OARS - Open Access Rewards System
DOI : 10.2240/azojomo0293

Gel Casting of Porous Alumina and Zirconia Bodies

Jean-Marc Tulliani, Valentina Naglieri, Mariangela Lombardi and Laura Montanaro

Copyright AD-TECH; licensee AZoM.com Pty Ltd.
This is an AZo Open Access Rewards System (AZo-OARS) article distributed under the terms of the AZo-OARS https://www.azom.com/oars.asp which permits unrestricted use provided the original work is properly cited but is limited to non-commercial distribution and reproduction.

AZojomo (ISSN 1833-122X) Volume 6 August 2009

Topics Covered

Abstract
Keywords
Introduction
Method and Materials
Results and Discussion
Conclusions
References
Contact Details

Abstract

An improved gel-casting procedure was successfully exploited to produce porous ceramic bodies based on two commercial powders, TM-DAR Taimicron alumina and TOSOH TZ-3YS zirconia, with controlled porosity features in terms of mean pore size, total pore volume as well as pore geometry. The gel-casting process in which a gelatine for food industry is used as gelling agent was firstly set-up to prepare dense alumina and zirconia components. Commercial polyethylene (PE) spheres, sieved to select proper dimensional ranges, were then added to the ceramic slurries as a pore-forming agent to prepare porous components.

In the case of zirconia, components having 50 vol. % porosity were manufactured, while for alumina porous materials, three different porosity percentages were studied, precisely 40, 50 and 60 vol. %. A suitable thermal treatment was set-up for burning-out the organic phases (PE and gelatine) and subsequent densification of the ceramic network without collapsing of the structure. The fired components presented spherical pores surrounded by highly dense ceramic walls and struts, having a homogeneous and fine microstructure. Image analysis performed on SEM micrographs was used for a full characterisation of the pores features and of their distribution

Keywords

Gel casting, gelatine, porous ceramics, controlled porosity, image analysis, pore distribution

Introduction

The importance of porous ceramic materials is continuously increasing because of a variety of interesting applications, such as, for example, molten metals filtration, high-temperature thermal insulation, support for catalytic reactions, filtration of particulates from diesel engine exhaust gases and of hot corrosive gases in various industrial processes and bone substitutes [1,2].

The porous features influence the peculiar characteristics of the produced components, thus two crucial steps are involved in the production of cellular materials: firstly, the selection and set-up of a reliable forming method allowing a strict control of pore size, volume, distribution and morphology; secondly, a proper characterisation of the actual porosity features as well as of the mechanical/functional performances [1,2].
This paper mostly deals with the exploitation of a novel gel-casting process to develop porous materials having controlled porosity characteristics.

Gel casting is now a well-known wet forming method based on the combination of ceramic processing and polymer chemistry. This process involves the dispersion of a ceramic powder into a monomer solution and casting of this suspension into a non-porous mould. Polymerisation is then promoted and consequently ceramic particles are entrapped into the rigid and homogenous polymeric network [3]. After gel formation, gel-cast green materials can be easily demoulded and are then dried in controlled conditions.
The main advantages of this method are the high strength of the green bodies, allowing their machinability, the low amount of organic additives, avoiding a preliminary thermal treatment as in the injection moulding process [4], and also its versatility which allowed to extend such process to a variety of ceramic materials.

Different monomers were exploited for gel-casting, starting from acrylamide systems, now withdrawn due to their neurotoxicity [5], as well as alternative gelling agents, such as, for example, agar [6], agarose [6-8], carrageenan gums [6], egg white [9], gelatine [10-12] and polyvinyl alcohol [12]. Also in this paper a natural gelatine, precisely a pig-derived product for food industry, was used. The natural gelling agents present the advantages of a polymerisation just promoted by a temperature change, without the use of a catalyst and an initiator, as synthetic monomers do, and of a low environmental impact.

