Journal of Materials Processing Technology 209 (2009) 5313–5318
Contents lists available at ScienceDirect
Journal of Materials Processing Technology journal homepage: www.elsevier.com/locate/jmatprotec
Flexible polyurethane foams as templates for cellular glass–ceramics M.W. Quintero a , J.A. Escobar a , A. Rey a , A. Sarmiento a , C.R. Rambo b , A.P. Novaes de Oliveira b , D. Hotza b,∗ a Research Center on Materials Properties and Structures (CIPP/CIPEM), Departments of Chemical and Mechanical Engineering, Los Andes University (Uniandes), Bogotá, Colombia b Group of Ceramic and Glass Materials (CERMAT), Departments of Chemical and Mechanical Engineering (ENQ/EMC), Federal University of Santa Catarina (UFSC), Florianópolis, SC, Brazil
a r t i c l e
i n f o
Article history: Received 20 May 2008 Received in revised form 3 March 2009 Accepted 27 March 2009 Keywords: Polyurethane foams Cellular glass–ceramics Permeability Cell diameter
a b s t r a c t In this work flexible polyurethane (PU) foams were obtained with varying air permeability, cell diameter and morphology. The addition of up to 1.5 g anti-foamer with a mixing time of 75 s resulted in larger cell diameters, higher air permeability and lower distorted area. PU foams obtained were used as templates to produce glass–ceramic (GC) foams by the replication method. Glass powder of the LZSA (Li2 O–ZrO2 –SiO2 –Al2 O3 ) system was used to infiltrate the PU foams. The retention capacity of the ceramic suspension in PU foam is increased with a reduction of the cell diameter. In contrast, the infiltration capacity increases with raise of the cell diameter. The permeability reduction of GC foams with respect to PU foams varied from 3% to 25% when 2.7 and 0.8.0 mm cell diameters were used, respectively. The results of mechanical characterization were coherent with the morphological characteristics in both PU and GC foams. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Cellular ceramic structures obtained from polymeric foam templates present high porosity, low density, high chemical stability, structural uniformity, and high surface area. These properties make these structures interesting for a variety of applications such as filters, catalyst supports, membranes, thermal insulators, among others (Gibson and Ashby, 1997; Scheffler and Colombo, 2005). The performances of cellular ceramic structures depend on cell morphology and on material composition. This leads to the conception of several fabrication technologies according to the application. The polymeric sponge method, also known as replication method, offers a simple, inexpensive and versatile way for producing ceramic foams. This method consists of dipping the polymeric sponge into a slurry containing ceramic particles and appropriate additives (binders and dispersants) followed by drying to evaporate the solvent, heating to burn out the organic part, and sintering to form an open-cell ceramic skeleton (Schwartzwalder and Somers, 1963). The final product is a ceramic with the initial form of the polymeric foam but with a volumetric shrinkage or expansion. The most common applications for open-cell ceramics are as supports for catalysts, and filters for molten metal and hot exhaust gases (Studart et al., 2006). Glass–ceramic (GC) materials have found applications in many fields thanks to important properties such as low coefficient of
∗ Corresponding author. Tel.: +55 48 3721 9448. E-mail address:
[email protected] (D. Hotza). 0924-0136/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jmatprotec.2009.03.021
thermal expansion, high abrasion and scratch resistance, and good chemical and thermal shock resistance. Due to their unique characteristics, glasses belonging to the Li2 O–ZrO2 –SiO2 –Al2 O3 (LZSA) have been used in powder technology processing by various forming methods such as extrusion (Montedo et al., 2004; Bertan et al., 2009), injection molding (Giassi et al., 2005), tape casting (Gomes et al., 2006), and rapid prototyping (Gomes et al., 2008). Particularly, the production of cellular glass–ceramics constitutes a new class of processing technology, which has been in recent times successfully explored (Sousa et al., 2005; Rambo et al., 2006). A critical variable for the fabrication process corresponds to the polymeric foam morphology (Silveira et al., 2007). The infiltration capacity of the ceramic suspension into the foam, the retention of the slurry, and the final characteristics of the ceramic structure depend directly on the polymeric foam cell diameter and distribution (Zhu et al., 2001; Sarmiento, 2006; Sousa et al., 2008). This paper deals with preparing and modifying the morphology of a polyurethane flexible foam, used as template in a parent glass suspension for the fabrication of cellular glass–ceramics, and evaluating the performance of the final cellular structures. 2. Experimental procedure 2.1. Raw materials for PU foam The formulation used for the foam preparation is shown in Table 1. The amounts of polyether polyol, water, silicone surfactant, amine and tin catalysts, and isocyanate were held constant, whereas the amount of anti-foamer was varied according to a fac-
5314
M.W. Quintero et al. / Journal of Materials Processing Technology 209 (2009) 5313–5318 Table 3 Composition of the parent glass suspension.
