An Additive Approach to Expanding the Performance

ANA MARIA CANO SIERRA

CHRISTIAN SCHEFFNER

ANDREA FRUTH

BRUCE ERNST

Rhein Chemie Rheinau GmbH,

Düsseldorferstraße 23, 68219 Mannheim, Germany

Abstract

Polyester-based polyurethane adhesives are often used in flexible packaging applications. The performance range of these adhesives goes from fairly low to very demanding. In most cases varying degrees of heat and hydrolysis resistance are required. In the medium to high performance segments this demand for increased heat and or aggressive media resistance is seen in the critical zipper and lid areas as well as for chemical applications. In the highest performance segments; retortable pouches for medical applications, concentrated chemical storage and outdoor exposure, the adhesive must survive the extended heat and moisture exposures (up to 120°C for 30 minutes) of the steam sterilization / retort process.

A family of carbodiimide additives has been developed, which added to various types of polyester, polyurethane adhesives (medium to high performance) will significantly increase the heat and hydrolysis resistance in standard mechanical and analytical tests. Evidence from environmental and retort testing show both an extension of the performance range of the medium-performance and the time to failure of the high-performance adhesives.

The use of these additives will enable formulators to either use more cost-effective backbone materials or to push their high-performance products into new segments based on their expanded properties.

Introduction

Polyurethane chemistry has been established throughout the world in many coatings, adhesive and sealants applications, and its potential for further development is by no mean exhausted. New applications are continually being discovered, and the polyurethane raw material product offerings are being systematically optimized to meet growing demands.

Films for food packaging have to fulfill a very demanding range of properties. The list is extensive and many of these properties have to be met simultaneously. The films have to be impervious to light, gases aromas, moisture and steam, resistant to sterilization and aggressive contents, impermeable to fat migration, printable, heat sealable, resistant to tearing and abrasion. There is no single film material able to meet this range of requirements. Consequently, films made from different materials are combined to produce composite films. The individual film layers are bonded together using very thin layers of special laminating adhesives, which are mainly based on polyurethane raw materials.

The adhesives are generally classified in terms of performance. Laminated films for packaging are classified into three performance classes based on the property profile required for the intended use (Table 1)

Three performance classes for laminated filmsPerformance Classes Laminate / Application Adhesive Chemistry

High Performance1) All laminates (film/film & film/aluminum) for

Solventborne/Solventless adhesives Two-component systems:

a) Agressive products (e.g. fruit juice, detergents)

b) High-strength laminates with high tear strength i: -OH-functional polymer with high molecular weight

c) Thermal exposure > 100°C for sterilizability (e.g.

animal feed) and boiling resistance (microwave)

-NCO crosslinker with low molecular weight

2) Aluminum (film/aluminum) laminates for thermal

exposure < 100°C (pasteurizable)

ii: -NCO prepolymer with high molecular weight

-OH crosslinker with low molecular weight

Medium Performance1) Film/film laminates for thermal exposure of <100°C

(pasteurizable)

- 50% solventborne adhesives Two-component systems

*OH-functional prepolymer with high molecular weight

* NCO-crosslinker with low molecular weight

2) "Simple" aluminum composite for thermal exposure of

<40°C (e.g. coffee packaging)

- 50% solvent free adhesives two component systems

*NCO prepolymer with high molecular weight

*OH crosslinker with low molecular weight

General Performance1) Film/film laminates for thermal exposure of < approx.

40°C (cheese packaging) and for low mechanical stress Two-component systems. The portion of solvent systems is much lower than of solvent free systems

2) Film/paper laminates Solvent free

Two componet systems

One component systems:

NCO-functional prepolymer with high molecular weight

(moisture curing)

Table 1: Requirements for laminated films in different performance classes

The adhesive are generally classified in terms of performance.

General performance (single or two component solvent based, solventless or water based)

Typical uses: for snack food and confectionery, bottle labels, bakery

Medium performance (two component solvent based, solventless or water based)

Typical uses: for coffee pouches, liquid soaps, cosmetics, sauces and condiments, laminates with moderate thermal treatments.

Requirements : Product resistance, high heat resistance for zipper installation

High performance: (solvent based or solventless) for retort applications as pouches, lidding, hot fill/boil-in-bag, concentrated chemical storage, outdoor exposure /agricultural bags, etc.

Requirements : Product resistance, high heat and or chemical

Industrial lamination: sail cloth, insulation laminates, cable wrap, metal laminates

Depending on requirements, Polyurethane laminating adhesives can be two-component or moisture curing one-component reactive systems, which for medium and high requirements, often also contain a solvent. Laminating adhesives can be classified by application type in solventborne, solvent less (100% solids), waterborne, radiation curable (100% solids), combination radiation curable.

