WO2025089240A1 - Colloidal silica and method for producing colloidal silica - Google Patents

Abstract

The present invention provides: a colloidal silica which is in a single-digit nanometer size or has a small particle diameter close to the single-digit nanometer size, and which is substantially free from aggregation; and a method for producing the colloidal silica.　This colloidal silica is characterized in that: the surface is modified; the BET diameter is 12 nm or less; and the colloidal silica is not substantially aggregated. This method for producing the colloidal silica is characterized by including: a starting material preparation step for preparing a starting material colloidal silica having a BET diameter of 12 nm or less by supplying an easily decomposable organosilicate to a reaction liquid that contains a hydrolysis catalyst composed of an organic amine and reacting the easily decomposable organosilicate with the reaction liquid; a concentration adjustment step for setting the solid content concentration of the starting material colloidal silica to 13 mass% or less and setting the concentration of an alcohol generated from the starting material preparation step to 1-25 mass%; a modification step for modifying the concentration-adjusted starting material colloidal silica; and a concentration step for concentrating the modified colloidal silica so that the residual organic solvent in the modified colloidal silica becomes 1 mass% or less.

Description

Colloidal silica and method for producing colloidal silica

This invention relates to colloidal silica and a method for producing colloidal silica.

　Methods that have been proposed and implemented as industrial methods for producing high-purity colloidal silica include ion-exchanging an aqueous solution of sodium silicate, thermal decomposition of silicon tetrachloride, and hydrolysis of organosilicate in a water-alcohol mixed solvent in the presence of an acid or alkali catalyst. However, the method of hydrolyzing organosilicate allows the use of high-purity organosilicate, catalyst, and solvent for the reaction, so that the amount of impurities derived from these raw materials is extremely small, and it is particularly suitable as a method for producing high-purity colloidal silica with low metal impurity content, and several methods have been proposed for hydrolyzing organosilicate.

Here, regarding colloidal silica used for various purposes, particularly colloidal silica used in the field of semiconductor wafer polishing, with the increasing integration of today's LSIs, various types of metal wiring and oxide films exist on a single wafer, and each semiconductor wafer requires its own appropriate polishing performance, so colloidal silica with various slightly different compositions and properties is required.

In addition, for colloidal silica used in applications where even the slightest alkali metal impurities are undesirable, such as binders for hard coatings and ceramics, chromate-based metal surface treatment agents, and ground improvement grouting agents, acidic colloidal silica is required, and several proposals for manufacturing such acidic colloidal silica are known.

For example, the applicant of the present application has investigated a method for easily producing colloidal silica having predetermined properties, such as spherical colloidal silica, which does not require special post-treatment such as acid treatment, ion exchange treatment, or further modification treatment, has an extremely low content of metal impurities, including alkali metals, and has an average particle size in the range of 5 to 500 nm, a standard deviation of 20 or less, and a polydispersity index of 0.15 or less, as determined by particle size distribution analysis using an electron microscope. As a result, the applicant has proposed that neutral colloidal silica with a pH of 5 to 8 can be easily produced without special post-treatment such as acid treatment, ion exchange treatment, or the like, by using an easily hydrolyzable organosilicate with a fast hydrolysis rate and a specific hydrolysis catalyst as the hydrolysis catalyst, and adding and reacting this hydrolysis catalyst so that at least the ratio of hydrolysis catalyst (A) to silica (B) in the reaction mixture at the end of the reaction {catalyst remaining molar ratio (A/B)} is a predetermined value or less (Patent Document 1).

Incidentally, even in the field of abrasives for semiconductor wafers, for example, there has been a need to meet various demands in recent years, and with the colloidal silica of several tens of nanometers (nm) that has been conventionally produced, it is difficult to make fine adjustments during polishing, leading to problems such as over-removal of the surface or failure to form a clean surface. Another factor is the ongoing miniaturization of semiconductors.

Therefore, there is a demand for colloidal silica with a particle size smaller than that of conventional colloidal silica, that is, a particle size of a few nm, or a particle size close to that.However, when the particle size is such, the tendency of particles to aggregate due to the solid content of colloidal silica increases, and even if the particle size is close to single nano size, there is a concern that the characteristics (e.g., polishability, improvement of the number of defects on the polished object, etc.) cannot be fully utilized.In this regard, according to the study by the inventors of the present application, it has been found that aggregation becomes more prominent, especially as the solid content concentration of colloidal silica increases.
That is, colloidal silica that has a particle size of single nano size or close to single nano size and has a stable dispersed state without causing aggregation even when the solid content concentration is relatively high has not been found so far.

JP 2007-153732 A International Publication No. 2016/117560 International Publication No. 2023/007938 JP 2004-315300 A JP 2023-072344 A JP 2018-002569 A International Publication No. 2012/161157

Shinjiro Aoki, Journal of the Society of Rubber Science and Technology of Japan, Vol. 59, No. 11, 1986, 610-619. M.-C. B. Salon, et al., Colloids and Surfaces A: Physicochem. Eng. Aspects, 2008, 312, 83-91.

The inventors of the present application therefore conducted extensive research to develop colloidal silica with a small particle size of single nano size or close to that size, which does not substantially aggregate even when the solid concentration is relatively high. As a result, they discovered that this could be achieved by devising a raw material preparation process and a concentration adjustment process, etc., and by subjecting colloidal silica to a modification treatment at a predetermined timing, and thus completed the present invention.

The object of the present invention is therefore to provide colloidal silica having a small particle size of single nano size or close to that size, which is substantially free of aggregation, and a method for producing the same.

Several techniques for modifying colloidal silica have been reported.
Here, Patent Document 2 is related to modified colloidal silica that can improve the stability of polishing rate over time when used as abrasive grains with few microparticles, but does not specifically teach colloidal silica with a single nano size or a small particle size close to that, and does not disclose particle size in its examples.In addition, in the manufacturing method described in Patent Document 2, it is essential to provide a process of concentrating the colloidal silica so that the residual organic solvent concentration is 1 mass% or less (organic solvent distillation process) before the modification process of modifying the colloidal silica.Therefore, in such a manufacturing process, when trying to obtain colloidal silica with a small particle size close to that of single nano size, there is a risk of aggregation before the modification process is performed.

Patent Document 3 relates to a chemical mechanical polishing composition capable of selectively polishing a silicon nitride film by sufficiently increasing the polishing rate of a silicon nitride film relative to the polishing rate of a silicon oxide film. In the examples of Patent Document 3, it is shown that a colloidal silica particle dispersion having a predetermined solid content concentration, primary particle size, and secondary particle size obtained by the method described in Example 8 or Example 1 of Patent Document 4 is mixed with 3-mercaptopropyltrimethoxysilane and heated under reflux to obtain a thiolated silica sol, and then hydrogen peroxide is added to the silica sol and heated under reflux to obtain a dispersion containing abrasive grains D or E having the predetermined solid content concentration, primary particle size, and secondary particle size. In the dispersions containing abrasive grains D or E, the primary particle size is 10 nm or less at the concentration (4% by mass) when obtained, the secondary particle size is not significantly different from the primary particle size, and the association ratio is about 1.2, so it is understood that there is little aggregation at that concentration. However, when the inventors of the present application verified this, as shown in Reference Experimental Examples 1 and 2 described below, it was found that these dispersions would likely aggregate and gel if the solids concentration was significantly greater than 10% by mass, and would not be stable over time.

Patent Document 5 provides a chemical mechanical polishing composition capable of polishing a molybdenum film and a silicon oxide film at a stable polishing rate and suppressing the corrosion of the molybdenum film and the occurrence of defects in the silicon oxide film. In the examples of Patent Document 5, a (3-triethoxysilyl)mercapto group-containing silane coupling agent is dropped into a colloidal silica dispersion and stirred, and then hydrogen peroxide is added and refluxed under normal pressure to obtain an aqueous dispersion J containing silica particles surface-modified with sulfo groups and having an average secondary particle diameter of 7.4 nm (see paragraph [0106]). The primary particle diameter of the silica particles was 7.0 nm. Patent Document 6 provides a method for converting the silica particles into nanoparticles having a diameter of 10 nm or less by mixing ordinary silica particles such as powder or granules, water, and a metal alkoxide compound having a chain group having an amino group to prepare a reaction mixture, and heating the reaction mixture at 75° C. or higher for a certain period of time. In the examples of this patent document 6, powdered silica gel, 3-aminopropyltrimethoxysilane, and distilled water are charged into a reaction vessel equipped with a reflux device, and the mixture is heated and stirred. The aqueous solution obtained after the reaction is a sol solution containing colloidal silica, and the particles observed with a tunneling electron microscope (TEM) are all nanoparticles of 10 nm or less, and are observed in a dispersed state without aggregation (see paragraphs [0057], [0061], etc.).
However, the aqueous dispersion obtained in Patent Document 5 and the sol solution containing silica obtained in Patent Document 6 are likely to aggregate when the solid content concentration is increased due to their production methods, and are also highly likely to lack stability over time.

As shown in Patent Document 7, an organic solvent-dispersed silica sol in which colloidal silica is dispersed in a dispersion medium containing an organic solvent is disclosed, in which a silane compound having at least one ether structure is bonded to the surface of colloidal silica particles. As confirmed in the examples, even after being kept in a thermostatic chamber at 50°C for one month, the particle size measured by dynamic light scattering remains unchanged and the particle is relatively stable. However, the BET diameter is not necessarily small. Furthermore, the organic solvent-dispersed silica sol described in Patent Document 7 uses a large amount of benzyl alcohol in the dispersion solvent, exceeding 80% by mass, and has a different purpose from the colloidal silica of the present application.

