**4.1. Experimental**

86 Polyurethane

discussed.

**catalyst** 

ratio, determine the properties of PUD.

understand the structure-property relationship. The properties of PUD are mainly determined by the interactions between the hard and soft segments, and by the interactions between the ionic groups. The ionic group content, solids content, segmented structure, molecular weight of the macroglycol, the type of chain extender and the hard/soft segments

In the following sections, we will review typical results to demonstrate the utility of TGA in deducing the structural and bonding information about waterborne PUD based on different polyols. The thermal stability of PU and poly(urethane –urea)s cast films with anionomer character, obtained from waterborne dispersions and based on isophorone diisocyanate (IPDI), dimethylolpropionic acid (DMPA), poly(propylene glycol) (PPG), polycarbonate diol (PCD) and glycolized products obtained from recycled poly(ethylene terephthalate)(PET) is also compared. Three types of chain extenders were used: ethylene glycol (EG), propylene glycol (PG) and ethylene diamine (EDA). The effect of type of polyols, chain extender, type of catalyst, ionic content, length of soft segment, hard segment content and the presence of urea or urethane linkages on the thermal stability of the waterborne anionic PUD are

**4. Water-based PUD based on poly(propylene glycol) and selective** 

not the isocyanate-water reaction (Colling et al., 2002; Blank &Tramontano, 1996).

reactive solvent with the metal content of 0.4%.

One of the inherent drawbacks of waterborne PU technology is the formation of carbon dioxide due to the side reactions of isocyanate with water. When an isocyanate reacts with water, the products are a urea linkage (via an amine intermediate) and carbon dioxide. The carbon dioxide formation is problematic in that it causes imperfections in the coating during cure, such as blistering and pin-hole formation. The main aspect in the development of waterborne PU is in the first place to find methods for preventing the undesired secondary reactions with water and achieving the best crosslinking. This reaction is reduced to a minimum by the use of non-tin catalysts. One novel approach to control the water side reaction is the use of catalysts which selectively catalyze the isocyanate-polyol reaction and

The relative selectivity (S) obtained from equation S = Purethane/Purea, was measured as urethane IR peak area (Purethane)/ urea IR peak area (Purea) ratio, by method given by Blank (Blank et al., 1999). After the integration of characteristic absorption max of urethane (1700 cm-1, 1540 cm-1) and urea (1640 cm-1, 1570 cm-1) was done, the relative selectivity was calculated. The manganese catalyst, a complex of Mn(III)–diacetylacetonatomaleate with various ligands based on acetylacetonate and maleic acid, used in some of the experiments (Stamenković et al., 2003; Cakić et al., 2006), has shown a high selectivity for the isocyanate– hydroxyl reaction in comparison to the commercially available zirconium catalyst (Blank et al., 1999). Zirconium catalyst is a proprietary zirconium tetra-dionato complex in the

TG is a suitable method to evaluate the thermal properties of several types of PU elastomers. The thermal stability of PU has been studied extensively because of the great importance of Poly(propylene glycol) (PPG) ), (Mn = 1000, hydroxyl value 111 mg KOH/g, dried under vaccuum, at 120 oC), isophorone diisocyanate (IPDI) and dimethylolpropionic acid (DMPA), were obtained from Aldrich Chemical Co. 1-methyl-2-pyrrolidone (NMP), dimethyl formamide (DMF) and triethyl amine (TEA) were received from Merck (Darmstadt, Germany). Ethylene glycol (EG) and propylene glycol (PG) obtained from Zorka Co.(Šabac Serbia). Dibutyltin dilaurate (DBTDL), was supplied by Bayer AG. Zirconium catalyst (ZrCat) was supplied by King Industries Inc., Norwalk, CT, USA. Manganese catalyst (MnCat) has been used in the reactive diluent with a metal content of 0.4% (Stamenković et al., 2003; Cakić et al., 2006 b). Water–based PUD from PPG with selective catalyst were prepared using the prepolymer method has been described in detail in our previously work (Cakić et al., 2009).

