Liquid polyurethane adhesives have been used for laminating flexible films, foils and paper in the flexible packaging (FlexPack) industry for over 30 years. This class of adhesives includes waterborne, solventborne and solventless adhesives, all of which are generally formulated from combinations of polyols and isocyanates.

In two-component adhesive systems, it is often imperative that the individual materials of the polyol side remain soluble to avoid separation after mixing and prior to use. Such solubility requirements may restrict the formulation latitude to components that are either soluble in one another or contain a cosolvent to maintain compatibility. While polyether and polyester polyols may each possess a unique set of benefits for an adhesive system, they are often insoluble in one another, creating barriers to simultaneously exploiting the advantages of both types of polyols. A workhorse, or “base,” polyol, compatible with either polyether or polyester polyols, would therefore offer the development chemist a broader formulating latitude to meet the diverse demands of various FlexPack markets.

Diethylene glycol phthalates, esters of phthalic anhydride and diethylene glycol, have demonstrated unique bonding advantages to various types of substrates.^1-3 In addition to their utility in adhesives, we have shown their superior resistance to degradation by water (or hydrolytic attack) in acidic conditions.⁴ However, these esters are incompatible with polypropylene glycols, a common trend with polyester polyols in general. Further, the relationship between the viscosity of diethylene glycol phthalates and molecular weight is approximately exponential, with a 1,000 molecular weight diethylene glycol-phthalate having a 25°C dynamic viscosity of approximately 200,000 cP and a 2,000 MW analogue exhibiting a viscosity over 3,000,000 at the same temperature.

To preserve the advantages of hydrolysis resistance and good bonding characteristics, while offering higher (>1,000 g/mole) molecular weight phthalates with manageable viscosities (<3,000 cP at 25°C) and a range of compatibility with polyethers and polyesters, we have developed a series of low viscosity (“LV”) phthalate esters using a proprietary technology. One product, STEPANPOL^® PD-90 LV, is a nominal 90 mg KOH/g hydroxyl value diol (or 1,247 g/mole, calculated) with a typical room temperature viscosity under 2,500 cP. In this work, we investigated the behavior of solventless two-component polyurethane adhesives prepared from blends of PD-90 LV, with each of two other diethylene glycol phthalates as coreactants (see Table 1).

Specifically, various blends of PD-90 LV with PS-3152 or PS-4002 were measured for viscosity and inspected for visual compatibility; the viscosity data was analyzed, modeled, and mapped as a designed experiment using Design-Expert software, version 6.0.6 (Stat-Ease, Minneapolis).

Our investigation of the resistance of phthalate esters to hydrolysis was extended here from previous work⁴ to provide direct comparisons of the polyols in Table 1 and esters commonly used in FlexPack polyurethane adhesives. Additional polyols used in this study are shown in Table 2.

Laminates of aluminum foil to either treated biaxial oriented polypropylene (BOPP), linear low-density polyethylene (LLDPE) or polyethylene terephthalate (PET) films were made by hand using blends of polyols from Table 1 as coreactants and a commercially available prepolymer as an isocyanate. Peel strengths of these laminates were compared to a commercial standard prepared in the same manner using a modified version of ASTM Method D 1876-95.

Experimental Details

GPC Analysis for Hydrolysis Studies

Gel Permeation Chromatography (GPC) analysis was performed on a system using equipment and software from Waters Corp. The company’s Millennium 32 GPC software provided peak information (e.g., retention times and area counts).

