Methodology Criteria Definitions Chemicals that Should Not be Profiled Home Start a New Profile Results Terms of Use Security and Anonymity Feedback, questions, problems, and comments

PBT Profiler Methodology
Persistence      Bioaccumulation      Toxicity
PBT Profiler Results for Known PBTs and POPs       PBT Profiler Results for Chemicals in an Evaluated Data Set

Introduction

The PBT profiler uses a well-defined set of procedures to predict the persistence, bioaccumulation, and toxicity of chemical compounds when experimental data are not available. The only user-required inputs for the PBT profiler are a unique identifier (e.g., a CAS Registry Number, product ID, or acronym) and a chemical structure. Chemical structures are entered into the PBT profiler using a SMILES notation [Weininger, D. SMILES, A Chemical and Information System. 1. Introduction to Methodology and Encoding Rules. Journal of Chemical Information and Computer Sciences 28: 31-6 (1988)]. An automatic look-up function based on the CAS Registry number simplifies this process by automatically retrieving a chemical’s SMILES notation using a pre-existing database containing over 100,000 records. The chemical structure is then passed to nine separate physical/chemical property estimation modules, and the results are converted electronically to a persistence, bioaccumulation, and toxicity value. The methodology used by the PBT profiler and its estimation modules is discussed in this section.

The persistence, bioaccumulation, and fish chronic toxicity values estimated by the PBT profiler are automatically compared to criteria published by the EPA. Those values that meet or exceed the criteria are flagged for the user on the PBT Profiler results page. When estimations meet or exceed criteria, that material should be evaluated as a potential PBT Chemical. A discussion of the persistence, bioaccumulation, and toxicity criteria used by the PBT Profiler is provided.

Persistence

Air

The PBT profiler calculates an atmospheric half-life by determining the importance of a chemical's reaction with two of the most prevalent atmospheric oxidants, hydroxyl radicals and ozone. The half-life is calculated directly from gas-phase hydroxyl radical and ozone reaction rate constants. These rate constants are obtained from a database of measured values or, if no experimental values are available, they are estimated using the method of Atkinson [Meylan, W.M. and Howard. P.H. Computer Estimation of the Atmospheric Gas-Phase Reaction Rate of Organic Compounds with Hydroxyl Radicals and Ozone. Chemosphere 26: 2293-9 (1993)]. The half-life is calculated from the rate constant and an average atmospheric concentration of these oxidants based on a 24 hour day [Prinn, R., Cunnold, P., Simmonds, R., Alyea, R., Boldi, A., Crawford, P., Fraser, D., Gutzler, D., Hartley, R., Rosen, R., and Rasmussen R. Global Average Concentration and Trend for Hydroxyl Radicals Deduced From ALE/GAGE Trichloroethane (Methyl Chloroform) Data for 1978-1990. Journal of Geophysical Research 97: 2445-61 (1992); Atkinson, R and Carter, W.P.L. Kinetics and Mechanisms of the Gas-Phase Reactions of Ozone with Organic Compounds under Atmospheric Conditions. Chemical Reviews 84: 437-70 (1984)].

The atmospheric half-life for each process is calculated as follows:

Hydroxyl radicals t1/2 = 0.693/(rate constant cm3/molecule-sec x 5x105 molecules/cm3 * 86400 sec/day)

Ozone t1/2 = 0.693/(rate constant cm3/molecule-sec x 7x1011 molecules/cm3 * 86400 sec/day)

and the overall half-life is obtained as:

1/t1/2overall = 1/t1/2 Hydroxyl radicals + 1/t1/2 Ozone

Water, Soil, and Sediment To Top
 

The half-life for degradation of a chemical in water, soil, and sediment is determined using the ultimate biodegradation expert survey module of the BIOWIN estimation program [Boethling, R.S., Howard, P.H., Meylan, W.M., Stiteler, W., Beauman, J., and Tirado, N., Group Contribution Method for Predicting Probability and Rate of Aerobic Biodegradation. Environmental Science and Technology 28: 459-65 (1994)]. This estimation program provides an indication of a chemical’s environmental biodegradation rate in relative terms such as hours, hours to days, days, days to weeks, and so on; the terms represent the approximate amount of time needed for degradation to be "complete". This output cannot be directly compared to established half-life criteria for purposes of identifying chemicals with PBT characteristics, nor can it be used directly by the level III multimedia mass balance model. The mean value within the estimated time range returned by the ultimate biodegradation survey model is converted to a half-life using a set of conversion factors These conversion factors consider that 6 half-lives constitute "complete" degradation of a chemical substance (assuming first-order kinetics). The resulting conversion factors for water are provided below.