In the first step, PPG and DMPA were dispersed in DMF to obtain a homogeneous mixture and heated at 70 oC. IPDI and DBTDL were added to the homogenized mixture at 80 oC. An NCO/OH equivalent ratio of 3.0 was used. Hard : Soft segment ratio was defined as a ratio between IPDI weight and polyol weight in the starting formulation and is calculated as 1.5. The reaction times were determined by the dibutyl amine back titration method. After obtaining completely NCO terminated prepolymer, the mixture was cooled to 60 oC and the carboxylic groups were neutralized with TEA (DMPA equiv) dissolved in NMP. In second step the chain extension was carried out with EG or PG. The selective catalysts, ZrCat or MnCat, at concentration of 2% relative to the resin solids, have been added to the reaction solution. Water was added to the mixture and stirred to obtain dispersion of organic phase in water. The waterborne PUD contains 40 wt% solids (Cakić et al., 2009). Films were prepared by casting the waterborne dispersions on leveled surfaces and allowing them to dry at room temperature, for 7 days, and then at 60 oC, for 12h (Coutinho, 1996, 2003). When chain extenders were EG and PG, the PUD had to be cast in Teflon surfaces due to the high adhesiveness observed on the glass surface, making demoulding impossible. After demoulding, the films were kept into a desiccator to avoid moisture.and polyurethanes were formed.

#### **4.2. Results and discussion**

Degradation profile of waterborne PUD is influenced by the variation of chain extender presented in Fig.1. It was verified that the thermal stability was influenced by chain extender type. In a general way, thermal stability was higher when EG chain extender was used, in comparison to PG, probably because of the higher reactivity of the primary hydroxyl groups. The onset temperatures calculated for the first stage for PU chainextended with EG were about 234 oC and 140 oC for PG. Unsymmetrical structure of IPDI enables easier diffusion of EG (Gunduz & Kisakurek, 2004).

Thermal Analysis of Polyurethane Dispersions Based on Different Polyols 89

higher when MnCat was used, in comparison with the use of ZrCat. This result suggested that all the residual NCO groups in PU particles did not react with the chain extender completely. Because the viscosity of particle is high at low temperature in the chain extension step, it would take long time for chain extenders to diffuse into the particle. Therefore, the efficiency of chain extension increased as total surface area of particles increased (Jang et al., 2002; Cakić et al., 2007 a). In general, the presence of more selective catalyst has also been found to have a stabilizing effect on the resultant PU, as can be observed in the curves obtained for the samples chain-extended with EG and PG, in comparison to the samples obtained without selective catalyst probably due to favoring of isocyanate-polyol reaction and not the isocyanate-water

The DTG curves show that there are different stages of degradation which are not perceptible in TG curves, showing the close relation and mutual influence between the

Fig. 3a shows the DTG curves corresponding to the TG degradation profiles presented in Fig. 2a, in which the catalysts were varied (MnCat or ZrCat). The chain extender employed was EG. Two peaks are observed. The first group of peaks, corresponding to the degradation of rigid segments formed by urethane and urea linkages, presents maximum of the peak from 200 to 250 oC. The second group, related to the degradation of PPG soft segment, varying maximum of the peak from 375 to 418 oC . The peaks shifting towards higher temperatures resulting from addition of more selective catalyst confirm the assumption that all isocyanate groups had not reacted with the added chain extender. Selective catalyst isocyanate-polyol reaction causes greater incorporation of chain extender in hard segments, which is reflected on higher thermal stability of hard segments (Jang et al.,

Fig. 3b depicts DTG curves related to TG profiles observed in Fig. 2b, in which the catalyst were varied (MnCat or ZrCat). The chain extender employed was PG. The first group of peaks presents maximum of the peak appearing in the range from 160 oC to 195 oC. The

**Figure 3.** DTG curves of PUD with EG as chain extender (a), PG as chain extender (b), and without

reaction. The PUD was formed with higher hard segment proportions.

degradation process of hard and soft segments.

