Chapter 2. Phytoremediation of Soils Contaminated with Organic Pollutants

S. D. CLTNNINGHAM E T AL.



56



WII. Current Phytorernediation Research and Development

A. Petroleum Contamination

B. Chlorinated Hydrocarbons

C. Pesticides

D. Biotechnological Improvements in Phytoremediation

IX. Conclusions

References



I. INTRODUCTION



A. “REMEDIATION” OF SOILSCONTAMINATED

WITH

ORGANIC

POLLUTANTS

The remediation of soils contaminated with organic pollutants is a global

problem that consumes considerable economic resources of industries and governments alike. Contaminated soils are widespread. They are found on all continents and are often associated with centers of population, petroleum handling

and storage areas, and areas of significant manufacturing activities. It is estimated that over the next 30 years 750 billion U.S. dollars will have to be spent in

the United States alone to remediate contaminated sites to current legal standards

(Russell et al., 1991). Although this figure contains both soil and water remediation costs, current breakdown (at 25% soils and 75% groundwater) suggests that

in the United States alone $6 billion/year will be required to remediate soils over

the next 3 decades.

For the most part soils contaminated with organic pollutants are remediated

using a diverse set of thermal, chemical, and physical methods that strip the

contaminants from the soil (Nyer, 1992). In addition to these physical and chemical treatment methods, microbial-based remediations have become more common

in the last 2 decades (USEPA, 1992). The first biotreatment systems targeted

petroleum contaminants in shallow groundwater. These early systems relied on

stimulation of naturally occurring aerobic populations to degrade the contaminants; for the most part this was accomplished by adding nutrients and increasing

oxygen flux within the contaminated zone. Microbial remediation of groundwater

contaminants has since expanded to target other organic pollutants and anaerobic

processes as well. Concurrent with these developments in groundwater bioremediation, some soil bioremediation techniques have been developed including

slurry reactors, composting, and “bioventing.” However, all biological-based

processes remain a relatively minor component of the total remediation market on

both a volume and a cost basis (Staff, 1993).

All remediation techniques are done either in place (in situ), or by removing



PHYTOREMEDIATION OF SOILS



57



the contaminated material for treatment (ex siru). The nature of the contaminant

and the soil matrix often determine the type and the location of remediation that

are appropriate. Certain contaminants are inherently more easily remediated than

others based on their chemical, physical, or biological properties. Contaminants

that are either water soluble or volatile are often remediated by pushing or

pulling air or water through the matrix. Techniques that fall into this class include

sparging, vacuum extraction, and leaching. Other contaminants, however, are

inherently much more difficult and expensive to remediate. These contaminants

tend to be relatively nonvolatile and non-water soluble. Some engineering-based

remediation techniques work to increase the extractability of the more difficult

contaminants by the application of heat, surfactants, or acids, or by physical

manipulation; however, intensive handling processes tend to greatly increase

costs and are fraught with difficulties.

All remediation techniques either remove the contaminant from the polluted

matrix in a process called “decontamination” or sequester the contaminant via

“stabilization.” Decontamination techniques include thermal desorption and matrix removal. Alternatively, the chemical and physical nature of the matrix may

be altered to sequester the contaminant in the matrix, a process known as stabilization. The most common stabilization technique is through the addition of cement. Stabilization techniques are often acceptable in cases where reduced biological availability and off site movement can be demonstrated to the satisfaction

of the regulatory and public community. All stabilization techniques are followed

by site management practices that continue to contain the pollutant on site and

further reduce future environmental risks. Although decontamination and stabilization techniques have different costs, and reach different endpoints, the term

“remediation” is accurately applied to both.

In many cases site managers would prefer decontamination processes. This

type of remediation allows for increased flexibility in future land use planning

and increases the value of the property. Certain sites, however, due to the nature

of the pollutant, site location, extent of contamination, or the human and environmental risks involved in excavating, are potentially more appropriately remediated through a sequestration/stabilization technique than through a decontamination technique.