Table 1 Formulation used for the flexible polyurethane foam. Material Poliol Polyether polyol initiated with glycerine OH number: 56 Theoretical functionality: 3 Producer: Dow Water Silicon surfactant Non-ionic, silicone-based Proprietary composition Producer: Goldschmidt Amine catalyst 33% triethylene diamine in dipropylene glycol 70% bis(dimethylaminoethyl)ether 30% dipropylene glycol Producer: Air Products
Parts per hundred polyol (pphp)
Amount
100
900 g
2.50
22.50 g
The air permeability was calculated based on the Forchheimer equation, described as:
0.80
7.20 g
P 2 = V+ V L k1 k2
0.15
1.35 g
Tin catalyst Stannous octoate Producer: Air Products
–
1.0 ml
Isocyanate TDI 80 Toluene diisocyanate 80/20 mixture of the 2,4 and 2,6 isomers Producer: Dow
–
340.20 g
Anti-foaming agent Linear chain of polydimethylsiloxanes Proprietary composition Producer: Goldschmidt
Variable
torial design, as explained in the next section. These raw materials were mixed in three different stages: first, polyol, silicone surfactant, amine catalyst and water; second, this blend with tin catalyst, and last, the whole polyol mixture with isocyanate. 2.2. Experimental design for PU foam A factorial design was applied, taking into account the possibility of modifying and controlling the morphology of the flexible polyurethane foam template for the production of porous ceramics. The experimental design established two factors to be analyzed: amount of anti-foaming agent (4 levels), and mixing time (3 levels). The latest factor makes reference to the first of the three mixing stages of a typical PU flexible foam fabrication process. Table 2 shows the factors and the levels selected. 2.3. Characterization of PU foams The response variables were air permeability, cell diameter, and distortion area. The cell diameter and permeability measurements were made according to ASTM D3576 (2004), and ASTM D3574 (2005), respectively, using a flowmeter (Uehling, model ZK, USA). Table 2 Factors and levels of the experimental design for PU foams. Factor Content of anti-foaming agent (g)
Mixing time (s)
Levels 0.7 0.9 1.1 1.3 15 45 75
Quantity (g)
Parent glass Bentonite Water Sodium silicate
100.0 5.94 100.63 1.07
(1)
where P is the pressure drop; L, the medium thickness; and , the viscosity and density of the fluid, respectively; and V, the volumetric flow rate per unit of cross section area. The parameters k1 and k2 , correspond to the Darcyan and non-Darcyan permeability parameters, respectively. Both constants depend exclusively on the medium characteristics (Moreira et al., 2004). The first term of Eq. (1) may be attributed to laminar flow; the second, to turbulent flow. In this case, only laminar flow was considered, since Re < 2000 was applied to the experimental measurements (Mills, 2005). Hence, the permeability can be determined using Eq. (2): k1 =
–
Component
·V P/L
(2)
The distorted area of foams was quantified from digital images with the help of the software Image J (Rasband, 2008). A picture of the superficial area of the inferior view of a 380 mm × 380 mm × 200 mm foam sample was taken. A percentage of occupied area was quantified by the cells with smallest cell diameter over the whole surface. An optical microscope (Metallux II, Leitz, Germany) was used. In addition to the morphological foam evaluation, a mechanical characterization of the foams was made with selected PU foams in a universal testing machine (Instron, Model 4202, USA) according to ASTM D3574 (2005), which involves standard methods for conditioning of foam samples and for testing the basic physical properties: density, tensile strength, tear resistance, airflow, resilience, indentation force deflection, compression force deflection and constant deflection compression set. 2.4. Raw materials for GC foams LZSA glass with nominal composition of 11.7Li2 O–12.6ZrO2 – 68.6SiO2 –7.1Al2 O3 was prepared from Li2 CO3 , ZrSiO4 and SiO2 and LiAlSi2 O6 (spodumene) as raw materials. The template powders were placed in a mullite crucible and melted at 1500 ◦ C for 2 h in a gas furnace. The melt was quenched in water, dried and subsequently milled. The formulation of the ceramic suspension was composed of a liquid phase (water and sodium silicate, used as a dispersant) and a solid phase (glass powder of LZSA, Li2 O–ZrO2 –SiO2 –Al2 O3 , and bentonite, used as a binder). A detailed study on formulation of the parent glass suspension according to rheological measurements can be found elsewhere (Rambo et al., 2006). Bentonite presents as main chemical components SiO2 (62.8 wt%), Al2 O3 , (20.3 wt%), Fe2 O3 , (3.8 wt%), Na2 O (2.4 wt%), MgO (2.3 wt%) and CaO (1.2 wt%), corresponding to montmorillonite as a main phase and quartz as an impurity (Bertan et al., 2009). According to an optimized composition, Table 3, the slurry was prepared with 48.5 wt% water, containing 48.2 wt% parent glass powder (LZSA), 2.9 wt% bentonite (Colorminas, Brazil), used as a binder, and 0.5 wt% sodium silicate (Merck, Natronwasserglas
M.W. Quintero et al. / Journal of Materials Processing Technology 209 (2009) 5313–5318
5315
apparatus air passes through a 25.4 mm × 50.8 mm × 50.8 mm sample when drawing a 25.4 mm of water differential pressure between the front and backside of the sample. Additionally, a qualitative evaluation was performed to analyze the impregnation and retention capacity of the foams. 3. Results and discussion 3.1. Characterization of PU foams
Fig. 1. Flow chart of fabrication of cellular glass–ceramics by the replication method.
105621, 7.5–8.5 wt% Na2 O, 25.5–28.5 wt% SiO2 ), used as dispersant. This optimized slurry presents a pseudoplastic behavior with apparent viscosities ranging from 25 to 18 mPa s for shear rates from 100 to 400 s−1 , respectively (Rambo et al., 2006). 2.5. Processing of GC foams The replication method was used in order to obtain the cellular glass–ceramic structure, as can be seen in Fig. 1. Water was first mixed with sodium silicate in a plastic vessel with alumina balls as grinding media for 12 h. Subsequently, the glass powder was added to the slurry and then milled for another 12 h. Bentonite was added to the slurry milled for another 12 h. The polymeric foams were cut in pieces of approximately 2.5 cm × 3.0 cm × 2.0 cm and immersed in the LZSA parent glass slurry. The impregnated templates were slightly compressed to remove the excess slurry and dried at room temperature for 24 h. After drying, the samples were submitted to the heat treatment, which was performed in an electrical furnace in air atmosphere. The samples were first heated at 1 ◦ C/min to 400 ◦ C to burn out the polymeric foam. Subsequently, the samples were heated to 700 ◦ C at 5 ◦ C/min to promote sintering of the glass powder and finally were heated at 900 ◦ C for crystallization at the same heating rate. The system was inertially cooled. 2.6. Experimental design for GC foam An experimental procedure was developed with a factor that evaluates the effect of the PU foam cell diameter on the final ceramic structure. The selected factor was the foam cell diameter, one of the response variables of the first part of the experimental process. Table 4 shows the selected factor and levels. The response variable used to evaluate this factor was the permeability to air flow, according to ASTM D3574 (2005), using a differential pressure air permeability tester developed at the Dow Chemical Polyurethane Laboratories (Freeport, TX, USA). In this
