C H APTER - Ill

THE ROLE OF THE CATALYST

IN SOL-GEL PROCESSING

OF SILICA GLASS

CHAPTER Ill

THE ROlE OF THE CATAlYST IN SOl-GEl

PROCESSING OF SiliCA GlASS

3.1 INTRODUCTION

The first recorded sol- gel synthesis of silica was conducted

in 1866, and since then sol-gel processing of inorganic solids has been the

subject of widespread among the material scientists (Friedel and Craft 1866,

Post 1943). Despite this fact, however, minimal understanding of the actual

mechanisms of hydrolysis and polymerization in the presence of different

catalysts has been developed. Unlike conventional glass and ceramic

synthesis, in which powders are reacted at high temperature, the sol- gel

process relies on a low temperature condensation reaction in liquid solution,

similar, to that utilized in the manufacture of some organic polymers.

In the sol- gel processing of silica, a silicon-containing raw

material, a solvent, water, and a catalyst are generally utilized. By accepted

definition, a constituent in a chemical reaction is considered a catalyst if :

1) The catalyst is unchanged chemically at the end of the reaction;

2) A small amount of catalyst is sufficient to bring about a

considerable extent of reaction, and ;

3) The catalyst does not affect the position of equilibrium in a

reversible reaction.(ller 1979)

A number of papers have been published on the effects of

pH on the properties of gels and the effect of rapid vs. slow hydrolysis, a

function of both pH and the water content of the solution.(Brinker et al

1982, Nogami and Moriya 1980, Yamane and Kojima 1981, Zarzycki 1982.

Klein and Garvey 1980, Mukherjee 1980, Rabinovich etal 1982, Sakka and

Kamiya 1980, Yamane and Okano 1979, Majumdar and Mahajan 1999,

Majumdar and Singh 1998). For example, Brinker reported that " under

comparable conditions, base- catalyzed hydrolysis proceeds much faster than

the acid- catalyzed reaction." (Brinker et al 1982). In a subsequent paper,

however, Brinker conducted another series of experiments from which he

postulated that base-catalyzed hydrolysis proceeds more slowly relative to the

polymerization reaction than acid- catalyzed hydrolysis. (Brinker et a! 1984).

Table 3.1 presents some examples of previous work in the authors'

conclusions under the column heading "comments".

No one has conducted a systematic study of the effects of a wide

variety of catalysts on sol-gel processing of silica under standardized

conditions. The gelation process of metal- alkoxides involves both hydrolysis

and polymerization reactions is complex, depending not only upon pH, but

also upon the reaction mechanism(s) of each catalytic agent. The aim of

this work is to systematically examine a range of catalysts to provide a

foundation upon which future work can be predicted. The objectives are

•

•

•

To determine how different catalysts affect gelation rate

To determine how different catalysts affect the properties of dried

and fired gels

To propose a tentative reaction mechanisms for the catalysts

considered.

3.2 GEL PREPARATION

In order to conduct a systematic study of the role of

catalysts, a standardized solution composition was utilized for all of the

results presented. The compositional ratio was four moles of enthanol and

four moles of water to one mole of tetraethoxysilane. Table 3.2 presents

this standard solution composition. The ethanol used in this study was

dehydrated, 200 proof supplied by Aldrich Chemicals Company U.S.A. The

tetraethoxysilane was 99.9 percent purity from Alfa Chemicals Ventuon

D1vision. The water was high purity, distilled, and double- deionized provided

by the Department of Chemistry, Government Science College, Raipur Figure

3.1 schematically shows the procedural steps in the preparation of the gel

samples produced in this study.

After mixing, solutions containing catalysts were allowed to

gel at 25°C. The time of gelation was measured by the "standardized

time of penetration method." It is important to determine exactly when a

solutions gels and, therefore, select some arbitrary viscosity value that can

be defined as the "gelled" state. In order to quantitatively determine when a

solutions gels, a viscosity of 10,000 poise was selected. A viscosity probe,

consisting of a thin glass rod was developed. A callibration curve for this

probe of viscosity vs. time penetration was prepared using Brookfield

viscosity standards, molten sucrose, and molten glucose at fixed

temperatures. A mark one inch from the tip of this probe represents the

fixed penetration distance. For a viscosity of 10,000 poise, the viscosity

probe requires about seventy seconds to reach the mark (Fig. 3.2).