BIOWIN Output Converted Half-Life (days)
Hours 0.17
Hours to Days 1.25
Days 2.33
Days to Weeks 8.67
Weeks 15
Weeks to Months 37.5
Months 60
Recalcitrant 180

Note that the maximum value returned by this model is 180 days even though the half-life of recalcitrant molecules in the environment is likely to be substantially longer. During the development of the PBT Profiler, it was determined that use of a longer half-life for recalcitrant molecules did not have a significant affect on the results.  

It is known that ultimate biodegradation is generally slower under anaerobic conditions than under aerobic conditions. The BIOWIN ultimate survey model that is used in this Profiler assumes aerobic conditions, but deeper layers of aquatic sediments are usually anaerobic. To account for the slower rate of ultimate biodegradation in sediment, the PBT Profiler uses a conversion factor developed in EPA’s P2 framework. The PBT Profiler assumes that sediments are anaerobic and that the rate of ultimate biodegradation in sediment is on average one-ninth (1/9) of that in the water column (which is assumed to be aerobic).

Similarly, the PBT Profiler makes an adjustment for the biodegradation rate in soil.  It is generally believed that the biodegradation rate for a chemical in soil is, on average, one-half (1/2) that in water.  The PBT Profiler, therefore, assigns the half-life in soil to be twice that estimated for water.

Percentage in Each Medium To Top
   

The PBT profiler uses the Level III multimedia mass balance model (also called a fugacity model) of Mackay [Mackay D., Paterson S., and Shiu W.Y. Generic Models for Evaluating the Regional Fate of Chemical. Chemosphere 24: 695-718 (1992)] to determine the percentage of a chemical in each medium. The Level III fugacity model is a multimedia model that uses a chemical’s physical/chemical properties and degradation half-lives for air, water, soil, and sediment. The Level III fugacity model was chosen for the PBT profiler because it was available in an electronic format, could be readily modified for use in a web-based application, and has been evaluated in several studies [for example, see Kuhne R. et al. Environmental Toxicology and Chemistry 16: 2067-9 (1997); Matoba Y et al. Journal of the Air and Waste Management Association 48: 969-78; Suzuki N et al. Chemosphere 37: 2239-59 (1998)]. 

The Level III fugacity model is a non-equilibrium, steady-state multimedia fate model that is designed to provide information on environmental partitioning and inter-media transport at the screening level. This additional information is consistent with the basic purpose of the Profiler, which is to provide the user with valuable information to help establish whether a more rigorous investigation of PBT characteristics is required.

The multimedia mass balance model requires a series of physical/chemical properties and environmental half-lives as input. Environmental half-lives are determined for air, water, soil, and sediment as discussed above. Physical/chemical properties are provided by EPIWINTM suite of structure-based estimation programs. Physical properties are used directly by the fugacity model to determine the transport between environmental compartments. These properties, and citations to the methodology used for each, are:

  • Henry’s Law constant [Meylan, W.M. and Howard, P.H. Bond Contribution Method for Estimating Henry's Law Constants. Environmental Toxicology and Chemistry 10: 1283-93 (1991)];
  • Vapor pressure [Lyman, W.J., Reehl, W.F.and Rosenblatt, D.H. Handbook of Chemical Property Estimation Methods. Washington, DC: American Chemical Society (1990)];
  • Melting point [Reid, R.C., Prausnitz, J.M., and Poling, B.E. The Properties of Gases and Liquids. Fourth edition. NY: McGraw-Hill, Inc., Chapter 2 (1987); Joback, K.G. 1982. A Unified Approach to Physical Property Estimation Using Multivariate Statistical Techniques. Stevens Institute of Technology, submitted to the Dept. of Chem. Eng. for M.S. Degree at the Massachusetts Institute of Technology in June 1984];
  • Octanol/water partition coefficient [Meylan, W.M. and Howard, P.H. Atom/Fragment Contribution Method for Estimating Octanol-Water Partition Coefficients. Journal of Pharmaceutical Science 84: 83-92 (1996)]; and
  • Molecular weight (determined from the SMILES notation).
In addition, the PBT Profiler estimates a chemical's water solubility [Meylan, W.M. and Howard, P.H. Improved Method for Estimating Water Solubility from Octanol/Water Partition Coefficient. Environmental Toxicology and Chemistry 15: 100-6 (1996)] as part of the Toxicity module.

