CFC-free polyether polyurethane systems for footwear applications.

There has been pressure on the polyurethanes industry to find alternative methods to blow foam. In those areas where the thermal conductivity of the blowing agent is not relevant the replacement is viewed as essential. Consequently considerable effort has been expended by chemical companies and by the end users, in order to eliminate CFCs from footwear applications. This article will discuss some of the work being done by ICI to remove CFCs from polyether polyurethane footwear systems.

The majority of MDI based polyurethane systems currently being used for footwear fall into the two broad categories of polyester and polyether systems. Each has a particular set of advantages and disadvantages and so together they offer a wide selection base to meet customer requirements. Table 1 lists some of the main advantages and disadvantages of each family of products. It can be seen that in general polyethers are easier to handle than polyesters, but that polyesters have superior mechanical properties.

The reasons why there are differences in properties between polyether and polyester based polyurethane systems are not fully understood, but there are a number of possible explanations. It is not the purpose of this article, however, to go into these in detail and so only one of the most well accepted explanations will be briefly discussed. The discussion will also show why CFCs are currently required and proceed to list two possible ways of eliminating them.

During the production of elastomeric polyurethane a number of simultaneous chemical reactions are occurring. Further, during these major chemical changes a number of

  Table 1 - polyether and polyester based PU
soling systems - advantages and disadvantages
Polyethers
Advantages
Liquid components
Good blend stability
Good hydrolytic stability
Good resistance to
 microbial attack
Good low temperature
 properties
Good damping properties
Disadvantages
Cooling units on component
 tanks (depending on
 climatic conditions)
Single density molding
Mold fouling
Poor abrasion resistance
Poor tear strength
Poor tensile strength
Polyesters
Dual density molding
   possible
Short demold times
Self adhesion
Good abrasion resistance
Good tear strength
Good tensile strength
Heated tanks
Melt out of components
    often needed

physical processes are also occurring. One such physical process is that of phase separation into the hard and soft block domains. The phase separation can begin at different times during the reaction and proceed at different rates.

When and how fast, depends upon reaction rate, rate of molecular weight build-up, viscosity, temperature and compatibility between hard and soft segment phases. All of these processes are in a non-equilibrium state and therefore it is difficult to generalize the situation. Indeed even simple specific systems can only be described by complicated mathematics.

However, it has been shown experimentally (ref. 1) that 4,4 MDI based hard segments in polyester polyurethane systems are relatively compatible with the soft polyester segments in the initial stages of reaction. As the reaction progresses a slow phase separation occurs. This allows the formation of relatively small, well distributed and crystalline hard blocks. In polyether systems, however, the hard segments are relatively incompatible with the soft segments even at the early stages of reaction. This causes the rapid formation of relatively large and amorphous hard blocks.

The resultant reinforcing effects in the two systems are different. A large number of small, well distributed hard blocks reinforce the polymer structure more efficiently than a few larger ones (comparing systems at the same overall hardblock content).

The shortcomings in the mechanical properties of polyethers have traditionally been compensated for the use of a high density surface layer, or integral skin layer. This integral skin is a natural consequence of the means of production.

Physical blowing agents, usually CFCs such as CFC-11, are added to the polyol blend. This blend is then mixed with the isocyanate bearing component to effect reaction and produce polyurethanes. During this exothermic polyurethane reaction, the heat generated is absorbed by the CFCs so vaporizing them and producing a foam. The temperature generated is particularly high in the bulk of the moldings and so low density foam results. Adjacent to the mold wall, however, two features result in reduced or even no blowing at all. First, the heat of reaction is conducted away from the reacting material into the mold so significantly less blowing results. Hence a low density core with high density skin (typically 2-5 mm thick) results. Secondly, in conjunction with this heat extraction, the pressure generated adjacent to the mold surface causes a condensation of the blowing vapors (CFC), resulting in a higher density boundary layer.

