Registration Dossier

Diss Factsheets

Environmental fate & pathways


Currently viewing:

Administrative data

Link to relevant study record(s)

Type of information:
experimental study
Adequacy of study:
key study
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
comparable to guideline study with acceptable restrictions
The study was not conducted in compliance with GLP.
equivalent or similar to guideline
OECD Guideline 111 (Hydrolysis as a Function of pH)
GLP compliance:
Analytical monitoring:
Details on sampling:
Individual kinetic experiments were conducted in one of two modes, depending on the expected half-life of the parent substance. The first mode was used for half-life < ca. 45 min (pH <5 or >8) and involved a separate reaction aliquot for each unique reaction time to be sampled. Immediately prior to analysis of a particular sample, the hydrolysis reaction was quenched by rapid adjustment to pH 6.7±0.5 by addition of acid or base. The final pH of a subset of samples was verified.

The second mode was used for pH 5-8 and involved up to four staggered reaction solutions. The separate reactions were repeatedly sampled in alternating fashion over several hours to collect hydrolysis data spanning approximately 3 half-lives of the parent substance.

On average, data were collected at 12 discrete times for each kinetic experiment.
Buffer solutions of known pH and concentration were prepared by titration of 1M glacial acetic acid or tris(hydroxymethyl)-aminomethane (99.9%) solution with 1M sodium hydroxide (99.998%) or hydrochloric acid solution (37 wt%), respectively. A constant ionic strength of 0.30 M was maintained by addition of an appropriate volume of 2M sodium chloride solution. Buffer solutions were made to known final volume in polypropylene volumetric flasks with deionized water (>18 MO cm). If necessary, final pH adjustments were made by dropwise addition of sodium hydroxide or hydrochloric acid using a calibrated pH meter. Prior to use all buffer solutions were sparged with argon for at least 15 min. As the test material was capable of altering solution pH through the basic primary amine, the pH reported for a given experiment was taken as that which was measured following silane addition.
Details on test conditions:
5*10-2 M stock solutions of the test material in acetonitrile were prepared in a nitrogen-purged glove bag and stored in 22 ml plastic vials having septum lined open-top caps. When not in use, the vials were stored in a secondary airtight container filled with Drierite.
Kinetic experiments were conducted over the pH range 4.7-9.0 with buffer concentrations varying from 20 to 200 mM for acetic acid/sodium acetate and 20 to 300 mM for Tris-HCl/Tris. As the hydrolysis reactions were expected to show general base catalysis, buffer concentrations were selected to give particular concentrations of the conjugate base over the range of pH covered by each buffer.
Experiments were conducted at 9.6 to 34.8°C, thermostatted to ±0.1°C.
The starting concentration was not varied as previous studies have demonstrated that the reaction rate in dilute aqueous solution is first order in silane concentration.
The reactions employed initial silane concentrations of 5*10-4 M (1 part silane stock solution + 100 parts buffer solution).
Statistical methods:
The changes in peak area associated with each of the four components of the reaction mixture (parent, intermediates and product) over time contain kinetic information pertaining to the rates of the three consecutive hydrolysis reactions. Unconstrained nonlinear regression analysis was used to obtain estimates for the rate constants k1, k2 and k3 by simultaneously fitting the dataset to a kinetic model based on pseudo-first order kinetics for each reaction. A parameter was added to account for the varying sensitivity of the instrument to each component. The initial silane concentration was treated as a fixed parameter.

The analysis was performed using Origin 6.0 data analysis software, which employs the Levenburg-Marquardt minimization algorithm. The software varied the software parameters iteratively. The tolerance was set at 0.01%. Convergence was typically reached in 3-4 iterations, although in one case 8 interations were required.
Preliminary study:
No preliminary study was carried out.
Transformation products:
Details on hydrolysis and appearance of transformation product(s):
Trialkoxysilanes undergo hydrolysis in dilute aqueous solution via a series of consecutive pseudo first order reactions:
RSi(OR')3 ¿ RSi(OR')2(OH) ¿ RSi(OR')(OH)2 ¿ RSi(OH)3
One mole of alcohol (in this case ethanol) is released at each hydrolysis step.