The PBT profiler will automatically use experimental values for Henry’s Law constant and octanol/water partition coefficient through an automatic database look-up function in preference to estimated values. All physical/chemical properties are estimated at 25 degrees C. It is important to note that the temperature in the environment fluctuates and may vary considerably from this value. However, methodologies that cover the full range of temperatures that may be encountered in the environment are not sufficiently robust for use in an automated tool nor are they required for a screening level evaluation.

If the Henry's Law constant can not be estimated from chemical structure, the PBT Profiler then determines if a water solubility and vapor pressure are available.  If they are, the Henry's Law constant is determined by dividing the vapor pressure by the water solubility (after converting each value to the proper units).

The PBT profiler uses the default settings of the Level III fugacity model [Mackay D., Paterson S., and Shiu W.Y. Generic Models for Evaluating the Regional Fate of Chemical. Chemosphere 24: 695-718 (1992)] for its calculations, and displays the results of these calculations as the percentage of the chemical expected at steady state in the different environmental media. Default emission rates are equal amounts (1,000 kg/hr) to air, soil, and water (direct discharges to sediment are unlikely). The model treats a generic environment of 100,000 square km with 10% water; 90% soil surface; water depth 20 m; soil depth 20 cm; sediment depth 5 cm; atmospheric height 1000 m.  The Level III fugacity model provides a representative environment about the size of the state of Ohio. It is important to note, however, that the percent in each medium will change as a function of the size of the compartments chosen. Moreover, the results of this model are calculated at steady state; a condition that may not occur in the environment (on a global scale). Nevertheless, this model provides useful information to compare chemicals in a screening-level assessment.

The PBT Profiler also provides the user with different release scenarios for a chemical to aid in identifying P2 opportunities that may be more consistent with its expected life cycle.  On the P2 considerations page, the emission rate is set at either 0 or 1,000 kg/hr for soil, water, and air, and the resulting distribution and half-life in each environmental compartment is provided.

The Level III multimedia mass balance model, by design, does not consider a chemical's potential to migrate to groundwater. Therefore, the PBT Profiler does not explicitly consider the fate of a chemical substance in groundwater. Because a complete P2 assessment should consider a chemical's potential to migrate to groundwater, the PBT Profiler contains a simple set of algorithms to highlight chemicals that may travel through soil into an underground aquifer based on its octanol/water partition coefficient (Kow).   The PBT Profiler only highlights chemicals that are expected to have the potential to persist in groundwater. If the log Kow of an individual chemical is < 4 and it is found to be persistent in soil or sediment (half-life > 2 months), then the PBT Profiler flags this chemical as one that has the potential to migrate to and be persistent in groundwater.  This information is provided to the user on the P2 considerations page.

Persistence Summary To Top
 

The PBT Profiler provides a persistence summary (also called the persistence ranking) as the initial tier of its integrated output. The persistence summary is based on the results of the physical/chemical properties estimates (as provided in the second tier of the PBT Profiler output - the detailed results). The persistence summary is derived using the following methodology.

PBT strategies typically consider the persistence of a chemical in water, soil, and sediment because these are the media associated with bioaccumulation. The PBT Profiler first determines the amount of the chemical expected to be found in water, soil, and sediment (expressed as a percentage of the total amount in the environment) using a level III multimedia mass balance model. It then determines which of these three compartments the chemical is most likely to partition to (the one with the highest percentage). Using this predominant compartment, the half-life in that compartment is then compared to the EPA criteria to determine the persistence summary. If the half-life in the predominant compartment exceeds the EPA criteria, the chemical is designated as persistent or very persistent in the summary output.

The advantage of this methodology is that it accounts for the two most important components of a chemical's fate in the environment; its removal and its partitioning. If only one aspect of a chemical's estimated environmental fate were considered, the potential for misclassification increases. For example, a chemical that partitions only to water and soil may have an estimated half-life in water of 15 days, an estimated half-life in soil of 30 days, and an estimated half-life in sediment of 135 days. This chemical would exceed the EPA persistence criteria for sediment based upon its half-life; however, the results of the multimedia model may indicate that it is not expected to be present in sediment. Therefore, its half-life in either soil or water is more representative of its persistence in the environment.