After gelation, samples were air dried at 25°C in semi- open

samples containers until they no longer exhibited weight loss due to

evaporation of residual water, catalyst, and solvent. The semi- open

containers were constructed such that one percent of the surface area of

the container lid was exposed to the air. After drying, physical properties of

the gels were measured. These included bulk density, apparent density,

porosity, volume percent shrinkage upon drying, and Vicker,s hardness.

Subsequent heat treatment was conducted at 600°C for 18 hours. after

which the physical properties of the fired gels were also measured.

3.3 THE EFFECT OF THE CATALYST ON GELATION TIME

AND PHYSICAL PROPERTIES

In the previous section, the preparation procedures were

described for the silica gels produced in this study. In this section, the

effect of varying the catalysts on gelation time and such physical properties

as porosity, bulk density, apparent density, volume shrinkage upon drying,

and Vicker's hardness are presented and discussed.

In table 3.3, gelation times and apparent initial pH of

solution values are presented for six different catalysts at equivalent mole

concentrations in solutions of the silica standard composition. Three of the

acid -catalyzed solutions exhibit low initial pH of solution values and long

gelation times. Acetic acid shows a significantly higher pH in solution which

could be attributed to a lower degree of dissociation for acetic acid. This

lower degree of dissociation has been documented in the C.R.C. Handbook

and basic chemistry texts for a wide variety of aqueous and alcohol

solutions. Another possible explanation for acetic acid will be discussed in

the next section.The higher pH for HF, on the other hand, cannot be

attributed to a lower degree of dissociation. It has been observed that for

solutions in which additional HF has been added to bring the pH below

0.5, gelation occurs in less than three minutes. The pH of solutions values

were measured with a glass electrode with a Systronix digital pH meter.

Figure 3.3 shows the pH of solution for four different catalysts during

the first hour after mixing. Subsequent pH measurements were conducted

near gelation and the pH was found not to vary significantly after the first

hour. The profiles for HN03 and H2S04 match almost identically the HCl

profile (Fig. 3.3). Gelation times do not appear to correlate with the pH of

solution for different catalysts, despites the fact that it is widely recognized

that for the same catalyst, decreasing the pH of solution also decreases the

time of gelation.(Aelion et al 1950) An example of this is presented by

Yamane and Kojima for the Sr0-Si02 system.(Yamane and Kojima 1981)

Table 3.4 presents some of the properties of the gels

examined in this study. The 25°C values are for gels dried in air with no

heat treatment., and the 600°C values are for gels heat treated at 600°C

for 18 hours. The apparent density is the density of the matrix including

closed pores and the bulk density is the total density including both open

and closed pores. The bulk densities of the HF and NH40H catalyzed gels

are significantly lower than those of the four other gels in both the heat­

treated and unheat-treated samples.

In all properties measured, the HF catalyzed gels appear

similar to the NH40H catalyzed gels. The low porosities and high bulk

densities of the HCl, HN03 and H2S04, and acetic acid catalyzed gels

appear similar to separate results obtained by both Brinker and Zarzycki, for

HCl catalyzed gels, in what have been described as slow polycondensation

reactions. The high porosity and low bulk density of the NH40H catalyzed

gels has been attributed to hydrolysis due to nucleophilic substitution of OH

groups. (Brinker 1982) Despite the similarities in properties, the HF catalyzed

gels are transparent while the NH40H catalyzed gels are white/ opaque.

Figure 3.4 is a photograph of a highly transparent sample of HCl catalyzed

silica glass. Large plates, rods, and other more complex shapes have been

obtained from HCl catalyzed gels far more readily than any of the other

catalyzed silica gels so far examined. Unlike most typical gel samples

HF catalyzed gels can be heated directly from room temperature to

in a few hours with minimal possibility of cracking during the firing process.

In table 3.4, the Vicker's hardness was measured by the

micro- indentation method using a Tukon microhardness tester. The decrease

in hardness as a function of porosity seems to follow an inverse exponential

relationship similar to that proposed for the fracture strength of sintered

ceramics and metals produced by powder metallurgy.(Kingery et al 1976) In

fact, the data presented was used to develop the following quantitative

relationship between hardness and porosity.

H = H0 exp (-BP) ... ( 1)

where; H0 = 750 (measured for silica)

B 4.64

P volume fraction pores.

and H = hardness of heat- treated silica gel.

It is interesting to note that for strength vs. porosity, the

constant "B" ranges between 4 and 7 (Kingery et a! 1976). The constant

calculated for the hardness of silica gels falls within this range. The acetic

acid catalyzed gel, possessing only 1.9 percent porosity after heat treatment,

has a Vicker's hardness of 667, which is considerably harder than window

glass and almost as hard as melt cast fused silica.

