Industrial Crops and Products 13 (2001) 135 143 www.elsevier.

com/locate/indcrop

Properties of starch-based foam formed by compression/explosion processing

Abstract Single-use foam packaging is used by manufacturers to protect and preserve a wide array of food and industrial products. Starch is one possible alternative material for making foam products. Starch-based foam was made using a compression/explosion process to study its properties and potential for single-use packaging. A feedstock was rst prepared which consisted of wheat (WS), corn (CS) or potato starch (PS) that was formed into aggregates (1 3 mm) and conditioned to moisture levels ranging from 8 to 20%. The conditioned aggregates were loaded in an aluminum compression mold heated to 230C and compressed for 10 s with 3.5 MPa force. The force was instantaneously released resulting in an explosive release of steam as the starch feedstock expanded and lled the mold. The moisture content of the feedstock inuenced the density and compressive properties of the foam. Wheat, corn and potato starch feedstock with 17, 17 and 14% moisture content, respectively, produced foam with some physical and mechanical properties similar to those of commercial food containers. The starch foam had the general shape of the mold and appeared similar to polystyrene. The microstructure of the foam revealed a cellular structure with mostly closed cells less than 1 mm in diameter. However, some regions of the foam had a microstructure similar to that of expanded polystyrene except that the cells were much smaller ( B 0.1 mm). 2001 Published by Elsevier Science B.V.
Keywords: Food packaging; Expanded polystyrene; Starch utilization

1. Introduction Single-use foam packaging is used by manufacturers to protect and preserve a wide array of food and industrial products. Expanded
* Corresponding author. Tel.: +1-510-5595677; fax: +1510-5595936. E-mail address: [email protected] (G.M. Glenn).

polystyrene (EPS) is used extensively in single-use packages due to its low density, thermal insulation, strength and low cost (AFPR, 1999). More than 1.12 105 tonnes of EPS were used for post-consumer protective packaging in the US in 1996 (AFPR, 1999). In spite of the excellent functional properties of EPS foam, there are concerns about the impact of the manufacture and disposal of EPS on the environment. These concerns have

0926-6690/01/$ - see front matter 2001 Published by Elsevier Science B.V. PII: S 0 9 2 6 - 6 6 9 0 ( 0 0 ) 0 0 0 6 0 - 1

136

G.M. Glenn, W.J. Orts / Industrial Crops and Products 13 (2001) 135143

lead to improvements in the manufacturing process and the implementation of recycling programs that have reduced the extent of EPS foam in our landlls. Notwithstanding, there continues to be an interest in developing alternative foams that are derived from renewable resources and are more degradable. Starch is one possible alternative material for making foam. Starch has been processed in various ways to make products with some properties similar to petroleum-based plastics (Otey et al., 1980; Doane, 1992; Glenn and Hsu, 1997). Starch may also be made into foam with insulating properties (Glenn and Irving, 1995) and densities (Lacourse and Altieri, 1989) that are similar to polystyrene foam. Starch is the lowest priced and most abundant worldwide food commodity (Whistler, 1984). Starch-based foam has long been a mainstay of the baking industry. The use of leavening agents for making bread, a starch-based foam, dates back to about 2600 BC (Shellenberger, 1971). A commercially successful starchbased foam product designed for an industrial application is loose-ll packaging made by extrusion. Lacourse and Altieri (1989) developed the extruded starch-based foam as an alternative to polystyrene-based loose-ll packaging. Extrusion processing made it possible to cook a starch feedstock of relatively low moisture content (14 16%) in a matter of seconds rather than minutes as required by conventional baking methods that use a feedstock with much higher moisture content (Chinnaswamy and Hanna, 1988). Extruded starch-based foam would have an even greater commercial potential if the extrusion process were more amenable to making molded foam articles such as food and beverage containers and if it were compatible with ber or other reinforcing materials. Tiefenbacher (1993) developed a technology for producing molded starch-based foam containers using a baking technology similar to that used for making wafer cookies. The process involves placing a measured amount of a dough-like material containing 7080% (w/w) moisture in a heated mold. Steam generated from the moisture of the dough acts as the blowing agent to create foam inside the mold. The high moisture content of the

