Properties related to linear and branched network structure of hydroxyl terminated polybutadiene

2.1 Materials

HTPB and N100 polyisocyanate were manufactured by Liming Institute of Chemical Technology of China (Luoyang, China). The hydroxyl value of HTPB was 0.78 mmol/g and the average molecular weight was 2800 g/mol. The average molecular weight of N100 was 728 g/mol and contained 5.32 mmol/g of the isocyanate group. Dioctyl adipate (DOA) and 2,4-TDI was purchased from Sinopharm Chemical Reagent Co., Ltd. Triphenyl bismuth (TPB) was used as the catalyst and synthesized by the Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences (Shanghai, China). Raw materials were dried in a vacuum oven at 50°C for 48 h before using.

2.2 Curing kinetic and viscosity measurement

The proportion of HTPB and curing agents in the compositions is listed in Table 1, which adheres to the stoichiometric ratio (R_NCO/OH) of 1:1. The concentration of TPB was 0.015 wt% as is commonly used in PBXs (6). The mixture of HTPB, TPB and curing agents were stirred using a single paddle mixer for 15 min to ensure proper dispersion and degassed for 10 min. After mixing, the mixture was moved to a Netzsch (Selb, Germany) STA449C calorimeter to perform the non-isothermal DSC measurement immediately to investigate the curing kinetics. Heating was done from ambient temperature to 300°C under an argon flow of 10 mL/min at heating rates of 5, 10, 15 and 20°C/min. Samples of 10 mg were used for the measurements.

A Brookfield (Middleboro, USA) LVDV-II⁺ digital viscometer combined with HA-7 spindle were used to measure the viscosity of the polymer compositions. The temperature was maintained at 60°C by a Brookfield TC-550AP temperature controller.

2.3 Preparation of HTPB-based polyurethanes and swelling test

Compositions were put into a convection oven at 50°C for 5 days to obtain PU films. The equilibrium swelling method was used to measure the swelling capacity and crosslink density of HTPB-based PUs with TDI and N100. Films were cut into specimens with dimension of 7 × 7 × 2 mm and immersed in selected solvents (Table 2) with different solubility parameters for about 72 h after weighing (W₁) and weighted again (W₂) until the mass of film remained constant. Subsequently, the solvents absorbed were removed by putting the swelled films in convective oven at 100°C, and the constant weight (W₃) was recorded.

2.4 Tensile test

HTPB-based PUs with different R_NCO/OH were prepared for determining the mechanical properties. Tensile test was processed using a Shimadzu (Kyoto, Japan) AGS-J electronic mechanical universal testing machine at a crosshead speed of 10 mm/min according to ASTM D882-10. Standard dimension specimens were also made by this standard test method, width of the specimen is 5 mm, length is 70 mm and thickness is about 0.5 mm. The abovementioned operational processes were performed at 25°C and a relative humidity of 50%. Tests were replicated for 5 times to obtain the average values of stress and strain.

2.5 Thermal analysis and isothermal accelerated aging

Thermal stability was investigated by differential scanning calorimetry-thermogravimetry-derivative thermogravimetry (DSC-TG/DTG) coupling analysis using a Netzsch STA449C calorimeter. The heating was done from 100 to 500°C at 10°C/min under an argon flow of 10 mL/min.

For the determination of the migration of plasticizers, HTPB-based PUs with different curing agents were prepared using DOA as the plasticizer. The mass ratio of HTPB and DOA in aging samples was 1:1. Aging was performed according to the “China test methods for the qualification of military composite explosive formulation”. A specimen with a diameter of 24 mm and a length of 125.7 mm was wrapped by monolayer filter paper and settled in a mold with corresponding dimensions. The specimen was heated in an oven at 71°C for 320 h. The mass change of the grain and the filter paper during heating was recorded. After the aging test, specimens were cut manually equidistance from the center of the diameter as shown in Figure 9. Methanol as the solvent was added to equiponderate specimens for 24 h. The extraction was moved to a volumetric flask, and the residue was washed with methanol and the washing was also removed to the volumetric flask. The prepared solution was performed using an Agilent (Santa Clara, USA) Technologies 1200 Series high performance liquid chromatograph (HPLC). The concentration of DOA was calculated to evaluate the migration in parallel with mass loss. A Hitachi TM3000 scanning electron microscope (SEM) was used to check the surface of specimens aging in the same conditions without being wrapped by filter paper.

3.1 Curing kinetic of HTPB/isocyanates

The DSC curves of the curing process of HTPB with TDI and N100 curing agents at different heating rates are shown in Figure 1 with some of the characteristic exothermic temperatures. It can be directly observed that the curing temperatures of HTPB/TDI are lower than that of HTPB/N100 at all the heating rates. It suggests that TDI is more active, allowing the urethane reaction between HTPB and TDI to occurr rapidly. However, the heat release of the HTPB/N100 curing reaction is higher due to the difference in the temperatures of formation between the isocyanate curing agents (23).