2002; Lee et al., 1995; Cakić et al., 2007 b).

catalyst (1), ZrCat (2), (3) MnCat

**Figure 1.** TG curves of PUD without catalyst with EG (1) and PG (2) as chain extender

Fig. 2 shows the degradation profile of PUD with variation of catalyst using catalysts of different selectivity. EG and PG formed urethane linkages by reaction with terminal NCO groups. The initial onsets observed are: 234 oC when EG was employed, 275 oC and 311 oC, when the chain extender was EG with ZrCat and EG with MnCat, respectively. The initial onsets observed are: 140 oC when PG was employed, 290 oC and 305 oC, when the chain extender was PG with ZrCat and PG with MnCat, respectively. The thermal stability was

**Figure 2.** TG curves of PUD with EG as chain extender (a), PG as chain extender (b), and MnCat (1), ZrCat (2), without catalyst (3)

higher when MnCat was used, in comparison with the use of ZrCat. This result suggested that all the residual NCO groups in PU particles did not react with the chain extender completely. Because the viscosity of particle is high at low temperature in the chain extension step, it would take long time for chain extenders to diffuse into the particle. Therefore, the efficiency of chain extension increased as total surface area of particles increased (Jang et al., 2002; Cakić et al., 2007 a). In general, the presence of more selective catalyst has also been found to have a stabilizing effect on the resultant PU, as can be observed in the curves obtained for the samples chain-extended with EG and PG, in comparison to the samples obtained without selective catalyst probably due to favoring of isocyanate-polyol reaction and not the isocyanate-water reaction. The PUD was formed with higher hard segment proportions.

88 Polyurethane

extender type. In a general way, thermal stability was higher when EG chain extender was used, in comparison to PG, probably because of the higher reactivity of the primary hydroxyl groups. The onset temperatures calculated for the first stage for PU chainextended with EG were about 234 oC and 140 oC for PG. Unsymmetrical structure of IPDI

enables easier diffusion of EG (Gunduz & Kisakurek, 2004).

**Figure 1.** TG curves of PUD without catalyst with EG (1) and PG (2) as chain extender

Fig. 2 shows the degradation profile of PUD with variation of catalyst using catalysts of different selectivity. EG and PG formed urethane linkages by reaction with terminal NCO groups. The initial onsets observed are: 234 oC when EG was employed, 275 oC and 311 oC, when the chain extender was EG with ZrCat and EG with MnCat, respectively. The initial onsets observed are: 140 oC when PG was employed, 290 oC and 305 oC, when the chain extender was PG with ZrCat and PG with MnCat, respectively. The thermal stability was

**Figure 2.** TG curves of PUD with EG as chain extender (a), PG as chain extender (b), and MnCat (1),

ZrCat (2), without catalyst (3)

The DTG curves show that there are different stages of degradation which are not perceptible in TG curves, showing the close relation and mutual influence between the degradation process of hard and soft segments.

Fig. 3a shows the DTG curves corresponding to the TG degradation profiles presented in Fig. 2a, in which the catalysts were varied (MnCat or ZrCat). The chain extender employed was EG. Two peaks are observed. The first group of peaks, corresponding to the degradation of rigid segments formed by urethane and urea linkages, presents maximum of the peak from 200 to 250 oC. The second group, related to the degradation of PPG soft segment, varying maximum of the peak from 375 to 418 oC . The peaks shifting towards higher temperatures resulting from addition of more selective catalyst confirm the assumption that all isocyanate groups had not reacted with the added chain extender. Selective catalyst isocyanate-polyol reaction causes greater incorporation of chain extender in hard segments, which is reflected on higher thermal stability of hard segments (Jang et al., 2002; Lee et al., 1995; Cakić et al., 2007 b).

Fig. 3b depicts DTG curves related to TG profiles observed in Fig. 2b, in which the catalyst were varied (MnCat or ZrCat). The chain extender employed was PG. The first group of peaks presents maximum of the peak appearing in the range from 160 oC to 195 oC. The

**Figure 3.** DTG curves of PUD with EG as chain extender (a), PG as chain extender (b), and without catalyst (1), ZrCat (2), (3) MnCat

second group can be observed in the range, varying maximum of the peak, from 267 to 347 oC, for PPG soft segments. A marked difference can be observed, promoted by changing the type of chain extender in DTG profiles, especially in the first stage of weight loss, corresponding just to urethane (EG or PG as chain extender) linkage degradation. The soft segment, formed only by PPG degradation step seemed to be also affected. The rigid segment formed from EG retarded the weight loss of PPG chains (peak at 375 oC), while PG showing peaks at 267 oC (Cakić et al., 2006, 2007 c). All DTG curves showed that there are different stages of degradation which are not perceptible in TG curves, showing the close relation and mutual influence between degradation of hard and soft segments.

