Naphthalene-2,6-dicarboxylic acid

Adsorption / desorption

Link to relevant study record(s)

Reference

Reference 1

Endpoint:

adsorption / desorption: screening

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model and falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE
EPI Suite 4.1, developed by the United States Environmental Protection Agency (US EPA) copyright © 2000 U.S. Environmental Protection Agency

2. MODEL (incl. version number)
KOCWIN v2.00 (2010) implemented in EPI Suite

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL
The SMILES notation was used as identifier

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
A method of estimating soil adsorption Koc involves correlations was developed by Doucette with log octanol-water partition coefficient (log Kow). Since an expanded experimental Koc database was available from the new MCI regression (see endpoint study record Meylan et al (1992)), it was decided to develop a log Kow estimation methodology that was potentially more accurate than existing log Kow QSARs for diverse structure datasets.
Effectively, the new log Kow methodology simply replaces the MCI descriptor with log Kow and derives similar equations.  The derivation uses the same training and validation data sets.  The training set is divided into the same non-polar (no correction factors) and correction factor sets.  The same correction factors are also used.

5. APPLICABILITY DOMAIN
- Descriptor domain:
According to EPI Suite, there is currently no universally accepted definition of a model domain. However, EPI Suite claimes that the substance might be out of the range of the domain, if the molecular weight is outside the range of the training set. For the MCI method the range of molecular weights in the training set is between 32.04 and 665.02 (average 224.4). With a molecular weight of 216.19, 2,6-naphthalend dicarboxylic acid is close to the average of the molecular weight of the training set substances. Hence, it falls within the applicabilita of the domain.
Statistics derived from the training set: n = 68, r² = 0.877, std dev = 0.478, avg dev = 0.371
- Similarity with analogues in the training set: Substances with a naphthalene basic structure were part of the training set.

Qualifier:

no guideline required

Principles of method if other than guideline:

See field "Justification for type of information"

Key result

Type:

log Koc

Value:

1.776 dimensionless

Reference 2

Endpoint:

adsorption / desorption: screening

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model and falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE
Chemprop(TM) Main Module 6.6 developed by the Helmholtz Centre for Environmental Research (UFZ) in Leipzig, Germany

2. MODEL (incl. version number)
The model of Huuskonen is implemented in the software

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL,
The CAS-No was used as identifier

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL AND APPLICABILITY DOMAIN
A group contribution approach based on atom-type electrotopological state indices for predicting the soil sorption coefficient (log Koc) of a diverse set of 201 organic pesticides is presented. Using a training set of 143 compounds, for which the log KOC values were in the range from 0.42 to 5.31, multiple linear regression (MLR) and artificial neural networks were used to build the models. The models were validated using two test sets of 20 and 38 chemicals not included in the training set. The statistics for a linear model with 12 structural parameters were, in test set 1, r2 = 0.79, s = 0.45 and, in test set 2, r2 = 0.74, s = 0.65. These results clearly show that soil sorption coefficients can be accurately and rapidly estimated from easily calculated structural parameters.
The log Koc is calculated according to the following formula: log Koc = 0.350chi + Sum(ai•Si) + 0.622
with the connectivity index chi and the sums Si of particular electrotopological states times factors ai

Qualifier:

no guideline required

Principles of method if other than guideline:

See field "Justification for type of information"

Key result

Type:

log Koc

Value:

2.59 dimensionless

Reference 3

Endpoint:

adsorption / desorption: screening

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model and falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE
Chemprop(TM) Main Module 6.6 developed by the Helmholtz Centre for Environmental Research (UFZ) in Leipzig, Germany

2. MODEL (incl. version number)
The model of Huuskonen is implemented in the software

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL
The CAS-No was used as identifier

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL AND APPLICABILITY DOMAIN
A correlation study based on simple structural descriptors for predicting the soil sorption coefficient, log Koc, of a diverse set of 568 organic compounds was developed. Using a training set of 403 compounds, in which the log Kocvalues were in the range 0−6.5, multiple linear regression (MLR) was utilized to build the models. The models were validated using a test set of 165 chemicals not included in the training set. The statistics for a linear regression model with calculated aqueous solubility, log S, were r2 = 0.80 and s = 0.51 in the training set, and r2 = 0.76 and s = 0.61 in the test set. The model parameters used allow rapid and accurate calculation of log Kocvalues for a diverse set of organic chemicals, and propose the importance of molecular solubility, lipophilicity, size, flexibility, and ionization for a chemicals' sorption to organic soil material.