The GPC system used in this work consisted of the following elements. THF (solvent grade material from VWR Scientific, stabilized with 250 ppm BHT) was used as the mobile phase at a 1.0 ml/min solvent flow rate delivered by a Waters model 515 dual-piston solvent pump. The samples were then injected by a Waters 717+ autosampler onto a column bank. The bank was held at 45˚C and consisted of four columns in series: first, two Styragel HR0.5s; one Styragel HR1.0; and finally, one Ultrastyragel 500 Å column. After separation on the column, detection was initially accomplished with a Waters 486 UV detector (set at 254 nm wavelength) and then a Waters 2410 RI detector (at 35˚C). Data acquisition was done on a personal computer equipped with a Pentium II processor using Waters’ Millennium 32 GPC software package. Approximately 125 ± 75 mg of each sample was dissolved in sufficient quantities of THF to produce sample concentrations of 2.0% by weight. About 3 ml of each sample solution was then filtered with a 0.45 mm syringe filter and placed into vials for injection into the GPC apparatus. Each analysis required 45 minutes of run time; the instrument was equilibrated for approximately 2.5 hours prior to the first injection of each day. Injections were made by the autosampler within 24 hours of loading into the vials.

Viscosity and Solubility

Separate blends of polyols were made with PD-90 LV and either PS-3152 or PS-4002 in weight ratios of 90:10, 75:25, 50:50, and 25:75, respectively, for both sets of blends. The viscosities at 25, 40 and 55˚C were measured using a Brookfield Thermosel (number 31 spindle). Hydroxyl values (ASTM E 222-73) were algebraically calculated from the equivalent weights of the individual polyols. A summary of the viscosity and hydroxyl value results are shown in Table 3. All mixtures were clear and remained homogenous at all these conditions.

Viscosity as a function of hydroxyl value and temperature of these polyol blends was evaluated using design of experiment (DOE) software.

Hydrolysis Resistance

General polyol hydrolysis resistance was evaluated by monitoring the acid value (ASTM D 4662) of 2% aqueous mixtures at 95°C as previously reported.4 For the base hydrolysis studies, 30% by weight of aqueous KOH (0.5 N) was mixed into the polyol and the mixtures were placed into a 95°C oven. Molecular modeling calculations to rationalize base hydrolysis phenomena were performed with PC Spartan Plus (Version 1.0) software.

Laminate Preparation and Adhesion Studies

All the laminates prepared in this investigation used a commercial prepolymer component applied for general-purpose flexible adhesive applications. This prepolymer had a % NCO content of 16.4% (ASTM D 2572-91).

Polyurethane adhesives consisted of coreactants made from blends of PD-90 LV with either PS-3152 or PS-4002 and the prepolymer. The prepolymer was mixed with the PD-90 LV mixtures so that the isocyanate:hydroxyl (NCO:OH) ratio was 1.2:1. The commercial laminate adhesive was made using the same NCO:OH ratio.

Laminates were prepared by bonding aluminum foil (0.0015” thickness, Metal Foils LLC) to polymer films with the experimental adhesives. The films used were: corona-treated, biaxial- oriented polypropylene (18 microns thickness, AQS BOPP, AET Films); linear, low-density polyethylene (2.4 mil thickness, medium slip, one side corona treated, Pliant); or polyethylene terephthalate (MYLAR LBT 2/48 gauge, DuPont).

Laminates were prepared by a laboratory method using a thermal laminator, file folders, tape and a draw-down roller. Films (approximately 8 x 11 inches) were secured to half of a file folder by taping at the edges. Adhesives were prepared by mixing prepolymer and coreactant and subsequently dissolving with acetone to make a 3% by weight solution. Approximately 1-2 mL of this solution was applied to an aluminum foil sample (cut to the same dimensions as the film), and a 3 mil draw-down roller was used to make a thin film of adhesive solution. Acetone was allowed to evaporate from the foil for 20 minutes prior to lamination with a film. The file folder was then closed, sandwiching the laminate inside, and the folder was passed through a thermal laminator at room temperature. The adhesive was allowed to cure at ambient conditions in the flat file folder for at least seven days before testing.

The amount of adhesive was determined by difference in weight using the average of two samples. Two one-inch square samples of bonded aluminum and plastic were weighed and compared with the average weight of two samples of non-bonded aluminum/plastic prepared in the same manner. The amount of adhesive was calculated to be the difference between the bonded and non-bonded squares.

