A brief summary of the proposed selenium metabolic pathway in a plant, adapted from Terry et al.29

- 3- Arsenite (AsO2 ) and arsenate (AsO4 ) are the commonly found forms of arsenic in terrestrial plants and are considered phytotoxic. Arsenate has been shown to compete for

17 phosphate transport sites suggesting arsenic induction by the phosphate pathway. However, arsenite does not result in a decreased flux of phosphate depicting another mechanism is involved during arsenic accumulation.30 Arsenate is enzymatically reduced to arsenite by arsenate reductase or non-enzymatically by glutathione.31, 32 Once reduced, it is suggested that an arsenite phytochelatin complex forms allowing transportation to the vacuole.33 Efforts toward arsenic phytoremediation have generally utilized the high natural accumulation of the Pteris vittata (Chinese brake fern). The simultaneous overexpression of γ-glutamlcysteine synthetase and arsenate reductase caused a 5-fold increase in arsenic accumulation of the Arabidopsis suggesting potential benefits via genetic modification.34

Organomercurials and ionic mercury are toxic to plants and, to date, no mercury hyperaccumulating plants have been identified.17 The mechanism for phytoextraction of mercury from the soil is not well understood. As a result, phytoremediation studies have focused on volatilizing mercury as Hg0 through the cell surface after metabolism. By expressing the bacterial genes merA, which encodes mercuric reductase, and merB, which encodes organomercury lyase, in plants, increased mercury reduction and volatilization have been accomplished.35, 36 Genetically modified poplar and cottonwood trees via merA and/or merB expression show potential as phytoremediators of mercury, and do not require regular seasonal replanting.37, 38 Unfortunately, the ultimate oxidation fate, which could potentially lead to toxic oxidized forms, and the location of mercury are impossible to control with the proposed methodology.

Common techniques for selenium and mercury phytoremediation utilize volatilization as the predominant method for contaminant removal. While successfully diluting the concentration of the contaminant in a specified location as well as putative toxicity reduction, the overall

18 concentration is not reduced for the biological cycle of the contaminant. This dissertation sought a varying method of phytoremediation termed phytoretirement by the author. Phytoretirement involves the extraction and metabolism of a contaminant by a plant where the resulting product is biologically inactive. This effectively removes the contaminant from the biological cycle, thereby lowering the overall biologically active concentration. Chapters 6 and 7 describe a potential candidate utilizing the protective effect of selenium with mercury in conjunction within a plant.

1.2 Inductively coupled plasma mass spectrometry (ICPMS)

Inductively coupled plasma mass spectrometry is a relatively new technique with the first journal article published in 1980.39 The first commercial instrument was introduced just three years later and the technique is currently considered to be mature due to the incredibly rapid development academically and industrially.40 Inductively coupled plasma mass spectrometry is an element-specific detector, which can quantitatively monitor most elements in the periodic table virtually simultaneously using a quadrupole detector. Detection limits at sub- to low part per trillion levels have been achieved with a dynamic range up to nine orders of magnitude. Further information regarding experimental uses for ICPMS is found in Chapters 2 through 6.

Inductively coupled plasma mass spectrometry is currently utilized in a variety of fields and applications, some of which are highlighted below. An increasing number of clinical laboratories are utilizing ICPMS for the quantification of As, Cr, Cd, Hg and Pb in blood and/or urine; Al, Se, Cu and Zn in serum; and Pt, Au and other trace elements for therapeutic purposes.

In addition, separation techniques such as liquid chromatography or capillary electrophoresis are

19 coupled with ICPMS for speciation.41 Measurements of enriched stable isotope ratios for mineral and trace elements utilized as tracers in biological systems were traditionally accomplished using thermal ionization mass spectrometry or fast atom bombardment mass spectrometry, but are now secondary to analysis by ICPMS.42 Metal-based anticancer agents and general drug metabolite profiling are frequently monitored by ICPMS due to the high sensitivity, quantitative analysis and speciation capabilities of the element of interest.43, 44 While generally considered a metals detector, ICPMS has been shown to effectively monitor nonmetals such as phosphorus and sulfur with application to the quantification of proteins through a natural element tag, alleviating the need for stable isotope labels or addition of chemical tags.45, 46 Furthermore, discussion regarding theoretical and experimental applications of sulfur analysis by ICPMS are given in Chapters 4 and 6. Of more pertinence to this dissertation, ICPMS has also been shown effective in monitoring a variety of elements in a multitude of environmental matrices including several food applications.47, 48 As the field of ICPMS application continues to grow and expand the experimental capabilities in scientific research, this dissertation will predominantly focus on varying applications utilizing the ICPMS to solve scientific queries in the field of environmental contaminant monitoring and remediation.

While ICP has been shown generally effective for a host of previously mentioned sample types and areas of scientific interest, it should be mentioned that alternative sources such as chemical flames, arcs, sparks, microwave plasmas, glow discharges, electrospray sources, thermal-ionization sources, and novel atmospheric-pressure plasmas have been explored. In addition, alternatives to the commonly used quadrupole mass filters for elemental determinations such as Fourier-transform mass spectrometers, ion traps, time-of-flight mass spectrometers, and multi-channel detectors systems have been coupled to ICP for varying applications.49

1.2.1 Instrument setup and theory

A brief summary of the theoretical aspects of ICPMS will be explained from left to right in accordance with the general schematic in Figure 1.3. Further information regarding current updates on fundamental studies involving ICPMS can be found in several reviews.50, 51

Figure 1.3 A block diagram depicting the general schematic of an inductively coupled plasma mass spectrometer.

Several common methodologies are utilized for sample introduction to ICPMS and include laser ablation/sampling, flow injection analysis, electrothermal vaporization, desolvation systems and several chromatography separation techniques. Further reading regarding current studies in sample preparation and introduction are listed.51, 52 In the context of this dissertation, chromatographic separation techniques commonly used for elemental speciation are discussed further in section 1.3. Liquid samples are passed through a nebulizer, which is generally concentric in design and comprised of an internal capillary for liquid flow which is surrounded by a high velocity argon gas stream. The applied gas pressure forces the solvent out into aerosol droplets which are sorted in a cooled spray chamber by size differences, where the larger

21 droplets condense and are discarded as waste. The smaller droplets, 5 µm to 10 µm, are desolvated to particles in 1.0 ms to 1.5 ms, vaporized in less than 150 µs, followed by atomization and ionization to a positive charge on the order of µs in the plasma.53 For particles less than 1 µm in diameter, their complete vaporization takes place over a 1 mm to 2 mm spatial region.54 Common radio frequency generated argon plasma temperatures range between 6000 K to 10,000 K depending on the spatial region monitored.55

The formed ions then pass through the sampler and skimmer cones in the interface region, which is marked by a transition from atmospheric pressure to the vacuum region provided by a turbomolecular pump. Influencing electrodynamic forces between passing ions, termed the Debye length, which is typically 10-4 mm to 10-3 mm, are minimized with cone orifice diameters of 0.4 mm to 1.0 mm.56 The ions are then focused via ion optics and subsequently enter an octopole ion guide enclosed in a collision/reaction cell for interference removal, which is described further in section 1.2.2. Ions possessing sufficient kinetic energy enter the quadrupole for selective mass filtering, which is accomplished by placing a direct current on one pair of rods and a radio frequency on the opposite pair, resulting in an electric field where only the selected ion of interest possess a stable trajectory as the voltages are scanned. General parameters allow for scanning through a mass range of 2 Da to 260 Da and sampling times for selected masses are generally set to 0.1 s. Filtered ions are detected via an electron multiplier.

Further information can be found in the cited reference.57

1.2.2 Theory of interference removal with the collision/reaction cell

The detection capability of a traditional quadrupole mass analyzer was historically comprised by the formation of polyatomic spectral interferences generated by argon (plasma

22 gas), solvent selection, or unique sample-based polyatomic or isobaric species. Through modern instrumental advances, the advent of the collision/reaction cell considerably lowered the detection limits for many elements by enabling the discrimination of the afore-mentioned interferences in ICPMS.58 A general comparison of the more commonly known triple quadrupole mass filter and the commonly used octopole and quadrupole mass filter is shown in

Figure 1.4.

After selectively passing parent ions through the first quadrupole, inert gas is added in the second quadrupole and subsequently, the parent ions are dissociated forming characteristic daughter ions. The third quadrupole is used for selectively passing daughter ions to the detector.

While generally less sensitive, the combination of parent and daughter information allows for molecular identification. In contrast, prior to passing elemental parent ions to the detector in

ICPMS, the octopole, which is comparative to the second quadrupole in a triple quadrupole mass spectrometer, is used for discrimination of polyatomic or isobaric interferences.

Method development requires the selection of an appropriate cell gas which is generally a nontrivial process that is dependent on a number of parameters including the isotope of interest, unique sample matrix, elution solvent and formation of isobaric adducts. Common adduct formations interfering with a desired analyte include argides, oxides, hydrides, chlorides and bromides. A brief list of commonly used cell gases in ICPMS for a wide range of applications is found in Table 1.2.

Table 1.2 Summary of commonly used cell gases for ICPMS adapted from Koppenaal et al.59

Collision He, Ar, Ne, Xe

Charge exchange H2, NH3, Xe, CH4, N2

Oxidation reagent O2, N2O, NO, CO2

Reduction reagent H2, CO

Other reaction (adduction) CH4, C2H6, C2H4, CH3F, SF6, CH3OH

Varying mechanisms causing the removal of analyte interferences in the

32 collision/reaction cell have been elucidated. Three potential mechanisms for O2 interference removal allowing for the exclusive detection of 32S, which is experimentally addressed in

Chapter 4, are illustrated in Figure 1.5.

32 Figure 1.5 A block diagram representing the three potential mechanisms for O2 discrimination allowing for 32S monitoring.

Voltage discrimination utilizes the difference in spatial size of the analyte and polyatomic interference. The comparatively larger polyatomic interference statistically undergoes more collisions with the selected cell gas causing an overall reduction in its kinetic energy. By setting an energy barrier via a difference in octopole and quadrupole bias voltages, the analyte of interest, possessing sufficient kinetic energy, is selectively passed while the polyatomic ion with reduced kinetic energy cannot cross the energy barrier formed by the more positive charge on the quadrupole. Charge transfer may theoretically take place from an interfering ion possessing a higher ionization potential to a neutral cell gas atom possessing a lower ionization potential, thus removing the interference. Collision induced dissociation takes place when a collision between

25 an interfering ion and a neutral cell gas atom results in the sufficient transfer of kinetic energy to internal energy allowing for bond cleavage thus removing the interference. Although not utilized in this dissertation, reaction gasses can also be effectively used for elemental detection depending on the application. Further reading on specific applications of varying cell gases in addition to an extensive theoretical analysis of the collision/reaction cell can be found in the cited reference.60

1.3 High performance liquid chromatography (HPLC)

It is generally known that the toxicity, bioactivity, bioavailability, transportability in the organism, and the eventual fate and impact of an element environmentally and biologically in mankind is dictated by the particular forms or species of an element rather than its total concentration. Although ICPMS has gained popularity due to its ability to rapidly quantify multi-elements virtually simultaneously at trace levels, the basic instrumental design cannot discern an element’s oxidation state or the overall molecular weight of its parent molecule. High performance liquid chromatography has successfully been coupled with ICPMS enabling the desired speciation capabilities in varying sample matrices including environmental, bioanalytical, clinical, food and nutraceutical samples. Further information can be found in several comprehensive reviews.61-63

Coupling HPLC to ICPMS is generally a simple process and requires PEEK capillary tubing to connect the chromatography column exit to the ICPMS nebulizer entrance. A major consideration in sample introduction is the efficiency of nebulization at the selected eluent flow rate and the quantity of total dissolved solids. A major disadvantage when using ICPMS as the chromatographic detector is the low concentration of organic displacer that can be used. Ionized

26 organic solvents can cause plasma instability, potentially resulting in extinguishing the plasma and carbon build up, which alters the sampler and skimmer cone orifice diameter causing a significant loss in sensitivity.

Depending on the analyte of interest, a multitude of chromatographic separation techniques can be interfaced with ICPMS including normal phase, reversed phase, ion-pairing reversed phase, ion-exchange and size exclusion in addition to capillary electrophoresis. In the context of this dissertation, further information is given on the mechanisms utilized in size exclusion and ion-paring reversed phase chromatography as these methodologies are commonly utilized in Chapters 2, 3, 5 and 6.

1.3.1 Size exclusion chromatography (SEC)

Size exclusion chromatography, also known as gel permeation or gel filtration chromatography, is generally considered the primary method for determining molar mass averages and distribution patterns of polydisperse macromolecules and is utilized in Chapters 3 and 5 in conjunction with ICPMS. A general illustration of the separation mechanism utilized in

SEC can be found in Figure 1.6.

In a brief summary, the dissolved analytes varying in a distribution of hydrodynamic sizes are injected in a column which is packed with a porous, inert material. The analytes of smaller hydrodynamic size are able to permeate the stationary phase pores, while the analytes of larger hydrodynamic size are excluded. The effective path length of the smaller analytes is longer resulting in a later elution time when compared with the larger molecules. For a homologous series of linear macromolecules, a direct relationship correlating size with molar mass can be drawn. General consideration is given to the salt concentration of the buffer

(typically around 100 mmol l-1), which partially alleviates the interaction between charged substituents of the analytes and the stationary phase. An inverse correlation between the analyte size and its interactions with the stationary phase is generally observed. Further discussion on current advancements in SEC is given in a recent review.64

Figure 1.6 An illustration depicting the general mechanism in size exclusion chromatography resulting in the separation of analyte mixtures.

1.3.2 Ion-pairing reversed phase chromatography (IPRP)

Considering the array of previously mentioned complications after introduction of organic solvent to the ICPMS, conventional reversed phase chromatographic methods using an organic solvent to displace the analyte of interest cannot be used at conventional flow rates (0.5

28 ml min-1 to 1 ml min-1). As a result, ion-pairing reversed phase chromatography is hyphenated with ICPMS detection and typically used to separate non-volatile charged analytes, which vary in hydrophobicity. Further discussion and experimental incorporation of IPRP can be found in

Chapters 2, 3 and 4. An illustration of the varying retention mechanisms in conventional IPRP

(Figure 1.7) and after an alteration in eluent buffer (Figure 1.8) is depicted from a recent study.65

Figure 1.7 An illustration depicting the general mechanisms in ion-pairing reversed phase chromatography.

Figure 1.8 An illustration depicting the general mechanisms in ion-pairing reversed phase chromatography after an alteration of eluent buffer.

In a brief summary, dissolved analytes varying in charge and hydrophobicity are injected into a conventional C18 reversed phase column and eluted with a buffer containing an ion-pairing agent. Common ion-pairing agents for anions include alkylsulfonates, trialkylammonium salts and tetraalkylammonium salts. Separation occurs through several mechanisms including hydrophobic interactions between the analyte and stationary phase, and hydrophobic interactions between the neutral ion-pair (formed between the ion-pairing agent and the analyte of interest) and stationary phase. If the retention time of the neutral ion-pair is too extensive, the ion-pairing reagent, stationary phase selection, and/or buffer must be altered often with mediocre results. An alternative method is introduced in Chapter 2 where a secondary buffer of greater charge is

30 utilized resulting in the formation of a neutral ion-pair between the ion-pairing agent and the buffer, which releases the analyte of interest allowing for shorter elution times.

1.4 X-ray absorption fine structure (XAFS)

With the appearance of a number of synchrotron sources throughout the world in the last

15 years, x-ray absorption fine structure is currently used in a wide array of applications elucidating local atomic and electronic structures of absorbing analytes of interest in fields such as chemistry, biology, physics, material science and geophysics.66 Further reading from current review articles are listed with topics of research including peptide folding, metal-binding mechanisms and metal-binding site structures;67 thin organic films and liquids;68 and molten salts and ionic liquids.69 X-ray absorption fine structure analysis is utilized in Chapter 6 to provide pertinent metabolic information regarding selenium and mercury metabolism in the Allium fistulosum (green onion).

1.4.1 Instrument setup and operation

A brief summary of the theoretical aspects of XAFS will be explained from left to right in accordance with the general schematic in Figure 1.9. The synchrotron ring accelerates electrons in a circular path producing a polychromatic x-ray beam tangential to the ring. The resulting x- ray beam passes through a double crystal monochromator where a computer controlled incident angle allows for stepwise tuning through the energy range of interest according to the Bragg condition (nλ = 2d sin θ). The monochromatic x-rays then pass through the sample. If the analyte of interest is highly concentrated, the difference between the signal acquired from the incident flux monitor and the transmitted flux monitor detected with gas ionization chambers can

31 be used. For analyte concentrations in the high part per million range, a fluorescence photon- counting detector is used.

Figure 1.9 A block diagram depicting the general schematic of x-ray absorption fine structure analysis.

1.4.2 X-ray absorption fine structure theory

In XAFS, an x-ray is absorbed by a core-level electron and propagates away from the atom into the continuum, which is termed the photoelectric effect and is depicted in Figure 1.10.

During a scan, generally increasing in x-ray energy, no absorption occurs until the incident x-ray possess energy equal to the ionization energy of 1s electrons for K edge monitoring or 2s or 2p electrons for L edge monitoring. These core electrons are then excited into the continuum.

Subsequently, an electron in the excited state decays and emits a characteristic energy wave which can be element specific. Synchrotrons generally provide a spectral distribution between 2 keV to 25 keV resulting in excitation of K shell electrons through the second transition series and excitation of L shell electrons for the remainder of the periodic table. As a result, all the elements starting from sulfur through the rest of the periodic table can be sampled.

When neighboring atoms are present, the excited photo-electron can scatter due to neighboring electrons and return to its original absorbing atom. The identity and location of these neighboring atoms can alter the absorption coefficient due to back scattering resulting in constructive or deconstructive interference as shown in Figure 1.11. These absorption modulations can be mathematically fit using data analysis software to elucidate the information below. Further information regarding fundamental theoretical aspects of XAFS may be found in the cited review articles.66, 70

A typical XAFS spectrum can be divided into two energy regions: x-ray absorption near edge structure (XANES) and extended x-ray absorption fine structure (EXAFS). Analysis in the

XANES region can be essentially thought of as a spectroscopic method where the oxidation state and site symmetry may be elucidated. Analysis in the EXAFS region can provide structural information of the atoms surrounding the analyte of interest including bond length, coordination

33 numbers and ligand identity. Considering the photon beam penetrates through the entire sample from the photoelectric effect, a projection showing all the locations of the analyte of interest are given. If the analyte environment is heterogeneous, spatial resolution may be lost and extensive data analysis is required for proper assignment of the analytes oxidation state and immediate coordinating environment.

Figure 1.11 A diagram depicting an x-ray absorption pattern through the photoelectric process.

1.5 References

1. FY 2007 superfund annual report; Environmental Protection Agency: 2008; p 33. 2. Dermont, G.; Bergeron, M.; Mercier, G.; Richer-Lafleche, M., Metal-Contaminated Soils: Remediation Practices and Treatment Technologies. Practice Periodical of Hazardous, Toxic, and Radioactive Waste Management 2008, 12, (3), 188-209. 3. Li, D.; Quan, X.; Zhang, Y.; Zhao, Y., Microwave-Induced Thermal Treatment of Petroleum Hydrocarbon-Contaminated Soil. Soil & Sediment Contamination 2008, 17, (5), 486- 496. 4. Merino, J.; Bucala, V., Effect of temperature on the release of hexadecane from soil by thermal treatment. Journal of Hazardous Materials 2007, 143, (1-2), 455-461. 5. Guo, Z.; Wei, Z.; Wu, Q.-T.; Long, X., Chelator-Enhanced Phytoextraction Coupling with Soil Washing to Remediate Multiple-Metals-Contaminated Soils. Practice Periodical of Hazardous, Toxic, and Radioactive Waste Management 2008, 12, (3), 210-215. 6. Ko, I.; Lee, C.-H.; Lee, K.-P.; Lee, S.-W.; Kim, K.-W., Remediation of soil contaminated with arsenic, zinc, and nickel by pilot-scale soil washing. Environmental Progress 2006, 25, (1), 39-48. 7. Himmelheber, D. W.; Pennell, K. D.; Hughes, J. B., Natural Attenuation Processes during In Situ Capping. Environmental Science & Technology 2007, 41, (15), 5306-5313. 8. Ricart, M. T.; Pazos, M.; Gouveia, S.; Cameselle, C.; Sanroman, M. A., Removal of organic pollutants and heavy metals in soils by electrokinetic remediation. Journal of Environmental Science and Health, Part A: Toxic/Hazardous Substances & Environmental Engineering 2008, 43, (8), 871-875. 9. Rutigliano, L.; Fino, D.; Saracco, G.; Specchia, V.; Spinelli, P., Electrokinetic remediation of soils contaminated with heavy metals. Journal of Applied Electrochemistry 2008, 38, (7), 1035-1041. 10. Yuan, C.; Chiang, T.-S., Enhancement of electrokinetic remediation of arsenic spiked soil by chemical reagents. Journal of Hazardous Materials 2008, 152, (1), 309-315. 11. Reddy, K. R.; Chinthamreddy, S., Electrokinetic remediation of heavy metal- contaminated soils under reducing environments. Waste Management (Oxford) 1999, 19, (4), 269-282. 12. Gratao, P. L.; Prasad, M. N. V.; Cardoso, P. F.; Lea, P. J.; Azevedo, R. A., Phytoremediation: Green technology for the clean up of toxic metals in the environment. Brazilian Journal of Plant Physiology 2005, 17, (1), 53-64. 13. Macek, T.; Kotrba, P.; Svatos, A.; Novakova, M.; Demnerova, K.; Mackova, M., Novel roles for genetically modified plants in environmental protection. Trends in Biotechnology 2008, 26, (3), 146-152. 14. Van Aken, B., Transgenic plants for phytoremediation: helping nature to clean up environmental pollution. Trends in Biotechnology 2008, 26, (5), 225-227. 15. Wenzel, W. W.; Lombi, E.; Adriano, D. C., Biogeochemical processes in the rhizosphere: role in phytoremediation of metal-polluted soils. Heavy Metal Stress in Plants 1999, 273-303. 16. Eapen, S.; D'Souza, S. F., Prospects of genetic engineering of plants for phytoremediation of toxic metals. Biotechnology Advances 2005, 23, (2), 97-114. 17. Peer, W. A.; Baxter, I. R.; Richards, E. L.; Freeman, J. L.; Murphy, A. S., Phytoremediation and hyperaccumulator plants. Topics in Current Genetics 2006, 14, (Molecular Biology of Metal Homeostasis and Detoxification), 299-340.

18. Salt, D. E.; Smith, R. D.; Raskin, I., Phytoremediation. Annual Review of Plant Physiology and Plant Molecular Biology 1998, 49, 643-668. 19. Zhou, J.; Goldsbrough, P. B., Structure, organization and expression of the metallothionein gene family in Arabidopsis. Molecular & General Genetics 1995, 248, (3), 318- 28. 20. Brouwer, M.; Hoexum-Brouwer, T.; Cashon, R. E., A putative glutathione-binding site in cadmium-zinc-metallothionein identified by equilibrium binding and molecular-modeling studies. Biochemical Journal 1993, 294, (1), 219-25. 21. Cobbett, C.; Goldsbrough, P., Phytochelatins and metallothioneins: roles in heavy metal detoxification and homeostasis. Annual Review of Plant Biology 2002, 53, 159-182. 22. Kramer, U.; Pickering, I. J.; Prince, R. C.; Raskin, I.; Salt, D. E., Subcellular localization and speciation of nickel in hyperaccumulator and non-accumulator Thlaspi species. Plant Physiol FIELD Full Journal Title:Plant physiology 2000, 122, (4), 1343-53. 23. Zhang, R.-Q.; Tang, C.-F.; Wen, S.-Z.; Liu, Y.-G.; Li, K.-L., Advances in research on genetically engineered plants for metal resistance. Journal of Integrative Plant Biology 2006, 48, (11), 1257-1265. 24. Shibagaki, N.; Rose, A.; McDermott, J. P.; Fujiwara, T.; Hayashi, H.; Yoneyama, T.; Davies, J. P., Selenate-resistant mutants of Arabidopsis thaliana identify Sultr1;2, a sulfate transporter required for efficient transport of sulfate into roots. Plant Journal 2002, 29, (4), 475- 486. 25. van Huysen, T.; Terry, N.; Pilon-Smits, E. A. H., Exploring the selenium phytoremediation potential of transgenic Indian mustard overexpressing ATP sulfurylase or cystathionine-g-synthase. International Journal of Phytoremediation 2004, 6, (2), 111-118. 26. Ellis, D. R.; Sors, T. G.; Brunk, D. G.; Albrecht, C.; Orser, C.; Lahner, B.; Wood, K. V.; Harris, H. H.; Pickering, I. J.; Salt, D. E., Production of Se-methylselenocysteine in transgenic plants expressing selenocysteine methyltransferase. BMC Plant Biology 2004, 4, No pp given. 27. LeDuc, D. L.; Tarun, A. S.; Montes-Bayon, M.; Meija, J.; Malit, M. F.; Wu, C. P.; AbdelSamie, M.; Chiang, C.-Y.; Tagmount, A.; DeSouza, M.; Neuhierl, B.; Boeck, A.; Caruso, J.; Terry, N., Overexpression of selenocysteine methyltransferase in Arabidopsis and Indian mustard increases selenium tolerance and accumulation. Plant Physiology 2004, 135, (1), 377- 383. 28. Kubachka, K. M.; Meija, J.; LeDuc, D. L.; Terry, N.; Caruso, J. A., Selenium Volatiles as Proxy to the Metabolic Pathways of Selenium in Genetically Modified Brassica juncea. Environmental Science & Technology 2007, 41, (6), 1863-1869. 29. Terry, N.; Zayed, A. M.; De Souza, M. P.; Tarun, A. S., Selenium in higher plants. Annual Review of Plant Physiology and Plant Molecular Biology 2000, 51, 401-432. 30. Wang, J.; Zhao, F.-J.; Meharg, A. A.; Raab, A.; Feldmann, J.; McGrath, S. P., Mechanisms of arsenic hyperaccumulation in Pteris vittata. Uptake kinetics, interactions with phosphate, and arsenic speciation. Plant Physiology 2002, 130, (3), 1552-1561. 31. Delnomdedieu, M.; Basti, M. M.; Otvos, J. D.; Thomas, D. J., Reduction and binding of arsenate and dimethylarsinate by glutathione: a magnetic resonance study. Chemico-Biological Interactions 1994, 90, (2), 139-55. 32. Mukhopadhyay, R.; Shi, J.; Rosen, B. P., Purification and characterization of Acr2p, the Saccharomyces cerevisiae arsenate reductase. Journal of Biological Chemistry 2000, 275, (28), 21149-21157.

