The International Information Center for Geotechnical Engineers





Main concept/description of the technique

Sometimes grouped under the more encompassing remediation technology of “bioremediation”, phytoremediation is the process of using vegetation to break down and/or remove environmental contaminants. While extensive planning and design work may be required, the main process is driven by nature and often requires little to no additional maintenance inputs. It is one of the most natural and passive techniques available, and is often less expensive as a result. With rising energy and labor costs, it is being reconsidered as a viable treatment technique, especially in cases of strong public awareness. The “green” nature of the process helps fill a niche role and can favorably influence the public’s perception of a once toxic site (Mudhoo, 2011). It has grown from a $20 million per year to a $350 million per year industry in less than 10 years (Rock et al, 1998). This remediation strategy has been aggressively pursued by industry and academic scientists as a sustainable technology with widespread application in both developed and developing countries.


Theoretical Background

The word “phyto” means plant in Greek. This process involves utilizing plants in remediating environmental contaminants. It generally refers to the use of plants without additional excavation or soil processing. Many different actions occur to absorb or degrade contaminants across a variety of scales. The plant’s root zones must be in contact with the contaminated soil, as this is where the contaminant removal occurs. The root membranes act as a filter in a process termed “rhizofiltration” and eventually absorb the pollution. Two driving forces then determine the contaminant fate depending on the nature of the chemical. “Phytodegradation” occurs when metabolic processes within the plant break down the organic chemicals, while “phytoaccumulation” occurs when typically inorganic compounds are locked into the plant’s structure (Sharma and Reddy, 2004). A somewhat combined process, called “rhizodegradation” occurs in root rhizomes in which mutually beneficial bacteria or fungus breaks down and/or incorporates pollutants on the surface of or within the plant’s roots. The driving energy force can be likened to a miniature pump and treatment system in which evapotranspiration during summer months causes large amounts of soil moisture to be processed. This natural process of “phytoextraction” can draw the soil’s pollution from the ground all the way to the leaves, however the bulk of the material can either be degraded, or accumulates where a local metabolic process is unable to break it down (Sharma and Reddy, 2004). The soil-moisture removal creates a small “cone of depression”, and limits the migration of a plume in a very miniature way. The organic attraction combined with the groundwater immobilization effectively limits its movement in a process called “phytostabilization”. “Phytovolatilization” is an emerging process called out by Mudhoo (2011), which involves the conversion of contaminants to volatile forms and directly releasing them to the atmosphere.

Table 1. Phytoremediation processes, mechanisms, and related pollutants/plant species (Gupta et al, 2000)


Through tracing organic pollutants with radioactive carbon, it has been found that plants like poplar trees transform compounds into trichloroethano, trichloroacetic acid, and dichloroacetic acid metabolites, sorbed by the roots, and “mineralized” to carbon dioxide (Sharma and Reddy, 2004). Around 15% of pesticides are able to be permanently affixed to the plant material, and almost every pollutant released to the atmosphere quickly break down due to hydroxyl radicals.

The efficiency of the phytoremediation process is ultimately determined by the fraction of organic matter in the soil and the pollutant’s hydrophobicity (Sharma and Reddy, 2004). If the pollutant strongly prefers organic material, or if there is an overabundance of organic content, it may be nearly impossible for plants to remove enough of the compound. Extremely hydrophilic pollutants will remain in solution and often pass through plant tissues without significant uptake. Contaminants with octanol-water (KOW) ratios of 1 to 3.5 have the greatest potential for phytoremediation.


Soil with shallow contamination, generally less than ten feet deep, can be cleansed through vegetation. This method works best for a “polishing treatment” with low-levels of existing contamination. Target contaminants generally include hydrophilic or hydrotolerant heavy metals such as cadmium, nickel, zinc, arsenic, selenium and copper. Additional contaminants include radioactive nuclides, polycyclic aromatic hydrocarbons, pesticides, petroleum byproducts, and explosive residues (Sharma and Reddy, 2004).


