Professional Documents
Culture Documents
23 (2006) 94–100
Malcolm Yates*
(Received: 28 September 2005; accepted: 24 January 2006)
DOI: 10.1002/ppsc.200601015
Abstract
Ceramic based monolithic supports are used in many ary agents, which when removed by thermal treatments,
catalytic reactions, especially in environmental remedia- generate a higher porosity. This results in an increase in
tion to avoid pressure drop limitations that may arise in the interconnectivity of the pore structure of the final
the treatment of large effluent gas volumes. In these sys- material. However, an upper limit is reached since an in-
tems, the active phases or their precursors can be incor- creased porosity negatively affects the mechanical
porated into the paste prior to kneading, extrusion and strength of these ceramic materials. In this paper, the
calcination, in order to achieve a uniform deposition mechanical strength and porosity of ceramic catalyst
throughout the final structure. However, these types of supports are related to the primary particle size of the
incorporated monolithic catalysts can suffer from diffu- raw materials, initial composition and final heat treat-
sion limitations that necessitate the inclusion of tempor- ment temperature.
with 0, 10 or 20 wt.-% PGA incorporated. The other two sion of the ceramic pastes, has a fibrous structure
series were also produced with 0, 10 or 20 wt.-% of PGA, ranging from 0.2 to 2 lm in length, and 0.1 to 0.3 lm in
in which the ratio of clay and either titania or alumina diameter for the bundles of fibers [6]. In the composite
was maintained at 1:1 by weight. In all of these prepara- materials these fibers act as a matrix in which the alu-
tions, the raw material powders were premixed in a dry mina or titania particles can be dispersed and stabilized
rotary mixer to ensure homogeneity. This powder mix- [7,8]. The phase changes that the silicate binder under-
ture was then added to a kneading machine where a paste goes with heat treatment have been thoroughly studied
with adequate rheological properties was formed with [9,10]. Loosely physisorbed water is lost from 80–150 °C.
the addition of water. This dough was extruded as honey- Between 300 °C and 350 °C the loss of zeolitic water and
comb monolithic structures with parallel channels of water of crystallization associated with the collapse of
square section with a cell density of 8 cells cm–2 and a wall the internal structure by a rotation to produce anhy-
thickness of 0.9 mm. The honeycomb monoliths were drous sepiolite occurs. At this point the structure can
dried at room temperature and then heated in air to still be rehydrated back to its original state. Heating be-
150 °C, 500 °C or 850 °C at a heating rate of 3 °C min–1 tween 400 °C and 500 °C causes the loss of constitutional
and maintained at this final temperature for 4 h. The water and the material cannot be rehydrated. Further
shrinkage of the materials with drying and calcining was weight losses between 600 °C and 700 °C are due to the
also determined. This shrinkage was designated as the loss of water present as hydroxyl groups. A very well de-
dimensional stability, in terms of the percentage change fined exothermic peak observed at about 825 °C, is due
relative to the original dimensions of the paste extruded to the phase change to enstatite, although no further
from the die, before drying and heat treatment. weight losses are produced. The heat treatment tem-
Mercury intrusion porosimetry (MIP) analyses were em- peratures chosen in this work were 150 °C, 500 °C and
ployed to determine the pore size distribution and pore 850 °C, respectively. The lower temperature leads to the
volume over a range of ca. 100 lm down to 7.5 nm loss of any loosely bound physisorbed water, and while
diameter, utilizing CE Instruments Pascal 140/240 appa- it causes a significant shrinkage, it does not bring about
ratus, on samples previously dried overnight at 150 °C. any significant improvement in the mechanical strength.