In general, surface soils contaminated with either volatile or water-soluble

organics tend to be poor targets for the development of new remediation technologies. This is because many of the existing techniques are relatively inexpensive

and the extent of such soils is relatively limited. Therefore, we recommend that

new remediation technologies satisfy the following criteria: ( 1 ) they must have

relatively low cost; (2) they must address larger surface areas of soil; (3) they

must deal with relatively immobile contaminants; (4) they should address those

problems which currently have the greatest economic significance; and ( 5 ) they

must meet regulatory requirements.



S. D. CUNNINGHAM ET AL.



58



With those criteria in mind, and as a background for our ongoing research

programs and this chapter, we surveyed current industrial and governmental

remediation expenditures, assessed future liabilities, and examined the potential

for accidental releases in the future. This body of information indicates that

appropriate targets for the development of new in situ, surface-soil remediation

technologies for soils contaminated with organics are (in rank order of volumes

to be remediated as well as expected economics):

1.

2.

3.

4.



petroleum products and by-products

industry-specific chlorinated organics (PCBs, dichlorobenzenes, etc.)

industry-specific nitroaromatic compounds (TNT, DNT, etc.)

pesticide residues that are historic, “off label,” or accidental spills.



The development of an effective remediation system based on green plants, for

these contaminants, under the above criteria would seem to be technically feasible, economically viable, and socially responsible.

It should also be mentioned here that in terms of remediation opportunities,

surface contamination by inorganics represents an excellent opportunity for similar new technology development. This is especially true as there are still fewer

viable technologies for the remediation of inorganic contaminants of surface

areas. Progress in the phytoremediation of inorganic compounds has recently

been reviewed (Salt et al., 1995; Kumar et al., 1995; Baker et al., 1991) and is

currently the subject of numerous laboratory and field programs.



B. THEPROCESS

AND GOALS

OF SOILREMEDIATION

Air and water pollution is almost universally regulated in the developed world.

Air and water pollutants are transported across individual property and national

boundaries. There is a long-standing public mandate to regulate pollutants that

affect the air and water that we share. In addition to the public mandate, there is

strong legal precedent for such regulation. On this legal and popular basis many

countries have passed “Clean Water” and “Clean Air” regulations. These regulations provide a solid regulatory framework from which developers of new technology can derive targets and measure effectiveness.

The legal precedent for regulating soil contaminated with only sparingly soluble compounds is more ambiguous. The legal precedent in many countries is

that soil is owned by the individual and not held for the “communal good.” At

times, the regulation of contaminated soils has directly clashed with historic

property rights. The regulation of contaminated soils is therefore more complex.

In cases where the soil can be demonstrated to have a clear impact on surface or

ground water, this often sets the standards for legal debate. For soils with rela-



PHYTOREMEDIATION OF SOILS



59



tively immobile contaminants the regulatory framework is more ambiguous. It is

not surprising therefore that there is considerable variation in the definitions of

both “contaminated’ and “clean” soils.

The United States, like many countries, lacks a “Clean Soils” Act. This

nonuniform legal situation places developers of new technology in an uncomfortable place with no clear idea of the “problem” definition or when “acceptable

goals” or “endpoints” have been obtained. Similarly contaminated soils are

treated in remarkably dissimilar manners depending on the state, the proximity to

population and water Bow, and the method of contamination. A petroleum spill at

a well head is regulated differently than the same material spilled in a refinery. In

the first case the most liberal state’s clean-up requirement goes into effect when

petroleum levels exceed 1% (10,000 mg/kg) by weight. In the most stringent

case, clean-up is required when the petroleum levels exceed 10 mg/kg. These

two scenarios present two vastly different remediation challenges with potentially identical contaminated matrices. Remediating soil from 50 pg/g TPH (total

petroleum hydrocarbon) down to a level of 9 p.g TPHlg soil presents a very

different design criteria than remediation from 4% oil down to 1%. In addition,

these regulatory inconsistencies dramatically affect the total surface area to be

remediated at a given site.

Differences in clean-up levels also exist depending on the physical location of

the site within a state. One of the greatest concerns of regulatory agencies is risk

and risk management. When the contaminants have been identified for a given

site, risk is often determined based upon the potential health hazard of the

chemicals and the use of the land/soil. Noncarcinogenic contaminants are rated

according to a hazard index which is generally based on acute toxicity. Carcinogenic compounds are ranked according to the concentrations necessary to induce

cancer. The “point of departure” for the carcinogenic compounds is the concentration that causes one extra case of cancer in a population of one million

assuming long-term consumption or exposure. Land use is a very important

component of the determination of exposure. Soil near residential housing may

be viewed in a different light than soil that will be used as industrial fill. In some

states, for example, maximum allowable TPH content for residential areas is 100

mg/kg, whereas concentrations of 300 mg/kg are acceptable for industrial areas.