Since more, if not all, properties are related to density as shown in figure 1, this higher density boundary zone of polyether systems results in acceptable performance properties such as tear strength and flex life. The figure further shows that while CFC blown polyethers can be processed at a slightly lower density, A, than polyether system, B, the resulting mechanical properties at these extreme densities, A and B, are inadequate. A certain minimum tear strength, as indicated in figure 1, is needed by the industry. Consequently higher densities than A and B have to be used. For

  Table 2 - solubility of blowing agent N
Solvent               Temperature   Amount
Monoethylene glycol    23[degrees]C  =25%
1,4 Butandiol          23[degrees]C  <2%
Daltocel PA38 (*)     23[degrees]C   <2%
                       50[degrees]C  >10%
Diethylene glycol      23[degrees]C  <10%

(*) Daltocel PA38is a commercially available polyoxyalkylene alcohol of [OH.sub.v] 33-38 mg KOH/g

  Table 3 - processing behavior of blowing
 agent N containing polyether polyurethane
                  systems
                        Reaction profile
                   Cream   Gel    Demold   Free rise
                    time   time    time     density
Blowing agent         (s)   (s)      (s)    ([gcm.sup.-3])
CFC                   10    15     220       0.317
[H.sub.2O] (0.3%)     13    16     220       0.365
  (fresh)
[H.sub.2O] (0.3%)     13    16     220       0.365
  (10 days)
Blowing agent N       14    16     220       0.420
  (eq. 0.3% [H.sub.2O]
   (fresh)
Blowing agent N       14    16     220       0.460
  (eq. 0.3% [H.sub.2O]
  (10 days)

In all cases Daltoped 411 RNNN -VMO 21 were used. Daltoped 411 RNN is a conventional polyether formulation based on Daltocel PA 38 and MEG. VM 021 is a conventional prepolymer for polyether polyurethane systems; [NCO.sub.v] = 23%

polyesters the minimum molding density to give acceptable properties is P.

However, the integral skin effect allows molding of polyether based systems at say X (overall density), the skin density then being Y and the core being Z.

Replacement of CFCs by water in polyether systems is, however, not a simple matter and in fact results in three undesirable effects:

* The integral skin is not produced. With carbon dioxide blowing, the pressures generated are not sufficient to condense the gas in the described manner, and so the thick skin layer is not formed.

* Polyether systems with direct replacement of the CFC by water do not show good dimensional stability. Hence very long demold times are required to avoid afterblow and even then, post demold shrinkage remains a problem.

* The viscosity of the water containing polyether polyol blend is higher than the equivalent CFC containing blends. This can make mixing less efficient.

The existing CFC blown polyether systems have been modified to minimize the deleterious effects of changing from CFC to water blowing. Further two new separate approaches have been followed, both of which have been successful. The description of these forms the basis of the article.

First, the approach was taken to maintain an integral skin but generate it by use of an alternative blowing agent. These may be physical or chemical blowing agents which are thermally activated causing blowing in thecore of the molding but not adjacent to the cooler mold surface.

Second, to improve the properties of polyether based systems, shifting the property-density profile up (figure 1) to ultimately coincide with that for polyester based systems. One possible way to achieve this is to modify the compatibility of the reacting components and hence influence the rate and degree of phase separation and final polymer morphology. One approach in each area will be discussed.

Experimental concepts

Alternative blowing agents

This article will limit itself to the family of blowing agents which produce chemical blowing. The range of alternative physical blowing agents, HCFCs etc., which are bing increasingly used, especially in rigid foam applications, will not be discussed. Rather, the family which behave in the following way will be discussed:

A ------ B + XOH (1) R - NCO + XOH ----- [CO.sub.2] + R - NHX (2)

The interior of the reacting mixture generates heat. In this case, however, the heat cleaves or decomposes product A. The thermal degradation products of A include either an isocynate reactive species, which reacts to give [CO.sub.2] (or another gas), or, is a gaseous material itself. Material adjacent to the mold wall does not generate enough retain heat to affect thermal decomposition of A and so no, or very little, blowing occurs. This creates the required integral skin.