Gel-casting was firstly set-up for preparing dense components, and more recently modified to fabricate also porous ceramics, by combining it with foaming techniques, or replica methods, or even the addition of a sacrificial phase [1, 11].
In this paper, a fugitive phase, made of commercial polyethylene spheres in a size range selected by sieving, was added to the ceramic suspension before gelling. The shape, size and size distribution of these spheres as well as their volume fraction with respect to the ceramic content into the slurry can allow a strict control of many porosity features of the final components. The feasibility of the new process was demonstrated on two different materials having different mean pore sizes and also, for one of them, by developing porous components with a pore volume percentage ranging from 40 to 60 vol. %.

Method and Materials

Two commercial ceramic powders were used: a pure - alumina powder (TM-DAR Taimicron) [13], supplied by Taimei Chemicals Co., and a zirconia powder doped with 3 mol.% yttria (TZ-3YS) [14], produced by Tosoh Co. A pig skin-derived gelatine produced by Italgelatine (Italy) was chosen as gelling agent; its melting point is 32.5°C and its viscosity in aqueous solution (6.67 wt. %) at 60°C is 42.5 mPa.s with a related pH of 5.08. This gelatine calcined up to 550°C for 18 hours yields a limited amount of residual ashes (0.2 wt. % of the starting mass), as confirmed by the thermogravimetric analysis (TGA; Netzsch STA 409).

A polyethylene (PE) powder, supplied by Clariant Italia SpA, having a density at 23°C of about 0.92 - 0.94 g/cm3 was used as fugitive phase. Its thermal decomposition is almost completed after calcination at 550°C, as confirmed by thermogravimetric analysis and characterised by three main mass loss steps, precisely between 220°C and 360°C, 370°C and 470°C, 490°C and 550°C, respectively.

PE powder was made of almost spherical particles, presenting some irregularities on their surface (Figure 1). Polyethylene spheres were firstly sieved to obtain several fractions: in particular, the fraction in the range 224 - 300 µm was used to produce porous alumina bodies, whereas a larger one, precisely in the range 125 - 300 µm, was exploited for the zirconia materials. It must be also noted that during sieving few smaller particles remained adherent to the larger ones, so that a limited fraction of pores smaller than those expected from the sieved range could be observed in the final materials (Figure 1).

Figure 1. SEM micrograph of the PE spheres used as pore-forming agent.

In a first step, to set-up the procedure, dense components were produced by adding ceramic powders to de-ionised water to prepare ceramic suspensions having 60 wt. % solid content. Alumina slurries required a prolonged time (80 hours) under magnetical stirring, whereas a suitable dispersion of zirconia powder was reached just by ultrasonication for 10 min, as checked by laser granulometry analyses (Fritsch model Analysette 22 Compact).

The gelatine was dissolved in de-ionised water at 60°C and then added to the ceramic suspensions at the same temperature under stirring, so that an amount of gelling agent of 2 wt. % for alumina and of 3 wt. % for zirconia respectively, with respect to the final water content, was reached.

The rheological behaviour of the ceramic slurries was investigated by viscosimetric analyses (viscosimeter Brookfield HBDV-II) at 60°C, at which the gelatine aqueous solution shows the lower viscosity. A deairing step was required to remove air bubbles entrapped in the ceramic suspensions: casting of the suspensions was carried out under vacuum (at about 10-2 Pa), by pouring into non-porous Plexiglas moulds at 60°C and then cooling down to room temperature for promoting gelification. Cylindrical moulds having different dimensions (internal diameters from 12 up to 18 mm and height in the range 30 - 55 mm) were used. In a first moment, the cast samples were slowly dried at room temperature under controlled humid atmosphere and then, after demoulding, in static air.

Dense gel-cast zirconia ceramics were sintered at 1500°C (heating rate of 5°C/min, soaking time of 1 hour at the maximum temperature and cooling rate of 10°C/min down to room temperature), while alumina materials have been heat treated at 1400°C (heating rate of 2°C/min, soaking time of 1 hour at the maximum temperature and cooling rate of 10°C/min down to ambient temperature). The above sintering cycles were set-up on the ground of dilatometric studies performed on gel-cast bars as well as on uniaxially pressed bodies (Netzsch 402E), submitted to heating up to 1500°C with a soaking time of 3 hours [15].