33. Meharg, A. A.; Hartley-Whitaker, J., Tansley review no. 133. Arsenic uptake and metabolism in arsenic resistant and nonresistant plant species. New Phytologist 2002, 154, (1), 29-43. 34. Dhankher, O. P.; Li, Y.; Rosen, B. P.; Shi, J.; Salt, D.; Senecoff, J. F.; Sashti, N. A.; Meagher, R. B., Engineering tolerance and hyperaccumulation of arsenic in plants by combining arsenate reductase and g-glutamylcysteine synthetase expression. Nature Biotechnology 2002, 20, (11), 1140-1145. 35. Bizily, S. P.; Rugh, C. L.; Summers, A. O.; Meagher, R. B., Phytoremediation of methylmercury pollution: merB expression in Arabidopsis thaliana confers resistance to organomercurials. Proceedings of the National Academy of Sciences of the United States of America 1999, 96, (12), 6808-6813. 36. Rugh, C. L.; Wilde, H. D.; Stack, N. M.; Thompson, D. M.; Summers, A. O.; Meagher, R. B., Mercuric ion reduction and resistance in transgenic Arabidopsis thaliana plants expressing a modified bacterial merA gene. Proceedings of the National Academy of Sciences of the United States of America 1996, 93, (8), 3182-7. 37. Che, D.; Meagher, R. B.; Heaton, A. C. P.; Lima, A.; Rugh, C. L.; Merkle, S. A., Expression of mercuric ion reductase in Eastern cottonwood (Populus deltoides) confers mercuric ion reduction and resistance. Plant Biotechnology Journal 2003, 1, (4), 311-319. 38. Rugh, C. L.; Senecoff, J. F.; Meagher, R. B.; Merkle, S. A., Development of transgenic yellow poplar for mercury phytoremediation. Nature Biotechnology 1998, 16, (10), 925-928. 39. Houk, R. S.; Fassel, V. A.; Flesch, G. D.; Svec, H. J.; Gray, A. L.; Taylor, C. E., Inductively coupled argon plasma as an ion source for mass spectrometric determination of trace elements. Analytical Chemistry 1980, 52, (14), 2283-9. 40. Potter, D., A commercial perspective on the growth and development of the quadrupole ICP-MS market. Journal of Analytical Atomic Spectrometry 2008, 23, (5), 690-693. 41. Parsons, P. J.; Barbosa, F., Jr., Atomic spectrometry and trends in clinical laboratory medicine. Spectrochimica Acta, Part B: Atomic Spectroscopy 2007, 62B, (9), 992-1003. 42. Stuerup, S.; Hansen, H. R.; Gammelgaard, B., Application of enriched stable isotopes as tracers in biological systems: a critical review. Analytical and Bioanalytical Chemistry 2008, 390, (2), 541-554. 43. Brouwers, E. E. M.; Tibben, M.; Rosing, H.; Schellens, J. H. M.; Beijnen, J. H., The application of inductively coupled plasma mass spectrometry in clinical pharmacological oncology research. Mass Spectrometry Reviews 2008, 27, (2), 67-100. 44. Gammelgaard, B.; Packert Jensen, B., Application of inductively coupled plasma mass spectrometry in drug metabolism studies. Journal of Analytical Atomic Spectrometry 2007, 22, (3), 235-249. 45. Prange, A.; Proefrock, D., Chemical labels and natural element tags for the quantitative analysis of bio-molecules. Journal of Analytical Atomic Spectrometry 2008, 23, (4), 432-459. 46. Rappel, C.; Schaumloeffel, D., The role of sulfur and sulfur isotope dilution analysis in quantitative protein analysis. Analytical and Bioanalytical Chemistry 2008, 390, (2), 605-615. 47. Butler, O. T.; Cook, J. M.; Harrington, C. F.; Hill, S. J.; Rieuwerts, J.; Miles, D. L., Atomic spectrometry update. Environmental analysis. Journal of Analytical Atomic Spectrometry 2008, 23, (2), 249-286. 48. Taylor, A.; Branch, S.; Day, M. P.; Patriarca, M.; White, M., Atomic spectrometry update. Clinical and biological materials, foods and beverages. Journal of Analytical Atomic Spectrometry 2008, 23, (4), 595-646.

49. Hieftje, G. M., Emergence and impact of alternative sources and mass analyzers in plasma source mass spectrometry. Journal of Analytical Atomic Spectrometry 2008, 23, (5), 661- 672. 50. Bacon, J. R.; Linge, K. L.; Parrish, R. R.; Van Vaeck, L., Atomic spectrometry update. Atomic mass spectrometry. Journal of Analytical Atomic Spectrometry 2008, 23, (8), 1130-1162. 51. Beauchemin, D., Inductively Coupled Plasma Mass Spectrometry. Analytical Chemistry (Washington, DC, United States) 2008, 80, (12), 4455-4486. 52. Beauchemin, D., Inductively Coupled Plasma Mass Spectrometry. Analytical Chemistry 2004, 76, (12), 3395-3416. 53. Olesik, J. W., Investigating the fate of individual sample droplets in inductively coupled plasmas. Applied Spectroscopy 1997, 51, (5), 158A-175A. 54. Horner, J. A.; Chan, G. C. Y.; Lehn, S. A.; Hieftje, G. M., Computerized simulation of solute-particle vaporization in an inductively coupled plasma. Spectrochimica Acta, Part B: Atomic Spectroscopy 2008, 63B, (2), 217-233. 55. Thomas, R., A beginner's guide to ICP-MS Part III: The plasma source. Spectroscopy (Duluth, MN, United States) 2001, 16, (6), 26,28-30. 56. Thomas, R., A beginner's guide to ICP-MS Part IV: The interface region. Spectroscopy (Duluth, MN, United States) 2001, 16, (7), 26-28,34. 57. Thomas, R., A beginner's guide to ICP-MS: Part V: The ion focusing system. Spectroscopy (Duluth, MN, United States) 2001, 16, (9), 38, 40, 42, 44. 58. Thomas, R., A beginner's guide to ICP-MS part IX - mass analyzers: collision/reaction cell technology. Spectroscopy (Duluth, MN, United States) 2002, 17, (2), 42-44, 46, 48. 59. Koppenaal, D. W.; Eiden, G. C.; Barinaga, C. J., Collision and reaction cells in atomic mass spectrometry: development, status, and applications. Journal of Analytical Atomic Spectrometry 2004, 19, (5), 561-570. 60. Tanner, S. D.; Baranov, V. I.; Bandura, D. R., Reaction cells and collision cells for ICP- MS: a tutorial review. Spectrochimica Acta, Part B: Atomic Spectroscopy 2002, 57B, (9), 1361- 1452. 61. Michalke, B., The coupling of LC to ICP-MS in element speciation: I. General aspects. TrAC, Trends in Analytical Chemistry 2002, 21, (2), 142-153. 62. Neubauer, K., Innovations in speciation analysis using HPLC with ICP-MS detection. Spectroscopy (Duluth, MN, United States) 2008, 23, (4), 22, 24-31. 63. Wang, T., Liquid chromatography-inductively coupled plasma mass spectrometry (LC- ICP-MS). Journal of Liquid Chromatography & Related Technologies 2007, 30, (5-7), 807-831. 64. Striegel, A. M., Size-exclusion chromatography: smaller, faster, multi-detection, and multi-dimensions. Analytical and Bioanalytical Chemistry 2008, 390, (1), 303-305. 65. Afton, S.; Kubachka, K.; Catron, B.; Caruso, J. A., Simultaneous characterization of selenium and arsenic analytes via ion-pairing reversed phase chromatography with inductively coupled plasma and electrospray ionization ion trap mass spectrometry for detection. Journal of Chromatography, A 2008, 1208, (1-2), 156-163. 66. Aksenov, V. L.; Koval'chuk, M. V.; Kuz'min, A. Y.; Purans, Y.; Tyutyunnikov, S. I., Development of methods of EXAFS spectroscopy on synchrotron radiation beams: Review. Crystallography Reports 2006, 51, (6), 908-935. 67. Duncan, K. E. R.; Ngu, T. T.; Chan, J.; Salgado, M. T.; Merrifield, M. E.; Stillman, M. J., Peptide folding, metal-binding mechanisms, and binding site structures in metallothioneins. Experimental Biology and Medicine (Maywood, NJ, United States) 2006, 231, (9), 1488-1499.

68. Hahner, G., Near edge X-ray absorption fine structure spectroscopy as a tool to probe electronic and structural properties of thin organic films and liquids. Chemical Society Reviews 2006, 35, (12), 1244-1255. 69. Hardacre, C., Application of EXAFS to molten salts and ionic liquid technology. Annual Review of Materials Research 2005, 35, 29-49. 70. Koningsberger, D. C.; Mojet, B. I.; Van Dorssen, G. E.; Ramaker, D. E., XAFS spectroscopy; fundamental principles and data analysis. Topics in Catalysis 2000, 10, (3,4), 143- 155.

Chapter 2

Simultaneous characterization of selenium and arsenic analytes via ion-pairing reversed phase chromatography with inductively coupled plasma and electrospray ionization ion trap mass spectrometry for detection; applications of river water, plant extract and urine matrices

2.1 Abstract

With an increased awareness and concern for varying toxicities of the different chemical forms of environmental contaminants such as selenium and arsenic, effective methodologies for speciation are paramount. In general, chromatographic methodologies have been developed using a particular detection system and a unique matrix for single element speciation. In this study, a routine method to speciate selenium and arsenic in a variety of “real world” matrices with elemental and molecular mass spectrometric detection has been successfully accomplished.

Specifically, four selenium species, selenite, selenate, selenomethionine and selenocystine, and four arsenic species, arsenite, arsenate, monomethlyarsonate and dimethylarsinate, were simultaneously separated using ion-pairing reversed phase chromatography coupled with inductively coupled plasma and electrospray ionization ion trap mass spectrometry. Using tetrabutylammonium hydroxide as the ion-pairing reagent on a C18 column, the separation and re-equilibration time was attained within 18 minutes. To illustrate the wide range of possible applications, the method was then successfully applied for the detection of selenium and arsenic species found naturally and spiked in river water, plant extract and urine matrices.

2.2 Introduction

Selenium, an element essential for human survival, has been shown to have a narrow range between deficiency and toxicity.1 Needed at only trace levels in the human body, selenium provides many pertinent biological properties including evidence for cancer chemoprevention2-4 and protection against oxidative damage through enzymes such as glutathione peroxidase.5 The

IV common forms of selenium in the environment are selenite (Se ), selenocystine (SeCys2), selenomethionine (SeMet) and selenate (SeVI),6 ordered in decreasing toxicity.7 When considering the range of toxicities from the different chemical forms of selenium and the growing concern drawn from high levels found in various contaminated environments,8 selenium speciation is vital for proper analysis and assessment in the environment and biological systems.

While there is possible evidence for the essentiality of arsenic at trace levels,9 its toxic effects are well known. Anthropogenic activities, such as wood treatment and mining/processing of ore,10, 11 have increased arsenic levels beyond natural concentrations thereby causing worldwide concern for the environment. The most common arsenic species include arsenite

(AsIII), arsenate (AsV), monomethlyarsonate (MMA) and dimethylarsinate (DMA), which are listed with decreasing toxicity.12 Due to the significantly different levels in toxicity of arsenic forms, speciation of arsenic is paramount to effectively assess the risk of a given environmental or biological sample.

In general, speciation methods have been developed for varying forms of single elements.

With the knowledge of accumulating environmental contamination from different sources, a need for routine multi-element speciation methods is apparent in order to meet the impending financial and time requirements of today. In the past, several separation techniques have been developed to speciate different selenium and arsenic forms in various matrices with some of the

42 more recent given here.13, 14 In addition to the separation of arsenic species, which has been typically accomplished using anion exchange chromatography,15-17 simultaneous speciation of selenium and arsenic species has also been shown, but in far fewer publications.18-20 With these methods, only the inorganic selenium species were resolved along with varying arsenic species.

This study addresses this shortcoming by including both organic and inorganic species of these metalloids in one chromatographic run.

Ion-pairing reversed phase liquid chromatography (IPRP) has been utilized for organic selenium speciation in prior studies.21-23 In addition, a few studies have incorporated IPRP for the simultaneous separation of selenium and arsenic species. Le et al. separated 13 arsenic and selenium species using hexanesulfonate as the ion-pairing reagent;5 however, when the method was tested with a urine matrix, a different column and mobile phase were needed along with elevated temperatures to achieve the desired separation. Do et al. separated 10 arsenic and selenium species employing tetrabutylammonium phosphate as the ion-pairing reagent;24 although, online compatibility with inductively coupled plasma mass spectrometry was not investigated. More recently, Pan et al. separated 12 selenium and arsenic species using tetrabutylammonium hydroxide as the ion-pairing reagent;25 however, SeVI, a major species in nature, had an elution time of approximately 40 minutes. Currently, there remains a need for the simultaneous separation of the eight predominant selenium and arsenic species: AsIII, AsV,

IV VI MMA, DMA, Se , Se , SeMet and SeCys2, in a timely and sensitive manner with online detection.

Several instruments have been used for the detection of selenium and arsenic including atomic fluorescence spectrometry,26, 27 atomic absorption spectrometry28, 29 and inductively coupled plasma mass spectrometry30, 31 (ICPMS). Of the frequently used techniques listed,

ICPMS has proven to be one of the most sensitive analytical techniques for fast multi-element determination of metals at ultra trace concentrations in different sample matrices, and is amenable to coupling with various chromatographic methodologies.32 A key reason for the enhanced speciation capability with ICPMS was the advent of the collision/reaction cell, which gave rise to the removal of polyatomic interferences, through collision, reaction and energy discrimination, from the analyte of interest (i.e. 40Ar40Ar for 80Se and 40Ar35Cl for 75As). This allowed for a greater signal to noise ratio and improved detection capabilities.

Newer speciation methodologies should include molecular mass spectrometry techniques in order to provide specific molecular weight information. Electrospray ionization ion trap mass spectrometry (ESI-ITMS) has been used as an advantageous companion to the element specific detection of ICPMS.33, 34 While ESI-ITMS suffers from high detection limits when compared to

ICPMS, ESI-ITMS can provide structural information to confirm unknown compounds by mass

(MSn) as a compliment to retention time matching and sample spiking assignments of ICPMS.

To date, the chromatographic methods mentioned above have not investigated the feasibility of

ESI-ITMS in conjunction with the ICPMS for species identification.

In past studies, chromatographic methodologies for the speciation of varying arsenic and selenium analytes generally have been limited to a single application in “real world” matrices.

The differences in composition of different matrices often cause insurmountable challenges in developing a universal separation; however, the ability of a speciation method to successfully separate the analytes of interest within various sample matrices is imperative to the overall achievement of a robust and valuable separation. Comprehensive studies reviewing arsenic and selenium separations listed water, plant and urine samples as common environmental and biological matrices for method applications.13, 14 Currently, a single method for the simultaneous

44 separation of common selenium and arsenic analytes within these matrices has not yet been shown.

In this study, a method was developed to accommodate the growing demands of multi- elemental speciation. Eight common environmentally and biologically observed arsenic and selenium standards were baseline separated including re-equilibration time within 18 minutes on a C18 chromatography column via ion-pairing reversed phase chromatography. Both ICPMS and

ESI-ITMS were incorporated as detectors without modifying the separation methodology, which lead to retention time matching and confident structural identification. The chromatographic method was then successfully applied to three environmental and biological matrices: river water, plant extract and urine to illustrate the wide range of potential applications.

2.3 Experimental

2.3.1 Instrumentation

High-performance liquid chromatography: Chromatographic separations were accomplished with an Agilent 1100 liquid chromatograph by Agilent Technologies (Santa Clara,

CA) equipped with a binary HPLC pump, an autosampler, a vacuum de-gasser system and a thermostated column compartment. Reversed phase chromatography was carried out with a

ZORBAX Eclipse XDB-C18 column (5 μm x 4.6 mm id x 250 mm) from Agilent Technologies

(Santa Clara, CA).

Inductively coupled plasma mass spectrometry: The ICPMS used for specific element detection was an Agilent 7500ce by Agilent Technologies (Santa Clara, CA). The instrument was equipped with a microconcentric nebulizer made by Glass Expansion (Pocasset, MA), a

Scott double channel spray chamber (2°C), a shielded torch, a CE lens stack, an octopole

45 collision/reaction cell with hydrogen gas pressurization (purity of 99.999%), a quadrupole mass analyzer and an electron multiplier for detection.

Electrospray ionization ion trap mass spectrometry: The ESI-ITMS used for molecular identification was a 6300 Agilent LC/MSD Trap XCT Ultra from Agilent

Technologies (Santa Clara, CA) used in the conventional LC introduction mode. The effluent from the column was sent directly, via a 40 cm length 0.25 mm ID PEEK tube from the column outlet, to the ESI interface with a high-flow spacer installed. The ESI-ITMS spectra were acquired in negative ion mode. For tuning purposes, individual arsenic and selenium standards diluted in buffer A were introduced via a syringe pump obtained from KD Scientific (Holliston,

MA) at a rate of 10 μl min-1.

For all instrumental conditions, see Table 2.1.

2.3.2 Reagents and standards

All the solutions were prepared in 18 MΩ cm-1 doubly deionized water (DDW) processed by Sybron/Barnstead (Boston, MA). Standards used for sample spiking and identification were the following: disodium methyl arsonate hexahydrate (MMA) purchased from Chem Service

(West Chester, PA); L(+)-selenomethionine (SeMet) was selected as it is the form found within biological samples such as plants35 and obtained from Acros Organics (Morris Plains, NJ);

III sodium (meta)arsenite (As ), Cacodylic acid (DMA) and seleno-L-cystine (SeCys2) were acquired from Fluka (Milwaukee, WI); potassium arsenate (AsV), potassium selenate (SeVI) and sodium selenite (SeIV) were purchased from Sigma-Aldrich (St. Louis, MO).

Table 2.1 Instrumental conditions ICPMS Forward power 1500 W Plasma gas flow 15.0 l min-1 Carrier gas flow 0.96 l min-1 Makeup gas flow 0.14 l min-1 -1 Collision gas 3.5 ml min H2 Quadrupole bias -16.0 V Octopole bias -18.0 V Monitored isotopes 75As, 77Se, 78Se, 80Se, 82Se Dwell time 100 ms per isotope ESI-ITMS Drying Temperature 350 °C Nebulizer gas 70.0 psi He -1 Drying gas 12.0 l min N2 Capillary voltage 3200 V Accumulation time 194 μs Scan range 50-400 m z-1 IPRP-HPLC Mobile phase A 5 mmol l-1 TBAH in 2.5 mmol-1

(NH4)3PO4 (pH 6.0) -1 Mobile phase B 10 mmol l (NH4)2SO4 (pH 6.0) Flow rate 1.0 ml min-1 Injection volume 100 μl Gradient program Time (min) 0 0.5 1.5 5 6 18 % A 100 100 0 0 100 100 % B 0 0 100 100 0 0

Claritas PPT arsenic and selenium elemental standards used for quantification were acquired from SpexCertiPrep (Metuchen, NJ). Calibration standards of 1.0 µg l-1 - 500 µg l-1 were prepared through dilution from a stock solution with 5 % (v/v) HNO3.

The mobile phase for general plant biomass extraction, was made by dissolving tris(hydroxymethyl) aminomethane hydrochloride (Tris-HCl) obtained from Fisher Scientific

(Fair Lawn, NJ) in DDW and adjusted the pH with hydrochloric acid. For reversed phase chromatography, mobile phase A contained 5 mmol l-1 tetrabutylammonium hydroxide (TBAH)

Fluka (Milwaukee, WI) and 2.5 mmol l-1 ammonium phosphate Sigma-Aldrich Co. (St.Louis,

MO) at a pH of 6.0. Mobile phase B contained 10 mmol l-1 ammonium sulfate purchased from

Sigma-Aldrich Co. (St. Louis, MO) at a pH of 6.0. The pH was adjusted with phosphoric acid for mobile phase A and ammonium hydroxide for mobile phase B. A summary of the mobile phase conditions is depicted in Table 2.1. All samples were filtered through 0.2 μm membrane syringe filters by Econofilters obtained from Agilent Technologies, Inc. (Santa Clara, CA) before being injected onto the HPLC-ICPMS or HPLC-ESI-ITMS.

2.3.3 Sample acquisition

River water: The river water sample was collected in polypropylene tubes from the Ohio

River. After homogenization via vortexing, 1 ml of solution was centrifuged at 5000 rpm for 15 minutes. The supernatant was decanted and filtered through a 0.2 µm syringe filter to remove particulates. Of the resulting solution, 100 µl were injected onto the IPRP-ICPMS.

Plant extract: The Alluim fistulosum (green onion) plants were purchased from a local supermarket. Subsequently, the roots were excised from the plants, washed with DDW, lyophilized using a Flexi-Dry MP (Stoneridge, NY) and homogenized into a powder with a

48 mortar and pestle after being frozen with liquid nitrogen. A mild extraction procedure was incorporated in order to preserve the labile compounds within the A. fistulosum plant tissue and adapted from Ref.36 In summary, ~25 mg of plant roots were combined with 1.5 ml of 20 mmol l-1 Tris-HCl (pH 7.5) and stirred at room temperature for 1.5 h. The solution was then centrifuged at 5000 rpm for 15 min. The supernatant was decanted, filtered through a 0.2 μm syringe filter and 100 μl were injected into the IPRP-ICPMS.

Urine: The urine sample was collected in a polypropylene tube from a human subject, who had not previously consumed any selenium dietary supplements. After homogenization via vortexing, 1 ml of solution was diluted to 10 ml with DDW. The resulting solution was vortexed, filtered through a 0.2 µm syringe filter and 100 µl were injected onto the IPRP-

ICPMS.

All sample matrices were spiked with ~50 ng ml-1 of arsenic and selenium standards.

2.3.4 Chromatographic optimization

Although many of the monitored selenium and arsenic species are predominately charged in solution, ion-pairing reversed phase chromatography was chosen for the separation technique because the previously attempted methods utilizing anion exchange chromatography could not facilitate simultaneous elemental speciation. Considering the anionic character of the analytes, various cationic ion-pairing reagents were tested. Tetraethyl ammonium perchlorate (TEAP) was initially investigated as an ion-pairing reagent but resulted in decreased analyte retention time and coelution of the selenium and arsenic species, respectively. This was attributed to a lack of hydrophophic character (shorter length of the hydrocarbon chain) of the ion-pairing reagent. Tetrabutyl ammonium hydroxide (TBAH) was then investigated and provided sufficient

49 retention necessary for a baseline separation. Hydroxide was chosen as the counter ion in the ion-pair considering other common counter ions such as bromide and chloride would create an elevated signal baseline resulting in higher detection limits due to the isobaric interference of

79Br1H and 81Br1H for 80Se and 82Se or 35Cl40Ar for 75As.

Table 2.2 The studied species in this work with their pKa values Compound Structure pKa Ref.

III Arsenenous Acid (As ) O As OH 9.3 38

V As Arsenic Acid (As ) HO OH 2.2 38 OH 7.0 O 11.5 As Monomethylarsonic Acid HO OH CH (MMA) 3 4.0 38 O 8.7 As H C OH 3 Dimethylarsinic Acid (DMA) CH3 1.8 38 O 6.1

Se Selenious Acid (SeIV) HO OH 2.4 38 O 7.9 Se VI O OH Selenic Acid (Se ) OH <0 38 OH 1.7

Se O CH3 Selenomethionine (SeMet) 2.6 37 NH2 O 8.9

HO Se Selenocystine (SeCys ) 2.4 37 2 NH 2 2 8.9

37, 38 The wide range in acid dissociation constants (pKa) for the analytes of interest, displayed in Table 2.2, deemed the separation pH dependent. As noted in prior studies, pH 5.5-

6.5 was found to be optimal for the separation of selenium and arsenic species and provided a starting point for further optimization.24, 25 An increase in pH above 6.0 shifted the analytes to a greater elution time due to an increase in the anionic concentration of the analytes, which allowed for increased ion-pairing. As expected, a decrease in pH below 6.0 brought about the counter effect, which resulted in a general loss of retention time. Baseline separation on the front side of the chromatogram without extensive retention for the later eluting compounds was observed at pH 6.0, therefore chosen as optimal.

In this study, the choice of buffer had a significant effect on the chromatographic resolution. When using ammonium citrate at pH 6.0, the selenium species were resolved, but the arsenic species were not all baseline resolved. On the contrary, when using ammonium phosphate at pH 6.0, the arsenic species were resolved; however, the selenium species could not be resolved in a timely manner, due to a significant increase for the elution time of SeVI.

Previous studies have shown the same extended elution time for SeVI.24, 25 Unfortunately, high concentrations of organic solvent cause plasma instability of the ICP; therefore, a secondary option utilizing a buffer gradient was necessary. Considering this, ammonium phosphate was incorporated as the primary buffer and a secondary buffer was added for overall reduction in analyte retention time.

Secondary buffer systems such as ammonium phosphate and ammonium sulfate, without the addition of ion-pairing reagent, were tested. Ammonium phosphate caused very poor peak shape for SeVI and extended elution time. On the contrary, ammonium sulfate was able to effectively elute SeVI in a timely manner. The acid dissociation constants for phosphate and

51 sulfate are pKa1 1.97, pKa2 6.82, pKa3 12.5 and pKa1 -3.0, pKa2 1.9, respectively. At pH 6.0, sulfate was more effective at eluting compounds due to its doubly charged state, which could compete for the ion-pair more effectively than the predominately singly charged phosphate. In summary, the ion-pairing reagent introduced through buffer A was incorporated for analyte retention and subsequently eluted from the column using the greater anionic charge of buffer B, which resulted in a decrease of the overall analyte retention time and an increased resolution on the front side of the separation.

2.4 Results and discussion

2.4.1 Chromatographic performance

The successful separation of eight arsenic and selenium standards within 18 minutes can be observed in Figure 2.1. In order to show the dynamic range of the separation, the standard concentrations were 250 ng ml-1 in Figure 2.1 (A, B) and reduced to 1 ng ml-1 with SeMet at 5 ng ml-1 in Figure 2.1 (C, D). The elution time and baseline separation were not compromised by the analyte concentration. In order to show the potential feasibility of the separation in a matrix of high chloride concentration, such as in urine, Figure 2.1 (A) displays the 35Cl profile at a concentration of 1000 µg ml-1. Although the collision/reaction cell greatly reduces the interferences of 35Cl40Ar on 75As, the chloride interference also is chromatographically resolved from the arsenic species of interest, thereby lowering the detection limits.

The reproducibility of the chromatographic retention times and peak areas were calculated by taking the standard deviation of six replicates. For the arsenic and selenium species, the retention time reproducibility ranged from 0.1-0.9 % RSD. For the arsenic and selenium species, the peak area reproducibility ranged from 0.5-1.5 % RSD.

The chromatographic detection limits were calculated by taking three times the standard deviation of seven replicates for the blank peak areas (3σ) divided by the slope of the calibration curves (IUPAC). The detection limits of arsenic and selenium ranged from 19-22 ng l-1 for arsenic species and 312-442 ng l-1 for selenium species. The calibration curve R2 value ranged between 0.993 and 1.000 for arsenic and selenium species. Column recovery for the reversed phase chromatographic run was calculated to be 97 ± 6 % for combined arsenic species and 98 ±

3 % for combined selenium species. The chromatographic figures of merit are summarized in

Table 2.3.