doughs results in long processing times (12 min) relative to the processing time needed to make polystyrene containers. In addition, excess steam generated within the sample forges channels throughout the matrix of the foam that compromise the strength of the nished product. Improvements in the structural integrity and processing time of the foam would improve its commercial potential, but the need for high moisture levels in the feedstock is a limiting factor. Explosion pufng is an alternative process to extrusion for creating starch-based foam from starch feedstock containing low moisture levels. Popcorn is an example of a feedstock that explosion puffs naturally at about 177C and requires only 1015% moisture to achieve maximum volume (Hoseney et al., 1983). A continuous explosion pufng process has been developed to puff grains or other foodstuffs that do not puff naturally when heated (Sullivan and Craig, 1984). Explosion pufng equipment has also been developed for batch processing rice cakes or puffed grain products (Touba, 1972). Explosion pufng can produce low-density starch-based foam within several seconds, but it is not suited for making molded products as with the extrusion process. The present study describes an alternative concept for making starch-based foam that utilizes a compression/explosion process. The process consists of a compression step in which a starch feedstock is compressed in a heated mold. The clamping force is then instantaneously released allowing the feedstock to expand and ll the mold. The objective of this work was to develop a compression/explosion method for making molded starch-based foam with physical and mechanical properties similar to the foam used in commercial food packaging.

2. Materials and methods

2.1. Materials 2.1.1. Commercial containers Commercial clamshell-type containers made of EPS, corrugated cardboard (CC) and a prototype starch-based clamshell made by wafer baking

G.M. Glenn, W.J. Orts / Industrial Crops and Products 13 (2001) 135143

137

technology (WB) were obtained from Tenneco Packaging, Canandaigua, NY.

2.1.2. Starch Unmodied wheat starch (WS (Midsol 50)) was purchased from Midwest Grain Products, Atchison, KS. Regular unmodied Dent corn starch (CS (Meloga)) was obtained from the National Starch and Chemical Company, Bridgewater, NJ. Unmodied potato starch (PS) was purchased from Avebe-America, Inc., Princeton, NJ. 2.2. Methods 2.2.1. Starch feedstock preparation A suspension of WS, CS or PS in water (10%, (w/w)) was placed into a 12-l stainless steel container and heated with constant stirring in a boiling water bath. Viscosity readings were taken intermittently with a viscometer (Brookeld, model RVT, Stoughton, MA). The suspension was heated until the viscosity no longer increased. The starch melt was covered and chilled (5C) overnight to allow gelation to occur. Starch gel (1.8 kg) and dry starch powder (3.2 kg) were thoroughly mixed and added into the hopper of a single-screw food extruder (Ambrette Machinery Corp., Brooklyn, NY). The starch was extruded without adding heat to the extruder barrel section although the mechanical work of the extrusion process itself increased the temperature of the starch to some extent. The starch mixture was extruded through a die and chopped by a rotating blade producing aggregates the size and shape of rice grains (approximately 2 7 mm). The starch aggregates were dried in a forced-air oven (60C) for 24 h. Starch feedstock was prepared by reconditioning the aggregates to the desired moisture content. Reconditioning involved adding predetermined amounts of water as a mist to starch aggregates that were sealed in a plastic bag, agitated periodically and incubated for at least 24 h to allow equilibration. 2.2.2. Compression/explosion process An aluminum mold was constructed where the

top half had an insert with greater depth than the cavity in the corresponding lower half of the mold (Fig. 1ac). The design ensured that a predetermined amount of force would concentrate solely on the starch feedstock. The overall dimensions of the mold were 152 mm in length and width (square) and 75 mm thick. The cavity of the bottom half (Fig. 1b and c) was 10 mm deep, 38 mm wide and 102 mm in length. The top-half was 13 mm deep, 36.5 mm wide and 100.5 mm long allowing it to slide into the cavity. The mold was mounted in a hydraulic press (Carver Model C) between two temperature-controlled platens that heated the cavity to 230C. The mold was lled with 10 g of feedstock aggregate and immediately closed with 3.5 MPa clamping force for 10 s. Upon releasing the hydraulic pressure and allowing the mold to open, the starch feedstock expanded into a foam that partially lled, completely lled or overlled the mold cavity depending upon the moisture content of the feedstock. The foam materials had a nal moisture content of approximately 2% and were stored in plastic bags while awaiting mechanical tests.

