Polymer Degradation and Stability 81 (2003) 19–27

www.elsevier.com/locate/polydegstab

Degradation profiles of cast films of polyurethane and

poly(urethane-urea) aqueous dispersions based on hydroxy-
terminated polybutadiene and different diisocyanates
Fernanda M.B. Coutinhoa,b,*, Marcia C. Delpechb, Thais L. Alvesa, Antonio A. Ferreirac
a
Instituto de Macromoléculas Professora Eloisa Mano- Universidade Federal do Rio de Janeiro, IMA/UFRJ, C.P. 68.525, 21945-970,
Rio de Janeiro, RJ, Brazil
b
Departamento de Processos Quı´micos/Instituto de Quı´mica- Universidade do Estado do Rio de Janeiro, IQ/UERJ, RJ, Brazil
c
Petroflex Indústria e Comércio S/A, RJ, Brazil

Received 2 December 2002; accepted 20 January 2003

Abstract
New formulations of waterborne polyurethanes (wPU) and poly(urethane-urea)s (wPUU) were produced based on hydroxy-ter-
minated polybutadiene (HTPB), poly(propylene glycol) (PPG), dimethylolpropionic acid (DMPA), and three different types of
aliphatic diisocyanates—isophorone diisocyanate (IPDI), 4,40 -dicyclohexylmethane diisocyanate (HMDI) and hexamethylene dii-
socyanate (HDI)—and the aromatic tolylene diisocyanate (TDI). Three types of chain extender were used in some of the formula-
tions: ethylene glycol (EG), forming polyurethanes, and hydrazine (HYD) or ethylenediamine (EDA), resulting in poly(urethane-
urea)s. In the formulations, HTPB content and NCO/OH ratio were varied. The thermal stability of the materials, obtained as cast
films prepared from aqueous dispersions was evaluated by thermogravimetry (TG). It was observed that initial degradation tem-
peratures were above 200  C, with two- or three-step degradation profiles. An increase in HTPB content, the presence of urea lin-
kages, and HMDI as diisocyanate in the formulations led to materials with higher thermal stability. DTG curves exhibited stages
not perceptible in the curves of weight loss, which were mainly influenced by the differences in the formulations.
# 2003 Elsevier Science Ltd. All rights reserved.
Keywords: Polyurethanes; Aqueous dispersions; Hydroxyl-terminated polybutadiene; TG; DTG

1. Introduction The most common wPU and wPUU produced in

industry and mentioned in literature are those based on
The development of waterborne polyurethane (wPU) polyethers or polyesters and very few papers relate the
and poly(urethane-urea) (wPUU) formulations has been use of hydroxyl-terminated polybutadiene (HTPB) as
increasing due primarily to the control of volatile polyol due, probably, to its hydrophobic nature, which
organic compound emissions to atmosphere and also to makes the dispersion in aqueous medium difficult [1,2].
the superior properties presented by these resins when HTPB is generally used in the preparation of adhe-
compared with the similar ones obtained in organic sives, coatings, elastomers, encapsulates, propellants, in
medium. Polyurethane aqueous systems can be tailor- many fields such as automobile, coatings, building-
made and present many of the features related to con- trade, electronic, aerospace and medicine [9,10].
ventional solvent-born coatings and several technologi- Another important field, in which HTPB-based poly-
cal advantages, including low viscosity and good urethanes are employed, is in the manufacture of gas
applicability as coatings for different types of substrates, separation membranes because of their low temperature
wood, concrete, leather, metal and some polymers [1–8]. flexibility and high segregation between hard and soft
segments due to the nonpolarity of HTPB [1,11,12].
* Corresponding author. Tel.: +55-2587-7322; fax: +55-21-2587-
The thermal stability of polyurethanes and poly
7227. (urethane-urea)s has been extensively studied because of
E-mail address: fern@uerj.br (F.M.B. Coutinho). their great importance and their wide range of applica-
0141-3910/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0141-3910(03)00058-2
20 F.M.B. Coutinho et al. / Polymer Degradation and Stability 81 (2003) 19–27