Table 2.3 Chromatographic Figures of Merit Retention Peak Se or As Calibration Time Area Detection Curve (% RSD) (% RSD) Limits (ng l-1) R2 Value As III 0.9 1.5 19 0.999 DMA 0.2 1.1 29 0.999 MMA 0.1 0.7 21 0.999 AsV 0.1 0.9 22 0.999

SeCys2 0.9 0.9 340 1.000 SeMet 0.6 0.5 442 0.993 SeIV 0.1 1.1 321 0.999 SeVI 0.1 1.5 312 0.999

2.4.2 Electrospray ionization studies

Considering that the only the predominant arsenic and selenium species were chromatographically separated with the developed method, the complementary addition of ESI-

ITMS provides additional information when desired for a given application. Due to the relatively low salt concentrations utilized by the separation method, complementary use of ESI-

ITMS was possible. Figure 2.2 displays the analysis of selenium and arsenic standards at a concentration of 10 µg ml-1. The IPRP-ESI-ITMS chromatograms in the top half of the figure are extracted ion chromatograms (EIC) of m/z 107, 137, 139 and 141 relating to the M-H- of each arsenic analyte of interest. It was important to note that in negative ion mode, three major interferences derived from the mobile phase constituents were prevalent in the mass spectra of

- - the analytes. The interference at m/z 97 could be either phosphate (H2PO4 ) or sulfate (HSO4 ) with subsequent losses of H2O, which resulted in an interference at m/z 79, plus m/z 195 was a

- - phosphate or sulfate cluster relating to the molecular formula of H5P2O8 or H3S2O8 and m/z 177

Figure 2.2 IPRP-ESI-ITMS extracted ion chromatograms of the arsenic standards (top): m/z 107, 137, 139 and 141 and the selenium standards (bottom): m/z 129, 145, 196 and 335 at 10 µg ml-1.

55 represented a loss of H2O from either. Along with the observed M-H- for each analyte peak, the loss or addition of H2O was observed with each exhibiting subsequent mobile phase adducts of

- - H2PO3 or HSO3 . The IPRP-ESI-ITMS chromatograms in the bottom half of the figure are EICs of m/z 129, 145, 196 and 335 relating to the M-H- of each selenium analyte of interest. The selenium containing peaks were easier to identify because of their distinct isotope ratio pattern for a single Se atom, which matched well to the theoretical pattern, (data not shown). Similar mass losses/additions discussed previously for arsenic were also observed in the mass spectra of the selenium compounds. It was important to note that the interference of m/z 195 distorted the selenium isotopic pattern of SeMet. Also, the mass spectrum of the SeCys2 exhibited reduction of the dimer into the monomer, which was apparent at m/z 168. The peak at m/z 248 correlated to the monomer with the second selenium still attached rather than a phosphate adduct, which was confirmed by the distinct isotopic pattern of Se2.

2.4.3 Varying matrix applications

A sample was collected from the Ohio River in order to test the method in a complex water matrix. The acquired river water sample (pH 7.6) after centrifugation and filtration was injected onto the IPRP-ICPMS and the resulting chromatograms illustrating 75As and 78Se profiles can be found in Figure 2.3 (A, B), respectively. Of the arsenic and selenium species monitored, only SeVI was predominant in the sample. This observation was not surprising given the oxidative environment. In previous studies, SeVI has been reported to be the predominant species found in water at pH values near 7.2.6 For standard identification and chromatographic verification of the separation methodology, the river water was spiked with ~50 ng ml-1 of arsenic and selenium standards. The resulting chromatograms after injecting spiked river water

56 onto the IPRP-ICPMS can be found in Figure 2.3 (C, D). All eight arsenic and selenium standards were baseline separated in less than 18 minutes without a significant change in elution time given the complexity of the river water matrix.

Figure 2.3 IPRP-ICPMS chromatograms of river water, (A) 75As profile and (B) 78Se profile, and ~50 ng ml-1 spiked river water (C) 75As profile and (D) 78Se profile.

The A. fistulosum was selected in order to test the developed chromatographic method in a plant matrix. In particular, the A. fistulosum has been studied prior for the identification of selenium metabolites.39 The supernatant from the A. fistulosum extraction was injected onto the

IPRP-ICPMS and the resulting 75As and 78Se chromatograms can be found in Figure 2.4 (A, B),

57 respectively. In the plant root extract, predominantly inorganic arsenic (AsIII, AsV) and selenium

(SeIV, SeVI) species were present at concentrations below 50 ng ml-1. Organic forms were not observed and may account for the A. fistulosum’s ability to withstand small concentrations of toxic inorganic arsenic and selenium species without further metabolic changes. The apparent split peak observed in the selenium profile for SeVI was due to an interference caused by the chromatographic gradient used. In most cases, this is easily background subtracted using a blank injection as the background. However, this was intentionally not done in order to illustrate the possibility of signal interference.

For standard identification and chromatographic verification of the separation methodology, the A. fistulosum extract was spiked with ~50 ng ml-1 of arsenic and selenium standards. The resulting chromatograms after injecting spiked plant extract onto the IPRP-

ICPMS are depicted in Figure 2.4 (C, D). All eight arsenic and selenium standards were baseline separated in less than 18 minutes without a significant change in elution time considering the complexity of the root extract matrix from the plant. As seen in the natural root extract, an interference from the chromatographic gradient used in the method produced a shoulder before the SeVI peak. Although noteworthy, the selenium shoulder did not significantly interfere with the SeVI signal.

A urine sample was collected in order to test the developed chromatographic method in a high chloride matrix as well as a sample of high general interest. The collected urine sample was injected onto the IPRP-ICPMS and the resulting chromatograms illustrating 75As and 78Se profiles can be found in Figure 2.5 (A, B), respectively. Although significant amounts of arsenic were not excreted by the subject, the predominant selenium species, SeIV and to a lesser extent SeCys2, represented the subject’s metabolism toward a reduced form of selenium. For standard identification and chromatographic verification of the separation methodology, the urine sample was spiked with ~50 ng ml-1 of arsenic and selenium standards and the resulting chromatograms after injecting spiked urine to the IPRP-ICPMS can be found in Figure 2.5 (C,

D). As seen for other sample types, all eight arsenic and selenium standards were baseline separated in less than 18 minutes without a significant change in elution time for the urine matrix. The ability to chromatographically separate the chloride interference (40Ar35Cl) from arsenic (75As), as can be observed in Figure 2.1 (A), in combination with the collision/reaction cell, allowed for method flexibility.

Figure 2.5 IPRP-ICPMS chromatograms of urine, (A) 75As profile and (B) 78Se profile, and ~50 ng ml-1 spiked urine (C) 75As profile and (D) 78Se profile.

2.5 Conclusion

With the ever increasing levels of arsenic and selenium in the environment due to natural and anthropogenic activities and the growing awareness of the need for element speciation not

IV VI just quantification, a method to speciate four selenium species (Se , Se , SeMet, SeCys2) and four arsenic species (AsIII, AsV, MMA, DMA) via IPRP-ICPMS and IPRP-ESI-ITMS has been developed. The results show a highly sensitive and selective baseline separation including column re-equilibration time of the eight species using TBAH as the ion-pairing reagent on a C18

60 column within 18 minutes and detection limits in the ng l-1 range. This novel methodology utilized a fast gradient addition of ammonium sulfate, which effectively displaced the later eluting selenium and arsenic species allowing for shorter elution times without compromising the front end resolution of the separation. In order to show the flexibility of the designed method, selenium and arsenic compounds naturally and spiked within river water, plant extract and urine matrices were effectively speciated. Future investigations will employ other environmental and biological matrices for additional method testing.

2.6 Acknowledgements

The authors would like to acknowledge support from the NIEHS-SBRP grant ES04908 and Agilent Technologies for their instrumentation and continuing support.

2.7 References

1. Brown, K. M.; Arthur, J. R., Selenium, selenoproteins and human health: a review. Public Health Nutr FIELD Full Journal Title:Public health nutrition 2001, 4, (2B), 593-9. 2. Combs, G. F., Jr.; Clark, L. C.; Turnbull, B. W., An analysis of cancer prevention by selenium. BioFactors 2001, 14, (1-4), 153-159. 3. Duffield-Lillico, A. J.; Dalkin, B. L.; Reid, M. E.; Turnbull, B. W.; Slate, E. H.; Jacobs, E. T.; Marshall, J. R.; Clark, L. C., Selenium supplementation, baseline plasma selenium status and incidence of prostate cancer: An analysis of the complete treatment period of the nutritional prevention of cancer trial. BJU International 2003, 91, (7), 608-612. 4. Medina, D.; Thompson, H.; Ganther, H.; Ip, C., Se-methylselenocysteine: A new compound for chemoprevention of breast cancer. Nutrition and Cancer 2001, 40, (1), 12-17. 5. Le, X. C.; Li, X. F.; Lai, V.; Ma, M.; Yalcin, S.; Feldmann, J., Simultaneous speciation of selenium and arsenic using elevated temperature liquid chromatography separation with inductively coupled plasma mass spectrometry detection. Spectrochimica Acta, Part B: Atomic Spectroscopy 1998, 53B, (6-8), 899-909. 6. Bujdos, M.; Mulova, A.; Kubova, J.; Medved, J., Selenium fractionation and speciation in rocks, soils, waters and plants in polluted surface mine environment. Environmental Geology (Berlin, Germany) 2005, 47, (3), 353-360. 7. Shibata, Y.; Morita, M.; Fuwa, K., Selenium and arsenic in biology: their chemical forms and biological functions. Advances in Biophysics 1992, 28, 31-80. 8. Zawislanski, P. T.; Benson, S. M.; TerBerg, R.; Borglin, S. E., Selenium Speciation, Solubility, and Mobility in Land-Disposed Dredged Sediments. Environmental Science and Technology 2003, 37, (11), 2415-2420. 9. Uthus, E. O., Evidence for arsenic essentiality. Environmental Geochemistry and Health 1992, 14, (2), 55-8. 10. Khan, B. I.; Jambeck, J.; Solo-Gabriele, H. M.; Townsend, T. G.; Cai, Y., Release of arsenic to the environment from CCA-treated wood. 2. Leaching and speciation during disposal. Environmental Science and Technology 2006, 40, (3), 994-999. 11. Nriagu, J. O., Arsenic in the environment. Part I: cycling and characterization. 430 pages, hard cover. Part II: human health and ecosystem effects. 292 pages, hard cover. 1995; Vol. 21, p 277. 12. Wang, S.; Mulligan, C. N., Occurrence of arsenic contamination in Canada: Sources, behavior and distribution. Science of the Total Environment 2006, 366, (2-3), 701-721. 13. B'Hymer, C.; Caruso, J. A., Arsenic and its speciation analysis using high-performance liquid chromatography and inductively coupled plasma mass spectrometry. Journal of Chromatography, A 2004, 1045, (1-2), 1-13. 14. B'Hymer, C.; Caruso, J. A., Selenium speciation analysis using inductively coupled plasma-mass spectrometry. Journal of Chromatography, A 2006, 1114, (1), 1-20. 15. Chen, Z.; Akter, K. F.; Rahman, M. M.; Naidu, R., Speciation of arsenic by ion chromatography inductively coupled plasma mass spectrometry using ammonium eluents. Journal of Separation Science 2006, 29, (17), 2671-2676. 16. Heitkemper, D. T.; Vela, N. P.; Stewart, K. R.; Westphal, C. S., Determination of total and speciated arsenic in rice by ion chromatography and inductively coupled plasma mass spectrometry. Journal of Analytical Atomic Spectrometry 2001, 16, (4), 299-306.

17. Kohlmeyer, U.; Kuballa, J.; Jantzen, E., Simultaneous separation of 17 inorganic and organic arsenic compounds in marine biota by means of high-performance liquid chromatography/inductively coupled plasma mass spectrometry. Rapid Communications in Mass Spectrometry 2002, 16, (10), 965-974. 18. Orero Iserte, L.; Roig-Navarro, A. F.; Hernandez, F., Simultaneous determination of arsenic and selenium species in phosphoric acid extracts of sediment samples by HPLC-ICP-MS. Analytica Chimica Acta 2004, 527, (1), 97-104. 19. Vassileva, E.; Becker, A.; Broekaert, J. A. C., Determination of arsenic and selenium species in groundwater and soil extracts by ion chromatography coupled to inductively coupled plasma mass spectrometry. Analytica Chimica Acta 2001, 441, (1), 135-146. 20. Wang, R.-Y.; Hsu, Y.-L.; Chang, L.-F.; Jiang, S.-J., Speciation analysis of arsenic and selenium compounds in environmental and biological samples by ion chromatography- inductively coupled plasma dynamic reaction cell mass spectrometer. Analytica Chimica Acta 2007, 590, (2), 239-244. 21. Kotrebai, M.; Tyson, J. F.; Block, E.; Uden, P. C., High-performance liquid chromatography of selenium compounds utilizing perfluorinated carboxylic acid ion-pairing agents and inductively coupled plasma and electrospray ionization mass spectrometric detection. Journal of Chromatography, A 2000, 866, (1), 51-63. 22. Muniz-Naveiro, O.; Dominguez-Gonzalez, R.; Bermejo-Barrera, A.; Bermejo-Barrera, P.; Cocho, J. A.; Fraga, J. M., Selenium speciation in cow milk obtained after supplementation with different selenium forms to the cow feed using liquid chromatography coupled with hydride generation-atomic fluorescence spectrometry. Talanta 2007, 71, (4), 1587-1593. 23. Wrobel, K.; Wrobel, K.; Kannamkumarath, S. S.; Caruso, J. A.; Wysocka, I. A.; Bulska, E.; Swiatek, J.; Wierzbicka, M., HPLC-ICP-MS speciation of selenium in enriched onion leaves - a potential dietary source of Se-methylselenocysteine. Food Chemistry 2004, 86, (4), 617-623. 24. Do, B.; Robinet, S.; Pradeau, D.; Guyon, F., Speciation of arsenic and selenium compounds by ion-pair reversed-phase chromatography with electrothermal atomic absorption spectrometry. Application of experimental design for chromatographic optimization. Journal of Chromatography, A 2001, 918, (1), 87-98. 25. Pan, F.; Tyson, J. F.; Uden, P. C., Simultaneous speciation of arsenic and selenium in human urine by high-performance liquid chromatography inductively coupled plasma mass spectrometry. Journal of Analytical Atomic Spectrometry 2007, 22, (8), 931-937. 26. Ipolyi, I.; Fodor, P., Development of analytical systems for the simultaneous determination of the speciation of arsenic [As(III), methylarsonic acid, dimethylarsinic acid, As(V)] and selenium [Se(IV), Se(VI)]. Analytica Chimica Acta 2000, 413, (1-2), 13-23. 27. Wietecha-Posluszny, R.; Dobrowolska, J.; Koscielniak, P., Method for determination of selenium and arsenic in human urine by atomic fluorescence spectrometry. Analytical Letters 2006, 39, (15), 2787-2796. 28. Niedzielski, P., The new concept of hyphenated analytical system: Simultaneous determination of inorganic arsenic(III), arsenic(V), selenium(IV) and selenium(VI) by high performance liquid chromatography-hydride generation-(fast sequential) atomic absorption spectrometry during single analysis. Analytica Chimica Acta 2005, 551, (1-2), 199-206. 29. Viitak, A.; Volynsky, A. B., Simple procedure for the determination of Cd, Pb, As and Se in biological samples by electrothermal atomic absorption spectrometry using colloidal Pd modifier. Talanta 2006, 70, (4), 890-895.

30. Fitzpatrick, S.; Ebdon, L.; Foulkes, M. E., Separation and Detection of Arsenic and Selenium Species in Environmental Samples by HPLC-ICP-MS. International Journal of Environmental Analytical Chemistry 2002, 82, (11-12), 835-841. 31. Li, B.; He, H.; Shi, S.; Ma, X.; Wen, H.; Lu, C., Simultaneous determination of iodine, bromine, selenium and arsenic in geological samples by inductively coupled plasma mass spectrometry. Journal of Analytical Atomic Spectrometry 2002, 17, (4), 371-376. 32. Rao, R. N.; Talluri, M. V. N. K., An overview of recent applications of inductively coupled plasma-mass spectrometry (ICP-MS) in determination of inorganic impurities in drugs and pharmaceuticals. Journal of Pharmaceutical and Biomedical Analysis 2007, 43, (1), 1-13. 33. Dernovics, M.; Giusti, P.; Lobinski, R., ICP-MS-assisted nanoHPLC-electrospray Q/time-of-flight MS/MS selenopeptide mapping in Brazil nuts. Journal of Analytical Atomic Spectrometry 2007, 22, (1), 41-50. 34. Dumont, E.; Ogra, Y.; Vanhaecke, F.; Suzuki, K. T.; Cornelis, R., Liquid chromatography-mass spectrometry (LC-MS): a powerful combination for selenium speciation in garlic (Allium sativum). Analytical and Bioanalytical Chemistry 2006, 384, (5), 1196-1206. 35. Iwaoka, M.; Ooka, R.; Nakazato, T.; Yoshida, S.; Oishi, S., Synthesis of selenocysteine and selenomethionine derivatives from sulfur-containing amino acids. Chemistry & Biodiversity 2008, 5, (3), 359-374. 36. Mestek, O.; Polak, J.; Juricek, M.; Karvankova, P.; Koplik, R.; Santrucek, J.; Kodicek, M., Trace element distribution and species fractionation in Brassica napus plant. Applied Organometallic Chemistry 2007, 21, (6), 468-474. 37. Chang, S. Y.; Chiang, H.-T., Simultaneous determination of selenium and antimony compounds by capillary electrophoresis with indirect fluorescence detection. Electrophoresis 2002, 23, (17), 2913-2917. 38. Kitagawa, F.; Shiomi, K.; Otsuka, K., Analysis of arsenic compounds by capillary electrophoresis using indirect UV and mass spectrometric detections. Electrophoresis 2006, 27, (11), 2233-2239. 39. Shah, M.; Kannamkumarath, S. S.; Wuilloud, J. C. A.; Wuilloud, R. G.; Caruso, J. A., Identification and characterization of selenium species in enriched green onion (Allium fistulosum) by HPLC-ICP-MS and ESI-IT-MS. Journal of Analytical Atomic Spectrometry 2004, 19, (3), 381-386.

Chapter 3

Elucidating the selenium and arsenic metabolic pathways following exposure to the non- hyperaccumulating Chlorophytum comosum, spider plant

3.1 Abstract

Although many studies have investigated the metabolism of selenium and arsenic in hyperaccumulating plants for phytoremediation purposes, few have explored non- hyperaccumulating plants as a model for general contaminant exposure to plants. In addition, the result of simultaneous supplementation with selenium and arsenic has not been investigated in plants. In this study, Chlorophytum comosum, commonly known as the spider plant, was used to investigate the metabolism of selenium and arsenic after single and simultaneous supplementation. Size exclusion and ion-pairing reversed phase liquid chromatography were coupled to an inductively coupled plasma mass spectrometer to obtain putative metabolic information of the selenium and arsenic species in the C. comosum after a mild aqueous extraction. The chromatographic results depict that selenium and arsenic species were sequestered in the roots and generally conserved upon translocation to the leaves. The data suggests that selenium was directly absorbed in the C. comosum root when supplemented with

SeVI, but a combination of passive and direct absorption when supplemented with SeIV due to partial oxidation of SeIV to SeVI in the rhizosphere. Higher molecular weight selenium species were more prevalent in the roots of plants supplemented with SeIV, but in the leaves of plants supplemented SeVI due to an increased translocation rate. When supplemented as AsIII, arsenic is proposed to be passively absorbed as AsIII and partially oxidized to AsV in the plant root.

Although total elemental analysis demonstrates a selenium and arsenic antagonism, a compound containing selenium and arsenic was not present in the general aqueous extract of the plant.

3.2 Introduction

In addition to the natural geological release of arsenic into ground water and soil, anthropogenic activities such as industrial production of pesticides, herbicides, wood preservatives and mining have increased arsenic levels beyond natural concentrations, causing worldwide environmental concern.1 Arsenic present in soil can enter the food chain via plant accumulation. Some of the most common arsenic species in the environment include arsenite

(AsIII), arsenate (AsV), monomethlyarsonate (MMA), and dimethylarsinate (DMA), in order of decreasing toxicities.2 General phytoremediation efforts, utilizing plants to remove toxins from the environment, have focused on hyperaccumulating plants for the depletion of arsenic.3 To accurately assess environmental risk and continue to develop more effective phytoremediation strategies using alternative plants, the metabolism of arsenic should be studied in a variety of plants.

Selenium is considered one of the most widely distributed elements on Earth, having an average soil abundance of 0.09 mg kg-1. Further, considerable concentration variability exists from one location to another, such as high selenium concentrations occurring in a few localized regions.4 As with arsenic, selenium contained in the soil environment can enter the food chain through plant accumulation. Although selenium has been identified as a necessary element to animal life and possesses cancer chemopreventive properties from clinical trials,5 its narrow range between deficiency and toxicity deem the uptake and accumulation of selenium worthy of extensive investigation.6 While essential to mammalian health, the question of selenium necessity as a micronutrient in plants remains unanswered.7 In order to properly assess environmental danger and continue to develop more effective phytoremediation strategies using alternate plants, the metabolism of selenium should be studied in a variety of plants.

In past studies, selenium has been shown to have an antagonistic affect on toxic elements in plants.8 Investigations over half a century ago provided evidence for a detoxifying or protective effect after toxic concentrations of selenium and arsenic were simultaneously administered to rats.9 More recently, the structure elucidated for the interaction of selenium and arsenic in a mammalian system was described as seleno-bis(S-glutathionyl) arsinium ion

- 10 [(GS)2AsSe] . Although the effects of selenium and arsenic have independently been studied in various plant matrices, little research has been devoted to provide information on a potential selenium and arsenic interaction at the molecular level within plants. If observed, an antagonism between selenium and arsenic may prove useful for further phytoremediation studies.

In general, extensive effort has been put forth to understand the metabolic pathways of contaminants such as arsenic11 and selenium12 in hyperaccumulating plants. However, few studies have investigated the metabolism of such contaminants in non-hyperaccumulating plants, which could act as a model for general environmental exposure. When considering the potential of contaminant remediation by genetically modified or native plants (wild type), the metabolic pathways and any variation in metabolism should be fully understood for accumulating and non- accumulating plants, as investigated in a previous study.13 Considering the increasing level of global contamination, studies on the metabolism of selenium and arsenic in non- hyperaccumulating plants are imperative to provide vital information about general environmental effects.

Size exclusion chromatography (SEC) provides a general molecular weight range of the varying species in the soluble portion of a plant matrix, such as extracted proteins.14 SEC has previously been used to monitor selenium and arsenic in various matrices such as Allium schoenoprasum (chives) and Antarctic krill.15, 16 While SEC can provide information on possible

68 interactions between molecules, poor analyte resolution causes the technique to be unsuitable for small molecule speciation. In the past, the two most frequently employed techniques to speciate and thus identify different selenium and arsenic species have been ion exchange and ion-pairing reversed phase chromatography (IPRP).17, 18 The most common arsenic and selenium species previously found in plants and soil were AsIII, AsV, MMA, DMA, selenite (SeIV), selenate (SeVI),

2, 19 selenomethione (SeMet) and selenocystine (SeCys2). A recent method displayed the ability to separate all eight species in a timely and sensitive manner using ion-pairing reversed phase chromatography coupled with inductively coupled plasma mass spectrometry (IPRP-ICPMS) for online detection.20 In addition, the fast, multi-elemental detection at trace levels allowing for the sensitivity and selectivity provided by ICPMS has been previously used in past studies for selenium and arsenic speciation in plant matrices.21, 22

In this study, the selected plant species is the Chlorophytum comosum, commonly known as the spider plant. C. comosum is generally known to be robust in varying cultivation conditions allowing for ease of care and possesses an extensive root system beneficial for nutrient and contaminant absorption. Further, earlier studies in this laboratory have shown preferential segregation of metal toxins in the plant roots.23, 24 The two main plant compartments, leaves and roots, were monitored for the absorption and translocation of selenium and arsenic metabolites. This study probes the potential effects of single and simultaneous addition of selenium and arsenic within C. comosum plants.

3.3 Experimental

3.3.1 Instrumentation

High-performance liquid chromatography. Chromatographic separations were accomplished with an Agilent 1100 liquid chromatograph by Agilent Technologies (Santa Clara,

CA) equipped with a vacuum de-gasser system, a binary HPLC pump, an autosampler and a thermostated column compartment. The column used for SEC was a Superdex Peptide 10/300

GL (10 mm x 300 mm x 13 μm) from Amersham Pharmacia Biotech AB (Uppsala, Sweden) and was calibrated with the following standards: cytochrome C, 12.5 kDa; insulin chain B oxidized,

3.5 kDa; and vitamin B12, 1.4 kDa obtained from Sigma-Aldrich Co. (St. Louis, MO). Reversed phase chromatography was carried out with a ZORBAX Eclipse XDB-C18 column (5 μm x 4.6 mm id x 250 mm) from Agilent Technologies (Santa Clara, CA).

Inductively coupled plasma mass spectrometry. The ICPMS used for specific element detection was an Agilent 7500ce by Agilent Technologies (Santa Clara, CA). The instrument was equipped with a microconcentric nebulizer made by Glass Expansion (Pocasset, MA), a

Scott double channel spray chamber (cooled to 2°C), a shielded torch, an octopole collision/reaction cell with hydrogen gas pressurization (purity of 99.999%), a quadrupole mass analyzer and an electron multiplier for detection.

Lyophilization and digestion. A Flexi-Dry MP lyophilizer (Stoneridge, NY) was used for freeze drying purposes. The microwave system used for digestion was an Intelligent

Explorer/Discover system produced by the CEM Corporation (Mathews, NC). The microwave system was programmable for time, temperature, power and pressure, and equipped with a 24 vial autosampler and a self contained microwave chamber.

A summary of all instrumental conditions can be found in Table 3.1.

3.3.2 Reagents and standards

All the solutions were prepared in 18 MΩ cm-1 doubly deionized water (DDW) processed by Sybron/Barnstead (Boston, MA). Standards used for supplementation and identification were the following: disodium methyl arsonate hexahydrate (MMA) purchased from Chem Service

(West Chester, PA); L(+)-selenomethionine (SeMet), the form commonly found within biological samples such as plants,25 obtained from Acros Organics (Morris Plains, NJ); sodium

III (meta)arsenite (As ), cacodylic acid (DMA) and seleno-L-cystine (SeCys2) acquired from Fluka

(Milwaukee, WI); potassium arsenate (AsV), potassium selenate (SeVI) and sodium selenite

(SeIV) purchased from Sigma-Aldrich (St. Louis, MO).

