Citation
Evaluating the mechanisms for phytoremediation of MTBE

Material Information

Title:
Evaluating the mechanisms for phytoremediation of MTBE
Creator:
Rubin, Ellen Gayle
Publication Date:
Language:
English
Physical Description:
249 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Butyl methyl ether ( lcsh )
Phytoremediation -- Evaluation ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 230-249).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Ellen Gayle Rubin.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
53906482 ( OCLC )
ocm53906482
Classification:
LD1190.E53 2003d R82 ( lcc )

Full Text
EVALUATING THE MECHANISMS FOR PHYTOREMEDIATION OF MTBE
by
Ellen Gayle Rubin
B.S., The Pennsylvania State University, 1991
M.S., The Pennsylvania State University, 1994
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Civil Engineering
2003


This thesis for the Doctor of Philosophy
Degree by
Ellen Rubin
has been approved
by
Dr. Anu Ramaswami
Dr. Ram Ramaswami


Z
Date


Ellen Rubin (Ph.D., Civil Engineering)
Investigating the Mechanisms for Phytoremediation of MTBE
Thesis directed by Associate Professor Anu Ramaswami
ABSTRACT
This thesis examines the dominant mechanisms and measures the engineering
parameters for phytoremediation of Methyl tert butyl ether (MTBE) employing bench
scale and tree scale experiments. Phase 1, hydroponic plant uptake experiments,
indicate 30% reduction in MTBE mass in water over a I-week period by small poplar
saplings. A mass balance indicated that MTBE was untransformed during transport
through the small poplar saplings and the transpiration stream concentration factor
(TSCF) was computed to be approximately 1, suggesting MTBE is actively transpired
by plants at the same concentration as in groundwater. The high TSCF and low RCF
indicated that phytovolatilization is the primary mechanism for phytoremediation of
MTBE in aqueous bench scale studies.
Phase 2 investigated the potential for microbial degradation of MTBE in the
rhizosphere of poplar trees. Ail studies indicated that MTBE degradation in plant
rhizosphere occurs minimally, if at all, and at rates that are very slow and
insignificant compared to the rapid rate of plant-assisted transpiration of MTBE
shown in phase 1.
Since the first two phases indicated that phytovolatilization is the primary mechanism
for MTBE removal from contaminated sites. Phase 3 focused on measuring and
predicting evapotranspiration (ET) of large 12 poplar trees since. Results inferred
that the Penman-Monteith equation can provide reasonable ET rates for evaluation of
chemical uptake during phytovolatilization in the field.
The last phase, Phase 4, researched the fate and transport of MTBE in large (12)
trees in an enclosed chamber allowing for MTBE mass balances to be completed. A
deficit of MTBE mass was observed in replicate experiments in a short time period
(2-weeks) indicating degradation was occurring within the mature tree. More
significantly, /er/-butyl alcohol a degradation product of MTBE was detected in
increasing amounts as MTBE was detected in decreasing amounts in the leaves as the
experiment progressed, further signifying MTBE degradation was occurring.
in


In conclusion, high uptake rates of MTBE by the poplar tree and degradation within
the poplar tree makes phytoremediation of MTBE a promising technology.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Signed
Anu Ramaswami
IV


DEDICATION
This thesis is dedicated to my parents whom I have the utmost of respect and
gratitude for. Without their guidance and endless support I could never have
accomplished this goal. When life got hard it is their words I heard that kept me
going. They told me I could do anything in life and made sure I believed it.
My parents have shown me through their lives that hard work and determination
can lead to great success. This has given me the courage to undertake challenges
and risks in life that have been the most rewarding experiences to me.
To my parents who have given me everything in life, words could never thank
you enough, but I hope that my actions in life make you proud and show you how
thankful I truly am to you.


ACKNOWLEDGEMENTS
I would like to extend my sincere thanks to my advisor, Anu Ramaswami, her
participation went above and beyond the requirements of an advisor and I am very
thankftil for her never ending support. The wealth of knowledge that Anu shared with
me along with advice not only helped me as an engineer but also as a person.
I would also like to thank all of my committee members, Mark Hernandez, Angela
Bielefeldt, Ram Ramaswami, and Greg Cronin for their help with my thesis.
Special thanks for my binding from:
The National Science Foundation 1998 POWRE Grant award number
9806228.
The University of Colorado-Denver Faculty Grant
Colorado Commission of Higher Education (CCHE)
Other people I would like to acknowledge are:
Vince Marti and Bo Meyers of the EPA Region 8 Laboratory for the many
samples they analyzed on behalf of my research.
Terry Buck of AHEC for helping with problems with the tree chamber.
The Calibration Laboratory personnel especially Ed Moss for endlessly
helping with broken equipment in the laboratory.
Larry Anderson and Jeff Boon of the Environmental Science Department
My supervisor and all my co-workers for their support especially Art
Palomares, Tami Thomas-Burton, Ron Rutherford, and Martin Hestmark.
Mehmet Iselyen, Kendra Morrisoa and Dale Brophy for their friendship and
help in the laboratory and all the other students in the laboratory.
My family for their encouragement especially Esther Rubin, Louis Rubin,
Julia Rubin and Matthew Rubin.
Bayard Yang and Mehmet Isleyen for the most entertaining car pool
experience.
Endless phone calls of support from Nancy Buck, Marissa Stein, and Michelle
Large.
Shelly Baxter for our walks, talks, and Fridays.
Brandon McDowell for his support and encouragement of this endeavor.
Jamie Clapper and Melissa Shapiro for Tuesday dinners.
Mindee Mitchell for adding some culture to my life.


TABLE OF CONTENTS
Figures..................................................................xiii
Tables...................................................................xviii
Chapter
1. Literature Review.....................................................1
1.1 Introduction..........................................................1
1.2 Methyl Tertiary Butyl Ether...........................................1
1.2.1 Physical Characteristics of MTBE......................................2
1.2.2 Occurrence of MTBE in Groundwater.....................................2
1.2.2.1 MTBE Releases.......................................................2
1.2.3 Health Effects of MTBE................................................5
1.3 Phytoremediation......................................................7
1.3.1 Advantages of Phytoremediation.......................................8
1.3.2 Disadvantages of Phytoremediation....................................9
1.3.3 Costs of Phytoremediation............................................9
1.4 Poplar Trees.........................................................10
1.5 Mechanisms o f Phytoremediation......................................12
1.5.1 Phytoextraction and Phytovolatilization..............................13
vii


16
18
18
20
20
21
21
21
22
24
24
24
25
26
26
27
27
27
29
30
1.5.2 Phytodegradation in the Rhizosphere..................
1.5.3 Phytodegradation within the Poplar Tree..............
1.6 Upscaling............................................
2. Objectives and Experimental Design...................
2.1 Phase 1..................................
2.2 Phase 2..................................
2.3 Upscaling............................................
2.4 Phase 3..................................
2.5 Phase 4..................................
3. Phase 1 Fate and Transport of MTBE in Poplar Cuttings
3.1 Introduction.........................................
3.2 Plant Uptake Studies.................................
3.3 MTBE Uptake by Vegetation............................
3.4 Experimental Design..................................
3.4.1 Objectives...........................................
3.4.2 Materials............................................
3.4.3 Plant Preparation....................................
3.4.4 Experiment 1.........................................
3.4.5 Experiment 2.........................................
3.4.6 Experiment 3.........................................
vm


31
31
31
36
39
41
41
43
43
43
45
49
49
51
55
57
59
60
61
61
64
3.4.7 Analytical Methods.............................................
3.5 Results.......................................................
3.5.1 Experiment 1..................................................
3.5.2 Transpiration Stream Concentration Factor Calculations........
3.5.3 Experiment 2..................................................
3.5.4 Root Concentration Factor Calculations........................
3.6 Discussion....................................................
4. Phase 2 Degradation of MTBE in the Rhizosphere of Poplar Trees
4.1 Introduction....................................................
4.2 The Rhizosphere...............................................
4.2.1 Degradation in the Rhizosphere................................
4.3 MTBE Degradation..............................................
4.3.1 Anaerobic MTBE Degradation....................................
4.3.2 Aerobic MTBE Degradation........................................
4.3.3 Cometabolie MTBE Degradation..................................
4.3.4 MTBE Degradation Field Studies..................................
4.3.5 MTBE Degradation Summary......................................
4.4 Experimental Procedures..........................................
4.4.1 Experiment 1: Soil Die-Away Studies...........................
4.4.2 Experiment 2: Chamber Studies.................................
4.4.3 Experiment 3: Aerobic Batch Biodegradation Experiments........
IX


4.5 Results...........................................................67
5. Phase 3 Evapotranspiration of Poplar Trees......................72
5.1 Introduction......................................................72
5.1.1 Evaporation.......................................................73
5.1.2 Transpiration.....................................................75
5.1.3 Stomata...........................................................76
5.1.4 Resistances.......................................................79
5.2 Evapotranspiration Models.........................................81
5.2.1 Introduction......................................................81
5.2.2 Direct Flux Measurements..........................................82
5.2.3 Empirical Equations...............................................83
5.2.4 Combination Method................................................85
5.3 Estimation of Parameters for the Penman-Monteith Equation.........99
5.3.1 Estimating Volumetric Water Flow Rate from Penman-Monteith ET.......
Estimates........................................................99
5.4 Validation of the Penman-Monteith Equation........................99
5.5 Measurement of Evapotranspiration................................103
5.5.1 The Lysimeter....................................................104
5.5.2 Satellite Remote Sensing.........................................104
5.5.3 Porometer........................................................104
5.5.4 Flow 2...........................................................104
x


5.5.5 TDP-32............................................................106
5.6 Calibration of Evaporation Measuring Devices.....................107
5.6.1 Flow 2...........................................................107
5.6.2 TDP-32...........................................................107
5.7 Transpiration in an Enclosed Chamber.............................108
5.8 Experimental Methods.............................................108
5.9 Results..........................................................113
6. Phase 4 Degradation of MTBE within a Large Tree..................125
6.1 Introduction.....................................................125
6.2 Mechanisms for Organic Degradation in Large Trees................126
6.2.1 Chemical Transport in a Woody Tree...............................127
6.2.2 Metabolism of Chemicals within Plants/Trees......................131
6.2.3 Sorption of Contaminants to Wood.................................135
6.3 Experimental Design..............................................138
6.3.1 Materials........................................................138
6.3.2 Methods..........................................................141
6.3.3 Analysis Methods.................................................144
6.3.4 Controls ........................................................145
6.4 Results..........................................................148
6.4.1 Follow up Experiments............................................158
7. Summary and Conclusions..........................................161
XI


7.1 Phase 1: - Fate and Transport of MTBE in Poplar Cuttings..........161
7.2 Phase 2: - Degradation of MTBE in the Rhizosphere of Poplar Trees... 162
7.3 Phase 3: - Evapotranspiration of Poplar Trees.....................163
7.4 Phase 4: - Degradation of MTBE within a Large Tree................164
Appendix
A. Phase 1 Articles......................................................167
B. Phase 2 Articles......................................................194
C. Calibration Curves and Raw Data for Phase 4...........................219
References................................................................230
xii


LIST OF FIGURES
Figure
1.1 Aerobic Biodegradation Pathway of MTBE...............................4
1.2 Decision Tree for the Use of Phytoremediation for a Remedial Technology. 11
1.3 Schematic Showing Various Mechanisms for Phytoremediation...........12
1.4 Root Concentration Factor vs. Log K<,w for Various Chemicals in Poplars
and Barley...........................................................14
1.5 Transpiration Stream Concentration Factor vs. Log K<,w for Various
Chemicals for Poplars and Barley.....................................15
2.1 The Phased Approach to Assess Phytoremediation of MTBE..............23
3.1 Drawing of Poplar Cutting inside 25-L Chamber.......................28
3.2 Small Cutting Uptake Study with Controls............................29
3.3 Small Cutting Mass Balance Experiment...............................30
3.4 Water Loss in Various Systems (Initial Aqueous MTBE Concentration =
1600 ppb)............................................................34
3.5 Final Aqueous MTBE Concentrations Various Experimental Systems
(Initial Aqueous MTBE Concentration = 1600 ppb)...................... 34
3.6 Percent Reduction in MTBE Concentration and Mass (Initial Aqueous
MTBE Concentration = 1600 ppb)
35


3.7 MTBE Concentration in Biomass (Initial Aqueous MTBE Concentration
= 1600 ppb)........................................................ 35
3.8 MTBE Mass in Biomass (Initial Aqueous MTBE Concentration^600 ppb) 36
3.9 Volume of Water Transpired vs. Leaf Surface Area....................38
3.10 Percent Mass Reduction of MTBE vs. Volume of Water Transpired....38
3.11 MTBE Mass Balance (Initial Aqueous MTBE Concentration^600 ppb).. 40
4.1 Drawing of Poplar Cutting Inside 25-L Chamber.......................62
4.2 Photo of Plant in Chamber .........................................63
4.3 Photo of Plant in Soil, Plant in Water, and Control................64
4.4 Picture of Experiment 3............................................66
4.5 Attenuation of MTBE Concentrations in Unvegetated Soils and Those
Planted with Poplar Trees...........................................68
4.6 Aqueous MTBE Concentrations in Control Bioreactors (no soil) and
Rhizosphere Bioreactors.............................................70
4.7 Dissolved Oxygen Concentrations in Control Bioreactors (no soil) and
Rhizosphere Bioreactors.............................................70
4.8 Low Aqueous MTBE Concentrations in Controls (no soil) and
Rhizosphere Bioreactors.............................................71
4.9 High Aqueous MTBE Concentrations in Controls (no soil) and
Rhizosphere Bioreactors.............................................71
5.1 The Hydrologic Cycle................................................73
5.2 Evapotranspiration Models...........................................74
5.3 Temperature vs. Evaporation.........................................75
XIV


5.4a Release of Water from the Leaf.
76
5.4b The Guard Cell....................................................77
5.5a Resistances.......................................................79
5.5b Stomata Resistance of the Leaf vs. Light Intensity................80
5.6 Inputs and Implementation for the Penman-Monteith Equation.......100
5.7 ET Estimated by the Penman-Monteith Method vs. Average Monthly
Lysimeter ET Measurements........................................101
5.8 Comparison of Time Course of Transpiration Rates Predicted by Penman-
Monteith Equation with Measured Values...........................102
5.9 Flow 2 Diagram....................................................105
5.10 Photo of Large Tree Inside Chamber..............................110
5.11 Comparison of Dynamax Flow 2 and TDP-32........................111
5.12 Comparison of Dynamax Flow 2 vs. TDP-32 Unfavorable Transpiration. 122
5.13 Comparison of Dynamax Flow 2 vs. TDP-32 Favorable Transpiration.123
5.14 Dynamax TDP-32 vs. the Penman and Penman-Monteith Unfavorable
Transpiration....................................................123
5.15 Dynamax TDP-32 vs. the Penman and Penman-Monteith Favorable
Transpiration....................................................124
6.1a Degradation Pathways in Vegetation...............................126
6.1b An Example of Plant Metabolism of 2,4 Dichlorophenoxyacetic Acid in
Soybean and Pentachlorophenol in Wheat Suspension Cells.........127
6.2 The Mature Tree...................................................128
6.3 The Phloem and Xylem..............................................129
XV


6.4 Movement of Chemicals in the Phloem and Xylem.....................131
6.5 The Tree Cell.....................................................133
6.6 Lignin-Octanol Linear Free Energy Relationship for Wood Particles.137
6.7 Photo of Small Chamber Compared to Large Chamber..................139
6.8 Schematic of Big Tree Chamber.....................................140
6.9 Photo of Ductwork in Large Chamber.................................141
6.10 Photo of Large Tree in Chamber...................................142
6.11 Mass Balance for Control 1.......................................146
6.12 Mass Balance for Control 2.......................................147
6.13 Measured Volume ofMTBE Transpired vs. Expected Volume of MTBE
Transpired for the First Experiment...............................150
6.14 MTBE Mass Balance for the First Experiment.......................151
6.15 MTBE Detected (ppb) in Vegetation Samples for Experiment 1.......152
6.16 Measured Volume ofMTBE Transpired vs. Expected Volume ofMTBE
Transpired for the Second Experiment..............................154
6.17 MTBE Mass Balance for the Second Experiment......................156
6.18 MTBE Detected in Vegetation Samples for Second Experiment........156
6.19 TBA Detected in Vegetation Samples for Second Experiment.........157
6.20 MTBE and TBA Detected in Vegetation Samples for Second
Experiment........................................................157
6.21a MTBE Sorption to Wood............................................158
6.21b TBA Sorption to Wood.............................................159
xvi


7.1 Results to the Phased Approach to Assess Phytoremediation of MTBE 166
!
i
i
XVI1


..3
..5
..6
..7
..8
17
32
32
40
41
44
47
50
54
56
59
68
78
LIST OF TABLES
1.1 MTBE Fate, Transport and Persistence Information..........
1.2 MTBE Characteristics and Properties.......................
1.3 MTBE Drinking Water Standards (ug/L)......................
1.4 Costs of Phytoremediation vs. Other Remedial Technologies.
1.5 Phytoremediation Applications.............................
1.6 Factors that Effect the Rhizosphere.......................
3.1a Mass Balance without the Presence of Poplar Trees........
3.1b Mass Balance Continued without the Presence of Poplar Trees
3.2 MTBE Mass Balance........................................
3.3 Measured vs. Predicted TSCF and RCF Values for MTBE......
4.1 Compounds Detected in Root Exudates.......................
4.2 Rhizosphere Degradation Applications.....................
4.3 Anaerobic Degradation of MTBE Found in Literature........
4.4 Aerobic Degradation of MTBE Found in Literature..........
4.5 Cometabolic Degradation of MTBE Found in Literature......
4.6 Degradation of MTBE in the Field Found in Literature.....
4.7 MTBE Mass Balance........................................
5.1 Typical Values of Transpiration Rates of Trees............
XVlll