Figure 1:

DSC curves of HTPB curing process with different curing agents.

(A) HTPB/TDI; (B) HTPB/N100.

The original Kissinger (24) and Crane method (25) were used to calculate the kinetic parameters, viz., apparent activation energy (E_a), pre-exponential factor (A) and the reaction order (n).

Kissinger equation

Crane equation

as E_a/nR≫2T_p, Crane equation can be rewritten as brief form.

where β is the heating rate, E_a is the apparent activation energy, n is the reaction order and R is the gas constant (8.314 J/mol/K). The kinetic parameters calculated according to Eq. [1–3] are listed in Table 3. It is noticed that E_a of HTPB/N100 is higher, which is consistent with the exothermic temperatures. However, the higher pre-exponential factor (A) suggests a higher probability of collision between HTPB and N100 under the same temperature due to the numerous branched isocyanate groups. As for the consistent n of curing reactions, it indicates that the elementary reactions with different curing agents were identical. The urethane reaction abides by the n order reaction model according to the research on the curing kinetic of the polymer (26), (27).

where k is the reaction rate constant, α is the curing degree, t is the reaction time and A is the pre-exponential factor in the Arrhenius equation. The E_a of curing reactions were also calculated at various curing degrees from 0 to 100% in steps of 10%. As can be seen from Figure 2, whereas the E_a of HTPB/TDI at different curing degrees remained almost the same, the gradual attenuation of E_a appeared in HTPB/N100 with the reaction processing.

Figure 2:

Apparent activation energy of the HTPB urethane reaction at different curing degrees.

3.2 Viscosity build-up

As shown in Figure 3, the initial viscosity of HTPB/TDI is slightly lower. It is observed that the viscosity of HTPB/TDI increases rapidly as the E_a is relatively lower according to the curing kinetic. However, a remarkable deceleration of viscosity build-up was observed after 200 min due to the steric hindrance effect (28) of two isocyanate groups on both sides of benzene in TDI. Whereas the viscosity of HTPB/N100 is lower, the viscosity build-up appeared as a gradual increase without apparent attenuation. It indicates that HTPB/N100 delivers a slow viscosity build-up and contributes to extend the pot life of HTPB-based composites.

Figure 3:

Viscosity build-up of HTPB/isocyanates mixtures at 60°C.

3.3 Crosslink density

From the weights of the swollen (W₂) and dried (W₃) specimens, the swelling capacity (Q) is given by

The weight and volume fraction of the polymer (w_p, v_p) and the solvent (w_s, v_s) can then be calculated by

The solubility parameters of HTPB-based PUs can be obtained from the peak fitting of the equilibrium swelling capacity at different solvents. The swelling curves of HTPB/TDI and HTPB/N100 are shown in Figure 4, hence, the δ_HTPB/TDI and δ_HTPB/N100 can be obtained.

Figure 4:

Swelling curves of HTPB polyurethanes.

δHTPB/TDI=17.33 J1/2/cm3/2 and δHTPB/N100=17.36 J1/2/cm3/2

The crosslink density (D_c) can be calculated by the Flory-Rhener equation (29),

where V_S is the molar volume of the solvent, and χ is the polymer-solvent interaction parameter, which is related to the solubility parameters (δ) of the polymer and the solvents,

where R is the gas constant and T is the temperature in Kelvin.

Crosslink density can be calculated using the corresponding parameters of different solvents according to Eq. [9, 10]. For amyl acetate, V_S is 0.14773 L/mol, hence, χ_HTPB/TDI and χ_HTPB/N100 are 0.352 and 0.377, respectively, at 298 K. D_{c (HTPB/TDI)} is 96 and D_{c (HTPB/N100)} is 205 as calculated from the parameters in amyl acetate. The crosslink density of PUs depicted in Figure 4 was obtained from the average value of D_c in different solvents. Although the network structures of HTPB with TDI and N100 were different, δ were almost the same. It appears that δ mainly depends on the polybutadiene soft segment instead of the hard segments formed by curing agents. Besides, swelling capacity decreases with the increasing crosslink density (30). Therefore, the higher crosslink density of HTPB/N100 was validated. It suggests that HTPB cured with N100 possesses high crosslink density, and hence, a complex network structure due to the highly branched isocyanate group in the N100.

3.4 Mechanical properties

The tensile curves of HTPB-based PUs with R_NCO/OH=1 are depicted in Figure 5 with the calculation result of tensile modulus at the 20% strain. The maximum tensile strengths of HTPB/N100 and HTPB/TDI were 0.455 MPa and 0.398 MPa with breaking elongations of 295% and 266%, respectively. It suggests that both elongation and tensile strength are improved by the branched N100 due to the higher crosslink density. The tensile modulus of HTPB/TDI was higher than that of HTPB/N100 at the initial part and decreases with the elongation until the elongation reaches 46%, after which, tensile modulus of HTPB/TDI remains in a low state. Crystallization can be observed during tensile tests as the transparency of the specimen was reduced. The difference of the degree of crystallinity between HTPB/TDI and HTPB/N100 changes the modulus.