The PBT Profiler methodology was developed with the aid of a database of experimental biodegradation rates for 136 chemicals. Comparison of the PBT Profiler summary results to an evaluated data set is provided below.

Overall Persistence To Top
 

It is important to distinguish between persistence in a single medium and overall environmental persistence. Persistence in an individual medium is controlled by transport of the substance to other media, as well as transformation to other molecular species (together, these are referred to as the residence time). Persistence in the environment as a whole is a distinct concept. It is based on the observations that the environment behaves as a set of interconnected media, and that a chemical substance released to the environment will become distributed in these media in accordance with its intrinsic (physical/chemical) properties and reactivity. Multimedia mass balance models offer the most convenient means to estimate overall environmental persistence from information on sources and loadings, chemical properties and transformation processes, and inter-media partitioning.

The PBT Profiler uses a model similar to the TaPL3 model, available from the Trent University website, to obtain an estimate of overall persistence in which all net loss from the defined model environment is due to reaction (e.g., biodegradation) only; i.e., there are no advective losses from the model environment. This is appropriate in screening-level review for PBT potential since advective losses merely relocate a chemical rather than alter it chemically. Overall persistence can be thought of as the integration of medium-specific half-lives with inter-media transport and partitioning to give a weighted average of the persistence in the individual media. The results are expressed as the environmental residence time tauR, which is equivalent to the total amount of the chemical in the defined environment divided by the total loss rate due to reaction (or the input rate, since these are equal in a steady-state, level III model). The overall persistence provides a meaningful way of expressing relative persistence when, for example, a chemical is discharged to one medium but rapidly partitions to another, and the half-life of the chemical is very different in each compartment. Users should consult the recent work by Webster et al. for further details [Webster E., Mackay, D., and Wania, F. Evaluating Environmental Persistence. Environmental Toxicology and Chemistry 17: 2148-2158 (1998)].

Bioaccumulation To Top
 

The PBT profiler determines a chemical’s potential to bioaccumulate directly from an estimated bioconcentration factor (BCF). The bioconcentration factor is estimated using EPISuites’s BCFWIN estimation program [Meylan,W.M., Howard, P.H., Boethling, R.S., Aronson, D., Printup, H., and Gouchie, S. Improved Method for Estimating Bioconcentration Factor (BCF) from Octanol-Water Partition Coefficient. Environmental Toxicology and Chemistry 18: 664-72 (1999)]. The estimated bioconcentration factors are compared to those contained in the EPA Criteria.  The BCFWIN program yields a screening-level prediction of BCF based on a chemical’s octanol/water partition coefficient and one or more chemical structure-based correction factors, if applicable. The model does not explicitly address a variety of factors that may influence bioaccumulation under field conditions, such as possible metabolism of the chemical in exposed organisms, which could lead to actual bioaccumulation being lower than predicted. Therefore, the user needs to exercise due caution when interpreting BCFWIN results.

Toxicity To Top
 

The PBT Profiler considers only the fish chronic toxicity and estimates it using the ECOSAR (Ecological Structure Activity Relationships) program. ECOSAR predicts the toxicity of chemicals to aquatic organisms such as fish, invertebrates, and algae by using Structure Activity Relationships (SARs). ECOSAR uses SARs to predict the aquatic toxicity of chemicals based on their structural similarity to chemicals for which aquatic toxicity data are available. SARs express the correlations between a compound's physical/chemical properties and its aquatic toxicity. SARs measured for one compound can be used to predict the toxicity of similar compounds belonging to the same chemical class. The SARs contained within the ECOSAR are based on test data and many of the SAR predictions have been evaluated. More information on ECOSAR, as well as the program itself, is available from the EPA.

For the PBT Profiler, the fish chronic (ChV) estimates from ECOSAR are used to predict toxicity because there is the potential for long-term exposure to persistent chemicals. If the ChV cannot be estimated using the QSAR equations available in ECOSAR, the PBT Profiler will return "Not Estimated." The PBT Profiler identifies chemicals that exceed the log octanol/water partition coefficient cutoffs for the QSARs used by ECOSAR. If the cutoffs are exceeded for a specific chemical, the PBT Profiler will return a ChV value of "Not Estimated" and it will not run the ECOSAR estimation.  Because of the algorithms used by ECOSAR, this program estimates a water solubility separately using the octanol/water partition coefficient.