That is, the gist of the present invention is as follows.
(1) A colloidal silica having a modified surface, a BET diameter of 12 nm or less, and being substantially free of aggregation.
(2) The colloidal silica according to (1), which has a BET specific surface area of 227 m ² /g or more.
(3) The colloidal silica according to (1), wherein the modification is anion-modified or cation-modified.
(4) The colloidal silica according to (3), wherein the anion-modification is performed with a sulfo group.
(5) The colloidal silica according to (3), wherein the cation modification is performed by a primary amino group.
(6) The colloidal silica according to (1), having a solid content of 12% by mass or more.
(7) The colloidal silica according to (1), which has a solid content concentration of 12% by mass or more, and which exhibits, with respect to a cumulant average diameter measured by a dynamic light scattering method, a rate of change in the cumulant average diameter after being kept at a temperature of 60° C. for one week, compared to that before the keeping, of within 20%.
(8) A method for producing the colloidal silica according to any one of (1) to (7), comprising the steps of:
a raw material preparation step of supplying and reacting a readily decomposable organosilicate to a reaction solution containing a hydrolysis catalyst made of an organic amine to prepare raw material colloidal silica having a BET diameter of 12 nm or less;
a concentration adjusting step of adjusting the solid content concentration of the raw material colloidal silica to 13% by mass or less and adjusting the concentration of alcohols generated in the raw material preparation step to 1 to 25% by mass;
a modification treatment step of modifying the raw colloidal silica having the adjusted concentration;
and a concentrating step of concentrating the modified colloidal silica so that a residual organic solvent in the modified colloidal silica is 1 mass % or less.
(9) The method for producing colloidal silica according to (8), wherein in the raw material preparing step, the reaction is carried out under conditions in which the feed rate of the easily hydrolyzable organosilicate is less than 1.5 mass%/min of the total amount of the easily hydrolyzable organosilicate fed, the reaction time is 6 hours or less, and the reaction temperature is 70° C. or lower.
(10) The method for producing colloidal silica according to (8), wherein in the concentrating step, the colloidal silica after the modification treatment step is concentrated so that the solid content concentration in the colloidal silica is 12 mass% or more.
(11) The method for producing colloidal silica according to (8), characterized in that the modification treatment step comprises a step of reacting a modifier having a functional group that can be converted into an anionic group with the colloidal silica after the concentration adjustment step, and a step of converting the functional group in the modifier after the reaction into an anionic group.
(12) The method for producing colloidal silica according to (11), wherein the modifying agent has a mercapto group and/or a sulfide group, and the mercapto group and/or the sulfide group are converted to a sulfo group by treating with an oxidizing agent to obtain colloidal silica having sulfo groups on the surface.
(13) The method for producing colloidal silica according to (8), wherein the modification treatment step comprises a step of reacting a modifying agent having a cationic group with the colloidal silica after the concentration adjustment step.

According to the present invention, it is possible to obtain colloidal silica that is substantially free of aggregation, even when the particle size is single nano size or close to that size and the solid content concentration is relatively high.

FIG. 1 shows an SEM image (magnification: 500,000) of the colloidal silica obtained in Example 1. FIG. 2 shows an SEM image (magnification: 500,000) of the colloidal silica obtained in Example 2. Figure 3 is a photograph showing the state of aggregated gel in Comparative Example 2. (Fig. 3A) is a photograph showing the state in which the gel was placed in a spherical glass container, (Fig. 3B) is a photograph showing the state in which a part of the gel was taken out and placed on a petri dish, and (Fig. 3C) is a photograph showing the state in which the gel was crushed. 4 is a photograph of the state in which a portion of the concentrated colloidal silica particle dispersion obtained in Reference Experimental Example 1 was taken, allowed to cool at room temperature, and then the container was turned upside down. In the figure, (*) indicates aggregated gelled matter. 5 is a photograph of the state in which a portion of the concentrated colloidal silica particle dispersion obtained in Reference Experimental Example 2 was taken, allowed to cool at room temperature, and then the container was turned upside down. In the figure, (*) indicates aggregated gelled matter.

<Colloidal Silica>
The colloidal silica of the present invention is surface-modified, has a BET diameter of 12 nm or less, and is substantially free of aggregation.

The colloidal silica used as the raw material is not limited as long as it has silanol groups on the surface, but considering that it does not contain metal impurities or corrosive ions such as chlorine, colloidal silica obtained by hydrolysis and condensation using hydrolyzable silicon compounds as raw materials (for example, easily hydrolyzable organosilicates or derivatives thereof, which will be described later) is preferred. The raw colloidal silica can be used alone or in a mixture of two or more types.

First, the BET diameter will be described.
As described above, the colloidal silica of the present invention has a small particle size of single nano size or close to single nano size, and has a BET diameter of 12 nm or less. As exemplified in the industrial applicability described below, small particle colloidal silica having a BET diameter of 12 nm or less is useful in that it can be easily prepared as an abrasive for semiconductor wafers and is less likely to be scratched during polishing.

The BET diameter is preferably 12 nm or less, more preferably 10 nm or less, in terms of polishing efficiency. There is no lower limit to the BET diameter, but it can be set appropriately taking into account the application and characteristics. For example, to ensure dispersion stability, it is preferably 1 nm or more, more preferably 5 nm or more.

The reason why the BET diameter is used to express a small particle diameter in this invention is that, for example, as will be explained later, it is possible for the colloidal silica to be not only monodispersed spherical products but also colloidal silica produced by two- or three-dimensional coalescence of multiple particles, and the production method of the present invention produces colloidal silica with an uneven surface and a relatively high specific surface area. In light of these circumstances, it is preferable to adopt the BET diameter assuming a spherical shape while also taking into account the specific surface area in judging the performance relative to the particle shape.

That is, in the present invention, the BET diameter is a particle diameter calculated from the BET specific surface area S (unit: ^m2 /g) of colloidal silica measured by the BET method and the true density ρ (unit: g/ ^cm3 ). Specifically, it can be calculated from the following formula (1).
BET diameter (nm) = 6000/(S x ρ) ... (1)
Here, ρ is the typical true density of _SiO2 , which is 2.2 g/ ^cm3 .

The colloidal silica of the present invention preferably has a BET specific surface area of 227 m ² /g or more. This BET specific surface area is suitable in terms of polishing efficiency. The preferred BET specific surface area is 230 m ² /g or more, more preferably 300 m ² /g or more. Although there is no upper limit for the BET specific surface area, it is preferably 500 m ² /g or less, more preferably 400 m ² /g or less, since this may cause concerns about stability.

Next, the modification will be described.
The colloidal silica of the present invention is obtained by subjecting raw colloidal silica to a modification treatment. Here, the modification treatment in the present invention is not limited as long as it contributes to the improvement or modification of the properties such as the improvement of dispersion stability, the suppression of aggregation, and the affinity with the object to be polished, which are the objects of the present invention, and can be appropriately selected from the modification treatments of colloidal silica known in the art. For example, nonionic modification, anionic modification, cationic modification, etc. can be mentioned. Specifically, it is preferable to subject colloidal silica to anionic or cationic modification, and more preferably anionic modification. As such a modification treatment, anionic or cationic modification is preferable, and although the reason why anionic modification is more preferable is not necessarily clear, it is presumed that the surface potential of colloidal silica is changed by modification, and the action of attracting with the object to be polished is expected, and the selectivity of the object to be polished is improved. In particular, compared with colloidal silica that has not been modified, the dispersion stability tends to be improved not only in alkaline conditions but also in acidic conditions where it is metastable and relatively prone to aggregation, so it is more preferable to perform anionic modification.

As mentioned above, the modification treatment can be appropriately selected from known methods, but as a preferred embodiment of the modification treatment, an anionic modification treatment and a cationic modification treatment are typically described below. For an anionic modification treatment, reference can be made to, for example, JP-A-2010-269985 and JP-A-2013-041992. For a cationic modification treatment, reference can be made to, for example, JP-A-2005-162533 and JP-A-2020-73445.

Specific methods for the anion modification treatment are not limited, but include, for example, a method in which a modifier having an anionic group is chemically bonded to the surface of colloidal silica. Another method includes a method in which a modifier having a functional group that can be converted to an anionic group by a chemical method or the like is chemically bonded to the surface of colloidal silica, and then a treatment is performed to convert the functional group to an anionic group, thereby forming an anionic group on the surface of colloidal silica. For reasons such as the efficiency of the modification treatment, the ability to stably introduce anionic groups into colloidal silica, and the difficulty of directly obtaining a modifier having an anionic group, a method using a modifier having a functional group that can be converted to an anionic group is preferable.

The compound (modifier) having a functional group that can be converted into an anion group is preferably, but not limited to, a silane coupling agent having a functional group that can be converted into an anion group. Here, the anion group is not limited to, but may be, for example, a sulfo group, a carboxy group, or a phosphate group. It may be ionic bonded with a cation to form a salt. Even with such an anion group, the cation is released in an aqueous solution and it functions as an anion group. Examples of cations that ionic bond with an anion group include alkali metal ions such as sodium ions and potassium ions, and alkaline earth metal ions such as calcium ions. From the viewpoint of improving the selectivity of the object to be polished, among the anion groups, the sulfo group and the carboxy group are preferable, and the sulfo group is more preferable.

A sulfo group will be taken as an example of a preferred embodiment of the anion group. Modifiers having a functional group that can be converted to a sulfo group include silane coupling agents having a sulfonate ester group that can be converted to a sulfo group by hydrolysis, and silane coupling agents having a mercapto group and/or a sulfide group that can be converted to a sulfo group by oxidation. Of these, a method using a silane coupling agent having a mercapto group and/or a sulfide group is more preferred because it is easy to modify the colloidal silica. In the present invention, only one type of modifier may be used, or two or more types may be used in combination.

Examples of silane coupling agents having a mercapto group include 3-mercaptopropyltrimethoxysilane, 2-mercaptopropyltriethoxysilane, 2-mercaptoethyltrimethoxysilane, and 2-mercaptoethyltriethoxysilane. Examples of silane coupling agents having a sulfide group include bis(3-triethoxysilylpropyl)disulfide.

A more specific method for the anion modification treatment will be described in detail in the explanation of the manufacturing method described later, but the amount of the above-mentioned modifier used is preferably 0.1 to 10 parts by mass, more preferably 0.5 to 7 parts by mass, and even more preferably 0.8 to 6 parts by mass, relative to 100 parts by mass of the solid content of the raw colloidal silica. In other words, the modifier is preferably 0.1 to 10% by mass, more preferably 0.5 to 7% by mass, and even more preferably 0.8 to 6% by mass in the solid content of the colloidal silica after modification. If the amount used is within this range, the particle surface of the colloidal silica can be sufficiently anionized. Furthermore, if the amount used is such, it is possible to produce colloidal silica that has been stably anion-modified without aggregation. Note that a solvent (water, a hydrophilic organic solvent, etc.) for dissolving the modifier may be used during the modification treatment.

As for cationic modification, examples include a method in which a modifying agent having a cationic group is chemically bonded to the surface of colloidal silica, and a method in which a modifying agent having a functional group that can be converted to a cationic group by a chemical method or the like is chemically bonded to the surface of colloidal silica, and then a process is carried out to convert the functional group into a cationic group, thereby forming a cationic group on the surface of colloidal silica. A preferred method is to chemically bond a silane coupling agent having a cationic group to the surface of colloidal silica.

Cation groups are not limited, but examples include primary amino groups, secondary amino groups, tertiary amino groups, quaternary ammonium groups, imino groups, and iminium groups. Primary amino groups, secondary amino groups, tertiary amino groups, and quaternary ammonium groups are preferred. Primary amino groups are more preferred. The cationic group may form an ionic bond with an anion to form a salt. Even with such cationic groups, the anion is released in the mixed liquid and the group functions as a cationic group. Examples of anions that form an ionic bond with a cationic group include fluoride ions, chloride ions, bromide ions, iodide ions, hydrochloride ions, acetate ions, sulfate ions, hydrofluoric acid ions, and carbonate ions.