The observed changes over time in the chromatographic peaks areas, together with the assumption that the components elute in order of decreasing hydroxyl substitution (polarity), served as a basis for the peak assignments.

The signal associated with the parent silane followed a simple exponential decrease over time. The peak corresponding to the first hydrolysis product (transformation product #1) appeared early in the reaction, closely followed by the concurrent appearance of the second intermediate (transformation product #2) and the silanetriol product (transformation product #3). The peaks of the intermediate products reached maxima part way through the hydrolysis process, followed by a gradual decrease that continued until completed hydrolysis was reached.

Over the pH range investigated, the intermediate silanol products (the mono- and di-ol) were observed to hydrolyze more rapidly than the original tri-alkoxysilane.  Consequently, these breakdown products can be considered transient. 
Key result
24.7 °C
0.8 h
(pseudo-)first order (= half-life)
Key result
24.7 °C
8.5 h
(pseudo-)first order (= half-life)
Key result
24.7 °C
0.15 h
(pseudo-)first order (= half-life)
Details on results:
The minimum hydrolysis rate at 24.7°C occurred at pH 6.6, with a half-life of 620 min (10.3 h). Extrapolating to 0°C, the maximum possible half-life was estimated as 150 h at pH 6.9. Under all conditions, it was observed that k1
Non-linear regression analysis was applied to the data describing changes in component peak area as a function of reaction time to obtain estimates of the consecutive hydrolysis rate constants. Very good agreement between experimental data and fitted curves was observed for all four components. Appropriate statistical tests indicated that the data adhere to the chosen kinetic model.

Nominal initial concentration = 5x10-4 M (~110 mg/L). The concentration was not directly measured; rate constants were extracted  from changes in analytical response for each component.

Table 1.  Observed rate constants for hydrolysis reactions of APTES



T / °C

[buffer], mM

k1* 104. s-1(uncertainty)

k2* 104. s-1


k3* 104. s-1






3.99 (0.18)

12.7 (3.1)

31.3 (12.7)





6.76 (0.19)

27.7 (5.2)

70.9 (49.7)





3.24 (0.18)

10.8 (3.3)

18.7 (8.1)





3.38 (0.09)

13.9 (2.2)

39.7 (13.4)





0.434 (0.017)

1.67 (0.31)

2.95 (0.78)





0.737 (0.020)

2.63 (0.38)

12.5 (5.4)





0.240 (0.003)

1.67 (0.06)






0.276 (0.013)

1.49 (0.53)






1.45 (0.05)

13.1 (1.5)






1.53 (0.04)

15.4 (1.4)






6.53 (0.26)

60.1 (6.9)






15.4 (0.7)

78.7 (7.6)






13.6 (0.6)

89.9 (8.5)






1.87 (0.11)

5.23 (1.39)

14.4 (7.2)





3.99 (0.18)

12.7 (3.1)

31.1 (12.7)





7.75 (0.21)

22.6 (2.3)

60.9 (11.3)





2.61 (0.14)

15.9 (3.0)






13.6 (0.6)

89.9 (8.5)






38.7 (1.9)

239 (27)


Effect of pH on the hydrolysis kinetics

For an acid-base catalysed hydrolysis reaction in aqueous buffered solution, the measured rate constant kobs is described by the general equation:

kobs = k0+ kH3O+[H3O+] + kOH-[OH-] + ka[acid] + kb[base]

where k0 refers to the spontaneous reaction with water and the latter two terms provide for possible catalysis by the conjugate acid and base of the particular buffer. kH3O+ and kOH- are the acid and base catalysed rate constants.

In order to understand the effect of pH on the stepwise hydrolysis of the test substance, a series of kinetic runs were conducted over a range of pH using acetate and Tris buffers of varying concentration. Varying the concentration of the buffer allows its catalytic effect to be elucidated; this is mainly interesting in terms of it impact on the investigation of pH effects. Non-linear regression analysis was used (as discussed in the methods section) to determine values of k1 and k2 and, if possible, k3 for each experiment corresponding to a particular pH and buffer composition. The results are shown in Table 1 above.