For total elemental analysis, digestion of plant biomass was accomplished using nitric acid (HNO3) obtained from Pharmco Products Inc. (Brookfield, CT) and hydrogen peroxide

(30%) from Fisher Scientific (Fair Lawn, NJ). Claritas PPT selenium and arsenic elemental standards used for quantification were acquired from SpexCertiPrep (Metuchen, NJ).

Calibration standards of 1.0 µg l-1 to 500.0 µg l-1 were prepared through dilution from a stock solution with 2% v/v HNO3.

The following depicts the preparation of mobile phases used for plant extraction and chromatographic separation. The mobile phase for SEC and general plant biomass extraction was made by dissolving tris(hydroxymethyl) aminomethane hydrochloride (Tris-HCl) from

Fisher Scientific (Fair Lawn, NJ) in DDW and adjusting the pH with hydrochloric acid. For

IPRP-ICPMS, mobile phase A contained 5 mmol l-1 tetrabutylammonium hydroxide (TBAH) from Fluka (Milwaukee, WI) and 2.5 mmol l-1 ammonium phosphate from Sigma-Aldrich Co.

(St.Louis, MO) at pH 6.0. Mobile phase B contained 10 mmol l-1 ammonium sulfate from

Sigma-Aldrich Co. (St. Louis, MO) at pH 6.0. The pH was adjusted with phosphoric acid for

71 mobile phase A and ammonium hydroxide for mobile phase B. A summary of the mobile phase conditions are depicted in Table 3.1. All samples were filtered through a 0.2 μm membrane syringe filter by Econofilters from Agilent Technologies, Inc. (Santa Clara, CA) before being injected into the HPLC-ICPMS.

3.3.3 Plant growth and supplementation

The C. comosum was cultivated from seed at the University of Cincinnati green house,

Department of Biological Sciences, Cincinnati, OH. The general purpose potting soil used to cultivate the plants was Premier Pro-Mix (Riviere-du-Loup, Quebec, Canada). During the growth period, plants were fertilized with 25% Hoagland solution as needed.26 After nine months of growth, the plants were split into six groups and supplemented with varying combinations of NaAsO2, K2SeO4, and Na2SeO3 at 25 ml a day for four days as depicted: Group

I, 30 mg l-1 SeIV; Group II, 30 mg l-1 SeVI; Group III, 20 mg l-1 AsIII; Group IV, 30 mg l-1 SeIV &

20 mg l-1 AsIII; Group V, 30 mg l-1 SeVI & 20 mg l-1 AsIII; Group VI, control. AsIII was chosen for supplementation based on prior studies depicting the formation of a selenium and arsenic complex within a mammalian system after simultaneous supplementation with selenium.10

Subsequently, the plants were allowed to mature one additional week before harvesting. The health of each plant was visually indifferent to the supplementation type given. During the process of harvesting, the plants were separated into roots and leaves, washed with DDW and lyophilized. Finally, the plants were homogenized into a powder and stored at -20 °C to prevent any further enzymatic activity leading to interspecies conversion, therefore changing the native distribution.

Table 3.1 Instrumental conditions in this study ICP-MS Forward power 1500 W Plasma gas flow 15.0 l min-1 Carrier gas flow 0.96 l min-1 Makeup gas flow 0.14 l min-1 -1 Collision gas 3.5 ml min H2 Quadrupole bias -16.0 V Octopole bias -18.0 V Monitored isotopes 75As, 77Se, 78Se, 80Se, 82Se Dwell time 100 ms per isotope HPLC SEC Mobile phase 100 mmol l-1 Tris-HCl (pH 7.5) Flow rate 0.60 ml min-1 Injection volume 100 μl IPRP Mobile phase (A) 5 mmol l-1 TBAH in 2.5 mmol l-1

(NH4)3PO4 (pH 6.0) -1 Mobile phase (B) 10 mmol l (NH4)2SO4 (pH 6.0) Flow rate 1.0 ml min-1 Injection volume 100 μl Gradient program Time (min) 0 0.5 1.5 5 6 18 % A 100 100 0 0 100 100 % B 0 0 100 100 0 0 Microwave Stage 1 Stage 2 Stage 3 Power (W) 125 125 150 Ramp (min) 1:00 1:00 1:00 Hold (min) 1:00 2:00 2:00 Temperature (°C) 120 175 170

3.3.4 Total selenium and arsenic determination

For the determination of total selenium and arsenic in C. comosum, a closed vessel microwave digestion system was employed. Three replicates of lyophilized plant biomass for

73 each supplementation type were subjected to the following three stage digestion program, which is summarized in Table 3.1. Briefly, 1 ml of HNO3 was added to approximately 50 mg of plant biomass and digested by Stage 1 and Stage 2 conditions. Subsequently, 0.2 ml of 30% H2O2 were added to the solution and digested by Stage 3 conditions. Following the microwave digestion sequence, the resulting solutions were diluted with DDW to 50 ml (should be 10 ml) and analyzed by ICPMS in continuous flow sample introduction mode. Of the selenium isotopes monitored, 78Se was found to give the lowest limits of detection.

3.3.5 Extraction procedures for plant tissues

A mild extraction procedure was incorporated in order to preserve the labile compounds in C. comosum plant tissue. In summary, 30 mg of homogenized plant biomass from the root or leaf were combined with 1.5 ml of 20 mmol l-1 Tris-HCl (pH 7.5) and stirred at room temperature for 1.5 h. The solution was then centrifuged at 5000 rpm for 15 min. The supernatant was decanted, filtered through a 0.2 μm filter and 100 μl were injected into the SEC-

ICPMS and IPRP-ICPMS. The chromatographic mobile phase conditions can be found in Table

3.1. In addition, total elemental analysis of the supernatant via ICPMS was performed.

Extraction efficiencies were calculated as a percentage of the total elemental analysis of the lyophilized plant tissue. A similar treatment was used for all plant supplementation types.

3.4 Results and discussion

3.4.1 Total element accumulation

Total C. comosum accumulation of selenium and arsenic was determined via microwave digestion and subsequent analysis by continuous flow ICPMS. The resulting selenium and

Figure 3.1 The C. comosum accumulation of arsenic and selenium for the varying supplementation types administered during the cultivation process.

75 arsenic concentrations of the leaves and roots for the varying supplementation types are depicted in Figure 3.1. The error bars represent one standard deviation of three replicates for each supplementation type. Overall, the results show a sequestering of selenium and arsenic species in the C. comosum roots, which concurs with previous studies demonstrating species sequestering in the roots after supplementation of selenium in Brassica oleracea22 and arsenic in

Brassica juncea.27

The total concentration of selenium in the roots of the SeIV supplemented plants was 18.7

µg g-1, which displays the inability of C. comosum to accumulate large concentrations of selenium. The difference in accumulation and translocation of selenium between different supplementation types was ascertained by the total selenium concentrations of 1.7 µg g-1 for the leaves and 18.7 µg g-1 for the roots after SeIV supplementation, whereas after SeVI supplementation, concentrations were 7.2 µg g-1 for the leaves and 10.8 µg g-1 for the roots.

These findings suggest an increased rate of selenium translocation from roots to leaves in the C. comosum after supplementation with SeVI versus SeIV, which is in agreement with previous plant studies.28 General consensus defines plants as non-accumulators that accumulate less than 25 µg g-1 of environmental contaminants, which classifies the C. comosum as a selenium nonaccumulator. In contrast to selenium uptake, greater arsenic accumulation was observed. The total concentration of arsenic in roots of the AsIII supplemented plants was 51.9 µg g-1, which demonstrates the capability of the C. comosum for arsenic accumulation. In the leaves of the

AsIII supplemented plants, the total concentration of arsenic was 9.1 µg g-1, therefore showing a considerable resistance to arsenic translocation. General consensus defines plants that accumulate 25 µg g-1 to 100 µg g-1 of environmental contaminants as secondary absorbers, the case for C. comosum.

For SeIV & AsIII supplemented plants, 10.4 µg g-1 of selenium and 38.9 µg g-1 of arsenic were observed in the roots, exhibiting a 44.4% and 25.0% decrease in accumulation, respectively, compared to single elemental supplementation. For SeVI & AsIII supplemented plants, 6.1 µg g-1 of selenium and 15.2 µg g-1 of arsenic were detected in the roots showing a

43.5% and 70.7% decrease, respectively, compared to single element supplementation. These findings suggest a mutual antagonism between selenium and arsenic upon simultaneous C. comosum supplementation. In accord with individual supplementation, the degree of accumulation in the roots or leaves of the C. comosum varied according to the form of selenium supplemented to the soil.

Overall, selenium and arsenic antagonism may occur by several pathways. The selenium and arsenic species may bind and form an insoluble complex, such as orpiment (As2Se3), resulting in a biologically unavailable selenium and arsenic species. Bacteria have been shown to reduce selenium and sulfur from selenate and sulfate to selenide and sulfide, respectively 29, 30.

It has also been demonstrated that sulfide, when produced abiotically or microbially, can

31 chemically reduce arsenic resulting in the formation of As2S3. These findings support a possible formation of As2Se3 in the soil environment after simultaneous supplementation of selenium and arsenic. Another possibility allowing for mutual detoxification of the two environmental contaminants may be through the formation of an arsenic selenium complex similar to that observed in the mammalian system: seleno-bis(S-glutathionyl) arsinium ion

- 10 [(GS)2AsSe] . In order to further investigate a possible selenium and arsenic containing species in the C. comosum, SEC-ICPMS and IPRP-ICPMS were utilized.

3.4.2 Root extract characterization of selenium and arsenic species

The utilization of SEC-ICPMS provided an overall molecular weight distribution of the selenium and arsenic containing compounds in the C. comosum. Plant roots from varying supplementation combinations were analyzed after a general extraction near physiological pH.

An example of the extraction efficiencies for the plant roots were calculated as 91 ± 6% (75As) and 31 ± 4% (78Se) with SeIV & AsIII supplemented plants (n = 3). Although these results display a near complete arsenic extraction, a large amount of selenium remained in the unextracted fraction of the root. The resulting chromatograms after injecting 100 µl of the water soluble plant supernatant from the Tris-HCl extraction into the SEC-ICPMS are represented in Figure

3.2. The SEC column recovery was calculated as 108 ± 17% (75As) and 102 ± 3% (78Se) for SeIV

& AsIII supplemented plants (n = 3) indicating negligible loss from analyte adsorption to the stationary phase. The predominant selenium and arsenic species eluted after the 1.4 kDa standard in all chromatograms, which indicates small molecules such as peptides or inorganic species. High molecular weight species were more prevalent in plants supplemented with SeIV than SeVI, which suggests an alteration in the selenium metabolism depending on the supplementation form. The lack of a void volume peak in the arsenic profiles illustrates arsenic exclusion from macromolecules such as proteins. Overall, the profile consistency demonstrates a general conservation of selenium and arsenic species, whether singly or simultaneously supplemented.

While the overall selenium accumulation was reduced when arsenic was supplemented simultaneously, the profile consistency suggests the metabolic pathway remains predominantly unaltered. This same phenomenon is also observed in comparing the arsenic profile of the root extract from plants supplemented with arsenic including or excluding selenium. After

Figure 3.2 78Se and 75As SEC-ICPMS chromatograms of the root extract from the C. comosum after varying supplementation combinations.

79 investigation of the selenium and arsenic chromatograms, a lack of profile overlap demonstrates that a selenium and arsenic containing molecule was not present in the plant roots regardless of the supplementation type. Whereas total elemental analysis provides evidence of a selenium and arsenic antagonism, the metabolic pathway of interaction did not result in a water soluble selenium and arsenic containing molecule in C. comosum.

To further characterize the selenium and arsenic containing compounds in the C. comosum root extract after varying supplementation combinations, IPRP-ICPMS was incorporated and the resulting chromatograms are shown in Figure 3.3. The calculated column recovery was 87 ± 3% (75As) and 55 ± 1% (78Se) for SeIV & AsIII supplemented plants (n = 3) indicating minimal loss of arsenic from analyte adsorption to the stationary phase; however, the selenium loss may be caused by non-eluting selenium macromolecular compounds. Although the amount of selenium and arsenic in the soil was not quantified, the control plants provide insight into the low molecular weight species metabolized after long term exposure to selenium and arsenic concentrations naturally found in commercial soil over the nine month cultivation period. Only inorganic selenium and arsenic species were observed in C. comosum control root.

Additionally, plants singly supplemented with selenium or arsenic showed a decrease in abundance for arsenic or selenium species, respectively, which supports the proposed antagonistic effect between the two.

Inorganic selenium species were predominately observed in the selenium supplemented

C. comosum root. Specifically in plants supplemented with SeIV, the concentration of SeIV and

SeVI in root extract was 14.3% and 74.6% of the total, respectively. The specific percentages reported in the manuscript for IPRP-ICPMS chromatograms are qualitative and used to aid visual interpretation. While the chromatograms were reproducible, no statistical analysis was

Figure 3.3 78Se and 75As IPRP-ICPMS chromatograms of the root extract from the C. comosum after varying supplementation combinations. 81 performed. The conversion of the selenium species to a more oxidized form than originally supplemented is contradictory to the suggested metabolic pathway of selenium in a plant.7 This finding suggests that oxidation occurred in the rhizosphere, the dynamic microenvironment immediately surrounding the plant roots, and may provide conditions significantly different from the adjacent bulk soil.32 The difference in bulk soil pH may be described by the pH values for the solutions administered during supplementation: NaAsO2 (9.15), K2SeO4 (7.17) and Na2SeO3

(8.77). In order to acquire necessary anions for biological processes, mmols of OH- can be released from the plant roots creating a potential difference between the root-soil interface,

- - 2- - which allows for absorption of anions such as NO3 , Cl , SO4 and H2PO4 to maintain the charge balance. The overall process generates rhizosphere alkalinity.33, 34 In addition, a prior study found SeVI to be the major form of selenium in environmental water sources at higher pH values.19 After initial supplementation with SeIV, the selenium species may have oxidized to

SeVI due to an alkaline pH shift during nutrient uptake, which would allow for direct absorption of selenium into the plant root through the sulfate pathway. Plants supplemented with SeVI revealed a similar selenium chromatographic profile in general; however, SeIV and SeVI made up

0.5% and 98.8% of the total concentration, respectively, which provides evidence for storage of inorganic selenium to favor SeVI. The lack of SeIV observed after SeVI supplementation suggests a passive induction of SeIV into the C. comosum root after SeIV supplementation instead of through a reduction pathway in the plant root. The findings suggest a direct absorption of selenium if the C. comosum is supplemented with SeVI, but a combination of passive and direct absorption of selenium if the C. comosum is supplemented with SeIV.

In the root extract of AsIII supplemented plants, AsIII and AsV made up 82.1% and 17.9% of the total concentration, respectively. These data suggest that the oxidation of the arsenic

82 species from AsIII to AsV may occur in the rhizosphere and subsequently be reduced to AsIII after absorption through the phosphate pathway in plant root. A prior study showed considerable amounts of AsIII found in L. esculentum (tomato), Z. mays (corn), P. sativum (pea) and C. melo

(melon) after supplementation with AsV.35 As an alternative metabolic pathway, AsIII may be passively absorbed in the root with subsequent partial oxidation to AsV. Previous work has shown that AsIII oxidation and AsV reduction can occur in plant roots.36 Although past studies have shown the production of phytochelatins as a means of arsenic detoxification within a plant,37 the C. comosum utilizes an alternate detoxification pathway. However, the production of phytochelatins may facilitate arsenic transport to the vacuole for storage in plant cells, as previously shown during a plant’s heavy metal detoxification process.38

The predominant species observed in root extract of the SeIV & AsIII supplemented plants

VI III IV V were Se , As and to a lesser extent, Se , SeMet, SeCys2 and As . The major metabolites detected in the root extract from the SeVI & AsIII supplemented C. comosum were SeVI and AsIII

V VI with As as a minor species. For selenium species, the overall concentration of Se was similar in the plants supplemented with SeIV compared with the SeIV & AsIII supplementation at 74.6% of total selenium concentration. A similar trend was noted for SeVI supplemented plants compared with SeVI & AsIII supplementation. However, the overall concentration of SeIV was reduced by more than half in plants supplemented with SeIV compared with SeIV & AsIII supplementation at 6.6% and 14.3%, respectively, of the total extracted selenium concentration.

These results suggest a greater restriction on the passive absorption of SeIV in the roots of the C. comosum than the direct absorption of SeVI, which may have been caused by an interaction with arsenic in the rhizosphere. For arsenic species, the overall concentration set as a ratio of AsV /

AsIII yielded 21.9% for AsIII supplemented plants, but 7.2% and 8.6% for AsIII & SeIV and AsIII

& SeVI supplemented plants, respectively. The observed loss of AsV suggests the metabolic pathway used by the C. comosum for arsenic absorption and metabolism. If AsIII was oxidized in the rhizosphere to AsV, then subsequently absorbed directly through the phosphate pathway and subsequently reduced to AsIII, a decrease in the arsenic concentration absorbed from the simultaneous addition of selenium should decrease the amount of AsIII observed. Since the contrary was found, the supplemented form of arsenic, AsIII, is suggested to be absorbed passively as AsIII and partially oxidized to AsV in the plant root.

3.4.3 Leaf extract characterization of selenium and arsenic species

In order to monitor the selenium and arsenic species after translocation and possible further metabolism in the leaf compartment, 100 µl from the Tris-HCl extraction of the C. comosum leaves were injected into the SEC-ICPMS and the resulting chromatograms are depicted in Figure 3.4. As noted in the chromatograms from the root extract, the major selenium and arsenic species in the leaf extract eluted after the 1.4 kDa standard, thus depicting small molecules such as peptides or inorganic species. Upon observing the selenium and arsenic chromatographic profile similarities and the decrease in elemental abundance from root to leaf regardless of supplementation type, it is suggested that compounds metabolized in the C. comosum roots are not readily translocated nor further metabolized in the leaves, which supports the earlier total elemental analysis results.

However, an exception was observed for plants supplemented with SeVI. In contrast to observations made from the plant root extracts, high molecular weight species were more prevalent in the leaves of plants supplemented with SeVI than SeIV indicating an alteration in the selenium metabolism. A reason may be simply due to the increased solubility of SeVI versus

Figure 3.4 78Se and 75As SEC-ICPMS chromatograms of the leaf extract from the C. comosum after varying supplementation combinations.

SeIV, which allows for greater mobility resulting in an increased rate of translocation. In addition, the lack of a void volume peak in the selenium plant profile when supplemented with

SeIV indicates sequestering high molecular weight selenium species (greater than 12 kDa) in the roots of the C. comosum.

In comparing the selenium and arsenic profiles of the plant leaves supplemented singly versus simultaneously with selenium and arsenic, several similar peaks were observed. While the overall concentration of selenium and arsenic was reduced during simultaneous supplementation, the chromatographic peak profile consistency illustrates that the metabolic

86 pathway remained predominantly unaffected. After further investigation of the selenium and arsenic profiles, a lack of chromatographic peak overlap reveals that a selenium and arsenic containing molecule was not present in the plant leaves regardless of the supplementation administered. Considering the low concentration and general conservation of translocated selenium and arsenic species in the C. comosum leaves, IPRP-ICPMS was not performed. A summary of the proposed metabolic pathways after arsenic or selenium supplementation in C. comosum can be found in Figure 3.5. Future studies will work towards a universal model by elucidating the metabolism of selenium and arsenic in other non-hyperaccumlating plants.

3.5 Acknowledgements

The author would like to acknowledge Pam Bishop (Rieveschl Green House, University of Cincinnati) for assistance in C. comosum cultivation, support from the NIEHS-SBRP grant

ES04908 and Agilent Technologies and the CEM Corporation for their instrumentation and continuing support.

3.6 References

1. Bhattacharya, P.; Welch, A. H.; Stollenwerk, K. G.; McLaughlin, M. J.; Bundschuh, J.; Panaullah, G., Arsenic in the environment: Biology and Chemistry. Science of the Total Environment 2007, 379, (2-3), 109-120. 2. Wang, S.; Mulligan, C. N., Occurrence of arsenic contamination in Canada: Sources, behavior and distribution. Science of the Total Environment 2006, 366, (2-3), 701-721. 3. Ma, L. Q.; Komar, K. M.; Tu, C.; Zhang, W.; Cai, Y.; Kennelley, E. D., A fern that hyperaccumulates arsenic. Nature (London, United Kingdom) 2001, 411, (6836), 438. 4. Kopsell, D. A.; Kopsell, D. E., Selenium. Handbook of Plant Nutrition 2007, 515-549. 5. Combs, G. F., Jr.; Clark, L. C.; Turnbull, B. W., An analysis of cancer prevention by selenium. BioFactors 2001, 14, (1-4), 153-159. 6. Brown, K. M.; Arthur, J. R., Selenium, selenoproteins and human health: a review. Public Health Nutr FIELD Full Journal Title:Public health nutrition 2001, 4, (2B), 593-9. 7. Terry, N.; Zayed, A. M.; De Souza, M. P.; Tarun, A. S., Selenium in higher plants. Annual Review of Plant Physiology and Plant Molecular Biology 2000, 51, 401-432. 8. He, P. P.; Lu, X. Z.; Wang, G. Y., Effects of Se and Zn supplementation on the antagonism against Pb and Cd in vegetables. Environment International 2004, 30, (2), 167-172. 9. Dubois, K. P.; Moxon, A. L.; Olson, O. E., Further studies on the effectiveness of arsenic in preventing selenium poisoning. Journal of Nutrition 1940, 19, 477-82. 10. Gailer, J.; George, G. N.; Pickering, I. J.; Prince, R. C.; Ringwald, S. C.; Pemberton, J. E.; Glass, R. S.; Younis, H. S.; DeYoung, D. W.; Aposhian, H. V., A Metabolic Link between Arsenite and Selenite: The Seleno-bis(S-glutathionyl) Arsinium Ion. Journal of the American Chemical Society 2000, 122, (19), 4637-4639. 11. Fayiga, A. O.; Ma, L. Q.; Rathinasabapathi, B., Effects of nutrients on arsenic accumulation by arsenic hyperaccumulator Pteris vittata L. Environmental and Experimental Botany 2008, 62, (3), 231-237. 12. Freeman, J. L.; Zhang, L. H.; Marcus, M. A.; Fakra, S.; McGrath, S. P.; Pilon-Smits, E. A. H., Spatial imaging, speciation, and quantification of selenium in the hyperaccumulator plants Astragalus bisulcatus and Stanleya pinnata. Plant Physiology 2006, 142, (1), 124-134. 13. Mounicou, S.; Vonderheide, A. P.; Shann, J. R.; Caruso, J. A., Comparing a selenium accumulator plant (Brassica juncea) to a nonaccumulator plant (Helianthus annuus) to investigate selenium-containing proteins. Analytical and Bioanalytical Chemistry 2006, 386, (5), 1367-1378. 14. Navaza, A. P.; Montes-Bayon, M.; LeDuc, D. L.; Terry, N.; Sanz-Medel, A., Study of phytochelatins and other related thiols as complexing biomolecules of As and Cd in wild type and genetically modified Brassica juncea plants. Journal of Mass Spectrometry 2006, 41, (3), 323-331. 15. Kapolna, E.; Shah, M.; Caruso, J. A.; Fodor, P., Selenium speciation studies in Se- enriched chives (Allium schoenoprasum) by HPLC-ICP-MS. Food Chemistry 2006, 101, (4), 1398-1406. 16. Li, B.; Bergmann, J.; Lassen, S.; Leonhard, P.; Prange, A., Distribution of elements binding to molecules with different molecular weights in aqueous extract of Antarctic krill by size-exclusion chromatography coupled with inductively coupled plasma mass spectrometry. Journal of Chromatography, B: Analytical Technologies in the Biomedical and Life Sciences 2005, 814, (1), 83-91.

17. B'Hymer, C.; Caruso, J. A., Arsenic and its speciation analysis using high-performance liquid chromatography and inductively coupled plasma mass spectrometry. Journal of Chromatography, A 2004, 1045, (1-2), 1-13. 18. B'Hymer, C.; Caruso, J. A., Selenium speciation analysis using inductively coupled plasma-mass spectrometry. Journal of Chromatography, A 2006, 1114, (1), 1-20. 19. Bujdos, M.; Mulova, A.; Kubova, J.; Medved, J., Selenium fractionation and speciation in rocks, soils, waters and plants in polluted surface mine environment. Environmental Geology (Berlin, Germany) 2005, 47, (3), 353-360. 20. Afton, S.; Kubachka, K.; Catron, B.; Caruso, J. A., Simultaneous characterization of selenium and arsenic analytes via ion-pairing reversed phase chromatography with inductively coupled plasma and electrospray ionization ion trap mass spectrometry for detection Applications to river water, plant extract and urine matrices. Journal of Chromatography A 2008, In Press. 21. Bluemlein, K.; Raab, A.; Meharg, A. A.; Charnock, J. M.; Feldmann, J., Can we trust mass spectrometry for determination of arsenic peptides in plants: comparison of LC-ICP-MS and LC-ES-MS/ICP-MS with XANES/EXAFS in analysis of Thunbergia alata. Analytical and Bioanalytical Chemistry 2008, 390, (7), 1739-1751. 22. Pedrero, Z.; Elvira, D.; Camara, C.; Madrid, Y., Selenium transformation studies during Broccoli (Brassica oleracea) growing process by liquid chromatography-inductively coupled plasma mass spectrometry (LC-ICP-MS). Analytica Chimica Acta 2007, 596, (2), 251-256. 23. Mounicou, S.; Shah, M.; Meija, J.; Caruso, J. A.; Vonderheide, A. P.; Shann, J., Localization and speciation of selenium and mercury in Brassica juncea-implications for Se-Hg antagonism. Journal of Analytical Atomic Spectrometry 2006, 21, (4), 404-412. 24. Yathavakilla, S. K. V.; Caruso, J. A., A study of Se-Hg antagonism in Glycine max (soybean) roots by size exclusion and reversed phase HPLC-ICPMS. Analytical and Bioanalytical Chemistry 2007, 389, (3), 715-723. 25. Iwaoka, M.; Ooka, R.; Nakazato, T.; Yoshida, S.; Oishi, S., Synthesis of selenocysteine and selenomethionine derivatives from sulfur-containing amino acids. Chemistry & Biodiversity 2008, 5, (3), 359-374. 26. Hoagland, D. R.; Arnon, D. I., Water-culture method for growing plants without soil. Calif. Agr. Expt. Sta., Circ. 1938, 347, 1-39. 27. Pickering, I. J.; Prince, R. C.; George, M. J.; Smith, R. D.; George, G. N.; Salt, D. E., Reduction and coordination of arsenic in Indian mustard. Plant Physiology 2000, 122, (4), 1171- 1177. 28. Shrift, A., Aspects of selenium metabolism in higher plants. Annual Review of Plant Physiology 1969, 20, 475-94. 29. Nelson, D. C.; Casey, W. H.; Sison, J. D.; Mack, E. E.; Ahmad, A.; Pollack, J. S., Selenium uptake by sulfur-accumulating bacteria. Geochimica et Cosmochimica Acta 1996, 60, (18), 3531-3539. 30. Zehr, J. P.; Oremland, R. S., Reduction of selenate to selenide by sulfate-respiring bacteria: experiments with cell suspensions and estuarine sediments. Applied and Environmental Microbiology 1987, 53, (6), 1365-9. 31. Stolz, J. F.; Oremland, R. S., Bacterial respiration of arsenic and selenium. FEMS Microbiology Reviews 1999, 23, (5), 615-627. 32. Wenzel, W. W.; Lombi, E.; Adriano, D. C., Biogeochemical processes in the rhizosphere: role in phytoremediation of metal-polluted soils. Heavy Metal Stress in Plants 1999, 273-303.