Examples of silane coupling agents having a cationic group include N-2-(aminoethyl)-3-aminopropylmethyldimethoxysilane, N-2-(aminoethyl)-3-aminopropyltrimethoxysilane, 3-aminopropyltrimethoxysilane, 3-aminopropyltriethoxysilane, 3-triethoxysilyl-N-(1,3-dimethyl-butylidene)propylamine, and N-phenyl-3-aminopropyltrimethoxysilane.

A more specific method for the cation modification treatment will be described in detail in the explanation of the manufacturing method below, but as described above, it is preferable to use a method using a modifying agent having a cationic group. The amount of the modifying agent having a cationic group used is preferably 0.1 to 1.5 parts by mass, more preferably 0.5 to 1.2 parts by mass, and even more preferably 0.6 to 1 part by mass, relative to 100 parts by mass of the solid content of the raw colloidal silica. In other words, the amount of the modifying agent having a cationic group in the solid content of the colloidal silica after cation modification is preferably 0.1 to 1.5% by mass, more preferably 0.5 to 1.2% by mass, and even more preferably 0.6 to 1% by mass. If the amount used is within this range, the particle surface of the colloidal silica can be sufficiently cationized. Furthermore, if the amount used is such, it is possible to produce a stably cation-modified colloidal silica without aggregation. In addition, a solvent (water, a hydrophilic organic solvent, etc.) for dissolving the modifying agent may be used during the modification treatment.

The modified colloidal silica of the present invention obtained by the above-mentioned modification treatment (hereinafter, sometimes referred to as "modified colloidal silica") has excellent dispersion stability and does not substantially aggregate even when the solid concentration is relatively high, and is suitable for various applications. The solid concentration is usually about 6% by mass due to the manufacturing process, etc., but it is preferably 12% by mass or more, more preferably 15% by mass or more, even more preferably 18% by mass or more, and particularly preferably 19% by mass or more, because it can improve polishing performance and reduce transportation costs. The solid concentration can be appropriately set depending on the application, etc. On the other hand, the upper limit of the solid concentration is not limited, but it is preferably 50% by mass or less because there is a tendency for the handleability to decrease due to an increase in viscosity, etc., and for the dispersion stability to decrease.

It is known that, in general, as the solid content increases, the distance between particles decreases, and therefore, colloidal silica becomes more susceptible to aggregation.
However, the modified colloidal silica of the present invention has a relatively small particle size as described above, and does not substantially aggregate even at a relatively high solids concentration as described above, and a practical colloidal silica of this kind has not yet been clearly identified.

As mentioned above, the colloidal silica of the present invention is substantially free of aggregation. Here, the wording "substantially" is used for the following reason. That is, in general, methods for confirming whether colloidal silica is aggregated include a method of confirming with an electron microscope, measuring the increase in viscosity, and measuring the cumulant mean diameter by dynamic light scattering (DLS). However, for the colloidal silica of the present invention, which is an extremely small particle with a BET diameter of 12 nm or less, it is difficult to completely grasp whether each and every particle is monodispersed. Since it is considered that the above-mentioned methods can only grasp or understand the dispersion state of the particles with a certain degree of certainty, as long as there is substantially no aggregation to the extent that it does not cause practical problems, the colloidal silica will not exhibit its functions or cause problems in use.

Then, in order to grasp more specifically that the above-mentioned "substantially not aggregated" is not present, it can be measured by the above-mentioned method, but since the particles are small, some parts may be difficult to grasp by an electron microscope. Therefore, it is preferable to use dynamic light scattering (DLS) and judge from the change in the cumulant mean diameter measured by the method. More specifically, it is preferable to confirm the change (or the lack of change) in the cumulant mean diameter by DLS measurement over time. For example, in the manufacturing method of the present invention, in the concentration step described below, the presence or absence of aggregation can be confirmed by confirming the DLS measurement results before and after concentration. In other words, by grasping the presence or absence of aggregation with an increase in solid concentration, it becomes possible to confirm whether the colloidal silica has a property that is likely to cause aggregation. This is because, as mentioned above, when the solid concentration is relatively low, aggregation does not occur much, and as the solid concentration increases, aggregation becomes more likely. This tendency usually tends to be noticeable when the solid concentration is 10% by mass or more, more specifically, 12% by mass or more.

Among the DLS measurement methods, a more practical method for grasping aggregation is to measure the change in particle diameter by DLS measurement after holding for a certain time (period) under heated conditions. With this method, it is possible to confirm whether the colloidal silica has the characteristics of being prone to aggregation without causing an active change in the solid content concentration. Specifically, as will be described in the examples below, the change in the cumulant mean diameter after holding for 7 days at at least 60°C, which is a temperature at which particles are prone to aggregation, is confirmed, and if the rate of change is within 20%, it can be determined that there is no aggregation. Here, the reason why the method of holding at 60°C or higher for 7 days is preferably applied is that it is known that the condition of 60°C for 7 days is approximately equivalent to a time-dependent change test in a state held at room temperature for one year, and this makes it possible to confirm the occurrence of aggregation in a time-dependent change test at least at room temperature for one year. And, preferably, the rate of change in the cumulant mean diameter by DLS measurement is within 10%, more preferably within 5%. Here, for the reasons described above, when performing such DLS measurement, the solid content concentration is preferably maintained or measured at 12% by mass or more, more specifically 18% by mass or more, even more specifically 19% by mass or more, and preferably 50% by mass or less, more preferably 40% by mass or less, and even more preferably 30% by mass or less. The solid content concentration at this time is not limited, and it is preferable that the DLS measurement is performed at a solid content concentration that matches the application and actual use.

The colloidal silica of the present invention may have any shape or other properties depending on the application or purpose, so long as it has the above-mentioned properties of BET diameter, modification, and substantial absence of aggregation. In particular, the colloidal silica obtained by the present invention preferably has the above-mentioned BET specific surface area, a relatively high particle surface area, and a large number of irregular small protrusions, so to speak, a confetti-like shape as a whole particle. Such a shape has a large BET specific surface area compared to the large SEM average particle diameter measured by measuring the arithmetic mean of particle images observed by SEM, and also has a high particle density (true specific gravity) measured by the liquid phase displacement method, in other words, a high hardness, and has an excellent polishing rate, making it suitable for use as an abrasive for CMP.

The shape of the colloidal silica particles of the present invention can be controlled by the feed composition, etc., to be monodisperse spheres (spherical products) or to have a shape in which the particles are bonded together and associated (associated products). For example, by adding a large amount of catalyst and relatively slowly introducing an organosilicate as a silica raw material to the reaction field, the organosilicate hydrolyzes quickly and uniformly and grows mildly, so that the seed particles grow gradually while maintaining their spherical shape, and a spherical product can be obtained. Also, by adding a small amount of catalyst and relatively quickly introducing an organosilicate as a silica raw material to the reaction field, the organosilicate hydrolyzes non-uniformly and acts like an adhesive between the particles, resulting in an associated product in which the particles are associated.

The viscosity of the colloidal silica of the present invention is preferably 1 to 100 mPa·s, more preferably 1 to 50 mPa·s, and even more preferably 1 to 20 mPa·s.

As described above, the colloidal silica of the present invention may be a monodisperse spherical product, or it may be an aggregated product (cocoon-shaped, chain-like, branched, etc.) that has a shape that appears to be formed by multiple particles bonding together in two or three dimensions when observed under an electron microscope.

The colloidal silica of the present invention may be adjusted to a pH that does not impair dispersion stability, depending on the above-mentioned modification treatment. For example, in the case of the anion-modified colloidal silica, the pH is preferably 1 to 5, and more preferably 2 to 3. For example, in the case of the cation-modified colloidal silica, the pH is preferably 8 to 11, and more preferably 8.5 to 10. Adjusting the pH to such a range is favorable in terms of the dispersion stability of the colloidal silica.

Furthermore, the colloidal silica of the present invention preferably has a metal impurity content of 1 ppm or less, more preferably 0.01 ppm or less, and even more preferably 0.0001 ppm or less. Although not limited thereto, such high purity colloidal silica can be achieved, for example, in the manufacturing method described below, by using silica source, hydrolysis catalyst, and water used as raw materials in the hydrolysis reaction to obtain the raw material colloidal silica, which satisfy the above metal impurity content.

<Method of producing colloidal silica>
The method for producing colloidal silica of the present invention essentially includes the following steps.
(a) a raw material preparation step of supplying and reacting a readily decomposable organosilicate to a reaction liquid containing a hydrolysis catalyst made of an organic amine, thereby preparing raw material colloidal silica having a BET diameter of 12 nm or less.
(b) a concentration adjusting step of adjusting the solid content concentration of the raw material colloidal silica to 13% by mass or less and adjusting the concentration of alcohols generated in the raw material preparation step to 1 to 25% by mass.
(c) a modification treatment step of modifying the raw material colloidal silica having its concentration adjusted in the step (b).
(d) a concentration step of concentrating the colloidal silica modified in the step (c) so that the residual organic solvent in the colloidal silica is 1 mass % or less.
These steps are described below.

[Step (a)]
First, in step (a), a colloidal silica having a BET diameter of 12 nm or less is prepared by supplying and reacting an easily decomposable organosilicate to a reaction solution containing a hydrolysis catalyst made of an organic amine. In the present invention, the colloidal silica prepared in step (a) is called raw colloidal silica. Although a commercially available product can be used as the raw colloidal silica, a hydrolysis method is used because it reduces metal impurities, can obtain colloidal silica with a relatively large surface area, and can produce particles with high uniformity. As the hydrolysis method, a method is used in which a silica source is supplied to a reaction solution containing a hydrolysis catalyst and hydrolyzed, and a method is used in which a easily decomposable organosilicate is supplied as a silica source to a reaction solution containing a hydrolysis catalyst made of an organic amine and hydrolyzed because it is easy to control the particle diameter.

Here, the silica source preferably used in step (a) is an easily hydrolyzable organosilicate with a fast hydrolysis rate. The easily hydrolyzable organosilicate is preferably one which is hydrolyzed within 1 hour by stirring 10 g of organosilicate and 100 g of pure water with impurities of 0.1 ppb or less at 25°C. Specific examples of such easily hydrolyzable organosilicates include trimethyl silicate (hydrolysis reaction time until hydrolysis reaction is completed: about 3 minutes), tetramethyl silicate (hydrolysis reaction time: about 5 minutes), triethyl silicate (hydrolysis reaction time: about 5 minutes), and methyl trimethyl silicate (hydrolysis reaction time: about 7 minutes). Tetraethyl silicate and organosilicates with a larger carbon number than tetraethyl silicate have a slow hydrolysis rate and tend to gel easily (hydrolysis reaction time: 24 hours or more for both), so the above-mentioned easily decomposable organosilicates are preferably used.