The popularity of solvent free systems has been growing for years. Nevertheless, solventborne systems are still strongly represented in the manufacture of laminated films for medium and maximum requirements. The reasons for this must be understood.

•Solventborne systems can be applied in thicker layers. At layer thicknesses of >3 g/m2, solvent free systems often have poorer optical properties (orange peel).

•Prepolymers with high molecular weight can be used in solventborne adhesives without causing viscosity problems.

•The solvent also allows the use of higher portions of high viscosity aromatic polyester diols which lend greater strength to the crosslinking adhesive film.

•Also facilitate the use of slow reacting aliphatic isocyanates.

•Any viscosity problems can be corrected by the choice and amount of solvent used.

Solvent-based adhesives cover the widest range of applications, from general purpose to high performance. Their main advantages are strong initial tack and final bond strength with good flexibility, and good optical properties.Due to the large number of different adhesives formulations and because the formulations available on the market represent the manufacturers proprietary formulators viewed across all systems and performance categories, laminating adhesives generally consist of approx. 30% polyether polyols, approx. 40% polyester polyols, and approx. 30 polyisocyanates moisture cure.

Technical selection criteria are product properties such as:

•Viscosity

•Compatibility of the polyol when mixed with other polyols

•Compatibility with the additives in the films

•Strength and elasticity of the crosslinked polymer

•Specific adhesion characteristics and resistance to hydrolysis

In the following table is shown a properties comparison between polyester and polyether based polyurethanes:

Table 2: Properties of polyester and polyether based polyurethanes

Adhesives based on highly crystalline polyesters yield durable bonds. Products containing less crystalline polyesters are characterized by their higher flexibility and better adhesion to non-polar substrates. Therefore, they are used for film lamination for manufacturing abrasive belts or for bonding polyurethane-based flexible foams.

But it is well known that a general weakness of polyesters is their poor resistance to the attack by moisture, particularly at elevated temperatures. The degradation or breakdown of polyester by water and acids is known as hydrolysis. The functional groups present in the chains are hydrolyzed resulting in both chain breaking and loss of crosslinking.

Deterioration may occur more quickly in a 95% RH environment than in liquid water because of more rapid permeation of the vapor. The hydrolysis increases with higher temperatures. This effect is usually much faster in flexible materials because water permeates it more easily.

Hydrolysis is catalyzed by acids. Acidic water solutions accelerate hydrolytic attack and, therefore, solutions of salts or acids are likely to degrade the polymer faster than pure water. The ester linkage of a polymer is cleaved by the action of water to produce carboxylic acid and alcohol. Once initiated, this process is autocatalytically accelerated and, in the absence of hydrolysis stabilizer, results in complete breakdown.

The reduction of the molecular weight caused by the hydrolytic degradation results in the irreversible loss of tensile strength, modulus and hardness. The decrease of the bond strength results in cohesive failure of the bond. However, before this occurs the adhesive usually swells and may cause deformation or bond failure before hydrolysis can completely take action.

Figure 1: Displacement of adhesives from the surface by other chemicals

Water can also permeate the adhesive and migrate to the interfacial region displacing the adhesive material at the interface. It is the most common cause of adhesive strength reduction in moist environments.

In flexible packaging is assumed that in the beginning of the aging in moist, high-temperature environments the mode of failure is truly cohesive. After about one week the failure will be caused by loss of adhesion. It is expected that water vapor permeates the adhesives through its exposed edges.

To improve the moisture and hydrolysis resistance the formulator generally has to choose a suitable base adhesive for the application to obtain the optimal performance. First step is to look for a base polymer that has a low diffusion coefficient and permeability to water. This reduces the rate of diffusion of moisture to the critical interphase between the substrate and the adhesive and also reduces the negative effect on the adhesive itself. Those polymers with moisture resistance are the rigid and crosslinked types. Unfortunately, they form brittle adhesives with poor peel and impact strengths. Therefore flexible packaging adhesives are formulated to provide the best compromise between environmental resistance and the desired mechanical properties.

Special additives, such as carbodiimides are recognized to improve the hydrolytic stability of many polymers like PUR (esterpolyol based elastomers, flexible foams, etc…), PET, PBT, TPU, TPE, PA and EVA.

The service life of a hydrolysis sensitive polymer containing a hydrolysis stabilizer such as a carbodiimide can be prolonged up to three times longer than unprotected products.

Chemical structures of industrial used carbodiimides have the following basic structure:

The common reactions of carbodiimides occur as addition reactions. The R groups attached to the carbodiimide functionality determine selectively what type of addition reaction occurs and under what kinetically driven conditions.