Plant roots stabilize soil and prevent movement of pollutants via runoff and windblown dust. The technique uses plants and natural resources and is therefore generally less expensive. The remediation is done in place, saving transportation and off-site processing costs. Compared to other systems, it is generally aesthetically pleasing and preferred by the public (Sharma and Reddy, 2004).  Mudhoo (2011) makes the claim that the shallow, but widespread nature of this technique makes it ideal for restoring agricultural soils damaged by dispersed industrial pollution.


Many government agencies have not fully embraced the benefits of this emerging technology. As a result, it is not considered for many projects listed on the National Priority List or Superfund list (Rock et al, 1998). Phytoremediation cannot treat deep contamination; grasses can cleanse up to three feet, shrubs up to ten feet, and deep-rooting trees up to 20 feet. The process is generally slow and can take three to five years to meet the targeted cleanup goals. A site-specific selection of plants must be chosen to process the mixture of chemicals while preventing death of the vegetation. The selection of plants and their combinations are very much in the experimental stage and in need of additional research. The process is highly dependent on local climatology and must be designed with local considerations. Additionally, large scale operations may require heavy agricultural grade equipment, which are commonly located far from our contaminated urban areas (Mudhoo, 2011). Because wildlife and people may consume the plants, precaution must be taken to prevent introduction of contaminants to the food-chain. If the contaminant is sorbed to the soil, it is generally not mobile enough to allow for phytoremediation. Lastly, the waste biomass must be disposed of properly, sometimes at elevated costs (Sharma and Reddy, 2004).

Field Setup, process involved

The general process outlined below, operates in a mainly linear fashion, with the possibility of two “loops” depending on the effect of the treatment.

Figure 1. General phytoremediation process flow (adapted from Sharma and Reddy, 2004)


Plant Selection and Positioning

Because the entire process takes so much time, it is critical to correctly determine the proper type and quantity of plants. The soil type generally determines which plants are effective, but additional laboratory testing is often performed to ensure toxins do not kill the plants. Varieties of plants with huge biomass yields (more than 3 tons/ acre) are generally chosen, in a general attempt to deplete the soil. Plants called “hyperaccumulators” are necessary for inorganic pollutants, as they will take up more than ten times background soil concentrations. When groundwater remediation is of concern, deep-rooting trees such as poplar, cottonwood, and willow trees are generally planted in rows perpendicular to flow, and monitoring wells are placed nearby (Sharma and Reddy, 2004).

Table 2. Hyperaccumulating plant species for various metals (Chaney, et al, 2000)


Table 3. Applications of phytoremediation (Rock, 1998)



Irrigation and Soil Amendment

Often soils need to be amended and routinely irrigated to ensure plant viability (Sharma/Reddy, 2004). Soil is sometimes flooded to encourage the dissolution of pollutants and increased evapotranspiration. If the same plant species is used in succession, the same rate-limiting nutrient may need to be continually replaced. The removal of nutrients and pollutants can drastically affect the pH, which may also need adjustment. Chelating agents, such as ethylenediaminetetraacetic acid (EDTA), (Baylock, 1999) can aide in keeping radionuclides and heavy metals in solution, increasing the ability to be taken up by vegetation.


Sampling of soil and/or groundwater should be routinely practiced. The changing contaminant concentration ensures that the current process is effective, and can resultingly be utilized to modified the current treatment plan. If the process is  too slow, a different type of plant can be substituted, or soil amendments can be adjusted to increase performance (Sharma and Reddy, 2004). This is the first “feedback loop” that can necessitate the modification of a previous process step.