The organic amines used as the hydrolysis catalyst in step (a) are not limited, but may be one or a mixture of two or more selected from quaternary ammoniums, tertiary amines, secondary amines, primary amines, and their carbonates, bicarbonates, and silicates. For example, quaternary ammoniums include tetramethylammonium hydroxide (TMAH), tetraethylammonium hydroxide (TEAH), trimethylethylammonium hydroxide, trimethylethanolammonium hydroxide (choline), triethylethanolammonium hydroxide, tetrapropylammonium hydroxide, butylammonium hydroxide, and other quaternary ammoniums, as well as their carbonates, bicarbonates, and silicates. Since a relatively high pH is desirable for the hydrolysis reaction, tetramethylammonium hydroxide (TMAH), choline, or tetraethylammonium hydroxide (TEAH) is preferred.

Furthermore, the primary amines, secondary amines, and tertiary amines of the organic amines used as hydrolysis catalysts are not limited, but examples thereof include aminoalcohols, morpholines, piperazines, aliphatic amines, and aliphatic ether amines. Here, various aminoalcohols including ethanolamine derivatives can be used as aminoalcohols, but ethanolamine derivatives are preferred, such as monoethanolamine, diethanolamine, triethanolamine, N,N-dimethylethanolamine, N,N-diethylethanolamine, N,N-di-n-butylethanolamine, N-(β-aminoethyl)ethanolamine, N-methylethanolamine, N-methyldiethanolamine, N-ethylethanolamine, N-n-butylethanolamine, N-n-butyldiethanolamine, N-tert-butylethanolamine, and N-tert-butyldiethanolamine.

Furthermore, for the morpholines of the organic amines used as the hydrolysis catalyst, various morpholine derivatives can be used, with preferred examples including morpholine, N-methylmorpholine, and N-ethylmorpholine. For the piperazines of the organic amines used as the hydrolysis catalyst, various piperazine derivatives can be used, with preferred examples including piperazine and hydroxyethylpiperazine. For the aliphatic amines and aliphatic etheramines of the organic amines used as the hydrolysis catalyst, preferred examples of the aliphatic amines include alkylamines having 1 to 8 carbon atoms, such as triethylamine, dipropylamine, pentylamine, hexylamine, heptylamine, and octylamine. Preferred examples of the aliphatic etheramines include aliphatic etheramines having 1 to 8 carbon atoms, such as 2-methoxyethylamine, 3-methoxypropylamine, 3-ethoxypropylamine, 3-propoxypropylamine, 3-isopropoxypropylamine, and 3-butoxypropylamine.

The above organic amines used as hydrolysis catalysts can be used alone or, if necessary, in a mixture of two or more kinds.

As described above, the reaction liquid must contain an easily decomposable organosilicate as a silica source and a hydrolysis catalyst, but other than that, water, alcohols, aldehydes, ketones, surfactants, etc. can also be used. It is preferable that the total content of the silica source, hydrolysis catalyst, and water is 90% by mass or more. More preferably, these are 95% by mass or more.

In the mixture after the reaction of the silica source with the hydrolysis catalyst (hereinafter, this may be referred to as the "reaction mixture"), or in the reaction mixture after the solids concentration or alcohol concentration has been adjusted as described below, or after the treatment of dispersion stabilization with acid (hereinafter, this may be particularly referred to as the "reaction concentrate"), it is preferable to add the hydrolysis catalyst to the reaction system and carry out the hydrolysis reaction so that the ratio of the hydrolysis catalyst (A) to the silica (B) {catalyst remaining molar ratio (A/B)} is 0.012 or less, more preferably in the range of 0.00035 to 0.012, and even more preferably in the range of 0.0035 to 0.011. This is preferable because it is possible to optimize the pH of the reaction mixture or reaction concentrate, and to suppress thickening and gelation.

The method for achieving such a catalyst remaining molar ratio is not particularly limited, but examples include a method of continuously or intermittently introducing a silica source calculated so that the final catalyst remaining molar ratio (A/B) falls within the above range into a reaction vessel charged with water and hydrolysis catalyst (A), a method of continuously or intermittently introducing a hydrolysis catalyst and a silica source calculated so that the final catalyst remaining molar ratio falls within the above range into a reaction vessel charged with only water, and a method of continuously or intermittently introducing a hydrolysis catalyst and a silica source calculated so that the final catalyst remaining molar ratio falls within the above range into a reaction vessel charged with water and a small amount of hydrolysis catalyst (A).

In addition, prior to the hydrolysis reaction of the silica source, colloidal silica seeds having particle growth properties may be charged into the reaction system for the hydrolysis reaction, and the silica source and hydrolysis catalyst may be gradually added to the reaction system so that the catalyst remaining molar ratio (A/B) falls within the above range. This is preferable because it allows the production of colloidal silica with uniform particles.

The silica source, hydrolysis catalyst, and water used as raw materials for the hydrolysis reaction preferably have a metal impurity content of 1 ppm or less, and more preferably have a high purity of 0.01 ppm or less. This ensures that the raw colloidal silica obtained and the modified colloidal silica obtained after modification also satisfy the above-mentioned range of metal impurity content.

In step (a), the method for making the BET diameter of the raw colloidal silica 12 nm or less is not limited, but it is preferable to adjust the rate at which the silica source is dropped into the reaction solution (feed rate), the reaction temperature, and the reaction time.

When a readily decomposable organosilicate is used as the silica source, the supply rate is preferably less than 1.5 mass%/min of the total amount of readily decomposable organosilicate added, more preferably 1.3 mass% or less/min, and even more preferably 1.2 mass% or less/min. If the supply rate is 1.5 mass% or more/min, the particles tend not to be uniformly dispersed or to not be nearly spherical. The reason why the total amount is used as the standard is that the amount added varies depending on the production scale, etc., and it is preferable to consider completing the supply within the reaction time described below. In other words, it is preferable to make the supply rate as slow as possible, which makes it easier to form the desired small particle size. There is no lower limit to the supply rate, but if the supply rate is too slow, there is a risk that the target particle size will not be achieved, so it is preferable to make it 1.0 mass%/min or more of the total amount added.

The reaction temperature is preferably 70°C or less, more preferably 65°C or less, and even more preferably 60°C or less. If the reaction temperature exceeds 70°C, the reaction liquid will volatilize more and the liquid composition will be more likely to change, which may make it difficult to control the particle size. On the other hand, the lower limit of the reaction temperature can be set as appropriate, but if the temperature is too low, the hydrolysis reaction tends to be slow and particle growth may be promoted, so it is preferably set to 20°C or more.

The reaction time is preferably 6 hours or less, more preferably 3 hours or less, and even more preferably 2 hours or less. If the reaction time exceeds 6 hours, there is a risk that the target particle size will not be achieved. On the other hand, the lower limit of the reaction time can be set as appropriate, but if the reaction time is too short, the hydrolysis reaction may not be completed, and particle formation and particle growth may not occur sufficiently, so it is preferably 20 minutes or more.

In step (a), it is generally preferable that the solids concentration during the reaction is 3 to 13% by mass. If the solids concentration is less than 3% by mass, particle formation and growth may not occur sufficiently. On the other hand, if the solids concentration is too high, particles are likely to aggregate. In step (a), it is preferable that the solids concentration of the colloidal silica after the reaction is in the range of 3 to 10% by mass.

[Step (b)]
In step (b), the solid content concentration of the raw colloidal silica obtained in step (a) is adjusted to 13% by mass or less. By adjusting the solid content concentration to 13% by mass or less, even if the BET diameter is 12 nm or less, the occurrence of aggregation can be reduced, and the next step (c) (modification treatment step) can be performed without causing aggregation gelation. Preferably, the solid content concentration is 10% by mass or less, more preferably 8% by mass or less. Here, although it is difficult to clearly define or classify, the term "aggregation gelation" generally refers to a state in which the aggregation of colloidal silica has progressed further and has become a jelly-like mass when viewed with the naked eye (for example, see FIG. 3 in Comparative Example 2 described later).

Here, if the solids concentration of the raw colloidal silica obtained in step (a) is already 13% by mass or less, there is no need to actively adjust the solids concentration in step (b), and in that case, it may be treated as a step of confirming the solids concentration, or a step of maintaining the solids concentration as it is. Ensuring that the solids concentration is indeed 13% by mass or less in step (b) before the modification treatment in the next step (c) is a characteristic step in obtaining the final modified colloidal silica of the present invention that is substantially non-aggregated and has a relatively high solids concentration.

The method for adjusting the solid content concentration in step (b) is not limited, but it is preferable to add the same solvent as that contained in the raw colloidal silica. The solvent to be added is preferably water and/or alcohols, more preferably water, methanol and/or ethanol, and even more preferably water and/or methanol. As described above, the solvent to be added may include water, alcohols, aldehydes, ketones, surfactants, etc.

In addition, in this step (b), the concentration of the alcohols produced from the step (a) is adjusted to 1 to 25% by mass. In the step (a), the hydrolysis reaction produces alcohols according to the easily decomposable organosilicate used as the silica source. For example, when the easily decomposable organosilicate used has a methoxy group, such as trimethyl silicate or tetramethyl silicate, methanol is produced as the alcohol. When the easily decomposable organosilicate used has an ethoxy group, such as triethyl silicate, ethanol is produced as the alcohol. The same applies to other cases.

Although the detailed mechanism behind the need to adjust the concentration of the alcohols resulting from the hydrolysis reaction in step (a) to 1-25% by mass is unclear, it is presumed that by maintaining the concentration of the organic solvent in this range, the occurrence of aggregation due to the polarity of the solvent can be reduced even if the BET diameter is 12 nm or less, and the next step (c) (modification treatment step) can be performed without causing aggregation and gelation. The concentration of the generated alcohols can be appropriately adjusted by the amount of easily decomposable organosilicate used in step (a) or by replacing the alcohols with pure water by distillation, but the preferred lower limit is 5% by mass or more, and the more preferred lower limit is 12% by mass or more. On the other hand, the preferred upper limit is 20% by mass or less.

As in the case of adjusting the solid content concentration (13% by mass or less) in this step (b), if the alcohol concentration of the raw colloidal silica obtained in step (a) is already 1-25% by mass, there is no need to actively adjust the alcohol concentration. In that case, it may be treated as a step of confirming the alcohol concentration, or as a step of maintaining the alcohol concentration produced in step (a) as it is. As in the case of adjusting the solid content concentration (13% by mass or less) described above, ensuring that the alcohol concentration is 1-25% by mass in this step (b) before the modification treatment in the next step (c) is a characteristic step in obtaining a substantially non-aggregated colloidal silica of the present invention with a relatively high solid content concentration as the final modified colloidal silica of the present invention. The alcohol may be part of the solvent in the modification treatment in the next step (c).

There are no limitations on the method for adjusting the concentration of the alcohols in step (b), but when adjusting to increase the concentration of the alcohols, it is preferable to add an organic solvent containing the same organic solvent as the produced alcohols as the main component (e.g., 95% by mass or more). As described above, the organic solvent added here may include water, alcohols, aldehydes, ketones, surfactants, etc. On the other hand, an example of a method for adjusting to decrease the concentration of the alcohols is to volatilize the alcohols by heating them while distilling them from a distillation tube with a condenser, as is done in Comparative Example 1 described later. In this case, pure water may be added.