Fig. 2 shows the cell diameters of polyurethane foams (average of two replicas) as a function of mixing time and anti-foamer amount. It can be noticed that by increasing anti-foamer amounts and mixing times, the cell diameter increases. In addition, with the use of 1.3 g anti-foamer (equivalent to 0.144 parts per hundred parts of polyol, pphp), the cell diameters obtained increased significantly when compared to other quantities of anti-foamer. The largest cell diameter was obtained with 1.3 g anti-foamer (0.144 pphp) and 75 s mixing time. The anti-foaming agent used in this work contains straight chains of polydimethyl siloxanes (HO[–Si(CH3 )2 O–]n H). This type of compound is also present in surfactants used for flexible polyurethanes (Snow and Stevens, 1999). When anti-foaming agents are added, there is an increase of surface tension of the liquid where bubbles are formed. The gas–liquid interface must follow the law of Laplace that expresses the balance of pressure difference and the force of surface tension as P =
2 r
(3)
where is the surface tension and r is the local radius of curvature of the surface (Weaire and Hutzler, 1999). The high surface tension of the system creates a higher differential pressure between the bubbles. Therefore, system stability is lower and average cell diameter is higher. Fig. 3 shows the air permeability values for the polyurethane foams, corresponding to the average of two replicas. An increase of anti-foamer amount and mixing time led to an increase in permeability. The maximum permeability value was reached with 1.3 g of anti-foamer (0.144 pphp) and a mixing time of 75 s. Higher values of permeability where always obtained with a mixing time of 75 s. As mentioned earlier, an increase of the anti-foaming agent causes instability in the foam. For this reason, it was necessary to use an appropriate quantity of tin catalyst (1 ml) in order to guarantee stability of the foams even with the use of large quantities of anti-foamer. Fig. 4 shows the quantification of distorted areas of PU foam surfaces obtained by an image analyzer, corresponding to the average of two replicas. It is possible to establish a relation between the quantity of anti-foamer and the distorted area on the foam surface, since an increase of the additive produced an increase of the
Table 4 Factor and levels of the experimental design for GC foams. Factor
Levels
Cell size (mm)
0.8 1.5 2.1 2.7
Fig. 2. PU foam cell diameters as a function of mixing time and anti-foamer amount.
5316
M.W. Quintero et al. / Journal of Materials Processing Technology 209 (2009) 5313–5318
Fig. 5. Comparison of the permeability for PU and GC foams.
Fig. 3. Darcyan permeability for PU foams as a function of mixing time and antifoamer amount.
Fig. 4. Distorted area for PU foams as a function of mixing time and anti-foamer amount.
distorted area. In addition, the reduction of the distorted area as a function of time is more perceptible for lower amounts of antifoamer. The lowest distorted area was obtained with 0.7 g addition of anti-foamer (0.078 pphp) and a mixing time of 75 s. It is assumed, based on the expected role of the anti-foamer on the surface tension increase, that the distorted areas, defined as those with the smallest cell diameter, are the places in which the anti-foamer does not act. Therefore, a small cell diameter zone is produced whereas the rest of the foam presented a larger cell diameter. It was visually observed that a qualitative but remarkable improvement was obtained in the cell diameter distribution with the increase in mixing time. However, there must be additional ways to obtain a more homogeneous distribution. From the obtained results of the response variables (cell diameter, permeability and distorted area), the foams mixed for 75 s and with better cell diameter distribution were selected. Additionally, to the selected foams, a conventional polyurethane foam was also chosen for comparison because it presents a much smaller cell diameter than the others. Moreover, a mechanical characterization was carried out with foams used for impregnation of the parent glass slurry (Table 5).
