Reinforcement of Castor Oil-Based Polyurethane with Surface Modification of Attapulgite

Polyurethane/attapulgite (PU/ATT) nanocomposites derived from castor oil were prepared by incorporation of 8 wt % ATT, acid-treated ATT, and KH560-treated ATT. The effects of three ATTs (ATT, acid-ATT, and KH560-ATT) on the comprehensive properties of PU/ATT nanocomposites were systematically investigated. The results showed that the incorporation of 8 wt % of three ATTs could produce an obvious reinforcement on the castor oil-based PU and that the silane modification treatment, rather than the acid treatment, has the more effective reinforcement effect. SEM images revealed the uniform dispersion of ATT in the PU matrix. DMA confirmed that the storage modulus and glass transition temperature (Tg) of PU/ATT nanocomposites were significantly increased after blending with different ATTs. For PU/KH560-ATT8 nanocomposites, the thermal stability of the PU was obviously enhanced by the addition of KH560-ATT. In particular, 8 wt % KH560-ATT loaded castor oil-based PU nanocomposites exhibit an obvious improvement in tensile strength (255%), Young’s modulus (200%), Tg (5.1 °C), the storage modulus at 25 °C (104%), and the initial decomposition temperature (7.7 °C). The prepared bio-based PU materials could be a potential candidate to replace petroleum-based PU products in practical applications.


Introduction
Recently, the use of renewable vegetable oil-based polyols to replace the commercial petroleum-based polyether or polyester polyols in the synthesis of bio-based polyurethane (PU) [1][2][3][4][5] has attracted great public attention as functional materials in many applications, such as thermoplastic [6,7], thermosets elastomers [8,9], adhesives [10][11][12], coatings [13][14][15], foams [16,17], medical fields [18,19] and nanocomposites [20,21]. The advantages of these bio-based materials are their biocompatibility, biodegradability, low cost, good thermal properties, and acceptable specific strength properties [22,23]. In comparison to other vegetable oils, castor oil has on average 2.7 hydroxyls per triglyceride and can therefore be directly used to synthesize PU materials [24]. However, the resulting pure castor oil-based PU is too soft to meet application requirements in some fields. This may be attributed to the intrinsic low OH value of castor oil and low cross-linking density of PU matrix [25,26]. sonicated for 60 min. The castor oil was added and stirred vigorously at room temperature (RT) for 24 h. The acetone was completely removed at 80 • C. The IPDI was then added and stirred for 30 min while cooling to RT. The mixtures with 8 wt % loading of three ATTs were poured to a PTFE (Teflon) mold of 100 mm diameter and 4 mm depth and degassed under vacuum until all bubbles disappeared. The samples were first cured at 90 • C for 2 h and then 110 • C for 24 h. Finally, the samples were slowly cooled down to RT and demolded. The resulting PU nanocomposites with 8 wt % ATT, acid-ATT, and KH560-ATT were denoted as PU/ATT8, PU/acid-ATT8 and PU/KH560-ATT8, respectively.

Characterization
X-ray diffraction (XRD) patterns was recorded by a XRD-6000 (Shimadzu, Kyoto, Japan) with a crystal monochromated Cu K α radiation (λ = 0.154 nm), the scanning rate was 6 • ·min −1 from 6 • to 60 • . Dynamic mechanical analysis (DMA) was measured by a DMA+450 (01dB-Metravib, Limonest, France) under tension mode at a fixed frequency (1 Hz) with a heating rate of 2 • C·min −1 from −70 to 50 • C for each sample. The mechanical performances of PU/ATT nanocomposites were performed at 23 • C using an 4466 universal testing machine (Instron, Norwood, MA, USA) at a strain rate of 50 mm·min −1 based on ASTM D638 standard. At least six specimens were tested for obtaining the average value of mechanical results. Thermogravimetric analysis (TGA) were performed under a N 2 atmosphere with a TGA/DSC 1 STAR e system (Mettler-Toledo, Greifensee, Switzerland) at a heating rate of 20 • C·min −1 in the temperature range 30-600 • C. Scanning electron microscopy (SEM) was observed by an S-4800 instrument (Hitachi, Tokyo, Japan), the acceleration voltage of the electron beam was 3 kV. The liquid nitrogen fractured surface of the samples was sputter coated with gold to provide enhanced conductivity before observation.