T-Peel testing was accomplished using a variant of ASTM Method D 1876-95. Inch-wide strips were cut from the bonded samples after curing. At least five samples were pulled to obtain average maximum T-Peel strengths.

Pot life studies were run using the same NCO:OH ratio of prepolymer and polyol blends as used in the laminate preparations. Polyol blends and the prepolymer were heated to 50˚C for 1 hour before being combined by hand-mixing. The 50°C viscosity was recorded with a Brookfield Viscometer (#31 spindle) and Thermosel arrangement over 25 minutes.

Results and Discussion

Figure 1. Contour Plots from the DOE Study of Viscosity vs. Hydroxyl Value and Temperature for (a) PD-90 LV + PS-3152 and (b) PD-90 LV + PS-4002 Blends

Viscosity and Solubility

Plots of viscosity as a function of composition and temperature for blends of PD-90 LV + PS-3152 and PD-90 LV + PS-4002 are shown in Figure 1. Hydroxyl values are used as the x-axis to indicate composition.

The contour lines in the graphs represent constant-viscosity mixtures. With the PD-90 LV + PS-3152 blend, the viscosities were typically higher than the PD-90 LV + PS-4002 blend at the same temperatures due to the inherently lower viscosity of PS-4002.

Coreactants (B-side) for flexible packaging adhesives are reacted with the isocyanate component (A-side), usually a prepolymer, to make a flexible packaging adhesive. The coreactants are single polyols or multiple-polyol blends. Typically, these adhesives are applied at temperatures between 40-70ºC, so the viscosities of both the coreactants and the isocyanates are important to processing these adhesives. Another significant factor is the hydroxyl value of the coreactant, as this determines the amount of isocyanate required.

All blends of PD-90 LV with PS-3152 or PS-4002 prepared for the DOE study appeared clear at 25, 40 and 55ºC. Also, clear mixtures were observed for 1:2, 1:1 and 2:1 weight blends of PD-90 LV and PPG-2000, showing PD-90 LV to be compatible with a standard polyether polyol as well as polyester polyols.

Figure 2. Acid Value Change for Ortho-Phthalate Polyol PS-2002 vs. AZA-DEG, SBA-DEG Polyols at 95ºC with 2% Water

Hydrolysis Resistance

We previously reported on the hydrolysis of ortho-phthalate polyols, showing their superior hydrolytic stability compared to similar molecular weight polyols made from either aliphatic acids or polycaprolactones.4 Here, we extended this work to compare the hydrolytic stability of azelaic acid- and sebacic acid-based polyols to ortho-phthalate polyols, and to compare the hydrolysis of ortho-phthalate polyols in a basic environment to adipic acid-based polyol.

Separately, 2% water by weight was added to the AZA-DEG, SBA-DEG and PS-2002 polyols; acid values of these solutions were monitored at 95°C over time. Figure 2 shows the results of these tests.

Figure 3. GPC Trace of AZA-DEG Polyol Before and After Hydrolysis with 2% Water at 95ºC

Clearly, the ortho-phthalate polyol shows superior hydrolytic stability (as less increase in acid value) as compared to the aliphatic polyols. Additionally, the ortho-phthalate polyol, AZA-DEG and SBA-DEG polyols (before and after exposure to water) were compared using GPC. The results for the AZA-DEG polyol are shown in Figure 3.

Figure 4. GPC Trace of DEG-Phthalate Polyol Before and After Hydrolysis with 2% Water at 95°C

Figure 3 shows an appearance of a new peak after hydrolysis (peaks to the left in these plots represent higher molecular weights). Thus it appears that some lower molecular weight species is forming during this experiment, which suggests a breakdown of product (hydrolysis). A similar observation was seen when the SBA-DEG polyol was analyzed using GPC. GPC traces for PS-2002, before and after hydrolysis, are shown in Figure 4.