The log Koc is calculated according to the following equation:
log Koc = 0.48 log Kow + 0.26 NAR - 0.07 ROT + 0.002 MW -0.77 Iacid + 0.56

where
HBA = number of O and N
NAR = number of aromatic 5 or 6 rings
MW = molweight
Iacid = 1 for COOH, 0 otherwise
ROT = number of rotational bonds

Reason / purpose for cross-reference:

reference to other study

Qualifier:

no guideline required

Principles of method if other than guideline:

See field "Justification for type of information"

Key result

Type:

log Koc

Value:

1.66 dimensionless

Reference 4

Endpoint:

adsorption / desorption: screening

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model and falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE
EPI Suite 4.1, developed by the United States Environmental Protection Agency (US EPA) copyright © 2000 U.S. Environmental Protection Agency

2. MODEL (incl. version number)
KOCWIN v2.00 (2010) implemented in EPI Suite

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL
The SMILES notation was used as identifier

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
Journal abstract:
"The first-order molecular connectivity index (MCI) has been successfully used to predict soil sorption coefficients (Koc) for nonpolar organics, but extension of the model to polar compounds has been problematic.  To address this, we developed a new estimation method based on MCI and series of statistically derived fragment contribution factors for polar compounds. After developing an extensive database of measured Koc values, we divided the dataset into a training set of 189 chemicals and an independent validation set of 205 chemicals.  Two linear regressions were then performed.  First, measured log Koc values for nonpolar compounds in the training set were correlated with MCI.  The second regression was developed by using the deviations between measured log Koc and the log Koc estimated with the nonpolar equation and the number of certain structural fragments in the polar compounds.  The final equation for predicting log Koc accounts for 96% and 86% of the variation in the measured values for the training and validation sets, respectively.  Results also show that the model outperforms and covers a wider range of chemical structures than do models based on octanol-water partition coefficients (Kow) or water solubility."

5. APPLICABILITY DOMAIN
- Descriptor domain: According to EPI Suite, there is currently no universally accepted definition of a model domain. However, EPI Suite claimes that the substance might be out of the range of the domain, if the molecular weight is outside the range of the training set. For the MCI method the range of molecular weights in the training set is between 32.04 and 665.02 (average 224.4). With a molecular weight of 216.19, 2,6-naphthalend dicarboxylic acid is close to the average of the molecular weight of the training set substances. Hence, it falls within the applicabilita of the domain.
Statistics derived from the training set: n = 69, r² = 0.967, std dev = 0.247, avg dev = 0.199
- Similarity with analogues in the training set: Substances with a naphthalene basic structure were part of the training set.

Qualifier:

no guideline required

Principles of method if other than guideline:

See field "Justification for type of information"

Key result

Type:

log Koc

Value:

2.924 dimensionless

Reference 5

Endpoint:

adsorption / desorption: screening

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model and falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE
Chemprop(TM) Main Module 6.6 developed by the Helmholtz Centre for Environmental Research (UFZ) in Leipzig, Germany

2. MODEL (incl. version number)
The model of Huuskonen is implemented in the software

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL
The CAS-No was used as identifie

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL AND APPLICABILITY DOMAIN
Abstract of the publication:
The solvation parameter model is used to construct models for the estimation of the soil–water and soil–air distribution constants and to characterize the contribution of fundamental intermolecular interactions to the underlying sorption processes. Wet soil is shown to be quite cohesive and polar but relatively non-selective for dipole-type, lone-pair electron and hydrogen-bond interactions. Using a comparison of system constant ratios chromatographic systems employing reversed-phase liquid chromatography on polar bonded phases are shown to provide suitable models for estimating soil–water distribution constants. No suitable gas chromatographic models were found for the soil–air distribution constant but the requirements for such a system are indicated. Models are also provided for adsorption at the air–water interface. Estimation methods based on either the solvation parameter model or chromatographic model reproduce experimental distribution constants for a wide variety of compounds with a similar error (0.2–0.3 log units) to that expected in the experimental data.

Qualifier:

no guideline required

Principles of method if other than guideline:

See field "Justification for type of information"

Key result

Type:

log Koc

Value:

2.2 dimensionless

Reference 6

Endpoint:

adsorption / desorption: screening

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model and falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. MODEL
The model of Sabljic et al is published in the Technical Guidance Document on Risk Assessment, PART III, published by the European Commission. It is also implemented in the software Chemprop(TM) Main Module 6.6 developed by the Helmholtz Centre for Environmental Research (UFZ) in Leipzig, Germany.

2. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL
The log Pow is used as input for the model

3. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
- Defined endpoint: Adsorption coefficient log Koc
- Unambiguous algorithm: logKoc = 0.60 logKow + 0.32 (for organic acids). A log Pow of 2.22 was used for the calculation of the log Koc. For detailes please refer to the endpoint summary of section 4.7
- Defined domain of applicability: According to Sabljic et al., this QSAR applies for all organic acids with a log Pow between -0.5 an 4.0
- Appropriate measures of goodness-of-fit and robustness and predictivity: n=23, r2=0.75, s.e.=0.34

4. APPLICABILITY DOMAIN
With a log Pow of 2.22, the substance falls into the applicability of the above defined domain

5. ADEQUACY OF THE RESULT
The log Koc is not relevant for classification and labelling

Reason / purpose for cross-reference:

reference to other study

Qualifier:

no guideline required

Principles of method if other than guideline:

See field "Justification for type of information"

Key result

Type:

log Koc

Value:

1.65 dimensionless

Reference 7

Endpoint:

adsorption / desorption: screening

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model and falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE
Chemprop(TM) Main Module 6.6 developed by the Helmholtz Centre for Environmental Research (UFZ) in Leipzig, Germany

2. MODEL (incl. version number)
The model of Schüürmann et al. is implemented in the software

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL ,
The CAS-No was used as identifier

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
Abstract of the publication
A new model to estimate the soil-water partition coefficient normalized to soil organic carbon, Koc, from the two-dimensional molecular structure is presented. Literature data of log Koc for 571 organic chemicals were fitted to 31 parameters with a squared correlation coefficient r2 of 0.854 and a standard error of
0.467 log units. The application domain includes the atom types C, H, N, O, P, S, F, Cl and Br in various important compound classes. The multi-linear model contains the parameters molecular weight, bond connectivity, molecular E-state, an indicator for nonpolar and weakly polar compounds, and 26 fragment corrections representing polar groups. The prediction capability is evaluated through an initial two-step development using an 80%:20% split of the data into training and prediction, cross-validation, permutation and application to four external data sets. The discussion includes separate analyses for subsets of H-bond donors and acceptors as well as for nonpolar and weakly polar compounds. Comparison with existing models including linear solvation energy relationships illustrates the superiority of the new model.

5. APPLICABILITY DOMAIN
Applicability domain:
Two separate aspects of the model application domain are available. The default application domain test considers the chemical space only, an additional descriptor domain test can be toggled on via the method parameters.

Descriptor Domain
The range confidence is checked for the molecular correction factors by comparing the frequency of these substructures in the molecule to the training set. The number of correction factors exceeding the training set frequency is checked and may optionally be shown in the output.

Chemical domain
Unless turned off by options or parameters, the compound will be checked versus the original training set compounds by means of 2nd order ACFs, and any mismatch of the chemical domain will be reported.

2,6-Naphthalene dicarboxylic acid is classified as category 2 "Borderline in". That means that either the frequency of at least one substructure of the compound exceeds the range of occurrences in the training set, or one substructure is not in the training set at all.

Qualifier:

no guideline required

Principles of method if other than guideline:

See field "Justification for type of information"

Key result

Type:

log Koc

Value:

2.33 dimensionless

Reference 8

Endpoint:

adsorption / desorption: screening

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model and falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE
Chemprop(TM) Main Module 6.6 developed by the Helmholtz Centre for Environmental Research (UFZ) in Leipzig, Germany

2. MODEL (incl. version number)
The model of Tao et al. is implemented in the software

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL AND APPLICABILITY DOMAIN
Abstract of the publication:
A fragment constant model for prediction of KOC was developed and evaluated with a diverse database of 592 chemicals belonging to 17 classes. The range of experimental KOC covered 7.65 log-units. The 592 chemicals were randomly divided into a training set and a testing set for model development and validation. A general model was then established using the entire database having 74 fragment constants and 24 correction factors. Statistically, the regression model
accounted for as much as 96.96% of the variation in the measured log KOC. The mean residual between the experimental and predicted KOC values was 0.366 log-units. In more than 74% of the chemicals studied the residual values were less than 0.5 log-units. The robustness of the regression model, with respect to either specific individual chemicals or particular compound classes, was evaluated through use of jackknife tests. The experimental results confirmed the ability of the fragment model to predict KOC for a wide variety of untested chemicals.

Qualifier:

no guideline required

Principles of method if other than guideline:

See field "Justification for type of information"

Key result

Type:

log Koc

Value:

1.98 dimensionless

Description of key information

The log Koc of 2,6 -Naphthalene dicarboxylic acid was estimated by in total 8 QSAR models and calcualtions. A summary of the results is given in the following table:

Author	log Koc
Sabljic et al. (1995)	1.65
Schüürmann et al. (2006)	2.33
Huuskonen (2003), ESI	2.59
Tao et al. (1999)	1.98
Huuskonen (2003), from log Pow	1.66
Poole & Poole (1999)	2.2
Meylan et al. (1992)	2.924
Doucette (2000)	1.776

All models are scientifically well recognised and published in scientific journals. However, as the publications are not available, the applicability of the domain cannot be proven for some of the models. But based on the weight of evidence approach, the models in a whole can be considered to be suitable for the estimation of the log Koc, even if some of them do not fall into the applicablity of the respective domain. As the log Koc values were estimated with different models, the median of the values was selected, as it is less sensitive for outliers than the arithmetic mean.

Key value for chemical safety assessment

Koc at 20 °C:

2.09