In the present chapter, the effects of varying the catalyst on

gels have been presented. It is evident that the effect of the catalyst on the

gelation process dramatically affects the properties of the resultant gel.

Possible mechanisms and further evidence is presented in the following

sections.

3.4 MECANISMS Of CATALYSIS

In !hl' JHt'VHHI.., '>t't l1o11, tfu• qt"ll1tlon ltttll'' .1nd JHOJU'tllt'" ot

,tlu" W<'l<' Jll<''<'IJII•cl '"' '"lut1"11' 111 wluch only till' ',,t,1Jv,r w." '''"'"'"

Tlw 'i'JIIIfico~nt v.lll<llion '>f ql'l.1t1on tunes ,md propl'tlll'' '.umot lw t•xpl.lllll'cl

"''''''! on till' h."" of pi I <IIlli till' wl<ltiv<' dt•qr<•t• of d1"'"'l•1t1on of <'·H·h

co~t.IIV't Tlw p.u-twul.u 1<'<11'11<>11 ""'' h.mis111 <'tllploved l>v ''""" , •lt.th)'t 11111't

"'"' lw con'"'''rl'd. l.1k<'ly tllod<•l'. based upon informaiJon ,,!n•.Hly in till'

litPr<lture, 111 conJutKtlon w1th tlw t•xperimental rl'sults lwrein cont.litwd c.1n

be proposed

Many authors have already proposed il pH dependency that

has been typically associated with general acid and general base hydrolysis.

(Klein and GaiVey 1980, Mukherjee 1980, Rabinovich eta! 1982. Sakka and

Kamiya 1980, Yamane and Okano 1979, Majumdar and Mahajan 1999.

Majumdar and Singh 1998). Acid catalyzed hydrolysis is and electrophilic

reaction that can be expressed by equations 2(a) and 2(b).

(R0) 3Si0R + Hp+ -------- (R0) 3SiOH + H+ + HOR . (2a)

... (2b)

In this kind of hydrolysis reaction, the reaction rate is

governed by the concentration of hydronium in solution. In base catalyzed

hydrolysis, a nucleophilic substitution of hydroxyl ions for OR groups occurs.

This reaction is presented in equations 3a and 3b.

(R0)3SiOH + H30+ -------- (R0)3SiOH + OR­

OK + H20 -------- HOR + OH-

... (3a)

... (3b)

Analogous to the acid catalyzed reaction, the rate of the

base catalyzed hydrolysis is a function of the hydroxyl concentration in

solution. (Aelion et al 1950)

The mechanisms proposed in equations 2 and 3 appear to

apply when one only considers HCl vs. NH40H catalyzed reactions, as has

been the case in most of the silica sol-gel literature. Expanding the study

to include more catalysts reveals that other factors besides pH alone

contribute to the gelation times and properties of silica gels prepared under

comparable conditions.

In the case of HF catalyzed reactions, it is evident that the

anion must play a significant role in the gelation process. In order to

confirm this observation, silica gels were prepared using equivalent molar

concentrations of potassium halide salts (Table 3.5). In only one of these

cases, using potassium chloride, did the salt appear not to dissociate fully in

solution, as evidenced by the salt remaining at the bottom of the beaker.

This is also the only instance in which solution immiscibility was observed

for this standard silica solution composition. The gelation times for the KBr

and KI catalyzed gels were three orders of magnitude greater than for the

case of KF catalyzed solution. The KF catalyzed gel was white and opaque

while the KBr and Kl gels were transparent. In addition to having the most

rapid gelation time, the KF solution exhibited a slightly basic character while

the KBr and KI solutions were mildly acidic.

Further confirmation of this trend can be seen in table 3.4,

comparing the gelation times for solutions catalyzed by four different acid

halides. Once again, the solution containing fluorine gelled most rapidly.

The hydrolysis and polymerization reactions presented in figures 3.6 and 3.7

are responsible for the rapid gelation of fluorine catalyzed gels. In the

hydrolysis reaction, a fluorine anion approaches a molecule of

tetraethoxysilane in solution forming a highly unstable pentacovalent activated

intermediate. This complex rapidly decomposes to form a partially fluorinated

silicon alkoxide plus by- products of water and alcohol. Another

pentacovalent complex is formed in the presence of water that decomposes

in to a partially hydrated silicon alkoxide plus regenerated fluorine and

hydronium. This process can continue until nearly all of the ethoxide bonds

are replaced by OH. More than likely, however, the polymerization process

begins before all of the tetraethoxysilane is hydrated.

