Perhydroimidazo[4,5-d]imidazole-2,5-dione

Partition coefficient

Link to relevant study record(s)

Reference

Reference 1

Endpoint:

partition coefficient

Data waiving:

study technically not feasible

Justification for data waiving:

the study does not need to be conducted because the substance does not dissolve in water or in octanol

Justification for type of information:

JUSTIFICATION FOR DATA WAIVING
The study does not need to be conducted since the substance does not dissolve in octanol.

Reference 2

Endpoint:

partition coefficient

Type of information:

(Q)SAR

Adequacy of study:

key study

Study period:

2018

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
Estimation Programme Interface (EPI) Suite programme for Microsoft Windows v4.11
Contact EPISuite:
U.S. Environmental Protection Agency
1200 Pennsylvania Ave.
N.W. (Mail Code 7406M)
Washington, DC 20460

2. MODEL (incl. version number)
KOWWIN v1.68
September 2010 (model development); November 2012 (model publication)

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL
See QPRF attached: ‘QPRF Title: Substance: Perhydroimidazo[4,5-d]imidazole-2,5-dione using the model KOWWIN v1.68 for the endpoint: Partition Coefficient (Log Kow)’ version 1; dated 12 April 2018.

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
Full details of the method are provided in the attached QMRF named ‘QMRF Title: KOWWIN v1.68 : n-Octanol/Water Partition Coefficient (Log Kow)’ version 1.02; date: 01 September 2015; updated 10 April 2018.

5. APPLICABILITY DOMAIN
See QPRF attached: ‘QPRF Title: Substance: Perhydroimidazo[4,5-d]imidazole-2,5-dione using the model KOWWIN v1.68 for the endpoint: Partition Coefficient (Log Kow)’ version 1; dated 12 April 2018.

6. ADEQUACY OF THE RESULT
1) QSAR model is scientifically valid. 2) The substance falls within the applicability domain of the QSAR model. 3) The results are adequate when taken under consideration of REACH Regulation (EC) 1907/2006 in a weight of evidence.

Guideline:

other: REACH Guidance on QSARs R.6, May/July 2008

Version / remarks:

Substance not soluble in Octanol. Calculation Used.

Principles of method if other than guideline:

Full details of the method are provided in the attached QMRF named ‘QMRF Title: KOWWIN v1.68 : n-Octanol/Water Partition Coefficient (Log Kow)’ version 1.02; date: 01 September 2015; updated 10 April 2018.
- The model applies the following methodology to generate predictions:
Fragment based (group contribution) QSAR; based on multiple linear-regression modelling

- The model and the training and validation sets are published by US Environmental Protection Agency (USA).
The experimental Log Kow values in the training set and validation set were measured using one or more methods equivalent or similar to the following guidelines:
- Shake Flask method (OECD TG 107)
- HPLC method (OECD TG 117)
- Slow Stirring method (OECD TG 123)
Plus relevant EU (1992 as amended) and US EPA OPPTS (1982 as amended) and ASTM (1993) methods may be also used where appropriate.
A full list of experimental Log Kow reference citations is provided in the KOWWIN help menu with additional reference citations.
- Justification of QSAR prediction: The result should be considered in relation to corresponding information presented and in accordance with the tonnage driven information requirements of REACH Regulation (EC) 1907/2006 in a weight of evidence.
Specifically the substance is hydrolytically unstable. The substance constituents have short hydrolytic half-lives (scale: hours at neutral pH, ambient temperatures, accelerated in acidic or basic pH conditions by catalytic hydrolysis) they would be considered as fulfilling the criteria for rapid (primary) degradability.
Provision of measured experimental log Kow has been determined as not technically possible under guideline OECD TG 117 HPLC method. It is also potentially not possible by other methods such as OECD TG 123 slow stir method.
Therefore, in accordance with the tonnage driven information requirements: the calculated method for Log Kow has been adopted and including consideration of modelled hydrolysis products and their respective Log Kow as relevant supporting information. There is evidence the substance and its hydrolysis products may be presented in representative blocks under the ‘block method’ of environmental risk assessment, given their comparable chemistries and physico-chemical properties. Specifically, the substance consists of blocks of constituents of Log Kow < 4.0 and/or relevant hydrolysis products have increasing water solubility and/or corresponding decreasing Log Kow.

Type of method:

estimation method (solubility ratio)

Partition coefficient type:

octanol-water

Analytical method:

other: calculation

Type:

log Pow

Partition coefficient:

ca. -3.28

Remarks on result:

not determinable because of methodological limitations

1. Defining the endpoint

QMRF 1. Physical Chemical Properties.

QMRF 1.6. Octanol-water partition coefficient (Kow).

Reference to type of model used and description of results: KOWWIN v1.68; integrated within the Estimation Programme Interface (EPI) Suite programme for Microsoft Windows v4.11; September 2010 (model development); November 2012 (model publication).

2. Description of results and assessment of reliability of the prediction

The predicted values are provided within the QPRF attached: ‘QPRF Title: Substance: Perhydroimidazo[4,5 -d]imidazole-2,5-dione using the model KOWWIN v1.68 for the endpoint: Partition Coefficient (Log Kow)' version 1.0; dated 10 April 2018.