5.2 Average Values of Stomata Resistance with Low Amount of Wind,
Cuticular Resistance, and Boundary Layer Resistance..................80
5.3 Summary of Evapotranspiration Models................................88
5.4 Examples of Roughness Lengths for Various Surfaces and Associated
Ratios of zom/h........................................................91
5.5 Resistances...........................................................93
5.6 Cuticular Resistance re and Minimum Leaf Resistance rlmin to
Transpiration Measured by Vapor Transfer Methods.......................97
5.7 Evapotranspiration Measurement Devices...............................103
5.8 Calibration of Flow 2................................................107
5.9 Meteorological Conditions for Phase 3................................112
5.10 Inputs to Calculate Penman and Penman-Monteith Equations and
Comparison with Flow 2 and TDP-32 for Unfavorable Transpiration
Conditions......................................................115
5.11 Flow 2 Calculations for the Unfavorable Transpiration Conditions...119
5.12 Example of the Excel Program for the TDP-32 Probes for Unfavorable
Transpiration........................................................120
5.13 Average Transpiration for the Poplar Trees..........................122
6.1 Known Plant Enzymes that Transform Organic Compounds..................134
6.2 Pant Enzyme Classes Known to Metabolize Specific Contaminants.......135
6.3 Wood Partition Coefficient for MTBE..................................137
6.4 MTBE-Water Solution Given to (12) Lance leaf Poplar Tree in the First
Experiment ...........................................................143
xix


6.5 MTBE-Water Given to (12) Noreaster Poplar Tree in the Second
Experiment........................................................144
6.6 Error Analysis...................................................145
6.7a Control 1 Mass Balance...........................................146
6.7b Control 2 Mass Balance...........................................147
6.8 MTBE in Transpiration Stream in Large (12) Lanceleaf Poplar for
Experiment 1......................................................149
6.9 MTBE Mass Balance for Large (12) Lanceleaf Poplar Tree for
Experiment 1 .....................................................150
6.10 MTBE in Transpiration Stream in Large (12) Noreaster in Second
Experiment.......................................................153
6.11 MTBE Mass Balance for Large (12) Noreaster Poplar Tree for Second
Experiment.......................................................155
XX


1. Literature Review
1.1 Introduction
In December of 1921 Thomas Midgley discovered that tetraethyl lead (TEL) could be
added to gasoline to reduce engine knocks and subsequently in 1923 TEL was being
manufactured. When manufacturing began, experts in the field of public health
expressed extreme concern over TEL and its poisonous characteristics. At the same
time research was being conducted on ethanol blended gasoline. In 1925, Charles
Kettering announced a new fuel. which was a mixture of alcohol and gasoline that
would double the gas mileage. However, the oil companies preferred TEL to ethanol
because the use of ethanol would reduce the gas consumption by 20-30% (Nadin,
2001). In 1970, after 50 years of leaded gasoline use, the Clean Air Act (CAA)was
signed and began the phase out of leaded gasoline. The use of oxygenates began in
1979, when TEL was substituted for Methyl-Tertiary-Butyl-Ether (MTBE) to
increase the octane rating. In 1981, the Environmental Protection Agency (EPA)
allowed 7% MTBE into gasoline. EPA's oxyfuels program mandated by the 1990
CAA introduced MTBE into mass-scale consumption as a gasoline additive to
improve fuel burning efficiency to reduce carbon monoxide emissions (NRC, 1996).
More specifically, in 1992. wintertime gasoline with 15% MTBE was introduced in
some cities; And in 1995, year round use of gasoline with 11% MTBE was
implemented in the smoggiest of cities.
1.2 Methyl Tertiary Butyl Ether
MTBE is manufactured by reacting methanol on isobutene produced by catalytic
cracking, stream cracking, or butane dehydrogenation (Fayolle, 2001). Though there
are many oxygenates such as ethyl tert butyl ether (ETBE), tert amyl methyl ether
(TAME), and alcohols like ethanol, MTBE is the most widely used oxygenate
(Ahmed, 2001). MTBE is a well liked oxygenate due to its low cost, ease of
production, and favorable transfer and blending characteristics (Squillance, 1996). In
the USA, winter oxyfuels program gasoline must contain 2.7% oxygen and
reformulated gasoline must contain 2.0% oxygen on average; this corresponds to a
MTBE content of 15% and 11% volume respectively. MTBE is also added to reach
the required octane number (Fayolle, 2001).
1


1.2.1 Physical Characteristics of MTBE
MTBE differs drastically from other gasoline hydrocarbons (i.e. benzene, toluene,
ethyl benzene, and xylene) in physical characteristics such as water solubility, vapor
pressure, Henrys Law Constant, and adsorption coefficient. Since MTBE is very
water-soluble and does not sorb well to soil, large MTBE plumes have been detected
nationwide. It is also essentially non-biodegradable and non-reactive in water (Suflita
et al., 1993), making MTBE a persistent and prevalent groundwater contaminant.
Once MTBE dissolves into the groundwater it is expected to travel at the same
velocity as the groundwater (EPA, 1998). Refer to Table 1.1 for the fate,
transformation, and persistence information on MTBE.
MTBE degradation is still under investigation and is fully addressed in Chapter 4.
However, current literature seems to suggest that MTBE will more likely degrade in
aerobic conditions to rerf-butyl-alcohol (TBA), tert-butyl-formate (TBF),
formaldehyde, and ultimately carbon dioxide (Steffen et al, 1997). The degradation
pathway is presented in Figure 1.1. Refer to Table 1.2 for a list of physical
characteristics of MTBE and TBA.
1.2.2 Occurrence of MTBE in Groundwater
MTBE was the second most prevalent contaminant detected out of 60 volatile organic
compounds (VOCs) analyzed for in urban areas during 1993-1994 as part of the
United States Geological Survey (U.S.G.S.) National Quality Assessment program.
Detections ranged from below a reporting limit of 0.2 pg/L to over 20,000 pg fL.
They also reported that in 1994, 6.2 billion kilograms (kg) of MTBE was produced in
the United States and between 1984 and 1994 there has been a 26% increase annually
of MTBE production (Squillance, 1996).
1.2.2.1 MTBE Releases
Releases can be from point sources such as direct release to ground and surface water
from underground storage tanks (USTs), pipelines, and tank cars or from non-point
sources such as vehicular or production emissions that dissolve in precipitation and
subsequently infiltrate to groundwater. Releases of MTBE from production sources in
the United States alone were approximately 1.7 million kg in 1996 with 97% being
released to air and 3 % to surface water.
In 1996, Santa Monica, CA had the most publicized spill of MTBE where upon levels
of 600 pg/L were detected in drinking water causing closure of half of the municipal
2


water supply (Fayolle, 2001). Other large releases have also occurred. For example,
approximately 11,000 gallons of gasoline from a tanker truck was released into the
Merrimack River, MA causing such elevated levels of MTBE that the water treatment
plants had to temporarily close (Fiorenza, 2002).
Happel et al (1998) analyzed data from 263 leaking USTs in California and found
that MTBE was detected at 78% of these sites. It is estimated that 7% of groundwater
resources could be contaminated by MTBE at concentrations higher than 0.2 pg/L in
the United States. Also, surface water could be polluted from recreational powered
water transport (Fayolle, 2001).
Due to the many reported releases of MTBE in the air and water and detections
nationwide in groundwater monitoring wells, irrigation wells and municipal supply
wells (Zogorski et al., 1997) the health effects of MTBE are being evaluated as
discussed next.
TABLE 1.1 MTBE Fate, Transformation and Persistence Information (Howard et al., 1991)
Air MTBE is not expected to persist in the atmosphere because of its rapid reaction with hydroxyl radicals. The half-life is in the order of 11 days.
Water MTBE is expected to volatilize from surface waters. Volatilization half-live of MTBE from surface water was established to range from 4 weeks to 6 months and from groundwater from 8 weeks to 12 months.
Soil MTBE is expected to volatilize rapidly from soil surfaces. However, MTBE resists degradation in soil. The half-life ranges from 4 weeks to 6 months.
Biota Bioconcentration factors indicate a low potential for bioconcentration
3


CH,
H3C-C-O-CH3
CH,
Meihvl ;en-butyl eiher i.MTBE)
O..
CvliH'hrome P-4?0

CH^
ten-Butyl for male tTBF)

-<0
O'
Formate

OH
i'
H,0


CH, OH
h,c-c-o-ch,
CH,
/en-Butoxy methanol

CHjO
Formaldehyde
CH,
H,C-C-OH
CH,
ten- Butyl alcohol (TBA)
CH,
H^C-C-CHj-OH
OH
2-Melhvl-2-hydroy-1-propanol (T)
CH,
h,c-c-c:
OH
,0
OH
2-Hyriroxy isobutvnc acid

Fun her
metabolism
FIGURE 1.1 Aerobic Biodegradation Pathway of MTBE
4


TABLE 1.2 MTBE Characteristics and Properties (USEPA, 1994)
MTBE TBA
CAS No. 1634-04-4 76-65-0
Molecular Formula c5h12o C4H10O
Physical State Colorless Liquid
Molecular Weight 88.15 g/mole 74.12 g/mole
Melting Point -109 C
Boiling Point 55.2 C 82.4 C
Water Solubility 51.26 g/L (a). 25 C Miscible
Density 0.74 g/mole 0.79 g/mole
Vapor Pressure 245 mm Hg @ 25 C 41 mm Hg (5) 25 C
Flash Point Flammable
Log Kow 1.24 0.35
Log IQ 1.035-1.091
Henry's Law Constant 0.02 dimensionless @ 25 C 5.03 x 10^ dimensionless 25 C
Taste threshold 10 -130 pg/L
Fish B ioconcentration Factor <2 (measured), <4 (estimated)
1.2.3 Health Effects of MTBE
Inhalation of MTBE is the most likely human exposure route (Ahmed, 2001).
Complaints about headaches, dizziness, irritated eyes, disorientation, and nausea were
reported from residents in Fairbanks and Anchorage, Alaska, Missoula, MT, and
Milwaukee, WI when MTBE was introduced into their fuel (Squillance, 1996).
However, more controlled inhalation studies indicated that a correlation between
MTBE and complaints were inconclusive. They hypothesized that the conflicting
results may be due to: 1.) media exposure skewing data, 2.) specific individuals being
hypersensitive to MTBE, and 3.) the unpleasant odor may cause an increase in
symptom awareness (Ahmed, 2001).
5


Health effects and toxicity have been evaluated by both oral and inhalation routes in
male rats with the primary target organ of MTBE being the kidney and testes. The
inhalation route in female mice has been assessed with the results showing the
primary organ affected was the liver (Ahmed, 2001). Even though the risk analysis for
MTBE is not complete, EPA has established MTBE as a possible carcinogen, for
which the EPA Drinking Water Provisional Health Advisory is 20-40 pg/1 (USEPA,
1997). Many states have also developed there own drinking water standard for MTBE
shown in Table 1.3.
TABLE 1.3
MTBE Drinking Water Standards (pg/L)
(Day et al. 2002)
USEPA 20-40
California 13
Delaware (proposed) 10
Maine 35
Massachusetts 70
New Hampshire 13
New Jersey 70
New York 10
Wisconsin 60
Researchers have experimentally assessed the toxicity of MTBE to freshwater
organisms. Results from these experiments have indicated that MTBE concentrations
of 57 -1000 mg/L (invertebrates) and 388-2600 mg/L (vertebrates) are toxic to
various aquatic organisms; luckily the current surface water MTBE levels (0.1 mg/1)
are not at levels that would be toxic to aquatic life. Also, results indicated that MTBE
does not appear to bioaccumulate in fish (Werner et al, 2001). Likewise, Mancini et
al (2002) reported toxicity for freshwater and marine life to be below the current
levels of MTBE.
The possible health effects and significant occurrence of MTBE has necessitated
remedial technologies. However, once MTBE gets into groundwater, an MTBE
plume develops rapidly which is difficult to remediate by conventional methods.
More information is needed on cost-effective techniques, such as phytoremediation,
to treat MTBE at contaminated sites.
6


13 Phytoremediation
Phytoremediation is an emerging, innovative, "green" technology that employs plant
systems for environmental remediation and restoration at waste contaminated sites.
It is an aesthetically-pleasing, passive, solar-energy driven cleanup technique (EPA,
1998). There are about 24,000 abandoned hazardous waste sites recorded by the
Environmental Protection Agency (EPA) and almost all of them coexist with plants
(Chang et al., 1998). Plants have grown at hazardous waste sites naturally or planted
for stabilization or aesthetic value. The life cycle of plants is known to drastically
differ the immediate vicinity biologically, chemically, and physically (Cunningham,
1996). However, not until recently have plants gained attention for the possibility as a
remediation technique. It is estimated that the cost of cleaning-up the tens of
thousands of wastes sites in the United States alone will surpass $700 billion dollars
(Revkin, 2001). This cost has been the catalyst for developing emerging technologies,
such as phytoremediation, that are capable of cleaning up waste sites economically.
Refer to Table 1.4. Phytoremediation has been successfully implemented in the field
to treat some volatile organic compounds (VOCs), but has not yet been evaluated for
MTBE. See Table 1.5. The focus of this thesis is to examine the potential of
phytoremediation of the fuel oxygenate MTBE. The advantages and disadvantages of
phytoremediation are discussed next.
TABLE 1.4 Costs of Phytoremediation vs. Other Remedial Technologies
Type of chemical waste Phytoremediation Other Technologies Reference
Metals $80.00 per cubic yard $250.00 per cubic yard (Black, 1995)
Site contaminated with petroleum hydrocarbons $70,000 $850,000 (Plummer, 1997)
10 acres lead contaminated land $500,000 $12 million (Richman, 1997)
7


TABLE 1.5 Phytoremediation Applications (USEPA, 2000)
Mechanism Contaminant Media Plant Reference
Degradation Atrazine, nitrates Surface water Poplar Schnoor, 1995
Degradation Landfill leachate Ground water Poplar Licht, 1990
Degradation TCE Ground water Poplar cottonwood Rock, 1997
Degradation TNT Wetlands Various Bader, 1996, Camera, 1996, McCutcheon, 1995
Degradation TPH Soil Grasses, crops Banks, 1997, Drake, 1997
Extraction Lead Soil Indian mustard Blaylock, 1997
Extraction Uranium Surface water Sunflower Dushenkov, 1997
Extraction Selenium Soil, surface water Various Banuelos, 1996, Terry, 1996
1.3.1 Advantages of Phytoremediation
There are many advantages to phytoremediation as can be seen by the bulleted list
below:
Minimized emissions and effluents
Controls erosion, runoff, rain infiltration, and dust emissions
Promotes biodiversity
Creates habitats
Sequesters greenhouse gases (CO2)
Acceptable Brownfields applications
Aesthetics
Green technology
8


Increased regulatory approval
Low maintenance
Passive
Solar powered, energy efficient
Can be utilized for multiple and mixed contamination and mediums
In-situ
Mineralization can occur
1.3.2 Disadvantages of Phytoreinediation
As with all remedial technologies there are disadvantages as can be seen by the
bulleted list below.
Toxicity to plant
Plant biomass may have to be harvested
Metal hyperaccumulators are generally slow-growing
Limited by root depth
Time consuming, slow
Additional fertilization needs
Susceptible to weather and infestation
Toxic intermediates or degradation products may be formed
Contamination could be released to the atmosphere through the plant
Accumulation of contamination in plants could be passed along the food
chain.
See Figure 1.2 for the decision tree that evaluates if phytoremediation of groundwater
is a suitable remedial technology for the field condition being assessed (USEPA,
1999).
1.3.3 Costs of Phytoremediation
In the United States the costs of remediation is astronomical, with an estimate of
surpassing 700 billion dollars for the tens of thousands of contaminated sites that
need to be cleaned-up (Revkin, 2001). So far, 410 Superfund Sites (32%) on the
National Priority List (NPL) have been remediated of hazardous waste to levels safe
for human health and the environment. The most common technologies used in these
clean-up projects was excavating and removing hazardous soil and solid waste (45%),
covering the landfill with a protective cap (39%) and pumping and treating
contaminated groundwater (34%). These technologies are very costly. Cost estimates
for excavation and disposal range from $270.00 to $460.00 per ton depending on the
9


nature of hazardous materials and methods of excavation. Approximate industry costs
for capping a contaminated site are $175,000 to $225,000 per acre (www.Jrtr.gov).
Not only are these two technologies costly they do not eliminate the contamination,
but move the waste in an area that has no access to the public. Actual costs of
pumping and treating a chlorinated solvent, volatiles, and selenium contaminated site
was $27,600,000, which corresponds to $23.00 per 1000 gallons of groundwater
extracted and $64.00 per pound of contaminant removed.
Estimates of costs for phytoremediation of a one acre site, including site preparation,
planting, and removal (harvest) of plant material, range from $2000.00 to $5000.00
(Phytokinetics). US AEC estimated that the cost for phytoremediation of one acre of
lead-contaminated soil to a depth of 50-cm was $60,000 to $100,000, whereas
excavating and land filling the same soil was $400,000 to $1,700,000. Growing a
green crop on an acre of land can be completed significantly less (2-4 orders of
magnitude) than excavation and reburial (Cunningham, 1996).
One out of many success stories for phytoremediation will be presented next. A
phytoremediation company used sunflowers and Indian mustard to remediate lead-
contaminated soil in Detroit. The lead contamination was reduced by 43% with a
project cost of $900,000. It was estimated that the costs of hauling off the 5,700 cubic
yards of lead-contaminated soil would have been more than a million dollars (Revkin,
2001). Table 1-2 compares some of the costs of other remedial technologies to
phytoremediation.
1.4 Poplar Trees
Poplar trees are typically used in phytoremediation of organic pollutants because they
are long lasting (between 25 and 50 years), fast growing, hardy, and transpire large
quantities of water. Poplar trees can grow six eight feet per year, reaching heights of
30 feet depending on species. For first two years of the tree life the expected
transpiration could be 200 gallons per tree per year. Grown poplars can uptake up to
100 liter per day of groundwater (Sutherson, 1997).
Trees can uptake water from the top of the saturated aquifer. As in a pump and treat
system, evapotranspiration due to the tree will draw down the water table in the areas
below the tree. However, a disadvantage of phytoremediation is that the roots must be
able to reach the contaminated groundwater for remediation. Therefore, making
phytoremediation an unfeasible remedial technology for deep contaminated aquifers.
Some companies such as Treemediation have patented systems to treat deep
contaminated soil and groundwater.
10


