REVUE DE PHYSIQUE APPLIQUÉE
Colloque C4, Supplément au n°4, Tome 24, Avril 1989 C4-23
FREEZE DRYING OF SILICA GELS PREPARED FROM SILICIUMETHOXID
E. DEGN EGEBERG and J. ENGELL
Industry, Technical University of Denmark, DK-2800
Résumé - Les gels étudiés ont été prépares par une hydrolyse en cata-
lyse acide de Si(0Et>4 dilué.; dans Bu^OH. Les acides employés sont H3PO4 et
acide oxalique. L'acide oxalique fonctionne aussi comme DCCA. Le séchage a
froid des gels sans échange du solvant resuite en une désintégration en
morceaux de l'ordre du millimètre. La détérioration du gel devient moins
importante quand la concentration de EtOH dans la phase liquide est réduite par
l'échange du solvant. Une série d'expériences indiquent que le Bu^-OH sature en
air montre une expansion (5 vol%) durant la congélation â l'opposé du Bu^-OH pur
qui ne montre qu'un petit changement de volume. Cette expansion du volume est
probablement responsable de la fracture des pièces monolithiques.,
Abstract - The freeze dried gels were prepared by acid catalysed hydrolysis of
Si(0Et)4 diluted by Bu^H. The acids used include H3PO4 and oxalic acid.
Oxalic acid has DCCA-properties. Freeze drying without prior solvent exchange
results in disintegration of the gels into mm size flakes. The fracturing of
the gels becomes less severe as the concentration of EtOH in the liquid phase
is reduced by solvent exchange. Pure Bu*-OH shows only a very small volume
change upon crystallization, but air saturated Bu*-OH expand about 5 vol% upon
freezing. This volume expansion is presumably the cause of the crack-pattern
observed in the solvent exchanged monolithic pieces.
The drying process is a critical step in the preparation of sol-gel derived monoliths (15,
23). During drying stress caused by capillary forces in the pore structure of the gel results
in shrinkage and fracturing unless special precautions are taken.
Monolithic silica xero-gels (ca. 10 cm-^J, with bulk densities in the range 1.5-1.8 g/crrH,
can be produced by controlled conventional drying in about 5 days using acid-catalysed hydro-
lysis of Si(0Et)4 and a relative high casting density (9). The use of Drying Controlling
Chemical Additives (DCCA) like formamide, oxalic acid or glycol permit a relatively rapid
drying (5, 18). Thus, monoliths up to >100 cirH in size of silica xero-gel with bulk densities
in the range 1.2-1.4 g/cm^ have been made within 2 days using formamide and acid catalysed
hydrolysis of Si(0Me)4 (5). The use of DCCA results in the formation of a narrow pore size
distribution by elimination of the smallest pore (1, 5, 18). Hereby the capillary forces are
reduced and the drying process facilitated. Mizuno et al. (12) have recently succeeded in
producing small monoliths of silica xero-gels with bulk densities as low as 0.73 g/cm^ by
conventional drying of Si(0Me>4 derived gels. Contrary to conventional wisdom these results
were obtained by aqueous solvent exchange before drying. This was done in order to strengthen
the gel by eliminating organic groups and thus, favour crosslinking.
Monolithic silica aero-gels with bulk densities in the range 0.02-0.3 g/cm^ can at present
only be produced by supercritical drying (6, 16, 19) where the liquid-vapour interface is
eliminated and capillary stresses thereby avoided.
The present work concerns the preparation of monolithic silica gels by freeze drying, i.e.
drying by sublimation of a frozen solvent. Freeze drying is used on a large scale in food and
medical technology (21). The technique has also proved very useful for laboratory preparation
of high quality ceramic powders (3, 14, 22). In principle the process makes it possible to
conserve the structure of the material being dried. However, monoliths can not be made from
aqua-gels by this method. The freezing process results in the formation of small flakes, or if
directional freezing is used fibres (11). Important factors are here, firstly the large volume
expansion upon freezing of water (10 vol%) and secondly the surface tension between the phases
Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1989404
REVUE DE PHYSIQUE APPLIQUEE
Table 1 . Freezing point (20) and calculated
upper limit for the rate of subli-
mation (Gmax) for water, BU~OH and
Table 2 . Physical properties of ButOH
Fig. 1 . Density of BU~OH
: Kuss, E . (10).
+ : Beilstein (2).
x : Shell (17).
: Air saturated (17).
0 : Air saturated (this work).
as a function
Structural Formula CH3- C -OH
~elting Point (Phase I)
Crystalline Polymorphs (7, 13)
Density, g/cm3 (10; 2 &
new data Fig. 1)
80 OC< T < 25.6OC
2 5 . b ° C <
p = 0.8095 - O.OO1lxT°C
T < 0 OC p = 0.8066 - O.OOIOxT°C
Viscosity (10) 3 0 ° C
Vapor Pressure (20)
Torr 1nP =22.9308 - 5764.2/T°K
-10 0 10 20 30 40 50 60 70
involved. The latter affects the degree to which gel-material is expelled from the growing
"solvent"-crystals during freezing.