Improved properties

Increasing the initial compatibility of components has been

Table 4 - summary of physical properties of
alternatively blown polyether polyurethane
                             systems
                                    Blowing agent
                                                    Blowing
Property                        CFC       [H.sub.2.O]  agent N
Overall density ([gcm.sup.-3)   0.580     0.577        0.868
Skin density ([gcm.sup.-3]      0.862     0.657        0.868
Hardness (Sh.A)                 75        65           75
Tensile strength                4.5       4.5          6.0
  @ break (MPa)
Elongation @ break (%)          320       280          300
Tear strength ([Nm.sub.-1]      10.10     10.27        15.11
Flex life (% cut growth) (*)    200       22,000 (**)  150

In all cases Daltoped 411 RNN - VM 021 were used.

(*) After 30,000 cycles (Deggen Flex)

(**) Cycles to break

     Table 5 - characteristics of PBA 2393
                 (typical results)
Appearance                          : clear to slightly yellow
                                    liquid
Isocyanate component                : MDI
NCOv                                = 19%
Processing temperature              = 25-35 [deg] C
Viscosity @ 25 [deg] C              = 1350 mPas
Freezing point                      = <15 [deg] C
Heat stability (80 [deg] C/24 hrs)
  NCOv drop                         = 0.05% max.
  vixcosity increase (@ 25 [deg] C) = 115 mPas

achieved by the development of a new prepolymer technology. This has provided an improvement of the physical performance of polyether based polyurethane footwear systems, and so allows the full replacement of CFCs by other typical blowing agents used in the polyurethane industry, such as water. Since physical properties have improved, the integral skin formation is no longer required and so the typical water blown type foam boundary formation which results is sufficient.

Discussion

Alternative blowing agents

Blowing agent N is a product which, under certain conditions, will thermally decompose to generate water:

N -- N' + [H.sub.2.O] (90-130[degrees]C)

Disclosure of the composition of blowing agent X is not necessary for the purposes of this article, but, it should be noted that it is a solid, soluble in low molecular weight alcohols and diols at room temperatures and in polyethers at moderately high temperatures of ca. 50[degrees]C (see table 2).

The product may simply be added as a solid and disproved into the polyol blend. The handling conditions and techniques are the same required for conventional polyurethane systems.

Further, the polyol blend containing blowing agent N, unlike those containing CFCs, is easy to handle - additional cooling units on the component tanks, or special extraction units, other than those used for conventional polyurethance systems, are unnecessary.

Table 3 lists a typical performance profile of a system containing blowing agent N in terms of the processability of the system. The table shows a comparison of essentially the same system but with three different blowing agents.

The first, blown with CFC, is the reference system. The second replaces the CFC by water, in this case 0.3 pbw on polyol. Finally, blowing agent N is used in the third system. The level used is that which, if fully thermally decomposed, would produce the equivalent of 0.3 pbw water.

As can be seen, the free rise density of this system is slightly higher than that with the water blown system. However, since blowing agent N is not decomposing at the surface of the foam, the product has a free rise skin density of around 0.75 [gcm.sup.-3] while the water blown system does not have a skin at all.

Table 3 also shows the blend stability of a blowing agent N containing system. After 10 days storage of the full blend at room temperature the reaction profile is unchanged. This is of course what would normally be expected for a polyether system.