Decision Tree for Phytoremediation
Groundwater
YES I VUl the climate support th6 proposed plants? I NO
NO |ls time or space a constraint^ YES
1s the contaminant physically within the range of the proposed plant
(typically less than 10-20 feet bgs for Sahx species willows, cottonwoods, poplars)?
Will the plants be used for hydraulic
control ONLY (prevent water from
REACHING the contaminated zone)?
WH the water be mechanically pumped and
applied to the phytoremediation system?
is the contaminant at phytotoxic-concentrations
(this may require a greenhouse dose-responseiest)'
Will stale regulations allow
this type of phytoremediation?
VWI the rhizosphere microbes and plant-exuded enzymes degrade the target
contaminants in the rhizosphere and are the metabolic products acceptable?
W
ts (he log Kow ot the contaminant or metabolic
products between 1 and 3.5 (will uptake occur)?
VWI the plant degrade the
contaminant after uptake and are
the metabolic products acceptable?
J Will the plants transpire the
~j nq | contaminant or metabolic products?
Is the quantity and rate of transpiration acceptable for this site? NO
YES

\ NO Wl! the plant accumulate the contaminant or metabolic products after uptake? YES


Can controls be put in place to prevent
the transfer of the contaminant or metabolic
products from a plant to humans/anlmaW?
YES I Can engineering controls make it acceptable?! NO
NO I
YES tor this site throughout the growth of the plant' T
p
is the final disposition of the contaminant
or metabolic products accepts Die?

pEf~
NO I Does the plant material constitute a waste if harvested?! YES
£
YES j Can the plant waste be economically disposed? j NO
Can the contaminant or metabolic product
be immobilized to acceptable levels?

Phyioremeoiation has the potentia'
beetieci've at the

Phytoremediation is NOT an option
ai the site consider other options
FIGURE 1.2 Decision Tree for Use of Phytoremediation for a Remedial
Technology
11


1.5 Mechanisms of Phytoremediation
Phytoremediation of organic pollutants primarily occurs by four mechanisms: 1.)
Phytoextraction: the uptake and translocation of pollutants from groundwater into
plant tissue; 2) Phytovolatilization: the transfer of the pollutant to air via the plant
transpiration stream; 3). Rhizosphere Degradation: the breakdown of organic
pollutants within the microbe-rich rhizosphere; and 4). Phytodegradation: the
breakdown of organic pollutants within the tree. See Figure 1.3 for a depiction of the
phytoremediation mechanisms.
Atmospheric VOC
transport and degradation
VOC DEGRADATION WITHIN TREE
VOC UPTAKE VTA PLANT TRANSPIRATION
Groundwater Flow
FIGURE 1.3 Schematic Showing Various Mechanisms for Phytoremediation
12


1.5.1 Phytoextraction and Phytovolatilization
Uptake of volatile organic contaminants (VOCs) by poplar trees depends on uptake
efficiency for the specific VOC and tree transpiration rate. Plant uptake of organic
compounds is found to depend upon the contaminant's hydrophobicity, quantified by
its octanol-water coefficient KoW. The process is most effective for pollutants with
log Kow in the optimum range of 0.5 to 3.0 (Briggs et al, 1982 and Burken and
Schnoor, 1998). When the log Kow is greater than 3, the contaminant is sorbed so
strongly to the root surface that the chemical is not translocated into the shoots. In
contrast, chemicals with a log Kow less than 0.5 are very water-soluble and will stay in
the water and not be sorbed to roots and therefore not transported through the plant.
The transpiration stream concentration factor (TSCF) and the Root Concentration
Factor (RCF) are two critical parameters in determining the uptake and transpiration
of organics by plants/trees. The RCF addresses sorption of organic pollutants from
soil to the lipophillic root tissues. Equation 1.1 defines the RCF.
concentrat ion of pollutant in roots
RCF =____________________________________________________ Equation 1.1
concentrat ion of pollutant in external solution
The RCF can be estimated from equation 1.2. The RCF increases as the pollutant's
Kow increases. Refer to Figure 1.4. Therefore the more hydrophobic the contaminant
is, the more likely it is to sorb to the roots.
log[/?CF 3.0] = 0.65 log Kow -1.57 Equation 1.2
(Burken and Schnoor, 1998)
The TSCF represents the translocation of groundwater contaminants to the plants'
transpiration stream as defined below in Equation 1.3.
concentration of pollutant in the transpiration stream ^ _
IbLr =------------------------------------------------------Equation 1.3
concentration of pollutant in the external solution
The TSCF can be estimated from Equation 1.4. As shown in Figure 1.5, the TSCF
ranges from 0 to 1. A TSCF of 1 represents that the tree is transpiring the
contamination at the same concentration that it exists in the groundwater. The TSCF
for poplars is optimum in the range of log K<>w of 0.5 to 3.0. Since the Log K<,w for
MTBE is 1.24 (Chemfate, 1994), which is in the range that is readily transpired by
13


RCF, mL/g
plants and trees it is hypothesized that MTBE will be readily uptaken by poplar trees
via their transpiration stream.
TSCF = 0.756 exp
-(log ^.-2.50)
2.58
Equation 1.4
(Burken and Schnoor, 1998)
Log Kow
RDX
phenol
X nitrobenzene
o benzene
A TCE
atrazine
o toluene
A ethylbenzene
m-xylene
TCB
PCP
RCF
Briggs [2]
FIGURE 1.4 Root Concentration Factor vs. Log for Various Chemicals in
Poplars and Barley
14


RDX
X aniline
phenol
o benzene
TCE
atrazine
o toluene
A ethylbenzene
m-xylene
X TCB
+ PCP
TSCF
Br>ggs [2]
FIGURE 1.5 Transpiration Stream Concentration Factor vs. Log K0* for
Various Chemicals for Poplars and Barley
The rate at which a pollutant is taken up by a tree via transpiration can be estimated
as in Equation 1.5.
R transpiration = Qt TSCF C Equation 1.5
where:
R transpiration is the removal rate of the contaminant due to uptake (mg/day).
Qt is the transpiration rate of the poplar tree (L/day).
TSCF is the transpiration stream concentration factor for the pollutant.
C is the aqueous phase pollutant concentration in groundwater (mg/L).
15


Equation 1.5 quantifies the tree as a sink for removal of a pollutant from groundwater.
Accurate estimates of TSCF as well as Qt are needed to predict removal of MTBE
from groundwater. Plant physiologists have developed several empirical and
physically based methods for estimating evapotranspiration (ET) from measured
climatic data such as, wind speed, sunlight, humidity, temperature, and leaf surface
area (Ward, 1995). These predictive models are mostly based upon energy balances
and mass transfer concepts. These mechanistic correlations, such as the Penman
equation or the modified Penman-Monteith equation, as well as more empirical and
aggregated predictive equations (SCS Blaney-Criddle Method, Jensen-Haise,
Thomwaite Method), predict the maximum evapotranspiration possible by plants
when an unlimited supply of water is available for uptake (potential ET). These
models are not correlated to actual field conditions, and do not represent the actual
ET but the potential ET. Since measurement of ET is not always practical or
affordable, if a predictive ET model can be shown to closely predict ET of poplar
trees this will make evaluation of phytoremediation less complex. Chapter 3 will
present uptake information and experiments with quantification of TSCF and RCF.
Chapter 5 will address ET and the correlation of measured ET to predictive ET
equations.
1.5.2 Phytodegradation in the Rhizosphere
The rhizosphere is a zone of increased microbial activity and biomass at the root-soil
interface. Plant roots secrete and slough substances such as carbohydrates, enzymes,
and amino acids which microbes can use as substrate and therefore increase their
mass. In turn, this area of increased microbial concentration is an ideal environment
for degradation of xenobiotics. There may be an increase of microbial biomass by an
order of magnitude or more in the rhizosphere compared to unvegetated soils. Factors
that affect the rhizosphere are shown in Table 1.6. Pollutant degradation in the
rhizosphere is further supported by additional oxygen transferred from the root
system into the soil causing enhanced aerobic mineralization of organics and
stimulation of cometabolic transformation of chemicals such as benzene,
polychlorinated biphenyls, and polyaromatic hydrocarbons to less toxic substances
(Anderson et ai, 1993).
Enhanced degradation in the rhizosphere has been demonstrated for a range of
chemicals including pesticides, polyaromatic hydrocarbons, oil, surfactants and
chlorinated alkanes (Anderson et al, 1993). However, no study to date has assessed
the degradation of MTBE in the rhizosphere of poplar trees. Chapter 4 addresses
degradation in the rhizosphere, MTBE degradation, and experiments and results that
investigate rhizosphere degradation of MTBE in the rhizosphere of poplar trees.
16


TABLE 1.6 Factors that Effect the Rhizosphere
Plant species Plants species produce various types of root exudates. Certain plant species may release phenols which help support the growth of PCB degrading bacteria ( Fletcher, 1995)
Plant age Generally, the older the plant the larger the microbial population in the rhizosphere.
Soil type Plant exudates will differ depending if the roots are in sand or soil. This would be specific to the plant, (i.e. pea roots exuded greater amino acids into quartz sand then into culture solutions! Wilding, 1966) and root exudates from potatoes growing in soil was more active than in sand (Widdowson, 1958)).
Exposure to xenobiotics The effects of contaminants will alter the root exudates.
Microorganisms Microorganisms affect the root exudates by several ways. 1. They affect the permeability of root cells 2. They affect the metabolism of roots 3. They affect the absorption of certain compounds in root exudates by microorganisms and excretion of other compounds.
Soil moisture Usually the moister the soil is the more root exudates are found.
Root damage More exudates are released if there is no root damage.
Temperature Different temperatures will grow different bacteria as well as alter the release of exudates.
Light Root exudates are affected by light. Different plants will perform optimal in different light conditions.
Plant nutrition i The higher the plant nutrition the more root exudates were found.
(Burken and Schnoor, 1996 and Rovira, 1969)
17


1.5.3 Phytodegradation within the Poplar Tree
Phytodegradation within the poplar tree refers to the breakdown of contaminants
taken up by trees and broken down by metabolic processes occurring within a large
woody tree. Pollutants are degraded and may be used as nutrients and incorporated
into plant tissues. Plants and trees metabolize chemicals in three stages very similar to
the liver in animals. Stage 1 is transformation (Sandermann, 1992); tree cells have a
degree of active enzymes that are able to degrade contaminants (Trapp and
McFarlane, 1995). Burken and Schnoor (1997) provided evidence for uptake and
vegetative detoxification of atrazine in contaminated soils. Likewise, TCE was shown
to bio transform in poplar trees (Newman et al, 1997). Stage 2 is conjugation, with
glutathione, sugars, or amino acids, and stage 3 is storage. Storage occurs in the cell
vacuole, extracellular spaces, and lignin and other cell wall components
(Sandermann, 1992). In a similar manner, MTBE could be transported from the
transpiration stream of the tree to the wood or lignin of the tree and be transformed or
stored. Chapter 6 presents the evaluation and results of degradation of MTBE within
the large woody tree.
1.6 Upscaling
Phytoremediation assessments can be conducted at several scales. Laboratory bench-
scale chamber studies with small trees (saplings) are useful in elucidating
phytoremediation uptake parameters in providing contaminant mass balances within
the enclosed water-root-shoot-air system. However, bench-scale results can not be
directly extrapolated to the field due to the differences in contaminant metabolism in
large trees (versus small saplings), as well as environmental factors that affect
transpiration, such as ambient wind, rain, sunlight, temperature and humidity, which
are not dominant in laboratory-scale chamber studies.
Newman et al, 1997 performed small-scale hydroponic laboratory tests with poplar
cuttings finding measurable amounts of the TCE were transpired to the air. They
continued their research and investigated phytoremediation of TCE with mature
hybrid poplar trees in a series of constructed lined artificial aquifers for three years.
The mature trees were able to remove 99% of the TCE from the groundwater and less
than 9% of the TCE was transpired to the air for the first two years and after year two
TCE was not detected in the air stream. They deduced that degradation in the
rhizosphere was not contributing to the TCE loss and that the mature tree was de-
chlorinating the TCE.
18


On the other hand, direct field scale studies with large trees make mass balances and
elucidation of mechanisms very difficult. An evaluation was conducted on
chlorinated ethenes from groundwater in tree trunks. Examination of cores collected
at various trunk heights of a bald cypress growing in an area of groundwater
contaminated with TCE showed that concentration of TCE decreased significantly
with increasing trunk height. However, the mechanism causing the decreased
concentration of TCE with increasing height is unknown (Vroblesky, 1999).
Controlled closed chamber investigations need to be conducted with larger trees to
quantify contaminant degradation within the tree. Therefore, I am proposing the use
of a unique laboratory-to-field approach in which phytoremediation measurements
are made at the small and large scale, as described in further detail in the following
sections.
19


2. Objectives and Experimental Design
The overall objective of this thesis is to identify the mechanisms and measure the
parameters for MTBE phytoremediation by poplar trees. Based on the overview of
phytoremediation presented in Chapter 1, there are several challenges and
uncertainties in identifying and quantifying mechanisms for phytoremediation of
MTBE. These challenges include:
MTBE is very volatile, making uptake of MTBE by vegetation difficult to
isolate and measure.
Literature is inconclusive about rhizosphere degradation of MTBE.
No research has been conducted on degradation of MTBE within mature
woody trees.
Mature poplar trees behave biochemcially very differently than small poplar
cuttings, making small cutting laboratory results for phytoremediation of
MTBE different than results with mature trees.
Laboratory results for phytoremediation do not always correlate to
phytoremediation field results.
Therefore, to fully evaluate the potential of phytoremediation of MTBE this thesis is
divided into four phases. The objectives of each phase are listed below and presented
in Figure 2-1:
2.1 Phase 1
Bench scale hydroponic transpiration experiments quantified MTBE uptake via the
transpiration stream of small poplar saplings in a hydroponic system. The main
challenge in demonstrating plant-assisted phytoremediation of MTBE lies in
separating live plant transpiration from physical diffusion and volatilization losses of
MTBE. which can be very high because of the high vapor pressure (Pv = 0.032 Pa at
25 C). Therefore, an innovative system separating the roots and shoots of the poplar
tree was employed. This system was placed in a closed chamber, allowing for
completion of a mass balance of MTBE in the coupled plant-water-air system to
assess the fate and transport of MTBE within the small poplar cuttings. These
experiments will allow determination of the engineering parameters Transpiration
Stream Concentration Factor (TSCF) and Root Concentration Factor (RCF) which are
needed to quantify and predict the effectiveness of phytoremediation. The
experimental design and results for this phase are detailed in Chapter 3.
20


2.2 Phase 2
Bench scale rhizosphere experiments investigated MTBE degradation rates in the
rhizosphere of poplar trees in three different scenarios. The first scenario will
compare the degradation of MTBE in the soil of planted (2-'/2 foot poplar cuttings)
and unplanted pots. The second experiment will utilize the same closed chamber
system as in phase 1, in which an MTBE mass balance deficit in planted systems (6 -
12 inch cuttings) is compared across hydroponic and rhizosphere environments to
determine any enhanced biodegradation of MTBE in the presence of soil. The third
experiment will assess degradation of MTBE in aerobic batch bioreactors utilizing
rhizosphere soil from mature trees (12 feet, 1 year old). Though there is much
literature on degradation of MTBE in soil, it is not known conclusively if MTBE will
degrade in the rhizosphere of poplar trees in a timely manner for phytoremediation.
These experiments are discussed in Chapter 4.
2.3 Upscaling
The first two phases utilized small cuttings in very controlled settings, and provided
the needed engineering parameters (TSCF, RCF, and kbio) for phytoremediation.
However, much care needs to be taken when extrapolating bench scale laboratory
results to field conditions. Larger more mature trees behave biochemically different,
have higher transpiration rates, increased enzymatic capabilities, and longer residence
time for the chemical. Furthermore, the lack of controlled settings in the field makes
it difficult to fully assess the fete and transport of contaminants in vegetation. To
address these issues, intermediate scale tests were performed with mature (1 year old)
12-foot tall trees in an enclosed chamber in Phases 3 and 4. This intermediate step
provides useful information that is generally ignored in most phytoremediation
studies.
2.4 Phase 3
Large scale tree transpiration experiments measured transpiration of water in large
12 poplar trees incorporating environmental factors such as sunlight, wind speed and
humidity in a controlled setting. The transpiration rates (Qt) as a function of varying
meteorological parameters will be calibrated to a predictive transpiration model such
as the Penman Monteith Equation. The idea is that measurement of
evapotranspiration (ET) is not always practical or affordable and that a predictive ET
model can be used to closely predict ET of poplar trees. This will make evaluation of
phytoremediation as a potential remediation technology less complicated to the
consultant and more acceptable to the regulator. Measuring and modeling of ET is
presented in Chapter 5.
21