tions. Thermogravimetry (TG) is a suitable method and the water. After the addition of EG, the reactions
widely used to evaluate the thermal properties of these were maintained at 45  C, for 1 h. Finally, water was
materials [3,13]. added, with high speed stirring, and aqueous dispersions
In a previous study, the degradation profile, based on with solids content of 35 2.5% (w/w) obtained.
TG and DTG curves, of polyurethane and poly(urethane- In the polyurethane formulations, the amount of
urea) cast films produced from aqueous dispersions was DMPA was fixed at 60% in equivalent-number of total
discussed. The formulations were based on poly(propylene OH groups (number of OH groups present in the
glycol) (PPG), 4,40 -dicyclohexylmethane diisocyanate hydroxylated compounds: DMPA, and the polyols PPG
(HMDI) and dimethylolpropionic acid (DMPA). Fur- and HTPB). The variations performed were: the type of
thermore, three types of chain extender were used: diisocyanate, the HTPB content (from 10 to 50% in
ethylene diamine (EDA) and hydrazine (HYD), forming equivalent-number relative to polyols OH groups), and
poly(urethane-urea)s, and ethylene glycol (EG), produ- NCO/OH prepolymer ratio (2.0 and 3.0), which were
cing polyurethanes [3]. the ratio between the equivalent-numbers of NCO
In this work, degradation profiles of some aqueous groups and total OH groups. It is important to empha-
dispersion-cast films based on hydroxy-terminated size that as the HTPB content increased, the dispersion
polybutadiene (HTPB), PPG, DMPA, and four differ- in water became more difficult, due to both the higher
ent diisocyanates were evaluated. The diisocyanates viscosity of the prepolymer and its non-polar character.
were the aromatic tolylene diisocyanate (TDI) and the Films were obtained by casting the aqueous disper-
aliphatics: isophorone diisocyanate (IPDI), 4,40 -dicyclo- sions on leveled polyethylene surfaces and allowing
hexylmethane diisocyanate (HMDI) and hexamethylene them to dry at room temperature, for 7 days, and then
diisocyanate (HDI). In some formulations HYD, EDA at 60  C, for 12 h. The thickness of the films were below
or EG were used as chain extenders. 1 mm. The cast films obtained from the aqueous dis-
persions became more flexible as the HTPB content in
the formulation increased.
2. Experimental
2.3. TG analysis
2.1. Reagents
Thermogravimetric experiments were performed in a
The following reagents were used as received: TA Instruments TG 2950 Analyser. Film samples ran-
dibutyltin dilaurate (DBTDL), Aldrich; 4,40 -dicyclo- ging from 2 to 6 mg were placed in a platinum sample
hexylmethane diisocyanate (HMDI), Hüls; dimethyl- pan and heated from 25 to 600  C, under N2, at a heat-
olpropionic acid (DMPA), Aldrich; ethylene diamine ing rate of 10  C/min. During the heating period, the
(EDA), Reagen; ethylene glycol (EG), Vetec; hexa- weight loss and temperature difference were recorded as
methylene diisocyanate (HDI), Bayer; hydrazine a function of temperature.
(HYD), 64%, Bayer; hydroxy-terminated polybutadiene
(HTPB) (M  n =3500, determined by VPO), Petroflex;
isophorone diisocyanate (IPDI), Hüls; poly(propylene 3. Results and discussion
glycol) (PPG), M  v =1200, Dow Chemical; tolylene di-
isocyanate (TDI), Bayer; and triethylamine (TEA), Polyurethane chains are composed of flexible seg-
Union Carbide. ments, derived from polyols, and rigid or hard segments
formed by the urethane linkages. Urea linkages may
2.2. Synthesis of aqueous dispersions and films also be present in the rigid segment, leading to the for-
preparation mation of poly(urethane-urea) chains [14].
The thermal stability of these thermoplastic elasto-
The synthesis of the prepolymer based on HTPB, mers is generally not high, specially above their soft-
PPG, DMPA and the diisocyanate, catalyzed by ening temperatures [15], and the degradation
(DBTDL), was carried out in bulk, at 100  C, for 2 h, mechanism is very complex due to the variety of pro-
under N2 atmosphere. The prepolymer neutralization ducts formed in the process [3,16,17]. It was proposed
with TEA was conducted at 50  C, for 1 h. The tem- that the thermal degradation of polyurethane is pri-
perature was reduced to 25  C and the dispersion of the marily a depolycondensation process, which starts at
bulk in water was carried out by high speed agitation. about 250  C [3,18], and presents a bimodal profile. The
NCO-terminated prepolymer chains were reacted with first mode is related to the hard segments, poly(urethane-
EDA or HYD, in a chain extension step, at 35  C, for 1 urea)s being more stable than polyurethanes [19].
h, forming urea groups. The chain extender EG was Usually, at a low heating rate, the degradation process
added before the dispersion, in order to avoid the com- results in differential weight loss (DTG) curves with
petition between the hydroxyl groups of the extender several peaks, which indicates the complexity of the
 F.M.B. Coutinho et al. / Polymer Degradation and Stability 81 (2003) 19–27 21