The GPC traces for the ortho-phthalate polyol are practically identical; no additional peaks are seen. From these results, it is apparent that the ortho-phthalate polyols are more hydrolytically stable than both the AZA-DEG and SBA-DEG polyols commonly used in flexible packaging adhesives. Presumably, this is due to the inherent tendency of the ortho position of the esters to hinder hydrolysis.⁴

Figure 5. Hydrolysis Results of PD-90 LV, PS-2002, PS-3152, PS-4002 and Adipic Acid/DEG Polyols

The hydrolysis results of the three polyols used in the laminate adhesive study were examined as well. Figure 5 shows the results of these ortho-phthalate polyols compared to the adipic acid-based polyol. These three polyols have better hydrolytic stability than an adipic acid-based polyol.

Figure 6. GPC Traces of 195 OH Value (a) DEG-Phthalate and (b) DEG-Adipate Polyols Before and After Exposure to Aqueous KOH

Our previous work on hydrolysis resistance⁴ involved polyols in neutral or acidic conditions. Following base hydrolysis by acid value changes is difficult; this process can be better tracked with chromatography. For base hydrolysis studies, 30% by weight of 0.5N KOH solution was added to PS-2002 and AA-DEG, respectively. These two mixtures were aged at 95°C for twenty days. The GPC traces of Figure 6a-b confirm the superior hydrolytic stability of the ortho-phthalate polyol.

Figure 7. (a) Optimized Geometry of the AA/DEG Diester with (a) LUMO Representation (White = Hydrogen, Red = Oxygen, Grey = Carbon) and (b) Electron Density LUMO Electrostatic Potential Surface

Computational Analysis of Base Hydrolysis

We previously used molecular modeling calculations to fundamentally explore why ortho-phthalate polyols have better hydrolytic stability compared to aliphatic acid polyols in neutral or acidic conditions.4 We now use a similar strategy to understand the hydrolytic stability of ortho-phthalate polyols in the presence of a strong base.

During ester hydrolysis, nucleophilic attack occurs on the carbonyl portion of an ester. Electrons from the nucleophile enter the lowest unoccupied molecular orbital (LUMO) as hydrolysis begins. Accordingly, the susceptibility to nucleophilic attack of a carbonyl portion of a molecule may be understood by analyzing the LUMO characteristics of an optimized conformation. Two molecular structures, an adipic acid-diethylene glycol diester (AA-DEG) and an ortho-phthalate diethylene glycol diester (OP-DEG), were optimized by semi-empirical methods using a conventional PM3 modeling algorithm.⁵

Figure 7a is a representation of an optimized structure with LUMO superimposition of the AA-DEG diester; the red and blue areas represent the LUMO. In Figure 7b, the electrostatic potential is mapped onto an isodensity surface that allows quantification of locations where nucleophilic attack would occur. Blue areas represent regions where nucleophilic attack is most likely to occur. The modeling program also allows estimation of the electrostatic potentials on these surfaces. The higher the number, the higher the susceptibility for nucleophilic attack. For the AA-DEG diester, this value was 0.179 kcal/mol at the carbonyl region of the ester.

Figure 8. Optimized Geometry of the Ortho-Phthalate with (a) LUMO Representation and (b) Electron Density LUMO Electrostatic Potential Surface

A similar analysis was done for the OP-DEG diester structure; Figure 8a shows the optimized structure and LUMO results. Note that most of the LUMO orbital is located over the benzene ring in contrast to the AA-DEG polyol where the majority of the LUMO is located on the carbonyl portion of the molecule. The electrostatic potential surface (Figure 8b) suggests the most susceptible area of attack resides on the benzene ring (0.053 kcal/mol). Nucleophilic attack, under the experimental conditions, does not typically occur in this region of the molecule. The carbonyl portion of the molecule is more susceptible for nucleophilic attack under these conditions and only had an electrostatic potential of 0.42 kcal/mol at the carbonyl region. Thus the aliphatic system, with LUMOs in the carbonyl region and comparatively higher electrostatic charges, should exhibit greater ease of hydrolysis than the phthalate analogue. This is in agreement with the experimental results.