2.5 Phase 4
Large scale tree degradation experiments determined potential metabolism of MTBE
into benign products within a much larger tree (12'-feet) enclosed in a chamber. The
degradation product of MTBE, tert butyl alcohol (TBA), will be analyzed for during
laboratory gas chromatograph analysis. The experiment is similar to Phase 1, but on a
much larger scale to evaluate the effect of longer residence time in the woody tree
along with increased transpiration and enzymatic capabilities. Therefore, any scale
effects in the phytoremediation parameters of MTBE will be evaluated in this phase,
which is described in Chapter 6.
The connections between the various phases of research are shown in Figure 2.1.
22


FIGURE 2.1 The Phased Approach to Assess Phytoremediation of MTBE


3. Phase 1 Fate and Transport of MTBE in Poplar
Cuttings
3.1 Introduction
Phytoremediation can be an effective remediation technology if the fete and transport
of the chemical within the vegetation is known. Uptake and accumulation rates of
xenobiotics in vegetation are important to quantify because they influence the food
chain and are potential hazards of bioaccumulation. Furthermore, there is a possibility
of moving a contaminant from one medium to another (i.e. from groundwater to air).
Depending on the scenario this may be a desirable or an undesirable condhion.
Likewise degradation either within the soil or vegetation may create contaminants of
even more concern than the parent compound causing further health hazards. For all
these reasons, the fate and transport of a contaminant in vegetation must be quantified
before phytoremediation can be used as a remediation technology in the field.
3.2 Plant Uptake Studies
Briggs et al (1982) represented the first predictive correlation for contaminated
uptake potential, focusing on the relationship between pesticide uptake and crops.
They reported the fundamental discovery that the hydrophobicity, quantified by its
octanol-water partition coefficient KoW was related to the contaminant root uptake
(transpiration). The transpiration stream concentration factor (TSCF) is an important
parameter for quantitative design of phytoremediation systems; it represents the
translocation of groundwater contaminants to the plants' transpiration.
Following in this line of research Burken and Schnoor (1998) investigated the uptake
of organic contaminant by hybrid Poplar trees. They also confirmed the correlation
between the TSCF and the log Kow. The uptake process is most effective for
pollutants with log Kow in the optimum range of 0.5 to 3.0 (Briggs et al, 1982 and
Burken and Schnoor, 1998). When the log KoW is greater than 3, the contaminant is
sorbed so strongly to the root surface that the chemical will not be translocated into
the shoots. In contrast, chemicals with a log KoW less than 0.5 are very water-soluble
and will not be sorbed to roots and therefore will not be transported through the plant.
Burken and Schnoor (1998) also developed a correlation between log Kow and the
root concentration factor (RCF). The RCF addresses sorption of organic pollutants
from soil to the lipophillic root tissues. The RCF increases as the pollutant's log KoW
24


Li et al (2002) evaluated the uptake of both trifluralin and lindane by ryegrass in
aqueous laboratory experiments. Uptake of lindane was promising but no metabolism
nor contaminants bound to plant tissue were identified; while trifluralin was uptaken,
metabolized, and bound to ryegrass tissue.
Many researchers have evaluated the uptake and transpiration of trichloroethylene
(TCE) in vegetation. TCE was reported to be taken up, transformed, and bound to
plant tissue of carrots, spinach, and tomatoes (Schabel et al, 1997) and in poplar trees
(Newman et al, 1997). Likewise Burken and Schnoor (1998) reported TCE along with
benzene, toluene, ethyl benzene, nitrobenzene, m-xylene were readily taken up and
transpired from the leaves of hybrid poplar trees. However, Orchard et al (2000)
investigated the uptake of TCE by hybrid poplar trees grown hydroponically and
presented results unlike those previously mentioned. They suggest that rhizosphere
degradation is the method of TCE metabolism and phytovolatilization was minimal.
They reported a TSCF of 0.02-0.22 in comparison to 0.75 that Burken and Schnoor
(1998) presented while Briggs et al (1982) correlation would estimate a TSCF of
0.62. Since research in this area has only been conducted in the last decade, there is
still much research that needs to be conducted to fully understand the uptake transfer
and fate of organic compounds in vegetation.
3.3 MTBE Uptake by Vegetation
The understanding of MTBE uptake is essential to evaluating MTBE mobility in the
ecosystem and exposure to humans. Since the Log K<>w for MTBE is 1.24 (Chemfate,
1994), which is in the range that is readily transpired by plants and trees,
phytoremediation may be a feasible remedial technique for subsurface MTBE
plumes. Also, alfalfa plants grown in MTBE contaminated soil uptook the MTBE
contaminated soil and transported it to the stems. A mean diffusion coefficient of
MTBE radial transport across the alfalfa plant was reported to be 1.23 x 10' cm /s
(Zhang et al, 2001). While MTBE does not readily degrade in soil (e.g. laboratory
microcosm tests showed no degradation over 300 days) (Suflita et al., 1993),
atmospheric MTBE reacts with hydroxyl radicals, yielding a MTBE half-life of the
order of days in air (USEPA, 1993). Thus, transpiration of MTBE from the subsurface
to air via plants can offer a low-cost method for destruction of the pollutant to benign
products. In addition, MTBE may be metabolized within plants to benign constituents
that contain ether linkages (Ramsden, 2000) as is discussed in Chapter 6.
Controlled greenhouse studies are typically conducted to understand the importance
of the fate and transport pathways of organic chemicals in the plant system. However,
such studies are difficult to conduct and interpret for very volatile compounds. Much
25


of the difficulty in demonstrating plant-assisted phytoremediation of MTBE lies in
separating live-plant transpiration from passive, physical volatilization losses of
MTBE, which can be very high because of the high vapor pressure ( Pv = 0.032 Pa at
25C) of the chemical.
Currently, there is need for information on transpiration or transformation of MTBE
from water due to plants. Quantitative data on MTBE uptake and transpiration is
required for design of engineered MTBE phytoremediation systems. This chapter
examines MTBE uptake by poplar plants in carefully controlled systems and provides
first quantitative estimates of parameters (TSCF, RCF) that are needed to develop
engineered phytoremediation of MTBE.
3.4 Experimental Design
3.4.1 Objectives
The objective of phase 1 is to implement experiments that enable quantification of the
various pathways for MTBE removal from water by plant systems, separating the
passive, physical process of volatilization from living plant processes such as
transpiration and degradation. Engineering parameters relevant for describing MTBE
phytoremediation are then computed to enable design and planning of engineered
MTBE phytoremediation systems. The engineering parameters of interest are the
transpiration stream concentration factor (TSCF) and the root concentration factor
(RCF). Estimation equations for both parameters are described in Briggs et al (1982)
and Burken and Schnoor (1998) and are defined in the introduction.
To achieve the above objectives the three following areas were investigated:
Removal of MTBE from water and accumulation in plant biomass was studied
at two different MTBE concentrations (300 and 1600 ppb) in triplicate plant
systems and compared with carefully designed controls, to assess the degree
of MTBE removal from water by plants. This information allowed for
calculation of the TSCF.
A mass balance of MTBE was computed for the plant-water-air system in a
closed chamber wherein measurement of MTBE in the air stream enabled
closure of the mass balance and assessment of potential MTBE
transformation.
The RCF was studied in microcosm batch equilibrium reactors containing
macerated roots in contact with MTBE-laden water.
26


The three experiments were conducted at the two MTBE concentrations of 300 and
1600 ppb, representative of the range observed at MTBE-contaminated sites (Johnson
et al, 2000).
3.4.2 Materials
MTBE solutions of approximately 300 and 1600 ppb (pg/L) were prepared from
dilutions made from neat MTBE obtained from Supelco, Belefonte, PA. The MTBE
solutions were tested with poplar trees grown from small cutting obtained from
Colorado State University Forestry Services.
3.4.3 Plant Preparation
The poplar tree cuttings were inserted inside the septa of lids that fit 250-ml glass
septa jars. The roots and shoots developed above and below the lid, respectively; the
stem grew in diameter forming a tight seal in the septum (Burken and Schnoor, 1998).
Note the jar was not fitted to the lids at this time. The plants were grown in this
manner in greenhouse conditions using commercial soil, sprayed with pesticide and
nutrients as necessary. When the plants had grown at least 6 inches above the septum
they were uprooted from the soil; the roots were washed so that the plant was ready to
be immersed in MTBE-solution. The lid, within which the plant had grown from
cutting stage, could then be screwed onto the jar to create an airtight seal.
3.4.4 Experiment 1
Removal of MTBE from water and accumulation in plant biomass were studied in
250-ml glass jars that had a small hole drilled in their side and were then fitted with a
plastic bag (Figure 3.1). The prepared MTBE solution was poured into the plastic
bags fitted into the 250-ml jars. The poplar-fitted lids were then screwed onto the jars
to form a tight seal with no headspace. Silicone was used on and around the lids to
reinforce airtight conditions. The weight of the jar with plant and MTBE solution was
recorded at this time; weight loss was measured over the course of the experiment
(one-week) and used to track the water mass and volume lost from the system due to
transpiration. The flexible plastic bag compressed as transpiration took place, thereby
maintaining zero headspace throughout the experiment and minimizing MTBE
volatilization. The hole in the side of the glass jar enabled bag compression to occur.
The jars were monitored daily to check for the undesirable development of
headspace. Approximately 50% of the plant-filled jars were successful in maintaining
zero headspace. Jars with headspace were eliminated from the experiment. The
experiment was terminated after one week, at which time the water, roots and shoots
were analyzed for MTBE. Forty milliliters (ml) of water were withdrawn from each
jar and placed into airtight glass vials for aqueous MTBE analysis. The poplar roots
27


and shoots were separated, quickly patted dry, dipped into liquid nitrogen, pulverized,
and placed into bottles for analysis of MTBE in biomass. All samples were
refrigerated until analysis, with a maximum hold time of 14 days.
The experiment utilized three different controls (Figure 3.2) for differentiating
between volatilization, passive transpiration and active transpiration. The first control
was a 250-ml jar with no insert in the septum, which simulated a no loss mechanism.
The second control consisted of inserting solid copper tubing through the septa. This
simulated a non-porous insert, therefore allowing for minimal volatilization. The third
control was dead balsa wood inserted through the septa of the 250-ml jar,
representative of a porous but non-living material, which accounted for passive
transpiration. All controls were treated the same as the live plant systems. Triplicates
were used for the controls as well as the plant systems.
FIGURE 3.1 Drawing of Poplar Cutting Inside 25-L Chamber
28


FIGURE 3.2 Small Cutting Uptake Study with Controls
3.4.5 Experiment 2
A mass balance of MTBE was computed by placing one 250 ml plant-jar, prepared as
described in experiment 1, into a 25 L chamber. Tubing from the outflow from the
chamber is connected to two carbon tubes in series that would trap any airborne
MTBE (Figure 3.3). The carbon tubes were changed daily during the five day long
experiment, after which the water, roots, shoots, and all carbon tubes were analyzed
for MTBE. A mass balance was computed to assess if MTBE lost from water could
be recovered either in plant biomass and/or in the air transpired by the plant, or if
significant degradation of MTBE had occurred. MTBE degradation product, tert butyl
alcohol (TBA) (Steffan, 1997) was also analyzed for during analysis.
29


FIGURE 3.3 Small Cutting Mass Balance Experiment
3.4.6 Experiment 3
The RCF was measured in 40- mL glass vials with no headspace, in which 2.5 g of
poplar root mass was contacted with MTBE-laden water at 300- and 1600- ppb levels.
Contacting macerated roots with contaminated water is a process aimed at isolating
equilibrium between root mass and contaminant and follows procedures established
with barley (Briggs et al, 1982) and poplars (Burken and Schnoor, 1998). The
experiment was initiated with 12 replicate vials at each concentration level. Vials
were periodically monitored to confirm the approach to equilibrium over a 1-week
period and sacrificed thereafter (six vials were sacrificed). Upon equilibration, the
water and roots in the six remaining vials were analyzed for MTBE. The RCF for
MTBE was then reported as the ratio of the final equilibrium MTBE concentration in
root mass to that in water.
30


3.4.7 Analytical Methods
MTBE in water was analyzed by gas chromatograph (GC) analysis using a purge and
trap method with an MXT-1 column (60m, 0.53mm, 5.0 pm) from SRI, Instruments.
The analysis method was based upon EPA Method 8260. Method detection limits for
MTBE were 10 ppb and the Relative Reference Factor (RRF) was 7.4%. MTBE in
plant biomass was analyzed using the same technique as water. An experiment to
assess MTBE recovery in the presence of plants showed an average of 87 4%
recovery.
MTBE in the air was analyzed based on NIOSH Method 1615, which is a direct
injection GC analysis using carbon disulfide to extract MTBE from the carbon tubes.
The front and back halves of each tube were extracted with 2-ml carbon disulfide
(CS2) and analyzed separately by Method 1615 with a method detection limit of 5
ppm (in CS2) and a relative reference factor of 8.3%.
A preliminary experiment was conducted to verify mass balance of MTBE within the
chamber system in the absence of plants. MTBE volatilized from a jar was recovered
from the air stream. The degree of recovery, i.e., the closure of mass balance with the
physical process of volatilization alone, was determined in five replicate experiments
yielding 103 16% recovery, indicating successful design of the air chamber. Refer
to Table 3.1. The next section will present the results for the three experiments
described in this section.
3.5 Results
3.5.1 Experiment 1
Results from experiment 1, removal of MTBE from water and accumulation in plant
biomass, at the 1600-ppb level, are shown in Figures 3.4 to 3.8. Figure 3.4 shows
water loss in the various controls and in the plant systems at the end of the one-week
experiment. The no-insert and copper insert controls showed minimal (2-ml and 0.8-
ml respectively) water loss; passive transpiration resulted in a water loss of 12-ml in
the dead balsa system, while active plant transpiration produced the greatest water
loss of 24-ml. Figure 3.5 shows the change in average MTBE concentrations in the
poplar and control systems after one week. Figure 3.6 shows the percent reduction in
MTBE concentration and mass in the various systems. Concentration reduction of
MTBE in the water due to the live poplar saplings was in the range of 20% and mass
reduction was in the range of 30%. Concentration reduction of MTBE in the water
due to the dead balsa wood was in the range of 10 % and mass reduction was in the
range of 15%. Thus passive transpiration produced approximately half the reduction
in concentration and mass caused by the active plant system. Concentration reduction
31


TABLE 3.1a
Mass Balance without the Presence of Poplar Trees
Jar Initial MTBE Water soln volume Final MTBE Water soln volume Initial MTBE carbon volum e Initial MTBE in Water area Initial MTBE in Water cone. Final MTBE in Water area Final MTBE in Water cone.
ml ml ml PPb PPb
1 239 237 2 2762 408 1728 243
2 230 230 2 2832 420 1776 251
3 231 230 2 2776 411 2072 298
4 237 236 2 2725 402 1624 227
5 226 226 2 2838 421 2056 296
TABLE 3.1b Mass Balance Continued without the Presence of Poplar Trees
Jar MTBE in carbon area MTBE in carbon cone. Initial MTBE mass in water Final MTBE mass in water Mass change Final MTBE mass in carbon % recovered
ppm Pg Pg Pg Pg
1 81 16 98 58 40.0 39.7 99
2 78 15 96 58 38.8 38.8 99
3 53 13 95 69 26.3 29.2 1.11
4 64 12 95 53 41.0 33.6 80
5 68 14 95 67 28.2 34.7 1.23
32


of MTBE in the water from the copper control was in the range of 2 % and mass
reduction was in the range of 1%, indicating little or no loss from these systems,
confirming integrity of the experiment. A t-test showed that the plant-assisted mass
and concentration reductions were significantly different from all of the controls at
the 10% level of significance using a one sided t-test.
Analysis of plant biomass showed significant concentration of MTBE in roots and
shoots (Figure 3.7), of the order of 160 ppb in roots and 60 ppb in shoots confirming
passage of MTBE through the plant. In contrast much less MTBE was detected in the
balsa. Mass of MTBE in the roots of the poplars was approximately 1.8 pg and in the
shoots of the poplars was in the range of 0.6 pg (Figure 3.8). Mass of MTBE in the
balsa wood was in the range of 0.02 pg.
Similar results were obtained at the 300- ppb level experiments. Concentration
reduction of MTBE in the water due to the live poplar saplings was in the range of
25% and mass reduction was in the range of 50%. Concentration reduction of MTBE
in the water due to the dead balsa wood was in the range of 10 % and mass reduction
was in the range of 30%. Thus again passive transpiration produced approximately
half the reduction in concentration and mass caused by the active plant system.
Analysis of plant biomass showed 0.2 pg of MTBE in the roots and 0.1 pg in the
shoots again verifying transport of MTBE through the poplar tree.
The percent MTBE removal from water, at both initial concentrations of 1600 and
300 ppb was strongly correlated with volume of water transpired and leaf surface area
of the plant as shown in Figures 3.9 and 3.10. All the above results support the
hypothesis that MTBE is actively transpired in significant quantities by plants, at
levels approximately double that of passive transpiration.
33