PBT chemicals are those that persist in the environment. They generally occur in low concentrations, and can be transported throughout the biosphere. They can be bioconcentrated in aquatic organisms and transported up food chains to humans, birds, and wild mammals. Exposure to PBT chemicals is generally through the diet and is the result of chronic exposure. Chronic exposures lead to chronic toxicity, and not acute toxicity. That is, chronically toxic chemicals affect processes other than survival. Therefore, it would be possible for organisms to survive the acute effects of such chemicals (i.e., not die) yet still undergo adverse effects from long-term exposure (e.g., chronic exposure may adversely effect growth or reproduction). In order to best estimate the effects of PBT chemicals on the environment, the fish chronic toxicity is used in the PBT Profiler. The ECOSAR program also estimates a variety other aquatic toxicity endpoints depending on the structure of the chemical. These include the acute toxicity of a chemical to fish (both fresh and saltwater), water fleas (daphnids), and green algae. For some chemical classes, endpoints for other organisms may be estimated (e.g., earthworms). A complete P2 assessment for a chemical substance should consider these other toxicity estimates, when available. To aid in this exercise, the PBT Profiler provides the complete ECOSAR output for each substance profiled. The link to the ECOSAR output is available from the P2 Considerations page.

The PBT Profiler compares the ChV (in mg/l) of each chemical to its water solubility. If the water solubility is less than the ChV, then there are no effects at saturation. For chemicals that the PBT Profiler identifies as potentially persistent and bioaccumulative, the no effects at saturation flag is not used in the overall summary as these chemicals may accumulate to higher levels over time. For all other chemicals that may display no effects at saturation, the PBT Profiler returns a green Toxicity summary.

For the chemicals profiled, the chronic fish ChV for neutral organic compounds is estimated. For some chemical classes, a class-specific ChV may also estimated. Below are some examples of chemical classes that have QSARs that estimate fish chronic ChVs:

  • Acrylates;
  • Aldehydes;
  • Anilines;
  • Anilines, dinitro;
  • Benzenes, dinitro;
  • Esters, mono;
  • Esters, di;
  • Phenols; and
  • Phenols, dinitro;
For a complete list of classes that have fish chronic ChV models, as well as a comprehensive description of how these QSARs were developed, please see the ECOSAR technical manual. The full ECOSAR output for each chemical is available from the P2 Considerations page.

For chemicals where more than one class-specific ChV is estimated, the PBT Profiler uses the lowest value (most toxic) in its Toxicity ranking.

PBT Profiler Results for Known PBTs and POPs To Top
 

The summary results of the PBT profiler were compared to organic chemicals generally recognized as being PBTs; the 64 chemicals in EPA's final rule on Persistent Bioaccumulative Toxic (PBT) Chemicals and the 12 United Nations Environment Programme (UNEP) Persistent Organic Pollutants (POPs).

For the 64 chemicals in EPA's TRI Rule, the PBT Profiler flagged 49 as PBTs and 13 as persistent and bioaccumulative with the toxicity "not estimated." As indicated on the "Interpreting the PBT Profiler Results" Page, persistent and bioaccumulate chemicals may accumulate in the environment to relatively high levels. Therefore, persistent and bioaccumulative chemicals without an estimated toxicity should be reviewed carefully (and using the same techniques as one estimated to be a PBT) when identifying P2 opportunities.

Similar results were obtained for the 12 POPs (using a representative structure for the PCBs, polychlorinated dioxins, and polychlorinated furans). All but 3 of the POPs are listed in the final TRI rule.

The results described above are summarized in the following tables.

Persistent, bioaccumulative, and toxic organic chemicals in EPA's final PBT Rule for TRI.
 