Adhesion Profiles

This study evaluated several coreactant mixtures using PD-90 LV, PS-3152 and PS-4002. These blends were made at room temperature and then characterized. All blends did not separate and were clear. A commercial prepolymer was used in all of these laminates and had 16% NCO content. Table 4 shows the various properties of each mixture.

As Table 5 indicates, the commercial coreactant gave PET film tear. In the PD-90 LV blends, lower amounts of PS-3152 and PS-4002 gave maximum peel strengths comparable to the commercial product with both cohesive and substrate failure. Approximately equal amounts of PS-3152 or PS-4002 (50:50) provided results similar to the commercial product while using less adhesive.

Table 6 shows the results of laminates that were prepared with aluminum foil and BOPP. Here, the coreactant mixtures suggest that increasing either PS-3152 or PS-4002 gives higher maximum peel strengths; these laminates had similar adhesive amounts.

Table 7 documents the results for the aluminum/LLDPE films. In all of these cases, the adhesive appeared to migrate to the edges of the bonded substrates, leaving little in the middle of the samples. The amount of adhesive shown in Table 7 was measured at the edges where the adhesive was concentrated. This is one of the limitations of hand laminations vs. commercial machine laminations.

As with the Al and BOPP results above, Table 7 shows increasing maximum peel strength with increasing amounts of PS-3152 or PS-4002.

From these studies, it is clear that coreactant mixtures of PD-90 LV with either PS-3152 or PS-4002 provided comparable —or at times superior— performance to a commercial coreactant product.

Figure 9. Pot Life of Coreactant Blends with Commercial Prepolymer at 50ºC

Pot life results for the 50:50 blends compared with the commercial sample are shown in Figure 9. The pot life curves show that the ortho-phthalate polyol blends increased in viscosity faster than the commercial and PD-90 LV coreactant mixtures, which was expected as these blends had higher hydroxyl values than the commercial product.

Conclusion

In this study we demonstrated the versatile compatibility of a unique polyester polyol, PD-90 LV, with both polyester and polyether polyols. The temperature- and composition-dependency of viscosity of two series of blends were also mapped to enable chemists to match viscosity and formulation requirements.

Hydrolysis resistance of ortho-phthalate polyols was also demonstrated under acidic and basic conditions and was found to be superior to azelaic and sebacic acid-based polyols. GPC was used to confirm hydrolysis phenomena under both basic and acidic environments. Molecular modeling provided a rationalization for the remarkable resistance to base hydrolysis of the phthalates.

Laboratory preparations also showed ortho-phthalate polyols to give satisfactory bonding results between aluminum and TEP, BOPP and LLDPE films.

As a workhorse polyol, PD-90 LV, when used with either PS-3152 or PS-4002, gave a coreactant system that offered bonding characteristics comparable to a commercial system. Additionally, PD-90 LV offered the benefits of compatibility with polyesters and a polyether polyol , as well as a superior hydrolysis resistance not generally observed with conventional polyester polyols.

For more information, contact Stepan Co., 22 W. Frontage Road, Northfield, IL 60093; phone (847) 446-7500; fax (847) 501-2466; or e-mail mobrien@stepan.com .

STEPANPOL^® is a registered trademark of the Stepan Co.

This article is based on a paper presented at the Fall 2003 ASC Meeting in Cleveland.

SIDEBAR: Abbreviations

PU: Polyurethane

OHV: Hydroxyl value

AZA-DEG: Azelaic acid-diethylene glycol polyester polyol

SBA-DEG: Sebacic acid-diethylene glycol polyester polyol

PET: Polyethylene terephthalate

BOPP: Oriented polypropylene

LLDPE: Linear low-density polyethylene

AA-DEG: Adipic acid-diethylene glycol polyester polyol

OP-DEG: Ortho-phthalate diethylene glycol polyester polyol