As the last potential step, harvesting can also be the most controversial. It involves the removal of the plant, sometimes including the roots. Through soil sampling, a mass balance can be performed to approximate the efficiency of contaminant removal (Sharma/Reddy, 2004). Depending on the nature of the contaminant and concentrations within the plant tissue, the vegetation may be composted, or require more expensive disposal. One emerging option includes incineration, in which the biomass provides electricity and heat, while pollutants are either oxidized or captured in flue-stack scrubbers. Some “contaminant-laced” biomass actually has economic value. A crop used to remediate selenium can be sold to livestock farmers as “bio-enhanced” and defray their costs of purchasing selenium salts to add to their normal livestock feed (Chaney, et al, 2000). Ash from the nickel-accumulator “Alyssum murale” can be economically processed into ore if the content is above 20% by weight (Chaney, et al, 2000). Generally hyperaccumulators retain inorganic compounds in the following “compartments” (in decreasing order): root, leaf, stem, fruit. Nightshade plants, such as cucumbers and tomatoes, however,  accumulate compounds in their leaves and stems more readily than roots, which make them better for above-ground harvest (Gupta et al, 2000). Following annual or semi-annual harvesting, testing determines whether additional treatment through another growing season is warranted.

Other important considerations

Regulatory Concerns

As with other remediation techniques, cleanup time and the extent of the cleanup is often under regulatory considerations. The treatment objectives are usually selected at the onset of the process, and with the extended amount of time, care must be taken to incorporate any changing regulations. Many studies have shown that atmospheric releases are not of great concern, as they are mostly metabolized and in almost immeasurable concentrations (Sharma and Reddy, 2004). The biggest emphasis has been on control of the potentially toxic plant material, as it may have significant amounts of organic compounds and need additional treatment.


Each case is unique, but soils with low levels of contaminants can generally be cleansed for $10-$35 per ton (Sharma and Reddy, 2004). Removal, off site treatment, and the return of processed residue can cost $200-$600 per ton (Gupta et al, 2000). This is dependent on the amount of soil amendment and irrigation, number of growing cycles, and difficulty of disposal. Because the method uses no industrial fuel or products, it is considered to be passive, saving operating and maintenance costs. Localized waste treatment options determine where the plant biomass is processed. Zodrow (1999), makes a first order comparison of remediating 500 ppm lead contaminated soil, and states that excavation and disposal costs $300,000 per acre, while phytoextraction costs approximately $110,000 per acre.

Table 4. Cost comparison with other methods (Rock, 1998)


Due Care

Care should be taken to prevent wind or water-based soil erosion, as soil contaminants may migrate off site and cause additional problems. In addition, pests and rodents must be monitored and deterred, as they can limit plant growth and cause unnecessary exposure of pollutants.

Predictive Modeling

Four main modeling techniques have been developed based on first-order kinetics and different schemes of partitioning. They are listed below with their associated assumptions and limitations.

Table 5. Phytoremediation modeling techniques (adapted from Sharma and Reddy 2004)

Combination of Phytoremediation With Other Methods

Phytoremediation relies on a complex interaction between large plants and their associated microbes. Because bioremediation and phytoremediation both rely on natural, living processes, they are often used in conjunction for a synergistic effect (Sharma and Reddy, 2004). Sometimes soil is inoculated with microbes before plants begin growing to either prep the soil or encourage increased growth within the root zone. The two methods are most often combined when rhizodegradation is the driving process in treatment. As the table 6 below shows, enzymes play a vital role in degrading organic compounds. These enzymes are also often found in microbes that mutually benefit the plant, and through encouraging this relationship, treatment efficiencies can be increased (Gupta et al, 2000).

Table 6. Plant enzymes with degrading compounds (Gupta et al, 2000)


Case histories

Heavy Metal and Radinuclide Phytoextraction in Chernobyl (Dushenkov et al., 1999)

In 1986 the nuclear fission plant commonly known as Chernobyl suffered a catastrophic meltdown. Cesium-137 was released as far away as Sweden, but the majority of the fallout occurred immediately north of the reactor site. Most of this pollutant quickly percolated into the sandy topsoil following rainstorms. Twenty different surfactants were used to increase the bioavailability of the cesium. While Indian mustard, corn, peas, artichoke, and sunflower were all used, it was found that only artichokes and sunflowers yielded substantial results. Even so, after only three weeks cesium levels ceased to decrease, and harvesting was prescribed. A modest 0.3% decrease in radioactivity was observed. Dushenkov et al. relied on results from previous studies and treated the lead-laden soil with chelating agents to mobilize it further (Dushenkov et al. 1999). These chelating agents caused a 20-fold increase in the lead uptake levels. Ultimately, incineration was used to reduce the volume of plant waste to less than 10% of its original.