degradation [3,13]. Those peaks correspond to the tem- finally at 208  C, when IPDI was employed. HDI, which
peratures at maximum rate of weight loss in the corre- contains the most flexible portion, –(CH2)6–, seems to
sponding step. facilitate the thermal degradation. On the other hand,
In the present study, the onset degradation tempera- the cyclic structures seem to be more resistant, specially
ture (Tonset) of the monomers, polyurethanes and HMDI, which is more symmetrical than IPDI and,
poly(urethane-urea)s was defined as the initial tempera- consequently, promotes stronger interactions between
ture of degradation, corresponding to the intercept of the chains by hydrogen bonding. This way, the former
the tangent drawn at the inflection point of the decom- imparts higher thermal resistance than the latter.
position step with the horizontal zero-line of the TG Fig. 2 compares the thermal resistance of formula-
curve [3,20,21]. Table 1 shows the TG parameters tions using diisocyanates with the aliphatic structures
obtained for the hydroxyl-terminated monomers. HTPB (HMDI and IPDI) and the aromatic TDI, at same
showed the highest thermal resistance. NCO/OH ratio (2.0), 10% HTPB content, and HYD as
TG curves for some formulations are presented in chain extender. Degradation in three discrete stages
Figs. 1–5 and some corresponding DTG curves are starts at 199  C for TDI, 205  C for IPDI and 208  C
depicted in Figs. 6–8. TG curves show two or three dis- for HMDI. At the beginning of the process, the aro-
crete stages of degradation. In general, the materials matic diisocyanate resulted in polymer chains with
obtained in the absence of HTPB presented a two- lower thermal stability. The aromatic rings, with 2,4- and
degradation-step profile and those synthesized with 2,6-isocyanate groups formed more rigid chains but also
HTPB, three stages, which are clearly distinguished in more asymmetrical, which seems to make the interactions
derivative curves (DTG). between the hard segments difficult, reducing their ther-
It is in the hard segment, composed of urethane lin- mal resistance. However, the presence of TDI seemed to
kages, that the initial thermal decomposition occurs. impart higher thermal stability to the flexible segments,
When urea linkages are present, the thermal stability is compared with the aliphatic structures.
higher, probably because of the higher hydrogen-bond- Figs. 3 and 4 show the influence of HTPB content on
ing capacity of this group when compared with urethane the thermal stability of formulations prepared with
[3,14]. The subsequent stages of degradation correspond HMDI (NCO/OH=2.0) and IPDI (NCO/OH
to the soft segments, formed from polyols [3,15,16]. ratio=3.0), respectively. HYD was the chain extender
Figs. 1 and 2 present the influence of diisocyanate used. The calculated onsets for the beginning of degra-
type on thermal stability for some formulations. Fig. 1 dation process are listed in Table 2. It can be observed
shows the degradation profiles of films obtained from that the presence of a low content of the flexible chain
aqueous dispersions, synthesized with the aliphatics of HTPB in the formulation (10%) did not influence the
HDI, HMDI and IPDI, at same NCO/OH ratio (3.0), thermal stability, comparing with the formulation with-
and in the absence of HTPB. The chain extender out HTPB. On the other hand, for the formulations
employed was HYD. The profiles show two discrete with higher contents of HTPB, the thermal stability was
steps of degradation when HDI is present and three notably improved.
when the cyclic diisocyanates make part of the struc- We also observed higher values for the initial degra-
ture. The first step, corresponding to the rigid segments, dation temperature of HMDI series, in which NCO/OH
begins at 197  C, for HDI, at 202  C, for HMDI and ratio or rigid segment content was lower than in the IPDI

Table 1
TG data obtained for the di-hydroxyl monomers

Monomers T1onset( C)a Tmax ( C)b m (%)c

HTPB 418 443 99

PPG 313 351 100

T1onset ( C)d T2onset ( C)e T1m ( C)f T2m( C)g m1 (%)h m2 (%)i

DMPA 204 293 234 302 71 26

a
Initial temperature of degradation.
b
Temperature at maximum rate of weight loss.
c
Percentage of weight loss.
d
Initial temperature of degradation of first step.
e
Initial temperature of degradation of second step.
f
Temperature at maximum rate of weight loss of first step.
g
Temperature at maximum rate of weight loss of second step.
h
Percentage of weight loss of first step.
i
Percentage of weight loss of second step.
22 F.M.B. Coutinho et al. / Polymer Degradation and Stability 81 (2003) 19–27

Fig. 1. TG curves of poly(urethane-urea)s synthesized with different types of aliphatic diisocyanates, HYD as chain extender, in the absence of
HTPB, and at NCO/OH ratio=3.0.

