Shale oils

Partition coefficient

Link to relevant study record(s)

Reference

Reference 1

Endpoint:

partition coefficient

Type of information:

(Q)SAR

Adequacy of study:

key study

Study period:

14 January 2015

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

other: see 'Remark'

Remarks:

The partition coefficient of the substance was estimated using a validated QSAR model. Sufficient information on the model training set is available to evaluate the applicability of the model to this substance. A KOWWIN prediction of representative components based on strong validation data was carried out.

Justification for type of information:

The Log octanol-water partition coefficient (log P) of Distillates (shale oil), middle fraction was estimated using the Log Octanol-Water Partition Coefficient Program (KOWWIN version 1.68, US EPA).
The composition of the material is very complex, being made up of a wide range of substances. The substances can be separated into four distinct groups and a representative substance from each group was therefore used to estimate the Log Octanol-Water Partition Coefficient for that specific group.

KOWWIN estimates the partition of an organic compound by identifying fragments of the structure and summing the known partition coefficients of these fragments.
The substance fell within the molecular weight range of the training set of the model. The functional groups of the substance were included in the training set and the instances of each fragment of the substance did not exceed the maximum count of that fragment in the training set data.

- Domain of the model: There is no domain for the KOWWIN model, however accuracy of the prediction is improved if the molecular weight of the substance is within the molecular weight of the training or validation set, and the number of instances that the fragments occur does not exceed the maximum number of instances these fragments occurred in any of the substances in the training set. Furthermore, accuracy is improved where all fragments of the substance have appropriate coefficients/fragments included in the database.
Training dataset molecular weight range: 18.02-719.92 (Average 199.98)
Validation dataset molecular weight range:27.03-991.15 (Average 258.98)

Qualifier:

according to guideline

Guideline:

other: Guidance on information requirements and chemical safety assessment, Chapter R.6: QSARs and grouping of chemicals (May 2008, ECHA)

Principles of method if other than guideline:

The Log octanol-water partition coefficient (log P) of Distillates (shale oil), middle fraction was estimated using the Log Octanol-Water Partition Coefficient Program (KOWWIN version 1.68, US EPA).
The composition of the material is very complex, being made up of a wide range of substances. The substances can be separated into four distinct groups and a representative substance from each group was therefore used to estimate the Log Octanol-Water Partition Coefficient for that specific group.

KOWWIN estimates the partition of an organic compound by identifying fragments of the structure and summing the known partition coefficients of these fragments.
The substance fell within the molecular weight range of the training set of the model. The functional groups of the substance were included in the training set and the instances of each fragment of the substance did not exceed the maximum count of that fragment in the training set data.

GLP compliance:

no

Remarks:

not applicable as no laboratory work took place

Type of method:

other: computer estimation

Partition coefficient type:

octanol-water

Type:

log Pow

Partition coefficient:

5.19

Remarks on result:

other: estimated log Pow for butylnaphthalene

Type:

log Pow

Partition coefficient:

3.11

Remarks on result:

other: estimated log Pow for diethylresorcinol

Type:

log Pow

Partition coefficient:

8

Remarks on result:

other: estimated log Pow for dihexylbenzene

Type:

log Pow

Partition coefficient:

8.2

Remarks on result:

other: estimated log Pow for hexadecane

Details on results:

METHODOLOGY
KOWWIN uses a "fragment constant" methodology to predict log P. The structure is divided into fragments (atom or larger functional groups) and coefficient values of each fragment or group are summed together to yield the log P estimate. Coefficients for individual fragments and groups were derived by multiple regression of 2447 reliably measured log P values.
Two separate regression analyses were performed. The first regression related log P to atom/fragments of compounds that do not require correction factors (i.e., compounds estimated adequately by fragments alone).
The correction factors were then derived from a multiple linear regression that correlated differences between the experimental (expl) log P and the log P estimated by Equation 1 below with the correction factor descriptors. This regression did not utilize an additional equation constant.
The general regression equation (first regression) is expressed as follows:
log P = Σ(fini ) + b
where:
Σ(fini ) = the summation of fi (the coefficient for each atom/fragment) times ni (the number of times the atom/fragment occurs in the structure)
b= the linear equation constant.
This initial regression used 1120 compounds of the 2447 compounds in the total training dataset.