The density of green and fired components was evaluated by using weight and geometrical measurements as well as Archimede’s method; their microstructures were observed by scanning electron microscopy (SEM, Hitachi S2300).
To fabricate porous ceramics, PE spheres were added to the powder suspensions after a preliminary dispersion in water, in suitable amounts to obtain alumina materials having 40, 50 and 60 vol. % porosity, respectively, and zirconia bodies having 50 vol. % porosity. The ceramic content of the slurries was the same adopted for the dense materials, except in the case of the alumina porous samples having 60 vol. % porosity, for which it was decreased to 50 wt. % to maintain the suitable rheological behaviour.

Another difference between dense and porous samples consisted in the modification of the overall thermal cycle during heating up to 600°C: various intermediate steps, on the basis of the TGA results above described, have to be introduced to control the thermal decomposition of the polyethylene spheres, without collapsing of the structure.

Porous ceramics were characterised as previously detailed for dense materials. Moreover, they were also submitted to Hg porosimetry (Carlo Erba Porosimeter 2000) on fracture surfaces, to avoid the wall effect due to the contact with the mould surfaces, as well as to an extensive SEM investigation on polished surfaces coupled to image analysis.

Results and Discussion

Table I reports the diameters corresponding to 10 % (d10), 50 % (d50) and 90 % (d90) of the particle size distributions before and after dispersion, for the used alumina and zirconia powders.

Table 1. Agglomerate size (µm) corresponding to 10 % (d10), 50 % (d50) and 90 % (d90) of the particle size distributions.

Powders

 

d10 (µm)

d50 (µm)

d90 (µm)

Alumina

as received

8.7

29.5

66.7

dispersed

n.d.

0.5

1.0

Ziconia

as received

0.4

1.2

7.8

dispersed

0.33

0.6

1.0

n.d. = not detectable

The wide particle size distribution of the as-received alumina is characterised by large agglomerates which completely disappeared after 80 hours magnetic stirring and the mean size is reduced to about 500 nm. Zirconia, just after ultrasonic dispersion for 10 min, reached a narrow distribution characterised by a mean particle size (about 600 nm) close to the value declared by the supplier.

A good dispersion of the ceramic slurry and a stable suspension is required, before adding gelatine and casting, to avoid gradients of concentration in the microstructure due to powder sedimentation (Figure 2).

Figure 2. Sedimentation of the suspension before gelation.

Viscosimetric analyses were carried out at 60°C on the ceramic suspensions before and after addition of the gelling agent. After a pre-shear, the apparent viscosity was measured at shear rates ranging from 6.12 to 245 s-1, maintaining each value constant for 10 sec. The higher the gelatine amount, the higher is the viscosity of the slurries having the same solid load.

Several tests were carried out to determine the optimal gelatine amount: on one hand, slurries having gelatine content less than 4 wt. % showed an apparent viscosity lower than 1000 mPa.s at 20 s-1, which allowed the casting process [10], on the other hand, suspensions with a gelatine content less than 1 wt.% did not form a rigid skeleton and, consequently, green gel cast bodies did not have a sufficient strength to be demoulded.

The pure zirconia suspension presented a very low viscosity and behaved as a Newtonian fluid, whereas it became pseudoplastic after gelatine addition. The viscosities of alumina suspensions with and without gelatine were higher than the corresponding ones of zirconia slurries, but still suitable for casting. More details on the rheological characterisation are reported elsewhere [16].

During casting into cylindrical moulds, it was possible to appreciate that samples made by using larger moulds, thus having a higher height to diameter ratio, needed the higher content of gelatine for a successful demoulding and handling, so that alumina suspensions added with 2 wt. % of gelatine were cast into the smaller cylindrical moulds (12 × 30 mm2), whereas the zirconia ones added with 3 wt. % gelatine were poured into larger moulds (18 × 55 mm2).