The polymerization process, which for simplicity will be

modeled with a completely hydrated silicon, required a hexacovalent

intermediate. This hexacovalency is required in order for the neighboring

monomer species to approach the silicon. Jler postulates that the

effectiveness of the fluorine in th epolymerization reaction, whichhe proposed

for silicic acid, is due to the smaller ionic radius of the fluorine anion

versus that of the hydroxyl, which performs the same function of temporarily

increasing the coordination of silicon.(Iler 1979)

It has been observed that nucleophilic catalysis of the gelation

reaction tends to microstructures of large spherical particles of silica with

large porosities and pore diameters. In as much as the proposed fluorine

catalysis mechanism is also a nucleophilic substitution reaction, it is not

unreasonable to expect that a similar structure might also be observed,

Figure 3.8 is a scanning electron micrograph of an HF catalyzed gel fired

to 700°C for 18 hours. This photograph shows a well polished surface with

a void in the center. Inside this void, many spherical particles of

approximately 1000 A in diameter are present. The polished surface also

reveals these particles but with mush less clarity.

In table 3.3, the apparent initial pH of solution and gelation

times were presented for six different catalysts. In the case of acetic acid,

it was pointed out that the pH of solution was significantly higher than for

HCl, HN03 , and H2S04 catalyzed solutions (3. 70 vs. 0.05). One obvious

explanation for this difference might be the lower degree of dissociation

documented for acetic acid.

The fact that the gelation time was shorter than these other

three catalysts would suggest that another possible explanation is necessary.

In fact, anionic substitution of the acetyl radial has been documented for

hydrolysis involving silicon alkoxides, including tetraethoxylilane (Andrianov

1965). The difference between this and other catalytic reactions is that true

catalysis, as defined in the introduction, would not be occuring (Figure 3. 9).

In this reaction, the acetyl groups are being consumed in the reaction to

form ethyl acetate as a by- product. This mechanism would also explain the

significantly higher pH observed for acetic acid catalyzed solutions.

Although traditional acid and base catalyzed gelation

reactions, typified by HCI and NH40H, have been discussed in the literature,

these reaction mechanisms cannot be dismissed. The total rate of gelation is

a functions of both the hydrolysis and polymerization reactions. The slowest

of these will be the rate determining step. Aelion, et a!., have shown that

the concentration of HCI and base (Aelion 1950). As a consequence, the

rate of hydrolysis without a catalyst and, therefore, the rate of gelation, is

extremely slow (Table 3.3). Iler demonstrated that the polymerization rate of

silicic acid, a fully hydrolyzed silicon, in an aqueous solution exhibits a

complex sinusoidal shape due to the interaction of several competing

processes. (Jler 1979)

In Iler's model for the gelation of silicic acid, three main

factors were considered to explain the roughly sinusoidal rate of gelation (i.e.

polymerization). These three factors are :

•

•

•

The isoelectric point (lEP) for silica in a stable solution is about

pH=2

At low pH, trace impurities of fluorine, introduced with the acid

catalysts, will greatly accelerate the polymerization reaction, and

As the OH concentration in solution is increased, the

electronegative surface charge on particles suspended in solution

will increase, thereby, retarding gelation.

Neglecting the differences in solvent medium between the

;stems studied by Aelion, ller, and the standardized solution herein

iscussed, the generalized inverse rate of reaction vs. pH of solution model

f Figure 3.10 can be proposed. This model applies for systems in which

eneral acid and general base catalysis are predominate.

Moreover, the relative rates of these reactions to one another

tnd, hence, the total rate of gelation, will be affected by such factors as

olution concentration, water content, and the type of silicon-containing raw

naterial utilized. The dashed lines in Fig. 3.10 represent the rate of reaction

iSsuming no trace impurities of fluorine introduced into the solution. Since

Jnly a few parts per million fluorine can affect the rate of reaction

:lramatically, the "with impurity" reaction rate has been drawn solid

(ller 1979).

In total, twelve different catslysts and their effect on gelation

have been examined in this study. These catalysts fall into four distinct

categories:

• general acid catalysts

•

•·

•

general base catalysts

salts

fluorine containing compounds .