Lead Phytoremediation at a New Jersey Brownfield Site (Baylock et al., 1999)

Phytoremediation was used over a 4500 sq ft area outside an abandoned lead-acid battery factory in Trenton, New Jersey. This contaminated site is in close proximity to schools, churches, and residences, and therefore needed to be visually appealing. A portable x-ray fluorescence (XRF) meter was used to quickly inventory the site’s lead levels, while additional samples were sent off-site for more detailed analysis. The metal-accumulating crop “Brassica juncea L. Czern”, also known as Indian mustard was employed to extract lead from the soil. The soil was fertilized and planted with 3.5 in. diameter pots of the mustard plant. EDTA was applied at varying rates to assess its affect on lead solubility and resulting plant uptake. Soil moisture was tracked with four sensors, and irrigation was provided with overhead sprinklers. Each growing cycle took 6 weeks, plants were harvested by cutting them at ground level, the site was rototilled, and three growth cycles could occur in one growing season. Although multiple growing season treatments were used for an overall 13% reduction in lead levels, 72% of the planted area was treated to below the EPA direct contact criteria of 400 ppm in one growing season.

Figure 2. Results of New Jersey Phytoremediation Project, Baylock et al. (1999)



OneSITE Wastewater Treatment Plant in Woodburn, Oregon (Zadrow , 1999)

Ten thousand poplar trees “populous trichocarpa deltoides” were planted over a ten acre abandoned sludge lagoon in an attempt to stabilize the waste and open over 400 acres for a treatment buffer. This buffer is currently working as an alternative to releasing over 5 million gallons of untreated wastewater to a small seasonal stream. The ten acre pilot plot was aimed at preventing the construction of a $2.5 million storage lagoon. EPA awarded $250,000 for the initial planting of poplar trees, which grow twice as fast as pine species and can be sold to the lumber industry every 10 years.  After the $2.5 million dollar tree planting project is implemented over all 300 acres, the municipality estimates the income of $800,000 every ten years upon harvesting of the trees.



Sharma, H.D., Reddy K.R. (2004). “Geoenvironmental Engineering.” Jon Wiley & Sons, Hoboken, New Jersey, 478-485 

Doty, S.L. (2008). “Enhancing phytoremediation through the use of transgenics and endophytes.” New Phytologist (2008) 179: 318–333

Blaylock, M.J., Elless, M.P., Huang, J.W., Dushenkov, S.M. (1999). “Phytoremediation of Lead-Contaminated Soil at a New Jersey Brownfield Site.” Remediation, summer 1999; 93-101

Chaney, R.L., Broadhurst, L., Centofanti, T. (2000) “Phytoremediation of Soil Trace

Elements.” Bioavailability, Risk Assessment and Remediation; 311-352

Rock, S.A., Sayre, P.G. (1998) “Phytoremediation of Hazardous Wastes: Potential Regulatory Acceptability.” Remediation, autumn 1998; 5-17

Zadrow, J.J. (1999). “Recent Applications of Phytoremediation Technologies.” Remediation, spring 1999; 29-36

Mudhoo, A. (2011). “Phytoremediation of Cadmium: A Green Approach.”

Gupta et al. (2000) “Phytoremediation: An Efficient Approach for Bioremediation of Organic and Metallic Ions Pollutants.” Bioremediation and Sustainability; 213-240

Dushenkov, S., Mikheev, A., Prokhnevsky A., Ruchko, M., and Sorochinsky, B., Phytoremediation of radiocesium-contaminated soil in the vicinity of Chernobyl, Ukraine, Environ. Sci. Technol., Vol. 33, pp. 469-475, 1999.

Add comment

NOTE: The symbol < is not allowed in comments. If you use it, the comment will not be published correctly.

Security code
*Please insert the above-shown characters in the field below.