Fig. 2. TG curves of poly(urethane-urea)s synthesized with different aliphatic and aromatic diisocyanates, HYD as chain extender, 10% HTPB, and
at NCO/OH ratio=3.0.
 F.M.B. Coutinho et al. / Polymer Degradation and Stability 81 (2003) 19–27 23

Fig. 3. Influence of the amount of HTPB in the degradation profiles of poly(urethane-urea)s materials synthesized with HMDI, HYD as chain
extender, and at NCO/OH ratio=2.0.

Fig. 4. Influence of the amount of HTPB in the degradation profiles of poly(urethane-urea)s materials synthesized with IPDI, HYD as chain
extender, and at NCO/OH ratio=3.0.
24 F.M.B. Coutinho et al. / Polymer Degradation and Stability 81 (2003) 19–27

Fig. 5. Influence of the chain extender type in the degradation profiles of polyurethane and poly(urethane-urea)s synthesized with HDI, at NCO/OH
ratio=3.0, in the absence of HTPB.

Fig. 6. DTG curves of poly(urethane-urea)s synthesized with different aliphatic and aromatic diisocyanates, HYD as chain extender, 10% HTPB,
and at NCO/OH ratio=3.0.
 F.M.B. Coutinho et al. / Polymer Degradation and Stability 81 (2003) 19–27 25

Fig. 7. DTG curves of poly(urethane-urea)s materials synthesized with HMDI, HYD as chain extender, at NCO/OH ratio=2.0, employing different
amounts of HTPB.

Fig. 8. DTG curves of polyurethane and poly(urethane-urea)s synhesized with HDI, different chain extenders, at NCO/OH ratio=3.0, in the
absence of HTPB.
26 F.M.B. Coutinho et al. / Polymer Degradation and Stability 81 (2003) 19–27

Table 2 Three peaks are observed. The first group of peaks,

Initial degradation temperature (onset 1) observed for cast films of corresponding to the degradation of rigid segments
poly(urethane-urea) aqueous dispersions, with different HTPB content formed by urethane and urea linkages, presents max-
HTPB contenta (%) T1onset ( C)b imum temperatures varying from 233 to 249  C. The
second group, related to the degradation of PPG soft
HMDI//2.0c IPDI//3.0c segment, varies from 316 to 358  C, and the third group,
0 213 208 observed in the range from 419 to 436  C, corresponds
10 213 208 to the degradation of HTPB soft segment.
30 230 216 Fig. 7 depicts DTG curves related to TG profiles
50 230 219
observed in Fig. 3, in which HTPB content varied. The first
a
Percentage in equivalent-number relative to polyol OH groups. group of peaks appears in the range from 233  C to 255  C.
b
Initial temperature of degradation. The second group can be observed in the range varying
c
NCO/OH ratio. from 319 to 330  C, for PPG soft segments, and the third
group, related to HTPB ones appears from 427 to 444  C.
series. Once more it can be observed that the former pro- Fig. 8 presents DTG curves correspondent to TG
motes more symmetry to the rigid segments than the latter ones shown in Fig. 5. A marked difference can be
and, consequently, stronger interactions between the observed, promoted by changing the type of chain
chains by hydrogen bonding, even in smaller quantities. extender in DTG profiles, especially in the first stage of
Thus, TG curves exhibited in Figs. 3 and 4 show that weight loss, corresponding just to urethane (EG as
HTPB has a marked influence on the degradation, chain extender), and urea (HYD or EDA as chain
especially for HMDI series, in which the interactions extenders) linkage degradation. The soft segment,
between rigid segments are stronger. formed only by PPG, degradation step seemed to be
Fig. 5 presents TG curves of formulations synthesized also affected. The rigid segment formed from hydrazine
with HDI (NCO/OH ratio=3.0) and without HTPB, in retarded the weight loss of PPG chains (peak at 331  C),
which the chain extender was varied. Ethylene glycol while EDA and EG promoted similar profiles of weight
(EG) formed urethane linkages by the reaction with loss, both showing peaks at 315  C.
terminal NCO groups and both hydrazine (HYD) and All DTG curves showed that there are different stages
ethylene diamine (EDA) lead to the formation of urea of degradation which are not perceptible in TG curves,
groups. The initial onsets observed are: 197  C when showing the close relation and mutual influence between
EG was employed, 211  C and 233  C, when the chain degradation of hard and soft segments [3].
extenders were HYD and EDA, respectively. The ther-
mal stability was higher when a diamine chain extender
was used, in comparison with the use of the diol, prob- 4. Conclusion
ably because of the higher hydrogen-bonding capacity
of urea groups [14]. It should be expected that HYD as The degradation profiles of poly(urethane-urea)s and
chain extender would impart initial higher thermal sta- polyurethane cast films obtained from aqueous disper-
bility than EDA because of the hard segment formed sions were influenced by the type of rigid segment
(–HN–CO–HN–NH–CO–NH–), in contrast with the formed. HMDI (4,40 -dicyclohexylmethane diisocyanate),
one formed with EDA (HN–CO–HN–CH2–CH2–NH– with a more symmetrical structure, improved the interac-
CO–NH) [3]. However, an unexpected behaviour was tions between chains leading to formulations with higher
observed, the initial degradation stage for HYD-based thermal stability. The presence of HTPB (hydroxy-termi-
formulation started before that of EDA. On the other nated polybutadiene), which formed part of the flexible
hand, the former degradation process, as a whole, showed segments, had a marked influence on the degradation of
higher thermal resistance than the latter, which began to the polymers, especially in elevated quantities, improving
decompose at a higher temperature but continued to lose the thermal stability of the materials. The type of chain
mass at lower temperatures than the HYD-based film. extender, forming urethane or urea or linkages, affected
In general, the presence of urea groups has also been the whole process of degradation and the presence of urea
found to have a stabilizing effect on the resultant poly- groups improved the thermal resistance of the chains.
urethane, as can be observed in the curves obtained for
the samples chain-extended with HYD and EDA, in
comparison with the samples obtained with EG as chain Acknowledgements
extender probably due to the liberation of amine during
the first stage of degradation [3,10,11]. The authors thank Conselho Nacional de Desenvol-
Fig. 6 shows the DTG curves corresponding to the vimento Cientı́fico e Tecnológico (CNPq) and FAPERJ
TG degradation profiles presented in Fig. 2, in which for financial support and Petroflex Indústria e Comércio
the diisocyanates were varied (HMDI, TDI or IPDI). for TG analysis.
 F.M.B. Coutinho et al. / Polymer Degradation and Stability 81 (2003) 19–27 27