FIGURE 3.4 Water Loss in Various Systems (Initial Aqueous MTBE
Concentration = 1600 ppb)
FIGURE 3.5 Final Aqueous MTBE Concentrations Various Experimental
Systems (Initial Aqueous MTBE Concentration = 1600 ppb)
34


c
o
*S
O
3
o
0)
a.
&
60
50
40
30
20
10
0
-

$ 7

, i"i &!, t .<.!! s
Smass
cone
initial closed jar cooper balsa poplar
tube
FIGURE 3.6 Percent Reduction in MTBE Concentration and Mass (Initial
Aqueous MTBE Concentration = 1600 ppb)
FIGURE 3.7 MTBE Concentration in Biomass (Initial Aqueous MTBE
Concentration = 1600 ppb)
35


FIGURE 3.8 MTBE Mass in Biomass (Initial Aqueous MTBE Concentration =
1600 ppb)
3.5.2 Transpiration Stream Concentration Factor
Calculations
The transpiration stream concentration factor, TSCF, is an important parameter for
quantitative design of phytoremediation systems (Briggs et al, 1982 and Burken and
Schnoor, 1998). This chapter presents a first numeric estimate of the TSCF for
MTBE, using the experimental data reported in the previous section. The observed
TSCF for MTBE is computed from the plant uptake experiments in two different
ways.
First, the MTBE concentration in the transpiration stream was computed from
knowledge of MTBE lost to air and the volume of water transpired; the TSCF was
computed as the ratio of the MTBE concentration in the transpired water to the time-
averaged MTBE concentration in the water remaining in the jar. This method yielded
an average TSCF of 2.2. A TSCF value greater than one implies either some sort of
concentration mechanism occurring within the plant transpiration stream; or an
anomaly caused by the inclusion of passive volatilization of MTBE. Since results
with balsa wood indicate that passive volatilization through porous membranes is
36


roughly half that of active plant uptake, a corrected TSCF estimate is one half of 2,
i.e., approximately 1, indicating that plants transpire MTBE at much the same
concentration as it exists in groundwater. TSCF measured values versus predicted
values from Burken and Schnoor equations are provided in Table 3-3.
In the second method, Figure 3-10 was used to determine the relationship between
percent MTBE mass removed from solution and volume of water transpired by the
plants. The TSCF was computed from the slope of the best-fit line in Figure 3-10 as:
volume of water transpired, and, V^ai is the initial volume of water in the
hydroponic system. The factor of 100 incorporates for the % values plotted in the
graph. [Note, multiplying and dividing the right hand side of equation 1 by the initial
mass of MTBE in water yields the concentration ratio definition of TSCF]. The slope
in the graphical technique isolates MTBE mass reduction achieved by plant
transpiration; the non-zero intercept indicates volatilization loss of MTBE with zero
transpiration. The observed TSCF for MTBE computed from the graphical method
yielded an estimate of 1.1, again suggesting that the MTBE is actively transpired by
plants at the same concentration as in groundwater. Models for MTBE loss from
groundwater would need to separately account for passive volatilization of MTBE
through porous biomass.
Equation 3.1
where dM'
transp
is the slope of the line plotting % MTBE mass reduction versus
37


Volume of Water Transpired
50 i-----------------------------=--------------------

Leaf surface Area ( sq. in.)
FIGURE 3.9 Volume of Water Transpired vs. Leaf Surface Area
0 20 40 60 80 100
Volume of Water Transpired (ml)
FIGURE 3.10 Percent Mass Reduction ofMTBE vs. Volume of Water
Transpired
38


The work of Burken and Schnoor (1998) indicated that the TSCF of different organic
chemicals can be estimated by correlation with compound KoW, but must be supported
by experimental data since the chemistry of individual compounds can create
anomalies, e.g., nitro compounds (Burken and Schnoor, 1998). Burken and Schnoor
(1998) correlation formula provide a much smaller TSCF value for MTBE of 0.3.
However, these correlations were not developed for highly hydrophilic solvents of
low molecular weight or high polarity. Due to the chemical differences of MTBE
from the chemicals used to develop the equation presented in the introduction, these
correlations may not be valid. Hong et al (2001) approximated a TSCF for MTBE of
0.5 to 0.8, again much higher than the formulas developed from Burken and Schnoor
(1998).
3.5.3 Experiment 2
Results of the mass balance test with one plant at high MTBE concentration (1600
ppb), one plant at low MTBE concentration (300 ppb), and two controls are presented
in Table 3.2 and Figure 3.11. Good recovery of MTBE was obtained in both systems,
suggesting no degradation of MTBE in the small plants. All the MTBE lost from
water was primarily recovered from air, thereby suggesting that plants act as a
conduit for MTBE transfer from groundwater to air. Laboratory scale tests with TCE
showed similar non-transformation of the pollutant in small trees, while field tests
showed significant metabolism of TCE in large trees with larger travel time from
roots to leaves (Newman et al, 1999). Metabolism of MTBE within a larger tree is
addressed in Chapter 6
39


Mass of MTBE
in vegetation
(ug) 0.4
FIGURE 3.11 MTBE Mass Balance (Initial MTBE Concentration = 1600 ppb)
TABLE 3.2 MTBE Mass Balance
Sample Initial MTBE mass in water (Pg) Final MTBE mass in water (Pg) Mass of MTBE in roots (Pg) Mass of MTBE in shoots (Pg) Mass of MTBE in air (Pg) Total Mass recovered (Pg) % MTBE recovered
Poplar with High Cone. MTBE (1600 ppb) 361.2 251.1 0.1 0.3 83.2 334.1 93
Poplar with Low Cone. MTBE (300 ppb) 44.5 25.4 ND ND 30 55.4 124
Control no plant (1600 ppb) 375.5 378.5 NA NA 0 378.4 101
Control Balsa (300 ppb) 60 40 ND ND 22 62 103
40


3.5.4 Root Concentration Factor Calculations
The RCF for MTBE computed from experimental data ranges from 0.7 to 1.4 (Table
3-3), which is about a factor of 2-3 times lower than that predicted through Kw
correlation equations. The ether structure in MTBE causes its chemistry to be much
different from all other compounds studied in the plant uptake experiments of Briggs
et al (1982) and Burken and Schnoor (1998), from which the correlation equations are
derived. This may explain the differences in quantitative estimates of MTBE
phytoremediation parameters observed experimentally versus those predicted from
correlations.
TABLE 3.3 Measured vs. Predicted TSCF and RCF Values for MTBE (Values shown are the mean +/- standard deviation computed from n data points)
TSCF n = 8) RCF ( n = 6)
System Measured Predicted Measured Predicted
Low Cone. (300 ppb) 1.12+/-0.12 0.39 1.38+/-0.83 3.17
High Cone. (1600 ppb) 1.02+/-0.20 0.39 0.68+/- 0.16 3.17
3.6 Discussion
In summary, a novel membrane jar unit enclosed in a flow-through air chamber was
designed to examine the potential for phytoremediation of MTBE, distinguishing
between plant uptake of MTBE by means of transpiration versus passive
volatilization losses. Results from this study indicate that MTBE was readily taken-up
from the water by hybrid poplar saplings, yielding 25% reduction in aqueous MTBE
concentration and 30% reduction in MTBE mass over a 1-week period. These
reductions in plant systems were significantly greater than in controls, showing that
phytoremediation is a promising remedial technique for the cleanup of MTBE-
contaminated ground water. Passive volatilization of MTBE accounted for
approximately half the reduction in mass and concentration seen in living plant
systems, correcting for which yielded an estimate of 1 for the TSCF. Thus,
experimental data suggest that MTBE is transpired by plants at the same
41


concentration as it exists in groundwater, indicating great potential for
phytoremediation in the field.
The experimentally derived RCF for MTBE was observed to range from 0.7 to 1.4,
observations were lower than those predicted from correlations with K<,w. Mass
balance studies show that MTBE was untransformed during transport through small
saplings used in this study. The high TSCF and low RCF indicated that
phytovolatilization is the primary mechanism for phytoremediation of MTBE in
aqueous bench scale studies. These results are very promising for the potential
remediation of MTBE by phytoremediation.
Hydroponic small sapling aqueous studies have indicated that phytovolatilization is
the primary route for phytoremediation of MTBE, however, for further evaluation of
MTBE phytoremediation Chapter 4 will address rhizosphere degradation, while
Chapter 6 addresses MTBE metabolism in much larger trees.
42


4. Phase 2 Degradation of MTBE in the
Rhizosphere of Poplar Trees
4.1 Introduction
Laboratory-scale research on MTBE has shown that phytovolatilization, i.e. uptake of
MTBE during transpiration, is an important pathway for removal of MTBE from
groundwater by poplar trees (Rubin and Ramaswami, 2001; Hong et al, 2001). Phase
1 studied hydroponic systems planted with poplar saplings. Results from the
experiment showed 30% removal of MTBE mass from water over a 1-week period
due to transpiration, at both high (1600 ppb) and low (300 ppb) initial aqueous MTBE
concentrations; almost all of the MTBE removed from water was recovered in the air
stream (Rubin and Ramaswami, 2001). The work of Hong et al (2001) also showed
that uptake and evapotranspiration of MTBE from poplar cuttings was the
predominant pathway for MTBE removal in hydroponic systems. In addition to
poplar trees, alfalfa plants grown in MTBE contaminated soil have also been shown
to uptake MTBE through the roots into stem tissue; an MTBE diffusion coefficient
across the stem was estimated to be in the range of 8.43-16.2 x 108 cm:/s (Zhang et
al, 2001). Laboratory studies thus indicate that uptake of MTBE by plants and trees
provide a mechanism for removal of MTBE from contaminated groundwater.
However, no research to date has evaluated the potential for degradation of MTBE in
the rhizosphere of trees. The focus of phase 2 is to examine the potential for
degradation of the fuel oxygenate MTBE in the rhizosphere of poplar trees in a timely
manner for phytoremediation. This chapter will first define the rhizosphere, second
discuss degradation in the rhizosphere, third identify MTBE degradation studies
(aerobic, anaerobic, cometabolic, and field studies), fourth describe the experimental
design to assess MTBE degradation in the rhizosphere of poplar trees, and last present
the results.
4.2 The Rhizosphere
The rhizosphere, 1-3 mm of soil surrounding each root, is a zone of increased
microbial activity and biomass. This increased microbial biomass is due to roots
secreting and sloughing substances such as carbohydrates, enzymes, and amino acids
which microbes can use as substrate and therefore increase their mass. See Table 4.1
from the Technology Evaluation Report TE-02-01 for a list of root exudates. Possibly
10,000 cells per plant per day can be sloughed from the root cap to the soil. Also, a
gelatinous substance, mucigel, is secreted by the root cells for lubrication to enhance
root penetration through the soil during growth (Anderson et al, 1993). There may be
43


an increase of general population microbes in the rhizosphere by 1-2 orders of
magnitude compared to unvegetated soils. Also, specific microbes are increased by 3-
4 orders of magnitude in the rhizosphere compared to unvegetated soils (IRTC, 2001).
TABLE 4.1 Compounds Detected in Root Exudates
COMPOUNDS EXAMPLE OF COMPOUND REFERENCES
Carbohydrates Glucose, fructose, sucrose, maltose, galactose, xylose, oligosaccharides Curl and Truelove, 1986
Amino Acids Gylcine, glutamic acid, aspargine, serine, amine, lysine, arginine, threonine, homoserine Curl and Truelove, 1986
Aromatics Phenols, 1-carvone, p-cymene, limonene, isoprene Fletcher and Hedge, 1995 Gilbert and Crowley, 1997
Organic Acids Acetic acid, propionic acid, citric acid, butyric acid, valeric acid, malic acid Curl and Truelove, 1986
Volatile Compounds Ethanol, methanol, formaldehyde, acetone, acetaldehyde, pfopionaldehyde, methyl sulfide, propyl sulfide, allyl sulfide Curl and Truelove, 1986
Vitamins Thiamine, biotin, niacin, riboflavin, pyridoxine, panotothenic acid Curl and Truelove, 1986
Enzymes Phosphatase, dehydrogenase, peroxidase, dehalogenase, nitroreductase, laccase, nitrilase Curl and Truelove, 1986 Newman, 1995 Vaughan et al., 1994
(Schnoor, 2002)
In turn, this area of increased microbial concentration is an ideal environment for
degradation of xenobiotics. Pollutant degradation in the rhizosphere is further
supported by additional oxygen transferred from the root system into the soil causing
enhanced aerobic mineralization of organics and stimulation of cometabolic
transformation. Laboratory seedlings can transport 0.5 moles of O2 per m3 per day to
the rhizosphere (Burken And Schnoor, 1996).
44


There are many advantages to rhizosphere degradation such as: destruction of
xenobiotics in-situ, translocation to the atmosphere is less likely, complete
mineralization of contaminant, and low cost. However there are also many
disadvantages to using this technology: a small percentage of rhizosphere soil is in
contact with the contaminant making degradation very limited, it can be a very timely
process, additional fertilizer needs, possible toxicity to the roots, and root length can
limit degradation potential (Wilson, 2000). Root radius was shown to be a more
important factor in one phytoremediation model than contaminant properties (Yoon-
Young and Corapciogla, 1998).
4.2.1 Degradation in the Rhizosphere
Due to the increased microbial activity, degradation of contaminants in the
rhizosphere has been widely studied in the last decade. The findings have reported
enhanced degradation in the rhizosphere for a range of chemicals including
pesticides, explosives, PCBs, PAHs, oil, surfactants and chlorinated alkanes
(Anderson et al, 1993), however the degradation of MTBE in the rhizosphere of trees
has not yet been studied. Refer to Table 4.2 for a list of rhizosphere degradation
studies which are detailed next.
Benzene, toluene, and xylene (BTX) are a major constituent of lead-free gasoline and
commonly detected in soil and groundwater due to leaking underground storage tanks
and commonly in MTBE plumes. Jordahl et al (1997) reported higher populations of
BTX degraders along with total heterotrophs, denitrifiers, pseudomonads, and
atrazine degraders in the rhizosphere of poplar trees (populus deltoids x nigra DN-34,
Imperial Carolina). One study found after 245 days, degradation of TPH was 50% in
clover plots, 45% in fescue plots, 40% in Bermuda grass pots, and 30% in
unvegetated pots (AATDF, 1998).
The organic particles of greatest concern are polycyclic aromatic hydrocarbons
(PAHs); many of which are carcinogenic. Much research has been conducted on PAH
degradation in the rhizosphere. Fang et al (2001) found no significant effects for the
degradation of atrazine and phenanthrene in the rhizosphere of five grass species;
while April and Sims (1990) reported the loss of PAHs was greater in eight types of
prairie grass vegetated soil than in non-vegetated soil. Confirming degradation of
PAHs by grasses, Pradham et al (1998) and Reilly et al (1996) reported PAH
degradation was greater in grass vegetated soil than in non-vegetated soil. Ferro et al
(1999) reported that fluoranthene, pyrene, chrysene had greater mineralization in
vegetated soils compared to non-vegetated soils. Past day 14, Yoshitomi and Shann,
(2001) reported 14C-pyrene mineralization was significantly greater by com root
exudates as compared to non-vegetated soil.
45


