Additives for Improved Processing of Microcellular Polyurethane Shoe Sole Systems

Gary D. Andrew Air Products & Chemicals, Inc 7201 Hamilton Blvd Allentown, PA 18195 James D. Tobias Air Products & Chemicals, Inc 7201 Hamilton Blvd Allentown, PA 18195

ABSTRACT The manufacturers of microcellular polyurethane articles require better processing aids to meet the escalating pressures of cost efficiency and complicated part designs. Additives such as surfactants and catalyst play a key role in the production of shoe sole systems. Intrinsic characteristics such as flowability and demoldability are affected by the type and amount of additives used during the production cycle. Surfactants and catalysts play a significant role in controlling foam cellular structure, reactivity, and physical properties. Densities in polyurethane shoe sole systems are trending downward to provide lower cost and increased comfort. One of the technical challenges that the industry currently faces with the reduction of density is maintaining dimensional stability and part integrity. This paper will discuss and demonstrate the utilization of new surfactant and existing catalyst technologies impact to the polyurethane shoe sole producer to meet the current processing demands of today. Furthermore, Air Products and Chemicals, Inc. has developed new surfactant technology that will enhance dimensional stability and improve skin definition to produce good quality shoe sole parts using all water blown polyester systems, while maintaining or improving key physical property characteristics.

INTRODUCTION Since they were first introduced in the 1960s polyurethane materials, with their unique blend of properties, durability and lightweight characteristics have played an important role in the development of many footwear component applications. These component applications have grown to include outsoles, midsoles, and sockliners. Judicious selection of catalyst and surfactant additives for material considerations, performance requirements, and processing characteristics

of polyurethane footwear must be considered when formulating these types of production systems. The reduction in density of polyurethane systems has required a fundamental change in the cellular nature of the foam. From the processing point of view, polyurethane foams molded at or below 0.30 g/cc densities requires controlling the cell structure to avoid shrinkage. Although the concept is indeed simple, the complex nature of polyurethane elastomers requires that careful consideration be given to the relationships between the structure, properties and processing characteristics of the material to achieve density reductions and maintain mechanical properties. Global production of shoes is expected to be at greater than 12.5 billion pairs for 2002. Polyurethane is used in 6% of the global production of shoes [1]. The footwear market remains an important polyurethane market segment that continues to grow. Microcellular polyurethane elastomers have long been used as soling materials because of their high wear resistance and light weight. Recent pressures to improve costs and increase wear comfort, as well as competition from non-PU materials, have presented both challenges and opportunities for this industry segment. Soling designs have become more complex, requiring polyurethane systems to have excellent flow to fill mold cavities and produce an aesthetically attractive surface. Lower densities are required to further reduce cost and provide increased wear comfort. At lower densities, polyurethane processing becomes difficult as a result of excessive shrinkage, poor skin definition and poor flow characteristics. Additives such as catalyst and surfactants play a key role in the production of microcellular systems and can be used to address many of the problems facing our industry today. The requirement for polyurethane catalysts, which provide delayed initiation and accelerated cure to maintain good flow, in combination with shortened demold times, has been long standing. This paper will also highlight the use of existing catalysts as well as new

surfactant chemistry designed to achieve these goals in all water blown, ethylene glycol extended polyester shoe sole systems. Additionally, we will review the role surfactants play to improve dimensional stability, and surface quality. Utilizing experimental design protocols, we have established surfactant structure-property relationships and identified new surfactant structures that provide improved dimensional stability and abrasion resistance in typical densities as well as low-density polyurethane systems. EXPERIMENTAL All Water Blown Microcellular System Handmix techniques and customer trials were used to evaluate existing catalyst technology and demonstrate new surfactant chemistries in all water blown microcellular systems. Isocyanate and premix temperatures were maintained at 45C for the laboratory segments of the evaluations. Laboratory experiments also utilized molded parts that were cast for physical property testing using an aluminum mold heated to 55C. The measurements of the mold were 25cm(l) x 15.2cm(w) x 1.0 cm(h). Index latitudes were run for each surfactant and catalyst package evaluated to determine the optimum index for performance as measured by pull time and demold times. Eight plaques were made at this optimum index and physical properties were determined using the following methods: Parts were cast in a mold with the measurements of 25cm(l) x 15.2cm (w) x 1.0 cm(h). A robotics arm was Dimensional used to accurately measure part Stability dimensions after 48 hours. The difference was reported as shrinkage or post blow. Tensile ASTM-D638 Split Tear ASTM-D3574 Test F Abrasion ASTM-1034 Elongation D-214 Shore A ASTM D-2240 The following materials were used for laboratory and customer trials: 33% DABCO crystal in 67% Dabco E.G. ethylene glycol. 25% DABCO crystal in 75% Dabco Dabco S-25 BDO. Proprietary amine catalyst for ethylene glycol extended systems. Dabco 1027 Designed to reduce demold and/or extend initiation times. Proprietary amine catalyst for 1,4 butanediol extended systems. Dabco 1028 Designed to reduce demold and/or extend initiation times.