X-ray Diffraction (XRD) Analysis
XRD has been a useful tool for characterizing the structure of clay and its nanocomposites. The XRD patterns of different ATTs and PU/ATT nanocomposites were presented in Figure 1.  121) and (061) planes of the ATT, respectively. The XRD patterns of the neat ATT clearly exhibited the six-plane characteristic peaks. The characteristic reflection of both acid-ATT and KH560-ATT do not change at all. The results indicated that the crystal structure of ATT did not alter or distort after acid-treated and silylated surface modification. As shown in Figure 1, the castor oil-based PU exhibited a strong broad diffraction peak at 2θ = 20.0 • , which is caused by the short-range order in the arrangement of amorphous PU chain segments. For the PU/ATT nanocomposites, the new characteristic diffraction peak at 2θ = 9.0 • is attributed to the characteristic diffraction peak of ATT in the PU matrix. The results demonstrated that the blended ATTs in the PU matrix could keep their intrinsic crystal structures.

Dynamic Mechanical Properties
DMA can provide insight into the effect of surface modified ATTs on the thermo-mechanical performances of the obtained PU/ATT nanocomposites. Figure 2 showed the temperature dependence of the storage modulus (logarithmic E ) and the loss tangent (tan δ) for neat PU and PU nanocomposites containing 8 wt % ATT, acid-ATT and KH560-ATT. The dynamic mechanical spectra regarding the logarithmic E were very similar for castor oil-based ATT-reinforced PU nanocomposites. A steep decrease of logarithmic E is observed, corresponding to improved chain mobility in the PU matrix upon transition through glass transition temperature (T g ). The logarithmic E curves that demonstrated the logarithmic E value of PU/ATT nanocomposites within the whole test temperature range were indeed larger than that of neat castor oil-based PU, which indicated the effective reinforcement of three ATTs in the PU matrix. The neat castor oil-based PU had an E at 25 • C (E 25 ) of 2.64 MPa. The E 25 of PU/ATT8 and PU/acid-ATT8 was increased 70% and 77% compared with that of neat PU, respectively. The maximum improvement (increased 104%) in E 25 is observed for PU/KH560-ATT8 with 8 wt % KH560-ATT. Such significant improvement in E 25 is likely to be due to good compatibility and interphase interaction between KH560-ATT and the PU matrix.

Dynamic Mechanical Properties
DMA can provide insight into the effect of surface modified ATTs on the thermo-mechanical performances of the obtained PU/ATT nanocomposites. Figure 2 showed the temperature dependence of the storage modulus (logarithmic E′) and the loss tangent (tan δ) for neat PU and PU nanocomposites containing 8 wt % ATT, acid-ATT and KH560-ATT. The dynamic mechanical spectra regarding the logarithmic E′ were very similar for castor oil-based ATT-reinforced PU nanocomposites. A steep decrease of logarithmic E′ is observed, corresponding to improved chain mobility in the PU matrix upon transition through glass transition temperature (Tg). The logarithmic E′ curves that demonstrated the logarithmic E′ value of PU/ATT nanocomposites within the whole test temperature range were indeed larger than that of neat castor oil-based PU, which indicated the effective reinforcement of three ATTs in the PU matrix. The neat castor oil-based PU had an E′ at 25 °C (E′25) of 2.64 MPa. The E′25 of PU/ATT8 and PU/acid-ATT8 was increased 70% and 77% compared with that of neat PU, respectively. The maximum improvement (increased 104%) in E′25 is observed for PU/KH560-ATT8 with 8 wt % KH560-ATT. Such significant improvement in E′25 is likely to be due to good compatibility and interphase interaction between KH560-ATT and the PU matrix.
The Tgs of neat PU and PU nanocomposites containing 8 wt % ATT, acid-ATT and KH560-ATT were presented in Figure 2b and Table 1. It is well known that Tg reveals the segmental mobility of polymer chains in the matrix. Generally, an enhanced Tg indicated that the segmental mobility of the polymer chains decreased [51]. It is noted that three ATTs in general increased the Tg of PU/ATT nanocomposites. Compared with the Tg of the neat PU (5.5 °C), the Tgs of PU/acid-ATT8 and PU/KH560-ATT8 were respectively improved by 5.6 and 5.1 °C. The improvement can be attributed to both acid-ATT and KH560-ATT decreased segmental mobility of the castor oil-based PU chains and the enhanced interfacial interaction between the ATTs and the PU matrix. Usually, the thermo-mechanical properties of thermosets usually were strongly associated with crosslink density (νe). Based on the kinetic theory of rubber elasticity [52], the νe of the cured crosslinking systems can be calculated from the rubbery E′ using Equation (1): The T gs of neat PU and PU nanocomposites containing 8 wt % ATT, acid-ATT and KH560-ATT were presented in Figure 2b and Table 1. It is well known that T g reveals the segmental mobility of polymer chains in the matrix. Generally, an enhanced T g indicated that the segmental mobility of the polymer chains decreased [51]. It is noted that three ATTs in general increased the T g of PU/ATT nanocomposites. Compared with the T g of the neat PU (5.5 • C), the T gs of PU/acid-ATT8 and PU/KH560-ATT8 were respectively improved by 5.6 and 5.1 • C. The improvement can be attributed to both acid-ATT and KH560-ATT decreased segmental mobility of the castor oil-based PU chains and the enhanced interfacial interaction between the ATTs and the PU matrix. Usually, the thermo-mechanical properties of thermosets usually were strongly associated with crosslink density (ν e ). Based on the kinetic theory of rubber elasticity [52], the ν e of the cured crosslinking systems can be calculated from the rubbery E using Equation (1): where E represents the storage modulus of cured crosslinking systems in the rubbery plateau region, T is the absolute temperature, and R is the gas constant. In this study, the rubber modulus at T g + 40 • C was selected for the calculation of ν e . The data for E at 50 • C (E 50 ) and the calculated ν e of PU/ATT nanocomposites are summarized in Table 1. The value of ν e for castor oil-based PU was 269.2 mol/m 3 . It is noted that the ν e of PU/KH560-ATT8 was increased to 513.6 mol/m 3 (increased by 91%). Furthermore, the ν e of PU/ATT8 and PU/acid-ATT8 was respectively increased 69% and 40%, compared with that of neat PU.