Thermal Analysis of Polyurethane Dispersions Based on Different Polyols 91

*Example 2 of glycolysis reaction:* In the second experimental runs of depolymerisation, appropriate amount of PET waste were added to 396.1 g PG, 750 g TEG or 2173 g PEG 400, so that molar ratio of PET repeating unit to glycol was 1:10. These mixtures (with different molar ratio PET/glycol) and 0.5 wt.% zinc acetate (based on the weight of PET as transesterification catalyst) were charged to a glass reactor, which was fitted with stirrer, reflux condenser, nitrogen inlet and temperature controller. This reactor was immersed in an oil bath and the content of the reaction kettle was heated at 190 oC for 2 h, subsequently

The obtained glycolyzed oligoester polyols were analysed by the hydroxyl value (HV) determination according to the conventional acetic anhydride/pyridine method (Cakić et al., 2011). The hydroxyl value of the oligoester polyol obtained in the glycolysis reaction based on molar ratio of PET repeating unit to glycol, 1:2, with PG was HBPG=490 mg KOH/g, TEG

The hydroxyl value of the oligoester polyol obtained in the glycolysis reaction based on molar ratio of PET repeating unit to glycol, 1:10, with PG was HBPG=201 mg KOH/g, TEG

**5.2. Synthesis of PUD based on glycolyzed products with molar ratio PET/glycol,** 

Anionic PUD based on glycolyzed products with molar ratio PET/glycol, 1:2, were prepared by modified acetone process. Acetone was added to the prepolymer and the dispersion is formed by the addition of water to this solution. Procedure for synthesis of anionic PUD has been developed adjusting the molar ratio of DMPA as a hydrophilic monomer to IPDI as 1:3.3. Mass of oligoester polyol, obtained by PET glycolysis, according to example 1, with a hydroxyl number which is equivalent to the hydroxyl number of 0.06 mol of poly(propylene glycol) PPG1000 (110 mg KOH/g), was for PG 15 g, TEG 20 g, poly (ethylene glycol) (PEG 400) 25 g.

The oligoester polyol and hydrophilic monomer (8 g, equ. 0.06 mol) was mixed in the cosolvent DMF (50:50 w/w), in a 250-ml round four-neck glass reactor connected to a stirrer, a thermometer, a reflux condenser and a nitrogen gas inlet. The reaction was carried out at 70 oC for 30 min to obtain a homogeneous mixture and the uniform distribution of hydrophilic monomer to PU backbone. IPDI (44.4 g, equ.0.2 mol) and catalyst DBTDL (0.03% of the total solid) were added to the homogenized mixture at 80 oC for about 4h until the amount of residual NCO groups reached a theoretical value, as determined by the dibutyl amine backtitration method. To reduce the viscosity and obtain a homogenous mixture of NCO prepolymers, acetone was added 50 wt% to the solid reaction mass. The theoretical value of NCO groups for PUD based on oligoester polyol obtained from the glycolysis with molar ratio of PET repeating unit to glycol, 1:2, was 19.2% for PG, 17.7%, for TEG and 16.1% for PEG.

After obtaining completely NCO terminated prepolymer, the mixture was cooled to 60 oC and the carboxylic groups in hydrophilic monomer were neutralized with TEA (DMPA

equ). TEA was dissolved in NMP by stirring the solution for 60 min.

the temperature was raised to 210 oC until all the solids disappeared.

HBTEG=370 mg KOH/g and PEG HBPEG400=297 mg KOH/g.

HBTEG=209 mg KOH/g and PEG HBPEG400=192 mg KOH/g.

**1:2** 

The degradation profiles of PU cast films obtained from water-based dispersions were influenced by the type of chain extender, length of the hard segment and type of catalysts. The presence of more selective catalysts, which formed urethane linkages with higher hard segment proportions, had a marked influence on the degradation of the polymers, especially in elevated quantities, improving the thermal stability of the materials. The DTG curves showed that the length of the hard segment had a strong influence on the thermal profile of the samples as a whole. The type of chain extender, forming urethane linkages, affected the whole process of degradation and the presence of more selective catalyst improved the thermal resistance of the chains.