Dabco DC-193

Dabco DC-3043

Dabco DC-3042 RCT 3358 Polyester Polyol MDI pre-polymer

Silicone copolymer surfactant commonly used in shoe soles. New silicone copolymer surfactant designed to improve dimensional stability used as a co-surfactant with DC-3042. New silicone copolymer surfactant designed to improve cell structure. Used as a co-surfactant with DC3043. Solvent based release agent from Chem Trend. Polyester polyol with an OH# of 45.5. MDI pre-polymer with an NCO of 17.0%.

Catalysts for Processing Improvements Two key objectives for the catalysis of polyurethane footwear are to increase the cream time, with a general reduction of the initial foaming rate, and to decrease the demold time, by increasing the rate of cure after the foam has stopped flowing within the mold. Both of these objectives help to increase the productivity of polyurethane footwear parts made in a manufacturing production process. In a typical footwear environment, multi-cavity molds are filled by multiple shots from a single mix head on a low-pressure foam machine. A longer delay to cream time and a slower foaming rate allows more cavities to be filled with ample time to close the mold lid before rising foam expands above the parting line. Decreasing the demold time by increasing rate of foam cure, demoldability can be reduced which allows for higher production rates. It is critical that the catalyst used to improve the curing rate does not adversely affect the cream time and the flowability of the polyurethane system. With many types of catalysts these two objectives are difficult to obtain simultaneously, since affecting one objective positively affects the other objective negatively. As the polyurethane industry competes for market share against alternative materials, manufacturers need to continue to develop lower cost, higher performance products. Lowering overall processing costs is one way to achieve this. Dabco 1027 and 1028 are two catalysts designed to give the formulator the ability to dramatically change processing characteristics and adjust overall system reactivity to meet equipment, productivity and design needs. When used with a co-catalyst they can provide longer initiation times and/or decreased demold times. By extending initiation time, lower repair and rejection rates can be achieved through more complete mold fill, even with complex mold designs. Extended initiation also allows for longer intervals between screw cleaning which increases productivity by minimizing

production down time. Decreasing demold times, while maintaining equal initiation times, can lower cycle times, thereby raising production outputs with no loss in product quality. Both Dabco 1027 and Dabco 1028 catalysts were designed to impact the back end reactivity in microcellular systems. The catalytic component is chemically identical in Dabco 1027 and 1028 catalysts. The catalysts differ only in their concentration and carrier type to tailor their performance to different extender systems. Dabco 1027 was developed for use in ethylene glycol extended systems, and is often used as a co-catalyst with Dabco EG. Dabco 1028 was developed for 1,4-butanediol extended systems, and is used as a co-catalyst with Dabco S-25. The work conducted for this paper utilizes Dabco 1027 and Dabco EG in ethylene glycol extended formulations to demonstrate catalyst performance and formulating flexibility. Similar results can be seen for Dabco 1028 and Dabco S-25 using 1,4-butanediol extended systems. Catalysts were evaluated in a representative polyester 0.30 g/cc mid-sole formulations shown in Table 1. Dabco 1027 catalyst is unique because it strongly affects backend reactivity or demold time but only has a slight effect on initiation time. In contrast, Dabco EG affects both initiation time and demold times. Figure 1 graphically illustrates the effect of simply lowering Dabco EG levels from 1.5 to 1.0 parts to achieve longer initiation times. Initiation time was extended from 11 seconds to 16 seconds, but demold time has also been extended from 5 minutes to 6 minutes. The reactivity data shown in Table 2 demonstrates how Dabco 1027 is utilized to alter reactivity profiles to achieve reduced demold or extended initiation times. Demold times can be reduced by 30%