Also, many chlorinated compounds have been mineralized from rhizosphere
degradation. High numbers of methanotrophic bacteria which have been shown to
degrade trichloroethylene (TCE) were detected in rhizosphere soils and on roots of
Lespedeza cuneata and Pinus taeda compared to unvegetated soils (Brigmon et al,
1999). Anderson and Walton (1995) reported TCE mineralization was greater in the
Chinese lespedeza, loblolly pine, and soybean than in non-vegetated soil. Possible
loss of TCE and 1,1,1 trichloroethane (TCA) was observed with alfalfa rhizosphere
(Narayanan et al, 1995) while no degradation of TCE was identified in the
rhizosphere of hybrid poplars (Newman et al, 1999). Chlorinated pesticides were
shown to have enhanced degradation in the rhizosphere (Shann, 1995).
Other pesticides have also been researched for superior degradation in the
rhizosphere. Enhanced degradation was reported for: atrazine grown in Kochia
species (Anderson et al, 1994), carbofuran found in the roots zone of com
(Buyanovsky et al, 1995), and parathion and diazinon (organophosphate insecticides)
grown with bush beans (Hsu and Bartha, 1979). Also, 2,4 Dichlorophenoxyacetic
acid (2,4-D) degraders were more abundant in sugarcane rhizosphere soils than non-
rhizosphere soils (Sandmann and Loos, 1984). Consistently, biodegradation of 2,4 D
and 2,4,5 trichlorophenoxy-acetic acid (2,4,5-T) was higher in rhizosphere soil as
| compared to non-rhizosphere soil (Boyle and Shann, 1995).
Polychlorinated biphenyls (PCBs) have extremely high physical and chemical
stability which have led to their use in numerous applications including hydraulic
j fluids, heat transfer fluids, and dielectrics. In the late 1960s, it was discovered that
j PCBs persisted in the environment longer than other manmade chemicals (Donnelly
! et al, 1994). This use and persistence has led to PCBs throughout our environment
and a fundamental need for remediation. Alfalfa was shown to be effective in
decreasing PCBs (Aroclor 1242, 1248, 1254, and 1260) as compared to non-vegetated
soil (Mahmannavaz, 2002). It has also been demonstrated that plant phenolics could
support the growth of PBC degrading organisms (Donnelly et al, 1994), especially
mulberry trees (Fletcher and Hedge, 1995). Upon death of the mulberry tree the fine
I roots have been shown to act as a source of substrate for PCB-degrading bacteria
j (Leigh et al, 2002).
Rhizosphere degradation has also been found to be efficient for the elimination of
explosives which are toxic and recalcitrant in nature. The most widely used
nitroaromatic compound is 2,4,6 trinitrotoluene, commonly known as TNT, which is
more recalcitrant than the mono dinitrotoluenes due to the location of the nitro group
I (Estev-Nunez, 2001). TNT degradation has been reported to be enhanced in the
rhizosphere (Kreslavski et at, 1999 and Siciliano & Greer, 2000) and shown to
46


express a nitroreductase which has the increased ability to detoxify, tolerate and take-
up TNT (Hannink et al, 2001). The explosives hexahydro-1,3,5 trinitro-1,3,5, triazine
(RDX) and octahydro-1,3,5,7 tetranitro 1,3,5,7 tetrazocine (HMX) were
investigated for rhizosphere degradation with two plants (myriophyllum aquaticum
and axenic hairy root cultures). Results suggest that RDX was degraded by both
plants and HMX was only slightly mineralized by the one axenic plant (Bhadra,
2001).
In summary, degradation of contaminants in the rhizosphere is very common.
However, only a few studies have evaluated the rhizosphere of polar trees finding no
TCE degradation, but higher populations of total heterotrophs, denitrifiers,
pseudomonads, BTX, and atrazine degraders. Moreover, no studies have evaluated
the capacity of the poplar tree rhizosphere to degrade MTBE. But first, the potential
of MTBE to degrade must be evaluated as is discussed in the next section.
TABLE 4.2 Rhizosphere Degradation Applications
PLANT/TREE RESULTS REFERENCE
Gasoline Contaminants (TPH, BTEX, ETC.)
Tall fescue, Bermuda grass, ryegrass, white clover After 245 months, degradation of TPH was: 50% in clover plots, 45% in fescue plots, 40% in Bermuda grass pots, and 30% in un vegetated pots. AATDF, 1998
Hybrid poplar trees A 5:1 ratio was found of BTX degraders in rhizosphere soil compared to unvegetated soil. Jordahl et al, 1997
PAHs
Com Reported higher degradation of pyrene by com root exudates compared to non- vegetated. Yoshitomi and Shann, 2001
Five grass species Found no significant degradation of atrazine and phenanthrene. Fang et al, 2001
Eight types of prairie Grasses The mineralization of PAHs (chrysene, benz(a)anthracene, benzo(a)pyrene, and bibenz(adi)anthracene) was greater in vegetated soil than non-vegetated soil. April & Sims, 1990
Perenial Rye Grass Fluoranthene, pyrene, chrysene had greater degradation in vegetated soils than non- v eg elated soils. Ferro et al, 1999
Fescue, sudangrass, switchgrass, alfalfa PAH (anthracene and pyrene) dissipation was greater in vegetated soil than n on- vegetated soil. Reilley et al, 19%
47


TABLE 4.2 (Contd) Rhizosphere Degradation Applications
PLANT/TREE RESULTS REFERENCE
Fescue, sudangrass, switch grass and alfalfa PAHs dissipation was greater in vegetated soil than non-vegetated soil at a MGP site. Pradham et al., 1998
Kochia species Enhanced degradation of atrazine was found in rhizosphere soil. Anderson et al, 1994
Clover and ryegrass Reported 66% and 42% reduction in chrysene and dibenz (ajilanthracene consecutively in the presence of arbuscular mycorrhiza (fungi) as compared to 56% and 20% in non -rhizosphere soils. Joner et al, 2001
Mulberry, Bermuda grass, and common sunflowers Microbially active rhizosphere fostered PAH mineralization. Olson et al, 2001
Chlorinated compounds
Lespedeza cuneata and Pinus taeda Detected high numbers of methanotrophic bacteria which are known to degrade TCE. Brigmon et al, 1999
Chinese lespedeza, a composite herb, loblolly pine, soybean TCE mineralization was greater in the Chinese lespedeza, loblolly pine, and soybean than in non-vegetated soil. An da-son & Walton, 1995
Alfalfa Possible loss of TCE and TCA was found with alfalfa rhizosphere. Narayanan et al, 1995
Hybrid poplars No degradation of TCE was found in hybrid poplars. Newman et al, 1999
Meadow bromegrass, perennial ryegrass, sweet vemalgrass Inoculated meadow bromegrass degraded TNT in soil up to 70% as compared to unplanted soils. Siciliano & Greer, 2000
Wheatgrass After 155 days, 22% of PCP was mineralized in vegetated soil while only 6 % was mineralized in non-vegetated soil. Ferro et al, 1994
Pesticides/herbicides
Sugarcane 2,4 D degraders were greater in rhizosphere soils than non-rhizosphere soils. Sandmann and Loos, 1984
Bush bean Mineralization of diazinon and parathion was enhanced in rhizosphere soils compared to non-rhizosphere soils. Hsu and Bartha, 1979
Rice Detected increased numbers of Gram- negative bacteria that could degrade propanil herbicide Hoagland, 1994
48


TABLE 4.2 (Contd) Rhizosphere Degradation Applications
PLANT/TREE RESULTS REFERENCE
Field grown plants Biodegradation of 2,4-D and 2,4,5-T was higher in rhizosphere soil as compared to non-rhizosphere soil. Boyle and Shann, 1995
Com Mineralization of carbofuran was enhanced in the roots zone in both laboratory studies and held studies. Buyanovsky et al, 1995
PCBs
Osage orange and | Rhizosphere supported growth of PCB mulberry | degraders Fletcher et al, 1995
Surfactants
Cattail and duckweed LAS and LAE (surfactants) was enhanced in the rhizosphere of the cattail as compared to non-vegetated soils. Duckweed soils only mineralized LAS. Federle and Schwab. 1989
Explosives
Com Detected decrease of TNT from 225 to 1 mg/kg in rhizosphere soil and to 11 mg/kg in non rhizosphere soil after 15 days. Kreslavski et at, 1999
Inoculated meadow bromegrass Degraded TNT in soil up to 70% as compared to unplanted soils Siciliano & Greer, 2000
myriophyllum aquaticum, axenic hairy root cultures RDX was degraded by both plant species and HMX was only slightly mineralized by the one axenic plant species. Bhadra, 2001
4.3 MTBE Degradation
Initial studies in the early 1990s found MTBE to be recalcitrant in nature. However,
more recently many researchers have examined the degradation of MTBE in
laboratory settings with positive findings. However, it is still not known conclusively
if MTBE will naturally degrade in the field. The degradation products of MTBE have
been identified as /er/-butyl-alcohol (TBA), tert-butyl-formate (TBF), formaldehyde
and ultimately carbon dioxide (Steffan et al, 1997). See the degradation pathway in
Figure 1-1.
4.3.1 Anaerobic MTBE Degradation
In 1993, Suflita and Mormile found that after 182 days there was no evidence of
degradation of MTBE under anaerobic conditions (sulfate-reducing and nitrate
49


reducing conditions). Yeh and Novak (1994) investigated the biodegradation of TBA,
MTBE and ETBE under the influence of nutrient availability, the addition of easily
degraded organics, and pH on unamended, denitrifying and methogenic conditions.
TBA was found to be the easiest to degrade, however MTBE was only found to
degrade within low organic content soil and a pH around 5.5. With the addition of
degradable organic compounds in the soil with low organic value degradation of
MTBE was inhibited. No degradation of MTBE was found in organic-rich soil.
Other researchers, Landmeyer et al (1998) and Finneran and Lovley (2001) also
reported no degradation of MTBE under anaerobic conditions. However when
Finneran and Lovley amended the sediment with humic substances and in which
Fe(III) was available as an electron acceptor there was variable degradation. The EPA
published a report in 2000 assessing natural attenuation of MTBE in the subsurface
under methanogenic conditions (Wilson et al, 2000). After 385 days of incubation,
removal of MTBE was significant both with supplemental alkylbenzenes and with
none. However, they also reported that for the site they investigated, with an average
concentration of 1200 ug/L MTBE, it would take 60 years to reduce the concentration
to 30 ug/L thereby making bioremediation an unfeasible remedial technology due to
the long time period for degradation. However, Somsamak et al (2001) and Bradley et
al (2001) showed anaerobic degradation of MTBE. Somsamak saw degradation
within 167 days under sulfate reducing conditions after 1160 days of incubation.
Bradley et al reported anaerobic degradation of MTBE was dependent upon electron
receptors. In the presence of methanogenic activity degradation was minimal but in
the presence of SO4 and NO3 there was significant mineralization of MTBE to CO2.
See Table 4-3 for a synopsis of MTBE anaerobic degradation studies.
TABLE 4.3 Anaerobic Degradation of MTBE Found in Literature
YEAR AUTHOR RESULTS
1993 Suflita and Mormile Results showed that anaerobically MTBE is recalcitrant.
1994 Yeh and Novak MTBE was only found to degrade within low organic content soil and a pH around 5.5 under unamended, denitrifying and methogenic conditions.
1998 Landmeyer et al Reported no degradation of MTBE under anaerobic (Fe(III) reducing conditions using aquifer material from a gasoline contaminated aquifer from 1993-1998.
50


TABLE 4.3 (Contd) Anaerobic Degradation of MTBE Found in Literature
YEAR AUTHOR RESULTS
2000 Wilson et al It was reported that after 385 days of incubation, removal of MTBE was significant both with supplemental alkylbenzenes and with none. However, they also reported that for the site with an average concentration of 1200 ug/L MTBE it would take 60 years to reduce the concentration to 30 ug/L.
2001 Finneran and Lovley It was reported that after 275 days sediment showed no degradation. Sediment amended with humic substances and in which Fe(III) was available as an electron acceptor showed variable degradation
2001 Somsamak et al Showed degradation within 167 days of MTBE to TBA under sulfate reducing conditions after 1160 days of incubation, but not under denitrification, Fe(III) reduction or methanogenesis.
2001 Bradley et al Anaerobic degradation of MTBE is possible for surface water sediments depending upon electron receptors.
4.3.2 Aerobic MTBE Degradation
Though initial studies assumed that MTBE was recalcitrant anaerobically, recent
studies have shown that under certain conditions MTBE may be biodegradable
aerobically. Many studies of aerobic degradation of MTBE have been reported with
mixed bacterial cultures. Salanitro et al (1994) conclusively showed aerobic
degradation of MTBE employing a mixed bacterial culture (BC-1) from a chemical
biotreater sludge plant. A lag time of 80-90 days was observed before removal
commenced. Likewise, Park and Cowan (1997) discovered a mixed culture from a
petroleum refinery activated sludge wastewater treatment facility that could degrade
MTBE. Salanitro et al (2000) reported MTBE biodegraded in microcosm studies
prepared from Port Hueneme, CA soil, a known MTBE contaminated site, for both
microcosms amended with a bacterial consortium MC-100 and unamended. For
unamended microcosms, there was a lag period of 2-3 weeks and it took 5-9 weeks
for MTBE level to be non-detect (<10 ppb), while for amended microcosms with 50
mg/L of MC-100 10-12 mg/L of MTBE was degraded in 2-3 weeks. Also, MTBE
was shown not to degrade in anoxic conditions. Another study reported a microbial
consortium (RS24) obtained from soil samples collected at the vadose zone,
groundwater interface, and below the groundwater surface at a MTBE spill site could
degrade MTBE both in the presence and absence of 2-propanol as a co-substrate. It is
hypothesized the low dissolved oxygen in the plume is inhibiting MTBE degradation
I
51


(Kern et al, 2002). The last two studies demonstrate that there can be naturally
occurring bacteria at a MTBE spill site that can degrade MTBE in the laboratory.
Mo et al, (1997) isolated a pure bacterial culture from activated sludge and the
Gingko tree that was shown to minimally degrade MTBE, but completely degrade
TBA. Hanson et al (1999) isolated a bacterial strain (PM1) from a mixed microbial
consortium in a compost bio filter that converted 46% of the MTBE to CO2 and 19%
to 14C labeled cells within 120 hours. They also inoculated the PM1 strain into
sediments from a groundwater plume at Port Hueneme, CA and found degradation of
MTBE when 20 ug of MTBE was amended to the sediment material. Hatzinger et al
(2001) evaluated a pure strain ENV735 and found it grew on MTBE slowly as a sole
source but it was greatly enhanced by the addition of a small amount of yeast. The
third pure bacterial strain, IFP 2012, which is able to grow on MTBE as a sole carbon
and energy source is gram positive and was isolated by using enrichment medium
containing TBA not MTBE; isolation came from an urban wastewater treatment plant
located neat Paris, France. The growth of IFP 2012 on MTBE alone was slow and
degradation was twice as high when grown on TBA. They reported the degradation of
MTBE was 100% with 68.7% mineralization after a 30 day incubation period
(Francois, 2002).
Some researchers have found that the addition of oxygen (Borden et al, 1997 and
Bradley et al, 1999/01) or oxygen and a mixed bacteria culture into sediments shows
MTBE degradation (Kane et al, 2001, Fortin et al, 2001,and Wilson et al, 2001).
Kane et al found degradation in microcosm studies from two out of the four MTBE
contaminated site in California that was investigated. The two sites showed
degradation in 15 days after a 4 day lag period, whereas the other 2 sites showed no
significant degradation of MTBE after 75 days. Thus, indicating that MTBE
degradation can vary appreciably between diverse field sites. Fortin et al used a
consortium degrading MTBE bacteria enriched with oxygen that showed a 79%
conversion of MTBE to carbon-carbon dioxide. MTBE (100 mg/L) degradation was
fast and followed zero order kinetics. The consortium was characterized as having a
very slow growth rate, a high affinity for MTBE, a low affinity for oxygen, and a low
biomass yield. Fortin et al suggests that the low affinity for oxygen could be the
reason that MTBE biodegradation in situ is slow or non-occurring.
In 1999, Bradley et al reported significant aerobic mineralization of MTBE sediments
to CO2 within 105 days. No degradation was detected under anaerobic conditions.
Bradley et al (2001) continued these studies and collected sediment from 11 sites in
South Carolina (both MTBE contaminated and non-contaminated) and discovered
significant aerobic mineralization of MTBE to CO2 within 50 days. The magnitude of
mineralization was related to grain size; high mineralization was detected within
52


I
sediments with a sand size grain between 0.125 and 2.0 mm. The lowest reduction of
MTBE was identified with clay-sized grains greater than 0.125 mm. Inhibition of
MTBE degradation was found in areas with higher organic content, which is
consistent with other researchers. It is proposed that this degradation in lake and river
sediments is due to their high aerobic condition.
Wilson et al (2001) reported 99.99% MTBE removal in the laboratory using a
continuously stirred batch reactor with high biomass retention (VSS > 600 mg/L), but
high concentration of TBA inhibited MTBE degradation. Pruden et at (2001), also
reported 99.9% MTBE removal with high biomass under 5 different substrate
conditions. Mixed cultures developed under the different substrate conditions,
however one organism, cytophaga-flexibacter, bacteriodes, was found in all cultures.
Unfortunately, this organism may not be found in general environmental conditions,
but may be found in some contaminated sites.
Deeb et al (2000 and 2001) and Sedran et al (2002) have investigated the
biotransformation of MTBE in the presence of BTEX. Many leaking underground
storage tanks house both MTBE and BTEX leading to commingled plumes. Deeb and
Alvarez-Cohen (2000) reported that two pure cultures, Rhodococcus sp. RR1 and
RR2 isolated from a mixed culture, shown to degrade BTEX had no degradation
effects on MTBE. Moreover, Deeb et al (2001) demonstrated that BTEX and MTBE
degrade independently from each other when using a pure culture PM-1; however
ethyl benzene and xylene explicitly inhibit the degradation of MTBE. Therefore, it is
surmised that MTBE degradation will not occur until it has migrated past the BTEX
plume. Conversely, Sedran et al reported that BTEX did not interfere with the
degradation of MTBE or TBA using a mixed bacterial culture (similar to PM1)
enriched on MTBE and BTEX. In contrast, the BTEX may assist in the development
of the biomass for TBA which would increase MTBE degradation. Sedran et al
presumed that the difference in these 2 studies is due to the use of a mixed bacterial
consortium compared to a pure culture that Deeb et al used. See Table 4.4 for a
synopsis of MTBE aerobic degradation studies.
I
53