Green gel-cast components presented densities of about 2.1 and 2.6 g/cm3 corresponding to 53 % and 43 % of the theoretical value, assumed to be 3.96 g/ cm3 and 6.05 g/ cm3 for alumina and zirconia, respectively, also in agreement with the data given by the suppliers. SEM observations showed a well-packed microstructure, quite free from defects such as large pores and voids due to entrapped air bubbles [15, 16].

Both sintered materials reached about 96 % of their respective theoretical densities and showed a fine, homogeneous microstructure. More details about sintering and morphological characterisation of dense materials have been presented elsewhere [15, 16].
Linear shrinkage of a sintered zirconia bar respect to a fresh cast one was over 40 % (Figure 3), confirming the importance of the drying and sintering steps. Also porosimetric analysis on fracture surfaces revealed densities close to the theoretical values for both materials, with a residual open porosity less than 1 vol. %.

Figure 3. Shrinkage of a sintered zirconia bar respect to a fresh one.

Green porous materials presented a quite homogeneous distribution of PE spheres in a well-packed green ceramic microstructure, also all along the longitudinal cross-sections of the samples, as a consequence of the starting suspension homogeneity as well as of the fact that they were not able to float from the bulk to the upper surface of the cylindrical moulds, due to the suspension viscosity and to the fast gelation after casting.

The theoretical densities of the green components were estimated by calculation using the rule of the mixture for composite systems, as a function of the PE volume fraction, and considering a PE density of 0.93 g/ cm3: the green densities of alumina bodies were then 55 % (1.5 g/cm3), 62 % (1.5 g/cm3) and 56 % (1.2 g/cm3), of the theoretical values (2.75, 2.44, 2.14 g/cm3, respectively, for the materials containing 40, 50 and 60 vol. % of PE). The lowering of the green density in the 60 vol. % material should be reasonably imputed to the decrease in solid content of the starting slurry (from 60 to 50 wt. %) needed to maintain the suitable viscosity for casting, as already discussed before. Green porous materials made of zirconia, presented a density of about 54 % (1.9 g/cm3) of the theoretical value (3.49 g/cm3). The lower green density reached by zirconia samples if compared to the respective alumina materials could be probably imputed to the difference in processing (casting performed into larger moulds also requiring a higher amount of gelatine), since the green particle packing seemed to be similarly effective for both powders. In any case, this difference in green density was completely recovered during densification.

PE decomposition yielded pores having a quite spherical shape (Figure 4a), surrounded by ceramic walls and struts showing a dense and fine microstructure (Figure 4b). After sintering alumina porous bodies presented a density of about 59 % (2.32 g/cm3), 55 % (2.29 g/cm3), 39 % (1.54 g/cm3) of the theoretical value (3.96 g/cm3) for 40, 50 and 60 vol.% materials, respectively. Zirconia porous materials reached a density of about 60 % (3.6 - 3.7 g/cm3) of the theoretical value (6.05 g/cm3).

Figure 4. SEM micrographs at low (a) and high (b) magnificationof a porous zirconia component after sintering.

The open porosity percentage values measured by Hg intrusion were about 28, 39 and 47 % for the alumina materials having 40, 50 and 60 vol. % porosity, respectively. Two main families of pores were detected. The former is given by pores having radii ranging from about 1 micron to few tens of microns, associable to the interconnections between the large pores and to the almost constant contribution of the small PE spheres which remained adherent to the larger one surfaces during sieving. The latter family of pores is minority, located at significantly lower pore radii (0.01 - 0.1 m), and imputable to the small intergranular pores on the dense ceramic walls. More details on porosimetric results are reported in ref [16].

However, the porosimetric measurements cannot give information concerning the large spherical pores yielded by the PE spheres decomposition, since their dimensions are over the upper size limit detectable by the Hg intrusion. To investigate such porosities, SEM observations on many polished surfaces coupled to image analysis (Image-Pro Plus ver. 4.5 by Media Cybernetics, Inc.) were effectively exploited to add information about porosity features, in particular the distribution and the shape of the voids. Some SEM images of the sintered bodies are illustrated in Figure 5.