CASName PBT Profiler Results
1115-32-2Dicofol P B T
2118-74-1Hexachlorobenzene P B T
31582-09-8Trifluralin P B T
41746-01-62,3,7,8-Tetrachlorodibenzo-p-dioxin P B T
5189-55-9Benzo(r,s,t)pentaphene P B T
6189-64-0Dibenzo(a,h)pyrene P B T
7191-24-2Benzo(g,h,i)perylene P B T
8191-30-0Dibenzo(a,i)pyrene P B T
9192-65-4Dibenzo(a,e)pyrene P B T
10193-39-5Indeno(1,2,3-cd)pyrene P B T
1119408-74-31,2,3,7,8,9-Hexachlorodibenzo-p-dioxin P B T
12194-59-27H-Dibenzo(c,g)carbazole P B T
13205-82-3Benzo(j)fluoranthene P B T
14205-99-2Benzo(b)fluoranthene P B T
15206-44-0Benzo(j,k)fluorene (fluoranthene) P B T
16207-08-9Benzo(k)fluoranthene P B T
17218-01-9Benzo(a)phenanthrene P B T
18224-42-0Dibenz(a,j)acridine P B T
19226-36-8Dibenz(a,h)acridine P B T
2029082-74-4Octachlorostyrene P B T
21309-00-2Aldrin P B T
2231508-00-62,3',4,4',5-Pentachlorobiphenyl P B T
2332598-13-33,3',4,4'-Tetrachlorobiphenyl P B T
2432598-14-42,3,3',4,4'-Pentachlorobiphenyl P B T
253268-87-91,2,3,4,6,7,8,9-Octachlorodibenzo-p-dioxin P B T
2632774-16-63,3',4,4',5,5'-Hexachlorobiphenyl P B T
2735822-46-91,2,3,4,6,7,8-Heptachlorodibenzo-p-dioxin P B T
283697-24-35-Methylchrysene P B T
2938380-08-42,3,3',4,4',5-Hexachlorobiphenyl P B T
3039001-02-01,2,3,4,6,7,8,9-Octachlorodibenzofuran P B T
3139227-28-61,2,3,4,7,8-Hexachlorodibenzo-p-dioxin P B T
3239635-31-92,3,3',4,4',5,5'-Heptachlorobiphenyl P B T
3340321-76-41,2,3,7,8-Pentachlorodibenzo-p-dioxin P B T
3440487-42-1Pendimethalin P B T
35465-73-6Isodrin P B T
3650-32-8Benzo(a)pyrene P B T
3751207-31-92,3,7,8-Tetrachlorodibenzofuran P B T
3852663-72-62,3',4,4',5,5'-Hexachlorobiphenyl P B T
3953-70-3Dibenzo(a,h)anthracene P B T
405385-75-1Dibenzo(a,e)fluoranthene P B T
415522-43-01-Nitropyrene P B T
4255673-89-71,2,3,4,7,8,9-Heptachlorodibenzofuran P B T
4356-49-53-Methylcholanthrene P B T
4456-55-3Benzo(a)anthracene P B T
4557117-31-42,3,4,7,8-Pentachlorodibenzofuran P B T
4657117-41-61,2,3,7,8-Pentachlorodibenzofuran P B T
4757117-44-91,2,3,6,7,8-Hexachlorodibenzofuran P B T
4857465-28-83,3',4,4',5-Pentachlorobiphenyl P B T
4957653-85-71,2,3,6,7,8-Hexachlorodibenzo-p-dioxin P B T
5057-74-9Chlordane P B T
5157-97-67,12-Dimethylbenz(a)anthracene P B T
5260851-34-52,3,4,6,7,8-Hexachlorodibenzofuran P B T
53608-93-5Pentachlorobenzene P B T
5465510-44-32',3,4,4',5-Pentachlorobiphenyl P B T
5567562-39-41,2,3,4,6,7,8-Heptachlorodibenzofuran P B T
5669782-90-72,3,3',4,4',5'-Hexachlorobiphenyl P B T
5770648-26-91,2,3,4,7,8-Hexachlorodibenzofuran P B T
5872-43-5Methoxychlor P B T
5972918-21-91,2,3,7,8,9-Hexachlorodibenzofuran P B T
6074472-37-02,3,4,4',5-Pentachlorobiphenyl P B T
6176-44-8Heptachlor P B T
6279-94-7Tetrabromobisphenol A P B T
638001-35-2Toxaphene P B T
641336-36-3Polychlorinated Biphenyls (PCBs) P B T

Persistent Organic Pollutants (POPs)

#CASName PBT Profiler Results
1309-00-2Aldrin P B T
257-74-9Chlordane P B T
350-29-3DDT P B T
460-57-1Dieldrin P B T
572-20-8Endrin P B T
676-44-8Heptachlor P B T
7118-74-1Hexachlorobenzene P B T
88001-35-2Toxaphene P B T
92385-85-5Mirex P B T
101336-36-3Polychlorinated biphenyls (PCBs) P B T
1157653-85-7Polychlorinated dibenzo-p-dioxins P B T
1260851-34-5Polychlorinated dibenzofurans P B T