2.2.3. Scanning electron microscopy Samples of starch foam were cut with a singleedged razor blade, using a slicing motion to avoid crushing the sample. The cut samples were glued to aluminum specimen stubs using twocomponent 5-min epoxy. Samples were gold coated in a Polaron E5100 sputter coater (Hateld, PA). These were viewed in a Hitachi S530 scanning electron microscope (SEM) (Tokyo, Japan). 2.2.4. Physical and mechanical tests Foam density was determined by volume displacement. Compression tests were performed using a universal testing machine (Instron, Model 4500, Canton, MA). The samples were deformed in compression (2.5 mm/min) until a yield point was reached or until the specimen was compressed beyond 10% deformation as specied in standard testing procedures for rigid cellular plastics (D1621-73, Annual Book of ASTM Standards, 1973).

138

G.M. Glenn, W.J. Orts / Industrial Crops and Products 13 (2001) 135143

2.2.5. Experimental design Properties of three commercial clamshell-type containers (EPS, CC and WB) were studied to provide a reference for starch-based foam made by the compression/explosion process. Starchbased foam was made using feedstock conditioned to 8, 11, 14, 17 and 20% moisture on a dry-weight basis. Three samples were prepared for each treatment made from WS, CS or PS. Data were analyzed (Sigma Stat, San Raphael, CA) by analysis of variance (ANOVA).

3. Results The commercial cartons, though functional, varied widely in their physical and mechanical properties (Table 1). The EPS container was approximately one-third the density of the WB product, but also had a much lower compression modulus and strength. The WB container was the densest, but had intermediate strength and modulus values. Compression/explosion processing was a rapid and efcient method of making molded starch-

Fig. 1. Top (a), side (b) and front (c) views of mold design used in compression/explosion process. See text for details of operation.

G.M. Glenn, W.J. Orts / Industrial Crops and Products 13 (2001) 135143 Table 1 Density, compressive modulus and compressive strength of commercial food cartons made of corrugated cardboard (CC), extruded polystyrene (EPS) and starch-based ingredients using a wafer baking (WB) process Sample Density (g/cc) Compressive modulus (MPa) Compressive strength (MPa) CC 0.16aa 3.6a 0.21a EPS 0.074b 0.48b 0.048b WB 0.23c 1.2b 0.13b

139

a Values within rows that are followed by a different letter are signicantly different at the 95% condence level.

based foam. The physical and mechanical properties of starch products varied depending on the moisture content of the feedstock. Wheat and corn starch feedstock with low moisture content ( 5 11%) yielded products with very little expansion that did not ll the mold cavity and had a yellow, translucent appearance. Similarly, samples made of potato feedstock with 8% moisture content were solid bars with a well-dened shape. Feedstocks with higher moisture content produced foams that were white, opaque, and that expanded readily and completely lled the mold. The lower surface of the molded foam was generally well formed, whereas the upper surface was often irregular. When the moisture content of the feedstock was raised to 20% some material escaped along the edges of the mold cavity, but a molded sample of precise dimensions was produced. Samples produced from feedstock with different moisture contents varied widely in mechanical properties. Stress/strain (Fig. 2) curves show that the samples failed in compression before 10% strain except for samples made from feedstock containing 20% moisture. Samples made of feedstock containing 20% moisture have a fairly linear increase in stress with strain up to 10% strain. A large decrease in stress:strain ratios occurred when the moisture content of the WS feedstock was increased from 14 to 17% (Fig. 2). Similar results were obtained for the CS samples. In contrast, the largest decrease in stress:strain ratios for samples made of PS feedstock was observed when the moisture content was increased from 11 to 14% (data not shown). PS foam had much