when using Dabco 1027 as a co-catalyst with Dabco EG, without effecting initiation. In contrast, initiation times can be extended by up to 35% without extending demold. Control formulation I, listed in Table 1, has an initiation time of 11 seconds with a demold time of 5 minutes at an optimum index of 98. Reducing Dabco EG to 1.0 part and adding 0.2 parts of Dabco 1027, formulation II, extends initiation time to 17 seconds while maintaining final demold at 5 minutes at a optimum index of 100. To achieve optimum demold times, without adversely affecting the initiation time 1.0 part of Dabco EG is used with 0.6 parts of Dabco 1027, as noted in formulation III, listed in Table 1. Using this formulation, demold time is reduced to 3 minutes while maintaining the 11 second initiation time at an optimum index of 100. In some systems using Dabco 1027 or Dabco 1028 catalysts, an index change is required to achieve optimal demold times. An upward shift of one to three index points may be necessary. Figure 2 shows this effect. Despite this shift, the shape of the curve which equates to the processing latitude of the catalysts, is approximately the same for either Dabco EG alone or a blend of Dabco EG and Dabco 1027 catalysts. It should be noted that the index shift is not seen in all systems, and the optimum index must be determined for each formulation. For the polyester system used in this paper, Dabco 1027 requires an increase in index by two points to maximize performance. Despite the dramatic reduction in demold achieved with Dabco 1027 catalyst, elastomer property data detailed in Table 3 show there is little effect on foam physical properties, even with the index shift.

Table 1: Ethylene Glycol Extended Shoe Sole Formulation Formulation Identification Dabco EG Dabco 1027 Dabco 1027 I II III Components PPHP PPHP PPHP Polyester Polyol 45.5 OH# 100.0 100.0 100.0 Ethylene Glycol 8.0 8.2 8.0 Water 1.0 1.0 1.0 DC-193 0.5 0.5 0.5 Dabco EG 1.5 1.0 1.0 Dabco 1027 -0.2 0.6 MDI-Pre-polymer Index

98

100

100

Index variability studies were conducted on each formulation to determine optimum index for reactivities.

Dabco EG 1.5 pphp Decrease in Dabco EG 1.0 pphp Increasing Processing Time
Figure 1. Reactivity changes with reduced Dabco EG levels. Dabco EG 1.5 parts versus Dabco EG 1.0 parts.

Formulation I II III

Index 98 100 100

Table 2: Reactivity Data Dabco EG/Dabco 1027 Initiation Top Of Cup Tack Free (Seconds) (Seconds) (Seconds) 11 22 65 17 28 68 10 18 54

Pull Or Pinch (Seconds) 140 143 118

*Demold (Minutes) 5 5 3

*Demold time is based on 90 degree bend test. No crack, part is considered ready for demolding.

Formulation

I II III

Ross Flex Average cut growth 50K cycles (mm) 21.25 11.50 13.50

Table 3: Physical Property Data Shoe Sole Formulation Tensile Tear Elongation Abrasion Measured (kPa) (Kg/cm) at break average Part (%) weight loss Shrinkage (grams 1K (% Shrink) cycles) 1981 4.50 404 .05 6 1823 5.25 344 .06 7 1925 5.45 349 .05 9

Density molded part (g/cc)

Shore A Hardness

0.30 0.31 0.30

25 28 27

160 Dabco EG 1.5 pphp

150

Pull 140 Time (sec)

130 Dabco EG 1.0 pphp Dabco 1027 0.6 pphp 92 94 96 98 100 102 104

120

Index
Figure 2. Index versus pull time reduction: DABCO EG 1.5 pphp versus DABCO EG/DABCO 1027 1.0/0.6 pphp

