Cyanate esters (CEs), developed in the early 1970s, are an increasingly important class of matrix resins. They have attracted widespread attention for their high performance in aerospace applications. Their advantages include excellent heat resistance, good mechanical properties, and very low moisture absorption. These attractive features render CEs suitable for use in higher-performance composites such as printed circuit boards, structural composites, and radomes.^1–3 However, applications have been limited by poor flexibility, e.g., the brittleness of CE resins due to high crosslink density. Addressing this limitation, thermoset toughening with liquid reactive rubbers significantly increases modified CE system toughness.^4–7

Hydroxyl-terminated polybutadiene (HTPB), developed in the 1960s, is a reactive liquid rubber that may be useful for simultaneously overcoming the limitations and retaining the advantages of CE resins. Since there is neither an ester nor an ether bond in the HTPB molecular skeleton, it features good stability and moisture resistance. HTPB rubber further possesses excellent low-temperature and dielectric properties. Consequently, an HTPB-modified CE resin should exhibit similar superior properties.

We used the monomer of bisphenol A dicyanate ester (BADCy) as received (purity >99%, melting point 79°C). The rubber was HTPB (number-average molecular weight, M_n, of 3000), and the catalyst was dibutyl tin dilaurate (DBTDL). HTPB was dissolved in butyl acetate, and mixed with BADCy as well as 0.02 weight percent (wt%) DBTDL, with vigorous stirring at room temperature. The solution was heated in an oil bath to drive off most of the butyl acetate. The mixture was then placed in vacuum to remove the residual solvent and air bubbles. This was followed by casting into a mold for curing at 130, 150, 180, and 200°C for 2h at each temperature. The resulting cured BADCy resin was demolded and cooled to room temperature. The samples were subsequently stored in a desiccator for testing.

We measured tensile strength and elongation at break with an Instron-4200 digital stress-strain tester at a crosshead speed of 10.0mm/min, using the American Society for Testing and Materials (ASTM) D790 method. Figure 1 illustrates typical stress-strain responses of the cured BADCy/HTPB resin systems. The tensile strength and modulus decreased with increasing HTPB content in the modified resins, due to a reduction in crosslinking density of the resin system and the introduction of flexible molecular chains.

We carried out hermogravimetric analysis (TGA) using a TA 2100-SDT 2960 instrument (Perkin-Elmer) from 25 to 750°C at a heating rate of 20°C/min under nitrogen atmosphere. Figure 2 presents TGA curves and corresponding derivative (differential thermogravimetric) curves for BADCy and BADCy/HTPB blends with different HTPB contents. The onset decomposition temperature clearly decreased with addition of HTPB relative to neat BADCy. However, the onset thermal decomposition temperatures of BADCy and BADCy blended with 10, 20, 30, and 40 HTPB were 446.2, 448.5, 430.2, 446.3, and 433.1°C, respectively (i.e., all above 400°C), indicating good thermal stability. The maximum degradation rate temperatures of BADCy and BADCy/HTPB blends were 452.4, 469.7, 471.6, 471.2, and 469.1°C, respectively. Both the onset degradation temperatures and maximum degradation rate temperatures changed slightly when the content of HTPB was low relative to that of neat BADCy. Generally, greater HTPB content imparted lower thermal stability to the blend. This is due to smaller rubber particles and a more homogeneous distribution in the matrix, both of which are beneficial for heat absorption.

We used scanning electron microscopy (SEM, Hitachi S-3400) to examine the fracture surfaces of the cured resins (see Figure 3). The samples were coated with a thin layer of gold prior to imaging. Neat BADCy possessed only one phase: see Figure 3(a) and (b). A few straight cracks were present in the matrix. The smooth surface illustrates a brittle fracture behavior. Cured resin morphology changed drastically with the addition of HTPB. Inclusion of HTPB led to phase separation, which is primarily dependent on HTPB content and determines fracture toughness: Figure 3(c) and (d). With increasing HTPB content, morphology varied from dispersed rubber domains in a crosslinked cyanate ester matrix to co-continuous structures: see Figure 3(e–h). When the HTPB content reached 50wt%, small particles were again observed, indicating phase separation: see Figure 3(i) and (j).

In summary, CE resin blend thermal stability slightly decreases with increasing HTPB composition. The modified resin systems essentially maintain their excellent thermal stability. With increasing HTPB content, morphology varies from dispersed rubber domains in a crosslinked BADCy matrix to co-continuous structures and dispersed BADCy domains in a rubbery matrix. The primary toughening mechanism of BADCy/HTPB blends is shear-yielding and crack-branching. These results will be useful for future efforts toward improving the properties of CE resins intended for high-performance composite applications.