Registration Dossier

Physical & Chemical properties

Water solubility

Currently viewing:

Administrative data

Link to relevant study record(s)

Reference
Endpoint:
water solubility
Remarks:
New water solubility study and review / conclusions regarding existing data
Type of information:
other: Experimental study and review of existing water solubility data points
Adequacy of study:
key study
Study period:
30-01-2017 - 10-02-2017
Reliability:
2 (reliable with restrictions)
Justification for type of information:
Despite having multiple water solubility endpoints and additional test was requested by the authorities. This report reviews this existing and newly generated data with the aim of clarifying the most suitable water solubility value for use in the registration dossier.
Qualifier:
according to
Guideline:
OECD Guideline 105 (Water Solubility)
Version / remarks:
Slow stir conditions were applied due to the varying data high speed stirring has caused in previous tests.
Deviations:
yes
Remarks:
see "principles of method"
Principles of method if other than guideline:
Additional extractions took place to better understand substance behaviour. Stirring time was also extended and final samples were analyzed before and after an extended rest period to demonstrate a disolved not dispersed material.
GLP compliance:
no
Remarks:
Due to the endpoint already containing multiple GLP studies and due to this study following an adapted version of the guideline as well as containing a review of existing data a GLP study was not conducted.
Other quality assurance:
other: Internal departmental control of raw data and correctness of reporting only.
Type of method:
other: Slow stir method as reccommended by authorities
Specific details on test material used for the study:
EC name: Di-tert-butyl 3,3,5-trimethylcyclohexylidene diperoxide
Trade name: Trigonox 29
CAS: 6731-36-8
Batch/Lot no.:1504447036
Purity: 93.4%
Composition: CoA included in report
Appearance: Clear colorless liquid
Expiry date: March 7, 2025
Storage: 10°C
Key result
Water solubility:
93 µg/L
Conc. based on:
test mat.
Remarks:
measured concentration.
Temp.:
20 °C
pH:
>= 7 - <= 8
Details on results:
Preliminary test
In the preliminary test samples were taken after 4 hours and analyzed before and after centrifugation, as carried out in existing GLP data. Existing data indicated maximum solubility was reached within 4 hours. The results generated in the definitive test do not support this and indicate the maximum water solubility is reached after 10 days.

Definitive test
In the definitive test, samples were taken 4 hours, 1, 2, 3, 4, 7, 9 and 10 days after addition of the test substance.
In the definitive test the determined result for water solubility of the test substance was an average value of two samples taken after 10 days (10 days shaking and 24 hours settle period). The average result was calculated to be 93 µg/L

Quality Criteria

The following quality criteria were met:
• The test temperature during the study was measured to be within the criteria, 20.0 ± 1.0°C
• The water concentration determined in at least 2 subsequent time points differed not more than 15% during the final definitive test.
• Analytical quality criteria demonstrated that the method used to have been sufficiently robust for the measurements made.


Analytical procedure

 

Introduction

A method is described to determine the concentration of Di-tert-butyl 3,3,5-trimethylcyclohexylidene diperoxide.Procedures and instrumentation are based on Gas Chromatography (GC) with MS detection.

 

Analytical procedure

The following conditions were found to be suitable for the determination of the test substance. The conditions are given in the table below.

 

Table I: Parameters for GC-MS 

Parameter

Setting

Instrument

Thermo Scientific 1310 GC

Ionization

EI-positive

Column:

DB1, 30m x 0.25mm; 0.1µm

Sourcetemperature

250 °C

Interfacetemperature

200 °C

Injector temperature

100 °C

Carrier flow

2.5 mL/min.

Injection volume:

1 µL

Injection

Splitless

Liner

Glass inlet liner with a plug of silanized glass wool

Oven program

40 °C (2 min.)→25°C/min→ 150°C (2 min.)

 

Preparation standard solutions

For preparation of the calibration standards, a stock solution of the test substance was made in hexane. From this stock solution, the calibration standards were made in hexane, in the concentration range 15 – 500 µg/L (n≥5). The GC-MS system with the settings described above was used for analysis.

 

Preparation test samples

Water samples of 10 mL were transferred to centrifuge tubes and extracted by adding 2 mL hexane and shaking intensively for 1 min. After approximately 10 minutes (to allow for good phase separation) the hexane layer was analyzed.