Log Pow: ca. -3.28.

It is noted by the applicant there is no universally acknowledged applicability domain for the model. However, assessment of the substance within the applicability domains recommended by the developers is documented within the corresponding QMRF named ‘QMRF Title: KOWWIN v1.68 : n-Octanol/Water Partition Coefficient (Log Kow)’ version 1.02 – section 5; indicates the substance (constituents):

(i) All constituents fall within the Molecular Weight range domain.

(ii) No substances have functional groups or features not in the training set of the model and/or for which no fragment constants and correction factors available. No constituents contain multiple fragment instances than the maximum of the training set (see QMRF title section 9.3 for more information).

Expert review of the data on epoxy or epoxide structural analogues in the validation dataset indicate that in general they over predict Log Kow compared to measured data. Additionally, there are several or more hydroxy fused rings in the training and validation sets. 

3. Uncertainty of the prediction and mechanistic domain

The training set of the model has the following statistics and coefficients of determination:

Total Training Set Statistics: number in dataset = 2447 ; correlation coef (r2) = 0.982 ; standard deviation = 0.217 ; absolute deviation = 0.159 and avg Molecular Weight = 199.98

The model has been externally validated on a set of 10,946 substances and the following statistics and coefficients of determination are presented:

Total Validation Set Statistics: number in dataset = 10946 ; correlation coef (r2) = 0.943 ; standard deviation = 0.479 ; absolute deviation = 0.356 ; avg Molecular Weight = 258.98

Data for the training set are available via external validation (see attached QMRF prepared by the applicant for full citations).

There is no overt mechanistic basis for the model. The model correlates thermodynamic relationships of surrogates to chemical activity. The KOWWIN v1.68 run in standalone mode allows Log Kow to be estimated based on measured values of analogues within the training set (if available). Then the model applies by adding/subtracting fragment constants and correction factors from the measured value. This therefore improves prediction since calculations are based on structural differences between target and analogue. The model domain ideally has at least one or more structurally similar substances to target substances on which to then apply ACF methodology. Whilst there appears to be no direct analogues within the training set. The model has been has been extensively validated externally (using > 10,000) substances with a correlation coefficient (r2) = 0.943. The model is non-proprietary and the training sets and validation sets can be downloaded from the internet. A summary of this information is presented by the applicant. Expert review of the data on relevant structural analogues in the validation dataset indicate that in general they over predict Log Kow compared to measured data.

Model predictivity could be improved by the assignment of additional substances into the training set. Inclusion of additional structural fragments and expansion of sub-structure correction factors and related rules. In addition, rules for stereochemical effects could feasibly improve modelling.

Conclusions:

The results are adequate for the regulatory purpose.

Executive summary:

KOWWIN v1.68 (model publication: November 2012)

Log Pow: ca. -3.28.

Adequacy of the QSAR:

1) QSAR model is scientifically valid. 2) The substance falls within the applicability domain of the QSAR model. 3) The prediction is fit for regulatory purpose.

The prediction is adequate for the Classification and Labelling or risk assessment of the substance as indicated in REACH Regulation (EC) 1907/2006: Annex XI Section 1.3. No substance constituents exceed the Log Kow = 4.0 cut-off. The assessment indicates that the prediction is suitable for the regulatory conclusion in accordance with the tonnage driven information requirements.

Description of key information

In accordance with REACH Regulation (EC) No. 1907/2006 Annex VII, column 2 section 7.8 the study does not need to be conducted if the test cannot be performed (e.g. the study is not techically feasible - substance does not dissolve in octanol). A calculated value for LogKow is provided as ca. -3.280 based on the (Q)SAR model KOWWIN v1.68.

KOWWIN v1.68 (model publication: November 2012)

Log Pow: ca. -3.28.

Adequacy of the QSAR:

1) QSAR model is scientifically valid. 2) The substance falls within the applicability domain of the QSAR model. 3) The prediction is fit for regulatory purpose.

The prediction is adequate for the Classification and Labelling or risk assessment of the substance as indicated in REACH Regulation (EC) 1907/2006: Annex XI Section 1.3. No substance constituents exceed the Log Kow = 4.0 cut-off. The assessment indicates that the prediction is suitable for the regulatory conclusion in accordance with the tonnage driven information requirements.

Key value for chemical safety assessment

Log Kow (Log Pow):

-3.28

Additional information

KOWWIN v1.68 (model publication: November 2012)

Log Pow: ca. -3.28.

Adequacy of the QSAR:

1) QSAR model is scientifically valid. 2) The substance falls within the applicability domain of the QSAR model. 3) The prediction is fit for regulatory purpose.

The prediction is adequate for the Classification and Labelling or risk assessment of the substance as indicated in REACH Regulation (EC) 1907/2006: Annex XI Section 1.3. No substance constituents exceed the Log Kow = 4.0 cut-off. The assessment indicates that the prediction is suitable for the regulatory conclusion in accordance with the tonnage driven information requirements.