Carbodiimides act as acid and water scavengers. The reaction with acids is highly preferred. Through these reactions they convert and neutralize both water and acids into non-hazardous urea structures.

Besides the important reactions with water and acids there are generally a lot of other reactions with low molecular weight organic and inorganic moieties possible. Examples are amines, isocyanates, etc…

The objective of this paper is to discuss about a family of carbodiimide additives that has been developed, which added to various types of polyester, polyurethane adhesives (general medium to high performance) will significantly increase the heat and hydrolysis resistance in standard mechanical and analytical tests showing an extension of the performance range of the medium-performance and the time to failure of the high-performance adhesives. The use of these additives will enable formulators to either use more cost-effective backbone materials or to push their high-performance products into new segments based on their expanded properties.

Carbodiimide Additives

Carbodiimide Additives

Materials

To show the performance of this family of carbodiimides (CDIs), a model solvent based two component formulation was used. Raw materials selected:

- A linear hydroxyl polyurethane prepolymer, Desmocoll® 140 from Bayer Material Science AG.

- Light branched Polyesterpolyol, Baycoll AS 2060 from Bayer Material Science AG.

- A solution of Tris(p-isocyantephenyl)thiophosphate in ethylacetate, Desmodur RFE from Bayer Material Science AG.

- CDI 1, oligomeric aromatic carbodiimide, NCN-content minimum12,5%

- CDI 2, monomeric aromatic carbodiimide, NCN-content minimum 10%

- CDI 3, low fogging monomeric aromatic carbodiimide, with low volatile organic compounds content (VOC), NCN-content minimum 10%

- CDI 4, oligomeric carbodiimide, NCN-content = 6-7,5%

CDIs A-D were added in a molar ratio, ensuring that examples 1-4 contains the same N=C=N content.

The next table shows the different ratios added in phr.

Raw material/Formulation S0 S1 S2 S3 S4

Desmocoll® 140 14 14 14 14 14

Baycoll® AS 2060 7 7 7 7 7

Desmodur® RFE 4 4 4 4 4

Ethylacetate 75 75 75 75 75 Carbodiimide A x 0,25 x x x Carbodiimide B x x 0,32 x x Carbodiimide C x x x 0,32 x Carbodiimide D x x x x 0,48

Table 3: Model solvent based two component adhesives formulations

The adhesive samples were manufactured as follows:

Variant 1: 14 phr Desmocoll® 140 was dissolved at 85°C in 75 phr water-free ethyl acetate. After cooling of the solution at room temperature the remaining components mentioned in the table were added.

Variant 2: 14 phr Desmocoll® 140 was dissolved at 85°C in 75 phr water-free ethyl acetate. After cooling the solution at room temperature the remaining components except Desmodur® RFE were added. Afterwards the solution was heated during the period of 5 hours at 80°C and afterwards Desmodur® RFE was added.

Production of sample test specimens (test panels)

Samples S0 and S1-S4, were manufactured according to the variants 1 and/or 2 formulations.

For the production of the specimens a commercial PET film, 23µm thickness, with aluminum foil of 100 µm thickness were used. Both substrates were manufactured by Goodfellow GmbH. The foils were not pre-treated.

The above mentioned adhesive samples S0 and S1 - S4, were produced in accordance with the formula of variants 1 and/or 2, the adhesive was applied on the aluminum foils in a layer of 50µm thickness, after 10 minutes of solvent flash-off, the PET foil was applied to the respective adhesive layer and hardened at 40°C for one hour with a contact pressure of 3 kilograms.

The specimen size for the PET/aluminum foils was 20 x 30 centimeters.

The long-term stability of the adhesive layer in at least 2 layers (PET/aluminum foils) was demonstrated with two tests:

1.Pressure cooker test (PCT) - The effects of sterilization were measured in this test. The specimens were

aged 20 minutes at 121°C and 2 bars pressure. A role peeling test was performed before and after aging.

2.Aging test (climate test) at 85°C and 85% relative humidity - Role peeling test before and after aging. In

the role peeling test the peeling resistance of an adhesive in N/cm or N/mm was measured. In this test the end of the PET aluminum foils was bent around 90°, so that the sample looked like a T-fitting. The free ends were clamped in universal test equipment and pulled apart.

Results

I.Determination of the curing temperature from model formulation through DSC (Differential

scanning calorimetry).

The curing temperature was determined according to the DSC analysis. (See Figures 3, 4, 5).