SURFACTANTS FOR MICROCELLULAR SHOE SOLE SYSTEMS Typical surfactants used in polyurethane footwear systems need to provide uniform cell structure, good foam stability and impart good dimensional stability. Many common microcellular foam surfactants can accomplish the first two requirements. However, foam dimensional stability is often a problem when using conventional surfactants especially in low-density systems. Problems with foam dimensional stability typically are part shrinkage or post expansion, which in extreme cases can cause foam splitting. Post expansion may take a permanent set if the foam is slightly uncured. If gas pressure within the cells is extreme, then the foam may split to relieve the internal stresses. Post expansion generally becomes more severe as foam-packing density is increased. Conversely, when packing density of molded parts is closer to free-rise density, shrinkage may occur as the hot gas within the cells cools, contracts, and diffuses out of the cells creating a partial vacuum. Therefore, foam dimensional stability can be improved by opening the cells of the foam. By allowing the internal pressure within the foam cells to equilibrate with atmospheric pressure, the majority of forces causing dimensional instability are relieved. Accordingly, surfactants that allow cell opening to occur without detrimental effects on cell structure are useful in many types of foams.

As pressures increase on producers of polyurethane microcellular parts to achieve lower densities, surfactant performance is becoming more critical. Surfactants can play a key role in modifying system reactivity and flow, as well as improving elastomer dimensional stability and surface quality. By modifying surfactant structure, it is possible to tailor performance for specific applications. Dabco surfactants are members of the general class of nonhydrolyzable, coupled polydimethylsiloxane /polyalkyleneoxide graft copolymers as illustrated in Figure 3 [2]. The true chemical identity of a silicone surfactant is actually a complex reaction product of two polymeric materials, namely polydimethylsiloxane and polyalkyleneoxide, each having their own characteristic number and mass averaged molecular weights. For the idealized structure represented in Figure 3, x and y correspond to the average number of dimethylsiloxy and methylpolyethersiloxy groups per molecule, respectively. Similarly, m and n indicate the average number of polyethylene oxide and polypropylene oxide units, respectively; and R is the polyether-capping group. DP (Degree of Polymerization) is simply the average number of silicon atoms in the silicone backbone of the surfactant and is calculated using the formula x + y + 2. The x/y ratio, (commonly referred to as the D/D ratio), DP, (the length and composition of the polyether), and the capping group are important for optimizing surfactant performance [3].

Me Me | | M e 3 S i O ( S i O )x ( S i O )y S i M e3 | | Me ( C H 2)3 | O | ( C H 2 C H 2 O )m | ( C H C H 3 C H 2 O )n | R
Figure 3. General surfactant structure. Design Set 1

Design Set 2
E

Control A DC-193

F 0.0

D/D Ratio

Figure 4. Surfactant experimental design.

emulsify the reacting components and provide more intimate mixing [4]. Key physical properties from both design sets are provided in Table 5. Surfactant A (DC-193) gave satisfactory results when physical properties were evaluated. However, surfactants B through E did not exhibit good dimensional stability in the 0.30 g/cc lower density system. Dimensional stability was greater than or equal to 8% with these surfactants. Based on the results of the first design set, we concluded that lower density shoe sole systems may need very different surfactant structures to achieve the required dimensional stability and skin definition. The second design set, again illustrated in Figure 4, is based on a different DP and D/D structural range and utilizes a new polyether. Initial screening of various surfactant types indicated that this design space would provide improved performance in low-density water blown polyether and polyester systems. Physical property data for polyester parts made with surfactants F through H are found in Table 5. This second design set showed improvements in dimensional stability, Ross-Flex and abrasion resistance. All of the surfactants in this second design set gave zero cut growth after 50,000 Ross-Flex cycles. The most significant improvement was noted in dimensional stability. Surfactant samples I and J showed the best dimensional stability. Shrinkage was reduced significantly when compared to the control surfactant. Demold times were also improved with the second design set. Demold times were reduced by as much as 30 seconds without comprise to physical properties. Demold reductions are attributed to improved emulsification of all components, thereby allowing better utilization of the catalyst, and increasing catalyst mobility throughout the foaming process.

