Registration Dossier

Environmental fate & pathways

Bioaccumulation: aquatic / sediment

Currently viewing:

Administrative data

Link to relevant study record(s)

Description of key information

C16 amine is a model compound for the primary alkyl amines. Therefore it is proposed to use for the primary alkyl amines a BCF of 173 L/kg as estimated for C16 by the ADME Model of Arnot & Gobas (2003) on basis of a Weight of Evidence.

Key value for chemical safety assessment

BCF (aquatic species):
173 L/kg ww

Additional information

Introduction

As Primary alkyl amines are readily biodegradable it can be expected that metabolism will occur in aquatic species like fish. This is confirmed by in vitro measurement of the metabolic rate of 1-Hexadecanamine (C16 amine) which was selected as model substance for the Primary alkyl amines. This means that establishing of the Bioconcentration factorBCFshould address Adsorption, Distribution, Metabolism and Excretion (so called ADME process).

Primary alkyl amines are cationic surfactants which sorb strongly to solid phases by van der Waals and ionic interactions (e.g. ion pair formation, cation exchange etc). This makes it extremely difficult or even impossible to ensure a constant concentration of the test item in a flow through system because of sorption but also by biodegradation as the test system is not sterile. Additionally as fish mucous is negatively charged the cationic surfactant is sorbing on the fish surface as in the test water used for aBCFtest no dissolved organice carbon (DOC) or suspended matter is present as it would under environmental conditions.

Usually for the determination of theBCFan OECD 305 BCF test would be carried out as the ADME process would be addressed completely. But due to the inherent properties of cationic surfactants technical issues arise and the measurement is likely to be highly uncertain and the result even not valid. Therefore also estimation procedures for BCF have to be considered and evaluated for use in the context of this assessment under REACH.

 

Attempt to measure theBCFin an OECD 305 Test with 1-Hexadecanamine

Despite a high technical effort a constant flow through with acceptable analytical recoveries could not be established (Akzo Nobel, 2007). In addition due to the negatively charged fish surface continous sorption of C16 amine to the fish mucous took place. Therefore real steady state could not be observed. The calculated BCF taking into account the adsorbed fraction in the mucous gave high values which are unrealistic. Removing the adsorbed C16 amine with methanolic hydrogen chloride resulted in BCF values around 300 L/kg but the validity of this figure cannot be judged. The conclusion is that the conditions proposed in the OECD 305 Guideline do not fit to the intrinsic properties of the n-Primary alkyl amines in general. A test in natural river water which would mimic realistic environmental conditions may be more appropriate but could lead to analytical issues like insufficient sensitivity even when using 14C material.

Conclusion: This preliminary study shows that the conditions described in the OECD 305 Guideline will not be applicable for n-Primary alkyl amines. But alternative conditions e.g. use of river water might create other issues like insufficient analytical sensitivity.

 

Use of a Mechanistic model addressing ADME

Adsorption, Distribution, Metabolism and Excretion Models like the one from Arnot & Gobas (2003) describe the ADME process for a fish with help of mathematical algorithms. When such a model is applied to the unprotonated n-Primary alkyl amines using measured fish metabolic rates low BCF values will be calculated.

 

BCF   =  (1 – LB)  + (kuptake* fdiss/ (kelimin + kegestion+ kgrowth+ kmetabol.))

 

LB                          = Lipid fraction in organism

Kuptake    = uptake rate (estimated by: 1/(0.01 + 1/Kow)* Weight0.4)

fdiss                         = fraction of dissolved substance

kelimin                     = elimination rate (estimated by: kuptake/ LB* Kow)

kegestion                   =  faecal egestion rate (estimated by: 0.02*Weight-0.15* e-0.06T/(5.1*10-8*Kow+2)*0.125

kgrowth                     =  0.0005*Weight-0.2

kmetabol.                              =  measured rate

 

 

Table4.2.1    Parameter values used in the Arnot & Gobas Model for 1-Hexdecanamine

Parameter

Value used in modelling

Remark

Log Kow

6.7

Estimated with US EPA KOWWIN V. 1.67 (USEPA, 2008

L(lipid fraction)

0.2

Standard in model

Weight of fish (kg)

0.438

Av. Fish weight in study for carp metabolic rate (Bernard et al., 2006)

Temperature (deg C)

12

REACH Guidance R.16.4.3.1

ffreely diss

(freely dissolved fraction)

0.2

Estimated from the differences in ecotoxcity measured in tap and in river water

kmetabolism  (1/d)

0.152

Lowest value from in vitro study (Bernhard et al, 2006)

 

Metabolic rates may not only be measured (Perdu-Durand et al, 2006) but could be estimated from Property property estimation programs like US EPA BCFBAF (US EPA, 2008).

 

Table4.2.2    Compilation of Km (Metabolic rate fish) for unprotonated & protonated C16 amine                                                                     

C16 amine
unprotonated
Km (1/d)   C16 amine
protonated
Km (1/d)   Reference
             
10 g fish 0.076 estimated 10 g fish 3.12 estimated US EPA (2008)
100 g fish 0.043 estimated 100 g fish 1.75 estimated
  n.a. n.a. 438 g fish 0.15 in vitro Perdu-Durand (2006)
1 kg fish 0.024 estimated 1 kg fish 0.99 estimated US EPA (2008)
10 kg fish 0.014 estimated 10 kg fish 0.55 estimated

The estimation program predicts much higher metabolism rates for the protonated C16 amine when compared with the unprotonated. The in vitro measured Kmis lower than the comparable values which were estimated.