Parallel to the crosslinking reaction, which can be detected in the DSC as an exothermic peak at 40°C, the adhesion was also detected by a conversion of the crystalline regions of the adhesive into amorphous regions; it is possible to observe this conversion in the DSC as an endothermic peak. Figure 4 shows these endothermic thermal processes both with the formulation S0 and with the formulation S2 within the range of 28 °C. A crosslinking reaction could not be detected here; however the exothermic peak of reaction of the crosslinking is quite small and could have been overlapped.Maximal reaction temperature: 41°C

Figure 3: Differential scanning calorimetry (DSC) analysis from reference formulation SO

Special carbodiimides are also well known as efficient crosslinking agents in water based systems, used instead of isocyanates. In our formulation isocyanates were used as crosslinkers, carbodiimides were used as hydrolysis resistant agents.

In order to investigate the possible post-crosslinking reactions that could occur with carbodiimides, an isotherm DSC was performed. The isotherm DSC at 40 °C showed that no further reaction enthalpies can be recognized after 40 minutes. As expected crosslinking is only achieved by isocyanates. No undesired post-crosslinking with carbodiimides was observed. See Figure 5 (formulation S0 and S2). The different heights of the base lines are instrumentation conditioned.

Comparison DSC Adhesives formulations 0 and 2

Figure 4: Differential scanning calorimetry (DSC) analysis comparison between adhesives formulation S0 (bottom line) and S2 (upper line)

Figure 5: Isotherm DSC analysis at 40°C from reference formulation SO and formulation S2

A more sensitive method to follow the reaction in this case was infrared spectroscopy. It was used to evaluate the reaction kinetics between the different components. The reaction was followed with IR spectroscopy. IR spectrums showed that in the reaction of a mixture of Desmodur RFE (isocyanate hardener) and the polyol (Baycoll) at a hardening temperature of 40 °C after several days, there were still small quantities of unreacted hardeners; see Graphic 1 (left). Due to these results, two test rows were performed for the stability analyses (hydrolytic stability): In one row the specimens were tempered at 40°C/60h after curing at 40°C/1h and in the other row the specimens were not tempered after curing at 40°C/1h.

Graphic 1: Determination of parallel reaction as crosslinking, competition with the isocyanate.

The investigations of the mixtures Desmocoll and Desmodur RFE and CDI 2 and Desmodur RFE are represented in Graphic 1 (right). The curves do not show a constant decrease, like with the Baycoll, but a nearly periodic increase and decrease. An explanation for this could not be found without further investigations. However a small amount of isocyanate was found. This indicated that a competitive reaction between carbodiimide and isocyanate can be excluded in the reaction from Desmodur RFE and Baycoll.

No crosslinking reaction of the carbodiimide has been identified. The positive effect was that the newly developed CDI had no effect on the viscosity of the adhesive. The viscosity remained constant.The ageing tests were performed in three test rows:

•All components (except the hardener Desmodur RFE) were heated at 85 °C/5h, in order to ensure that the carbodiimide reacted with the polyester acid groups.

•The second test series was tempered after curing during an additional sixty hours at 40 °C (following hardening recommendations for the commercial systems and the results of the IR measurements) •And a third one, without tempering after curing.

The commercial formulations were prepared according to the manufacturer’s technical data sheets. The system 1 could be used in two different mixing proportions, both were examined. The tests specimens were manufactured according to the model formulation.

II.Ageing tests

A. Model formulation

A1. Pressure Cooker test

In the Pressure Cooker test (PCT) the materials performance during sterilization was tested. The samples were aged 20 minutes at 121 °C and 2 bar pressure. A rolling peeling test of the specimens (strips) was measured before and after sterilization. See Graphic 2.

Graphic 2: Pressure Cooker test (PCT) from model formulations with and without CDIs.

The initial values comparing the tempered and non-tempered adhesives systems showed a post crosslinking in the adhesive system. This was also seen in the IR results. This post-crosslinking / tensile strength increase is not unusual and can often be detected in adhesive systems.

A decrease of the role peeling strength was not seen in the Pressure Cooker test. The formulations S2 and S3 with monomeric CDIs, showed a higher role peeling strength than the formulations S1 and S4 with the oligomeric CDIs.

b. In case of the non-tempered samples, the following observations were made:

• The reference formulation S0 had the lowest strength. After 168 hours aging, the peeling strength rose

but after 500 hours it showed a low residual value of peeling strength.

• It was also possible to observe this effect with the other formulations containing CDIs. This result

indicated a post-crosslinking reaction of the adhesive.