[Step (c)]
In step (c), the raw colloidal silica, the solid content concentration of which and the concentration of the generated alcohols have been adjusted through step (b), is modified. As described above, the modification is not limited and can be selected from known modification treatments, but anion modification or cation modification is preferred, and anion modification is more preferred.
Here, the anion modification or the cation modification is as exemplified above, and the anion modification is preferred, and a suitable method for the anion modification is a method in which a modifying agent having a functional group that can be converted to an anion group by a chemical method or the like is chemically bonded to the surface of the colloidal silica. Among the anion modifications, the sulfo group exemplified as a preferred embodiment will be taken as an example. Also, among the cation modifications, the primary amino group exemplified as a preferred embodiment will be taken as an example.

(Embodiment of anion modification treatment using sulfo group as an example)
To form sulfo group on the surface of colloidal silica, it can be carried out by a known method as described above.Preferably, a method can be mentioned in which a modifying agent having a functional group that can be converted to a sulfo group by chemical method or the like is chemically bonded to the surface of colloidal silica, and then the functional group is converted to a sulfo group, and a modifying agent having a functional group that can be converted to a sulfo group by oxidation is preferable.Among these, a preferred modifying agent is a silane coupling agent having a mercapto group and/or a sulfide group as described above, and a representative one will be described below.

In this method, it is preferable to include a step (step c1) of reacting raw colloidal silica, the solid content of which has been adjusted through steps (a) and (b) in the presence of the silane coupling agent, and a step (step c2) of oxidizing the reaction product of step c1 to convert mercapto groups and/or sulfide groups to sulfo groups. Steps c1 and c2 can also be performed with reference to the above-mentioned JP-A-2010-269985 and JP-A-2013-041992. Note that other steps may be included as appropriate as long as they do not impair the object of the present invention. Examples of other steps include adjusting the viscosity of the reaction solution and adjusting the pH.

(Step c1)
In step c1, the raw colloidal silica, the solid content of which has been adjusted through steps (a) and (b), is reacted in the presence of a silane coupling agent having a mercapto group and/or a sulfide group, so that the silane coupling agent is chemically bonded to the surface of the raw colloidal silica.

The reaction in step c1 can be carried out within the range of temperatures suitable for use of the denaturant, and is not limited to a particular range, for example, 40°C or higher and below the boiling point of the reaction liquid (solvent). To improve reactivity, the reaction is preferably carried out at a temperature of 50°C or higher, more preferably 60°C or higher, and preferably below the boiling point of the reaction liquid (solvent), 100°C or lower. The reaction time is also not limited, but is preferably carried out for 10 minutes to 10 hours, and more preferably 1 to 8 hours.

In the reaction of step c1, a solvent for improving the solubility of the modifier or silane coupling agent can be added within a range that does not impair the object of the present invention. Such a solvent can be a hydrophilic solvent, for example, alcohols such as methanol, ethanol, isopropanol, etc., but is not limited to these. It is more preferable to use the same alcohols as those produced by the hydrolysis reaction to obtain the raw colloidal silica.

As mentioned above, the amount of the modifier (silane coupling agent) used is preferably 0.1 to 10 parts by mass, more preferably 0.5 to 7 parts by mass, and even more preferably 0.8 to 5 parts by mass, per 100 parts by mass of the solid content of the raw colloidal silica.

The reaction in step c1 can also be understood, for example, by analyzing the functional groups introduced into the colloidal silica. For example, pulsed NMR (TD-NMR) can be used to perform qualitative and/or quantitative analysis of the functional groups.

(Step c2)
In step c2, the functional groups introduced onto the surface of the raw colloidal silica in step c1 are converted into anionic groups by a chemical method. Here, a method for converting the functional groups introduced by a silane coupling agent having a mercapto group and/or a sulfide group into sulfo groups by oxidation treatment will be described.

Methods for converting mercapto groups and/or sulfide groups to sulfo groups by oxidation treatment include, but are not limited to, the use of an oxidizing agent. Examples of oxidizing agents include nitric acid, hydrogen peroxide, oxygen, ozone, organic peracids (percarboxylic acids), bromine, hypochlorite, potassium permanganate, chromic acid, etc. Among these, hydrogen peroxide and organic peracids (peracetic acid, perbenzoic acids) are preferred in terms of handling and reactivity (oxidation yield). It is most preferred to use hydrogen peroxide because it produces fewer by-products in the reaction. The amount of oxidizing agent added may be in excess of the amount of the modifying agent (silane coupling agent), but it is preferable to minimize the amount of residual oxidizing agent, and it is more preferred to use 3 to 5 moles of oxidizing agent per mole of silane coupling agent in order to ensure a sufficient oxidation reaction. There are no limitations on the oxidation reaction, but it can be carried out under suitable reaction conditions for the oxidizing agent used, as long as it does not impair the purpose of the present invention. For example, it is preferable to carry out the reaction at a temperature above room temperature and below the boiling point of the solvent used in the reaction solution (e.g., below 100°C) for a period of 3 to 5 hours.

By this step c2, colloidal silica having anionic groups (sulfo groups) on the surface can be obtained. Note that step c2 may include other steps after the oxidation treatment. For example, it may include a step for removing the oxidizing agent, a step for adjusting the pH of the solution after the reaction, etc., and these can be appropriately selected and performed within the scope that does not impair the object of the present invention.

(Embodiment of cationic modification treatment using amino group as an example)
To form amino groups on the surface of colloidal silica, known methods can be used as described above. A preferred method is to carry out a treatment in which a modifying agent having an amino group is chemically bonded to the surface of colloidal silica. As described above, a preferred modifying agent among these is a silane coupling agent having an amino group, which will be described below as a representative example.

In this method, it is preferable to include a step of reacting the raw colloidal silica, the solid content concentration of which has been adjusted through steps (a) and (b) as described above, in the presence of the silane coupling agent. Such steps can also be performed with reference to the above-mentioned JP-A-2005-162533 and JP-A-2020-73445. Other steps may be included as appropriate as long as they do not impair the object of the present invention. Examples of other steps include adjusting the viscosity of the reaction solution and adjusting the pH, but for dispersion stability, it is preferable to adjust the pH before performing the cation modification treatment. The pH adjusted at this time is preferably 8 to 11, and more preferably 8.5 to 10. The pH can be adjusted by a method using a known compound (pH adjuster), but a preferred method is to use the same compound as the hydrolysis catalyst used in step (a) as the pH adjuster.

In other words, the raw colloidal silica, whose solid content concentration has been adjusted through steps (a) and (b), is reacted in the presence of a silane coupling agent having an amino group. This causes the silane coupling agent to chemically bond to the surface of the raw colloidal silica.

This reaction can be carried out within the range of temperatures suitable for use with the denaturant, and is not limited, but can be carried out at a temperature range of, for example, 40°C or higher and below the boiling point of the reaction liquid (solvent). To improve reactivity, it is preferable to carry out the reaction at a temperature of 50°C or higher, more preferably 60°C or higher, and preferably below the boiling point of the reaction liquid (solvent), 100°C or lower. There is also no limit to the reaction time, but it is preferable to carry out the reaction for, for example, 10 minutes to 10 hours, and more preferably 1 to 8 hours.

In this reaction, a solvent for improving the solubility of the modifier or silane coupling agent can be added within a range that does not impair the object of the present invention. Such a solvent can be a hydrophilic solvent, for example, alcohols such as methanol, ethanol, isopropanol, etc., but is not limited to these. It is more preferable to use the same alcohols as those produced by the hydrolysis reaction to obtain the raw colloidal silica.

As mentioned above, the amount of the modifier (silane coupling agent) used is preferably 0.1 to 1.5 parts by mass, more preferably 0.5 to 1.2 parts by mass, and even more preferably 0.6 to 1 part by mass, per 100 parts by mass of the solid content of the raw colloidal silica.

This reaction can also be understood, for example, by analyzing the functional groups introduced into the colloidal silica. For example, pulsed NMR (TD-NMR) can be used to perform qualitative and/or quantitative analysis of the functional groups.

Furthermore, for the sake of dispersion stability, it is preferable to adjust the pH even after the cationic modification treatment. The pH adjusted in this case is preferably 8 to 11, and more preferably 8.5 to 10. As described above, the pH can be adjusted by a method using a known compound (pH adjuster), but a preferred method is to use the same compound as the hydrolysis catalyst used in step (a) as the pH adjuster.

[Step (d)]
After carrying out the step (c), in step (d), the residual organic solvent in the colloidal silica obtained after step (c) is concentrated to 1 mass % or less. Preferably, the residual organic solvent is 0.1 mass % or less, more preferably 0.05 mass % or less. That is, the colloidal silica obtained through the step (d) is dispersed in a dispersion medium having a residual organic solvent of 0.1 mass % or less. Such a dispersion medium is, for example, an aqueous solution having a residual organic solvent of 0.1 mass % or less, and is preferably substantially water, as in the embodiment described in the examples below. It is not excluded that the dispersion medium contains trace amounts of alcohols, aldehydes, ketones, surfactants, etc., and these are also preferably 0.1 mass % or less. By reducing the residual organic solvent to this range, the solid content concentration can be adjusted or increased to a predetermined range, and in the case of a highly volatile residual organic solvent, it is preferable because the concentration of colloidal silica can be prevented from fluctuating due to the evaporation of the organic solvent. In addition, when the colloidal silica of the present invention is used, there is an advantage that it is not necessary to consider the resistance of the material used in the application to residual organic solvents.

Here, the method for removing and concentrating the residual organic solvent is not particularly limited, and any known method can be used. For example, there is a method in which the residual organic solvent is distilled by heating using an apparatus equipped with a distillation tube with a condenser.

In this step (d), the solid content concentration can be adjusted to a level suitable for use, taking into account the intended use and purpose, by concentrating the process of removing the residual organic solvent and water. In the present invention, modified colloidal silica is obtained through the above-mentioned steps (a) to (c), and even if the solid content concentration is relatively high in this step (d), colloidal silica that does not substantially aggregate can be obtained. Considering such features of the present invention and the intended use and purpose of colloidal silica, as described above, the solid content concentration in the colloidal silica after the modification step (c) is preferably 12% by mass or more, more preferably 15% by mass or more, even more preferably 18% by mass or more, and particularly preferably 19% by mass or more. The upper limit of the solid content concentration is the same as described above, and is preferably 50% by mass or less, more preferably 40% by mass or less, and even more preferably 30% by mass or less.

The colloidal silica of the present invention can be produced by carrying out the above steps (a) to (d), but other steps may be included as appropriate even after step (d) as long as they do not impair the object of the present invention. For example, it is preferable to subject the colloidal silica after step (d) to dispersion stabilization. Publicly known treatments can be used for the dispersion stabilization treatment.