Based upon the information in the literature in combination

with the experimental evidence and models in this study, a summary of

catalysts and their principle catalytic mechanisms in the hydrolysis and

polymerization of tetraethoxysilane are presented in table VII. lt is evident

that catalytic mechanisms based upon pH alone are inadequate to explain

the gelation process.

3.5 EFFECT OF THE CATALYST ON PORE AND CAPILLARY

EFFECTS DURING DRYING

It has been obsered during the drying of gels, that the

surface tension and vapor pressure of the solvent and pore diameter of the

gel have a pronounced effect upon the volume shrinkage and porosity of the

resultant gel. (Zaezycki 1982) It has also been observed that the choice of

the catalyst produces a wide variation in these same properties. The

purpose of this particular section is to explain these observations based upon

the effect of the catalyst and pH on the gelation process, as developed in

the preceding section.

The final shrinkage and porosity of a gel can be viewed as

the result of two opposing forces; the capillary pressure produced by the

volitilization of the liquid residuals pulling the gel matrix towards collapse

versus the ability of the matrix to resist deformation due to its rigidity, as

determined by its extent of crosslinking and mcrosturctural stiffness.

Fig. 3.11 shows these interrelated processes, and the main factors that

affect them.

Fig. 3.12 presents an idealized schematic of a fluid inside a

cylilndrical pore. The capillary pressure inside the pore is described by

equation 3.4.

where

2y,1 cos e P=----

' r

Y = liquid-vapor surface tension vi

r = the pore radius

and e = the wetting angle.

.... (3.5)

The welting angel can be determinde knowing the liquid­

vapor surface tension, the liquid-gel surface tension, and the vapor-gel

surface tension using equation 3.5.

cosO=---- . ... (3.6)

In chemical handbooks, however, only the liquid-vapor surface

tension is given. Values of the other two surface tensions for silica gels and

different solvents do not currently exist in the literature. An approximation

suggested by Zarzycki is to assume complete wetting, acondition for which

only the liquid- vapor surface tension os required (Zarzycki 1982). For

Zarzycki's approximation, equation 3.4 reduces to a simplified form given by

equation 3.6.

2 Yvg

P,= --­r

.... (3. 7)

ln actual cases, however, it is not realistic to expect the wetting angle to

approach zero for all liquids.

It is appropriate, at this point, to relate the catalytic

mechanisms of the gelation process and tje resulting microstructures with

the properties of the dried and fired silica gels. Many authors have

postulated a fine network structure of linear chains for acid catalyzed gels

and more dense spherical particles with large interstices between them for

base catalyzed gels. (Brinker et al 1982, Nogami and Moriya 1980, Yamane

and Kojima 1981, Zarzycki 1982, Klein and Garvey 1980, Mukherjee 1980,

Rabinovich eta! 1982, Sakka and Kamiya 1980, Yamane and Okano 1979,

Majumdar and Mahajan 1999, Majumdar and Singh 1998).

In view of Fig. 3.10, which combines ller's results on the

effect of pH on the polymerization reaction with Aelion's relationship on the

effect of catalyst concentration on hydrolysis, the microstructural differences

between acid and base catalyzed gels can be explained. For pH values

above the isoelectric point of the solution, which is approximately 2 for

silica, increasing the OH concentration increases the rate of

polymerization.This would be expected to produce a continuous decline in

gelation time with increasing OH concentration. This decrease in gelation

time,however, reaches a minima due to repulsive charge build- up on the

particulate clusters. As pH is increased beyond this minima point, the size

of the sol particles must corrospondingly increase in order to overcome this

repulsive surface charge effect (Her 1979). This results in gels with

microstructures consisting of large spherical particles for high pH catalyzed

solutions. In the extreme case, precipitation may occur before gelation.

In the low pH regime, linear chain growth seems to be

prefered. This may be the result of higher reactivity at chain ends

promoting continued linear growth due to a lower mass density that allows

for easier approach of reactive groups. Simply stated, the chain ends are

less crowded. A more sophistcated approach would be to suggest that the

electron charge density at the chain ends is perturbated, resultion in a

higher electropotential for reaction.

In Fig. 3.13, idealized drawings of acid and base catalyzed

microstructures are presented with comments. For acid catalyzed gels. a

network structure of linear chains with low cross-linking and, therefore, low

stiffness is formed. These thin chains have a greater degree of freedom to

bend, rotate, and plastically deform due to capillary forces, as discussed

earlier, than the interconnected spheres obtained for base catalyzed gels.