Mechanical Properties
For superior PU nanocomposites, the mechanical performances of PU nanocomposites are significantly influenced by the nanofiller dispersion and interfacial interaction between nanofiller with the PU matrix. The ATT surface modification via acid treatment and KH560 modification should be an effective strategy to improve ATTs' dispersion and load transfer efficiency. The stress-strain curves and tensile performances of neat PU and PU nanocomposites containing 8 wt % ATT, acid-ATT and KH560-ATT were presented in Figure 3 and the detailed data were listed in Table 2. From the curves in Figure 3a, neat PU and three PU nanocomposites showed similar linear elastic deformation behavior. The neat PU had a tensile strength of 2.0 MPa and a Young's modulus of 1.9 MPa. The tensile strength and Young's modulus of PU/ATT nanocomposites were dramatically improved with incorporation of different ATTs into the PU matrix. The maximum improvement in the tensile strength (255%) and Young's modulus (200%) was observed for PU/KH560-ATT8, which is likely to be due to better homogeneous dispersion and good compatibility between KH560-ATT and the PU matrix. The PU/ATT8 and PU/acid-ATT8 had a 30% and 170% increment in the tensile strength, and had a 105% and 137% increment in Young's modulus, respectively. The increase of tensile strength indicated that three ATTs have an obvious reinforcement effect on the PU matrix, which were consistent with the DMA results. The reinforcement effects of different ATTs are in the order of KH560-ATT > acid-ATT > ATT and thus the KH560-ATT has the best reinforcement effect among three ATTs. It is noted that the elongation at break of PU/acid-ATT8 and PU/KH560-ATT8 were higher than that of neat PU. The elongation at break slightly improved because the decrease of ductility was not significant after incorporation of homogeneous dispersed ATTs, which contributed to a tougher PU matrix.

Thermal Stability
The thermal stabilities of neat PU and PU nanocomposites containing 8 wt % ATT, acid-ATT and KH560-ATT were investigated with TGA in nitrogen. Figure 4 displays the TGA thermograms versus temperature of neat PU and PU/ATT nanocomposites. The TGA and DTG curves were analyzed and the data were listed in Table 3. The addition of 8 wt % KH560-ATT into the PU matrix remarkably improved the IDT (5% mass loss temperature), T10% and T50% of PU. Especially, the IDT of PU/KH560-ATT8 was delayed by 7.7 °C compared with that of neat PU. The TGA results indicated that the incorporation of KH560 surface modified ATT into the PU matrix significantly improved its thermal stability. The increase in thermal stability was due to the barrier effect of well dispersed KH560-ATT ( Figure 5), which delays the release of volatile degradation compounds from the PU matrix.