Fig. 2. Stress/strain curves for samples of wheat starch foam tested in compression. The compressive properties varied widely among samples made of feedstock equilibrated to moisture levels ranging from 11 to 20%. Note that the foam samples eventually fractured with increasing compressive strain except for samples made of feedstock containing 20% moisture.

lower stress at a given level of strain than either WS or CS samples made of feedstock containing 1417% moisture (Fig. 3). Compressive strength was greatest for WS and CS samples made from feedstock with 14 and 17% moisture, respectively (Table 2). WS and CS feedstock with 8% moisture content did not form a cohesive molded sample and could not be

Fig. 3. Stress/strain curves for foams made of CS, WS or PS feedstock containing 14% moisture.

140

G.M. Glenn, W.J. Orts / Industrial Crops and Products 13 (2001) 135143

Table 2 Compressive strength, density and Youngs Moduli of foam samples made from wheat, corn and potato starch feedstock conditioned to different moisture levels Water content (%) Physical/mechanical properties Compressi6e strength (MPa) Wheat Corn Potato Density (g/cm 3) Wheat Corn Potato Youngs Moduli (MPa) Wheat Corn Potato
a b

8 N/Aa N/A 2.14 0.86a 0.97a 0.81a N/A N/A 40

11 2.42ab 1.41a 1.33a 0.92a 0.73a 0.25b 70a 40a 45a

14 2.75a 2.43a 0.51b 0.36a 0.31a 0.093b 80a 104a 15b

17 0.259a 0.474b 0.169c 0.087a 0.13a 0.081a 6.7ab 14.5a 3.5b

20 0.127a 0.333b N/A 0.078a 0.089a N/A 1.9a 5.9b N/A

Not applicable. Values within columns and within each subsection that are followed by a different letter are signicantly different (0.05 level).

tested. In contrast, the PS feedstock with 8% moisture formed into a cohesive sample and was tested. However, PS feedstock with 20% moisture formed poor samples due to excessive expansion and sticking and was not tested. The compressive strength of samples made from feedstock containing 1720% moisture was higher with CS than WS (Table 2). The density of the starch samples ranged from 0.97 g/cm3 for samples made of feedstock with 11% moisture or less to 0.078 g/cm3 for samples made of feedstock containing 14% moisture or higher (Table 2). There was no signicant difference in the density of WS and CS samples made of feedstock equilibrated to the same moisture content. The density of WS and CS samples decreased the most as the moisture content of the feedstock increased from 11 to 17%. The density of PS samples decreased the most as the moisture content of the feedstock increased from 8 to 14% (Table 2). The Youngs modulus for the starch samples generally decreased as the sample density decreased below approximately 0.25 g/cc. (Table 2). The microstructure of the samples varied with the moisture content of the feedstock. The low moisture samples were relatively dense, had a

closed-cell structure with relatively large cells and thick cell walls (data not shown). Foam made of feedstock with at least 14% moisture content had thin cell walls and large cells that were non-uniform in size (Fig. 4a and b). The cells of the starch foam were generally larger than those of EPS foam (compare Fig. 4a and d). The starch material that ashed out of the mold during the compression step of the compression/explosion process formed into white, opaque, low-density foam with an appearance very similar to the EPS foam. The cell size of the ash-formed starch foam was generally much smaller than the cell size of EPS foam (compare Fig. 4c and d). The cell shape of the EPS foam was skewed in the direction of expansion (Fig. 4d). Some of the cells of the ash-formed starch foam were similarly skewed in the direction of expansion (Fig. 4c).