The large interstices formed by the packing of these spheres provide an

easy path by which volitiles can be removed.

In the present section we have made an altern! to explain,

the process by which volatilization of residual liquids in the gel results in

shrinkage. A better understanding of how the wet gel microstructure, as

determined by the catalyst employed in the hydrolysis and polymerization

reactions, affects the volume shrinkage during drying has been discussed.

The role of the catalyst in the sol- gel processing of silica

has been examined in a systematic approach. The catalyst affects both the

hydrolysis and polymerization reactions. The slowest of these reactions is the

rate determining step. Different catalysts produce significant variations in the

gelation time and properties of gels. Gelation times and properties do not

solely depend upon the pH of solution, but also depend upon the catalytic

mechanism in both the hydrolysis and polymerization reactions.

In the case of fluorine compounds, nucheophilic substitution of

fluorine catalyzes both the hydrolysis and polymerization reactions. For

general acid and general base catalyzed reactions, the rate of gelation vs.

pH of solution is complex, depending upon the relative rate of hydrolysis,

the isoelectric point of solution, electrorepulsive surface charge build- up on

particles with increasing OH concentrations, and trace impurities of fluorine

as introduced by the acid catalyst. It has been demonstrated that the

catalyst and pH of solution affect the microstructure of the wet gel and,

therefore, the volume shrinkage and drying properties of the gel.

PROCEDURAL STEPS IN PREPARATION OF GELS

MIX SOLUTION OF WATER,

ETHANOL, AND CATALYST

DECANT SOLUTION INTO

MEASURED QUANTITY

OF TEOS

MONITOR VISCOSITY BY

"STANDARDIZED TIME OF

PENETRATION METHOD"

AIR DRY AT 28°C

HEAT TREATMENT AT 700°C

FOR 24 HOURS

MEASURE PROPERTIES

FIG 3.1 SCHEMATIC REPRESENTATION OF STANDARDIZED GEL PREPARATION PROCEDURE.

1 X 106

5 X t05 VISCOSITY TIME (SEC)

t23 p 2

600 p 5

1000 p 9

J X 105 6600 p 40

28000 p 130 s x to• 500000 p 1980

&i' [JJ

0 !::-;... t x to• t: [JJ

0 u :!2

5 X t03

>

I X t03

s x to'

• lxt02 L-~------~---L--------~--_.---------L----L---~ l 5 10 50 100 500 1000

TIME OF PENETRATION (LOG SCALE)

FIG3.2 VISCOSITY VS. TIME OF PENETRATION STANDARDIZATION CURVE FOR VISCOSITY PROBE.

10.0

9.0

8.0

7.0

;.: 0

6.0 -F-;:, ...l 0 "' "" 5.0 0 :I: HOAC Q.

4.0 ~ ..--- • • •

3.0 HF

2.0 v 1.0

HCL ... 0.0 J

05 10 30 60

TIME (MINS.)

FIG. 3.3 APPARANT PH VS. TIME FOR SELECTED CATALYSTS.

I

FIG. 3.4 PHOTOGRAPH OF HCI CATALYZED PLANE

-----~~~~~~~~------

800 -(7)

(I) NH40H

(2) HF 700

(6) (3) H,so,

1:1: (4) HN03

"' 600 (5) HCI ~

:E ~ (6) HOAC z <Fl 500 (7) Fused Silica <Fl • (4)

,lj "' z • ~ 1:1: 400 < :t <Fl 1:1:

"' 300 ::.:: u -> • (3)

200

100

0 10 20 30 40 50 60 70 80

PERCENT POROSITY

FIG 3.5 \"ICKERS HARDNESS VS. POROSITY FOR 700°C HEAT TREATED SILICA GELS.

'

p 0

F "').I_ RO -S1- OR

I 0 R

Silicon Alkoxide Plus Fluorine

- F -H2 o H - ' I

' .

.,..