PBT Profiler Results for Chemicals in an Evaluated Data Set To Top
   

Introduction

The persistence summary from the PBT Profiler was compared to a published data set containing evaluated environmental persistence values [Mackay, D; Shiu, W.Y.; Ma, K. Physical-Chemical Properties & Environmental Fate on CD-ROM. CRC Press . ISBN/ISSN: 0849321921 (2000)].  In this compilation, published data were used to place a chemical's half-life into one of nine evaluated persistence classes as shown in the following table (for water, soil, and sediment).

Evaluated Class Mean Half-Life (d) Half-Life Range (d)
1 0.21 <0.42
2 0.71 0.42-1.2
3 2.3 1.2-4.2
4 7.0 4.2-12.5
5 23 12.5-42
6 70.8 42-125
7 229 125-417
8 708 417-1250
9 2292 >1250

Classes 1-5 correspond to the EPA “not persistent” criteria (Green, half-life < 2 months) and classes 8 and 9 correspond to the “very persistent” (Red, > 6 months) criteria. The half-life ranges for classes 6 & 7 overlap more than one EPA criteria and, therefore, chemicals in these classes can not be compared directly to the "persistent" category (Yellow, >2 months and < 6 months).

Comparison of the evaluated data set to the PBT Profiler results

Initial results were obtained by running the 293 chemicals in this data set through the PBT Profiler. The predominant compartment returned by the PBT Profiler was identified and the evaluated half-life class for that compartment was determined. The resulting half-life was then compared to the persistence summary estimated by the PBT Profiler. For example, the PBT Profiler predicts that acetone will be found predominately in water and that it is not persistent (half-life < 2 months). The evaluated persistence class for acetone in water is 4, which corresponds to a mean half-life of 7 days (range ~4-12 days). Therefore, the results for the evaluated data and the PBT Profiler are in agreement for acetone and this chemical is represented in the area shaded green in the following pie chart. The results for all 293 chemicals in the data set are summarized in the following pie charts (using the PBT Profiler paradigm) by evaluated class (or combined classes, as appropriate).

Of the 164 chemicals in the not persistent category, 133 (81%) were shown to have a half-life of < 2 months for both the evaluated data set and the PBT Profiler summary (Figure 1). However, if one reviews the likely fate of the chemicals that are categorized differently, the literature indicates that biodegradation is not their primary removal process in the environment. For the 2 chemicals (1%) predicted by the PBT Profiler to be very persistent and the 30 chemicals (18%) predicted to be persistent, all but 3 can be put into one of six chemical groups based on common functionality: chlorophenols and chlorophenyl ethers; beta-chloroethers; aromatic amines; carbothiomate pesticides; thiophosphates; and aromatic carbamates. These chemical classes, as well as the remaining three chemicals, are expected to undergo rapid chemical degradation (half-life < 2 months) in the environment via hydrolysis and/or photolysis (either direct or indirect photo-oxidation) based on the available literature.  Even though chemicals that undergo hydrolysis or other chemical degradation processes should not be run through the PBT Profiler, it is currently the responsibility of the user to identify those chemicals that meet this criteria.  For a screening level method developed for a wide range of potential users, it is, therefore, not appropriate to remove these 20 chemicals from this exercise.  Nevertheless, it emphasizes the high degree of agreement between the PBT Profiler summary and the evaluated data set for this category of chemicals.  Future versions of the PBT Profiler may incorporate structure recognition capabilities that will be able to identify functional groups that are known to undergo rapid chemical degradation.

Analysis of the PBT Profiler summary for persistent chemicals from classes 8 and 9 was also performed (Figure 2). For these 42 chemicals, the Profiler estimated that 36 chemicals would be very persistent and 6 would be persistent.

The 6 chemicals that the PBT Profiler estimated to be persistent include 3 chlorinated biphenyls (a mono, di, and trichloro congenor), pyrene, flouranthene, and trichlorofluoromethane. All of these chemicals should be reviewed carefully when considering P2 opportunities and, from this standpoint, the PBT Profiler summaries are in excellent agreement with the evaluated data.