Figure 5:

Tensile strain-stress curves of HTPB polyurethane films.

The correlation of mechanical properties with R_NCO/OH is shown in Figure 6. Higher tensile strength and breaking elongation of HTPB/N100 were observed in all R_NCO/OH. However, the tensile strength increased with the R_NCO/OH, and decreased after the R_NCO/OH exceeds 1.1. It suggests that the curing reaction was unfinished due to the inadequate quantitative consumption of isocyanate, and the excess isocyanate did not contribute to the mechanical properties. The breaking elongation maximized at the R_NCO/OH of 0.9, and afterwards decreased with the increase of R_NCO/OH. It seems that unexpended isocyanate may react with the moisture in the environment and generates gaseous CO₂ as the reactivity of water with isocyanate is between the primary and secondary hydroxyl group, and the micropores in the polymer matrix led to the attenuation of the mechanical properties.

Figure 6:

Mechanical properties of HTPB polyurethanes cured by TDI and N100 curing agents and different stoichiometric ratios of NCO/OH.

3.5 Thermal stability

Figure 7A depicts the DSC curves with the characteristic exothermic temperatures. Two major steps were observed in the decomposition of HTPB-based PUs. The first exothermic peak was thought to be the depolymerization and slight decomposition of hard segments according to the difference of heat release, which was also verified by the mass loss as shown in Figure 7B. Whereas the thermal stability of aromatic TDI is higher than that of aliphatic N100 (31), the first step decomposition temperature of HTPB/N100 shifted to a higher temperature by about 6°C compared to the HTPB/TDI with a linear network structure. It suggests that the greater crosslink density and branched network contributes to the structural stability of PUs, and hence the thermal stability. The peak temperatures of the second step decomposition are nearly the same due to the decomposition of the polybutadiene soft segment. However, the difference of heat flow might be due to the different mass proportions of HTPB and curing agents in PUs. There was a slight exothermic peak after 510°C without apparent loss of mass in the TG curve. It is speculated to be the decomposition of the butadiene gaseous products. The mass loss of HTPB/TDI occurred earlier as the temperatures of 5% mass loss shown in Figure 7B, which was consistent with the DSC curves. Unlike the research of Vieira et al. (32) on HTPB/TDI, there are two slight peaks in the first step decomposition of HTPB/TDI as observed from the DTG curve. It may be caused by the configuration difference of the diisocyanate group. Moreover, there is only one peak in the DTG curve of the first step decomposition of HTPB/N100 as N100 is an aliphatic isocyanate with branched isocyanate groups.

Figure 7:

Thermal decomposition of HTPB polyurethanes.

(A) DSC curves; (B) TG and DTG curves.

3.6 Migration of plasticizer

The HTPB/DOA-based PUs after aging and isothermal accelerated aging molds are shown in Figure 8. It is observed that the filter paper wrapped on the HTPB/TDI cylinder is saturated by DOA and becomes translucent. The mass loss of the cylinder and the mass gain of filter paper were 1.70 g and 1.66 g, respectively (total mass of HTPB/TDI/DOA is 51.62 g) while DOA in the HTPB/N100 cylinder (with total mass of 51.90 g) was still contained in the polymer matrix after aging with a slight mass change of 0.4 g due to the higher crosslink density. The SEM images of the surface of PU films after aging are shown in Figure 9. The dense phase separation points on the surface suggest that there was grievous migration of the plasticizer in the HTPB/TDI polymer matrix. The phase separation points in the HTPB/N100 films are far less than that of HTPB/TDI as the mass loss of HTPB/N100 is only 0.77%.

Figure 8:

Isothermal accelerated aging mold, HTPB/DOA/isocyanates polyurethanes and filter papers after aging.

Figure 9:

SEM images of the surface of HTPB/TDI and HTPB/N100 after aging.

The equiponderate specimens were manually cut and the concentration of DOA inside the HTPB-based PUs measured by HPLC after aging are shown in Figure 10. The concentration of DOA in HTPB/TDI decreased with the distance from the center. It suggests that the DOA on the surface was absorbed by the filter paper and the DOA inside migrated towards the surface. Whereas the concentration of DOA on the surface of HTPB/N100 was decreased as shown at 12 mm, the concentrations at other distances were almost unchanged. It indicates that the branched network structure of HTPB/N100 prevents the migration of DOA.

Figure 10:

The specimens cut equidistantly for HPLC and the concentration of DOA inside the HTPB polyurethanes with a different network after aging.