Regulatory assessments of contaminated soil will also determine whether or

not a new technology is acceptable based on the time required to complete the

remediation. The overriding factor in such a decision may be whether the time

required for the remediation will create an unacceptable risk. Two identical sites

with different land use (an area adjacent to a day care center or an industrial park)

can require different technologies based on speed of obtaining the endpoint. The

developers of new technology should be well aware that the goals, constraints,

and expected (or achievable) endpoints are currently highly variable.

Despite the variability in regulatory guidelines between states, the remediation



60



S. D. CLJNNINGHAM ET AL.



process will follow a fairly predictable chronology. The process always begins

with an event, an analytical result, findings of a construction project, etc. Analytical data is collected, and if analytical results are higher than predetermined cutoffs, a report is made to the regulatory authorities. The site is then sampled more

extensively and the nature, extent, and potential impacts of the contaminant are

assessed. Then, in discussion with the authorities, a decision on the need for

remedial action is made. At this point, the group responsible for the site further

investigates the site and proposes to the regulators a suggested remediation plan.

After a series of discussions is held, decisions are made on the method, timing,

endpoints, and monitoring that will occur.

If the clean-up occurs under a regulatory order, a legal document is produced

(e.g., “ROD or record of decision) which commits the parties involved to a

particular course of action. Once a ROD is signed there is considerably less

flexibility in what new technologies can be tried on a given site. There are

occasions, however, that a clean-up occurs under a voluntary arrangement and

considerably more freedom to experiment on these sites may be possible. The

developers of any new remediation techniques would be well advised to become

involved in the remediation process as early as possible.



C. THE

ECONOMICS

OF REMEDIATION

For the most part soils are remediated by engineering techniques. The nature

of these techniques depends on the volume of the soil to be remediated, the

physical and chemical properties of the pollutant(s), and the type of soil/sludge

to be remediated. Costs vary greatly with the remediation system, ranging over

three orders of magnitude. Processes that rely on in situ water flushing (“pump and

treat”) or vapor stripping tend to be the most inexpensive with total project costs

running at roughly $10 per ton treated. In general these processes are slower and

costs are spread out over multiple years. Unfortunately, they also may only be

containment techniques as many contaminants are only slowly removed from the

soil in this manner. Calculations show that these systems may have to run for

decades (and in some cases centuries) in order to finally remediate the contaminated area. Ex situ treatments are quicker, but are generally considerably more

expensive. Cost ranges for ex situ treatment projects range from $40/ton to over

$800/ton. Costs for small projects with high site management expenses and little

economy of scale are the highest. Petroleum-contaminated soils are often a

favorite remediation target. Costs associated with low-temperature thermal desorption of these soils run from $75 to $125/ton, and high-temperature thermal desorption ranges from $300 to $450/ton. Some of the more exotic treatments (e.g.,

in situ vitrification for radionuclides) can reach several thousand dollars per ton.

All costs are dependent on the total volume to be remediated and can vary with



PHYTOREMEDIATION OF SOILS



61



such factors as: (I)proximity to incineration unit, landfill, and stabilizing agents,

(2) state regulations, and (3) supply and demand. Like any business, the remediation business operates under free market economics, including volume discounts, sales, and “one time only if you act now” sales. Some readers might be

surprised to know that landfill operators actually have sales on landfill space,

while at other times shortages develop. Altogether, however, it is not uncommon

in the United States to spend 1 million dollars/hectare to remediate down to a

depth of half a meter.

The fact that remediation is expensive is the driving force behind the search for

new technology. If remediation were inexpensive, all the contaminated soil in

industrial countries would obviously be remediated. An inexpensive new technology would not only save money on the remediations that do occur, but also

vastly increase both the volume of soil that is remediated and the rate of soil

clean-up in industrialized countries, and possibly even decrease the number of

lawyers involved in “not remediating.”