No anti-foamer agent was added to the 8.0 mm cell diameter foam, meaning that this type of foam corresponds to the commercial one. An increase in hardness can be seen as the cell diameter increases. The addition of anti-foamers may cause some secondary effects such as higher foam hardness and shorter curing times. The resilience test was carried out by measuring the distance reached by a steel ball after bouncing over the foam. As the cell diameter increased, the foam resilience decreased. As already mentioned, an increase in the hardness of the foam was caused by the addition of anti-foamer. As the cell diameter increased the values for the tensile strength and elongation were reduced. With smaller cell diameters, the amount of reticulation points decreased, because fewer solid strut networks were formed (Herrington and Hock, 1991). In a similar way, when there are fewer reticulation points, the foam looses its elastic deformation capacity and hence its elongation will be decreased. For larger cell diameters, the value for tear strength and permanent deformation was reduced. With fewer reticulation points on the foam, the capacity to resist to shear stress was lower, taking into account that foams are 3D-structured. The reduction of the elastic deformation has an effect of lowering limit values between elastic and plastic zones (Gibson and Ashby, 1997). Therefore, it is expected that permanent or plastic deformation to be reduced with an increase on the cell diameter. 3.2. Characterization of GC foams Fig. 5 shows the results of the air permeability tests of PU and GC foams. The data shown are an average of two replicas. The maximum permeability for the GC foam is reached with 2.7 mm cells. It can be seen as well that an increase of the PU foam cell diameter will cause an increase on the ceramic foam permeability. This increase can be directly related to a larger cell diameter in the original polymeric foam (Scheffler and Colombo, 2005). The models presented in the literature describe permeability as non-linear. Nevertheless, in the present work, according to the regression analysis, a linear model fits the experimental data in an acceptable way (R2 = 96.15%). In this way, it was possible to use
Table 5 Mechanical characterization of PU foams. Cell size (mm)
0.8 1.5 2.1 2.7
Mechanical properties Hardness (kg)
Resilience (%)
Tear strength (N/mm)
Tensile strength (kPa)
Elongation (%)
Permanent deformation (%)
2.8 6.6 7.4 8.0
49 48 46 44
0.31 0.17 0.18 0.14
75.4 46.7 33.6 33.0
158 12 1 1
6 4 3 2
M.W. Quintero et al. / Journal of Materials Processing Technology 209 (2009) 5313–5318
Fig. 6. PU foams with 0.8, 1.5, 2.1, 2.7 mm cell diameter (a, c, e, g), respectively, and their correspondent GC replicas (b, d, f, h).
5317
5318
M.W. Quintero et al. / Journal of Materials Processing Technology 209 (2009) 5313–5318
linear models to describe permeability for cellular ceramics at low air flows (3.4–5.1 m3 /h). Fig. 6 shows micrographs of polymeric foams with their respective glass–ceramic replicas. A regular impregnation was obtained for the foams with cell diameters of 0.8 and 1.5 mm (Fig. 6a–d). The visual inspection suggests that for foams of 2.1 and 2.7 mm cell diameters (Fig. 6e–h) the impregnation was not homogenous. Thus, the most convenient polymeric foams to serve as template for the replication process would be the foam with a cell diameter of 1.5 mm or smaller. Additionally, ceramics that have a better capacity of retention will present a higher mechanical strength, since they have a better adhesion of the suspension to the struts of the foams cells (Scheffler and Colombo, 2005). It is important to mention that the increase in cell diameter of the PU foam facilitates the infiltration process. This can be explained from the migration phenomenon of a liquid within a pore (Eq. (4)).
v=
∗ cos RC ∗ 4L
(4)
where RC is the pore radius, is the contact angle between the liquid and the solid capillary material, is the surface tension of the liquid, is the viscosity, L is the length of liquid column of the pore, and v is the flow rate average on the column pore. For a certain column of liquid L, if the other variables remain constant (except for RC , as applied during experimentation), the radius of the pore increases and the average flow rate in the column increases. This means that the rate of penetration of the liquid in the cells of the foam is higher. 4. Conclusions Ceramic foams made are being considered for a whole range of applications including gas filters, interpenetrating composites and biomedical applications as well as thermal insulation, kiln furniture and catalyst supports. By the use of anti-foaming agents it was possible to modify and control the morphology (cell diameter) of flexible polyurethane foam. The retention capacity of the ceramic suspension in flexible polyurethane foam was increased with a reduction of the cell diameter. While the infiltration capacity increased with an increase in the cell diameter. The mechanical strength of the ceramics obtained by the infiltration of polymeric foams of smaller size was apparently better. For this reason it is necessary to carry out some type of modification to the suspension formula for foams of larger cell diameters. The decrease of permeability of GC foams with respect to PU foams can be hindered with an increase in the cell diameter of the impregnated foams. A correlation between PU and ceramic foams is necessary to be quantified in further investigations.
