Extract obtained from defatted powder of Theobroma cacao (Malvaceae) by extraction with ethanol

Partition coefficient

Link to relevant study record(s)

Reference

Reference 1

Endpoint:

partition coefficient

Type of information:

(Q)SAR

Adequacy of study:

supporting study

Study period:

run on 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 limited documentation / justification

Reason / purpose for cross-reference:

reference to other study

Principles of method if other than guideline:

QSAR estimation

Partition coefficient type:

octanol-water

Type:

log Pow

Partition coefficient:

>= -5.12 - <= 0.16

Remarks on result:

other: Temperature and pH not provided (QSAR estimation)

Remarks:

Range of logKow estimated on a the known constituents of the substance

Temperature and pH values are not considered in this model.

The estimation is based on the non-ionised form of the molecule.

No Experimental Value Adjustment was made.

The molecular weight of the constituents are between 46.0684 and 342.2965 g/mol, hence within the estimation domain of the validation set (27.03 -991.15 g/mol), all the functional groups are included in the list of descriptors, and, the occurence of each fragment is below the maximum number in the training/validation sets (Appendix D), such as:

		Training Set		Validation Set
Fragment Descriptor & Correction Factor Descriptor	Coef	Max	Number	Max	Number
-CH3 [aliphatic carbon]	0,5473	13	1401	20	7413
-CH2-   [aliphatic carbon]	0,4911	18	1083	28	7051
-CH   [aliphatic carbon]	0,3614	16	460	23	3864
-COOH   [acid, aliphatic attach]	-0,6895	3	126	3	672
Aromatic carbon	0,294	24	1790	30	8792
-C(=O)- [carbonyl, aliphatic attach]	-1,5586	2	54	2	241
-OH     [hydroxy, aliphatic attach]	-1,4086	6	187	9	1525
-O-   [oxygen, aliphatic attach]	-1,2566	5	108	12	1235
C-O-C-O-C Structure correction	0,5036	3	24	4	270
Multi-alcohol correction	0,4064	1	97	1	499
HO-CH-C(-O-)-CH-OH structure correction	1,0649	1	20	3	133
HO-CH-C(-OH)-CH-OH structure correction	0,5944	2	19	4	98
-O-C(-OH)-C-OH structure correction	0,2248	1	3	1	2
-O-C(-C-HO)-C-O- structure correction	0,85	2	3	3	55
HO-C-C(=O)-C-OH structure correction	2,0476	1	8	1	11
HO-C-C(-OH)-C(=O)- structure correction	1,2	1	2	1	2
-N< [aliphatic attach]	-1,8323	4	308	6	2304
-N [aliphatic N, one aromatic attach]	-0,917	3	541	5	2687
-C(=O)N- [aromatic attach]	0,1599	2	128	3	847
-NC(=O)N- [urea]	1,0453	2	157	2	758
Aromatic Nitrogen [5-member ring]	-0,5262	3	130	6	1746
C [aliphatic carbon - No H, not tert]	0,9723	3	242	11	1361
HO-C-COOH (alcohol-acid carbon)	0,3114	1	6	1	8
-NH- [aliphatic attach]	-1,4962	4	416	5	2371

Executive summary:

When the constituents of the substance were sufficiently known and were within the applicability domain (MW, descriptors), the partition coefficient of the known constituents of the substance were estimated using the recommended QSAR model KOWWIN v1.68, based on Atom/Fragment Contribution (AFC) method.

Log Kow of these constituents were estimated to be between -5.12 and 0.16 (estimated by QSAR) but this range is not representative of the major part of the substance.

Reference 2

Endpoint:

partition coefficient

Data waiving:

study technically not feasible

Justification for data waiving:

other:

Description of key information

No range of logKow representative of the test item can be determined (neither with an experimental approach nor with an estimation approach)

Additional information

The substance is a complex mixture only partially known and its physical state leads to difficulties for physico-chemical testing: It is described as a black-brown pasty solid, mixture of porous clumps.

GLP studies according to OECD 117 have been conducted on other substances belonging to the same group of substance (UVCB sub-type 3, where the source is biological and the process is refinement, such as : exudates, resinoids and concretes that are well-known to be very difficult to analytically assess). These studies showed that this method was not appropriate for such complex substances (the study lead to one peak or to a significant part of the substance with a log Kow < 0.3 while the substance was not soluble in water). In addition, % Area is not necessarily correlated to absolute concentration in the substance if UV detection is used. Therefore we cannot guarantee the repartition of the logPow in the composition of the substance and that the results obtained by HPLC/UV-DAD are representative of the whole substance.

Moreover the physical state, the limited knowledge on the composition of the substance and the diverse nature of the constituents of the substance (proteins, lipids, sucroses,...) lead to technical issues to quantify the substance dissolved in water or in octanol media (HPLC do not allow to detect the totality of the test item constituents). Then OCDE 107 and 123 are also considered as not appropriate.

Finally as reported in the EU/OECD guidelines and in Chapter R.7a,  v6.0, July 2017,R.7.1.8.3, the high performance liquid chromatography (HPLC) is applicable when the log Pow value falls within range 0 to 6 and the flask method is applicable when the log Pow value falls within range -2 to 4. Considering QSAR estimations on known constituents of the substance (EpiSuite estimations of logKow are between -5,12 and 0,16), we anticipate that a part of the constituents of the substance will be outside the applicability domain of the standard methods.

Therefore conventional standard tests for partition coefficient endpoint intended for single substances are considered as not appropriate for such complex substance. That is why the test has been waived for the registered substance.

 

Even if only a range of logKow on a limited part of the composition of the substance, not representative of the whole substance can be calculated using QSAR estimations, calculated values were provided on known constituents of the substance, when possible.

However it is considered that no range logKow representative of the whole test item can be determined (neither with an experimental approach nor with an estimation approach).