Multiple linear regression analysis was then used to model the effect of hydronium or hydroxide ion concentratoin and buffer concentration on the observed rates of hydrolysis. The results are shown in Table 2 (for pH 4.7 -5.9, dominated by hydronium ion catalysis) and Table 3 (for pH , dominated by hydroxide ion catalysis).

Table 2: Results of multiple linear regression analysis of APTES kinetic experiments in the pH range 4.7 -5.8 at 24.7°C

 Variable (units)  k1 (significance, P) k2 (significance, P)  k3 (significance, P)
 [H3O+] (M-1 s-1)  23.1 (0.0012)  71.1 (0.0020) 132 (0.0143) 
 [HOAc] (M-1 s-1)  2.43E-03 (0.0124)  1.59E-02 (0.0025) 5.19E-02 (0.0035) 
 Intercept (s-1)  5.5E-06 (0.7815)  8.7E-06 (0.9054) 2.4E-04 (0.4051) 
 Adjusted r2  0.9891 0.9909   0.9812

Table 3: Results of multiple linear regression analysis of APTES kinetic experiments in the pH range 7.0 -9.0 at 24.7°C

 Variable (units)  k1 (significance, P) k2 a (significance, P)  k3 b (significance, P)
 [OH-] (M-1s-1)  125 (0.0000)  1130 (0.0005) -
 [Tris] (M-1s-1) 3.24E-04 (0.0051) 4.75E-03 (0.0547)  -
 Intercept (s-1)  7.8E-06 (0.0702) 5.9E-06 (0.9188)  -
 Adjusted r2  0.9999 0.9991

a pH 9.0 data not included in the model for k2 due to poor initial fit with a large standardized residual for this observation.

b Final hydrolysis step was too rapid to measure quantitatively

It can be seen from the above tables that [H3O+] and [OH-] are very significant (P<0.01) in all cases. The coefficients are the second order catalytic constants, kH3O+ and kOH-, for the first, second and (for kH3O+) third hydrolysis steps. At the higher pH, the third hydrolysis step was too rapid to measure quantitatively in all cases. The adjusted r2 for the final model is >0.98 in all cases.

The buffer concentrations, described by [HOAc] and [Tris] were found to be significant, indicating that buffer catalysis is occuring.

kH3O+ and kOH- both increase for successive hydrolysis steps, with kOH- increasing to a much greater extent.

A statistically significant intercept term (P<0.1) was obtained for the intercept of k1 in the higher pH experiments. This represents k0 for the first hydrolysis step.

There was good agreement between measure values of k1, k2 and k3 and those predicted based on the linear regression analyses from the two catalytic regimes. This indicates that the model results accurately represent the experimental data and that the chosen variables account for most of the variance in the data.

Effect of temperature on the hydrolysis kinetics

To determine the effect of temperature on the rate of hydrolysis of the parent silane and the intermediate hydrolysis products, additional kinetic runs were made at 10 and 35°C and pH 4.7 and 9.0, using the lowest buffer concentrations from the respective 25°C runs. Under these conditions, the change in the observed rate constant with temperature should relate predominantly to the specific acid and base catalysed mechanisms. The results are shown in Table 1 above. These rate constants were used to construct a series of three-point Arrhenius plots from which pre-exponential factors (A) and activation energies (Ea) were estimated for the specific acid and base catalysed reactions. The results are given in Table 4 below.

Table 4: Arrhenius parameters for the Hydronium and Hydroxide Ion Catalyzed Hydrolysis Reactions of APTES

   A, s-1 Ea, kJ/mol  r2
 k1 H3O+  5.00E03  40.3  0.9900
 k2 H3O+  2.80E04  41.8  0.9997
 k3 H3O+  4.72E04  40.7  0.9901
 k1 OH-  6.09E10  77.8  0.9999
 k2 OH-  5.00E11  78.5  0.9996
 k3 OH-  -  -  -

For each plot r2 was found to exceed 0.99, suggesting that a single dominant reaction pathway (ie specific acid or specific base catalysis) is being observed at each extreme of pH. There is not enough data to draw conclusions on the significance of the variation in Ea among the stepwise reactions, although they do appear to be very similar. However, it is clear that the activation energies are approximately a factor of 2 larger for the hydroxide catalysed reaction.