• S4 lost the majority of its peeling strength upon further aging (to 1000 hours), while the formulations

S2 and S1 experienced a further rise of the peeling strength.

c. The 5 hours at 85°C pre-reacted formulations can be compared with the reference S0 non-tempered, since they did not experience a further tempering during the adhesion process:

• As in the case of the non-tempered samples, the peeling strength of the 5hours pre-reacted

formulations rose during the first 168 hours of aging. After 500 hours the model formulations modified with CDIs showed a clearly higher peeling strength than the reference formulation S0, whereby monomerics CDIs in formulations S2 and S3 showed higher peeling strength than oligomeric ones

• After 1000 hours aging all CDI formulations showed peeling strength of 0,1 N/mm, which

corresponded to a middle level. The reference system S0 does not show any peeling strength after 1000 h.

All CDI Types showed clearly positive effects in the peeling test, influenced by the adhesive formulation process, curing temperature and time.

Monomeric CDI types showed longer hydrolytic stability and higher mechanical properties up to 500h compared to the Polymeric CDI types, with a little lower mechanical strength. However, over a longer time period, up to 1000h, they showed better hydrolytic stability.

To show the influence of temperature before curing and tempering after curing, the results from reference formulation S0 and S2 (formulation containing monomeric CDI) are represented in the following graphic.

Graphic 7: Comparison peeling strength under the influence of temperature before curing and tempering after curing, results from reference formulation S0 and formulation S2

tempered P e e l i n g S t r e n g t h (N /m m )

Time (h)

tempered 5h

B. Commercial formulations

>Climate aging test at 85 °C/85% relative humidity of the commercial formulations

The results of the commercial formulations are represented in the Graphic 8

Graphic 8: Aging test (85°C / 85% relative humidity) comparison between commercial formulations with and without CDIs

The commercial formulations modified with the four CDI types showed higher peeling strength and a longer, up to threefold increase in its service life, than the reference system. CDIs 3 and 2, monomeric ones, were the most effective in regards to peeling strength; however the oligomeric-types showed a longer protective effect against aging.

Conclusions

Depending on the hardening time and temperature, different peeling strengths are reached by both model and commercial formulations. This is a well-characterized property of these adhesives. The different strengths are based essentially on differently formed polymers networks. Within the range of structural adhesives this characteristic is still more highly pronounced than in the formulations tested here. An extrapolation to other types of adhesive formulations is possible but must still be experimentally verified.

From the kinetics investigations it was shown that the crosslinking reaction takes place between the isocyanate and polyol. The investigations with the added CDI-types did not indicate any further reactions, such as crosslinking, only the desired stabilization reaction produced by the reaction between polyol acid groups and PU-components with the CDI-types.

The model formulations showed that after the Pressure Cooker test there was an increased peeling strength. This suggests a post-crosslinking and re-orientation of the formed polymer net. In general, it was shown that formulations stabilized with CDIs provide higher values of peeling strength than unstabilized ones. The monomeric types are in general better than the polymeric types.

The Pressure Cooker test did not show clearly improved resistance to hydrolysis, also the initial values were not decreased. However, no negative influences were determined. For this reason a more aggressive aging test was selected to show the performance of CDIs.These statements were also confirmed in the CDI modified commercial formulations in the comparison to the not modified commercial reference system.

The results from the Climate aging test at 85°C/ 85% r.h. show a positive influence by the CDI types on the resistance to hydrolysis, both in the model and in the commercial formulations. All CDI Types showed clearly positive effects in the peeling test, influenced by the adhesive formulation process, curing temperature and time.

Monomeric CDI types showed longer hydrolytic stability and higher mechanical properties up to 500 hours compared to the Polymeric CDI types, with a little lower mechanical strength but over a longer time period, up to 1000hours, showed better hydrolytic stability.

The objective of this paper has been achieved. It has been demonstrated that stabilizing an adhesive with the family of carbodiimide additives that has been developed, brings about an up to threefold increase in its service life, showing an extension of the performance range of the medium-performance and the time to failure of the high-performance adhesives.

When carbodimides react with the cleavage products carboxylic acids and water, it gives rise to urea compounds which have no negative impact on the stabilized material.

Using this family of carbodiimides improves the cost/benefit ratio in well established applications and will enable formulators to either use more cost-effective backbone materials or to push their high-performance products into new segments based on their expanded properties. Considerable raw material-related cost advantages may also be achieved in applications which are normally only open to substantially higher cost materials.

References

/1/ Market forces put pressure on lamination adhesives manufactures, Flexo & Gravure Int'l 1- 2006, 56-58

/2/ Marktübersicht Verpackungsklebstoffe, Adhäsion 9 04, 30-36

/3/ Formulierung von Kleb- und Dichtstoffen, Bodo Müller / Walter Rath, Vincentz Verlag, 2004, 48-51