33. Hedley, M. J.; Nye, P. H.; White, R. E., Plant-induced changes in the rhizosphere of rape (Brassica napus var. Emerald) seedlings. II. Origin of the pH change. New Phytologist 1982, 91, (1), 31-44. 34. Nye, P. H., Changes of pH across the rhizosphere induced by roots. Plant and Soil 1981, 61, (1-2), 7-26. 35. Nissen, P.; Benson, A. A., Arsenic metabolism in freshwater and terrestrial plants. Physiologia Plantarum 1982, 54, (4), 446-50. 36. Tu, S.; Ma, L. Q.; MacDonald, G. E.; Bondada, B., Effects of arsenic species and phosphorus on arsenic absorption, arsenate reduction and thiol formation in excised parts of Pteris vittata L. Environmental and Experimental Botany 2004, 51, (2), 121-131. 37. Schulz, H.; Haertling, S.; Tanneberg, H., The identification and quantification of arsenic- induced phytochelatins - comparison between plants with varying As sensitivities. Plant and Soil 2008, 303, (1-2), 275-287. 38. Shaw, B. P.; Prasad, M. N. V.; Jha, V. K.; Sahu, B. B., Detoxification/defense mechanisms in metal-exposed plants. Trace Elements in the Environment 2006, 291-324.

Chapter 4

Investigating the mechanisms and feasibility of xenon as a cell gas in inductively coupled plasma mass spectrometry for multi-element detection including sulfur with application to bottled water

4.1 Abstract

Previous studies have shown Xe as an effective collision/reaction cell (CRC) gas for attenuating the interferences that previously prohibited sulfur detection in inductively coupled plasma mass spectrometry (ICPMS). However, the practical feasibility for multi-element detection upon addition of Xe CRC gas has not been extensively tested. In this study, the optimized Xe CRC gas flow rate for multi-element monitoring, including sulfur, is obtained via the 34S+ signal. Increasing the difference between the voltage on the extraction lens and the octopole, plus using a positive voltage between the octopole and quadrupole provided the lowest sulfur detection limits. When monitoring a standard solution comprised of 52 elements, including sulfur, the range of optimal octopole and quadrupole bias settings were altered when compared with single element sulfur detection. Similar detection limits were acquired when comparing Xe to He or H2 CRC gas with a general trend comparable to detection levels for He.

32 + 32 + Energy discrimination is the predominant mechanism for removing the O2 interference of S , rather than charge transfer or collision-induced dissociation. Xenon is effective in CRC-ICPMS for quantifying elements in commercially available bottled water.

4.2 Introduction

Sulfur detection has historically provided many complications using analytical instrumentation. The predominant sulfur isotope, 32S, lacks nuclear spin and is, therefore, not useful for nuclear magnetic resonance spectroscopy, while the much less abundant isotope, 33S, has a spin I = 3/2 but generally produces weak, broad signals. X-ray absorption near edge structure can be used to monitor sulfur using the K edge at 2472 eV; however, speciation of sulfur compounds have only reached detection limits of a few 100 µg g-1 in natural samples and routine analysis is quite costly.1 While traditionally considered a metals detector, new avenues of research utilizing inductively coupled plasma mass spectrometry (ICPMS) have included nonmetal detection such as iodine,2 selenium,3 phosphorus4 and sulfur5 for varying environmental and biological applications. The high ionization potential of sulfur, 10.36 eV, and

32 + 32 the predominant polyatomic interference, O2 , have caused significant barriers for viable S detection via ICPMS as shown in Table 4.1.

Table 4.1 Pertinent ionization potentials and isotopic distributions

IP (eV) 32 Da (%) 33 Da (%) 34 Da (%)

S 10.36 95.02 0.75 4.21

O2 12.07 99.53 0.08 0.40

O2H 12.62 99.51 0.09

NO 9.26 0.20

Xe 12.13

He 24.59

Traditionally, two fundamentally different approaches have been incorporated for sulfur detection by ICPMS. Past studies have employed O2 collision reaction cell (CRC) gas in order to successfully react with 32S+ causing a mass shift to 48SO+ without a parallel mass shift of the

32 + 48 + 5-7 major polyatomic interference O2 to O3 . However, the O2 addition may cause new isobaric interferences when multi-element detection mode is employed requiring careful consideration of the sample matrix. Using Xe as an alternative CRC gas has been attempted for sulfur detection with successful results.8, 9 The proposed interference removal mechanism in

32 + sulfur detection has generally been thought to occur through the charge transfer of O2 to the neutral Xe, thus eliminating the polyatomic interference.10 In contrast, an alternative mechanism

32 + 32 + involving the collision-induced dissociation (CID) of O2 by Xe, thereby allowing for S detection, has also been suggested.11 To date, an extensive study has not been conducted to fully optimize the instrumental parameters, such as Xe CRC gas flow rate, octopole bias and quadrupole bias settings, in order to provide the lowest limits of detection for sulfur.

In addition to single element sulfur analysis, the virtually simultaneous multi-element detection by quadrupole ICPMS can be utilized for multi-element analysis in a variety of matrices from low to high molecular weight. It has been estimated that one third of all mammalian enzymes contain a metal cofactor.12 The ability to contemporaneously monitor sulfur and biologically relevant metals would provide enhanced capabilities in studying metalloproteins, since bonding through sulfur is highly likely. Furthermore, the metal/sulfur ratio is useful as a screening tool for selecting altered proteins. ICPMS has been used to quantify many impurities in aqueous matrices such as ground water and bottled water.13, 14 The quantification of sulfur in a multi-element suite either has not been done or necessitated using the less available high resolution mass analyzer. The addition of sulfur in multi-element analysis via

94 a quadrupole mass analyzer will provide additional risk/benefit information on environmental contaminants.

When utilizing the multi-element capability of ICPMS, selection of a CRC gas to discriminate between matrix induced interferences and the desired analytes can be a complex process. Koppenaal et al. compiled a brief list of common CRC gases used in previous studies.15

The combination of reported potential problems with the use of highly reactive cell gases has led to the conclusion that no single reaction gas or gas mixture provides a reliable method of interference removal across a wide range of analytes and sample matrices.16 Due to the additional complications with using highly reactive gases for multi-element monitoring by

ICPMS, He, a chemically inert gas, has been shown as a possible universal CRC gas.16

However, the feasibility for multi-element detection including sulfur was not investigated.

In this study, the instrumental parameters such as Xe CRC gas flow rate, octopole bias and quadrupole bias, were optimized to provide the lowest limits of detection for sulfur. In addition, the optimization parameters for a solution comprised of 52 elements, including sulfur, are provided. A comparative detection limit study for 52 elements utilizing Xe, He and H2 was

32 + performed. A mechanism for O2 interference removal by Xe is suggested as well as for the discrimination of other isobaric interferences with selected analytes. Using Xe CRC gas to monitor elemental composition in commercially available bottled water was examined.

4.3 Experimental

4.3.1 Instrumentation

The ICPMS used for specific element detection was an Agilent 7500ce by Agilent

Technologies (Santa Clara, CA). The instrument was equipped with a microconcentric nebulizer

95 made by Glass Expansion (Pocasset, MA), a Scott double channel spray chamber (2°C), a shielded torch with a sampling depth of 7 mm, nickel sampling and skimmer cones, an octopole ion guide enclosed in a CRC pressurized with H2, He or Xe gas, purity of 99.999%, from

Matheson Gas Products (Parsippany, NJ), a quadrupole mass analyzer with a dwell time of 100 ms per isotope and an electron multiplier used for element detection. The octopole ion guide operating in only rf mode prevented signal losses due to scattering of both reactant and product ions whether formed through the classical inelastic collision or the recently demonstrated stop- and-go interaction.17 General instrumental parameters were as follows: forward power, 1500 W; plasma gas flow rate, 15.0 l min-1; auxiliary gas flow rate, 1.00 l min-1; carrier gas flow rate, 0.99 l min-1; makeup gas flow rate, 0.14 l min-1; extraction lens, 3.5 V. Instrument tuning parameters were optimized daily with a solution of Li, Y, Ce and Tl (1 ng ml-1) without CRC gas and additional experimental adjustments other than octopole bias, quadrupole bias or CRC gas flow rate were not made in order to ease instrument operation.

4.3.2 Reagents, standards and materials

All the solutions were prepared from water purified by a NanoPure (18 MΩ cm) doubly deionized water (DDW) system from Sybron Barnstead (Boston, MA). Metal free polypropylene tubes for sample preparation and storage were purchased from CPI International

(Santa Rosa, CA). Sodium sulfate (Na2SO4) was purchased from Sigma-Aldrich (St. Louis, MO) for CRC flow rate studies. Elemental standards used for quantification were Claritas PPT Multi- element Solution 2, 3, 4 and mercury purchased from Spex CertiPrep (Metuchen, NJ).

Calibration standards from 1.0 ng ml-1 to 500 ng ml-1 were prepared through dilution from a stock solution of mixed calibration standards. Four varying types of commonly available bottled

96 water were purchased and are represented as A, B, C and D, ordered by increasing cost. Three samplings were taken from each of the seven varying batches of bottled water from each manufacturer.

4.3.3 Analytical protocol

+ Several experiments were performed to elucidate the mechanism of O2 interference discrimination by Xe. Voltage discrimination studies were performed with a sulfur standard of

-1 250 µg ml Na2SO4 and a blank of DDW. Charge transfer studies were carried out using the introduction of DDW to the ICPMS in continuous flow mode. In both experiments, the flow rate was adjusted in increments of 0.005 ml min-1 from 0.05 ml min-1 to 0.25 ml min-1 for Xe and increments of 0.1 ml min-1 from 0.5 ml min-1 to 5.0 ml min-1 for He with the resulting signal plotted.

-1 Xe gas flow optimization was accomplished using a Na2SO4 standard (250 µg ml ) and

DDW as the blank. The flow rate was adjusted in increments of 0.005 ml min-1 from 0.05 ml min-1 to 0.35 ml min-1 and the resulting signal was plotted. The following octopole and quadrupole voltage (octopole voltage, quadrupole voltage) combinations were used to find a universal CRC gas flow rate: (-18,-16), (-16,-8), (-18,-1), (-34,-12), (-30,-1), (-50,-16), (-40,-1),

(-50,-6) and (-50,-1).

For multi-element analysis, the following 90 isotopes of 52 elements were monitored throughout the study: 6Li, 7Li, 9Be, 11B, 23Na, 24Mg, 26Mg, 27Al, 28Si, 29Si, 30Si, 31P, 32S, 33S, 34S,

39K, 42Ca, 43Ca, 44Ca, 46Ti, 47Ti, 49Ti, 51V, 52Cr, 53Cr, 55Mn, 56Fe, 57Fe, 59Co, 60Ni, 62Ni, 63Cu,

65Cu, 66Zn, 68Zn, 69Ga, 71Ga, 72Ge, 73Ge, 75As, 77Se, 78Se, 80Se, 85Rb, 86Sr, 88Sr, 90Zr, 91Zr, 93Nb,

95Mo, 97Mo, 99Ru, 101Ru, 103Rh, 105Pd, 107Ag, 108Pd, 109Ag, 111Cd, 115In, 118Sn, 119Sn, 121Sb, 123Sb,

125Te, 126Te, 133Cs, 137Ba, 138Ba, 177Hf, 178Hf, 181Ta, 182W, 183W, 184W, 185Re, 187Re, 191Ir, 193Ir,

194Pt, 195Pt, 197Au, 200Hg, 202Hg, 203Tl, 205Tl, 207Pb, 208Pb, 209Bi and 238U. The octopole and quadrupole bias settings were varied using the following combinations, in addition to those used in the CRC gas flow rate optimization: (-16,-12), (-8,-1), (-16,-4), (-34,-16), (-34,-8), (-34,-4) and

(-50,-12). Each experiment was run in triplicate and the detection limits were calculated using the IUPAC definition, which is three times the standard deviation of the blank divided by the slope of the calibration curve, via the ICPMS ChemStation software package.

For CRC gas comparison studies, all 16 octopole and quadrupole bias combinations with

-1 a CRC gas flow rate of 0.17 ml min were used for Xe. For He and H2, a common -18 V octopole bias, -16 V quadrupole bias (+2 V energy discrimination) and CRC gas flow rates of 3.5 and 4.5 ml min-1 were used. All the isotopes previously mentioned were monitored.

4.4 Results and discussion

+ 4.4.1 Studying the O2 removal mechanism by Xe

+ -1 To investigate the mechanism used by Xe for O2 interference removal, a 250 ng ml

Na2SO4 standard and DDW as a blank were used to measure m/z 32 after Xe or He CRC gas addition with a constant -16 V quadrupole bias and a varying octopole bias. The resulting plots

32 + are shown in Figure 4.1. The predominant m/z 32 interference in the blank (DDW) is O2 , which can be seen in Table 4.1. With the octopole bias set at -15 V, the net discrimination barrier between the octopole and quadrupole was -1 V attracting the positive ions, which allowed

32 + the investigation of charge transfer and CID of O2 with Xe and He. The lack of separation

32 + 32 + between the S standard signal and O2 background signal when using Xe CRC gas, suggests

+ that charge transfer and polyatomic dissociation are not the predominate mechanisms for O2

98 removal. As both mechanisms are independent of the energy barrier between the octopole and the quadrupole, a separation between the signal and blank should have been observed. In

+ addition, charge transfer and CID of O2 were not observed with He CRC gas.

+ 18 For O2 CID, 6.66 eV is needed. The collision energy obtained for polyatomic interference dissociation was calculated according Equation 1:11

푚푘 퐸푐푚 = ∗ 퐸푙푎푏 푚푘 + 푚푝

99 where Ecm represents the collision energy at the center of mass, mk symbolizes the CRC gas mass, mp stands for the target ion mass and Elab represents the kinetic energy resulting from the potential difference between the extraction lens voltage and octopole bias voltage. Low energy

19 + CID is commonly classified as less than 100 eV. Using Xe CRC gas (mk 131.29 Da), O2 as the target ion (mp 32.00 Da), an extraction lens setting of 3.5 V and an octopole bias at -15 V

(Elab of 18.5 V), yields an Ecm of 14.9 eV, which theoretically provides sufficient energy to

+ 11 dissociate the O2 interference if a collision with Xe occurs, previously reported, but not supported by the results given here.

Although Xe has been effectively shown in CID of diatomic molecules in past studies,20,

21 these data suggests an inefficient transfer of available kinetic energy into internal energy for bond dissociation. In agreement, studies of small, strongly bound inorganic ions have highlighted that the energy is inefficiently transferred from collisional kinetic energy to internal

22 energy of the ion. In comparison, using He CRC gas (mk 4.003 Da) and an octopole bias at -15

V (Elab of 18.5 V), yields an Ecm of 2.1 eV. Theoretically, He does not provide sufficient energy,

+ even with complete transfer of kinetic energy into internal energy, to dissociate the O2 interference. This is supported by these experimental results as noted in Figure 4.1 (lower He charts) by the lack of separation between background and standard.

To investigate the effect of a minor net energy discrimination barrier, the octopole bias was adjusted to -18 V resulting in a net +2 V between the octopole and quadrupole (Figure 4.1 -

32 + 32 + center). This resulted in an increase in separation between the S standard signal and O2 background signal when using Xe CRC gas, therefore, increased sensitivity. On the contrary, the

32 + 32 + S standard signal and O2 background signal overlap when using He CRC gas. Further, by adjusting the octopole bias to -50 V, yielding +34 V between the octopole and quadrupole as

100

32 + 32 + shown in Figure 4.1 (right), a distinct difference between the S standard signal and O2 background signal is seen using Xe CRC gas. As observed for all voltage adjustments, the 32S+

32 + standard signal and O2 background signal overlap when using He CRC gas.

32 + Although CID of O2 is not the primary mode of interference removal, the increased kinetic energy of Xe compared with He, may explain the difference in the plots shown in Figure

4.1. Using Equation 1, a change in octopole bias from -18 V to -50 V causes an alteration in the

Ecm value from 17.3 eV to 43.0 eV using Xe CRC gas and 2.4 eV to 5.9 eV using He CRC gas.

+ The data suggest that the primary mode for O2 interference removal by Xe is attributed to energy discrimination brought about by an increase in the difference between the octopole and quadrupole bias voltages, shown by the increased collisional kinetic energy of Xe when compared with He.

+ 10 The O2 interference removal by charge transfer to Xe was previously investigated.

Experimentally, 4He+ and 129Xe+ signals were monitored upon addition of each individual gas into the CRC with the ICPMS in continuous flow mode and passing DDW through the nebulizer.

The resulting plots are shown in Figure 4.2. While the 4He plot does not display an increase in signal upon He introduction in the CRC, Xe addition shows a distinct increase in the 129Xe signal. This supports previous indications that Xe possesses charge transfer capabilities.10

However, prior studies also have shown that effective charge transfer occurs from a species possessing a higher ionization potential to a species possessing a lower ionization potential.23

The ionization potentials listed in Table 4.1 depict that charge transfer should not readily take

+ + place from O2 (12.07 eV) to Xe (12.13 eV), but instead from the highly abundant Ar (15.7 eV) to Xe (12.13 eV). Furthermore, the ionization potential for He (24.5 eV) does not allow for

101

+ + charge transfer from either O2 or Ar . These suggestions agree with the experimental results in

+ + Figure 4.2 and suggest that charge transfer exists from Ar to Xe, but not readily from O2 to Xe.

Figure 4.2 A graph overlaying the 4He and 129Xe responses after percent stepwise addition of He (20 ml min-1) and Xe (1 ml min-1) CRC gas, respectively.

4.4.2 Sulfur optimization via ICPMS

-1 To optimize the Xe CRC gas flow, a 250 µg ml Na2SO4 standard was used and the resulting signal, blank (DDW) and background equivalent concentration (BEC) plots are shown in Figure 4.3. General signal attenuation was observed upon increasing the CRC gas flow as expected. A first attempt was made to locate the optimized flow rate for Xe CRC gas via the

BEC minimum. The BEC minimum for the 32S+ plot is at 0.305 ml min-1. While plausible for monitoring sulfur alone, the signal attenuation will significantly raise the detection limits of other elements, thereby compromising multi-element analysis possibilities. When viewing the BEC

102 plot for 33S+, the chaotic behavior is attributed to the minimal difference between the sulfur standard and blank signals, which ultimately resulted in an undefined optimized Xe flow rate.

Although the 34S abundance (4.21%) is significantly lower than the 32S abundance

(95.02%), depicted in Table 4.1, the 34S+ plot was the most advantageous for Xe gas flow rate optimization. The BEC for 34S+ showed a net improvement of three orders of magnitude lower than 33S+ and three times lower than 32S+. The minimum BEC is found at 0.23 ml min-1 and 0.32

103 ml min-1; however, neither flow rate was selected due to the background instability, which resulted in the chaotic fluctuation of the BEC observed. The IUPAC definition for the detection limit of an analyte is equal to three times the standard deviation of the background signal divided by the slope of the calibration curve. Therefore, increased background fluctuation will result in an increased detection limit. The 34S+ plot inset displays the blank signal where the optimal flow rate of 0.17 ml min-1 was selected before significant background signal fluctuation and was reproducible among voltage settings with a standard deviation of ~ 0.01 ml min-1. This flow rate was then used for multi-element optimization including sulfur via octopole and quadrupole voltage bias adjustment.

Figure 4.4 A contour plot of detection limits for 32S as a function of the alteration in quadrupole and octopole bias.

104

For optimization of the net voltage for an energy discrimination barrier needed to achieve the lowest detection limit for sulfur, a contour plot of 32S+ with differing octopole and quadrupole bias settings depicting the change in detection limits can be found in Figure 4.4.

Overall, the detection limits ranged over ca. one order of magnitude. The unique optimization range for 32S, which provided the lowest detection limits, was obtained with an octopole bias setting below -37 V and the difference between the octopole and quadrupole bias, of ca. +28 V to

+42 V.

These relatively extreme voltages, allowing or discriminating for both the analyte and interference into the mass analyzer, respectively, are likely responsible for the 32S+ detection limits. The highly negative octopole bias voltage needed for lower 32S+ detection limits may be explained using Equation 1. The range of octopole bias settings used varied from -18 V to -50

V yielding an Ecm of 17.3 eV and 43.0 eV, respectively. The increased kinetic energy at the center of mass allows for greater energy attenuation during ion-neutral collisions in the CRC.

4.4.3 Multi-element detection via sulfur optimization

Instrument optimization for multi-element analysis including sulfur with Xe CRC gas was performed via octopole and quadrupole bias adjustments and the resulting contour plot can be found in Figure 4.5. In order to best represent the 4,320 individual detection limits obtained from the differing isotopic masses for the elements monitored, a single isotope of each element giving the lowest detection limit for that element was selected. The 52 isotopes represented are the following: 7Li, 9Be, 11B, 23Na, 24Mg, 27Al, 28Si, 31P, 32S, 39K, 44Ca, 47Ti, 51V, 52Cr, 55Mn, 56Fe,

59Co, 60Ni, 63Cu, 66Zn, 69Ga, 72Ge, 75As, 78Se, 85Rb, 88Sr, 90Zr, 93Nb, 95Mo, 101Ru, 103Rh, 105Pd,

105

107Ag, 111Cd, 115In, 118Sn, 121Sb, 125Te, 133Cs, 138Ba, 178Hf, 181Ta, 182W, 185Re, 193Ir, 195Pt, 197Au,

202Hg, 205Tl, 208Pb, 209Bi and 238U.

Each element was included in the overall percentage of elements listed in Figure 4.5 if the element’s detection limit was within 3x the lowest detection limit obtained for that element using all 16 unique octopole and quadrupole bias combinations. For example, the lowest detection limit obtained for 32S using all 16 octopole and quadrupole bias combinations was 4.39 ng ml-1. For a particular octopole and quadrupole bias combination, the detection limit for 32S

106 would have to be less than 13.17 ng ml-1 in order to be included in the overall percentage at that setting. Then, as indicated above for each octopole and quadrupole bias combination, the percentage of elements exhibiting a detection limit within 3x of the lowest detection limit obtained for each element is shown. The value of 3x was selected for graph presentation clarity and statistically yielded 4.11 σ when averaging all of the standard deviations from triplicate analysis for each of the 52 elements at each octopole and quadrupole bias combination.

In general, a net gain of an order of magnitude was acquired by using settings within the optimized octopole and quadrupole bias region. As compared with exclusive sulfur monitoring using Xe (Figure 4.4), the optimization location varied for multi-element detection. The observed results are attributed to a number of factors including inter-element interferences, ionization suppression from a high concentration of easily ionized elements, increased space- charge effects due to the presence of a high concentration of high mass ions in the extracted ion beam, and/or interference from polyatomic overlap.16

While previous studies using organic analytes have shown that the conversion or transfer efficiency of available kinetic energy into internal energy for bond dissociation increases with the size of the ion,24, 25 a general trend was not found relating analyte mass to the octopole and quadrupole bias combination providing the lowest limits of detection. This is attributed to the unique interferences for the varying isotopic masses of the elements monitored. Identifying the origin for the unique pattern of each element is not yet explained, given the sample matrix complexity.

107

4.4.4 Evaluation of Xe versus He or H2 as a CRC gas

In order to test the general application of Xe as a universal CRC gas for multi-element analysis, a comparison was performed with common CRC gases, He and H2, and the resulting bar graph is shown in Figure 4.6. From the 90 isotopes of the 52 elements tested, only the isotopes possessing the lowest detection limits were used in this figure and a summary is given in the prior section describing Figure 4.5. Figure 4.6 is comprised of elements which possessed a difference in detection limits greater than 2.6x the lowest detection limit obtained per element when using Xe vs. He or H2 as the CRC gas. The value of 2.6x was selected for graph presentation clarity and statistically yielded 3.56 σ when averaging all of the standard deviations from triplicate analysis for each of the 52 elements at each octopole and quadrupole bias

108 combination when using Xe CRC gas. It should be noted if 3x (4.11 σ) was used as in Figure

4.5, only 39K would not be included in Figure 4.6. In general, the detection limits obtained after using Xe CRC gas closely reflected the results acquired using He CRC gas with none of the 52 elements possessing a detection limit greater than 2.6x when comparing Xe to He. Only the

44 66 78 125 detection limits for Ca, Zn, Se and Te were greater than 2.6x when comparing Xe to H2 as a CRC gas. These results demonstrate the general efficacy of Xe as a CRC gas for multi- element applications.

Due to an isobaric overlap of 40Ar, 99.60% isotopic abundance, with 40Ca, 96.94% isotopic abundance, 44Ca was monitored. However, the higher detection limit observed for 44Ca when using Xe CRC gas is attributed to the low isotopic abundance of 2.09%. The increased dissociation energy provided by Xe, as compared with the decreased atomic size of He, attenuates the analyte and background signal to a greater extent, resulting in the observed

44 decrease in signal to noise. The lowest detection limits for Ca resulted when using H2 CRC gas. This finding is attributed the effective attenuation of the polyatomic interference 44ArHe,

99.60% isotopic abundance. The general ability for effective discrimination of polyatomic

26, 27 interferences formed from the Ar plasma by H2 CRC gas is also noted in prior studies.

The elevated 66Zn detection limit when using Xe CRC gas is attributed to the lower isotopic abundance of 27.90% and the Ar plasma formed polyatomic interference 66MgAr. For

78 Se, a marked improvement in the detection limit was observed after the use of H2 CRC gas when compared with He or Xe. Past studies have noted this finding and is attributed to the

78 28 effective reduction in the polyatomic interference Ar2, 0.125% isotopic abundance. The elevated detection limit obtained for 125Te when using Xe CRC gas is attributed to the lower isotopic abundance of 7.07% and the isobaric overlap of 125XeH, 0.10% isotopic abundance. In

109 summary, the resulting increased dissociation energy of Xe, when compared with He or H2, increased the detection limits of monitored isotopes with decreased natural abundance.

39 28 32 Elements K, Si and S resulted in lower detection limits when Xe vs. He or H2 CRC

39 gas was used. The proposed mechanism for the observed detection limits of K with H2 CRC gas is the effective attenuation the Ar plasma formed polyatomic interference 39ArH, 0.063% isotopic abundance. In contrast, 39ArH discrimination, when using Xe CRC gas, is attributed to the ability for increased collision induced energy attenuation as compared with He CRC gas.