Acknowledgments The authors are grateful to the Brazilian Agencies CAPES and CNPq, and to the Colombian Agency COLCIENCIAS for funding this work. References ASTM D3574, 2005. Standard Test Methods for Flexible Cellular Materials—Slab, Bonded, and Molded Urethane Foams. ASTM D3576, 2004. Standard Test Method for Cell Size of Rigid Cellular Plastics. Bertan, F.M., Montedo, O.R.K., Rambo, C.R., Hotza, D., Oliveira, A.P.N., 2009. Extruded ZrSiO4 particulate-reinforced LZSA glass-ceramics matrix composite. J. Mater. Process. Technol. 209, 1134–1142. Giassi, L., Montedo, O.R.K., Hotza, D., Fredel, M.C., Oliveira, A.P.N., 2005. Injection moulding of LiO2 –ZrO2 –SiO2 –Al2 O3 (LZSA) glass ceramics. Glass Technol. 46, 277–280. Gibson, L., Ashby, M., 1997. Cellular Solids, 2nd ed. Cambridge University Press, New York. Gomes, C.M., Biscaia, F.N., Quinaud, J.T., Montedo, O.R.K., Oliveira, A.P.N., Hotza, D., 2006. Aqueous tape casting of LZSA glass ceramics. Ceram. Trans. 193, 9–16. Gomes, C.M., Oliveira, A.P.N., Hotza, D., Travitzky, N., Greil, P., 2008. LZSA glass–ceramic laminates: fabrication and mechanical properties. J. Mater. Process. Technol. 206, 194–201. Herrington, R., Hock, K., 1991. Flexible Polyurethane Foams. Dow Plastics, Midland, MI. Mills, N.J., 2005. The wet Kelvin model for air flow through open-cell polyurethane foams. J. Mater. Sci. 40, 5845–5851. Montedo, O.R.K., Reitz, G.M., Oliveira, A.P.N., Hotza, D., Siligardi, C., 2004. Extruded LZS glass–ceramics. Am. Ceram. Soc. Bull. 83, 9201–9206. Moreira, E.A., Innocentini, M.D.M., Coury, J.R., 2004. Permeability of ceramic foams to compressible and incompressible flow. J. Eur. Ceram. Soc. 24, 3209– 3218. Rambo, C.R., Sousa, E., Oliveira, A.P.N., Hotza, D., Greil, P., 2006. Processing of cellular glass ceramics. J. Am. Ceram. Soc. 89, 3373–3378. Rasband, W., 2008. ImageJ 1.36b. National Institute of Health, USA. Freeware (Public Domain). Available at http://rsb.info.nih.gov/ij/ as on 25 March 2008. Sarmiento, A., 2006. Mechanical Engineering Thesis. Universidad de Los Andes, Bogotá, Colombia. Scheffler, M., Colombo, P. (Eds.), 2005. Cellular Ceramics. Wiley-VCH, Weinheim, Germany. Schwartzwalder, K., Somers, A.V., 1963. Method for Making Porous Ceramic Articles. US Patent 3,090,094. Silveira, C.B., Escobar, J.A., Quintero, M.W., Sousa, E., Moraes, E.G., Oliveira, A.P.N., Rambo, C.R., Hotza, D., 2007. Thermal decomposition of polyurethane foams for manufacturing LZSA cellular glass ceramics. Quim. Nova 30, 1104–1107. Snow S.A., Stevens R.E., 1999. The science of silicone surfactant application in the formation of polyurethane foam. In Hill, Randal, M. (Eds.). Silicone Surfactants, Surfactants Science Series, vol. 86. CRC Press, New York, p. 138. Sousa, E., Rambo, C.R., Hotza, D., Oliveira, A.P., Novaes, Fey, T., Greil, P., 2008. Microstructure and properties of LZSA glass–ceramic foams. Mater. Sci. Eng. A 476, 89–97. Sousa, E., Silveira, C.B., Fey, T., Greil, P., Hotza, D., Oliveira, A.P.N., 2005. LZSA glass ceramic foams prepared by replication process. Adv. Appl. Ceram. 104, 22– 29. Studart, A.R., Gonzenbach, U.T., Tervoort, E., Gauckler, L.J., 2006. Processing routes to macroporous ceramics: a review. J. Am. Ceram. Soc. 89, 1771–1789. Weaire, D., Hutzler, S., 1999. The Physics of Foams. Clarenton Press, Oxford, p. 21. Zhu, X., Jiang, D., Tan, S., Zhang, Z., 2001. Improvement in the strut thickness of reticulated porous ceramics. J. Am. Ceram. Soc. 84, 1654–1656.
