Trimethoxysilane

Hydrolysis

Link to relevant study record(s)

Reference

Endpoint:

hydrolysis

Type of information:

experimental study

Adequacy of study:

key study

Study period:

2002-05-24 to 2002-5-30

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

guideline study

Qualifier:

according to guideline

Guideline:

OECD Guideline 111 (Hydrolysis as a Function of pH)

GLP compliance:

yes

Radiolabelling:

no

Analytical monitoring:

yes

Details on sampling:

- Sampling intervals for the parent/transformation products: Immediately after addition of the test substance to the buffer and then at 1 minute intervals.
- Sampling method: Approximately 800 µL of test solution was quickly transferred to an NMR tube using a clean dried 1000 µL syringe.

Buffers:

- pH: Target: 4.0; Measured: 4.02
- Type and final molarity of buffer: Formic Acid/Sodium Hydroxide, 0.05 M
- Composition of buffer: 0.237 g Acetic Acid, 1.24 mL 2 M Sodium Hydroxide solution, 1.323 g NaCl. Total volume 100 mL.

- pH: Target: 7.0; Measured: 7.00
- Type and final molarity of buffer: Sodium Phosphate, monobasic/Sodium Hydroxide, 0.05 M
- Composition of buffer: 0.600 g Sodium Phosphate, monobasic, 0.92 mL 2 M Sodium Hydroxide solution, 0.953 g NaCl. Total volume 100 mL.

- pH: Target: 9.0; Measured: 9.00
- Type and final molarity of buffer: Boric acid/Sodium hydroxide, 0.30 M
- Composition of buffer: 0.312 g boric acid, 0.440 mL 2 M Sodium Hydroxide solution, 1.413 g NaCl. Total volume 100 mL.

- Buffer solutions were sparged with argon gas for a minimum of 5 min to exclude oxygen prior to conducting the test.
- The temperature of each buffer was 2.0±0.1°C
- The pH meter was calibrated for use with D2O.

Estimation method (if used):

Not applicable.

Details on test conditions:

oTemperature:  2.0±0.1°C.  The rationale for performing the hydrolysis at one temperature was to provide the best opportunity to slow down the hydrolysis rate and obtain kinetic information.
oThe hydrolysis reactions employed an initial test substance concentration of 1.0 x10-3 M.
oDue to the hydrolytically unstable nature of trimethoxysilane (Kallos et al., 1991), a stock solution in acetonitrile-d3/acetonitrile was used. 
oConstant ionic strength of 0.25 M was maintained for buffers by the addition of sodium chloride.
o0.05 M buffer solutions were prepared using deuterated water (99.9 atom %D).  Deuterated water (D2O) rather than H2O was used to provide a reference frequency lock for the NMR spectrometer and minimize the dynamic range problem introduced by a large solvent peak.
oThe relationship between the pH and pD scales has been established (Glasoe and Long, 1960).  For solutions of comparable acidity or basicity, the pH meter reading in D2O solutions is 0.40 pH units lower than in H2O solutions when calibrated against aqueous buffer standards.  Therefore, pD = pH meter reading + 0.40 pH units.  The relationship is independent of whether the internal solution of the electrode contains H2O or D2O.
oBuffer solutions were sterilized by filtering through 0.20 um cellulose nitrate membrane. oThe pH of each buffer solution was measured with a calibrated pH meter (using aqueous pH buffers) at 2.0± 0.1°C and then converted to pD values.  This provided a D+ concentration equivalent to the H+ concentration at pH 4, 7, and 9.
oVessels:  50-mL sterile polypropylene centrifuge tubes.
oCo-solvent:  <1% acetonitrile (~50:50 mixture of deuterated and non-deuterated acetonitrile).