The pie charts for the chemicals in evaluated classes 6 (Figure 3)and 7 (Figure 4) are also provided. The half-life range in class 6 overlap the not persistent and persistent categories and those in class 7 overlap the persistent and very persistent categories. Because these two classes do not directly correspond to EPA criteria, the results should be viewed with caution. The general trend for the PBT Profiler results from

the class 6 and 7 chemicals is consistent with what one would expect for categories that straddle the persistent/not persistent threshold. There was a decrease in the percentage of chemicals that the PBT Profiler predicted to be not persistent with a concomitant (but smaller) increase in the number predicted to be very persistent. 

Comparison of the PBT Profiler Results with the Evaluated Data Set To Top
 

The PBT Profiler summaries were compared with 293 chemicals from the evaluated data set containing environmental persistence summaries [Mackay, D; Shiu, W.Y.; Ma, K. Physical-Chemical Properties & Environmental Fate on CD-ROM. CRC Press. ISBN/ISSN: 0849321921 (2000)] discussed above.  The purpose of this exercise was to evaluate the water, soil, and sediment half-life multipliers (1, 2 and 9) used by the PBT Profiler as well as using the predominant compartment instead of the media-specific half-life for the persistence summary.

The PBT Profiler summary results were used to separate the 293 chemicals into three groups (not persistent, persistent, and very persistent) and the evaluated half-life class for water, soil, and sediment were plotted for each chemical. The predominate compartment, as determined by the PBT Profiler for each chemical, is also provided for each chemical.  These graphs were then reorganized for presentation (using a primary sort on water half-life, secondary sort on soil half-life, tertiary sort on sediment half-life).  The numbers on the x-axis are simply place holders and each graph tick represents a different chemical.

As discussed above, the results must be interpreted with caution because some of the evaluated classes do not directly correspond with the EPA criteria.  In the following graphs, the area shaded red corresponds with the very persistent PBT Profiler category and the area shaded green corresponds with the not persistent category.  The PBT Profiler persistent category does not directly correspond to the evaluated classes 6 and 7. Figure 5 presents the data for the very persistent category, and the data for the persistent and not persistent categories are shown in Figure 7 and 8, respectively.

Analysis of these data reveals that, in general, the PBT Profiler summaries are in very good agreement with the evaluated data.  The results also indicate that the chemicals in the persistent category overwhelmingly (94%) had soil as the predominant compartment while those in the very persistent and not persistent category had essentially an even split of soil or sediment and soil or water as the predominant compartment.  These graphs suggest that using the predominant compartment to determine the persistence summary is not an overwhelming factor for assigning the overall persistence summary.  For the majority of the chemicals in this data set, selecting the half-life in soil instead of sediment would not have changed the persistence summary. Similarly, selecting the half-life in soil instead of water would not have changed the summary for the not persistent chemicals.

Using the predominant compartment to determine the persistence summary, however, appears to increase the overall predictive accuracy of the methodology.  This is most evident when the half-life in one (or more) of the three environmental compartments fall into different persistence categories.  For example, chemicals 10-15 in the very persistent group could be placed in either the very persistent or not persistent category based on their medium-specific half-lives in sediment and water, respectively.  Given that these chemicals partition predominantly to sediment, their half-life in this media (> 6 months) is likely to be more representative of their persistence in the environment.

An opposite trend is observed for chemicals in the not persistent group that have estimated sediment half-lives greater than 2 months and water (and possibly soil) half-lives less than 2 months.  If the persistence summary was based solely on the highest estimated half-life in water, soil, or sediment, the environmental persistence would likely be overestimated for these chemicals.  This arises directly from the water:soil:sediment half-life multipliers of 1:2:9 used by the PBT Profiler.  If the persistence summary were based on the highest medium-specific half-life, the value for sediment (a factor of 9 that of water) would always be used.  The expected outcome, therefore, would be a general trend that overestimates a chemical's persistence in the environment. Analysis of the available data indicates that using the predominant compartment half-life attenuates this bias and reduces the potential for overestimating a chemical's persistence in the environment.

[Methodology]       [Criteria]       [Definitions]       [Chemicals that Should Not be Profiled]
[Home]       [Start a New Profile]       [Results]       [Terms of Use]       [Security]       [Interpreting Results]

Developed by the Environmental Health Analysis Center under contract to the Office of Chemical Safety and Pollution Prevention , U.S. Environmental Protection Agency
Computer Resources Donated by SRC, Inc.          Ver 2.001     Last Updated September 28, 2016