This can be shown by the increased dissociation energy of ArH, 6.16 eV,29 which is similar in magnitude to the dissociation energy of O2, suggesting a similar interference removal mechanism as discussed previously. While it has been shown that CID is not the predominant mode for O2 removal, the effect of adding a higher mass collision gas has been studied, and has been shown to increase the deposited internal energy,20, 30 as predicted by Equation 1.

The differences between the detection limits observed for 28Si when using Xe or He CRC gas can also be explained utilizing an explanation similar to that given above. The predominant

28 28 Si interference, N2, possess an isotopic abundance of 99.27%, a dissociation energy of 9.90

31 eV and an ionization potential of 15.8 eV. The high dissociation energy of N2 is more effectively attenuated by Xe due to the increased collision energy when compared with He. In addition, the ionization potential of N2 allows for charge transfer with Xe, which further reduces the interfering signal. The greatest difference in detection limits for the 52 elements tested was observed with 32S when using Xe CRC gas, which resulted in the lowest detection limit. While the BEC for 32S is increased versus 34S (Figure 4.3), lower detection limits are generally observed for 32S, which is attributed to the increased natural abundance.

110

4.4.5 Quantification of bottled water impurities

Figure 4.7 A bar graph displaying the elemental profile concentration found in bottled water from four major manufacturers represented as A-D (lowest to highest priced).

In order to test the applicability of Xe in CRC-ICPMS for a wide range of elements and concentrations, the elemental concentration profiles found in commonly available bottled water from four major manufacturers is reported and the resulting bar graph is shown in Figure 4.7.

All of the elements represented in the bar graph were present in at least one of the bottled water brands with a concentration greater than 50 ng ml-1. The concentrations obtained from the

ICPMS had an instrumental standard deviation of less than 10% in general for concentrations over 5 ng ml-1. The elemental reproducibility between bottled water batches varied from 1.7% to

111

522% with the reproducibility generally inversely proportional to the calculated concentration.

Another trend derived from the bar graph equated the elemental profile concentration present in the bottled water samples to be inversely proportional to the bottled water cost. Without the use of Xe CRC gas, the quantification of sulfur would not have been possible for the bottled water from each manufacturer. With Xe, it was possible to sulfur and other nonmetals in conjunction with metalloids and metals using the same ICPMS conditions.

4.5 Acknowledgements

The author would like to acknowledge support from the University of Cincinnati

Metallomics Center of the Americas and Agilent Technologies for their instrumentation and continuing support.

112

4.6 References

1. Jalilehvand, F., Sulfur: not a \"silent\" element any more. Chemical Society Reviews 2006, 35, (12), 1256-1268. 2. Shah, M.; Wuilloud, R. G.; Kannamkumarath, S. S.; Caruso, J. A., Iodine speciation studies in commercially available seaweed by coupling different chromatographic techniques with UV and ICP-MS detection. Journal of Analytical Atomic Spectrometry 2005, 20, (3), 176- 182. 3. B'Hymer, C.; Caruso, J. A., Selenium speciation analysis using inductively coupled plasma-mass spectrometry. Journal of Chromatography, A 2006, 1114, (1), 1-20. 4. Shah, M.; Caruso, J. A., Inductively coupled plasma mass spectrometry in separation techniques: Recent trends in phosphorus speciation. Journal of Separation Science 2005, 28, (15), 1969-1984. 5. Wang, M.; Feng, W.; Lu, W.; Li, B.; Wang, B.; Zhu, M.; Wang, Y.; Yuan, H.; Zhao, Y.; Chai, Z., Quantitative Analysis of Proteins via Sulfur Determination by HPLC Coupled to Isotope Dilution ICPMS with a Hexapole Collision Cell. Analytical Chemistry (Washington, DC, United States) 2007, 79, (23), 9128-9134. 6. Divjak, B.; Goessler, W., Ion chromatographic separation of sulfur-containing inorganic anions with an ICP-MS as element-specific detector. Journal of Chromatography, A 1999, 844, (1 + 2), 161-169. 7. Jensen, B. P.; Smith, C.; Wilson, I. D.; Weidolf, L., Sensitive sulphur-specific detection of omeprazole metabolites in rat urine by high-performance liquid chromatography/inductively coupled plasma mass spectrometry. Rapid Communications in Mass Spectrometry 2004, 18, (2), 181-183. 8. Ellis, J. L.; Conklin, S. D.; Gallawa, C. M.; Kubachka, K. M.; Young, A. R.; Creed, P. A.; Caruso, J. A.; Creed, J. T., Complementary molecular and elemental detection of speciated thioarsenicals using ESI-MS in combination with a xenon-based collision-cell ICP-MS with application to fortified NIST freeze-dried urine. Analytical and Bioanalytical Chemistry 2008, 390, (7), 1731-1737. 9. Schaumloeffel, D.; Giusti, P.; Preud'Homme, H.; Szpunar, J.; Lobinski, R., Precolumn Isotope Dilution Analysis in nanoHPLC-ICPMS for Absolute Quantification of Sulfur- Containing Peptides. Analytical Chemistry (Washington, DC, United States) 2007, 79, (7), 2859- 2868. 10. Mason, P. R. D.; Kaspers, K.; van Bergen, M. J., Determination of sulfur isotope ratios and concentrations in water samples using ICP-MS incorporating hexapole ion optics. Journal of Analytical Atomic Spectrometry 1999, 14, (7), 1067-1074. 11. Proefrock, D.; Leonhard, P.; Prange, A., Determination of sulfur and selected trace elements in metallothionein-like proteins using capillary electrophoresis hyphenated to inductively coupled plasma mass spectrometry with an octopole reaction cell. Analytical and Bioanalytical Chemistry 2003, 377, (1), 132-139. 12. Hann, S.; Koellensperger, G.; Obinger, C.; Furtmueller, P. G.; Stingeder, G., SEC-ICP- DRCMS and SEC-ICP-SFMS for determination of metal-sulfur ratios in metalloproteins. Journal of Analytical Atomic Spectrometry 2004, 19, (1), 74-79. 13. Cheng, Z.; Zheng, Y.; Mortlock, R.; van Geen, A., Rapid multi-element analysis of groundwater by high-resolution inductively coupled plasma mass spectrometry. Analytical and Bioanalytical Chemistry 2004, 379, (3), 512-518.

113

14. Momani, K. A., Chemical assessment of bottled drinking waters by IC, GC, and ICP-MS. Instrumentation Science & Technology 2006, 34, (5), 587-605. 15. Koppenaal, D. W.; Eiden, G. C.; Barinaga, C. J., Collision and reaction cells in atomic mass spectrometry: development, status, and applications. Journal of Analytical Atomic Spectrometry 2004, 19, (5), 561-570. 16. McCurdy, E.; Woods, G., The application of collision/reaction cell inductively coupled plasma mass spectrometry to multi-element analysis in variable sample matrices, using He as a non-reactive cell gas. Journal of Analytical Atomic Spectrometry 2004, 19, (5), 607-615. 17. Greaves, S. J.; Wrede, E.; Goldberg, N. T.; Zhang, J.; Miller, D. J.; Zare, R. N., Vibrational excitation through tug-of-war inelastic collisions. Nature (London, United Kingdom) 2008, 454, (7200), 88-91. 18. Nonose, N. S.; Matsuda, N.; Fudagawa, N.; Kubota, M., Some characteristics of polyatomic ion spectra in inductively coupled plasma mass spectrometry. Spectrochimica Acta, Part B: Atomic Spectroscopy 1994, 49B, (10), 955-74. 19. Wells, J. M.; McLuckey, S. A., Collision-induced dissociation (CID) of peptides and proteins. Methods in Enzymology 2005, 402, (Biological Mass Spectrometry), 148-185. 20. Aristov, N.; Armentrout, P. B., Collision-induced dissociation of vanadium monoxide ion. Journal of Physical Chemistry 1986, 90, (21), 5135-40. 21. Hales, D. A.; Armentrout, P. B., Effect of internal excitation on the collision-induced dissociation and reactivity of the cobalt diatomic monopositive ion. Journal of Cluster Science 1990, 1, (1), 127-42. 22. Armentrout, P. B., Threshold collision-induced dissociations for the determination of accurate gas-phase binding energies and reaction barriers. Topics in Current Chemistry 2003, 225, (Modern Mass Spectrometry), 233-262. 23. Olesik, J. W.; Jones, D. R., Strategies to develop methods using ion-molecule reactions in a quadrupole reaction cell to overcome spectral overlaps in inductively coupled plasma mass spectrometry. Journal of Analytical Atomic Spectrometry 2006, 21, (2), 141-159. 24. Douglas, D. J., Applications of collision dynamics in quadrupole mass spectrometry. Journal of the American Society for Mass Spectrometry 1998, 9, (2), 101-113. 25. Marzluff, E. M.; Campbell, S.; Rodgers, M. T.; Beauchamp, J. L., Collisional Activation of Large Molecules Is an Efficient Process. Journal of the American Chemical Society 1994, 116, (15), 6947-8. 26. Boulyga, S. F.; Becker, J. S., ICP-MS with hexapole collision cell for isotope ratio measurements of Ca, Fe, and Se. Fresenius J Anal Chem FIELD Full Journal Title:Fresenius' journal of analytical chemistry 2001, 370, (5), 618-23. 27. Eiden, G. C.; Barinaga, C. J.; Koppenaal, D. W., Selective removal of plasma matrix ions in plasma source mass spectrometry. Journal of Analytical Atomic Spectrometry 1996, 11, (4), 317-22. 28. Diaz Huerta, V.; Hinojosa Reyes, L.; Marchante-Gayon, J. M.; Fernandez Sanchez, M. L.; Sanz-Medel, A., Total determination and quantitative speciation analysis of selenium in yeast and wheat flour by isotope dilution analysis ICP-MS. Journal of Analytical Atomic Spectrometry 2003, 18, (10), 1243-1247. 29. Houk, R. S.; Praphairaksit, N., Dissociation of polyatomic ions in the inductively coupled plasma. Spectrochimica Acta, Part B: Atomic Spectroscopy 2001, 56B, (7), 1069-1096.

114

30. Charles, M. J.; McLuckey, S. A.; Glish, G. L., Competition between resonance ejection and ion dissociation during resonant excitation in a quadrupole ion trap. Journal of the American Society for Mass Spectrometry 1994, 5, (12), 1031-41. 31. Kao, C. M.; Messmer, R. P., The origins of correlation effects in the valence ionization energies of molecular nitrogen: a generalized-valence-bond interpretation. Chemical Physics Letters 1984, 106, (3), 183-6.

115

Chapter 5

Comparing the molecular makeup of mercury in conjunction or exclusion of selenium in the Allium fistulosum, green onion

116

5.1 Abstract

Significant effort has been devoted to studying the mutually protective effect of selenium and mercury in mammals. However, a limited number of studies have investigated the potential selenium-mercury antagonism in plants, which may prove viable for phytoremediation purposes.

In this study, the Allium fistulosum, green onion, was used to investigate the metabolic fate of mercury in conjunction with or exclusion of selenium. The plants were grown in perlite media and supplemented with sodium selenite and mercuric chloride. Total elemental analysis via microwave digestion and inductively coupled plasma mass spectrometry (ICPMS) detection, showed that the A. fistulosum accumulated 50 times more mercury than selenium, both elements were predominately sequestered in the root fraction, and a mutual antagonism was observed.

Size exclusion and capillary reversed phase chromatography were coupled to ICPMS to investigate the plant root and leaf extracts. The data suggests a possible selenium-mercury association in a proteinaceous macromolecule which is not further metabolized upon translocation to the aerial plant regions. The selenium-mercury association and may be formed from small selenium molecules which are normally translocated to the aerial plant regions. Data from enzymatic and acid hydrolysis suggests the formation of mercury metabolites with greater robustness when supplemented in conjunction with selenium.

117

5.2 Introduction

Increased mercury accumulation in the environment has been a topic of ongoing concern.

Once in the soil, mercury can be biomagnified resulting in increased human exposure. Mercury is continually released naturally via atmospheric deposits, soil erosion and through anthropogenic activities including agricultural materials, mining, and urban release, combustion and industrial discharges.1 General phytoremediation efforts, utilizing plants to remove toxins from the environment, are an inexpensive and environmentally friendly method targeting soil clean-up, over conventional remediation methodologies. Genetic engineering has further increased the phytoremediation potential of plants for mercury removal.

In summary, mercury depletion from the soil is accomplished by expressing bacterial genes mer A and mer B in plants, which enable them to convert Hg2+ to Hg0 resulting in volatilization.2, 3 The obvious disadvantages of this methodology are the lack of control over the final destination of mercury and the potential hazards posed in the atmosphere. In addition, the introduction of genetically modified exotic plants may be invasive and detrimental long term to the indigenous plants of a region. Currently, there is a need to further understand the metabolism of mercury in plants and investigate alternative methods for effective remediation.

The allotted daily level of exposure for selenium is relatively narrow with adverse health effects seen with deficiency and excess.4 A large body of epidemiological evidence, including observational, case-control, cohort and randomized controlled clinical trials, support the proposition that selenium may prevent prostate cancer in humans.5 In addition, an initial study in

1967 showed that selenium effectively prevented renal necrosis and mortality caused by mercury in rats.6 This protective effect has also been shown in rabbit liver, dolphin renal cells (Sp1K cells) and human K-562 cells eluding to a universal occurrence in mammals.7-9 Structural basis

118

for the antagonism is postulated as a selenium-mercury core with glutathione attached via a Hg-

10, 11 Se-S(GSH) connection which is bound to selenoprotein P noted as {(Hg-Se)n}m-Sel P.

The selenium-mercury antagonism may prove viable for phytoremediation purposes.

There are a limited number of studies investigating the potential selenium-mercury antagonism in plants. Through quantitative results of a supplemented radish, selenium and mercury were postulated to form a complex in the soil-root environment.12 Only two studies thus far, have investigated the molecular interaction of selenium and mercury in a terrestrial plant. Using supplemented Brassica juncea (Indian mustard) and Glycine max (soybean), it is suggested that selenium and mercury form a high molecular weight, proteinaceous complex which is predominately sequestered in the plant roots.13, 14 Inferences to a selenium-mercury association were based on substituents ambiguously eluting in the void volume of a size exclusion chromatography (SEC) column. Therefore, further investigation is needed for verification of a selenium-mercury complex. In addition, there has not been a study to date monitoring the metabolites of mercury if supplemented alone or simultaneously with selenium in a terrestrial plant.

Simultaneous detection of selenium and mercury at trace levels has historically provided many complications with common analytical techniques. For total elemental analysis, neutron activation analysis of selenium and mercury has been shown in several types of biological samples such as hair, nails, fish reference materials, porcupine tissues and also in selenium supplements.15, 16 However, subcellular fractions must be collected and off-line analysis is required if chromatography is necessary. Total concentrations of selenium and mercury have been successfully determined by atomic absorption spectrometry and inductively coupled plasma mass spectrometry (ICPMS) with detection limits in the low µg l-1 range.17, 18 However,

119

pertinent chromatographic studies have utilized HPLC-ICPMS when investigating selenium- mercury associations at the molecular level. Size exclusion chromatography was coupled with

ICPMS to suggest a high molecular weight entity containing selenium and mercury in the cytosol of chicken livers and in the root extract from plants.13, 14, 19 Further instrumentation advances have enabled the use of capillary liquid chromatography coupled with ICPMS detection for analysis of molecules containing selenium.20, 21

This study seeks to further investigate the mercury metabolites in a plant after supplementation in conjunction with or exclusion of selenium. The plant species selected was the Allium fistulosum (green onion), because of its relative ease of growth and previously noted and investigated selenium accumulation.22 Quantitative analysis of the A. fistulosum plant compartments were analyzed via microwave digestion with ICPMS detection in continuous flow mode. For molecular information regarding general mercury metabolism and investigation of a possible selenium-mercury association, separation techniques such as size exclusion, conventional and capillary reversed phase liquid chromatography (capRPLC) were coupled to

ICPMS. Acid and enzymatic digestions were performed on the selenium and mercury metabolites elucidating structural stability and potential associations.

5.3 Experimental

5.3.1 Instrumentation

High-performance liquid chromatography. Conventional flow rate chromatographic separations were accomplished with an Agilent 1100 liquid chromatograph by Agilent

Technologies (Santa Clara, CA) equipped with a vacuum de-gasser system, a binary HPLC pump, an autosampler, a thermostated column compartment and a diode array detector set at 280

120

nm for monitoring eluting proteins. For purposes of clarity, the 280 nm profile is not shown, but described where pertinent.

The SEC columns used were a Superdex Peptide 10/300 GL SEC (13 μm, 10 x 300 mm) with a listed molecular weight separation range of 0.1 kDa to 7 kDa and an eluent flow rate of

0.6 ml min-1 and a Superdex 200 10/300 GL SEC (13 μm, 10 x 300 mm) with a listed molecular weight separation range of 10 kDa to 600 kDa and an eluent flow rate of 0.65 ml min-1 from

Amersham Pharmacia Biotech AB (Uppsala, Sweden). The Peptide 10/300 GL SEC was calibrated with the following standards: myoglobin, 16.7 kDa; cytochrome C, 12.5 kDa; vitamin

B12, 1.36 kDa; and hexaglycine, 0.36 kDa from Sigma-Aldrich Co. (St. Louis, MO).

Experimental conditions were adapted from previous studies.23

Reversed phase chromatography was carried out with a ZORBAX Eclipse XDB-C18 (5

μm, 4.6 x 250 mm) from Agilent Technologies (Santa Clara, CA) using an eluent flow rate of 1 ml min-1 and experimental conditions were adapted from previous studies.24 Capillary flow rate chromatographic separations were accomplished with an Agilent 1200 liquid chromatograph by

Agilent Technologies equipped with a vacuum de-gasser system, a binary HPLC pump and an autosampler. Capillary reversed phase chromatography was carried out with a ZORBAX 300

SB-C18 (5 μm, 0.5 x 150 mm) from Agilent Technologies using an eluent flow rate of 10 µl min-

1 with the following gradient program: 0-3 min, 10% B; 3-5 min 10-100% B; 5-12 min 100% B;

12-13 min 100-10% B; 13-20 min 10% B.

Scott double channel spray chamber (2°C) for conventional flow rates, a microflow total

121

consumption nebulizer DS-5 from CETAC (Omaha, NE) for capillary flow rates, a shielded torch with a sampling depth of 7 mm, nickel sampling and skimmer cones, an octopole ion guide enclosed in a collision/reaction cell (CRC) pressurized with hydrogen gas, purity of 99.999% from Matheson Gas Products (Parsippany, NJ), a quadrupole mass analyzer with a dwell time of

100 ms per isotope and an electron multiplier for detection. General instrumental parameters were as follows: forward power, 1500 W; plasma gas flow rate, 15.0 l min-1; auxiliary gas flow rate, 1.00 l min-1; carrier gas flow rate, 0.99 l min-1; makeup gas flow rate, 0.14 l min-1; hydrogen

CRC gas flow rate, 4.0 ml min-1; octopole bias, -18 V, quadrupole bias, -16 V for a net energy discrimination voltage of +2 V; monitored isotopes, 77Se, 78Se, 80Se, 200Hg, 202Hg. Instrument tuning parameters including CRC gas flow rate were optimized daily.

Lyophilization and digestion. A Flexi-Dry MP lyophilizer (Stoneridge, NY) was used for freeze drying purposes. The ultrasonic bath used for acid digestion was a Model FS30 from

Fisher (Pittsburgh, PA). The microwave system used for digestion was an Intelligent

Explorer/Discover system from the CEM Corporation (Mathews, NC). The microwave system was programmable for time, temperature, power and pressure, and equipped with a 24 vial autosampler and a self contained microwave chamber.

5.3.2 Reagents and standards

All the solutions were prepared in 18 MΩ cm doubly deionized water (DDW) by Sybron

Barnstead (Boston, MA). Standards used for supplementation and identification were the following: mercuric chloride (HgII) from Mallinckrodt Chemicals (Phillipsburg, NJ); L(+)- selenomethionine (SeMet) from Acros Organics (Morris Plains, NJ); seleno-L-cystine (SeCys2)

122

from Fluka (Milwaukee, WI); potassium selenate (SeVI), sodium selenite (SeIV), and Se-

(methyl)selenocysteine from Sigma-Aldrich Co. (St. Louis, MO).

For total elemental analysis, plant biomass was digested using nitric acid (HNO3) from

Pharmco Products Inc. (Brookfield, CT) and hydrogen peroxide (30%) from Fisher Scientific

(Fair Lawn, NJ). Claritas PPT selenium and mercury elemental standards used for quantification were acquired from SpexCertiPrep (Metuchen, NJ). Calibration standards from 1.0 µg l-1 to

-1 500.0 µg l were prepared through dilution from a stock solution with 5% v/v HNO3.

The following depicts the mobile phases used for the three types of chromatography in this study. The mobile phase for SEC and general plant biomass extraction was made by dissolving Tris(hydroxymethyl) aminomethane hydrochloride (Tris-HCl) from Fisher Scientific in DDW and adjusting the pH with hydrochloric acid from Pharmco Products Inc. For IPRP-

ICPMS, the mobile phase contained 0.1% heptafluorobutyric acid (HFBA) from Sigma-Aldrich

Co. and 5% methanol from Pharmaco Products Inc. For capRP-ICPMS, mobile phase A contained 0.1% formic acid from Fisher Scientific in DDW and mobile phase B contained 0.1% formic acid in 90% acetonitrile from Pharmaco Products Inc. and 10% DDW. All samples were filtered through a 0.2 μm membrane syringe filter by Econofilters from Agilent Technologies

(Santa Clara, CA) before being injected in the HPLC-ICPMS.

5.3.3 Plant cultivation and supplementation

The A. fistulosum plants were purchased from a local supermarket produce department.

The plants were dipped into Essential 1-0-1 from Growth Products (White Plains, NY) for one hour, which provided needed fertilization, then subsequently transplanted into perlite from

Therm-O-Rock West Inc. (Chandler, AZ), which is a porous media used for support. After one

123

month of growth at the University of Cincinnati green house, Department of Biological Sciences,

Cincinnati, OH, the plants were split into four groups containing eight plants each and supplemented with the following: Group I, 30 mg l-1 SeIV; Group II, 15 mg l-1 HgII; Group III, 30 mg l-1 SeIV and 15 mg l-1 HgII; Group IV, control. The plants were given 75 ml of their respective supplementation five times over a one week period and were harvested three days following the supplementation week. The health of each plant was visually indifferent to the supplementation type given. After harvesting, the plants were separated into roots and leaves, washed with DDW and lyophilized. Finally, the plants were homogenized into a powder and stored at -20 °C to prevent any further enzymatic activity leading to interspecies conversion, therefore changing the native distribution.

5.3.4 Total selenium and mercury determination

For determination of total selenium and mercury in the A. fistulosum, a closed vessel microwave digestion system was employed. Briefly, 0.5 ml of HNO3 and 0.5 ml of DDW were added to approximately 25 mg of plant biomass and digested via the following microwave digestion program: Power, 150 W; Ramp, 2:00 min; Hold, 2:00 min; Temperature, 160 °C.

Subsequently, 0.2 ml of 30% H2O2 were added to the solution and digested using the microwave digestion program mentioned previously. The resulting solutions were diluted with DDW to 10 ml and analyzed by ICPMS in continuous flow sample introduction mode.

5.3.5 Extraction procedures of plant tissues

A mild extraction procedure was incorporated in order to preserve the labile compounds in the A. fistulosum plant tissue. In summary, 30 mg of homogenized plant biomass from the

124

root or leaf were combined with 1.5 ml of 100 mmol l-1 Tris-HCl (pH 7.5) and stirred at room temperature for 1.5 h. The solution was then centrifuged at 5000 rpm for 15 min. The supernatant was decanted, filtered through a 0.2 μm filter and 100 μl were injected in the SEC-

ICPMS, IPRP-ICPMS and capRPLC-ICPMS. A similar treatment was used for all plant supplementation types.

5.3.6 A. fistulosum root extract digestion

To elucidate putative metabolic information regarding the selenium and mercury species, enzymatic proteolysis and acid hydrolysis were performed. In summary, 90 mg of A. fistulosum root biomass were combined with 3 ml of 50 mmol l-1 Tris-HCl (pH = 7.5) and stirred at room temperature for 1.5 h. The solution was then centrifuged at 5000 rpm for 15 min. The supernatant was decanted, filtered through a 0.2 μm filter and became the sample solution used in the following digestions.

Acid hydrolysis. For acid hydrolysis, four 100 μl aliquots of the sample solution were combined with HCl and DDW to make equal volumes of 0, 0.1, 1, and 3 M HCl solutions. The solutions were then sonicated for 30 min. and 100 μl amounts were injected into the SEC-

ICPMS.

Enzymatic proteolysis. For enzymatic proteolysis, two procedures were performed. In the first procedure, 200 μl of the sample solution were combined with 30 μl (4 μg) of trypsin from Promega (Madison, WI) and kept at 37 °C for 22 h. The solution was then diluted to 330 μl and 100 μl were injected in the SEC-ICPMS. In the second procedure, 200 μl of the sample solution were combined with 50 μl of 45 mmol l-1 dithiothreitol (DTT) from Sigma-Aldrich Co. and kept at 69 °C for 30 min. Afterward, 50 μl of 100 mmol l-1 iodoacetamide from Sigma-

125

Aldrich Co. were added to the solution and immediately placed in the dark for 20 min.

Subsequently, 30 μl (4 μg) of trypsin were added to the resulting solution and kept at 37 °C for

22 h. Finally, 100 μl of the solution were injected in the SEC-ICPMS.

5.4 Results and discussion

5.4.1 A. fistulosum total element accumulation

Figure 5.1 The A. fistulosum accumulation of selenium and mercury for the varying supplementation types administered during the cultivation process.

126

Total A. fistulosum accumulation of selenium and mercury was determined via microwave digestion of lyophilized plant biomass and subsequent analysis by continuous flow

ICPMS. The resulting selenium and mercury concentrations of the leaves and roots for the varying supplementation types are depicted in Figure 5.1. The total concentrations of selenium and mercury in the control plants were 2.8% and 0.26% of the supplemented plants, respectively.

This depicts that the accumulated selenium or mercury from the A. fistulosum’s initial growth environment is insignificant when compared with the supplementation conditions. Overall, the results show a sequestering of selenium and mercury in the A. fistulosum roots, which concurs with previous studies demonstrating species sequestering in the roots after supplementation of selenium in Brassica oleracea (kale) and mercury in Brassica juncea (Indian mustard) and

Glycine max (soybean).13, 14, 25

The total concentration of selenium in the roots of the SeIV supplemented plants was 227

µg g-1 and 77.8 µg g-1 in the leaves demonstrating a resistance to selenium translocation to the aerial plant regions. In contrast to selenium uptake, greater mercury accumulation was observed.