TABLE 4.4 Aerobic Degradation of MTBE Found in Literature
YEAR AUTHOR RESULTS
1994 Salanitro et al Reported a mixed bacterial culture (BC-1) from a chemical plant biotreater sludge that could grow on MTBE (120 1200 mg/L) with a lag time of 80-90 days before removal commenced.
1997 Park and Cowan Discovered a mixed culture from a petroleum refinery activated sludge wastewater treatment facility that could degrade MTBE.
1997 Borden et al A first order degradation rate of 0.001/day was reported after a 20-day lag period with the addition of oxygen to sediments from a contaminated groundwater plume. They detected no degradation in anaerobic or low initial oxygen studies.
1997 Mo et al A pure bacterial culture shown to grow on MTBE as a carbon and energy source was isolated from activated sludge and the Gingko tree. It was shown to minimally degrade MTBE, but completely degrade TBA.
1998 Eweis Reported almost complete disappearance of 35 ppmv MTBE in air by use of a MTBE degrading culture inoculated onto a biofilter.
1999 Hanson et al A bacterial strain (PM1) shown to grow on MTBE as a carbon and energy source was isolated from a mixed microbial consortium in a compost biofilter that converted 46% of the MTBE to C02 and 19% to 14C labeled cells within 120 hours.
1999 Bradley et al Bed sediments from 2 UST spill sites in South Carolina were shown to degrade of MTBE in 105 days.
2000 Salanitro et al Microcosm studies (collected from Port Hueneme) for both amended with MC-100 (for 50 mg/L of MC-100 10-12 mg/L of MTBE was degraded in 2-3 weeks) and unamended microcosms microcosm (lag period of 2-3 weeks and it took 5-9 weeks for MTBE level to be non- detect) showed biodegradation.
2001 Bradley et al Sediments from 11 sites (both MTBE contaminated and non-contaminated) were investigated and showed significant aerobic mineralization of MTBE to C02 within 50 days.
2001 Hatzinger They found a pure strain ENV735 grew on MTBE slowly as a sole source but it was greatly enhanced by the addition of a small amount of yeast.
54


TABLE 4.4 (Contd) Aerobic Degradation of MTBE Found in Literature
YEAR AUTHOR RESULTS
2001 Fortin Researched the degradation of MTBE by a microbial consortium enriched in the laboratory with the addition of oxygen. The rate was fast 0.5-2.8 mg/L-hr converting 79% of MTBE to C02.
2001 Kane et al After addition of oxygen and a bacterium closely related to PM1 MTBE degradation was detected at two out of the four sites analyzed. The other 2 sites showed no degradation after 75 days.
2001 Pruden et al MTBE was successfully degraded in reactors. One organism, cytophaga-flexibacter, bacteriodes, was found in all cultures but may not be found in general environmental conditions. They used polyethylene porous membrane for high biomass concentration.
2001 Wilson et al Found MTBE degradation in the laboratory using a continuously stirred batch reactor (CSTR) with biomass retention under aerobic conditions.
2001 Deeb Ethyl benzene and xyiene explicitly inhibit the degradation of MTBE.
2002 Sedran Reported that BTEX did not interfere with the degradation of MTBE or TBA using a mixed bacterial culture (similar to PM1) enriched on MTBE and BTEX.
2002 Magar et al Reported promising results that aerobic degradation at two field sites without the addition of nutrient or bacteria.
2002 Kern et al This study does demonstrate that there can be naturally occurring bacteria (a microbial consortium RS24) at a gasoline spill site that can degrade MTBE in the laboratory.
2002 Francois The isolated a third pure bacterial strain, IFP 2012, which is able to grow on MTBE as a sole carbon and energy source. They reported the degradation of MTBE was 100% with 68.7% mineralization after a 30 day incubation period
4.3.3 Cometabolic MTBE Degradation
Several researchers have found degradation of MTBE cometabolically in the
laboratory. MTBE degradation has been shown with bacteria that utilize propane
(Steffan et al, 1997), alkane (Hardison et al, 1999 and Hyman et al, 1998),
cyclohexane (Corcho et al, 2000), pentane (Gamier et al, 1999) and ethanol
(Hernandez-Perez et al, 2001 and Piveteau et al, 2000).
55


Several researchers have reported no significant degradation of MTBE with oxidizers
of methane, toluene, ammonia or propylene (Hyman and Reilly, 1999) or with two
toluene grown Rhodococcus sp. strains (Deeb and Alvarez-Cohen, 2000).
Cometabolic degradation of MTBE using butane oxidizing arthrobacter (ATCC
27778) was publicized to be able to degrade low levels of MTBE (100-800 mg/L),
however, tow other strains (streptomyces griseus and methylosinus trichosporium
OB3b) were evaluated and neither were able to degrade MTBE (Liu et al, 2001).
MTBE vapor cometabolism with pentane using a culture of pentane-oxidizing
bacteria was shown to degrade MTBE, however degradation was not as great as the
mixed culture pure strain that have been demonstrated to mineralize MTBE
(Dusquiere et al, 2002).
Soil and groundwater were collected from three field sites down gradient from the
spill at Port Hueneme, CA and down gradient and at the source zone of a Department
of Defense Housing Facility, Navato, CA for evaluation of aerobic, anaerobic and
cometabolic MTBE degradation. Aerobic results showed complete degradation of
MTBE aerobically within 65 days for both down gradient sites. Cometabolic
degradation with butane and propane had similar results to the MTBE only
microcosm showing that in this study there was no cometabolic benefit over direct
MTBE oxidation. Negligible degradation of MTBE was discovered anaerobically
(Magar, 2002). See Table 4-5 for a synopsis of MTBE cometabolic degradation
studies.
TABLE 4.5 Cometabolic Degradation of MTBE Found in Literature
YEAR AUTHOR RESULTS
1997 Steffan et al Results showed that both a laboratory strain and natural isolates were able to degrade MTBE after growth on propane. The enzyme thought to be involved in the degradation process is a soluble P-450 enzyme- PMO.
1997 Hardsion et al Results indicate that n-alkane grown Graphium mycelid can cometabolically degrade MTBE by ctyochrome P450 catalyzed reaction to first TBF and then TBA. Unsaturated hydrocarbons like acetylene ethylene inhibited degradation
1998 Hyman et al Showed cometabolic degradation of MTBE by an alkane grown xanthobacter strain to TBF and then TBA
56


TABLE 4.5 (Contd) Cometabolic Degradation of MTBE Found in Literature
YEAR AUTHOR RESULTS
1999 Gamier Identified a bacterial strain known as Pseudomonas aerugina that could cometabolically degrade MTBE in the presence of pentane.
2000 Deeb and Alvarez- Cohen, 2000 No significant degradation of MTBE with two toluene grown Rhodococcus sp. Strains.
2000 Corcho A mixed culture grown on cyclohexane was shown to cometabolize MTBE to TBA.
2000 Piveteau Gordinia terrae IFP 2007 was shown to degrade MTBE to TBA and formate in the presence of ethanol, but could not degrade either TBA or formate. However, Burkholderia cepacia IFP 2003 could degrade TBA and formate, but not MTBE. The mixed consortium would degrade both.
2001 Liu et al Butane oxidizing arthrobacter (ATCC 27778) was publicized to be able to degrade low levels of MTBE (100-800 mg/L), while neither streptomyces griseus nor methylosinus trichosporium OB3b was shown to degrade MTBE
2001 Hernandez-Perez Demonstrated cometabolic degradation of MTBE with ethanol as the carbon source. They selected a Gordonia terrae strain IFP 2001 from activated sludge since it is known to grow on ethyl-tert-butyl-ether (ETBE) and detected both TBA and TBF but no formaldehyde.
2002 Dusquiere et al MTBE vapor cometabolism with pentane using a culture of pentane-oxidizing bacteria was shown to degrade MTBE, however degradation was not as great as the mixed culture pure strain that have been demonstrated to mineralize MTBE.
4.3.4 MTBE Degradation Field Studies
MTBE degradation in the field has been researched minimally. The well-known
Borden field site in Canada performed extensive studies of BTEX and MTBE
degradation in 1988. At the time of the study it was concluded that MTBE was
recalcitrant and showed only a small decrease in mass after 16 months. In 1995/96
57


another in depth groundwater sampling was conducted to find only 3% of MTBE left.
Volatilization, absorption, abiotic, and phytoremediation were ruled out as the
mechanisms for the loss of MTBE thereby indicating that biodegradation occurred.
However, they reported that biodegradation was very slow and it is unlikely this
could be an effective and timely remediation process (Schirmer, 1998).
Landmeyer et al (1998) analyzed the fate of MTBE in a gasoline-contaminated
aquifer and concluded that the main reason for decreases in MTBE is due to the
natural attenuation process of dispersion and dilution. However, Landmeyer et al
(2001) also reported degradation of MTBE in a relatively permeable gravel bed under
natural and induced aerobic conditions at a UST site with MTBE contamination since
1991 near Beaufort, SC. Groundwater contaminated with MTBE discharges through
cracks in a concrete pad that overlies a gravel bed with visible biofilm. This biofilm
was used in laboratory microcosm studies and found complete mineralization of
MTBE after 80 days, no intermediate products were detected. Field study results
reported degradation in both natural oxic conditions and induced oxic conditions by
addition of a magnesium peroxide slurry. This is a unique setting that shows the
possible potential of aerobic degradation of MTBE in the field.
Salanitro et al (2000) investigated MTBE degradation at Port Hueneme, CA at three
test sites located 450-m down gradient of the source. The three test sites consisted of
a control an oxygenated site, and a site with added oxygen and a MTBE degrading
bacterial consortium (MC-100). Results indicated that both the oxygen only and
oxygen plus MC-100 site caused MTBE degradation as compared to the control. In
the oxygen only site, MTBE degradation was not observed until 173-230 days after
commencement of oxygen while MTBE degradation in the MC-100 plus oxygen site
was seen 30 days after addition of MC-100.
In 2002, Wilson et al continued their research by conducting experiments with
sediments from the Vandenburg Air Force Base, CA and in the field. Microcosm
studies conducted with sediments from the site and 3000 ug/L MTBE solution and
oxygen degraded significantly in 20 days and TBA was detected variably through the
experiment. They field pilot study was performed in 2 lined trenches while oxygen
was emitted through a well. They reported that adding oxygen to a MTBE
contaminated-site stimulated organism that led to MTBE degradation though no TBA
was detected. A lag time of 2 months and pseudo first order rate of 5.3/day were
reported. Refer to Table 4-6 for a synopsis of degradation of MTBE in the field.
58


I
TABLE 4.6
Degradation of MTBE in the Field Found in Literature
YEAR AUTHOR RESULTS
1998 Schirmer Reported that biodegradation was very slow and it is unlikely this could be an effective and timely remediation process.
2000 Salanitro Investigated MTBE degradation at Port Hueneme, CA (a known MTBE contaminated site). Results indicated that both the oxygen only and oxygen plus MC-100 site caused MTBE degradation as compared to controls.
2001 Landmeyer et al Reported degradation of MTBE in a relatively permeable gravel bad under natural and induced aerobic conditions. The rate was fast 0.5-2.8 mg'L-hr converting 79% of MTBE to C02.
2002 Wilson et al Reported that adding oxygen to a MTBE contaminated site (Vandenburg Air Force Base, CA) stimulated organisms that led to MTBE degradation after a 2 month lag period with a degradation rate of 5.3/day.
4.3.5 MTBE Degradation Summary
In summary, the overall results are inconclusive regarding the potential for microbial
degradation of MTBE in natural in situ conditions. Many studies in laboratory
settings show MTBE may degrade quite rapidly aerobically, particularly when
augmented with MTBE-degrading microbes and/or supplemented with co-substrates.
Some field studies suggest rapid biodegradation of MTBE (short half lives on the
order of days, e.g., Wilson et al (2002)) in sites with aeration and /or co-substrate
additions. Other field studies have shown that sites behave very differently and
simply adding oxygen will not always produce the desired degradation results.
Timing is also an issue along with adding non-indigenous bacteria in-situ. No study to
date has specifically and quantitatively studied the potential for MTBE degradation in
the rhizosphere of plants. It is unknown if the rich and diverse microbial communities
found in plant rhizospheres would support rapid MTBE degradation similar to that
which has been reported at some field sites.
59


It is also essential to know if MTBE degradation in rhizospheres is an vital
phytoremediation mechanism compared to phytovolatilization, which occurs rapidly
over a period of days {Rubin and Ramaswami, 2001; Hong et al, 2001), even in small
laboratory saplings. Additionally, MTBE lignification has been proposed as a
potential phytodegradation mechanism in large woody trees {Ramsden, 2000) which
is evaluated in Phase 4. Before embarking on a study of MTBE degradation in large
trees, it is important to quantify any confounding effect caused by degradation of
MTBE in rhizosphere soils.
4.4 Experimental Procedures
The objective of phase 2 was to investigate the potential for microbial degradation of
MTBE in the rhizosphere of poplar trees in a time frame relevant for
phytoremediation. The research was conducted in three experiments:
1. Soil MTBE die-away tests were conducted in which attenuation of MTBE
concentration in soil was compared across vegetated and non-vegetated
systems,
2. Closed chamber tests were conducted in which an MTBE mass balance
deficit in planted systems was compared across hydroponic and
rhizosphere environments to determine any enhanced biodegradation of
MTBE in the presence of soil, and
3. Aerobic batch biodegradation experiments were conducted in which soils
previously acclimated to MTBE in a rhizosphere setting were
subsequently exposed to a known MTBE concentration in a controlled
aerated bioreactor.
Thus, aggregated vegetated systems were studied in Experiment 1, while the
rhizosphere sub-system was isolated and studied in a more controlled fashion in
Experiments 2 and 3, to conclusively demonstrate the presence or absence of MTBE
biodegradation.
Aqueous MTBE and TBA solutions were prepared from dilutions made from neat
chemicals obtained from Supelco, Bellefonte, PA. Poplar cuttings were obtained from
The Colorado State University Forestry Services and were grown in a greenhouse and
fed nutrients and sprayed with pesticides as needed. Two and a half feet tall poplar
saplings were used in the first experiment and soil from these saplings in the third
experiment, while smaller saplings were used in the second experiment.
60


4.4.1 Experiment 1: Soil Die-Away Studies
Soil-MTBE die away studies were conducted in which attenuation of MTBE
concentration in soil was compared across unvegetated and vegetated systems. Three
planted pots were compared with three unvegetated pots; each pot contained 9 Kg of
soil (dry basis). All pots were irrigated with low concentrations of MTBE (10-20
ppb) for six weeks to acclimate the microorganisms in the soil to MTBE. Once the
soil was acclimated, each of the pots were sub-irrigated once with 400 ml of water
j containing 8.4 mg/L MTBE, thereby yielding ~ 400 ppb MTBE concentration in dry
soil. The soil was then covered with several layers of aluminum foil to prevent
| volatilization. The initial soil moisture average was 40%. Weight change of the pots
| was monitored and enabled assessment of evapotranspiration; neither planted nor
unplanted pots were given extra water during the experiment. Samples were
I collected every three days from three different locations in each pot. Core samples
, were collected from the top of the pot to the bottom of the pot and mixed before
putting into sample jars.
I
4.4.2 Experiment 2: Chamber Studies
Phytovolatilization of MTBE had previously been studied in hydroponic jar systems
designed to eliminate volatilization in Phase 1 (Rubin and Ramaswami, 2000). The
purpose of this study was to include rhizospheric soils into the hydroponic jar systems
to quantify any enhanced degradation of MTBE due to the presence of soil microbes.
In this experiment, poplar cuttings were grown by inserting the cuttings inside the
septa of lids that fit 250-ml jars so that their roots and shoots developed separately.
The same set-up as discussed thoroughly in Phase 1. There were two systems: a
hydroponic planted system in which a poplar sapling with fully washed roots (no
rhizosphere) was placed in a bag containing 186 ml of 3000 ppm MTBE solution,
and, a rhizosphere soil-plant system. The rhizosphere planted system jar was set-up in
the same manner as the hydroponic planted system, but the poplar sapling was
transplanted along with rhizosphere soils (200 g) and microbes associated with the
| roots and then dosed with the 3000 ppm MTBE solution. As in all the tests, the
rhizosphere soils had been acclimated to low levels of MTBE for six weeks prior to
| commencement of the experiment. Approximately 50-ml of air were left in the
| rhizosphere root system to enhance aerobic degradation. Each jar system was then
i placed within a larger 25 L glass chamber with air flowing through carbon tubes to
capture any MTBE released by the poplars to air. See Figure 4.1, 4.2, and 4.3.
The chamber test ran for five days during which the plastic bag-enclosed root system
compressed during water transpiration from the jars, thereby minimizing MTBE
volatilization from the root solution. Upon termination of the test after 5 days,
61


samples of the water, soil, roots, shoots, and air were collected and analyzed for
MTBE and TBA. MTBE mass recovered at the end of the experiment was computed
as the sum of the MTBE mass measured in the final aqueous or soil solution,
biomass, and air traps. A significant mass balance deficit between initial and final
MTBE masses in the systems would indicate degradation of MTBE. A significantly
greater mass balance deficit in the rhizosphere planted system compared with the
hydroponic planted system would suggest MTBE degradation in rhizosphere soils.
FIGURE 4.1 Drawing of Poplar Cutting inside 25-L Chamber
62