Thermal Stability
The thermal stabilities of neat PU and PU nanocomposites containing 8 wt % ATT, acid-ATT and KH560-ATT were investigated with TGA in nitrogen. Figure 4 displays the TGA thermograms versus temperature of neat PU and PU/ATT nanocomposites. The TGA and DTG curves were analyzed and the data were listed in Table 3. The addition of 8 wt % KH560-ATT into the PU matrix remarkably improved the IDT (5% mass loss temperature), T 10% and T 50% of PU. Especially, the IDT of PU/KH560-ATT8 was delayed by 7.7 • C compared with that of neat PU. The TGA results indicated that the incorporation of KH560 surface modified ATT into the PU matrix significantly improved its thermal stability. The increase in thermal stability was due to the barrier effect of well dispersed KH560-ATT ( Figure 5), which delays the release of volatile degradation compounds from the PU matrix.   As can be seen in Figure 4b, the decomposition processes of PU/ATT nanocomposites were similar with that of the PU. The decomposition of neat PU and PU/ATT nanocomposites showed the typical decomposition steps [53,54]: the first loss step (230-368 °C) corresponds to degradation of urethane and urea groups in hard segment, while the second loss step (368-520 °C) can be due to the   aggregations in the PU/ATT8 and PU/acid-ATT8, the KH560-ATT dispersion was uniform throughout the PU matrix. The rougher surface and KH560-ATT uniform dispersion in Figure 5d indicates that KH560 surface modification could improve compatibility and interfacial interaction between ATT with the PU matrix. Hence, among the three ATTs, KH560-ATT had the most reinforcement effect on the PU matrix, which demonstrated the mechanical performance results of neat PU and PU/ATT nanocomposites.

Conclusions
PU nanocomposites based on castor oil were prepared by incorporating neat ATT, acid-ATT, and KH560-ATT into the PU matrix. The results showed that the incorporation of 8 wt % of the three ATTs could result in a significant reinforcement on the castor oil-based PU. SEM revealed that KH560 surface modified ATT enhanced the dispersion of KH560-ATT in the PU matrix. DMA confirmed that the 8 wt % KH560-ATT loaded castor oil-based PU nanocomposite exhibits the obvious improvements in the storage modulus at 25 °C (104%) and Tg (5.1 °C). Remarkably, at 8 wt % KH560-ATT loading, the 255% increase in tensile strength, 200% increment in Young's modulus were obtained. Furthermore, the thermal stability of PU/KH560-ATT8 was significantly enhanced As can be seen in Figure 4b, the decomposition processes of PU/ATT nanocomposites were similar with that of the PU. The decomposition of neat PU and PU/ATT nanocomposites showed the typical decomposition steps [53,54]: the first loss step (230-368 • C) corresponds to degradation of urethane and urea groups in hard segment, while the second loss step (368-520 • C) can be due to the decomposition of the soft segment from castor oil. After the incorporation of different ATTs into the castor oil-based PU matrix, both of the decomposition steps of the PU/ATT nanocomposites were slightly subdued compared to that of neat PU. Besides, the T max2 (maximum rate of degradation temperature of second stage) and the char at high temperatures of PU/ATT nanocomposites were higher than that of the neat PU. The results indicated that different ATTs were well dispersed in the PU matrix and resulted in the shift of the T max2 to higher temperature. As shown in Table 3, the char at 600 • C (w char ) of neat castor oil-based PU is 1.1%. The w char of PU/ATT nanocomposites was larger than that of the neat PU, ranging from 6.7% to 8.4% as the 8 wt % different ATTs in the PU matrix. Usually, the thermal stability of polymer and its nanocomposites can be evaluated by the value of IDT and w char . Thus, the introduction of KH560-ATT into the PU matrix could obviously improve the thermal stability of castor oil-based PU nanocomposites, which is attributed to homogeneous KH560-ATT dispersion and strong interfacial interaction between KH560-ATT and the PU matrix.

Morphology
The morphologies and dispersion of the ATTs in the PU matrix were examined with SEM. The fractured micrographs are shown in Figure 5. The increase in roughness after addition of different ATTs into the PU matrix is the result of the deflection of the crack path induced by the presence of the rigid ATT in the path of the growing crack. The bright points were the broken ATTs, illustrating that most of the ATTs had broken apart from the fractured surface. Despite the existence of some aggregations in the PU/ATT8 and PU/acid-ATT8, the KH560-ATT dispersion was uniform throughout the PU matrix. The rougher surface and KH560-ATT uniform dispersion in Figure 5d indicates that KH560 surface modification could improve compatibility and interfacial interaction between ATT with the PU matrix. Hence, among the three ATTs, KH560-ATT had the most reinforcement effect on the PU matrix, which demonstrated the mechanical performance results of neat PU and PU/ATT nanocomposites.