Degree of Polymerization (DP)

Surfactant Experimental Designs Two surfactant experimental design sets shown in Figure 4 were evaluated in the low-density high hardness polyester midsole formulation found in Table 4. These design sets allowed us to identify structure versus performance benefits. Design set 1 (points A through E) encompasses a design space that utilizes typical DP and D/D ratios used in microcellular applications today, such as Dabco DC-193. The polyether for this surfactant design set is also typical for these surfactant types and was kept constant. Point A in this set corresponds to the industry standard, Dabco DC-193. Small differences in initiation and demold times were observed when evaluating this design set. Dabco DC-193 provided the best reactivity, with a 11 second initiation and 5 minute demold. Surfactants B, C, D and E gave extended demold times from 6 to 6 minutes. We hypothesize that the shorter demold time is a result of Dabco DC-193s ability to better

Table 4 : Surfactant Experimental Design Formulations Formulation Identification Dabco EG I Components PPHP Polyester Polyol 45.5 OH# 100.0 Ethylene Glycol 7.0 Water 1.0 Surfactant (Varied) 0.5 Dabco EG 1.5 MDI Pre-polymer (17.0% NCO) Index Density (g/cc)

98 0.30

Table 5 : Silicone Surfactant Study Physical Properties Ross-Flex Tensile Tear Elongation Abrasion average cut (kPa) (Kg/cm) at break Average weight growth (%) loss (grams 1K 50K cycles cycles) (mm) Control A 21.25 1981 4.50 404 0.05 B 10.80 1925 5.34 356 1.3 C 7.65 1850 4.56 356 1.2 D 12.50 1765 4.78 347 1.4 E 0 1995 6.56 419 0.9 F 0 1986 5.89 429 0.4 G 0 1956 6.78 410 0.35 H 0 1945 6.56 498 0.10 I 0 1935 6.35 510 0.55 J 0 1989 7.01 515 0.01 Index for surfactant study equals 98 for all tests. Surfactant

Measured part shrinkage (% shrink) 6 13 11 10 8 1.3 4 4 0.8 0.5

Shore A Hardness

25 29 32 32 28 34 35 36 34 37

NEW SURFACTANT TECHNOLOGY DEVELOPED Using design of experiment methodology, we were able to identify key surfactant structural performance characteristics. However, producers require both dimensional stability and surface improvements, which will require a totally new approach with surfactant technologies. Typically, producers of polyurethane shoe soles will use a single surfactant when producing polyurethane footwear. The result output of the design of experiment indicated dimensional stability and surface benefits could be achieved when two very differing surfactant chemistries were applied. This would allow the customer to tailor performance based on their specific processing needs and specific formulation. Formulation differences at customers field sites would require different ratios or use of only one new surfactant technology to allow for total optimization and provide

maximum performance. Surfactants DC-3042 and DC-3043 were developed to address these processing challenges and issues. DC-3042 used as a single surfactant, was designed to provide excellent fine cell structure and surface quality as evidenced by comparing the SEMs in Figures 5 and 6. DC-3042 surfactant was run at 0.5 phpp using formulation III listed in Table 1. DC-3042 also provided a slight improvement in reactivity reducing demold by approximately 15 seconds over the control surfactant DC-193. However, significant improvement, in surface and cell structure, without loss of key physical properties was obtained as shown in Table 6. Comparing DC-3042 to a competitive surfactant, surfactant X, a clear advantage in reactivity, Ross-Flex, surface quality and dimensional stability were obtained. Physical properties are outlined in Table 6.

Figure 5 : Control Surfactant

Figure 6 : Dabco DC-3042

Figure 7: Dabco DC-193

Figure 8 : Dabco DC-3043

DC-3043 was not initially designed for use as a single surfactant. However, trials at producers of polyurethane shoe soles have been able to optimize their system to use DC-3043 as their sole surfactant. The unique structure of DC-3043 has a significant impact on dimensional stability. To achieve good dimensional stability foam cell structure must be open. It is also important that the cells at the surface are collapsed and form a thick smooth skin to prevent water absorption and at the same time provide good abrasion resistance. Figure 7 shows cells at a 1,000 magnification for a part made with DC-193. Figure 8 shows cells made at a 1,000 magnification using DC-3043. As the polyurethane is reacting and the cell is forming when using DC-3043 the cell wall ruptures imparting excellent dimensional stability. Surface quality is also