The Arrhenius parameters can be used with the previously discussed catalytic constants to predict t1/2 for the disappearance of the test substance as a function of pH and temperature at zero buffer concentration. This is shown in Figure 5 (attached) for the three temperatures examined during the study. It should be noted that k0 is only included in the 25°C curve as the temperature dependence of this reaction pathway has not been determined. Therefore, the other curves represent conservative estimates of half-life particularly in the pH region where the rate is near minimum.

A hydrolysis half life for disappearance of parent substance of 8.5 h at pH 7 and 24.7°C was determined in a reliable study conducted according to an appropriate test protocol but not conducted according to GLP. The subsequent hydrolysis steps and the temperature and pH dependence of the hydrolysis kinetics were also investigated.
Executive summary:

The kinetics of the hydrolysis reactions of 3-aminopropyl-triethoxysilane in dilute aqueous solution were characterized over a range of environmentally relevant pH and temperature. The results are consistent with a series of consecutive pseudo-first order reactions having an increasing rate for each subsequent hydrolysis step (k1<k2<k3). Reaction rates are strongly influenced by pH, with catalysis by hydroxide ion being 5 times more effective than hydronium ion at promoting hydrolysis of the parent trialkoxysilane; this discrepancy increases for the subsequent reactions leading to formation of the silanetriol. In addition, the contribution of the solvent catalyzed reaction, k0, is significant to the overall rate of hydrolysis of ATPES extrapolated to zero buffer concentration. Given that the first hydrolysis reaction, k1, is rate limiting and using 10 half-lives as the definition of "complete", this study indicates that the trialkoxysilane will be exhaustively hydrolyzed to the silanetriol in =4.5 days at 25°C.

Description of key information

Hydrolysis half-life: 0.8 hours at pH 5, 8.5 hours at pH 7 and 0.15 hours at pH 9 and 24.7°C (OECD 111)

Key value for chemical safety assessment

Half-life for hydrolysis:
8.5 h
at the temperature of:
24.7 °C

Additional information

The hydrolysis half-lives of 3-aminopropyl(triethoxy)silane have been measured using a method similar to OECD 111 to be 0.8 hours at pH 5, 8.5 hours at pH 7, and 0.15 hours at pH 9 and 24.7°C. The quoted results relate to disappearance of parent substance. The result is considered to be reliable and is used as key study. The measured result is supported by predicted hydrolysis half-lives of 0.4 h at pH 4 and 0.1 h at pH 9 and 20 -25°C using a validated QSAR estimation method. In Beari et al (2001), a hydrolysis half-life of <1 hour at pH 6 was reported for the substance.

In the key study, the parent substance, two intermediates and the final hydrolysis product were observed and quantified. The substance was found to hydrolyse according to the following reaction scheme:

RSi(OEt)3 ¿ RSi(OEt)2(OH) ¿ RSi(OEt)(OH)2 ¿ RSi(OH)3

One mole of ethanol is released at each hydrolysis step.

Estimates for the rate constants for the first, second and third reaction steps (k1, k2 and k3 respectively) were obtained; over the range of pH and temperature investigated the intermediate silanol products were found to hydrolyse more rapidly than the original trialkoxysilane. Consequently, these intermediates can be considered transient. The rate constant values obtained for the hydroxonium ion catalysed reaction are: k1 = 23.1 M-1s-1, k2 = 71.1 M-1s-1, k3 = 132 M-1s-1. The values obtained for the hydroxide ion catalysed reaction are: k1 = 125 M-1s-1, k2 = 1130 M-1s-1, k3 = not measured (as the reaction was too fast).

The concentration of each hydrolysis product has been plotted against time (expressed as number of half-lives for degradation of parent substance) for the acid catalysed reaction. The graph is attached to this endpoint summary (Figure 1). The parent compound dominates during the time span <1 half-life of the parent compound. The final hydrolysis product starts to dominate after approximately 1.5 half-lives of the parent compound have passed. After approximately 4 half-lives, the final hydrolysis product represents 90% of the compound present. Under basic or neutral conditions, the concentrations of the intermediate hydrolysis products reach lower maxima and begin to decrease more quickly because the ratios of k2 and k3 to k1 are greater than for the acid catalysed reaction.