TEST SYSTEM
- Type, material and volume of test flasks, other equipment used: 50 mL sterile polypropylene centrifuge tubes with caps.
- Sterilisation method: Nalgene sterile filtration units with 0.2 µm cellulose nitrate membrane were used to sterilize the buffers.
- Measures to exclude oxygen: Prior to use, all buffer solutions were sparged with argon gas for a minimum of 5 min to exclude oxygen and carbon dioxide.
- Details on test procedure for unstable compounds: The test substance is unstable with respect to moisture. Items used to prepare the test substance stock solution were oven dried to remove trace moisture. Syringes were dried in a Hamilton syringe heater with aspirator. The acetonitrile used to make up the test substance solution was distilled over P2O5 and stored over molecular sieves to remove trace moisture. The stock solution was prepared inside a nitrogen gas purged glove bag and stored in 22-mL plastic vials with septum lined open-top caps. When not in use, the stock solution was stored in a secondary air tight container with Drierite. The test substance solution in acetonitrile was added directly to the buffer solution at the start of the hydrolysis experiment.
- Details of traps for volatile, if any: None
- If no traps were used, is the test system closed/open: Closed, although caps were removed for sampling.
- Is there any indication of the test material adsorbing to the walls of the test apparatus?: No

TEST MEDIUM
- Volume used/treatment: 800 µL of the 0.1035 M solution of the test substance in acetonitrile-d3/acetonitrile was added to 25 mL of each buffer.
- Kind and purity of water: Deuterated water, D2O.
- Preparation of test medium: A 0.1 M stock solution of the test substance in acetonitrile was prepared. This was injected directly into the buffer solution at the start of the hydrolysis experiment because the test substance is highly unstable in water.
- Renewal of test solution: Not applicable.
- Identity and concentration of co-solvent: Acetonitrile (0.98%)

Number of replicates:

Replicates:  Two at pD 4.02, 7.00, and 9.00.

Positive controls:

no

Negative controls:

no

Statistical methods:

Descriptive statistics were performed.

Preliminary study:

The substance is known to be unstable at environmentally relevant temperatures, therefore, no preliminary study was required.

Transformation products:

yes

No.:

#1

No.:

#2

Details on hydrolysis and appearance of transformation product(s):

- Formation and decline of each transformation product during test: Decrease in the 1H-NMR peak for MeO-Si (in the test substance) and increase in the peak due to methanol (transformation product) were measured during the test. Complete disappearance of the test substance was observed.
- Pathways for transformation: Due to the limitation imposed by the rapid hydrolysis, only the total hydrolysis could be studied.

% Recovery:

0

pH:

4

Temp.:

2 °C

Duration:

1.5 min

% Recovery:

0

pH:

7

Temp.:

2 °C

Duration:

2 min

% Recovery:

0

pH:

9

Temp.:

2 °C

Duration:

1.5 min

Key result

pH:

4

Temp.:

2 °C

DT50:

<= 0.2 min

Remarks on result:

other: This value represents an estimated upper limit of the hydrolysis half-life. It refers to disappearance of the test material.

Key result

pH:

7

Temp.:

2 °C

DT50:

<= 0.3 min

Remarks on result:

other: This value represents an estimated upper limit of the hydrolysis half-life. It refers to disappearance of the test material.

Key result

pH:

9

Temp.:

2 °C

DT50:

<= 0.2 min

Remarks on result:

other: This value represents an estimated upper limit of the hydrolysis half-life. It refers to disappearance of the test material.

Other kinetic parameters:

None determined.

Details on results:

TEST CONDITIONS
- pH, sterility, temperature, and other experimental conditions maintained throughout the study: Yes

MAJOR TRANSFORMATION PRODUCTS
The increase in methanol during the hydrolysis experiment was observed but not quantified.

In all kinetic experiments, trimethoxysilane was completely hydrolysed by the time the first 1H NMR spectrum was acquired. Initial spectra were acquired after 90-144 seconds.

Rate constants and half-lives could not be determined quantitatively, although the data is certainly adequate for estimating the upper limit of t_1/2. The half-life was estimated based on the elapsed time to the initial spectrum.

Table 1 shows the results for each experiment.

Table 1. Results

pH		Replicate	Elapsed time	Estimated half-life*	Average half-life
4		A	98	14.0	14.1
4		B	99	14.1	14.1
7		A	144	20.6	17.0
7		B	94	13.4	17.0
9		A	103	14.7	13.8
9		B	90	12.9	13.8

* [elapsed time / s] / 7

Since the hydrolysis was so rapid, there was insufficient data to determine rate constants (k1, k2, and k3) for the sequential hydrolysis reactions of each methoxy group by regression modelling.