Figure 5. SEM micrographs for 40 (a), 50 (b), and 60 (c) vol. % alumina, and 50 vol. % (d) zirconia porous sintered bodies.

The SEM images were then processed by means of “segmentation”, which is a process by which certain grey levels in the image can be visually identified, then isolated from the image as a whole. Therefore, it can be used for separating items or objects of interest from the "background noise" that naturally occurs in most acquired images. Further, segmented areas can be either kept in their original colour or turned into a single colour (grey masking in this case).

Masking an image facilitates or improves the performance of other operations, like, for example, determining the area of the detected objects (Figure 6). However, objects in contact with images borders were not considered (Figure 6b).

Figure 6. (a) SEM micrographs for 40 vol. % alumina porous body and (b) corresponding image after “masking” the detected objects

Artefacts such as cracks and scratches, probably due to sample cutting and polishing, characterised by a high aspect ratio were also manually removed from the analysed images. Moreover, attention was paid, when objects were not completely detected, due to a limited contrast respect to the background, to the fact that the number of porosities was not increased by the “segmentation” step and, in this case too, the “exceeding elements” were ignored.

Areas, X and Y coordinates of object centroid in the image and aspect ratio between major axis and minor axis of an ellipse having the same area as the detected porosities where determined for all the evidenced objects in each SEM image. Figures 7-10 illustrate the X and Y coordinates of the detected elements in function of their aspect ratio, for the alumina and the zirconia porous components: the projections into the XY plane seem to indicate a quite regular distribution of the porosities in the images and thus a good dispersion of the polyethylene spheres in the initial suspensions.

Figure 7. X and Y coordinate of object centroid vs aspect ratio of the pores detected in the SEM images related to 40 vol. % alumina sample.

The number of detected objects in the alumina sample having 50 vol. % of porosity was slightly minor respect to all the other ones, and this can explain the apparent not so regular dispersion (Figure 8 and Table 2).

Figure 8. X and Y coordinate of object centroid vs aspect ratio of the pores detected in the SEM images related to 50 vol. % alumina sample.

Figure 9. X and Y coordinate of object centroid vs aspect ratio of the pores detected in the SEM images related to 60 vol. % alumina sample.

Figure 10. X and Y coordinate of object centroid vs aspect ratio of the pores detected in the SEM images related to 50 vol. % zirconia sample.

Table 2 also reports the aspect ratio of the pores, for all the sintered samples, as well as for the polyethylene spheres: the calculated values, for the polymeric spheres alone and the heat treated samples, were rather similar, confirming the good distribution of the porosities within the sintered bodies, as there is no significant change of this parameter whatever the pore-forming agent content.

Table 2. Aspect ratio of polyethylene spheres as such and of the porosities in the sintered.

Sample

Detected objects

Aspect Ratio

Polyethylene spheres

84

1.54± 0.52

Al2O3 40 vol. % porous

422

1.65 ± 0.54

Al2O3 50 vol. % porous

243

1.81 ± 0.49

Al2O3 60 vol. % porous

312

1.69 ± 0.49

ZrO2 50 vol. % porous

625

1.64 ± 0.47

Conclusions

A novel gel-casting procedure employing a natural gelatine for food industry as gelling agent and commercial PE spheres as pore-forming agent was successfully exploited to fabricate alumina and zirconia porous components. The porosity features such as pore shape, size and size distribution were easily controlled by adding almost spherical PE spheres, suitably sieved in dimensional ranges more restricted than that commercially supplied. A good dispersion of the ceramic slurry before adding gelatine and casting into non-porous moulds is required to obtain almost dense ceramic walls and struts characterised by a fine microstructure.