Methanol (MeOH) and ethanol (EtOH) are not suitable for freeze drying because of their low
freezing points and sublimation rates (Table 1 ) . However, tert-butanol (BU~OH) appears to be
very suitable (Table 2 ) . The freezing point of pure BU~OH is 25.6OC and the volume change upon
crystallization of air free material appears to be very small (Fig. 1 ) . Gmax, the upper limit
for the rate of sublimation neglecting the effects of pore structure, thermal conductivity and
other physical limitations, can be calculated using a simplified formula ( 8 ) . These
calculations (Table 1) indicate that for any given material the drying time can be reduced
substantially if BU~OH is used instead of water. At temperatures above O ° C three crystalline
polymorphs of BU~OH
are known (7, 13). The freezing point of BU~OH decrease in the presence of
impurities (H20, EtOH, BuSOH, BU~OH). Furthermore, air saturated BU~OH shows a volume
expansion upon freezing (5-6 ~01%. 17 & Fig. 1 ) . Thus, in order to avoid melting and cracking
in the monoliths it is important to use high purity air-free BU~OH.
2 . 0 w t %
7 . 1 wt%
39.6 w t %
BU~OH + Hz0
3 0 . 9 + 11.7 wt%
Mixed at 4 0 ° C -
Stirred at 2 0 ° C
2 0 ° C , 10 min. I
CAST at 60°C in sealed
Gelation Time 1-2 h
>5 times, 60°C, 12 hours
removal of EtOH & Hz0
FREEZE DRYING I-i
24 h, 60°C, 0 . 1 Torr
Fig. 2 . Flow diagram showing the procedure used for the preparation of freeze dried
Monolithic silica gels were prepared by acid catalysed hydrolysis of Si(OEt)4 by the
procedure shown in Fig. 2 . The acids used include H3P04 and oxalic acid (C2H204). Two of the
acid groups on H3PO4 react immediately with Si(OEt)4 and form copolymers with one remaining
acid group (4). this prevent later evaporation of the acid. Oxalic acid was used for its DCCA-
properties. The chosen ratio between Si(OEtI4, H20 and BU~OH was inside the miscible area in
the ternary system (Fig. 4) in order to avoid phase separation. The EtOH released by hydro-
lysis lowers the freezing point of the solvent. Thus, solvent exchange with pure BU~OH is
necessary in order to ensure crystallization of the major part of the solvent above -18OC.
Analyses of the exchanged solvent by gas-chromatography showed that the chosen exchange
conditions (60°C, 12h) were sufficient for obtaining equilibrium.
After aging the gels were frozen either unidirectionally (O°C) or homogeneously in a deep-
freezer (-18OC) and freeze dried in a HETOSICC freeze dryer (CD 52-1). The gels were further
vacuum dried (60°C, 0.1 Torr, 24 h) in order to remove residual solvent.
The dry gels were characterized by He-Pycnometry (apparent density; Micromeritic Multi-
volume Pycnometer 1305). Bulk density and open porosity accessible to water were determined by
Archimedes method in water. From these data the open porosity accessible to He was calculated.
REVUE DE PHYSIQUE APPLIQUEE
A: Sample 8063
A: bample uuo5
B: Sample 8068
n: sampre uuau
Fig. 3 . Photographs of freeze dried silica monoliths reported in Table 3 .
Fig. 4 . Liquid immiscibility in the system Si ( O E ~ ) 4 - ~ u t ~ ~ - ~ 2 ~ - ~ ~ 2 ~ ~ ~ (pers.com. K . T .
Table 3 . Compositions, conditions and results of the freeze drying experiments on silica gels.
* Number of times gels has been exchanged (60°c, 12 hours); added volume of ButOH or
B U ~ O H - C ~ H ~ O ~
> volume of gel.
SAMPLE 1204 1904 0205
( CH3 ) 3COH
3 9 . 7
1 . 5
AGING at 60°C,
Linear Shrinkage, %
Residual EtOH, wt%
2 . 9
Condensor Temp. -50°C
After Vacuum Drying
24h, 60°C, 0 . 1 Torr
accessible to Hz0 %
accessible to He %
Organic Residual, wt% 73.4 40.0 30.0
RESULTS AND DISCUSSION
The results are listed in Table 3 and examples of the freeze dried materials obtained are
shown in Fig. 3 . All the gels were clear and homogeneous after gelation and aging. Attempts to
freeze a gel (1204; -18OC) without exchanging the solvent showed that EtOH-enriched liquid was
entrapped inside the frozen gel. During freeze drying at low pressure this liquid boil and the
gel disintegrate into small flakes. A reduction of the EtOH concentration to ca. 0.5 wt% by
solvent exchange appear to be sufficient to solve this problem (8063-8068).
The ButOH tend to form needle like crystals. Attempts to remove the EtOH by directional
freezing (zone refinement) prooved unsuccessful (1904 & 0205). Repeated freezing and thawing
of gels containing air saturated BU~OH results in development of an increasingly finer crack-
pattern. It is believed that this problem can be solved by using air-free BU~OH.
Gels prepared with oxalic acid, repeated solvent exchange and long aging time gave large
(6~3~0.5 cm) translucent pieces of low density gels even though air saturated ButOH was used.
It is encouraging to observe that the bulk densities of the dried gels are in the range 0.3-
0.5 g/cm3. Thus, with further improvements it should be possible to produce monolithic aero-
gels by freeze drying.