First order or pseudo-first order behaviour could not be confirmed because:  a) the analytical method was unable to follow the decrease of the parent peak intensity from the test substance or the increase of peak intensity from the hydrolysis co-product (methanol) due to a rapid hydrolysis reaction, b) no data points were obtained during the critical portion of the hydrolysis process (20 - 70% hydrolysed), and c) the relationship between k1, k2, and k3 is not known.

Validity criteria fulfilled:

yes

Conclusions:

A hydrolysis half-life of =17 s at pH 4, 7 and 9 and 2°C was determined for the substance in a reliable study conducted according to an appropriate test protocol, and in compliance with GLP.

Description of key information

Hydrolysis half-life: =0.2 minutes at pH 4 and pH 9, =0.3 min at pH 7 and 2°C (OECD 111)

Additional information

Trimethoxysilane, HSi(OMe)₃, is very unstable in the presence of water. The substance contains two reactive groups: Si-OMe and Si-H. The rate of Si-OMe hydrolysis has been measured in a reliable study conducted in accordance with OECD 111 and in compliance with GLP. Half-lives of =0.2 minutes at pH 4 and pH 9, =0.3 minutes at pH 7 and 2°C were determined for the submission substance. The result is considered to be reliable and is selected as key study. Methanol is produced by this reaction. If Si-OMe is hydrolysed, but Si-H is not, silanetriol (HSi(OH)₃) would be formed as an intermediate hydrolysis product.

The Si-H bond of silanetriol is expected to react in water, forming monosilicic acid, Si(OH)₄ as the ultimate hydrolysis product. The rate of this reaction is uncertain. Neither silanetriol nor monosilicic acid have been isolated; but only exist in dilute aqueous solution. They readily and rapidly (within minutes) condense to give insoluble polymeric species at concentrations above 100-150 mg SiO₂/L. Depending on the pH and concentration, solutions will contain varying proportions of monomeric silanol species, cyclic and linear oligomers and polymeric species of three-dimensional structure.

In a secondary source to which reliability could not be assigned, the stability of the substance in aqueous media under physiological conditions was investigated. The rates of hydrolysis of 1000 ppm trimethoxysilane were determined in water at pH 5.7, 0.15 molar (M) sodium phosphate buffer (PBS), and 10% rat serum in 0.15M PBS at pH 7.4 and 37.4°C in soft glass reactors. In this study, the substance was hydrolysed in PBS and PBS plus 10% rat serum at pH 7.4 and 37°C with half-lives of 0.23 minutes. At pH 5.7 and 37.4°C, half-life of 0.08 min was obtained. This is also supported by a result in a secondary literature to which reliability could not be assigned, which reports a half-life of 0.08 minutes at pH 5.7 and 37.4°C.

As the hydrolysis reaction may be acid or base-catalysed, the rate of reaction is expected to be slowest at around pH 7 and increase as the pH is raised or lowered. For an acid-base catalysed reaction in buffered solution, the measured rate constant is a linear combination of terms describing contributions from the uncatalysed reaction as well as catalysis by hydronium, hydroxide, and general acids or bases.

k_obs= k₀+ k_H3O+[H₃O⁺] + k_OH-[OH^-] + k_a[acid] + k_b[base]

 

At extremes of pH and under standard hydrolysis test conditions, it is reasonable to suggest that the rate of hydrolysis is dominated by either the hydronium or hydroxide catalysed mechanism.

Therefore, at low pH:

k_obs˜k_H3O+[H₃O⁺]

 

At pH 4 [H3O⁺] = 10^-4 mol dm-³and at pH 2 [H3O⁺] = 10^-2mol dm^-3; therefore, k_obs at pH 2 should be approximately 100 times greater than k_obs at pH 4.