- --~ o---si-oo I

H I \

I \

0 0 R R

Pentacovalent Complex (SNi-Si Reaction)

( - )

+

R 0 I ( + H3o J

R 0

I F---Si--- OR

II 0 0 R R

Pentacovalent Activeted Complex (SN2-Si Inversion Reaction)

<+H2o J ~

R 0 I

HO- Si- OR I 0 R

1-

F- Si- OR-tHOR+ H20 I 0 R

Partially Fluorinated Silicon Alkoxide Plus Alcohol and Water

+ H30 + F

Partially Hydrated Silicon Alkoxide Plus Regenerated Fluorine Anion

and Hydronium

FIG 3.6 MECHANISM OF FLUORINE CATALYZED HYDROLYSIS

-H F

~r HO- Si -OH

~\ H H HO 0

" / Si

H 0/ " OH

Monomer (Silicic Acid) Plus Fluorine and Another Monomer

--+

H H 0 0 ( ~)

' ,

' / HO- -Si- -F

/ ' H()" 'oH /OH

" Si

H 0/ 'o

H

Partially Activated Intermediate Dimer Ion

FIG 3.7 FLUORINE CATALYZED POLYMERIZATION

H H 0 0 I I

HO-Si-0-Si-OH I I 0 0 H H

.,.F + H20

Disilicic Acid Plus Regenreted Fluorine Anion and Water

(-HORl

Pentacovalent Compound Formation (S,i-Si reaction)

(R ~ C2H

5-)

R CH3 I I

0- -C = 0 I I R-- o

I RO- Si -OR

I 0 R

Reaction of Ethanol with Triethoxyacetoxysilane

0 • RO 0-C-CHJ " / Si

RO/ "oR

Formation of Triethoxyacetoxysilane

H 0

I

( +HORJ

RO- Si -OR + CH3 COOC2H5 I 0 H

Partially Hydrated Monomer Pins by Product

of Ethyl Acetate

FIG 3.9 POSSIBLE CATALYTIC MECHANISM OF ACETIC ACID.

0

I I I I I I I~ I

Total Rate of Reaction / E (Hydrolysis+ Polymerization)/ ~ I

I ~ I 1/ ~ J ~

= Without Impurities ~ ~

'/ ~

1

Rate of Hydrolysis

Rate of Polymerization

\

"

pH OF SOLUTIONS

14

FIG 3.10 IN\'ERSE REACTION RATES FOR GENERAL ACID AND GENERAL BASE CATALYZED HYDROLYSIS AND POLYMERIZATION.

I

FORMATION OF RIGID STIUICTliRE

Factors

• Extent of cross-linking during

gelation.

• development of particulate vs.

linear cham structure.

• amount of cross-linking within

network structure.

II

SHRINKAGE DUE TO

VOLATILIZATION OF

RESIDUAL LIQUIDS

Factors

• vapor pressure of solvent.

• evaporation rate as affected

by external environment.

• pore size effect on capillary

pressure.

• residual liquid surface

tension effect on capillary

pressure.

FIGURE 3.11 TWO INTERRELATED PROCESSES AFFECTING

SHRINKAGE DURING DRYING.

Y,, Vapor - Liquid Surface Tension

Y,, Vapor (v) Cos 9 Yvg y,, + "(,.,

2 Y,, Cos 9 Y,, P, r

'Yvl r Pore Radius

e Wetting Angle Gel (g) Liquid (I)

Pore Wall

FIG 3.12 CAPILLARY EFFECT IN GELS.

ACID CATALYZED

• network structure of linear chains with low functionality and low stiffness.

• greater freedom to bend, rotate and plastically deform.

• fine pore structure that resists flow of volatiles.

BASE CATALYZED

• spherical particles of high functionality and stiffness.

• large interastices between spheres that allow volatiles to escape during drying.

• freedom to deform primarily limited to abilioty of spheres to compress.

FIG 3.13 COMPARISON BETWEEN THE MICROSTRUCTURES OF ACID AND BASE CATALYZED GELS.

TABLE 3.1

SURVEY OF PREVIOUS WORK

AUTHOR(S) SYSTEM CATALYSTS COMMENTS

Brinker Si02 HCl Slow hydrolysis relative to the (J.N.C.S.) condensation reaction. Small 1982 pores.

NHpH Rapid hydrolysis relative to the condensation reaction. Large pores.

Nogami Si02 HCl High bulk density. No particles Moriya observed. (J.N.C.S.) 1980 NH40H Low bulk density. Spherical

particles observed.

Zarzycki Si02 Catalyst The rate of gelation is (J.M.S.) used but not proportional to the OH 1982 specified concentration above pH=2,

and proportional to the H+ concentration belowpH=2.