FIGURE 4.2 Photo of Plant in Chamber
63


I
4.4.3 Experiment 3: Aerobic Batch Biodegradation
Experiments
The size and complexity of the jar systems in the chamber tests did not allow for
tracking of oxygen consumption in the root zone, which could be correlated with
aerobic microbial activity. In order to conclusively prove the presence or absence of
MTBE degradation by actively thriving rhizosphere microbes in an aerobic
environment, a controlled slurry bioreactor study was conducted in which microbial
activity was tracked by oxygen consumption, while MTBE removal was
simultaneously tracked by both aqueous MTBE and TBA concentration
measurements. The aerobic batch biodegradation experiment consisted of placing 25
grams of rhizosphere soil (acclimated with 10-20 ppb MTBE solution for one month
prior to commencement of experiment) in 250-ml of autoclaved water spiked with
MTBE to yield a 150 ppb MTBE solution, in a closed 500 ml BOD bottle. The 25-
grams of rhizosphere soil included small roots along with the soil directly adjacent to
the roots, such that any root exudates that would promote MTBE degradation would
64


also be present. The dissolved oxygen in the aqueous suspension was measured
throughout the two-day experiment by Dissolved Oxygen Meter 840041 from Sper
Scientific. The experimental observations were compared to a control bioreactor that
consisted of the same set-up without any rhizosphere soil.
This experiment was repeated for a longer time period (10 days) with oxygen
sparging at the beginning and middle of the experiment at a low MTBE
concentration, 100 ppb, and a high MTBE concentration, 100 ppm. See Figure 4.4.
The aerobic batch biodegradation experiment consisted of again placing 25 grams of
rhizosphere soil (acclimated with ppm MTBE solution for two months prior to
commencement of experiment) in 250-ml of autoclaved water spiked with in a closed
500 ml BOD bottle. The 25-grams of rhizosphere soil included small roots along with
the soil directly adjacent to the roots, such that any root exudates that would promote
MTBE degradation would also be present. The experimental observations were
compared to a control. The control is the same set-up as the experiment but without
rhizosphere soil.
For all experiments MTBE and TBA in water, soil, and biomass were analyzed by a
SRI 8610 C GC (equipped with a flame ionization detector [FID]) using a purge and
trap method with a Restek MXT-1 Column (60, 0.53, 5.0 um) from SRI, Instruments.
This analysis method was based on EPA Method 8260. The initial program
temperature is 40C, with a 10 ramp and final temperature of 140. Method
detection limits were 10 ppb for MTBE and 25 ppb for TBA with a relative reference
factor of less than 10% and 15%, respectively. Roots and shoots were first placed in
liquid nitrogen and then blended before analysis. The MTBE in air that was trapped
onto carbon tubes was analyzed by GC Modified Method NIOSH 1615 with a carbon
disulfide extraction.
65


FIGURE 4.4 Picture of Experiment 3
66


4.5 Results
Results from experiment 1 are presented in terms of average concentration of MTBE
in soil (dry basis) plotted as a function of time. Rapid decrease in soil MTBE
concentrations were observed over a 2 week period in both vegetated and unvegetated
systems spiked with MTBE. Volatilization seems to be the significant reason for the
loss of MTBE. However, as shown in Figure 4.5, poplar tree soils showed no
significant difference in MTBE concentration when compared with the unvegetated
soils, based on a one tailed t-test at the 10% level of significance. No TBA was
detected in either vegetated or unvegetated soils.
Experiment 2 chamber tests enabled capture of any MTBE volatilized from
hydroponic planted and rhizospheric planted jar systems. By assessing MTBE
concentration and mass in all three media air, water, and biomass a mass balance
of MTBE in the system was computed as shown in Table 4.7. Note, due to the
presence of rhizosphere soil, a much smaller amount of aqueous MTBE solution
could be accommodated in the rhizospheric-planted jar system compared to the
hydroponic-planted jar system. Hence, although MTBE concentrations in solution
were the same, initial MTBE masses in hydroponic-planted and rhizospheric planted
system are quite different. MTBE mass recovery ranged from 86% to 111% in
hydroponic-planted and rhizospheric-planted systems. The mass recovery range of 86
to 110% is well within the analytical error of measurements, and consequently the
mass balance data also indicate no significant MTBE degradation either within plants
or within the rhizosphere soil. Much of the MTBE removed from water was
recovered from air, and thus MTBE uptake and phytovolatilization appears to be the
primary phytoremediation mechanism. Note that MTBE uptake and transpiration by
plants occurs very rapidly, with much of the MTBE transpired to air over a 5 day time
period.
67


FIGURE 4.5 Attenuation of MTBE Concentrations in Unvegetated Soils and
Those Planted with Poplar Trees.
TABLE 4.7 MTBE Mass Balance
System Initial MTBE in jar Final MTBE in jar MTBE in roots MTBE in shoots MTBE in air % MTBE Recovered
ug g g g g
Hydroponic- planted 545 261 0.65 0.15 205 86
Rhizo spheric -planted 60 0.44 0 0.01 66 111
In order to conclusively demonstrate that the microbes in the rhizosphere were alive
and respiring, yet unable to degrade MTBE, a controlled experiment 3 bioreactor
study was conducted which simultaneously tracked dissolved oxygen, MTBE and
TBA. Results from two day Phase three experiments are shown in Figure 4-6 and 4-7.
The bioreactor study progressed for about 2 days during which aerobic conditions
68


existed as shown by dissolved oxygen monitors. Figure 4.6 shows that the
concentration of MTBE in both the control and the rhizosphere bioreactors did not
change between the initial and final samples showing no MTBE degradation.
Furthermore, no TBA was detected with a detection limit in water of 25 ppb.
However, the dissolved oxygen in the rhizosphere bioreactor was consumed to almost
zero in two days showing that actively-respiring microorganisms were present in the
system, most likely metabolizing the root materials included with the soils. In
contrast, the dissolved oxygen in the control bioreactor (with no soils) remained the
same, demonstrating integrity (no leaks) of the closed bioreactor experiment.
The results from the longer bioreactor experiment for the low and high MTBE
concentrations are shown in Figure 4.8 and 4.9. Once again, even with oxygen
sparging and a longer running experiment the MTBE concentrations in the
rhizosphere bioreactors did not change between the initial and final samples and no
TBA was detected confirming no MTBE degradation. These results show that
microbes acclimated to low and high-levels of MTBE for periods of more than 1
month and then exposed to both high and low MTBE concentrations in aerated
aqueous systems are still unable to biodegrade MTBE at any significant rate
compared with the more rapid rate of MTBE phytovolatilization as indicated by the
large amount of MTBE recovered shown in Table 4.7.
In summary, for MTBE removal from groundwater and soil, phytovolatilization is the
dominant pathway in comparison to rhizosphere degradation. Therefore, the next
chapter will evaluate measurement and modeling techniques for transpiration for
quantification of MTBE removal from the subsurface due to phytovolatilization.
69


FIGURE 4.6 Aqueous MTBE Concentrations in Control Bioreactors (no soil)
and Rhizosphere Bioreactors
FIGURE 4.7 Dissolved Oxygen Concentrations in Control Bioreactors (no soil)
and Rhizosphere Bioreactors
70


Initial Control 1 Control 2 Final
FIGURE 4.8 Low Aqueous MTBE Concentrations in Controls (no soil) and
Rhizosphere Bioreactors
Initial Control 1 Control 2 Final
FIGURE 4.9 High Aqueous MTBE Concentrations in Controls (no soil) and
Rhizosphere Bioreactors
71


5. Phase 3 Evapotranspiration of Poplar Trees
5.1 Introduction
Trees are effectively working as solar pumping systems, thereby uptaking
groundwater and in many areas contaminated natural groundwater. It is essential to be
able to quantify this groundwater uptake by trees to design efficient phytoremediation
systems for remediation of contaminated groundwater. In phase 1, it was established
that MTBE in groundwater is taken up at approximately the same rate as water, with a
TSCF of 1. In phase 2, it was shown that there is no significant degradation of MTBE
in the rhizosphere of poplar trees at time scales relevant for phytoremediation.
Therefore the total removal of MTBE from groundwater can be determined from the
evapotranspiration (ET) rate of the tree. Phase 3 will evaluate ET of poplar trees.
ET can be determined by actual measurements or through predictive ET mathematical
models. This chapter will: 1.) define ET, 2.) discuss predictive mathematical ET
models, 3.) show estimation of parameters for the Penman-Monteith equation, 4.)
present validation of the Penman-Monteith equation, 5.) identify measurement
methods of ET, 6.) describe calibration of ET measuring devices, 7.) present
transpiration rates in an enclosed chamber, 8.) explain the experimental design to
measure ET, and 9.) present the results that compare measured ET rates to the
predictive models. The idea is that measurement of ET is not always practical or
affordable and that a predictive ET model can be shown to closely predict ET of
poplar trees. This will make evaluation of phytoremediation as a potential
remediation technology less complicated to the consultant and more acceptable to the
regulator.
Water evaporates from many surfaces such as: the oceans, lakes, streams, reservoirs,
and soil. Vegetation can also transpire water from the subsurface into the atmosphere.
This water, from either evaporation or transpiration, is transferred by wind and
weather and then condenses and can fall back down to earth somewhere many miles
away from its entrance into the hydrologic cycle. Once the precipitation falls back on
the earth there are several pathways the water can follow. These pathways consist of
depression storage; temporary storage on the ground as snow, ice, or water puddles;
overland flow, precipitation that drains across the land to a stream or water body; and
infiltration, which is the precipitation that seeps into the ground. All groundwater
eventually flows into surface water bodies that drain to the seas/oceans. Water from
the rivers, seas, and oceans evaporate to form clouds that cause precipitation. This
process is cyclical with no starting or ending point.
72


The hydrologic cycle is a dynamic process as shown in Figure 5.1.
FIGURE 5.1 The Hydrologic Cycle
5.1.1 Evaporation
Evapotranspiration is the combination of evaporation and transpiration. See Figure
5.2 (Ward and Elliot, 1995). Evaporation is the net occurrence of water escaping
from a liquid state to a gas state. Water molecules transport themselves to gas
molecules from energy gained from solar radiation. This energy is called the latent
heat of vaporization, which is used in many of the formulas that calculate ET as
discussed in the next section. Many studies have shown that when water is not a
limiting factor, the rate of evaporation depends heavily on solar radiation. Other
important parameters when estimating evaporation are the concentration gradient of
the water vapor in air, wind speed, air and water temperature, atmospheric pressure,
and quality of the water (Ward and Elliot, 1995). These parameters control the rate of
mass transfer of water vapor from the surface of a leaf to the bulk atmosphere.
The rate that water vapor will go into the air is determined by the difference in
concentration of water molecules at the water surface and the bulk air surface. When
there are limited water molecules in the air space the water molecules will move
faster from water to air. Wind continues to blow and disperse the water molecules
making room for more water molecules. Therefore, the greater the wind speed and the
73


there are limited water molecules in the air space the water molecules will move
faster from water to air. Wind continues to blow and disperse the water molecules
making room for more water molecules. Therefore, the greater the wind speed and the
larger the concentration gradient in air, the larger the water evaporation of a tree will
be (Linsley et al, 1949). As shown in Figure 5.3 the higher the temperature the larger
the evaporation rate (Nemec, 1972). In summary, the greater the wind, temperature,
and solar radiation the greater the evaporation of water will be.
Instruments that measure actual evaporation are called atmometers and are divided
into three classes: tanks or pans, porous porcelain bodies, and wet paper surfaces.
The porous porcelain body is the best device used for measuring evaporation from
plants, since the porous membranes most resemble plants (Linsley et al, 1949).
FIGURE 5.2 Evapotranspiration Models
74


FIGURE 5.3 Temperature vs. Evaporation
5.1.2 Transpiration
Transpiration is the movement of water through the plant from the subsurface. The
plant obtains the water by its roots; the water travels through the plant and is finally
evaporated through the leaves. The leaves let water out of the plant system which
allows the water to evaporate into the atmosphere. The more water that is released
from the leaves the more water the roots must uptake. The basic system in which
leaves release water to the atmosphere is shown on the next page in Figure 5.4a.
Water is spaced throughout the spongy mesophyll cells (see Figure 5.4a) in the leaf
and is released by the stomata cells on the outside of the leaf. For a tree, the stomata
cells are on both sides of the leaf, however, are generally larger and more abundant
on the bottom side of the leaf (Kowlowski and Pallardy, 1997). The stomata density
ranges from 50-500 per mm2, a length of 10-30 pm, a mean width of 0 (closed) -10
(frilly open) pm and a total pore area of 0.3-1% of the leaf area (Jarvis, 1976- voll).
A form of energy is needed to release the water from the leaf this energy form is
direct solar radiation.
75


Boundary Layer
T~J
n n
o%o o
r~nr
h2o
Stomata Pore
Upper Epidermis
Palisade Mesophyll
Cells
Lower Epidermis
Guard Cell
Boundary Layer
FIGURE 5.4a Release of Water from the Leaf
5.1.3 Stomata
Stomata open and close due to the changes in the guard cells turgor pressure. Turgor
is a response to the accumulation of solutes that decrease osmotic pressure. The
decrease in osmotic pressure decreases the guard cell water potential, resulting in
water uptake and increased turgor. The increased turgor causes stomata opening.
Closure of the stomata is the reverse of opening. Refer to Figure 5.4b for a picture of
the guard cell. Though the complete mechanisms of the stomata are not yet fully
understood. It is known that the stomata will close due to:
A lack of water (,Jarvis, 1976- voll),
Internal carbon dioxide is high (Gates, 1980),
Very high wind conditions (Gates, 1980), and
High humidity (Gates, 1980).
76


And, the stomata will open due to:
Atmospheric conditions below normal for carbon dioxide (Jarvis, 1976- voll),
and
Sunlight (Gates, 1980).
Even though the exact mechanisms of the stomata are not known, the stomata
resistance is affected by several environmental factors including water potential,
illumination, temperature, and humidity (Gates, 1980). There are even more factors
that influence transpiration on a whole such as: air temperature, air moisture, wind
velocity, precipitation amount, soil properties, root characteristics, solar radiation,
optical properties (reflectivity, emissivity, etc.), and stomata response (Janssen,
1994).
FIGURE 5.4b The Guard Cell
Studies completed by many researchers have shown that for cotton, grain sorghum
and alfalfa about 80% of water from the subsurface can be utilized before the stomata
starts regulating evaporation (Ritchie, 1973). It has been observed that for a single
tree with a minimal canopy, stomata movements do control water loss from the leaf.
However, in dense canopies radiation and wind speed are more relevant factors in
measuring ET (Hinckley et al, 1994). Most models assume that there is enough water
within the soil system; however the roots must be able to reach water (which at times
can be a problem for phytoremediation). During dry times, the maximum
transpiration will be related to the amount of water the roots can acquire (Monteith,
1995).
See Table 5.1 for a list of measured ET rates for different types of trees.
77


TABLE 5.1 Typical Values of Transpiration Rates of Trees
TYPE OF TREE LOCATION YEAR SAP FLOW (per tree) REFERENC E
Hybrid Poplar East Argonne, IL August & September 2000 1.2 to 10.5 L/day Ferro et al, 2001
Hybrid Poplar (2 years old) Arberdeen, TX May 1997 6.0 to 41.2 L/day 1.6 to 10.9 gallons/day Ferro et al, 2001
Hybrid Poplar Moore County, NC July 2000 1.0 to 1.8 L/day m2 leaf area 15600 m2 to 51700 m2 leaf area/acre Ferro et al, 2001
Eastern Cottonwood (1 year old) (Populus Deltoides Bartri) Carswell Air Force Base, TX May to October 1997 1.6 to 9.2 L/day Ferro et al, 2001
Imperial Carolina (1-5 years old) (Populus Deltoides x Populus Nigra, DN 34) Ogden, UT September & October 1998 6.47(1 year old) to 25 (5 year old) gallons/day Ferro et al, 2001
Hybrid Poplar (4 years old) Near Sumner, WA Midsummer 1992 20-51 kg/day Hinckley et al, 1994
Eucalyptus nitens ( 4 & 8 vears old) Northeastern Tasmania December 1996 to January 1997 4.5-17.71 L/day Hunt and Beadle, 1998
Carya illinoensis Central Texas 1990 100-159 kg/tree Steinberg et al, 1990
78


5.1.4 Resistances
There are many resistances to water movement in a leaf. The most important
resistances to evaluate when measuring ET are the stomata and aerodynamic
resistances. The aerodynamic resistance is the resistance offered by an atmospheric
boundary layer on the leaf surface that controls the convective heat transfer between a
canopy and the atmosphere. The leaf resistance includes the intercellular air-space
resistance, r (which is very small), the stomata resistance, rs, and the cuticular
resistance, rc. Resistances from the leaf (p) and from the aerodynamic resistance (ra)
are added in series, while the leaf resistances (rj, rs, rc) are added in parallel
(Shuttleworth, 1991). See Figure 5.5a for a figure of the resistances.
FIGURE 5.5a Resistances (Ramaswami etal, 2003)
The stomata resistance of the leaf is related to light intensity in a hyperbolic fashion
as shown in Figure 5.5b. As light is reduced to darkness the resistance becomes very
high and represents the cuticular resistance of the leaf. Therefore, when measuring
ET, the stomata resistance (rs) measurement is used for daytime calculations while the
cuticular resistance (rc) measurement is used for nighttime calculations. See Table 5.2
for average values of stomata resistance with low amount of wind, cuticular
resistance, and boundary layer resistance (Gates, 1980). The various mass transfer
resistances presented above may be combined to predict ET, as described next.
79


FIGURE 5.5b Stomata Resistance of the Leaf vs. Light Intensity (Gates, 1980)
TABLE 5.2 Average Values of Stomata Resistance with Low Amount of Wind, Cuticular Resistance, and Boundary Layer Resistance (Gates, 1980)
rs rc ra
Populus tremula 230 60
Betula Verrucosa 120 7000 80
Quercus robur 1160 29000 90
Acer plantanoides 790 10100 80
Circaea lutetiana 1240 9000 50
Lamium galeobdolon 940 3000 80
Helianthus annuus 40 50
Units are seconds per meter. All plants are hypostomatous. (From Holmgren et
al., 1965)
80