11. “PHYTOREMEDIATION”



A. CONCEPTSAND DEFINITIONS

The generic term “phytoremediation” consists of the Greek prefix phyto (plant)

attached to the Latin root remedium (to correct or remove an evil).

“Remediation” in this case encompasses all of the discussion in Section 1.

Although there is some variation in the term phytoremediation as currently used,

we define phytoremediation as the “use of green plants and their associated

microbiota, soil amendments, and agronomic techniques to remove, contain, or

render harmless environmental contaminants.” The term phytoremediation, like

the term remediation itself, is rather loosely divided into processes that decontaminate the matrix (extract, degrade, volatilize, etc.) and processes that stabilize

the contaminant in the soil to reduce or prevent further environmental damage

(sequester, solidify, precipitate, etc.). Phytoremediation technical concepts are

borrowed from many years of work by other researchers in the areas of land

reclamation, landfarming of oily wastes, waste water engineering, soil chemistry, plant physiology, and agricultural pesticides.

The use of plant-based (or “phyto”) remediation systems is not new. In fact the

first plant based system was installed over 300 years ago in Germany for the

treatment of municipal sewage (Hartman, 1975). Since that time, overland flow

systems, spray irrigation systems, and constructed wetlands are common for the

secondary treatment of municipal sewage waters. The concepts and principles

are well understood and there are numerous companies who actively design and



62



S. D. CUNNINGHAM ET AL.



install such systems. These include reed-bed filters (Green, 1992), constructed

and natural wetlands (Knight et al., 1992), and systems designed around floating

plants (Buddhavarapu and Hancock, 1991). Although all of these systems are

designed primarily to remove municipal waste water contaminants, there has

been some work at describing the efficiency of removing industrial Contaminants

as well (Wolverton and McDonald-McCaleb, 1986; Winter and Kickuth, 1989).

Significant decreases in concentration of a wide variety of industrial pollutants

have been shown to occur as the water passes through such vegetated beds.

In recent years there has been a concerted effort to extend this concept beyond

municipal water treatment systems. Using some of the same principles and plant

species familiar in water treatment, the concept has been expanded to purify

shallow groundwater in order to prevent off-site migration of the pollutants. There

are now dozens of demonstration projects underway. Collectively, these techniques can be referred to as bio-curtains, bio-filters, or rhizo-filters. Trees, with

deep roots and high transpiration rates, are being field tested to address landfill

leachates, pesticide contamination, and plant nutrient elements such as N and P

(Schnoor et al., 1995). Commercial names of such endeavors include EcolotreeB

cap, Treemediation, and Rhizofiltration. All of these processes have targeted

relatively water-soluble contaminants migrating in surface or shallow subsurface

water flows. In situ, surface soil decontamination by plants without a flowing

water phase is considerably less well documented, both in the lab and in the field.

The use of plants in the remediation of contaminants in the air is also receiving

significant attention. In one sense, we have long recognized our mammalian

dependence on plants for remediation of our air. Two centuries ago, Priestley’s

classic bell jar experiments containing a mouse and green plant proved that plants

“remedy an evil” in our air. Followed shortly by Lavoisier’s observations, these

early researchers concluded that animals give off “some sort of poison, and that

the green plant renioves this poison” (Winchester, 1965). The demonstration of

the phytoremediation of contaminated air (beyond CO,) is more recent. Plant

leaves, with their waxy surfaces, absorb lipophilic volatile compounds including

priority pollutants (Keymeulen et al., 1993) and PAHs (Simonch and Hites,

1994). They or their associated microflora have also been shown to act as biofilters for a number of indoor air contaminants. (Raloff, 1989). Over the last 2

decades a number of patents have been filed on novel designs of plant-based air

biofilters intended for the home and office environment. A patent search also

reveals one ambitious patent for “vegetating the external surfaces of buildings to

remediate urban and industrial air pollution” (Dittmar, 1976). In addition to

patent activity, a number of serious attempts at further quantifying the scientific

basis of the phenomena and the extent of the effect on air contaminants are being

made throughout the world. One particularly promising area of research is the

genetic engineering of plants to improve the organic uptake, as well as uptake

and metabolism of NO, and SO,, products of the internal combustion engine.