The total concentration of mercury in the roots of the HgII supplemented plants was 28.7 mg g-1, and 143 µg g-1 in the leaves demonstrating a significant resistance to mercury translocation upon metabolism to the aerial plant regions for accumulation or volatilization. The high mercury accumulation observed may render the A. fistulosum effective for phytoremediation.

For SeIV & HgII supplemented plants, 200 µg g-1 of selenium and 25.2 mg g-1 of mercury were observed in the roots, exhibiting a 12.2% decrease in accumulation for both elements compared to single elemental supplementation. These findings suggest a mutual antagonism between selenium and mercury upon simultaneous supplementation to the A. fistulosum. The selenium-mercury antagonism may occur as a result from a complex forming in the growth

127

media as previously suggested.12 Soil bacteria have been shown to form HgS through a reduction of sulfate to sulfide, which may similarly occur with selenium and mercury forming

HgSe.26 Another possibility allowing for mutual detoxification of the two environmental contaminants may be through the formation of an selenium-mercury complex similar to that

10, 11 observed in the mammalian system, {(Hg-Se)n}m-Sel P. Considering that the A. fistulosum accumulates 50 times more mercury than selenium, the 1:1 selenium to mercury ratio previously observed in mammals is either not applicable in plants or mercury is not completely metabolized with selenium.27, 28 In order to further investigate a plausible selenium-mercury association in the A. fistulosum, SEC-ICPMS, IPRP-ICPMS and capRPLC-ICPMS were utilized.

5.4.2 A. fistulosum extract analysis via SEC-ICPMS

The utilization of SEC-ICPMS provided an overall molecular weight distribution of the selenium and mercury containing species in the A. fistulosum. Plant roots and leaves from varying supplementation combinations were analyzed after a general extraction near physiological pH. The selenium extraction efficiencies for the A. fistulosum leaves were 87% and 89% for SeIV and SeIV & HgII supplemented plants, respectively, showing a near complete extraction of selenium. However, the selenium extraction efficiencies for the A. fistulosum roots were 42% and 33% for SeIV and SeIV & HgII supplemented plants, respectively, alluding to the increased difficulty in breaking the plant’s cell wall and the possibility of hydrophobic selenium containing compounds.

The mercury extraction efficiencies for the A. fistulosum leaves were 11% and 3% for

HgII and SeIV & HgII supplemented plants, respectively, showing a significant quantity in nonpolar compounds and/or within the cell wall of the plant. The mercury extraction efficiencies

128

for the A. fistulosum roots were 15% and 4% for HgII and SeIV & HgII supplemented plants, respectively, suggesting increased difficulty in breaking the plant’s cell wall and the possibility of hydrophobic mercury containing compounds. Significant efforts toward increased mercury extraction were not attempted as dilution of the overall solution would have been necessary due to the overly abundant mercury concentration, resulting in a decreased selenium concentration.

Figure 5.2 80Se and 202Hg SEC-ICPMS chromatograms (0.1 kDa to 7 kDa) of the leaf and root extract from the A. fistulosum after varying supplementation combinations.

The resulting chromatograms after injecting 100 µl of the water soluble plant supernatant from the Tris-HCl extraction in the SEC-ICPMS (0.1 kDa to 7 kDa) are represented in Figure

5.2. The selenium profile for SeIV supplemented A. fistulosum leaf and root extracts displays similar peaks suggesting a conservation of species upon translocation from root to leaf. The

129

selenium profile for SeIV and SeIV & HgII supplemented root extracts also show a very similar metabolite distribution. While the selenium profile for SeIV & HgII supplemented leaf and root extracts show similar molecular weight entities, the concentration is significantly altered when compared with SeIV supplemented plants. The most concentrated peak elutes after the 1.36 kDa standard depicting small selenium containing molecules such as peptides or inorganic species.

This data suggests that the small molecules containing selenium may be incorporated with mercury in the A. fistulosum root rather than translocated to the aerial plant regions.

While the mercury concentration is predominantly sequestered in the roots, the mercury profile for HgII and SeIV & HgII supplemented A. fistulosum leaf and root extracts display similar peaks suggesting a conservation of species upon translocation from root to leaf. The most concentrated peak elutes in the void volume of the column depicting a molecule larger than 16.7 kDa for both supplementation types. Upon observation of the mercury profile near the baseline, small quantities of mercury containing molecules, ranging from 1.36 kDa to 12.5 kDa, are found in the root extract of HgII supplemented plants, but absent in the root extract of SeIV & HgII supplemented plants. The data suggests an alteration in the mercury metabolism if supplemented in conjunction with selenium in the A. fistulosum.

The selenium and mercury profiles overlap in the column’s void volume for SeIV & HgII supplemented leaf and root extracts. This suggests that there is a possible interaction between selenium and mercury in a macromolecule that is not further metabolized upon translocation. In addition, smaller molecular weight molecules comprised of selenium and mercury, such as bis(methylmercuric) selenide previously noted in rats,29 are not considered as prominent antagonism associations in the A. fistulosum.

130

Figure 5.3 80Se and 202Hg SEC-ICPMS chromatograms (10 kDa to 600 kDa) of the root extract from the A. fistulosum after varying supplementation combinations.

For further information, 100 µl of the water soluble plant supernatant from the Tris-HCl extraction of the HgII and SeIV & HgII supplemented roots were injected in the SEC-ICPMS (10 kDa to 600 kDa) and the resulting chromatograms are depicted in Figure 5.3. The mercury profile shows distinctly different characteristics pending the original supplementation administered. This again alludes to the variation in mercury metabolism if selenium is present or absent in the growth media. The selenium and mercury profiles overlap in the column’s void volume for the SeIV & HgII supplemented root extract. This data suggests that there is a possible association between selenium and mercury in a macromolecule of relative hydrodynamic size greater than the molecular weight separation range of the SEC column (600 kDa). Considering that past studies also show a similar high molecular weight entity with selenium and mercury in

B. juncea and G. max,13, 14 these results suggest universal association in plants. The mercury profile for the HgII supplemented root extract also shows a peak in the column’s void volume

131

suggesting a similar metabolite existing in conjunction or exclusion of selenium, possibly involving sulfur.

5.4.3 Enzymatic proteolysis of the A. fistulosum root extract

Figure 5.4 80Se and 202Hg SEC-ICPMS chromatograms (0.1 kDa to 7 kDa) of the root extract following enzymatic digestion from the A. fistulosum after varying supplementation combinations.

For further verification of a selenium-mercury association within the same molecule, enzymatic proteolysis was examined. The resulting SEC-ICPMS (0.1 kDa to 7 kDa)

132

chromatograms after trypsin proteolysis in conjunction or exclusion with DTT of the Tris-HCl extraction from HgII and SeIV & HgII supplemented roots are represented in Figure 5.4. As the selenium-mercury association is hypothesized to be proteinaceous in nature, trypsin was utilized for protein proteolysis. Both chromatograms showing trypsin proteolysis illustrate the general inhibition of trypsin in chemically cleaving either the selenium or mercury containing compounds in the root extracts from HgII or SeIV & HgII supplemented plants. However, the lower molecular weight species containing mercury from the HgII supplemented plants are not observed in contrast to results shown in Figure 5.2. This suggests the potential formation of

HgS during proteolysis, which precipitated and was retained on the SEC column.

Dithiothreitol was then incorporated for disulfide and diselenide bond cleavage as shown in previous studies.30 The selenium profile for trypsin and DTT proteolysis of the SeIV & HgII supplemented A. fistulosum root extract shows a new selenium peak not previously seen at 39 min. While mercury has been shown to deactivate trypsin,31 the data suggests that inorganic mercury is not interfering with unique association of trypsin’s catalytic triad amino acids (His57,

Asp102, Ser195). The mercury profile for trypsin and DTT proteolysis of the HgII and SeIV & HgII supplemented A. fistulosum root extract shows a significant reduction in the void volume peak concentration and additional peaks with greater elution times were absent suggesting mercury adsorption to the column. Considering the pKsp values for HgSe and HgS are 56.6 and 51.05, respectively, this data suggests the formation of insoluble complexes which precipitated and remained in the column after proteolysis.32, 33

The mercury profile for the trypsin and DTT proteolysis of the HgII and supplemented A. fistulosum root extract shows several mercury containing molecules not previously seen. This suggests differences in stability and overall structure when compared to the high molecular

133

weight entity eluting in the void volume. In addition, selenium may aid in stabilizing these common mercury containing metabolites as they are not observed in the mercury profile for SeIV and HgII supplemented plants.

5.4.4 Acid hydrolysis of the A. fistulosum root extract

Figure 5.5 78Se and 202Hg SEC-ICPMS chromatograms (0.1 kDa to 7 kDa) of the root extract following acid digestion from the A. fistulosum after varying supplementation combinations.

Considering enzymatic proteolysis did not result in the profile overlap of selenium and mercury at an alternate elution time other than the column void volume, acid hydrolysis was

134

examined in hopes of further confirming selenium-mercury complex formation. The resulting

SEC-ICPMS (0.1 kDa to 7 kDa) chromatograms after acid hydrolysis at varying concentrations of the Tris-HCl extraction from HgII and SeIV & HgII supplemented roots are represented in

Figure 5.5. The 0 mol l-1 HCl was used as a control to demonstrate that a short period of ultra- sonication did not degrade the selenium or mercury containing molecules when compared with

Figure 5.2.

After exposing the A. fistulosum root extracts to 0.1 mol l-1 HCl, a reduction in the void volume peak concentration was observed in the selenium and mercury profiles from plants of both supplementation types. A greater signal reduction was observed for HgII supplemented plants (blue line) suggesting an added stability of the mercury macromolecule once selenium is introduced into its compound matrix. This coincides with a previous study demonstrating a greater bond affinity for mercury to selenium than sulfur.34 The selenium and mercury profiles for SeIV & HgII supplemented plants (red and black lines) after 0.1 mol l-1 HCl hydrolysis show a new overlap with an elution time of 15.5 minutes and an absorbance at 280 nm, which is not observed in the mercury profile for HgII supplemented plants. Additionally, an absorbance at

280 nm is not observed in the acid control sample. As the acid hydrolyzed molecule elutes later than the column void volume and with UV conformation, this data further supports the hypothesized formation of a proteinaceous macromolecule partially composed of selenium and mercury.

After exposing the A. fistulosum root extracts to 1 mol l-1 HCl, a reduced void volume peak concentration was observed in the selenium and mercury profiles from plants of both supplementation types. The hydrolyzed mercury compounds have retention times of 35.7 min and 43.5 min with corresponding absorptions at 280 nm. When comparing the profiles of

135

selenium and mercury for the root extract from the SeIV & HgII supplemented plant, there is a distinct association, or peak overlap, at 43.5 min and plausibly at 35.7 min. This data provides further evidence of a selenium-mercury association in the same proteinaceous macromolecule.

The mercury profile for the root extract from HgII supplemented plants also shows peaks with retention times of 35.7 min and 43.5 min. This data suggests a similar macromolecule metabolized by the A. fistulosum in conjunction or exclusion of selenium.

After exposing the A. fistulosum root extracts to 3 mol l-1 HCl, the selenium and mercury profiles generally show similar peaks from plants of both supplementation types when compared with 1 mol l-1 HCl hydrolysis. Due to a reduced analyte concentration, the plausible selenium- mercury complex for the root extract from SeIV & HgII supplemented plants (35.7 min) after 1 mol l-1 HCl hydrolysis is more clearly resolved after 3 mol l-1 HCl hydrolysis. The mercury peak’s retention time is 34.5 min, while the selenium peak’s retention time is 36.5 min, both with a corresponding 280 nm absorbance. This data suggests that mercury is also complexed by an organic molecule via other elements than selenium.

5.4.5 A. fistulosum extract analysis via IPRP/capRPLC-ICPMS

In order to provide additional support for a selenium-mercury containing metabolite in the A. fistulosum, IPRP-ICPMS and capRPLC-ICPMS were utilized. Although not shown, the

IPRP-ICPMS conditions fit for selenium speciation were ineffective in eluting the mercury metabolites from the column. This is attributed to the size of the mercury complexes (often greater than 600 kDa). The high concentration of organic solvent needed for mercury displacement caused plasma instability and ultimately extinguishing. Therefore, capRPLC-

136

ICPMS was used and the resulting chromatograms of the Tris-HCl extractions from HgII and

SeIV & HgII supplemented roots are represented in Figure 5.6.

Figure 5.6 78Se and 202Hg capRPLC-ICPMS chromatograms of the root extract from the A. fistulosum after varying supplementation combinations.

The mercury profile from the root extract of HgII supplemented plants confirms prior

SEC analysis displaying a high molecular weight complex with several lower molecular weight complexes containing mercury. The mercury profile from the root extract of SeIV & HgII supplemented plants also confirms prior SEC analysis displaying a high molecular weight complex coinciding with selenium. The data suggests a possible selenium-mercury association and a plausible sulfur-mercury association in a macromolecular complex upon metabolism of the supplemented inorganic forms in the A. fistulosum.

137

5.5 Acknowledgements

The authors would like to acknowledge the following: Dr. Santha Yathavakilla and Dr.

Monica Shah for their help in unpublished background work prior this manuscript, Agilent

Technologies and CEM Corp. for their instrumentation and continuing support, Pam Bishop for assistance in Allium fistulosum growth, Brittany Catron for assistance in graphical representation, and Dr. Lisa Afton for assistance in plant harvesting.

138

5.6 References

1. Wang, Q.; Kim, D.; Dionysiou Dionysios, D.; Sorial George, A.; Timberlake, D., Sources and remediation for mercury contamination in aquatic systems--a literature review. Environ Pollut FIELD Full Journal Title:Environmental pollution (Barking, Essex: 1987) 2004, 131, (2), 323-36. 2. Hussein, H. S.; Ruiz, O. N.; Terry, N.; Daniell, H., Phytoremediation of Mercury and Organomercurials in Chloroplast Transgenic Plants: Enhanced Root Uptake, Translocation to Shoots, and Volatilization. Environmental Science & Technology 2007, 41, (24), 8439-8446. 3. Ruiz, O. N.; Hussein, H. S.; Terry, N.; Daniell, H., Phytoremediation of organomercurial compounds via chloroplast genetic engineering. Plant Physiology 2003, 132, (3), 1344-1352. 4. Rayman, M. P., Food-chain selenium and human health: emphasis on intake. British Journal of Nutrition 2008, 100, (2), 254-268. 5. Klein, E. A., Selenium: Epidemiology and Basic Science. Journal of Urology (Hagerstown, MD, United States) 2004, 171, (2, Pt. 2), S50-S53. 6. Parizek, J.; Ostadalova, I., The protective effect of small amounts of selenite in sublimate intoxication. Experientia 1967, 23, (2), 142-3. 7. Frisk, P.; Wester, K.; Yaqob, A.; Lindh, U., Selenium protection against mercury-induced apoptosis and growth inhibition in cultured K562 cells. Biological Trace Element Research 2003, 92, (2), 105-114. 8. Naganuma, A.; Imura, N., Changes in distribution of mercury and selenium in soluble fractions of rabbit after simultaneous administration. Pharmacology, Biochemistry and Behavior 1980, 13, (4), 537-44. 9. Wang, A.; Barber, D.; Pfeiffer, C. J., Protective effects of selenium against mercury toxicity in cultured Atlantic spotted dolphin (Stenella plagiodon) renal cells. Archives of Environmental Contamination and Toxicology 2001, 41, (4), 403-409. 10. Gailer, J.; George, G. N.; Pickering, I. J.; Madden, S.; Prince, R. C.; Yu, E. Y.; Denton, M. B.; Younis, H. S.; Aposhian, H. V., Structural Basis of the Antagonism between Inorganic Mercury and Selenium in Mammals. Chemical Research in Toxicology 2000, 13, (11), 1135- 1142. 11. Yoneda, S.; Suzuki, K. T., Equimolar Hg-Se complex binds to selenoprotein P. Biochemical and Biophysical Research Communications 1997, 231, (1), 7-11. 12. Shanker, K.; Mishra, S.; Srivastava, S.; Srivastava, R.; Dass, S.; Prakash, S.; Srivastava, M. M., Study of mercury-selenium (Hg-Se) interactions and their impact on Hg uptake by the radish (Raphanus sativus) plant. Food and Chemical Toxicology 1996, 34, (9), 883-886. 13. Mounicou, S.; Shah, M.; Meija, J.; Caruso, J. A.; Vonderheide, A. P.; Shann, J., Localization and speciation of selenium and mercury in Brassica juncea-implications for Se-Hg antagonism. Journal of Analytical Atomic Spectrometry 2006, 21, (4), 404-412. 14. Yathavakilla, S. K. V.; Caruso, J. A., A study of Se-Hg antagonism in Glycine max (soybean) roots by size exclusion and reversed phase HPLC-ICPMS. Analytical and Bioanalytical Chemistry 2007, 389, (3), 715-723. 15. Vasconcellos, M. B. A.; Catharino, M. G. M.; Paletti, G.; Saiki, M.; Bode, P.; Favaro, D. I. T.; Baruzzi, R.; Rodrigues, D. A., Determination of mercury and selenium in biological samples by neutron activation analysis. Journal of Trace and Microprobe Techniques 2002, 20, (4), 527-538.

139

16. Zhao, J.; Chen, C.; Zhang, P.; Chai, Z.; Qu, L.; Li, M., Preliminary study of selenium and mercury distribution in some porcine tissues and their subcellular fractions by NAA and HG- AFS. Journal of Radioanalytical and Nuclear Chemistry 2004, 259, (3), 459-463. 17. Frias, S.; Diaz, C.; Conde, J. E.; Trujillo, P. J. P., Selenium and mercury concentrations in sweet and dry bottled wines from the Canary Islands, Spain. Food Additives & Contaminants 2003, 20, (3), 237-240. 18. Li, Y.-F.; Chen, C.; Li, B.; Wang, Q.; Wang, J.; Gao, Y.; Zhao, Y.; Chai, Z., Simultaneous speciation of selenium and mercury in human urine samples from long-term mercury-exposed populations with supplementation of selenium-enriched yeast by HPLC-ICP- MS. Journal of Analytical Atomic Spectrometry 2007, 22, (8), 925-930. 19. Cabanero, A. I.; Madrid, Y.; Camara, C., Study of mercury-selenium interaction in chicken liver by size exclusion chromatography inductively coupled plasma mass spectrometry. Journal of Analytical Atomic Spectrometry 2005, 20, (9), 847-855. 20. Ellis, J.; Grimm, R.; Clark, J. F.; Pyne-Gaithman, G.; Wilbur, S.; Caruso, J. A., Studying Protein Phosphorylation in Low MW CSF Fractions with capLC-ICPMS and nanoLC-CHIP- ITMS for Identification of Phosphoproteins. Journal of Proteome Research, ACS ASAP. 21. Schaumloeffel, D.; Encinar, J. R.; Lobinski, R., Development of a sheathless interface between reversed-phase capillary HPLC and ICPMS via a microflow total consumption nebulizer for selenopeptide mapping. Analytical Chemistry 2003, 75, (24), 6837-6842. 22. Shah, M.; Kannamkumarath, S. S.; Wuilloud, J. C. A.; Wuilloud, R. G.; Caruso, J. A., Identification and characterization of selenium species in enriched green onion (Allium fistulosum) by HPLC-ICP-MS and ESI-IT-MS. Journal of Analytical Atomic Spectrometry 2004, 19, (3), 381-386. 23. Afton, S.; Catron, B.; Caruso, J. A., Elucidating the selenium and arsenic metabolic pathways following exposure to the non-hyperaccumulating Chlorophytum comosum, spider plant. Submitted for review. 24. Kotrebai, M.; Tyson, J. F.; Uden, P. C.; Birringer, M.; Block, E., Selenium speciation in enriched and natural samples by HPLC-ICP-MS and HPLC-ESI-MS with perfluorinated carboxylic acid ion-pairing agents. Analyst (Cambridge, United Kingdom) 2000, 125, (1), 71-78. 25. Pedrero, Z.; Elvira, D.; Camara, C.; Madrid, Y., Selenium transformation studies during Broccoli (Brassica oleracea) growing process by liquid chromatography-inductively coupled plasma mass spectrometry (LC-ICP-MS). Analytica Chimica Acta 2007, 596, (2), 251-256. 26. Revis, N. W.; Osborne, T. R.; Holdsworth, G.; Hadden, C., Distribution of mercury species in soil from a mercury-contaminated site. Water, Air, and Soil Pollution 1989, 45, (1-2), 105-13. 27. Kuehl, D. W.; Haebler, R.; Potter, C., Coplanar PCB and metal residues in dolphins from the U.S. Atlantic Coast including Atlantic bottlenose obtained during the 1987/88 mass mortality. Chemosphere 1994, 28, (6), 1245-53. 28. Bjoerkman, L.; Palm, B.; Nylander, M.; Nordberg, M., Mercury and selenium distribution in human kidney cortex. Biological Trace Element Research 1994, 40, (3), 255-65. 29. Masukawa, T.; Kito, H.; Hayashi, M.; Iwata, H., Formation and possible role of bis(methylmercuric) selenide in rats treated with methylmercury and selenite. Biochemical Pharmacology 1982, 31, (1), 75-8. 30. Guenther, W. H. H., Methods in selenium chemistry. III. Reduction of diselenides with dithiothreitol. Journal of Organic Chemistry 1967, 32, (12), 3931-4.

140

31. Hong, F.; Wang, L.; Wang, X.; Lei, Z.; Chao, L., Effects of Ce3+, Cd2+, and Hg2+ on activities and secondary structure of trypsin. Biological Trace Element Research 2003, 95, (3), 233-240. 32. Goates, J. R.; Cole, A. G.; Gray, E. L., Free energy of formation and solubility product constant of mercuric sulfide. Journal of the American Chemical Society 1951, 73, 3596-7. 33. Mehra, M. C.; Gubeli, A. O., Complexing characteristics of insoluble selenides. 3. Mercuric selenide. Journal of the Less-Common Metals 1971, 25, (2), 221-4. 34. Sugiura, Y.; Tamai, Y.; Tanaka, H., Selenium protection against mercury toxicity: high binding affinity of methylmercury by selenium-containing ligands in comparison with sulfur- containing ligands. Bioinorganic Chemistry 1978, 9, (2), 167-80.

141

Chapter 6

Exploring the structural basis for mercury/selenium antagonism in Allium fistulosum - green onion

142

6.1 Abstract

While significant efforts have been devoted to studying the mutually protective effect of mercury and selenium in mammals, a limited number of studies have investigated the potential mercury-selenium antagonism in plants, which may prove viable for phytoremediation purposes.

In this study, the Allium fistulosum, green onion, was used to investigate the metabolic fate of mercury and selenium. The plants were grown in perlite media and supplemented with sodium selenite and mercuric chloride. Data from x-ray fluorescence mapping of a freshly excised root and capillary reversed phase chromatography, with inductively coupled plasma mass spectrometry detection of homogenized root extract, suggest the plausible formation of a mercury-selenium species and a similarly structured mercury-sulfur species predominantly residing in the cell wall of the epidermal root tissue. With x-ray absorption near edge structure analysis, the local environment of mercury and selenium qualitatively match the mercury- selenium species formed in a mammal via a Hg-Se-S(GSH) moiety. In addition, mercury also matches the local environment of (GS)2Hg, suggesting a difference in metabolism.

143

6.2 Introduction

The problem of mercury and selenium accumulation in soil has been a well known area of concern for some time. Once in the soil, mercury and selenium can be chemically altered to more toxic forms and/or biomagnified resulting in a more serious concern with human exposure.

Mercury and selenium are continually released into the environment naturally and from anthropogenic activities including soil erosion, atmospheric deposits, irrigation drainage water, mining, coal ash leachates, agricultural materials, and a host of general industrial discharges.1, 2

Investigations into how these elements interact in the environment as well as how to manage their fate is paramount in order to protect human and environmental health.

General phytoremediation efforts, utilizing plants to remove toxins from the environment, usually suggest an inexpensive and environmentally benign alternative to conventional methodologies targeting elemental remediation. Genetic engineering has further increased the phytoremediation potential of plants for mercury and selenium removal. Phytoremediation studies have focused on volatilizing mercury as Hg0 through the cell surface after metabolism.

By expressing the bacterial genes merA, which encodes mercury reductase, and merB, which encodes organomercury lyase, in plants, increased mercury reduction and volatilization have been accomplished.3, 4 Genetically modified poplar and cottonwood trees via merA and/or merB expression show potential as phytoremediators of mercury, and do not require regular seasonal replanting.5, 6 Phytoremediation studies have focused on volatilizing selenium as dimethyldiselenide after metabolism. By overexpressing ATP sulfurylase, which results in increased selenium accumulation, and SeCys methyltransferase, which results in the formation of dimethyldiselenide, increased remediation of selenium was accomplished in Brassica juncea

(Indian mustard).7 While successfully diluting the concentration of mercury and selenium in a

144

specified location as well as putative toxicity reduction, the final location and oxidation state of mercury and selenium are impossible to control with such a proposed methodology.

The mutual protective effect of simultaneously administered mercury and selenium in mammals has been known for four decades since an initial study in 1967 showed that selenium effectively prevented renal necrosis and mortality caused by mercury in rats.8 This mercury- selenium antagonism has also been shown in human K-562 cells, dolphin Sp1K renal cells, and rabbit liver, eluding to a universal occurrence in mammals.9-11 Structural basis for the antagonism is postulated as a mercury-selenium core with glutathione attached via a Hg-Se-

S(GSH) connection that is bound to selenoprotein P noted as [(Hg-Se)n]m-Sel P, which was elucidated utilizing data collected via x-ray absorption fine structure spectroscopy (XAFS).12, 13

A possible alternative approach using phytoremediation may be incorporated for mercury and selenium remediation. Termed phytoretirement, mercury and selenium would be extracted, sequestered and metabolized in plant tissues resulting in the formation of a metabolically stable mercury-selenium complex, which would effectively lower the overall biologically active concentration of both elements.

There are a limited number of studies investigating the potential metabolism of mercury and selenium in plants on a molecular level. Using supplemented Glycine max (soybean),

Brassica juncea (Indian mustard) and Allium fistulosum (green onion), it is suggested that mercury and selenium form a high molecular weight, proteinaceous complex that is predominately sequestered in the plant roots.14-16 Further, it was suggested that the formed mercury-selenium complex is more robust, but similar in composition to a plausibly formed mercury-sulfur complex.16 Inferences to a mercury-selenium association were suggested from overlapping chromatographic profiles after altering conditions were imposed. Further

145

investigation is needed for verification and speciation of a mercury-selenium complex and to gain possible additional metabolic information.