Colloidal silica that has been subjected to dispersion stabilization treatment using the method of the present invention exhibits excellent dispersion stability for a period of at least one week, and even for a period of several years, and does not undergo two-layer separation.

In the present invention, colloidal silica with a small particle size, such as a BET diameter of 12 nm or less, can be obtained as described above, but as in conventional methods, a method is also possible in which the obtained colloidal silica is used as seed particles, and a silica source is supplied to and reacted with a reaction liquid containing the seed particles and the hydrolysis catalyst to grow the particle size.

Below, we will explain in detail the preferred embodiment of the present invention based on examples and comparative examples.

[Example 1]
In a 5 liter (L) glass vessel equipped with a stirrer, a thermometer, a distillation tube with a condenser, and an organosilicate inlet tube, 3200 g of pure water having a metal impurity content of 0.1 ppb or less and 3.42 g of triethanolamine (boiling point (bp): 361°C) having a metal impurity content of 10 ppb or less were charged, and while the liquid temperature in the reaction vessel was kept at 60°C using a mantle heater, 706 g of tetramethylsilicate (manufactured by Tama Chemicals Co., Ltd.) having a metal impurity content of 10 ppb or less was continuously fed under stirring over a period of 90 minutes. The obtained raw colloidal silica was subjected to various analyses, and the results are shown in <1> of Table 1.

Next, 3900 g of the resulting reaction product (colloidal silica, solids concentration 6.45% by mass, methanol concentration 15% by mass) was added to a 5 L glass container, and no operations were performed to change the solids concentration or methanol concentration. Next, while maintaining the temperature at 80°C, 13.1 g of 3-mercaptopropyltrimethoxysilane dissolved in 117.9 g of methanol was continuously fed over a period of 6 hours. After feeding, 22.7 g of 30% by mass hydrogen peroxide solution was further continuously fed over a period of 4 hours while maintaining the temperature at 80°C, and the raw colloidal silica was modified.

After the hydrogen peroxide solution was supplied, the temperature in the reaction vessel was temporarily lowered to 40°C, the pressure in the system was reduced with a vacuum pump, and then heating was resumed, the reaction mixture in the reaction vessel was further heated to 52 to 68°C, and the generated methanol was distilled from a distillation tube equipped with a condenser at a distillation temperature of 32 to 67°C, and 250 g of pure water was further added while distilling off the water and methanol, thereby obtaining colloidal silica with a solid concentration of about 20% by mass. The residual organic solvent in the obtained colloidal silica was 0.1% by mass.
The obtained colloidal silica was subjected to various analyses, and the results are shown in <2> of Table 1. Furthermore, the obtained colloidal silica was stored at 60° C. for 7 days, and various analyses of the obtained colloidal silica were performed. It was confirmed, particularly from the results of DLS measurement, that substantially no aggregation had occurred. The results are shown in <3> of Table 1.

[Example 2]
In a 5 L glass vessel equipped with a stirrer, a thermometer, a distillation tube with a condenser, and an organosilicate inlet tube, 3374 g of pure water containing 0.1 ppb or less of metal impurities and 3.608 g of triethanolamine (bp: 361° C.) containing 10 ppb or less of metal impurities were charged, and while the liquid temperature in the reaction vessel was kept at 70° C. using a mantle heater, 1122 g of tetramethylsilicate containing 10 ppb or less of metal impurities (manufactured by Tama Chemicals Co., Ltd.) was continuously fed with stirring over a period of 270 minutes. Various analyses of the obtained raw colloidal silica were carried out, and the results are shown in <4> of Table 1.

Next, 2000 g of the resulting reaction product (colloidal silica, solids concentration 10.38% by mass, methanol concentration 20% by mass) was added to a 2 L glass container, and no operations were performed to change the solids concentration or methanol concentration. Next, while maintaining the temperature at 80°C, 6.71 g of 3-mercaptopropyltrimethoxysilane dissolved in 60.35 g of methanol was continuously fed over a period of 6 hours. After feeding, 11.63 g of 30% by mass hydrogen peroxide solution was further continuously fed over a period of 4 hours while maintaining the temperature at 80°C, and the raw colloidal silica was modified.

After the hydrogen peroxide solution was supplied, the temperature in the reaction vessel was temporarily lowered to 40°C, and the pressure in the system was reduced using a vacuum pump. Heating was then resumed, and the reaction mixture in the reaction vessel was further heated to 52-68°C. The methanol produced was distilled from the distillation tube with a condenser at a distillation temperature of 32-67°C. 320 g of pure water was then added while distilling off the water and methanol, yielding colloidal silica with a solid concentration of approximately 20% by mass. Various analyses were performed on the resulting colloidal silica, and the results are shown in <5> of Table 1. Furthermore, the resulting colloidal silica was stored at 60°C for 7 days, and various analyses were performed on the resulting colloidal silica, confirming that essentially no aggregation had occurred, particularly from the results of DLS measurements. The results are shown in <6> of Table 1.

[Example 3]
In a 5 L glass vessel equipped with a stirrer, a thermometer, a distillation tube with a condenser, and an organosilicate inlet tube, 3463 g of pure water containing 0.1 ppb or less of metal impurities and 0.336 g of triethanolamine (bp: 361°C) containing 10 ppb or less of metal impurities were charged, and while the liquid temperature in the reaction vessel was kept at 70°C using a mantle heater, 911.6 g of tetramethylsilicate (manufactured by Tama Chemicals Co., Ltd.) containing 10 ppb or less of metal impurities was continuously fed under stirring over a period of 90 minutes. Various analyses of the obtained raw colloidal silica were carried out, and the results are shown in <7> of Table 1.

Next, 4,200 g of the resulting reaction product (colloidal silica, solids concentration 8.82% by mass, methanol concentration 17% by mass) was added to a 5 L glass container, and no operations were performed to change the solids concentration or methanol concentration. Next, while maintaining the temperature at 80°C, 11.98 g of 3-mercaptopropyltrimethoxysilane dissolved in 107.8 g of methanol was continuously fed over a period of 6 hours. After feeding, 20.77 g of 30% by mass hydrogen peroxide solution was further continuously fed over a period of 4 hours while maintaining the temperature at 80°C, and the raw colloidal silica was modified.

After the hydrogen peroxide solution was supplied, the temperature in the reaction vessel was temporarily lowered to 40°C, and the pressure in the system was reduced using a vacuum pump. Heating was then resumed, and the reaction mixture in the reaction vessel was further heated to 52-68°C. The resulting methanol was distilled from the condenser-equipped distillation tube at a distillation temperature of 32-67°C. 700 g of pure water was then added while distilling off the water and methanol, yielding colloidal silica with a solids concentration of approximately 20% by mass. Various analyses were performed on the resulting colloidal silica, and the results are shown in <8> of Table 1. Furthermore, the resulting colloidal silica was stored at 60°C for 7 days, and various analyses were performed on the resulting colloidal silica, confirming that no aggregation had occurred, particularly from the results of DLS measurements. The results are shown in <9> of Table 1.

[Comparative Example 1]
In a 5 L glass vessel equipped with a stirrer, a thermometer, a distillation tube with a condenser, and an organosilicate inlet tube, 3200 g of pure water containing 0.1 ppb or less of metal impurities and 3.42 g of triethanolamine (bp: 361° C.) containing 10 ppb or less of metal impurities were charged, and while maintaining the liquid temperature in the reaction vessel at 60° C. using a mantle heater, 706 g of tetramethylsilicate (manufactured by Tama Chemicals Co., Ltd.) containing 10 ppb or less of metal impurities was continuously fed with stirring over a period of 90 minutes. Various analyses of the obtained raw colloidal silica were carried out, and the results are shown in <10> of Table 2.

Next, 3,900 g of the resulting reaction product (colloidal silica) was added to a 5 L glass container, and unlike in Examples 1 to 3 above, it was heated to 52°C to 72°C at this point, and the resulting methanol was distilled from a distillation tube with a condenser at a distillation temperature of 32 to 67°C. Furthermore, 1,800 g of pure water was added while distilling off the water and methanol.

3,900 g of the reaction product from which water and methanol had been distilled off (colloidal silica, solids concentration 6.45% by mass, methanol concentration 0.1% by mass) was added to a 5 L glass vessel, and no operation was performed to change the solids concentration. Next, while maintaining the temperature at 100°C, 13.7 g of 3-mercaptopropyltrimethoxysilane dissolved in 123.6 g of methanol was continuously fed over a period of 6 hours. After feeding, 23.8 g of 30% by mass hydrogen peroxide was further continuously fed over a period of 4 hours while maintaining the temperature at 100°C, and the raw colloidal silica was modified.

After the hydrogen peroxide solution was supplied, the temperature in the reaction vessel was temporarily lowered to 40°C, the pressure in the system was reduced using a vacuum pump, and then heating was resumed. The reaction mixture in the reaction vessel was further heated to 52-68°C, and colloidal silica with a solid content of approximately 20% by mass was obtained. Various analyses of the obtained colloidal silica were carried out, and the results are shown in <11> of Table 2. From the analysis results, particularly the DLS measurement results, the change in the cumulant mean diameter by DLS measurement exceeded 20%, confirming the presence of aggregation. The obtained colloidal silica was further stored at 60°C for 7 days, and various analyses of the obtained colloidal silica were carried out. The results are shown in <12> of Table 2. Because the colloidal silica was aggregated at the time of result <11>, the DLS result in this result <12> was also similar to result <11> and higher than result <10>.

[Comparative Example 2]
In a 5 L glass vessel equipped with a stirrer, a thermometer, a distillation tube with a condenser, and an organosilicate inlet tube, 3200 g of pure water containing 0.1 ppb or less of metal impurities and 3.42 g of triethanolamine (bp: 361° C.) containing 10 ppb or less of metal impurities were charged, and while keeping the liquid temperature in the reaction vessel at 60° C. using a mantle heater, 706 g of tetramethylsilicate containing 10 ppb or less of metal impurities (manufactured by Tama Chemicals Co., Ltd.) was continuously fed with stirring over a period of 90 minutes. Various analyses of the obtained colloidal silica were carried out, and the results are shown in <13> of Table 2.

Next, 4,500 g of the resulting reaction product (colloidal silica, solids concentration 6.01% by mass, methanol concentration 15% by mass) was added to a 3 L glass container, and no operations were performed to change the solids concentration or methanol concentration. Then, without carrying out a modification treatment process, it was heated at 52°C to 72°C, and the generated methanol was distilled from a distillation tube with a condenser at a distillation temperature of 32°C to 67°C. Pure water was further added while distilling off the water and methanol, and when the solids concentration reached 16% by mass, coagulation and gelation occurred. The state of coagulation and gelation is shown in Figure 3.