The equation for the second regression is:
lop P (expl) - log P (eq 1) = Σ(cjnj )
where:
Σ(cjnj ) = the summation of cj (the coefficient for each correction factor) times nj (the number of times the correction factor occurs (or is applied) in the molecule).

Results of the two successive multiple regressions (first for atom/fragments and second for correction factors) yield the following general equation for estimating log P of any organic compound:
log P = Σ(fini ) + Σ(cjnj ) + 0.229
(number = 2447, r² = 0.982, standard deviation = 0.217, mean error = 0.159)

- Domain of the model: There is no domain for the KOWWIN model, however accuracy of the prediction is improved if the molecular weight of the substance is within the molecular weight of the training or validation set, and the number of instances that the fragments occur does not exceed the maximum number of instances these fragments occurred in any of the substances in the training set. Furthermore, accuracy is improved where all fragments of the substance have appropriate coefficients/fragments included in the database.
Training dataset molecular weight range: 18.02-719.92 (Average 199.98)
Validation dataset molecular weight range:27.03-991.15 (Average 258.98)

1) KowWin results for butylnaphthalene:

Log Kow (version 1.68 estimate): 5.19

TYPE	NUM	LOGKOW FRAGMENT DESCRIPTION	COEFF	VALUE
Frag	1	-CH3 [aliphatic carbon]	0.5473	0.5473
Frag	3	-CH2- [aliphatic carbon]	0.4911	1.4733
Frag	10	Aromatic Carbon	0.2940	2.9400
Const		Equation Constant		0.2290

Log Kow = 5.1896

 

2) KowWin results for diethylresorcinol:

Log Kow(version 1.68 estimate): 3.11

TYPE	NUM	LOGKOW FRAGMENT DESCRIPTION	COEFF	VALUE
Frag	2	-CH3 [aliphatic carbon]	0.5473	1.0946
Frag	2	-CH2- [aliphatic carbon]	0.4911	0.9822
Frag	6	Aromatic Carbon	0.2940	1.7648
Frag	2	-OH [hydroxy, aromatic attach]	-0.4802	-0.9604
Const		Equation Constant		0.2290

Log Kow = 3.1094

 

3) KowWin results for dihexylbenzene:

Log Kow(version 1.68 estimate): 8.00

TYPE	NUM	LOGKOW FRAGMENT DESCRIPTION	COEFF	VALUE
Frag	2	-CH3 [aliphatic carbon]	0.5473	1.0946
Frag	10	-CH2- [aliphatic carbon]	0.4911	4.9110
Frag	6	Aromatic Carbon	0.2940	1.7640
Const		Equation Constant		0.2290

Log Kow = 7.9986

 

4) KowWin results for hexadecane:

Log Kow(version 1.68 estimate): 8.20

TYPE	NUM	LOGKOW FRAGMENT DESCRIPTION	COEFF	VALUE
Frag	2	-CH3 [aliphatic carbon]	0.5473	1.0946
Frag	14	-CH2- [aliphatic carbon]	0.4911	6.8754
Const		Equation Constant		0.2290

Log Kow = 8.1990

Conclusions:

The KOWWIN model estimated the partition coefficient of the four representative substances to be as follows:
The log Pow for butylnaphthalene was estimated to be 5.19.
The log Pow for diethylresorcinol was estimated to be 3.11.
The log Pow for dihexylbenzene was estimated to be 8.00.
The log Pow for hexadecane was estimated to be 8.20.

Executive summary:

Standard experimental test methods have been determined not to be suitable for this UVCB substance and hence it was concluded to be most appropriate to provide estimated log Pow values for four representative substances; these four substances were taken from the four classes which have been determined to make up the composition of the substance. Butylnaphthalene, diethylresorcinol, dihexylbenzene and hexadecane were selected as the four representative substances.

The Log octanol-water partition coefficient (log Pow) of these components of Distillates (shale oil), middle fraction was therefore estimated using the Log Octanol‑Water Partition Coefficient Program (KOWWIN model v1.68 (EPI Suite v4.11)).

KOWWIN estimates the partition of an organic compound by identifying fragments of the structure and summing the known partition coefficients of these fragments. The substances fell within the molecular weight range of the training set of the model. The functional groups of the substances were included in the training set and the instances of each fragment of the substances did not exceed the maximum count of that fragment in the training set data.