PHYTOREMEDIATION OF SOILS



63



The use of stationary plants to remediate streams of air and water that pass

around them is relatively well advanced in comparison to the use of plants in the

remediation of contaminated soil. This is partly due to the lower regulatory

oversight of contaminated soils, but also reflects the fact that there are significantly more challenging dynamics involved in the mass flow, kinetic, and analytical constraints which are present in soil remediation.

Within the research community concerned with the phytoremediation of contaminated soils, research on the use of plants to remediate inorganic contaminants is progressing at a more rapid pace than analogous research with organic

contaminants. There are two inherent reasons for this trend: (1) Soils contaminated with inorganics have far fewer remediation alternatives than soils similarly

contaminated with organics, and ( 2 ) with relatively few exceptions, inorganics

are immutable. Both regulators and scientists alike can follow the fate of inorganics in a soil-plant system with relative confidence. This is usually accomplished by either direct analysis (XRF) or acid digestion and then a spectrometric

analysis (ICP, ICP-MS, or AA).

Organic. contaminants present both an inherent analytical challenge and an

opportunity to the field of bioremediation. The analytical challenge is that following both the fate and the effect of a contaminant in a soil-microbe-plant

system can be exceedingly difficult. Not only does the compound undergo physical, biological, and chemical changes in the process, but subtle changes in the

matrix (plant and soil) alter chemical extraction efficiencies. Plants are themselves

complex organic matrixes that can bind or mask the presence of a contaminant.

ldentifying the contaminant (as well as its metabolites) in this background can be a

challenging task without a radiolabeled compound. This analytical quandry increases the cost of doing research, slows progress, and prevents duplicating much

of the field scale work currently occurring in inorganic research with organics.

Those doing research on the phytoremediation of organics envy the ability of

their inorganic counterparts to search exotic places for metal-accumulating flora.

With instruments as simple as an XRF or a strip of reactive paper they can test

plant tissue for metal accumulation and find a tree with sap that has Ni content in

excess of 25% by weight (Jaffre et a/., 1976). Unfortunately for those involved in

the phytoremediation of organics, simple analytical tools (e.g., a rapid, field

bioassay for benzene degradation) do not currently exist. It would be of great

benefit to organic phytoremediation to be able to search the flora of the world for

variations in degradative capacity with a rapid assay. Many of those involved

would love the opportunity to examine the metabolic activities of different ecosystems.

One such assay may be in the development stages for a few compounds (e.g.,

trinitrotoluene (TNT), hexachloroethane, and triaminotoluene). Working in association with the EPA, researchers have developed a rapid bioassay for plant

enzymatic activity (Wolfe et al., 1995) that can now run by high school students.



64



S. D. CUNNINGHAM ET AL.



This screening tool allows the rapid examination of hundreds of plants on a daily

basis rather than the relatively few plants that can be screened in the laboratory

setting on a weekly basis.

As mentioned previously, phytoremediation either removes the contaminant

from the matrix (decontamination) or sequesters it into the matrix (stabilization).

The latter process seeks to reduce potential environmental harm by reducing the

mobility and availability of the contaminant. Both processes are outlined below

and illustrated in Figs. 1 and 2. Literature searches on each of the words used in

the figures will generally uncover a wealth of knowledge on the biological and

chemical processes involved; however, for the most part, the data are phenomenological, or limited to a single well-studied compound that is not a current target

of phytoremediation for other reasons. Much of what we know about phytoremediation must be extrapolated from this literature base. Little is known

about actual field rates, kinetics, lower obtainable remediation limits, and vegetation and soil management practices to accelerate the processes involved. Although the development of a remediation technique can be accomplished through

trial and error (and many engineering techniques have been) phytoremediation R

& D represents an excellent area in which basic and applied science have begun

to work together to reach a potentially valuable technology.

1. Phytodecontamination



Phytodecontamination is a subset of phytoremediation in which the concentration of the contaminants of concern in the soil is reduced to an acceptable level

through the action of plants, their associated microflora, and agronomic soil

techniques. Figure 1 shows processes involved in this type of remediation. The

inherent processes behind these techniques are further described below.



Phvtodecontamination

Processes



Phvtovowir ation



\lCJ



\



Phvtoextractlon \

Harvest and

Destruction



(Phvtod-



‘ n



K


.