Klein Si02 HCl Accelerated hydrolysis and Garvey retarded polymerization. (Soluble Silicates) NHpH "Limited hydrolysis"

1982 NH4Cl "Postponed hydrolysis"

followed by rapid gelation

Yoldas Al20 3 16organic Fvaluated peptizing effects of

(Ceram. and inorganic acids on slurry of AI( OHh

Bull.) acids

1975

I

TABLE 3.2

STANDARD SILICA GEL SOLUTION COMPOSITION

CONSTITUENT CONCENTRATION WEIGHT (Mole : TEOS) PERCENT

Tctracthoxysilanc (TEOS) I 44.8

Ethanol 4 39.6

Water 4 15.6

Catalyst 0.05 ------

TABLE 3.3

GELATION TIMES AND APPAREMT INITIAL pH OF SOLUTION FOR SIX CATALYSTS

CATALYST CONCENTRATION APPARENT GELATION (Mole : TEOS) INITIAL pH TIME(HR)

OF SOLUTION

HF 0.05 1.90 12

HCI 0.05 0.05* 92

HN03 0.05 0.05* 100

H2S04 0.05 0.05* 106

HOAC 0.05 3.70 72

NH40H 0.05 9.95 107

No Catalyst ----- 5.00 1000

*Between 0.00 and 0.05

TABLE- 3.4

PROPERTIES OF GELS DRIED AT 25°C AND 600°C FOR DIFFERENT CATALYSTS

25°C PROPERTIES 25°C PROPERTIES

S.No. Catalyst Volume Bulk Apparent Percent Volume Bulk Apparent Percent Vickers

Shrikage Density Density Porosity Shrikage Density Density Porosi~· Hardness

(%) (grnlcc) (grn!cc) (%) (grn!cc) (gmfcc)

L HOAC 84 132 133 0.7 ----- 2.08 2.12 L9 666.5

2. HCI 81.3 ----- ----- ----- 85.2 2.06 2.12 2.8 429

3. HN03 79.9 1.14 1.16 !.7 85.2 !.82 2.02 10.0 470

4. H2S04 7!.6 ----- ----- ----- 80.0 !.46 2.12 3!.0 224

5. HF 78.4 0.54 !.24 56.7 82.7 0.71 2.13 67.0 75

6. NH40H 67.8 0.49 L13 57.0 7L7 0.70 2.21 68.0 28

7. No 87.5 0.95 2.09 54.6 ----- !.25 2.21 43A -----

Catalyst - -

TABLE 3.5

GELATION TIMES AND INITIAL pH OF SOLUTION VALUES FOR SILICA GELS CATALYZED BY POTASSIUM HALIDES

CATALYST CONCENTRATION APPARENT GELATION COMMENTS (Mole : TEOS) pH TIME

KF 0.05 8.50 6mins. Salt fully dissolved in solution

KCI 0.05 4.70 48 hrs. Immiscible solution

KBr 0.05 3.50 I 00 hrs. Clear solution, salt fully dissolved

KI 0.05 3.40 ISO hrs. Clear solution, salt fully dissolved

- -

TABLE 3.6

GELATION TIMES AND INITIAL pH OF SOLUTION VALUES FOR SILICA GELS CATALYZED BY ACID HALIDES

CATALYST CONCENTRATIOINI INITIAL GEJ,A TIOINI (Mole : TEOS) pH TIME

HF 0.05 1.90 12 hrs.

HCI 0.05 0.05 92 hrs.

HBr 0.05 0.20 285 hrs.

HI 0.05 0.30 400hrs.

TABLE 3.7

CATALYSTS AND THEIR PRINCIPLE CATALYTIC MECHANISM IN THE HYDROLYSIS AND POLYMERIZATION REACTIONS OF T.E.O.S

CATALYST HYDROLYSIS POLYMERIZATION (CONDITIONS) REACTION REACTION

HF, KF Nucleophilic substitution ofFiuorine Anion Nucleophilic substitution ofFiuorine (all pH) (SN2-Si) Anion (SN2-Si)

HCI, HBr, HI and Electrophilic reaction ofHydronium Ion Nucleophilic reaction ofHydronium NH03 (above pH=2) Ion

H2S04, HOAC Possible SNi-Si reaction in addition to " (abovepH=2) electrophilic substitution ofHydronium Ion !

NH40H Nucleophilic substitution ofHydroxyl Ion " I KCI,KBr,KI Nucleophilic substitution of respective anion " (abovepH=2) (SN2-Si)

HCI, HBr, HI, KCI, Nucleophilic substitution of fluorine introduced " KBr, KI, HOAC, tbrough trace impurities in addition to HN03, H2S04 mechanisms postulated for higher pH values. (below pH=2)

I

.._ ' .

~