4. Discussion Materials that include paper, corrugated paperboard, plastic, polystyrene, glass and metals are currently manufactured into a wide variety of packaging products. Packaging products protect goods from mechanical damage, contain the

G.M. Glenn, W.J. Orts / Industrial Crops and Products 13 (2001) 135143

141

goods and preserve their quality from environmental elements such as heat, cold, moisture, atmospheric gases, light, etc. Concern about the environmental impact of single use packaging has been raised due to the enormous quantities of packaging products consumed. More than 100 billion aluminum cans, billions of glass bottles, 5 million tonnes of paper and thousands of tonnes of plastic are consumed each year for packaging purposes (Andersen and Hodson, 1996). The impact of single-use containers on the environment has been the focus of public attention in recent years. Polystyrene products are made using hazardous solvents and blowing agents that diminish the ozone layer. In addition, polystyrene degrades very slowly, accumulates in the environment and is derived from a nonrenewable resource. Other packaging materials also have adverse affects on the environment. Paper products have been perceived as more compatible with the environment because they are degradable and derived from a renewable resource. However, studies indicate that the paper industry has a greater negative impact on the environment than the polystyrene industry (Andersen and Hodson, 1996). In 1991, the Environmental Protection Agency reported that the paper industry was the third largest source of toxic pollutants in the US,

with mills releasing over 91 000 tonnes of toxic materials (Clean Water Action Council, 1999). Metal, glass and ceramic containers ultimately, also have a negative environmental effect. Signicant air and water pollution and high-energy consumption can all be attributed to the manufacture of metal, glass and ceramic containers. Starch is an inexpensive biopolymer that is abundant and widely available. Starch production from crops such as wheat or corn exact a small toll on the environment since production requires the expenditure of fuel and the use of fertilizers and pesticides that can contaminate ground water. However, starch is a renewable resource and easily degrades in nature. Starch-based feedstock has been molded into containers using a wafer process (Tiefenbacher, 1993; Haas et al., 1996). Coatings must be applied to protect the containers from breakdown in the presence of moisture. Shogren and Lawton (1998) described a coating material and process that provided stability to starchbased containers exposed to moisture. Composites of starch, ber and other proprietary ingredients are being used to make commercial food containers using the wafer baking (WB) process (Andersen and Hodson, 1996). The starch-based containers were coated to provide moisture resistance. While the starch-based containers are

Fig. 4. Scanning electron micrographs of wheat starch foam made by the compression/explosion process (A C) and extruded polystyrene foam (D). Scale bars; A, 1 mm, B, 0.2 mm, C, 0.5 mm, D, 1 mm.

142

G.M. Glenn, W.J. Orts / Industrial Crops and Products 13 (2001) 135143

promising, the process is somewhat limited because it requires more than 1 min to make each container. The compression/explosion process used in the present study provided a marked advantage over the WB process for making starch-based products. The feedstock was easily fed into the mold and the processing time was only 10 s. The density of the WS, CS and PS samples prepared from feedstock equilibrated to 17, 17 and 14% moisture content, respectively, was similar to that of the EPS sample, while the compressive strength of the starch samples was greater. However, unlike the commercial packaging samples that deformed in compression without ever fracturing, the starch samples eventually fractured during deformation and had a much higher compressive modulus indicating greater stiffness. This property would make the starch foam less suited for containers such as clamshells that are designed to bend. Andersen and Hodson (1996) were able to form starch-based containers that could bend without fracturing by incorporating ber in their formulations and by providing a exible coating. It may be possible to improve the mechanical properties of compression/explosion formed foam by incorporating bers, llers, plasticizers and/or coatings to form containers that bend. Coatings would also provide a moisture barrier to make the foam moisture resistant. Moisture resistant coatings such as those used for WB containers would be necessary to make the starch-based foam functional in moist or humid environments. Finally, the surface quality of the starch-based foam should be improved to achieve commercial quality. The surface should be uniform and smooth, the molded shape should be precise and consistent and the amount of material lost to ashing should be minimized. In the present study, the starch foam was molded into a desired shape. However, the upper surface of the molded product was at times uneven and the corners of the mold did not always ll properly. One possible cause for the corners not lling is that the ow of the starch melt was inhibited by air trapped in the corners. Andersen and Hodson (1996) found with the WB technology, that failure to remove air from the mold resulted in air pockets in the