References [11] Huang S-L, Lai L-Y. J Appl Polym Sci 1995;58:1913.
[12] Huang S-L, Lai J-Y. J Membrane Sci 1995;105:137.
[1] Coutinho FMB, Delpech MC, Garcia MEF. Polym Test 2002; [13] Chang TC, Chiu YS, Chen HB, Ho SY. Polym Degrad Stab
21:719. 1995;47:375.
[2] Coutinho FMB, Delpech MC, Alves LS. J Appl Polym Sci 2001; [14] Hepburn C. Polyurethane elastomers. 2nd ed. London: Elsevier;
80:566. 1991.
[3] Coutinho FMB, Delpech MC. Polym Degrad Stab 2000;70:49. [15] Wang TL, Hsieh TH. Polym Degrad Stab 1997;55:95.
[4] Delpech MC, Coutinho FMB. Polym Test 2000;19:939. [16] Ferguson J, Petrovic Z. Eur Polym J 1976;12:177.
[5] Shao CH, Huang JJ, Chen GN, Yeh JT, Chen KN. Polym [17] Grassie, N., Scott, G. Polymer degradation and stabilisation
Degrad Stab 1999;65:359. England: Cambridge University Press; 1985. p. 41.
[6] Coutinho FMB, Delpech MC, Moura PMQ, Mello SDS. Polym [18] Grassie N, Zulfiqar M. J Polym Sci 1978;16:1563.
Bull 1996;37:1. [19] Bhattacharyya J. In: Cheremisinoff NP, editor. Physical chemical
[7] Coutinho FMB, Delpech MC. Polym Test 1996;15:103. crosslinking of polyurethane, vol. 2: handbook of polymer
[8] Satguru R, McMahon J, Padget JC, Coogan RG. J Coat Technol science and technology. New York: Marcel Dekker; 1989. p. 578.
1994;66(830):47. [20] Nakamura K, Hatakeyama T, Hatakeyama H. Polym Adv
[9] Scharpman F, Couvercelle JP, Bunel C. Polymer 1998;39(4):965. Technol 1992;3:151.
[10] Haska SB, Bayramli E, Pekel F, Özkar S. J Appl Polym Sci 1997; [21] Hatakeyama H, Izuta Y, Kobashigawa K, Hirose S, Hatakeyama
64:2347. T. Macromol Symp 1998;130:127.