[Comparative Example 3]
In a 5L glass vessel equipped with a stirrer, a thermometer, a distillation tube with a condenser, and an organosilicate inlet tube, 3463 g of pure water containing 0.1 ppb or less of metal impurities and 0.336 g of triethanolamine (bp: 361°C) containing 10 ppb or less of metal impurities were charged, and while the liquid temperature in the reaction vessel was kept at 80°C using a mantle heater, 1136 g of tetramethylsilicate (manufactured by Tama Chemicals Co., Ltd.) containing 10 ppb or less of metal impurities was continuously fed under stirring for 480 minutes. Various analyses of the obtained colloidal silica were carried out, and the results are shown in <14> of Table 2. In this Comparative Example 3, the BET diameter exceeded 12 nm due to the temperature at the time of adding the silica source.

[Example 4]
In a 5 L glass vessel equipped with a stirrer, a thermometer, a distillation tube with a condenser, and an organosilicate inlet tube, 3889 g of pure water containing 0.1 ppb or less of metal impurities and 3.42 g of triethanolamine (bp: 361°C) containing 10 ppb or less of metal impurities were charged, and while the temperature of the liquid in the reaction vessel was kept at 60°C using a mantle heater, 707 g of tetramethylsilicate containing 10 ppb or less of metal impurities (manufactured by Tama Chemicals Co., Ltd.) was continuously fed with stirring over a period of 90 minutes. The obtained raw colloidal silica was subjected to various analyses, and the results are shown in <15> of Table 3.

Next, 3,800 g of the resulting reaction product (colloidal silica, solids concentration 6.39% by mass, methanol concentration 13% by mass) and 20 g of triethanolamine (bp: 361°C) with a metal impurity content of 10 ppb or less were added to a 5 L glass container, and no operation was performed to change the solids concentration or methanol concentration. The pH of the colloidal silica at this time was 8.9. Next, while maintaining the temperature at 80°C, a modifying agent obtained by dissolving 1.71 g of 3-aminopropyltrimethoxysilane in 50 g of methanol was continuously supplied over a period of 6 hours to modify the raw colloidal silica.

After the modifier was supplied, the temperature in the reaction vessel was temporarily lowered to 40° C., 20 g of triethanolamine (bp: 361° C.) having a metal impurity content of 10 ppb or less was added, the pressure in the system was reduced using a vacuum pump, and then heating was resumed. The reaction mixture in the reaction vessel was further heated to 52 to 68° C., and the produced methanol was distilled from the distillation tube equipped with a condenser at a distillation temperature of 32 to 67° C., and 670 g of pure water was further added while distilling off the water and methanol, thereby obtaining colloidal silica with a solid content concentration of about 20 mass%.
The obtained colloidal silica was subjected to various analyses, and the results are shown in <16> of Table 3. Furthermore, the obtained colloidal silica was stored at 60°C for 7 days, and various analyses of the obtained colloidal silica were performed. It was confirmed, particularly from the results of DLS measurement, that substantially no aggregation had occurred. The results are shown in <17> of Table 3.

[Example 5]
In a 5 L glass vessel equipped with a stirrer, a thermometer, a distillation tube with a condenser, and an organosilicate inlet tube, 3890 g of pure water containing 0.1 ppb or less of metal impurities and 3.42 g of triethanolamine (bp: 361°C) containing 10 ppb or less of metal impurities were charged, and while the temperature of the liquid in the reaction vessel was kept at 60°C using a mantle heater, 707 g of tetramethylsilicate containing 10 ppb or less of metal impurities (manufactured by Tama Chemicals Co., Ltd.) was continuously fed with stirring over a period of 90 minutes. The obtained raw colloidal silica was subjected to various analyses, and the results are shown in <18> of Table 3.

Next, 4001 g of the resulting reaction product (colloidal silica, solids concentration 6.41% by mass, methanol concentration 13% by mass) and 20 g of triethanolamine (bp: 361°C) with a metal impurity content of 10 ppb or less were added to a 5 L glass container, and no operation was performed to change the solids concentration or methanol concentration. The pH of the colloidal silica at this time was 9.0. Next, while maintaining the temperature at 80°C, a modifying agent obtained by dissolving 1.83 g of 3-aminopropyltrimethoxysilane in 54 g of methanol was continuously supplied over a period of 6 hours to modify the raw colloidal silica.

After the modifier was supplied, the temperature in the reaction vessel was lowered to 40° C., 10 g of triethanolamine (bp: 361° C.) having a metal impurity content of 10 ppb or less was added to 1,975 g of the reaction mixture, the pressure in the system was reduced with a vacuum pump, heating was then resumed, and the reaction mixture in the reaction vessel was further heated to 52 to 68° C., and the produced methanol was distilled from a distillation tube equipped with a condenser at a distillation temperature of 32 to 67° C., and 340 g of pure water was added while distilling off the water and methanol, thereby obtaining colloidal silica with a solid content concentration of about 20 mass%.
The obtained colloidal silica was subjected to various analyses, and the results are shown in <19> of Table 3. Furthermore, the obtained colloidal silica was stored at 60°C for 7 days, and various analyses of the obtained colloidal silica were performed. It was confirmed, particularly from the results of DLS measurement, that substantially no aggregation had occurred. The results are shown in <20> of Table 3.

[Example 6]
In a 5 L glass vessel equipped with a stirrer, a thermometer, a distillation tube with a condenser, and an organosilicate inlet tube, 3890 g of pure water containing 0.1 ppb or less of metal impurities and 3.43 g of triethanolamine (bp: 361°C) containing 10 ppb or less of metal impurities were charged, and while the liquid temperature in the reaction vessel was kept at 60°C using a mantle heater, 707 g of tetramethylsilicate containing 10 ppb or less of metal impurities (manufactured by Tama Chemicals Co., Ltd.) was continuously fed with stirring over a period of 90 minutes. The obtained raw colloidal silica was subjected to various analyses, and the results are shown in <21> of Table 3.

Next, 2350 g of the resulting reaction product (colloidal silica, solids concentration 6.47% by mass, methanol concentration 13% by mass) and 12 g of triethanolamine (bp: 361°C) with a metal impurity content of 10 ppb or less were added to a 5 L glass container, and no operation was performed to change the solids concentration or methanol concentration. The pH of the colloidal silica at this time was 8.9. Next, while maintaining the temperature at 80°C, a modifying agent obtained by dissolving 1.08 g of 3-aminopropyltrimethoxysilane in 30 g of methanol was continuously supplied over a period of 6 hours to modify the raw colloidal silica.

After the modifying agent was supplied, the temperature in the reaction vessel was temporarily lowered to 40° C., 12 g of triethanolamine (bp: 361° C.) having a metal impurity content of 10 ppb or less was added, the pressure in the system was reduced using a vacuum pump, and then heating was resumed. The reaction mixture in the reaction vessel was further heated to 52 to 68° C., and the produced methanol was distilled from the distillation tube equipped with a condenser at a distillation temperature of 32 to 67° C., and 351 g of pure water was added while distilling off the water and methanol, thereby obtaining colloidal silica with a solid content concentration of about 20 mass%.
The obtained colloidal silica was subjected to various analyses, and the results are shown in <22> of Table 3. Furthermore, the obtained colloidal silica was stored at 60°C for 7 days, and various analyses of the obtained colloidal silica were performed. It was confirmed, particularly from the results of DLS measurement, that substantially no aggregation had occurred. The results are shown in <23> of Table 3.

Next, as described above, using a colloidal silica particle dispersion having a predetermined solid content concentration, primary particle size, and secondary particle size obtained by the method described in Example 8 or Example 1 described in Patent Document 4, a dispersion containing Abrasive D or Abrasive E described in Patent Document 3 was prepared in the following manner, and the dispersion was evaluated.
That is, the following [Reference Experimental Example 1] and [Reference Experimental Example 2] show the evaluation results of colloidal silica particle dispersions corresponding to Abrasives D and E prepared by procedures almost identical to those described in paragraph 0075 of Patent Document 3 and Example 8 (paragraph 0053) of Patent Document 4 described therein, and in paragraph 0076 of Patent Document 3 and Example 1 (paragraph 0045) of Patent Document 4 described therein, respectively.

[Reference Experimental Example 1]
(procedure)
A solution of 250 parts of methanol, 10 parts of pure water containing 10 ppb or less of metal impurities, and 2 parts of 29% aqueous ammonia was prepared in a reaction vessel, and after maintaining the temperature at 63° C., 27 parts of tetramethoxysilane (manufactured by Tama Chemicals Co., Ltd.) containing 10 ppb or less of metal impurities was added dropwise at a rate of 1 cm ³ /min·L under stirring conditions to carry out hydrolysis and polycondensation reactions. Here, the unit of the dropping rate was defined as described in paragraph 0029 of Patent Document 4. After the reaction was completed and aging was carried out for 1 hour, 80 parts were distilled out by simple distillation, and 200 parts of ultrapure water were added. As described in Patent Document 4, the colloidal solution became slightly white, so 5 parts of 29% aqueous ammonia were added to return it to a transparent state.

5 kg of the colloidal silica particle dispersion obtained above, with a silica concentration of 3.8% by mass, a primary particle diameter of 9.0 nm, and a secondary particle diameter of 10.1 nm, was mixed with 2 g of 3-mercaptopropyltrimethoxysilane and heated under reflux for 2 hours to obtain a thiolated silica sol. Hydrogen peroxide was added to the silica sol and heated under reflux for 8 hours to obtain a dispersion containing 3.7% by mass of abrasive grains with primary particles of 9.0 nm and secondary particles of 10.1 nm, with sulfo groups fixed to the surfaces of the silica particles.

This colloidal silica particle dispersion was distilled at a distillation temperature of 32-67°C from a distillation tube equipped with a condenser by simple distillation in the same manner as in Example 1 above, and concentrated to 12% by mass, resulting in aggregation and gelation, as shown in Figure 4. Figure 4 is a photograph of a portion of the colloidal silica particle dispersion concentrated to approximately 12% by mass taken and left to cool at room temperature in an inverted container, and shows that the colloidal silica particle dispersion has aggregated and gelled at the bottom of the container (i.e., the upper part of the photograph), and does not fall out.

[Reference Experimental Example 2]
(procedure)
A solution of 310 parts of methanol, 16 parts of pure water containing less than 10 ppb of metal impurities, and 3 parts of 29% aqueous ammonia was prepared in a reaction vessel, and after maintaining the temperature at 60° C., a mixed diluted solution of 38 parts of tetramethoxysilane (manufactured by Tama Chemicals Co., Ltd.) containing less than 10 ppb of metal impurities and 18 parts of methanol was dropped at a dropping rate of 1 cm ³ /min·L under stirring conditions to carry out hydrolysis and polycondensation reactions, thereby obtaining a colloidal silica particle dispersion with a solid content concentration of about 3.8 mass %. Here, the unit of the dropping rate was defined according to the definition described in paragraph 0029 of Patent Document 4.