The homogeneous distribution of spherical pores surrounded by the ceramic network was demonstrated by an extensive SEM investigation on various, both transversal and longitudinal, cross-sections of the sintered cylindrical bodies. Such result is achievable thanks to a good dispersion of the PE spheres into the starting ceramic slurry as well as to an appropriate suspension viscosity and to a fast gelation of the gelatine after casting into the moulds avoiding the PE spheres to float from the bulk to the upper surface of the components.

References

  1. Studart A.R., Gonzenbach U.T., Tervoort E. and Gauckler L.J., “Processing Routes to Macroporous Ceramics: A Review”, Journal of the American Ceramic Society, 89 [6] (2006) 1771-89.
  2. Colombo P., “Conventional and novel processing methods for cellular ceramics”, Philosophical Transactions of the Royal Society A, 364 (2006) 109-24.
  3. Omatete O.O., Janney M.A. and Strehlow R. A., “A new ceramic forming process”, American Ceramic Society Bulletin 70 [10] (1991) 1641-49.
  4. Young A.C., Omatete O.O., Janney M.A. and Menchhofer P.A., “Gelcasting of alumina”, Journal of the American Ceramic Society, 74 (1991) 612-8.
  5. Janney M.A., Omatete O.O., Walls C.A., Nunn S.D., Ogle R.J. and Wastmoreland G., “Development of low - toxicity gelcasting systems”, Journal of the American Ceramic Society, 81 [3] (1998) 581-91.
  6. Millán A.J., Moreno R. and Nieto M.I., Thermogelling polysaccharides for aqueous gelcasting—part I: a comparative study of gelling additives, Journal of the European Ceramic Society, 22 (2002) 2223-30.
  7. Santacruz I., Nieto M.I. and Moreno R., “Alumina bodies with near - to - theoretical density by aqueous gelcasting using concentrated agarose solutions”, Ceramics International, 31(2005) 439-45.
  8. Adolfsson E., “Gelcasting of zirconia using agarose”, Journal of the American Ceramic Society., 89 [6] (2006) 1897-1902.
  9. Dhara S. and Bhargava P., “Egg white as environmentally friendly low - cost binder for gelcasting of ceramics”, Journal of the American Ceramic Society, 84 [12] (2001) 3048-50.
  10. Chen Y., Xie Z., Yang J. and Huang Y., “Alumina casting based on gelation of gelatine”, Journal of the European Ceramic Society, 19 (1999) 271-5.
  11. Vandeperre L.J., De Wilde A.M. and Luyten J., “Gelatin gelcasting of ceramic components”, Journal of Materials Processing Technology, 135 (2003) 312-6.
  12. Ortega F.S., Valenzuela F.A.O., Scaracchio C.H. and Pandolfelli V.C., “Alternative gelling agents for the gelcasting of ceramic foams”, Journal of the European Ceramic Society, 23 (2003) 75-80.
  13. http://www.taimei-chem.co.jp
  14. http://www.tosoh.com/Products/SpecialtyMaterials
  15. Tulliani J.M., Naglieri V., Lombardi M. and Montanaro L., (2008), “Porous alumina and zirconia bodies obtained by a novel gel casting process” in vol. 29 (Procedings of the 32nd International Conference & Exposition on Advanced Ceramics and Composites), Narayan R. and Colombo P. Eds., Daytona Beach (FL) (2008)
  16. M. Lombardi, V. Naglieri, J.M. Tulliani and L. Montanaro, “Gelcasting of dense and porous ceramics by using a natural gelatine”, Accepted for publication in Journal of Porous Materials, Available online: http://www.springer.com/home?SGWID=0-0-1003-0-0&aqId=286910&checkval=46caef2903db2c56d7ac383e76c289af

Contact Details

Jean-Marc Tulliani, Valentina Naglieri, Mariangela Lombardi and Laura Montanaro
Dept. Materials Science and Chemical Engineering
Politecnico of Torino
INSTM-UdR PoliTO-LINCE Lab.
Torino, Italy

E-mail: [email protected]

This paper was also published in print form in “Advances in Technology of Materials and Materials Processing”, 11[1] (2009) 9-18.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this article?

Leave your feedback
Submit