nished product. High melt viscosity and surface tension could also affect the ability of the starch melt to ll the mold. One solution to the problem of air pockets is to include air vents in the mold design. Andersen and Hodson (1996) found that the proper placement of small vents in the mold helped eliminate trapped air and assisted the uniform ow of the starch melt into all regions of the mold. Improvements in mold design should improve the quality and uniformity of the molded starch foam. 5. Conclusion The compression/explosion process used in this study is promising because it produces starchbased foam in 10 s. The compressive properties and density of the foam are comparable with that of commercial products with the exception that the foam fractures under excessive strain. The compression/explosion process has an advantage over traditional pufng or extrusion technology in that the product can be molded during formation. Further advances in mold design and the incorporation of bers or llers may help provide for improved mechanical properties, more uniformity and better surface quality. Acknowledgements The authors thank Pauline Yu for her assistance with SEM micrographs. References
Alliance of Foam Packaging Recyclers (AFPR), 1999. EPS physical properties. 2124 Priest Bridge Drive, Suite 19, Crofton, MD 21114, http://www.presstar.com/afpr/ fnf.html. Andersen, P.J., Hodson, S.K., 1996. Molded articles having an inorganically lled organic polymer matrix. U.S. Patent No. 5,545,450. Annual Book of ASTM Standards, 1973. In: Compressive Properties of Rigid Cellular Plastics, Vol. 08.01. Philadelphia, PA, pp. 1621 1673. Chinnaswamy, R., Hanna, M.A., 1988. Optimum extrusioncooking conditions for maximum expansion of corn starch. J. Food Sci. 53, 834 840.

G.M. Glenn, W.J. Orts / Industrial Crops and Products 13 (2001) 135143 Clean Water Action Council, 1999. Clean paper-making. 1270 Main Street, Suite 120, Green Bay, WI 54302, http:// www.cwac.net/cleanpapermaking.html. Doane, W.M., 1992. USDA research on starch-based biodegradable plastics. Starch 44, 293295. Glenn, G.M., Irving, D.W., 1995. Starch-based microcellular foams. Cereal Chem. 72, 155161. Glenn, G.M., Hsu, J., 1997. Compression-formed starch-based plastic. Ind. Crops Prod. 7, 3744. Haas, F., Haas, J., Tiefenbacher, K. 1996. Process of manufacturing rottable thin-walled starch-based shaped elements. U.S. Patent No. 5,576,049. Hoseney, R.C., Zeleznak, K., Abdelrahman, A., 1983. Mechanism of popcorn popping. J. Cereal Sci. 1, 4352. Lacourse, N.L., Altieri, P.A., 1989. Biodegradable packaging material and the method of preparation thereof. U.S. Patent No. 4,863,655. Otey, F.H., Westhoff, R.P., Doane, W.M., 1980. Starch-based

143

blown lms. Ind. Eng. Chem. Prod. Res. Dev. 19, 592 595. Shellenberger, J.A., 1971. Production and utilization of wheat. In: Pomeranz, Y. (Ed.), Wheat Chemistry and Technology. American Association of Cereal Chemists, St. Paul, MN, pp. 1 14. Shogren, R.L., Lawton, J.W., 1998. Enhanced water resistance of starch-based materials. U.S. Patent No. 5,756,194. Sullivan, J.F., Craig, J.C., 1984. The development of explosion pufng. Food Technol. 38 (52-55), 131. Tiefenbacher, K.F., 1993. Starch-based foamed materials-use and degradation properties. J.M.S.-Pure Appl. Chem. A 30 (9-10), 727 731. Touba, A.R., 1972. High pressure process for making puffed food product and product. U.S. Patent No. 3,650,763. Whistler, R.L., 1984. History and future expectation of starch use. In: Whistler, R.L., Bemiller, J.N., Paschall, E.F. (Eds.), Starch Chemistry and Technology. Academic Press, San Francisco, pp. 1 9.