 

Calculation of concentrations

Quantification was done by measurement of peak areas. The concentrations of the test substance in the samples were calculated from the relation between concentration of standards (Cs) and peak area (PAs)obtained with quadratic regression analysis.

Method validation

Regression

The detector’s response was checked by analyzing calibration standards of the test substance and plotting a calibration graph of peak area versus concentration. Most calibration curves were found to be quadratic in the range of 15 – 500 µg/L test substance, R squared was >0.998. An overview of the regression data of the calibration curves and accuracy of the calibration standards are displayed in Table II. Figure 1 shows an example of the calibration curves. In Figure 2 and 3 examples of GC-MS chromatograms are displayed.

 

Table II: Regression and accuracy data

Date

Calibration curve

R2

Accuracy

30-01-2017

y = 18.446x2+ 1661.4x - 7444.9

0.999

93.7 – 103.5*

01-02-2017

y = 9.3768x2+ 3754.3x - 50678

0.998

89.0 – 101.4*

06-02-2017

y = 27.531x2+ 11726x - 24695

0.999

90.5 – 103.5

10-02-2017

y = 21476x - 62110

0.999

96.3 – 105.5

13-02-2017

y = 2.914x2+3671.3x - 81756

0.998

92.9 – 108.8*

*= one calibration standard did not meet the set criteria and therefore excluded from the calibration curve

Accuracy

The accuracy of the calibration curves was determined as the recovery of the calibration standard concentrations, calculated using the calibration curve, compared to the nominal standard concentrations. As shown in Table II all accuracies fell between 98.9 – 104.4% of the nominal concentrations which is within the limits required, i.e. 80 - 120%.

 

LOQ

The Limit of Quantification (LOQ) was set at the lowest calibration standard that complied with the criteria set, i.e.30 µg/L. However lower calibration standards could be analyzed.

 

System stability

Control standards with a nominal concentration of 150 µg/L were analyzed during and at the end of each sampling series. The deviation of the measured concentration of control standard and the calibration standard was calculated to be within 10%, the required limit. The results of the control standards are given in Table V.

 

Table V: Results of control standards

Analysis

date

Conc. Calibration
standard

µg/L

Conc.
Control
standard

µg/L

Deviation from
calibration standard

%

01-02-2017

152.1

161.7

6.3

06-02-2017

148.2

157.1

6.0

10-02-2017

154.3

147.5

-4.4

13-02-2017

144.3

144.0

-0.2

 

Extraction recovery

To determine the extraction efficiency; four samples were spiked at a concentration of 10 µg/L in water, a 10 mL sample was transferred to a centrifuge tube and subsequently 2 mL of n-hexane was added. Test substance was extracted by shaking intensively for approximately 1 minute and subsequently left standing for 10 minutes for good phase separation. Hexane layer was subjected to analysis. Extraction was considered to be well enough, that no correction for extraction efficiency was necessary.

 

Table VI: Recovery extraction

Sample (extraction)

Area

(counts)

Recovery (%)

Recovery sample (no extraction).

63351

n.a.

Centrifuge tube I

60326

95.2

Centrifuge tube II

61776

97.5

Glass centrifuge tube I

59637

94.1

Glass centrifuge tube I

60756

95.9

Average

 

95.7

 

Analytical quality criteria

Parameter

Limit

Determined

Regression (R2)

≥ 0.99

0.998

Extraction (% recovery)

80 – 120

95.7

Accuracy (%)

80 – 120

89.0 – 118.8

LOQ system (µg/L)

 

15 *

LOQmethod(µg/L)

 

3

System stability (deviation % from calculated)

≤ 10

-4.4 – 6.3

* The LOQ of 15 µg/L was sufficient to determine the concentration of the test substance during the study. The test substance was extracted from the water phase by hexane (introducing a fortification factor of five), therefore the test substance could be determined down to 3 µg/L in water. Except blank and control samples, all measured water concentrations were above 3 µg/L.   

 

Conclusions:
In conclusion all of the existing studies demonstrate that the material in question is of very low solubility with all data supporting a solubility of <0.6 mg/L. In practice maintaining a stable stock solution at this level was not possible in multiple toxicity tests that lead to doubts regarding this value. New GLP studies were therefore conducted as up-scaled slow stir solubility tests by Mead and Mullee (2013). This data showed results of between 20 and 100 µg/L. Due to this variability an emulsion was still presumed and subsequent centrifugation gave figures of <2-4µg/L. Considering all of this data stirring studies were conducted by Kean 2015 (Ref 6) demonstrating the influence a high material loading and fast stirring can have when the test material is above the water solubility. At lower loading rates and stirring speeds values of between 50-90 µg/L were found.