Approximately 0.2 g of sample was accurately weighed After treatment at 500 °C, the structure is stable when
into the sample holder, which was subsequently out- exposed to water, and thus typical impregnation meth-
gassed to a vacuum of < 10 Pa at room temperature for ods can be used to incorporate any active phases. After
five min, before filling with mercury and commencing heat treatment at 850 °C, with the formation of enstatite,
the analysis. The pressure/volume data were analyzed the structure undergoes further shrinkage, and a signifi-
by using the Washburn equation [3] assuming a cylindri- cant enhancement occurs in the mechanical strength.
cal nonintersecting pore model, taking the recom- The MIP curves corresponding to the four starting mate-
mended values for the mercury contact angle of 141° rials used in this study are presented in Figure 1. It was
and surface tension of 484 mN m–1 [4]. The primary par-
ticle sizes of the four raw materials were determined
1.5
from analysis of the intrusion curves in the interval
relating to the filling of the interparticulate pore space,
assuming spherical particle geometry [5]. For the mono- 1.2
lithic samples, a single-piece two channels in width and
Pore Volume (cm3g-1)
observed that the clay binder gave rise to a bimodal was observed between ca. 5 lm–100 lm. The primary
pore size distribution. With diameters of greater than particle sizes of the raw materials estimated from ana-
about 300 nm, the porosity was due to the loose aggre- lyses of these curves were 90% < 0.4 lm, 0.8 lm, 40.8 lm
gate structure of the powder; which was removed when and 76.6 lm for the clay, titania, alumina and PGA,
formed as a monolith due to the compaction of the sam- respectively.
ple during wet kneading and extrusion. Below 300 nm, The results for textural and mechanical strength ob-
the porosity was due to the interparticulate and intrapar- tained with the nine monolith compositions are given in
ticulate porosities. The bimodal porosity observed for Table 1. The changes in the porosity of the clay after
the titania was mainly due to interparticulate porosity in conformation as a monolith by drying at 150 °C, and
pore diameters between 100 nm and 10 lm, and intra- heat treatments at 500 °C and 850 °C, are shown in Fig-
particulate porosity in diameters below 100 nm. For alu- ure 2. Conformation of the clay as a monolith elimi-
mina, the interparticulate porosity was located between nated the porosity due to the loose aggregate structure
1–100 lm, while the intraparticulate porosity was very and that caused by the interparticulate porosity was
low due to the nature of this sample, with mesoporosity reduced and shifted to narrower pores due to the higher
lying just below the detection limit of the porosimetry degree of compaction during the wet kneading and ex-
technique, at ca. 6 to 7 nm [7]. For the starch sample, a trusion processes, and final shrinkage of the material
single distribution due to the interparticulate porosity formed by drying at 150 °C. The PSD of this material
0.5
c
0.5
0.4
3 -1
0.0
0.3
b
0.5
0.2
0.0
a
0.1
0.5
0.0
0.0
1 10 100 1000 10000
1 10 100 1000 10000
Pore Diameter (nm)
Pore Diameter (nm)
Fig. 2: Cumulative volume (filled symbols) and PSD (empty sym-
bols) curves of the Clay monolith treated at 150 °C (䊉), 500 °C Fig. 3: Cumulative volume curves for Clay: (䊉), Clay/PGA 90/10
(~) and 850 °C (䉬). (~) and Clay/PGA 80/20 (䉬) monoliths treated at (a) 150 °C, (b)
500 °C, and (c) 850 °C.
was bimodal due to contributions from both intraparti- the samples that had 10 wt.-% and 20 wt.-% PGA in
culate and interparticulate porosities, which may be dis- their original composition, compared to the sample pre-
tinguished by the minima at 18 nm. Heat treatment at pared without PGA, but had no effect on the porosity in
500 °C caused a slight shift in the intraparticulate poros- narrower pores. Similarly, for the samples treated at
ity to wider pores, shown by the minima being located 850 °C, it is observed in Figure 3c) that the porosity in
to 23 nm. Although the measured pore volume was pores of less than ca. 300 nm was equivalent for all
greater after treatment at 500 °C, the mechanical three, but increases of 25% and 50% in the pore vo-
strength also rose. With heat treatment at 850 °C, the lumes with respect to the clay monolith were produced
differentiation between intraparticulate and interparti- by elimination of 10 and 20 wt.-% PGA, respectively.