Table4.2.3gives theBCFvalues calculated with the ADME Model from Arnot & Gobas (2003) using the parameter listed inTable4.2.1before as well as specific data. Log Kowwere taken from Table 1.3 and it was assumed that the metabolic rate for the n-Primary alkyl amines (C12 to C18) is the same.

 

 

 

 

Table4.2.3    BCF calculated with an ADME Model using Parameters listed inTable4.2.1

Chain length n-Primary alkyl amines


BCFusing Log Kow(L/kg)


BCFusing Log Coct/Cwater(L/kg)

C12

162

168

C14

172

173

C16

173

173

C18

174

174

C18= (Oleyl)

173

175

 

The data from Table4.2.3 show that Log Kowdoes not influence theBCFmuch for this partitioning range. The reason is that the high measured metabolic rate controls the BCF output. The BCF range for the C12 to C18 amines is 162 to 174 L/kg with C16 amine 173 L/kg. So 1-Hexadecanmine (C16 amine) is a reasonable representative for the C12 to C18 n-Primary alkyl amines.

Conclusion: This model requires e.g. Kowas input parameter as well as a metabolic rate which can be measured either by in vitro methods (complex) or calculated by property estimation program like US EPA BCFBAF (US EPA, 2008). The adavantage of this model is that it addresses the ADME process for fish but can only cope with neutral molecules.

 

Use of BCF Property estimation programs

BCF property estimation programs like US EPA BCFBAF use correlations between Log KowandBCFe.g.

                       LogBCF= 0.6598 log Kow– 0.333 + correction

 

Such estimation programs address only the absorption process of ADME but not distribution, metabolism and excretion. For rapidly biodegrable substances this may overestimate the trueBCFby far as metabolism is not taken into account. When using the log Kow of 6.7 for the unprotonated and and Log Coct/Cwater of 1.48 for the protonated C16 amine the US EPA BCFBAF model estimates BCF of 500 for the unprotonated and a BCF of 4 for the protonated C16 amine.

Conclusion: Property estimation programs as the US EPA BCFBAF can supply quickly estimates but can address only the partitioning either for the unprotonated or the protonated amines but not for both in an equilibrium described by the pKa.

 

 

 

Use of a Model which can predict the BCF for acids and bases in equilbrium

Fu et al (2009) have published a model which can estimate the BCF of acid and bases as function of the pH. The fraction of the unprotonated amine fn can be calculated by the Henderson-Haselbalch equation 

                        fn = 1 / (1+10i(pKa-pH))   with i = 1 for bases

The apparent Kow for weak electrolytes also called D can be calculated by

                       D = fn* Kow (unprotonated) + fd* Kow (protonated)

Kow(protonated) can be either calculated by

                       Log Kow (protonated) = Log Kow (unprotonated) – 3.5

or the measured Log Coct/Cwater for the protonated can be used.

Fu et al. analyzed available data for strong bases and found the following regression

                       LogBCF       =  0.24 Log D + 0.87

For the C16 amine theBCFcan be estamated as function of pH 4, 7 and 9

              

Table4.2.4    BCFas function of pH for the C16 amine

BCF
pH4 pH7 pH9
C16 amine 43 50 124

 

Conclusion:The model of Fu et al (2009) is the only which can address theBCFof acids and bases as function of the pH but it cannot be judged if cationic surfactants were included in the regression of the model. The model can also not address metabolism in e.g. fish.

 

Weight of Evidence Approach (WoE)

None of the approaches described in this chapter used to derive theBCFof Primary alkyl amines can give a reliable results which addresses the full ADME process especially for fish. Therefore a Weight of Evidence Approach has to be applied.

1) The test design for an OECD 305 test for the measuring of theBCFis not suitable. The result from this preliminary test has an uncertainty which cannot be judged.

2) The ADME model of Arnot & Gobas (2003) can address the ADME process but only for the unprotonated amine. Due to the relatively high metabolic rate from an in vitro measurement lowBCFare predicted based on uptake of the unprotonated C16 amine which is considered as a worst case assumption.

3) Classical BCF estimation methods based on Log Kow, predict higher BCF values for the unprotonated than for the protonated C16 amine

4) The model of Fu et al (2009) is the only model which can address the coexisting protonated and unprotonated C16 amine as function of pH. Unfortunately it does address only the Adsorption of the ADME process and in addition it is not known if it is valid for cationic surfactants.

The most suitable approach to derive a BCF for Primary alkyl amines is the ADME Model of Arnot and Gobas (2003) for the unprotonated C16 amine. Most likely this is conservative when the values are compared with the pH dependend results from Fu et al (2009).

Overall conclusion: 

C16 amine is a model compound for the Primary alkyl amines. Therefore it is proposed to use for the Primary alkyl amines a BCF of 173 L/kg as estimated by the ADME Model of Arnot & Gobas (2003) on basis of a Weight of Evidence.