Finally, with regard to the start of reaction, the blowing agent N containing system is slightly slower for the same demold time (in this case, the time taken for no crack formation

 Table 6 - performance of PBA 2393 compared to
 conventional polyether based polyurethane unit
 sole systems
                                      System
                             Conventional   Unit sole
Property                      CFC-ether     PBA 2393
Cream time (s)                   7             6
Gel time (s)                    18            13
Tack free time (s)              20            17
End-of-rise (s)                 32            30
Pinch time (s)                  65            45
Free rise density ([gcm.sup.-3]) 0.220      0.267
Molding density ([gcm.sup.-3]) 0.540        0.560
Minimum demold time (s)        260           280
Table 7 - physical property performance of PBA
  2393 compared to conventional polyether
   based polyurethane unit sole systems
                                      System
                                 Conventional
                                  CFC-ether     Unit sole
Property                            (unit)      PBA 2393
Density ([gcm.sup.-3])              0.540         0.560
Hardness (SH[degrees]A)              72             70
Tensile strength (MPa)              4.52           4.90
Elongation (%)                      320            395
Tear strength ([Nm.sup.-1])         10.1           12.0
Abrasion (mg)                       320            255
Flex life (% cut growth) (*)         28            32
Ball rebound (%)         28                           32

(*) After 30,000 cycles (Deggen Flex)

on demolding). While this effect has not yet been fully evaluated, it potentially implies a longer screw fouling time and even better flow behavior than standard blown systems.

Physical properties of the derived foams of systems containing blowing agent N are shown in table 4. As is immediately obvious, the properties of the blown system are at least as good as those for the CFC blown. The properties of the water blown systems shown here are not particularly bad. However, flex life of the three systems does reveal a significant difference. The water blown system shows reasonable but inferior behavior.

Further, there is the problem of dimensional stability-afterblow and/or shrinkage. The CFC blown and N blown systems are dimensionally stable, showing no afterblow after demolding and less than 0.5% shrinkage after 24 hours. The water blown system shows excessive afterblow unless excessively long demold times (more than six minutes) are employed. If shorter demold times are used, for example, as quoted in table 3, the water blown systems show upward of 4% volume increase due to afterblow.

Improved properties

A new polyether prepolymer technology has been developed which provides improvement of physical properties.

Compatibility of hard segments and soft segments of polyether based polyurethane systems has been modified through this new prepolymer technology. Such changes are also achievable in other ways such as higher processing temperatures, but such alternative methods generally lead to impractical processing procedures or difficult component handling procedures.

This product technology concept essentially combines the benefits of the traditional prepolymer technologies of polyester based and polyether based polyurethane footwear

[TABULAR DATA OMITTED]

systems. The work reported below limits results to one such prepolymer, PBA 2393, which shows typical behavior for these types of products.

Table 5 lists the characteristics typical of this prepolymer. For those familiar with conventional polyether prepolymers, it is obvious that these characteristics are not too dissimilar to those of conventional products. The viscosity may be seen as being slightly higher than conventional products, such as Suprasec VM 022, but it is still comparable to others such as Suprasec VM 021.

The prepolymer may be processed with CFC free, but otherwise conventional, polyether based polyol blends. Processing is possible on all types of conventional processing machines - open pour or closed mold injection, as units or direct on.

Tables 6 and 7 show typical processing and physical property results respectively for this new type of polyether system (unit sole application) processed on a direct injection machine, Desma PSA 91. The reaction profile of the PBA 2393 based system is slightly higher than for the conventional system, yet the demold time is slightly longer. This aspect of behavior is currently being addressed in the ICI laboratories in Everberg. Formulation modifications may be able to overcome some of the problems, but optimization of the prepolymer is currently ongoing to avoid performance compromises. The properties, shown in table 7, are on the whole superior to the conventional system.

Processing and physical properties of the derived foams (unit systems) processed on a casting machine, Desma DS 20-20, are shown in table 8.

Demold times of the PBA 2393 based system are longer than the conventional polyether system. Again, as above, this problem is being addressed. Regarding the physical properties, as for the direct-on processing, those systems based on PBA 2393 (A and B) are at least as good as the conventional system.

Again, not all the problems of simple water replacement of CFCs are apparent from the figures. These problems being, as mentioned above, dimensional stability which manifests itself as afterblow and shrinkage. It is for this reason that the conventional but water blown system (table 8) has such a long demold time.