The half-life of a substance at pH 2 is calculated based on:

t_1/2(pH 2) = t_1/2(pH 4) / 100

The calculated half-life of the substance at pH 2 is therefore less than 2 seconds at 25°C. However, it is not appropriate or necessary to attempt to predict accurately when the half-life is less than 5-10 seconds. As a worst-case it can therefore be considered that the half-life for trimethoxysilane at pH 2 and 20-25°C is approximately 5 seconds.

Reaction rate increases with temperature therefore hydrolysis will be faster at physiologically relevant temperatures compared to standard laboratory conditions. Under ideal conditions, hydrolysis rate can be recalculated according to the equation:

DT₅₀(XºC) = DT₅₀(T°C) *e^(0.08.(T-X))

Where T = temperature for which data are available and X = target temperature.

Thus, for trimethyoxysilane the hydrolysis half-life at 37.5ºC and pH 7 (relevant for lungs and blood), and at 37.5ºC and pH 2 (relevant for conditions in the stomach following oral exposure), it is not appropriate to apply any further correction for temperature to the limit value and the hydrolysis half -life is therefore approximately 5 seconds.

The ultimate hydrolysis products in this case are silicic acid and methanol.

Si-H hydrolysis

There is no direct evidence for the rate of reaction of the Si-H bond in trimethoxysilane.

 

However, studies have been carried out for several siloxane and silane compounds containing one or more Si-H bonds. The data are presented in the table below. The experiments investigated the rate of Si-H reaction by measuring the formation of gas (H₂) using headspace analysis. The half-lives obtained are in the range of hours to days. There is some indication that substances which hydrolyse rapidly to silanol containing species (HD4, HD5, H2L2) produce hydrogen faster than siloxanes that hydrolyse slower or silanes; and of these the silanetriol forming species (HD4 and HD5) produce H₂ faster than the silanediol forming substance (H2L2). Therefore, the Si-H bond in silanetriol may react faster than any of these substances. The substance H4D4 is considered the most appropriate read-across for trimethoxysilane as it produces Me[SiH](OH)₂ (the closest analogue to [SiH](OH)₃) on a rapid timescale.

 

For this substance, 80% of the expected H₂ formed after approximately 20 hours; assuming (pseudo) first order kinetics, this can be translated into a half-life of approximately 9 hours.

 

Si(H)(OH)₃ might be expected to be more reactive than RSi(H)(OH)₂ (where R is an organic group, for example methyl) as the replacement of –H with –R increases the nucleophilicity of the Si centre and favours attack by, for example, OH^-. The presence of a C atom attached to Si will slow these types of processes, for electronic and steric reasons.

 

The half-life for Si-H reactivity of trimethoxysilane is estimated as <12 hours (possibly much less than 12 hours) at pH 7 and 25°C for the purposes of the current chemical safety assessment but further testing is considered necessary.

  

Hydrolysis data for Si-H containing substances

CAS	Substance Name (acronym)	Initial hydrolysis produc	Final hydrolysis product	Parent substance degradation t1/2 at pH 7 and 22.5-25°C	Timescale of full siloxane hydrolysis	Timescale of H2 evolution
2370-88-9	H4D4	H[OSi(Me)(H)]4OH	MeSi(OH)3	2.2 min	Estimated t1/2 for siloxanediol chain shortening reactions 0.2 – 1 h	80% expected H2 formed after ca. 20 h
6166-86-5	H5D5	H[OSi(Me)(H)]5OH	MeSi(OH)3	4.2 min	Estimated t1/2 for siloxanediol chain shortening reactions 0.1 – 1 h	80% expected H2 formed after ca. 20 h
3277-26-5	H2L2	Me2Si(H)OH	Me2Si(OH)2	11 min	n/a	Apparent half-life of 2.5 days for degradation of intermediate hydrolysis product observed
1873-88-7	HL3	Me3SiOSi(Me)(H)OH and Me3Si(OH)3	MeSi(OH)3(and Me3Si(OH)3)	2.1 d	Estimated t1/2 for hydrolysis of Me3SiOSi(Me)(H)OH is 8 h.	8.9 days to 41% conversion to H2. Half-life of 17 days calculated but rate of H2formation did slow over time.
993-07-7	Trimethylsilane	n/a	Me3SiOH	4.2 d	n/a	t1/2= 4.2 d