The pH dependence of the hydrolysis kinetics was investigated, by carrying out experiments at a range of pH values between 4.7 and 9.0. The reaction rate was found to be slowest at pH 6.5 - 7 and increase as the pH was raised or lowered. Estimates of kH3O+ (the hydroxonium ion catalysed rate constant) for the first, second and third reaction steps; kOH- (the hydroxide ion catalysed rate constant) for the first and second reaction steps; and k0 (the solvent catalysed rate constant) for the first reaction step were obtained. These can be used to estimate the reaction rate at any pH for zero buffer concentration.

The temperature dependence of the hydrolysis kinetic was investigated by carrying out experiments at 10, 25 and 35°C. The reaction rate was found to increase with temperature, to a greater extent for the hydroxide catalysed reaction than the hydronium ion catalysed reaction. Arrhenius parameters were calculated for the hydronium and hydroxide calculated reactions and these can be used to estimate the reaction rate at any temperature.

The authors of this summary have used the rate constants and Arrhenius parameters quoted in the study report to calculate the half-lives in the table below for a range of relevant temperature and pH values.

Table1: Estimated half-lives at a range of temperature and pH values.

 Temperature / °C  pH Relevance  Half-life 
 20  7  Ecotoxicology studies  11 h
 20  8  Ecotoxicology studies  2.6 h
 20  9  Ecotoxicology studies  0.3 h
 37.5  7  Toxicology, lungs and blood  3.4 h
 37.5  5.5  Toxicology, skin  1.4 h
 37.5  2  Toxicology, stomach  5 s a
 35  4.7  Boundary of experimental results  15 mins

a The calculated value is 2 s. However, it is not appropriate or necessary to attempt to predict accurately when the half-life is less than 5-10 seconds. Therefore, the value is reported as 5 s.

The measurements in the study were taken at pH 4.7 - 9.0 and 10 - 35°C; therefore, the estimates at 37.5°C and pH 2 represent extrapolations. The difference between 35°C and 37.5°C is very small, so this extrapolation is not considered to add significant uncertainty to the results. The estimated result at pH 2 does represent an extrapolation significantly outside the range of pH values studied (4.7 - 9.0). However, the study results are strongly supportive of a hydroxonium ion catalysed reaction being dominant in the acid pH range (4.7-5.9) and the reaction rate increasing as the hydroxonium ion concentration increases (pH decreases). Therefore, it is extremely unlikely that the hydrolysis at pH 2 is slower than that at pH 4.7.

The final hydrolysis products are 3-aminopropylsilanetriol (1 mole) and ethanol (3 moles).

The hydrolysis of other substances used for read-across in other sections are discussed below.

Hydrolysis of the read-across substance triethoxy(3-isocyanatopropyl)silane(CAS 24801-88-5)

Data for the substance triethoxy(3-isocyanatopropyl)silane (CAS 24801-88-5) are read-across to the submission substance 3-aminopropyl(triethoxy)silane for the toxicity to micoorganisms endpoint. The silanol hydrolysis products and the rate of hydrolysis of the two substances are relevant to this read-across, as discussed in the appropriate sections.


For triethoxy(3-isocyanatopropyl)silane, the isocyanate group is expected to hydrolyse very rapidly to form 3-aminopropyl(triethoxy)silane (CAS 919-30-2, the registered substance) as an intermediate hydrolysis product and carbon dioxide. The hydrolysis half-lives of 3-aminopropyl(triethoxy)silane have been measured in accordance with OECD 111 to be 0.8 h at pH 5, 8.5 h at pH 7, and 0.15 h at pH 9 and 24.7°C. In addition, hydrolysis half-lives of 0.4 h at pH 4 and 0.1 h at pH 9 were predicted for 3-aminopropyl(triethoxy)silane using a validated QSAR estimation method.