In both cases, it can be seen that the new range of polyether systems provide performances comparable to the conventional CFC blown polyether systems. Reaction profiles other than the slightly longer demold times are essentially unchanged. This combined with the safety aspect of the materials used suggests that sophisticated machine adaptations are unnecessary (e.g., as for the case of pentane blown systems).

Finally tables 9 and 10 show the results for mid and outsole systems processed with PBA 2393. These can be processed in situ to give dual density moldings, which is not normally possible with conventional polyether systems. The reasons for this are:

* Conventional "polyether" prepolymers fall into the generic group of hard blocK prepolymers. Such prepolymers have reaction profiles and cure characteristics unfavorable to dual density molding. For example, mold opening times of conventional nonblown polyether (outside) systems are too long for viable dual density production procedures.

* The CFCs in conventional polyether (midsole) systems provide a high density and relatively thick boundary zone. This again is unfavorable for the adhesion at the PU-PU interface.

Table 9 shows the processing characteristics of the two types of soling material. The values achieved show that mold filling is not problematic. Further the mold opening time for the outsole system is good even compared to polyester outsole systems which have mold opening times of 60-90 seconds.

Table 10 shows the properties of the midsole, the outsole

Table 9 - performance processing characteristics
  of PBA 2393 in dual density molding
                                     Type of sole
                             (based on PBA 2393)
Property                       Midsole       Outsole
Cream time (s)                    6      not measured
Gel time (s)                      12          17
Tack free time (s)                17          26
End-of-rise (s)                   30          NA
Pinch time (s)                    63          NA
Free rise density [(gcm.sup.-3.)]0.220        0.755
Molding density [(gcm.sup.-3.)]  0.440        1.000
Minimum mold opening
  time (s)                        NA            75
Minimum demold time (s)           350         NA
Table 10 - physical property performance of PBA
2393 in dual density polyether based PU systems
                                             Type of system
                                         Polyether     Polyester
Property                            Midsole     Outsole  Outsole
Density [(gcm.sup.-3)]              0.440        1.000     1.000
Hardness (SH[degree]A)                53           72        65
Tensile strength (MPa)               3.70         10.8      16.9
Elongation (%)                       385          750       550
Tear strength [(Nm.sup.-1.)]         7.0          25.0      20.2
Abrassion (mg)                       NA           221       120
Flex life (% cut growth(*))           0           NA        NA
Ball rebound (%)                     23           NA        NA
                                      Combisole
Adhesion strength ([Nm.sup.-1.])   7.2 (100% failure in midsole)
Flex life (%) cut growth)(*)        0
  (*) After 30,000 cycles (Deggen Flex)

and the combisole and compares the outsole performance to that of a polyester based system. The tear strength of the polyether based outsole system is good and comparable even to the best polyester systems, which can be up to 30 [Nm.sup.-1].

As a last point, the results for the combisole show that adhession between the two layers is superior to the cohesive strength of the midsole, as 100% failure occurred in this layer. This illustrates that the normal problems associated with dual density polyether moldings have been overcome

Since the new prepolymer employs alternative technology to the said hard block type prepolymer and the system is waterblown, and therefore has no thick skin formation, dual density molding is now readily achieveable.

Conclusions

It has been shown that two new families of polyether based polyurethane systems for shoe soling applications have been developed, which both eliminate the need of CFCs. The two families provide either an integral skin type molding as is achieved with CFCs, or have improved innate mechanical properties (and dimensional stabilities) to such levels that they do not require the integral skin and which overcome problems such as afterblow. Further, this latter family allows the possibility of dual density molding, not previously possible with conventional polyether systems.

Finally, the described effects are possible without the need of any machine modifications and with the handling conditions normally used for MDI based polyurethane systems.

Reference

[1] Van Bogart, J.W.C., A. Lilaonitkul and S. L. Cooper. 1979. "Morphology and properties of segmented copolymers," in Multiphase Polymers, S.L. Cooper and G.M Estes, eds.; Adv. Chem. Sci., 176, USA.: ACS.