The KOWWIN model estimated the partition coefficient of the four representative substances to be as follows:

The log Pow for butylnaphthalene was estimated to be 5.19.

The log Pow for diethylresorcinol was estimated to be 3.11.

The log Pow for dihexylbenzene was estimated to be 8.00.

The log Pow for hexadecane was estimated to be 8.20.

Reference 2

Endpoint:

partition coefficient

Type of information:

experimental study

Adequacy of study:

supporting study

Study period:

21st July - 7th September 2005

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

other: GLP study conducted in compliance with a standard protocol but the results are of limited value due to difficulties in performing the study on a UVCB test substance.

Qualifier:

no guideline followed

Principles of method if other than guideline:

Neither the HPLC-method according to OECD Guideline No. 117 nor the flask-shaking method according to OECD Guideline No. 107 were applicable for the determination of the partition coefficient of Shale Oil. Thus, the log POW-value for the test material was estimated from its solubility in n-octanol and in water, respectively.

GLP compliance:

yes (incl. QA statement)

Type of method:

other: estimation

Partition coefficient type:

octanol-water

Analytical method:

other: N/A

Type:

log Pow

Partition coefficient:

> 3.4

Temp.:

20 °C

Details on results:

In the present study, the n-octanol solubility of Shale Oil was estimated to be > 544 g/l by visual judgement.

log P_OW = log₁₀ (> 544 g/l / 0.22 g/l) > 3.4

Conclusions:

In conclusion, the partition coefficient, log Pow of Shale Oil was calculated from the solubility in n-octanol and the content of TOC (total organic carbon) in water, respectively, to be > 3.4

Executive summary:

A GLP study was performed to investigated the partition coefficient of the test material. Neither the HPLC-method according to OECD Guideline No. 117 nor the flask-shaking method according to OECD Guideline No. 107 were applicable for the determination of the partition coefficient of Shale Oil. Thus, the log P_OW-value for the test material was estimated from its solubility in n-octanol and in water, respectively.

The partition coefficient, log P_OWof Shale Oil was calculated from the solubility in n-octanol and the content of TOC (total organic carbon) in water, respectively, to be > 3.4

Description of key information

The partition coefficients of four representative components of the substance have been estimated by KOWWIN version 1.68 to vary between log Pow = 3.11 and log Pow = 8.20.

Key value for chemical safety assessment

Log Kow (Log Pow):

3.4

at the temperature of:

20 °C

Additional information

Standard experimental test methods have been determined not to be suitable for this UVCB substance and hence it was concluded to be most appropriate to provide estimated log Pow values for four representative substances; these four substances were taken from the four classes which have been determined to make up the composition of the substance. Butylnaphthalene, diethylresorcinol, dihexylbenzene and hexadecane were selected as the four representative substances.

The Log octanol-water partition coefficient (log Pow) of these components of Distillates (shale oil), middle fraction was therefore estimated using the Log Octanol‑Water Partition Coefficient Program (KOWWIN model v1.68 (EPI Suite v4.11)).

KOWWIN estimates the partition of an organic compound by identifying fragments of the structure and summing the known partition coefficients of these fragments. The substances fell within the molecular weight range of the training set of the model. The functional groups of the substances were included in the training set and the instances of each fragment of the substances did not exceed the maximum count of that fragment in the training set data.

The KOWWIN model estimated the partition coefficient of the four representative substances to be as follows:

The log Pow for butylnaphthalene was estimated to be 5.19.

The log Pow for diethylresorcinol was estimated to be 3.11.

The log Pow for dihexylbenzene was estimated to be 8.00.

The log Pow for hexadecane was estimated to be 8.20.

Supporting information includes a GLP study which was performed to investigated the partition coefficient of the registered substance. Neither the HPLC-method according to OECD Guideline No. 117 nor the flask-shaking method according to OECD Guideline No. 107 were applicable for the determination of the partition coefficient of Shale Oil. Thus, the log P_OW-value for the test material was estimated from its solubility in n-octanol and in water, respectively.

The partition coefficient, log P_OW of Shale Oil was calculated from the solubility in n-octanol and the content of TOC (total organic carbon) in water, respectively, to be > 3.4