culate porosities became unclear due to further sinter- Similar results were obtained for the samples produced
ing, with only a single maximum at 39 nm. However, the by maintaining a clay/titania weight ratio of 1/1. The cor-
total porosity was reduced, partly due to further shrink- responding porosimetry curves obtained with this series
age, doubling the mechanical strength. The variations in of samples, following treatment at 150 °C, 500 °C and
the pore size distribution of the clay binder with drying 850 °C, are presented in Figures 4a), 4b) and 4c), respec-
and heat treatment were important since they affect the tively. Thus, in Figure 4a) it is seen that after conforma-
mechanical strength development in these materials, tion and drying at 150 °C, pore volume increases of 6%
and also in the monolith composites produced with tita- and 41% respectively, were observed for samples that
nia and alumina. had 10 wt.-% and 20 wt.-% PGA in their compositions,
The MIP curves for the clay series treated at 150 °C, although the porosities of these three samples were
500 °C and 850 °C are presented in Figures 3a), 3b) and equivalent up to about 150 nm pore diameter. From the
3c), respectively. It is seen from Figure 3a) that for the porosimetry curves shown in Figure 4b) it is seen that
clay samples treated at 150 °C, the incorporation of 10 after heat treatment at 500 °C, the porosities of the ma-
or 20 wt.-% PGA caused increases of pores greater than terials in pores smaller than about 150 nm were equiva-
200 nm, with corresponding reductions in the mechani- lent. Increases of the pore volumes in the larger pores of
cal strength. These increases in total pore volumes were 13% and 30% for samples originally produced with
surprising since the soluble starch used as the PGA was 10 wt.-% and 20 wt.-% PGA, respectively, were ob-
expected to fill the interparticulate pore space, and thus served to the compared sample without PGA. Similarly,
cause a decrease in the total porosity. with heat treatment at 850 °C, as shown in Figure 4c),
From the curves presented in Figure 3b) it is observed the porosities in pores of less than 150 nm remained
that decomposition of the starch by heat treatment at practically unchanged. The increased porosities in the
500 °C led to well defined increases in the pore volumes wider pores in comparison with the sample produced
of about 19% and 36% in pores greater than 300 nm for without PGA were 11% and 31%, respectively.
0.0 ered [12,13], since fracture will occur due to the defect
b structure caused by the phase boundaries between the
0.5
compacted particles, rather than fracture of the indivi-
dual primary particles. When the clay used as the ag-
glomerating agent was dried and heat treated at 150 °C
0.0
a and 500 °C, it displayed bimodal distributions due to
contributions of both intraparticulate and interparticu-
0.5 late porosities that could be distinguished by the ob-
served minima in the PSD curves at 18 nm and 23 nm,
0.0 respectively. When this distinction between the two
1 10 100 1000 10000 100000 types of porosity disappeared following heat treatment
Pore Diameter (nm) at 850 °C, the limiting value was taken from the maxima
Fig. 5: Cumulative volume curves for: Clay/Alumina 50/50 (䊉), in the PSD curve at 39 nm. These three values were sub-
Clay/Alumina/PGA 45/45/10 (~) and Clay/Alumina/PGA 40/40/ sequently used to calculate the interparticulate poros-
20 monoliths (䉬) treated at (a) 150 °C, (b) 500 °C, and (c) 850 °C. ities of all the materials.
The relationships between the interparticulate porosity their interparticulate porosities. However, the relation-
and mechanical strength are presented in Figures 6a), ships were different for each system and each heat treat-
6b) and 6c) for the clay, clay/titania and clay/alumina ment since the mechanical strength development in
series, respectively. As the heat treatment temperature these composites is dependent on the size and bonding
was increased for the clay and clay/titania systems, the between the individual primary particles, the interparti-
mechanical strengths also rose, especially after treat- culate porosity, and temperature of the final heat treat-
ment at 850 °C. This development of mechanical ment. However, these curves are useful for estimating
strength was mainly due to phase transformations of the the expected mechanical strength for an unknown com-
magnesium silicate binder. These changes caused a position which falls within the measured range and pre-
slight increase in the mechanical strength coupled with dicting the maximum amount of PGA that could reason-
an increase in the interparticulate pore volume, follow- ably be added to any of the formulations without
ing treatment at 500 °C, and a reduction in the interpar- causing too great a reduction in the final mechanical
ticulate pore volume accompanied by a significant in- strength of the heat treated monolith, for its industrial
crease in mechanical strength, after treatment at 850 °C. use as a catalyst support.
c 4 Conclusions
25
Mechanical Strength (MPa)