5 kg of the colloidal silica particle dispersion obtained above, with a silica concentration of 3.8% by mass, a primary particle diameter of 5.9 nm, and a secondary particle diameter of 7.3 nm, was mixed with 2 g of 3-mercaptopropyltrimethoxysilane and heated under reflux for 2 hours to obtain a thiolated silica sol. Hydrogen peroxide was added to the silica sol and heated under reflux for 8 hours to obtain a dispersion containing 3.8% by mass of abrasive grains with primary particles of 5.7 nm and secondary particles of 7.0 nm, with sulfo groups fixed to the surfaces of the silica particles.

This colloidal silica particle dispersion was distilled at a distillation temperature of 32-67°C from a distillation tube equipped with a condenser by simple distillation in the same manner as in Example 1 above, and concentrated to 11% by mass, resulting in aggregation and gelation, as shown in Figure 5. Figure 5 is a photograph of a portion of the colloidal silica particle dispersion concentrated to approximately 11% by mass taken and left to cool at room temperature in an inverted container, and it can be seen that, like Figure 4, the colloidal silica particle dispersion has aggregated and gelled at the bottom of the container (i.e., the upper part of the photograph) and does not fall out.

[Results and Discussion]
As described above, when each of the colloidal silica particle dispersions containing abrasive grains corresponding to Abrasive grain D or Abrasive grain E described in Patent Document 3 is concentrated to a concentration of at least 12 mass %, the colloidal silica particle dispersions all aggregate and gel, and therefore the colloidal silica having a BET diameter of 12 nm or less and a solids concentration of 12 mass % or more shown in the above-mentioned embodiment of the present application is not obtained, and therefore it can be said that the same stability over time as that shown in the embodiment of the present application is not provided.

(Discussion)
Here, the reason why the colloidal silica particle dispersion prepared by the method shown in Patent Document 3 cannot exceed 12% by mass is considered to be influenced by the following two points (i) and (ii), that is, (i) the alcohol (methanol) content concentration of the silica particle dispersion during the thiolation reaction (modification treatment process) and (ii) the presence or absence of methanol dilution of the silane coupling agent (3-mercaptopropyltrimethoxysilane). Regarding this (i), when the alcohol contained in the silica colloid particle dispersion is replaced with pure water before the modification treatment process as in the above-mentioned Comparative Example 1, as explained in the above-mentioned Comparative Example 1 and shown in Table 2, the change in the cumulant average diameter by DLS measurement exceeded 20% before and after the modification treatment process, and a tendency to aggregate was observed. Since a simple distillation operation is performed after the preparation of the silica particles in Example 8 of Patent Document 4, it is considered that the abrasive D of Patent Document 3 using the silica particle dispersion could not be concentrated to 12% by mass even after the modification treatment.

In addition, gelation was also confirmed in the abrasive grain E of Patent Document 3 (using Example 1 of Patent Document 4), which was subjected to a modification treatment process without performing a simple distillation operation after the preparation of silica particles, and this is believed to be due to the above reason (ii). In the above-mentioned Example 1, the silane coupling agent is diluted with alcohol and then continuously supplied (i.e., added dropwise) for 6 hours during the modification treatment process, while in Patent Document 3, it is mixed without dilution before heating and refluxing. Regarding this, for example, in the above-mentioned Non-Patent Document 1 (Aoki Shinjiro, Journal of the Japan Rubber Association, Vol. 59, No. 11, 1986, 610-619.), which teaches the trend of modification using coupling agents, it is described in "3.1 Silane" on page 611 that "Structurally, RSi(OCH ₃ ) ₃ , R is an organic functional group, and a polar polymer is selected and combined. It dissolves in methanol and water, hydrolyzes, and condenses to form silane triol, which is chemically adsorbed to silica-based powders and fibers. (Diagram omitted) When the solvent is evaporated and weakly heated, it becomes a thin film with organic functional groups that chemically bond with the polymer. It undergoes hydrolysis and chemical reaction, and chemical adsorption and crosslinking/polymerization at the interface. ' It is known that in the reaction of general silane coupling agents, including 3-mercaptopropyltrimethoxysilane, after the alkoxy group of the silane coupling agent is hydrolyzed, the self-condensation reaction between the silane coupling agents (resulting in the formation of aggregates) and the modification reaction to the silica particles compete with each other, but in the modification treatment process, as in the above-mentioned Example 1, it is confirmed that an efficient modification reaction to the silica particles proceeds by diluting the silane coupling agent with methanol and continuously supplying it.

The effect of concentration and temperature on such modification reactions is shown, for example, in the above-mentioned non-patent document 2 (M.-C. B. Salon, et al., Colloids and Surfaces A: Physicochem. Eng. Aspects, 2008, 312, 83-91.). This document is a paper titled "Kinetics of hydrolysis and self condensation reactions of silanes by NMR spectroscopy" and, in "3.4.1. Silane concentration" on page 89, the effect of the concentration of the silane coupling agent on the reaction is investigated. The silane coupling agent is compared at 10% and 4% in the alcohol solvent. As a result of 4% compared to 10%, it is stated that "The hydrolysis rate decreased (complete hydrolysis within 10 h). The _{silanol formation T 0 H} ^was _also slowed down (only 80% of ^such species within 6 h reaction)." Furthermore, in an experiment in which a different silane coupling agent was used and the concentration was increased to 20%, it was stated that "However, the maximum of formed silanol units (T ⁰ _H ) was reduced to 50% probably because of the increase of self condensation reactions; the T ^{1 and} _T ² formations were accelerated, as presented in Figure 9b." The conclusion regarding the above concentration study was concluded as "These data show that a concentration of 10% gave optimal experimental conditions to enhance the maximum silanol formation and to lower self condensation reactions."

Furthermore, in "3.4.2. Effect of the temperature" on page 90, the effect of temperature on the reaction is investigated. That is, as described in "The hydrolysis reaction was started at room temperature (25℃) for 2h. Then the temperature was kept to 70℃ (corresponding to the boiling temperature of the solvent mixture), as displayed in Fig. 9c.", Fig. 9c shows the time course of hydrolysis and self-condensation of the silane coupling agent when heated to 70℃. The hydrolysis rate (decrease in T ⁰ _H ) is not affected by the temperature rise, but the rapid increase in T ¹ followed by the gradual increase in T ² indicates that the self-condensation reaction is greatly accelerated. Therefore, the authors report that, for example, when modifying an inorganic material with a silane coupling agent, the temperature must be kept as low as possible if self-condensation is to be suppressed.

Then, on page 91, in "4. Conclusion," it is stated that "The optimal concentration of silanes hydrolysis was found to be around 10% (w/w) with respect to the solvents and the temperature raise enhanced drastically the self condensation reactions."
Therefore, it is believed that the abrasive grain E of Patent Document 3 could not be concentrated to 12 mass % due to the method of adding the silane coupling agent during the modification treatment.

The physical properties of the resulting colloidal silica were evaluated by the following methods.
(1) BET specific surface area, BET diameter: Measured using a NOVA4200e (manufactured by Anton Paar). The particle diameter is calculated by the formula 2727/S using the BET specific surface area S ( ^m2 /g) measured by the nitrogen adsorption method (BET method), the true density of _SiO2 (2.2 g/ ^cm3 ), and the above formula (1).
(2) Cumulant mean diameter by dynamic scattering method: Measured using an SZ-100 (manufactured by Horiba, Ltd.) When measuring, the silica content in the measurement sample was adjusted to 1.13 g with pure water and ammonium nitrate, and the adjusted liquid was measured.
(3) Silica solids concentration: Using an SMS-70 (manufactured by A&D Co., Ltd.) as an apparatus, the residue obtained after evaporating the water content was taken as the silica concentration.
(4) pH: Measured at 25° C. using an instrument D-51 (manufactured by Horiba, Ltd.).
(5) Viscosity: Measured at 25° C. using a VM-10A (manufactured by Sekonic Corporation).
(6) Methanol concentration: Measured using a GC-2025 (Shimadzu Corporation).

The colloidal silica of the present invention is suitable for applications such as abrasives (silicon wafers, hard disks, etc.), coating agents (eyeglasses, displays, building materials, paper, etc.), and binders (ceramics, catalysts, etc.).

Claims (13)

Colloidal silica characterized by having a modified surface, a BET diameter of 12 nm or less, and being substantially non-aggregated. 2. The colloidal silica according to claim 1, which has a BET specific surface area of 227 m ² /g or more. The colloidal silica according to claim 1, characterized in that the modification is anion-modified or cation-modified. The colloidal silica according to claim 3, characterized in that the anion modification is by sulfo groups. The colloidal silica according to claim 3, characterized in that the cationic modification is by primary amino groups. Colloidal silica according to claim 1, characterized in that the solid content is 12% by mass or more. Colloidal silica according to claim 1, characterized in that the solid content is 12% by mass or more, and the rate of change in the cumulant mean diameter measured by dynamic light scattering after being kept at 60°C for 7 days is within 20% compared to before the keeping. A method for producing the colloidal silica according to any one of claims 1 to 7, comprising the steps of:
a raw material preparation step of supplying and reacting a readily decomposable organosilicate to a reaction solution containing a hydrolysis catalyst made of an organic amine to prepare raw material colloidal silica having a BET diameter of 12 nm or less;
a concentration adjusting step of adjusting the solid content concentration of the raw material colloidal silica to 13% by mass or less and adjusting the concentration of alcohols generated in the raw material preparation step to 1 to 25% by mass;
a modification treatment step of modifying the raw colloidal silica having the adjusted concentration;
and a concentrating step of concentrating the modified colloidal silica so that a residual organic solvent in the modified colloidal silica is 1 mass % or less. The method for producing colloidal silica according to claim 8, characterized in that in the raw material preparation process, the reaction is carried out under conditions in which the feed rate of the easily hydrolyzable organosilicate is less than 1.5 mass%/min of the total amount of the easily hydrolyzable organosilicate added, the reaction time is 6 hours or less, and the reaction temperature is 70°C or less. The method for producing colloidal silica according to claim 8, characterized in that in the concentration step, the colloidal silica after the modification treatment step is concentrated so that the solid content concentration in the colloidal silica is 12 mass% or more. The method for producing colloidal silica according to claim 8, characterized in that the modification treatment step includes a step of reacting a modifier having a functional group that can be converted into an anionic group with the colloidal silica after the concentration adjustment step, and a step of converting the functional group in the modifier after the reaction into an anionic group. The method for producing colloidal silica according to claim 11, characterized in that the modifying agent has a mercapto group and/or a sulfide group, and the mercapto group and/or the sulfide group are converted to a sulfo group by treating with an oxidizing agent, thereby obtaining colloidal silica having sulfo groups on the surface. 9. The method for producing colloidal silica according to claim 8, wherein the modification treatment step comprises a step of reacting a modifying agent having a cationic group with the colloidal silica after the concentration adjustment step.

Images