improved as shown in Figures 9 and 10. Improved surface quality will impart better abrasion resistance. When using DC-3043 as single surfactant, reactivity remains unchanged. Laboratory results demonstrated a slight loss of tensile and elongation properties, however, significant improvements in dimensional stability were obtained. Other physical property results with DC-3043 are provided in Table 6. Dabco DC-3042 and DC-3043 can be used as cosurfactants. Laboratory work and customer trials have demonstrated excellent physical properties and dimensional stability when these new surfactants are combined. However, the blend ratio of the two surfactants may change dependent on the formulation being used. Key physical properties for this work are recorded in Table 6.

Figure 9 : Dabco DC-193

Figure 10: Dabco DC-3043

Table 6: Impact of Surfactant Blends on Reactivity, Key Physical Properties, Surface and Dimensional Stability Surfactant Demold Ross-Flex Tensile Tear Elongation Cell Dimensional (pphp) Time average cut (kPa) (Kg/cm) at break structure stability (Minutes) growth 50K (%) size at (%) cycles surface (mm) (mm) DC-193 (0.5) 5.5 21.25 1981 4.5 404 1.5 6 DC-3042 (0.5) Surfactant X (0.5) DC-3043 (0.5) DC-3042 (0.4) DC-3043 (0.1) 5.25 5.75 5.5 5.25 16.43 36.75 18.95 15.45 1960 1990 1850 1995 4.6 4.7 4.8 4.4 380 410 365 415 1.0 1.5 2.0 <1.0 5 7 1 1

CONCLUSION Dabco DC-193 silicone surfactant remains an industry standard for shoe sole applications providing excellent properties and performance in high-density systems. New surfactant structures have been developed to meet growing needs to reduce densities while maintaining key physical properties. These new surfactants were developed using experimental design methodologies. DC3042 improves cell structure and surface quality. DC-3043 improves dimensional stability. When using these surfactants as co-surfactants we provide the customer the ability to adjust surfactant chemistries to meet their specific needs in both polyester and polyether systems. These new technologies can be used in both low and high-density systems. We have also demonstrated the ability of Dabco 1027 and Dabco 1028 to adjust reactivity

profiles to meet the demands placed on the producers of polyurethane systems for improved quality and reduced cost. This provides the manufacturer the ability to extend cream time as well as reduce the demold time to meet their production needs. Air Products remains committed to helping our customers solve problems through the application of both existing and newly developed additive technologies. Several programs are underway at Air Products to develop new additive technologies that will allow producers to meet the demand of their customers in density reduction as well as production of more complicated footwear.

REFERENCES 1. Steve L.; Proceedings at the 99 Asia Utech, Footwear Market Overview, Paper 1. 2. Frey, J.H.; Stevens, R.E.; Sudbury-Holtschlag, J.E.; Proceedings at the 97 Asia Utech Meeting, Paper 22. 3. Mansfield, K.F.; Miller, J.W.; Wong, W.; Proceedings at the 98 Asia Utech, New Additives To Improve Dimensional Stability In Rigid Polyurethane Foam Systems. 4. Andrew, G.D.; Smith, N.K.; Optimized Processing of Microcellular Polyurethane Systems Through Catalyst and Surfactant Technology, Polyurethanes: Exploring New Horizons, Proceedings of the SPI 30th Annual Technical/Marketing Conference. 1988.

BIOGRAPHIES Gary D. Andrew Gary Andrew is a Senior Principal Application Chemist for the Performance Chemicals Technology Department of Air Products. He is responsible for providing both technical and developmental support for Asia and Latin America. Since joining Air Products in 1983 he has authored several papers and patents in the area of polyurethane foam. James D. Tobias Jim Tobias is currently a Senior Principal Applications Chemist in the Performance Chemicals Technology Group at Air Products and Chemicals, Inc., located at the Corporate Headquarters in Allentown, Pennsylvania. He received his B.S. degree from Moravian College, Bethlehem, PA in 1983. He is currently responsible for applications development and technical service for catalysts and surfactants with emphasis on supporting the flexible market segment