The current study has taken all of these findings into account and has reproduced results found in most of the previous studies with the exception of Datta 2013 in which a high value of <0.6mg/L was reported. It is worthy of note that when considering all of the values found in Datta 2013 (53, 570, and 98 µg/L) that most of the values were concordant with the findings of the other water solubility studies. With the current knowledge that the material is hydrolytically stable and knowing that high loadings can result in dispersion of the test material. The highest peak found in this study is hypothesized to have been caused by fast stirring and the remaining two reported figures are likely a more accurate indication of the true solubility, which is supported by the majority of the test data.

This study demonstrates that what was thought to be emulsion loss after centrifugation can be attributed mainly to absorption to the centrifuge tube. After a 24 hour rest period (no stirring) identical dissolved concentrations were observed indicating a stable truly dissolved concentration.In addition this study demonstrates that with a slow stir a longer perod is needed before the solubility curve flattens and a definitive figure can be estimated.

Considering that the much GLP and non GLP data as well as some modelled results supports the findings of this study and considering that the anomalies of the existing data have been largely explained, a definitive solubility limit of 93 µg/L is concluded for 3,3,5-trimethylcyclohexylidene diperoxide
Executive summary:

All of the existing GLP and non GLP data was reviewed and combined with the findings of a new study based as close as possible on an appropriate study guideline with the approach requested by the authorites. When considering the total data package much of the existing data can be explained and there is also considerable overlap in the exisiting measured data that supports this new value. The newly determined value demonstrated in this study supported by the weight of evidence in the existing data allow a definative value of sufficient reliability to be concluded. Old data demonstrated solubility in the correct range but variation in results made a final value difficult to conclude. This study has demonstrated that values with sufficient reproducibility can be achieved when secure slow stir methodology is applied.

Description of key information

A large amount of solubility data (GLP and non GLP and models) exists. This has arisen due to the difficult to test properties the substance possesses. See additional information. On request of the competant authorites another solubility study was conducted. The aim being to explain the historical variation in data and come to a single usable water solubility endpoint whilst staying as true as possible to the test guideline.

Key value for chemical safety assessment

Water solubility:
93 µg/L
at the temperature of:
20 °C

Additional information

In conclusion all of the existing studies demonstrate that the material in question is of very low solubility with all data supporting a solubility of <0.6 mg/L. In practice maintaining a stable stock solution at this level was not possible in multiple toxicity tests that lead to doubts regarding this value. New GLP studies were therefore conducted as up-scaled slow stir solubility tests by Mead and Mullee (2013). This data showed results of between 20 and 100 µg/L. Due to this variability an emulsion was still presumed and subsequent centrifugation gave figures of <2-4µg/L. Considering all of this data stirring studies were conducted by Kean 2015 (Ref 6) demonstrating the influence a high material loading and fast stirring can have when the test material is above the water solubility. At lower loading rates and stirring speeds values of between 50-90 µg/L were found.

 

The current key study Van Dam 2017 study has taken all of these findings into account and has reproduced results found in most of the previous studies with the exception of Datta 2013 in which a high value of <0.6mg/L was reported. It is worthy of note that when considering all of the values found in Datta 2013 (53, 570, and 98 µg/L) that most of the values were concordant with the findings of the other water solubility studies. With the current knowledge that the material is hydrolytically stable and knowing that high loadings can result in dispersion of the test material. The highest peak found in this study is hypothesized to have been caused by fast stirring and the remaining two reported figures are likely a more accurate indication of the true solubility.

 

This study demonstrates that what was thought to be emulsion loss after centrifugation can be attributed to absorption to the centrifuge tube. After a 24 hour rest period (no stirring) identical dissolved concentrations were observed indicating a stable truly dissolved concentration.

 

Considering that the weight of evidence of both GLP and non GLP tests as well as some calculated data supports the findings of this study and considering that the anomalies of the existing data have been largely explained, a

definitive solubility limit of 93 µg/L is concluded for3,3,5-trimethylcyclohexylidene diperoxide