The ultimate products of the hydrolysis reaction of triethoxy(3-isocyanatopropyl)silane under dilute conditions are 3-aminopropylsilanetriol, ethanol and carbon dioxide. 

Hydrolysis of the read-across substance 3-(trimethoxysilyl)propyl isocyanate (CAS Number: 15396-00-6)

Data for the substance 3-(trimethoxysilyl)propyl isocyanate (CAS 15396-00-6) are read-across to the submission substance 3-aminopropyl(triethoxy)silane for the short-term toxicity to fish and short-term toxicity to aquatic invertebrates endpoints. The silanol hydrolysis product and the rate of hydrolysis of the two substances are relevant to this read-across, as discussed in the appropriate sections.


For 3-(trimethoxysilyl)propyl isocyanate, the isocyanate group is expected to hydrolyse very rapidly to form 3-aminopropyltrimethoxysilane (CAS 13822-56-6) as an intermediate hydrolysis product and carbon dioxide. The hydrolysis half-lives of 3-aminopropyltrimethoxysilane have been predicted using a validated QSAR estimation method to be 0.2 h at pH 4, 0.3 h at pH 5, 2.6 h at pH 7, and 0.1 h at pH 9 and 20-25°C. Also, hydrolysis half-life of (3-isocyanatopropyl)trimethoxysilane was found to be much less than 2.4 h at pH 4, pH 7 and pH 9 and 50°C. Therefore, the half-lives at 25°C and pH 4, pH 7 and pH 9 were estimated to be <1 day, which is consistent with the expected behaviour. The study was conducted according to OECD Guideline 111 (1981) and in compliance with GLP; the results are considered reliable. However, according to the most recent version of the test guideline, a higher-tier test should be carried out if a substance is found to be unstable in the preliminary test and this was not done in this study. However, in view of the extreme instability of the isocyanate group, in practice it may not be technically feasible to do so and obtaining a more accurate half-life value would be of limited use.

The final hydrolysis products are 3-aminopropylsilanetriol and methanol.


Hydrolysis of the read-across substance N-(3-(trimethoxysilyl)propyl)ethylenediamine (CAS 1760-24-3)

Data for the substance N-(3-(trimethoxysilyl)propyl)ethylenediamine (CAS 1760-24-3) are read-across to the submission substance, 3-aminopropyl(triethoxy)silane for the long-term toxicity to aquatic invertebrates endpoint. The hydrolysis half-life and formation of similar hydrolysis products are relevant to this read-across as discussed in the appropriate sections.

For N-(3-(trimethoxysilyl)propyl)ethylenediamine, measured hydrolysis half-life value of 0.1 h at pH 4 and 24.7°C, 0.025 h at pH 7 and 24.7°C, and 0.32 h at pH 5 and 24.7°C was determined for the substance in accordance with OECD 111. The result is considered to be reliable. At pH >7, the half-life became too rapid (<90 s) to measure using the methodology of this study.

In other secondary sources to which reliability could not be assigned, hydrolysis half-life of 0.016 h at pH 7 and 24.7°C was reported. Also, a hydrolysis half-life of 24.1 h at 25°C was reported, information on the pH was not stated.

The hydrolysis products are N-(3-(trihydroxysilyl)propyl)ethylenediamine and methanol.

The table below summarises all relevant hydrolysis half-lives used in this chemical safety assessment:

 Table: Summary of relevant hydrolysis half-lives



Half-lives at 20-25°C

Half-lives at 37.5°C




pH 4

pH 5

pH 7

pH 9

pH 2

pH 5.5

pH 7

2,2,4 (or 2,4,4)-Trimethylhexane-1,6-diisocyanate


3.81 minute*


4.88 minute*

1.93 minute*




3-(Trimethoxysilyl)propyl isocyanate


0.2 hour**

0.3 hour**

2.8 hour**

0.1 hour**






0.2 hour**

0.3 hour**

2.6 hour**

0.1 hour**






0.4 hour**

0.8 hour*

8.5 hour*

0.15 hour*

5 seconds

0.15 – 3 hour

3 hour



0.1 hour*

0.32 hour*

0.025 hour*

too rapid to measure






-not calculated

Categories Display