Simultaneous detection of the mercury-selenium complex has historically provided many complications with common analytical techniques. Electrospray ionization mass spectrometry, matrix assisted laser desorption ionization time-of-flight mass spectrometry, fast atom bombardment mass spectrometry, 77Se or 199Hg nuclear magnetic resonance signals, and Raman spectroscopy yielded inconclusive results for identification, which was likely due to the stability of the compound and the large molecular size.12 Capillary reversed phase chromatography was coupled to inductively coupled plasma mass spectrometry (capRPLC-ICPMS) for investigation of the mercury-selenium complex in the root extract of supplemented A. fistulosum.16 However, sulfur was not monitored, which is likely part of the complex composition. X-ray absorption fine structure spectroscopy has been utilized for mercury and selenium identification in prior studies.12, 17-21 As minimal sample preparation is required and mercury and selenium species may be analyzed in plant root tissue without the need for aqueous extraction, XAFS is suggested to be the most effective means for elucidating location and possible speciation of the mercury- selenium complex.

This study seeks to further investigate the metabolic fate of mercury and selenium upon accumulation in a plant after supplementation. The plant species selected for this study was the

A. fistulosum, because of its relative ease of growth and known accumulation of mercury and selenium.16, 22 The sulfur, selenium and mercury species in the root extract of the A. fistulosum were analyzed by capRPLC-ICPMS. X-ray fluorescence maps were constructed of freshly excised root tissue revealing the location of mercury and selenium. The elucidation of the local environments of mercury and selenium in vivo were performed via XAFS.

146

6.3 Experimental

6.3.1 Plant cultivation and supplementation

The A. fistulosum plants were purchased from a local supermarket produce department.

The plants were dipped into Essential 1-0-1 from Growth Products (White Plains, NY) for one hour, which provided needed fertilization, then subsequently transplanted into perlite from

Therm-O-Rock West Inc. (Chandler, AZ), which is a porous media used for support. After one month of growth at the University of Cincinnati green house, Department of Biological Sciences,

Cincinnati, OH, the plants were split into three groups containing eight plants and supplemented

II IV with HgCl2 (Hg ) from Mallinckrodt Chemicals (Phillipsburg, NJ) and Na2SeO3 (Se ) from

Sigma-Aldrich Co. (St. Louis, MO) according to the following: Group I, 20 mg l-1 HgII; Group

II, 30 mg l-1 SeIV and 20 mg l-1 HgII; Group III, control. The plants were given 100 ml of their respective supplementation six times over a period of two weeks and were harvested after 4 days following the supplementation week. All of the solutions were prepared in 18 MΩ cm doubly deionized water (DDW) by Sybron Barnstead (Boston, MA). The health of each plant was visually indifferent to the supplementation type given.

6.3.2 capRPLC-ICPMS analysis

For capRPLC-ICPMS analysis, the plant roots were separated, washed with DDW and lyophilized with a Flexi-Dry MP lyophilizer (Stoneridge, NY). Finally, the roots were homogenized into a powder and stored at -20 °C to prevent any further enzymatic activity leading to interspecies conversion, therefore changing the native distribution. A mild extraction procedure was incorporated in order to preserve the labile compounds in the A. fistulosum root tissue. In summary, 25 mg of homogenized plant biomass from the root were combined with 1.5

147

ml of 20 mmol l-1 Tris-HCl (pH 7.5) and stirred at room temperature for 1.5 h. The solution was then centrifuged at 5000 rpm for 15 min. The supernatant was decanted, filtered through a 0.2

μm membrane syringe filter by Econofilters from Agilent Technologies (Santa Clara, CA) and

100 μl were injected in the capRPLC-ICPMS.

Capillary flow rate chromatographic separations were accomplished with an Agilent 1200 liquid chromatograph by Agilent Technologies (Santa Clara, CA) equipped with a vacuum de- gasser system, a binary HPLC pump and an autosampler. Capillary reversed phase chromatography was carried out with a ZORBAX 300 SB-C18 (5 μm, 0.5 x 150 mm) from

Agilent Technologies using an eluent flow rate of 10 µl min-1 with the following gradient program: 0-3 min, 10% B; 3-5 min 10-100% B; 5-12 min 100% B; 12-13 min 100-10% B; 13-20 min 10% B as used in a previous study.16 Mobile phase A contained 0.1% formic acid from

Fisher Scientific (Fair Lawn, NJ) in DDW and mobile phase B contained 0.1% formic acid in

90% acetonitrile from Pharmaco Products Inc. (Brookfield, CT) and 10% DDW.

The ICPMS used for specific element detection was an Agilent 7500ce by Agilent

Technologies (Santa Clara, CA). The instrument was equipped with a microflow total consumption nebulizer DS-5 from CETAC (Omaha, NE), a shielded torch with a sampling depth of 9 mm, nickel sampling and skimmer cones, an octopole ion guide enclosed in a collision/reaction cell (CRC) pressurized with xenon gas, purity of 99.999% from Matheson Gas

Products (Parsippany, NJ), a quadrupole mass analyzer with a dwell time of 100 ms per isotope and an electron multiplier for detection. Instrumental parameters were as follows: forward power, 1500 W; plasma gas flow rate, 15.0 l min-1; auxiliary gas flow rate, 1.00 l min-1; xenon

CRC gas flow rate, 0.1 ml min-1; octopole bias, -20 V; quadrupole bias, -10 V; monitored isotopes, 32S, 34S, 78Se, 80Se, 200Hg, and 202Hg.

148

Several factors were considered when tuning the ICPMS in capillary mode for sulfur, selenium and mercury. When optimizing for sulfur, the most significant impact to the signal to noise ratio was made by increasing the Xe CRC gas flow rate. This is attributed to greater

34 + 34 + 23 dissociation energy of the predominant interference of S , O2 (6.66 eV), which is attenuated by voltage discrimination.24 When optimizing for selenium, the most significant impact to the signal to noise ratio was made by increasing the difference between the octopole and quadruple bias voltages. This is attributed to the lower dissociation energy of the

78 + 78 + 25 predominant interference of Se , Ar2 (1.5 eV to 2 eV), which is attenuated by the increased net energy barrier without the need of greater Xe CRC gas flow rates required for 34S analysis.

As mercury has few interferences due to its high elemental mass, all adjustments to increase Xe

CRC gas flow rate and/or the difference between the octopole and quadruple bias voltages resulted in a decreased signal to noise ratio. As multi-element monitoring was necessary, a comprised set of parameters were used in this study as described above.

6.3.3 XAFS analysis

For XAFS analysis and x-ray fluorescence mapping, the A. fistulosum roots were washed twice with DDW and once with 1 mmol l-1 calcium chloride as a means of removing any weakly absorbed mercury or selenium. Freshly excised root sections were removed just prior to analysis and mounted directly to the x-y-theta stage by attaching the ends of the root to Kapton tape from

DuPont (Wilmington, DE) with super glue from the Super Glue Corp. (Rancho Cucamonga, CA) and freezing the sample in liquid nitrogen. The x-y-theta stage was cooled to -120 °C to prevent beam induced reduction of the mercury and selenium species, which was previously observed at

-30 °C.

149

X-ray fluorescence mapping and XAFS experiments were measured at the Advanced

Photon Source (APS), Argonne National Laboratory, Chicago, IL on the GeoSoilEnviroCARS

Beamline 13-ID. The incident beam was further resolved by a Si(111) double-crystal monochromator. Harmonic rejection was accomplished by detuning one crystal of the monochromator to ca. 50% off-peak. The Hg LIII-edge XAFS spectra were scanned from 30 eV below and 85 eV above the edge threshold energy of 12.284 keV. The Se K-edge XAFS spectra were measured from 20 eV below and 70 eV above the edge threshold energy of 12.655 keV. X- ray absorption spectra were measured as the x-ray fluorescence excitation spectra by monitoring the Hg Lβ1 (11.82 keV) or Se Kα (11.21 keV) intensity using a 16-element Ge solid-state detector. All spectra were normalized for the intensities of the incident beam, I0, measured by an ionization chamber filled with N2 gas. The fluorescence intensities were measured for 2 s to 6 s per sampling point. All data were collected at -120 °C.

6.4 Results and discussion

6.4.1 Analysis of A. fistulosum root extract via capRPLC-ICPMS

As previous studies have noted, the predominant sequestering of mercury and selenium species in the root tissue of the A. fistulosum and other plants,14-16 only the root compartment was analyzed throughout the study. In order to provide a generalized profile of the sulfur, selenium and mercury species in the A. fistulosum root, a mild Tris-HCl extraction was performed on homogenized root tissue. As high concentrations of organic solvent needed for analyte displacement in reversed phase chromatography caused plasma instability and ultimately extinguished the plasma, capRPLC-ICPMS was utilized, which allowed a stronger solvent, but at much lower solvent load than with std. bore LC. The resulting capRPLC-ICPMS

150

chromatograms depicting the 34S, 78Se and 202Hg signals from plants supplemented with HgII or

SeIV & HgII can be seen in Figure 6.1.

Figure 6.1 34S, 78Se and 202Hg capRPLC-ICPMS chromatograms of the root extract from the A. fistulosum after varying supplementation with HgII or SeIV & HgII.

For HgII supplemented plants, a profile overlap of 34S and 202Hg is seen eluting later in the chromatogram suggesting a large molecule with a mercury and sulfur association, which was previously suggested to be greater than 600 kDa.16 Additional small molecules containing sulfur are also seen. For SeIV & HgII supplemented plants, a similar profile overlap of 34S, 78Se and

202Hg is seen eluting at the same time as the 34S and 202Hg profile overlap from the extract of HgII supplemented plants. In addition, previous work has shown a significant accumulation of mercury by the A. fistulosum versus selenium.16 This suggests the plausible formation of a mercury-selenium species and a similarly structured mercury-sulfur species. Additional small molecules containing sulfur and selenium are also seen with several overlapping profiles suggesting similar structures. 151

6.4.2 Mercury and selenium location in the A. fistulosum root

In order provide a general location of mercury and selenium, a 500 µm x 500 µm fluorescence map depicting the elemental distribution was made from ca. the center section of an excised SeIV & HgII supplemented A. fistulosum root 1.3 mm in diameter and can be seen in

Figure 6.2. The results support the capRPLC-ICPMS chromatograms from the SeIV & HgII supplemented root extract. Significant co-localization between mercury and selenium is observed in several sites (highlighted by circles). This suggests the possible formation of a mercury-selenium containing molecule in the root tissue. It can also be seen that there are unique areas of increased mercury and selenium concentration that do not coincide. This data supports the formation of differing mercury metabolites, possibly associated with sulfur, and the metabolism of selenium into molecules of decreased molecular size observed in the capRPLC-

ICPMS chromatograms.

152

Figure 6.3 X-ray fluorescence maps of the edge section from an excised SeIV & HgII supplemented A. fistulosum root depicting the elemental distribution of mercury and selenium.

The high photon flux density available at the APS provided a micro scale beam allowing for spatially resolved analysis of the mercury and selenium distribution. A 500 µm x 500 µm fluorescence map depicting the elemental distribution was made from the edge section of an excised SeIV & HgII supplemented A. fistulosum root 1.3 mm in diameter and can be seen in

Figure 6.3. The fluorescence map for selenium shows wide distribution in concentration suggesting the residing place of selenium to be throughout the plant roots. While selenium has not been proven essential for plants, the extensive distribution suggests that selenium is not readily toxic to the A. fistulosum.

In contrast, the fluorescence map for mercury shows the greatest concentration at the root edge (representing a higher concentration of the root circumference). In addition, an outline of a cell formed by a difference in concentration can be seen toward the center of the root. This suggests that the predominant residing location for mercury is in the cell wall of the epidermal

153

root tissue, which can be referenced in Figure 6.4. Previous studies have also noted that mercury is accumulated and sequestered in the root cell wall of the Triticum turgidum (durum wheat) and Salix spp. (willow).26, 27 While essential metals to plants such as copper, iron and zinc have been shown to be actively transported into the plant roots, general passive induction of metals is also common through alterations in the rhizosphere.28 The long-distance transport of nutrients requires xylem loading in the roots, movement to the aerial plant regions and efficient xylem unloading.28 The ability to store mercury in a large molecular species located in the epidermal tissue reduces the potential toxicity of mercury to the plant from ambiguous passive induction by alleviating the possibility for translocation to the aerial plant regions. In addition, the toxicity to the epidermal cells is reduced by mercury storage in the cell wall where few biological processes occur.

Figure 6.4 A diagram representing the cross section of a monocot root.

154

6.4.3 Mercury and selenium speciation in the A. fistulosum root

Considering the photon beam penetrates through the entire excised root sample from the photoelectric effect, a projection showing of the local environments of the element of interest are given in a XAFS spectrum. A typical XAFS spectrum can be divided into two energy regions: x- ray absorption near edge structure (XANES) and extended x-ray absorption fine structure

155

(EXAFS). Analysis in the XANES region is within 50 eV of the absorption edge and can be used to elucidate the oxidation sate and site symmetry of the element of interest. Analysis in the

EXAFS region is beyond 50 eV of the absorption edge and can provide structural information of the atoms surrounding the analyte of interest including bond length, coordination numbers and ligand identity.

For selenium speciation, the XANES region provided sufficient features for qualitative identification without using mathematical data fitting in the EXAFS region. Using the fluorescence absorption maps from Figure 6.2, spots of selenium and mercury co-localization were used to probe for XANES spectra of the Se K-edge. The resulting absorbance from the

SeIV & HgII supplemented A. Fistulosum excised root compared with previously noted standards can be seen in Figure 6.5. The selenium absorption profile for the A. Fistulosum closely matches the unknown mercury-selenium species found in rabbit plasma after simultaneous mercury and selenium supplementation and the synthesized mercury-selenium species which are hypothesized to contain similar local environments depicted as a Hg-Se-S(GSH) moiety.12

For mercury speciation, the absorption edge resulting in the greatest signal for mercury,

Hg LIII, was not utilized for EXAFS analysis due to the overlapping absorbance of the Se K- edge. In addition, the Hg LII-edge did not provide sufficient features for mathematical data fitting. Therefore, qualitative identification was made using the XANES region of the Hg LIII- edge. Using the fluorescence absorption maps from Figure 6.2, spots of selenium and mercury co-localization were used to probe for XANES spectra of the Hg LIII-edge. The resulting absorbance from the SeIV & HgII supplemented A. Fistulosum excised root compared with previously noted standards can be seen in Figure 6.6.

156

While there are few characteristic features, the mercury absorption profile for the A.

Fistulosum, Figure 6.6 (A), closely matches the unknown mercury-selenium species found in rabbit plasma after simultaneous mercury and selenium supplementation and the synthesized mercury-selenium species which are hypothesized to contain similar local environments depicted as a Hg-Se-S(GSH) moiety.12 However, the mercury absorption profile for the A. Fistulosum,

Figure 6.6 (B), closely matches mercury diglutathione (GS)2Hg. This data compliments the

157

capRPLC-ICPMS chromatograms and x-ray fluorescence maps suggesting two distinct associations of mercury with selenium and sulfur.

6.5 Acknowledgements

The authors would like to acknowledge the following: GSECARS grant (GUP9279) for beamline time; Matt Newville and Steve Sutton for their assistance and expertise operating the

GSECARS Beamline 13-ID at APS in Argonne National Laboratory. We also thank Agilent

Technologies for their instrumentation and continuing support.

158

6.6 References

1. Berken, A.; Mulholland, M. M.; LeDuc, D. L.; Terry, N., Genetic engineering of plants to enhance selenium phytoremediation. Critical Reviews in Plant Sciences 2002, 21, (6), 567-582. 2. Wang, Q.; Kim, D.; Dionysiou Dionysios, D.; Sorial George, A.; Timberlake, D., Sources and remediation for mercury contamination in aquatic systems--a literature review. Environ Pollut FIELD Full Journal Title:Environmental pollution (Barking, Essex: 1987) 2004, 131, (2), 323-36. 3. Bizily, S. P.; Rugh, C. L.; Summers, A. O.; Meagher, R. B., Phytoremediation of methylmercury pollution: merB expression in Arabidopsis thaliana confers resistance to organomercurials. Proceedings of the National Academy of Sciences of the United States of America 1999, 96, (12), 6808-6813. 4. Rugh, C. L.; Wilde, H. D.; Stack, N. M.; Thompson, D. M.; Summers, A. O.; Meagher, R. B., Mercuric ion reduction and resistance in transgenic Arabidopsis thaliana plants expressing a modified bacterial merA gene. Proceedings of the National Academy of Sciences of the United States of America 1996, 93, (8), 3182-7. 5. Che, D.; Meagher, R. B.; Heaton, A. C. P.; Lima, A.; Rugh, C. L.; Merkle, S. A., Expression of mercuric ion reductase in Eastern cottonwood (Populus deltoides) confers mercuric ion reduction and resistance. Plant Biotechnology Journal 2003, 1, (4), 311-319. 6. Rugh, C. L.; Senecoff, J. F.; Meagher, R. B.; Merkle, S. A., Development of transgenic yellow poplar for mercury phytoremediation. Nature Biotechnology 1998, 16, (10), 925-928. 7. Kubachka, K. M.; Meija, J.; LeDuc, D. L.; Terry, N.; Caruso, J. A., Selenium Volatiles as Proxy to the Metabolic Pathways of Selenium in Genetically Modified Brassica juncea. Environmental Science & Technology 2007, 41, (6), 1863-1869. 8. Parizek, J.; Ostadalova, I., The protective effect of small amounts of selenite in sublimate intoxication. Experientia 1967, 23, (2), 142-3. 9. Frisk, P.; Wester, K.; Yaqob, A.; Lindh, U., Selenium protection against mercury-induced apoptosis and growth inhibition in cultured K562 cells. Biological Trace Element Research 2003, 92, (2), 105-114. 10. Naganuma, A.; Imura, N., Changes in distribution of mercury and selenium in soluble fractions of rabbit after simultaneous administration. Pharmacology, Biochemistry and Behavior 1980, 13, (4), 537-44. 11. Wang, A.; Barber, D.; Pfeiffer, C. J., Protective effects of selenium against mercury toxicity in cultured Atlantic spotted dolphin (Stenella plagiodon) renal cells. Archives of Environmental Contamination and Toxicology 2001, 41, (4), 403-409. 12. Gailer, J.; George, G. N.; Pickering, I. J.; Madden, S.; Prince, R. C.; Yu, E. Y.; Denton, M. B.; Younis, H. S.; Aposhian, H. V., Structural Basis of the Antagonism between Inorganic Mercury and Selenium in Mammals. Chemical Research in Toxicology 2000, 13, (11), 1135- 1142. 13. Yoneda, S.; Suzuki, K. T., Equimolar Hg-Se complex binds to selenoprotein P. Biochemical and Biophysical Research Communications 1997, 231, (1), 7-11. 14. Mounicou, S.; Shah, M.; Meija, J.; Caruso, J. A.; Vonderheide, A. P.; Shann, J., Localization and speciation of selenium and mercury in Brassica juncea-implications for Se-Hg antagonism. Journal of Analytical Atomic Spectrometry 2006, 21, (4), 404-412.

159

15. Yathavakilla, S. K. V.; Caruso, J. A., A study of Se-Hg antagonism in Glycine max (soybean) roots by size exclusion and reversed phase HPLC-ICPMS. Analytical and Bioanalytical Chemistry 2007, 389, (3), 715-723. 16. Afton, S. E.; Caruso, J. A., Comparing the molecular makeup of mercury in conjunction or exclusion of selenium in the Allium fistulosum, green onion. Submitted for review. 17. Arai, T.; Ikemoto, T.; Hokura, A.; Terada, Y.; Kunito, T.; Tanabe, S.; Nakai, I., Chemical forms of mercury and cadmium accumulated in marine mammals and seabirds as determined by XAFS analysis. Environmental Science and Technology 2004, 38, (24), 6468-6474. 18. Skubal, L. R.; Biedron, S. G.; Newville, M.; Schneider, J. F.; Milton, S. V.; Pianetta, P.; O'Neill, H. J., Mercury transformations in chemical agent simulant as characterized by X-ray absorption fine spectroscopy. Talanta 2005, 67, (4), 730-735. 19. Lowry, G. V.; Shaw, S.; Kim, C. S.; Rytuba, J. J.; Brown, G. E., Jr., Macroscopic and Microscopic Observations of Particle-Facilitated Mercury Transport from New Idria and Sulphur Bank Mercury Mine Tailings. Environmental Science and Technology 2004, 38, (19), 5101- 5111. 20. Takaoka, M.; Yamamoto, T.; Fujiwara, S.; Oshita, K.; Takeda, N.; Tanaka, T.; Uruga, T., Chemical states of trace elements in sewage sludge incineration ash by using X-ray absorption fine structure. Water Science and Technology 2008, 57, (3), 411-417. 21. Shah, P.; Strezov, V.; Stevanov, C.; Nelson, P. F., Speciation of Arsenic and Selenium in Coal Combustion Products. Energy & Fuels 2007, 21, (2), 506-512. 22. Shah, M.; Kannamkumarath, S. S.; Wuilloud, J. C. A.; Wuilloud, R. G.; Caruso, J. A., Identification and characterization of selenium species in enriched green onion (Allium fistulosum) by HPLC-ICP-MS and ESI-IT-MS. Journal of Analytical Atomic Spectrometry 2004, 19, (3), 381-386. 23. Nonose, N. S.; Matsuda, N.; Fudagawa, N.; Kubota, M., Some characteristics of polyatomic ion spectra in inductively coupled plasma mass spectrometry. Spectrochimica Acta, Part B: Atomic Spectroscopy 1994, 49B, (10), 955-74. 24. Afton, S. E.; Catron, B.; Caruso, J. A., Investigating the mechanisms and feasiblity for xenon as a cell gas in inductively coupled plasma mass spectrometry for multi-element detection including sulfur with application to bottled water. Submitted for review. 25. Kebarle, P.; Haynes, R. M.; Searles, S. K., Mass-spectrometric study of ions in xenon, krypton, argon, and neon at pressures up to 40 torr: termolecular formation of the rare-gas molecular ions. Bond dissociation energy of Ar2+ and Ne2+. Journal of Chemical Physics 1967, 47, (5), 1684-91. 26. Cavallini, A.; Natali, L.; Durante, M.; Maserti, B., Mercury uptake, distribution and DNA affinity in durum wheat (Triticum durum Desf.) plants. Science of the Total Environment 1999, 243/244, 119-127. 27. Wang, Y.; Greger, M., Clonal differences in mercury tolerance, accumulation, and distribution in willow. Journal of Environmental Quality 2004, 33, (5), 1779-1785. 28. Meagher, R. B.; Heaton, A. C. P., Strategies for the engineered phytoremediation of toxic element pollution: mercury and arsenic. Journal of Industrial Microbiology & Biotechnology 2005, 32, (11-12), 502-513.

160

Chapter 7

Conclusions and future directions

161

Several methodologies have been developed and utilized in this dissertation in order to speciate and monitor the interactions of selenium and environmental contaminants, arsenic and mercury, in plants.

Four selenium species, selenite, selenate, selenomethionine and selenocystine, and four arsenic species, arsenite, arsenate, monomethlyarsonate and dimethylarsinate, were simultaneously separated using ion-pairing reversed phase liquid chromatography (IPRP) coupled with inductively coupled plasma mass spectrometry (ICPMS) and electrospray ionization ion trap mass spectrometry. Using tetrabutylammonium hydroxide as the ion-pairing reagent on a C18 column, the separation and re-equilibration time were attained within 18 minutes. To illustrate the wide range of possible applications, the method was then successfully applied for the detection of selenium and arsenic species found naturally and spiked in river water, plant extract and urine matrices.

The Chlorophytum comosum, spider plant, was used to investigate the metabolism of selenium and arsenic after single and simultaneous supplementation via size exclusion chromatography (SEC) and IPRP coupled to ICPMS after a mild aqueous extraction. Selenium and arsenic species were sequestered in the roots and generally conserved upon translocation to the leaves. Selenium was directly absorbed in the C. comosum root when supplemented with

SeVI, but a combination of passive and direct absorption when supplemented with SeIV. Higher molecular weight selenium species were more prevalent in the roots of plants supplemented with

SeIV, but in the leaves of plants supplemented SeVI. When supplemented as AsIII, arsenic is proposed to be passively absorbed as AsIII and partially oxidized to AsV in the plant root.

Although total elemental analysis demonstrates a selenium and arsenic antagonism, a compound containing selenium and arsenic was not present in the general aqueous extract of the plant.

162

The optimized Xe collision reaction cell (CRC) gas flow rate for multi-element monitoring, including sulfur, is obtained via the 34S+ signal. Increasing the difference between the voltage on the extraction lens and the octopole, plus using a positive voltage between the octopole and quadrupole provided the lowest sulfur detection limits. When monitoring a standard solution comprised of 52 elements, including sulfur, the range of optimal octopole and quadrupole bias settings were altered when compared with single element sulfur detection.

Similar detection limits were acquired when comparing Xe to He or H2 CRC gas with a general trend comparable to detection levels for He. Energy discrimination is the predominant

32 + 32 + mechanism for removing the O2 interference of S , rather than charge transfer or collision- induced dissociation. Xenon is effective in CRC-ICPMS for quantifying elements in commercially available bottled water.

The Allium fistulosum, was used to investigate the metabolic fate of mercury in conjunction with or exclusion of selenium. Total elemental analysis via microwave digestion and ICPMS detection, showed that the A. fistulosum accumulated 50 times more mercury than selenium, both elements were predominately sequestered in the root fraction, and a mutual antagonism was observed. Capillary reversed phase chromatography (capRPLC) and SEC were coupled to ICPMS to investigate the plant root and leaf extracts. The data suggests a possible selenium-mercury association in a proteinaceous macromolecule which is not further metabolized upon translocation to the aerial plant regions. The selenium-mercury association and may be formed from small selenium molecules which are normally translocated to the aerial plant regions. Data from enzymatic and acid hydrolysis suggests the formation of mercury metabolites with greater robustness when supplemented in conjunction with selenium.

163

An additional study utilized the Allium fistulosum to further investigate the metabolic fate of mercury in conjunction with or exclusion of selenium. Data from x-ray fluorescence mapping of a freshly excised root and capRPLC-ICPMS of homogenized root extract, suggest the plausible formation of a mercury-selenium species and a similarly structured mercury-sulfur species predominantly residing in the cell wall of the epidermal root tissue. With x-ray absorption near edge structure analysis, the local environment of mercury and selenium qualitatively match the mercury-selenium species formed in a mammal via a Hg-Se-S(GSH) moiety. In addition, mercury also matches the local environment of (GS)2Hg, suggesting a difference in metabolism.

Figure 7.1 The proposed metabolic pathways leading to the formation of the mercury-selenium and mercury-sulfur entities in the A. fistulosum root

164

While not experimentally conclusive, the proposed metabolic pathways leading to the formation of the noted mercury-selenium and mercury-sulfur species in the A. fistulosum root can be seen in Figure 7.1. Future studies would utilize Arabidopsis thaliana for further investigation into the unique metabolism of sulfur, selenium and mercury in plants. As the genome has been sequenced for A. thaliana, genetic modifications can be made to elucidate the plausible protein responsible for the macromolecular size mercury-selenium and mercury-sulfur species observed. In addition, genetic modifications could enhance the phytoremediation capabilities of plants toward mercury and possibly other environmental contaminant elements, which can be detoxified after simultaneous addition with selenium.

165