Contaminant



Microbial metaboiisr



Figure 1 Naturally occurring processes involved in phytodecontamination.



PHYTOREMEDLATION OF SOILS



65



Phytoextraction. Absorption of the contaminant into the plant tissue and

subsequent harvesting for destruction.

Phytovolatilization. Plants or their associated microbial activity help to

increase the rate of volatilization of a contaminant from the contaminated soil.

The volatilization occurs from the plant shoots or roots, as well as from the

soil surface.

Phytodegradation. Plants take up the contaminant and metabolize it to an

environmentally benign material.

Rhizo(sphere)degradation. Plant roots, their associated microflora and/or

excreted products destroy the contaminant in the root zone.

2. Phytostabilization



Figure 2 lists processes involved in the stabilization and sequestration of

contaminants in soil. The acceptability of this type of remediation rests on

processes that must be demonstrated as acceptable to the satisfaction of scientists, regulators, and the general public. These processes include the reduction in

biological availability to all potential receptors as well as the off-site migration of

the pollutant. In general, processes listed in Fig. 2 are well documented and

occur naturally, but are hard to quantify. The acceptance of reduced bioavailability in risk assessment and remediation is increasingly well accepted for

some metals (Pb, Cr, etc.). With organic contaminants, however, it is only

beginning to be an important factor in regulatory discussions. No site, contaminated with organics, has been, or is being intentionally remediated by phytostabilization at this time. Sites contaminated with inorganics, however, are undergoing phytostabilization.

There is growing regulatory acceptance of the concept of “relative bioavailability” of organics. The establishment of “environmentally acceptable



Phvtostabilization

Processes

Plants chosen for tolerance to

site conditions, erosion and

leachina control, and Door



Humification

Contaminant



I



contaminant

secluestration

sequestration

due to contaminant interaction cellwalllignins

with increasedorganic matter

soil mineral fraction upon

aging.and weathering



Plant and microbial

bind contaminant

into soil humus



Contaminant



Figure 2 NdtUrdy occurring processes involved in phytostabilization.



66



S. D. CUNNINGHAM ET hL.



treatment endpoints” relies on this concept. The regulatory community also

seems to be moving toward a risk-based framework for regulation of contaminated soils. As reduced bioavailability is a measurable and demonstratable phenomenon, the concept fits well into this emerging framework. In addition, the

emerging focus on “intrinsic or natural bioremediation” invites discussions of the

bioavailability concept. Ongoing research by academic, private industry, and

regulatory authorities suggests that processes which involve demonstrated reduction in bioavailability to human and environmental receptors will be of increasing

value. This research area is increasingly important in the determination of the

fate and effect of agricultural products and environmental toxicants, and will also

be important in pending legislation.

Definitions are provided here for clarification.

Humification. Incorporation of the contaminants into soil humus resulting in

lower bioavailability.

Lignification. Toxic components become irreversibly trapped in plant cell

wall constituents.

Irreversible binding (aging). Compounds become increasingly unavailable

due to binding into soil.



B. APPARENT

ADVANTAGES

OF PHYTOREMEDIATION

The traditional remediation community does not appreciate plants. In a technical and legal sense they fall under the category of “debris” which must be

removed and treated prior to remediation. For communication purposes, with a

somewhat skeptical engineering community, we have found it helpful to redefine

green plants as “solar-driven pumping and filtering systems” that have “measurable loading, degrading, and fouling capacities.” Roots may be described as

“exploratory, liquid-phase extractors that can find, alter, and/or translocate elements and compounds against large chemical gradients” (Cunningham and Berti,

1993). Such definitions serve as the basis for modeling efforts and economic

evaluations, and provide research directions to biologists. Involvement with

traditional remediation engineers is vital to the success of this new technology.

Their educations and backgrounds are beneficial and may help direct research in

areas not normally covered by plant biologists (e.g., supplying kinetic parameters for whole plant processes). They are also the eventual customers for any

technology that phytoremediation produces.

Cost effectiveness is believed to be one of the greatest apparent advantages of

phytoremediation. Agronomic techniques are considerably cheaper than costs

mentioned above. Farming costs are listed by county extension agents and often