Preparation of highly exfoliated epoxy/clay nanocomposites by clay grafted with liquid crystalline epoxy

ABSTRACT Epoxy/clay nanocomposites with a high degree of exfoliation were achieved by intercalating liquid crystalline epoxy into clay intragallery as well as using a so-called ‘solution compounding’ process. In this process, clay modified was first treated with trichloromethane to form organoclay-trichloromethane suspension followed by liquid crystalline epoxy modification. The liquid crystalline epoxy grafted clay was then mixed extensively with epoxy to form epoxy/nanoclay composites. The mechanism of exfoliation was explored by monitoring the change of morphology of organoclay during each stage of processing with X-ray diffraction (XRD). The liquid crystalline epoxy grafted clay synthesised was characterised by fourier transform infrared spectroscopy (FT-IR) and polarising optical microscopy (POM). The clay platelets uniformly dispersed and highly exfoliated in the whole epoxy matrix were observed using transmission electron microscopy (TEM) and FT-IR imaging system. The epoxy nanocomposites were fabricated by incorporating different liquid crystalline epoxy grafted clay loading. The results revealed that the incorporation of liquid crystalline epoxy grafted clay resulted in a significant improvement in glass transition temperature (Tg) derived from dynamic mechanical analysis (DMA) and thermal stability measured by thermogravimetric analysis (TGA). GRAPHICAL ABSTRACT


Introduction
Epoxies are among the best polymeric materials being used in a variety of applications, such as aviation industry, electronic encapsulates, in composites in the transportation and in structural applications as the matrix materials of fibre-reinforced composites. However, they have some drawbacks on account of the highly cross-linked structure which makes the epoxy resin's brittleness and poor resistance to crack propagation have limited their industrial applications and the limits of improving properties made of traditional micrometre-scale fillers have been reached. Since epoxy/clay nanocomposites have attracted extensive attention due to the superior properties of these socalled nanocomposites as compared to those of conventional epoxy-inorganic composites in mechanical properties (modulus, strength, fracture toughness and surface hardness), barrier property, flammability resistance, thermal stability and resistance to solvent swelling [1][2][3]. These enhancements are all achieved with only a few weight percent of nanoparticles and without sacrificing other properties. Therefore, cost, clarity and processability of epoxy resin are retained. This is in obvious contrast to conventional polymer fillers, which require high concentrations and thus induce processing difficulties, to provide a fraction of these enhancements [4][5][6].
From a structural point of view, intercalation and exfoliation are used to describe two general terms of clay structure that can be prepared [7]. Intercalated nanocomposites are formed when the clay is well-dispersed in a polymer matrix with polymer chains inserted into clay intragallery with fixed interlayer spacing. Exfoliated ones consist of highly dispersed clay platelets throughout the polymer, especially disordered and individually dispersed in the polymer matrix. In improving the physical property of nanocomposites, the latter case is known to be especially desirable. This can be assumed that structural distinction that delaminated clay layers have huge aspect rations and greater phase homogeneity than intercalated nanocomposites [8][9][10][11][12][13]. In order to achieve epoxy/clay nanocomposites with highly exfoliated structure, several efforts from both academic institutes and industrial sectors have explored different methods to facilitate clay exfoliation, including the cation exchange capacity (CEC) of clay, the property of clay modifier, the kinds of curing agent, the curing conditions and the processing methods, which exhibit certain impact on the synthesis and final structures of nanocomposites [14][15][16][17][18][19][20][21][22][23][24][25][26][27][28]. Nonetheless, most of these trials reported in the published literature have confirmed that the perfect nanocomposites structure is obtained rarely by these approaches.
Liquid crystalline epoxy resin is thermoset liquid crystalline polymer, which combines of thermoset and liquid crystal formation capability. Liquid crystalline thermoset (LCT) can be achieved by incorporating well-organised liquid crystal structures into the network during the curing process and the orientational order can be fixed in the final network. In comparison with conventional epoxies, liquid crystalline thermoset reveal a mass of enhanced properties, such as anisotropic orientation, high modulus and tensile strength and low coefficient of thermal expansion, because of the formation of many LC domains in the cured networks [29][30][31][32]. In addition, according to our own recent work [33], self-oriented alignment of liquid crystalline epoxy resin molecules resulted from rigidity of mesogen component will give rise to faster reaction rate than non-mesomorphic epoxy resin molecules during curing process.
In order to complete clay exfoliation in present work, we consider both dispersion of the clay in the epoxy matrix before curing reaction and polymerisation reaction of clay intragallery which is another important factor in the preparation of the nanocomposite [14,16]. In this study, dry organoclay powder modified by p-aminobenzoic acid was added to trichloromethane, which has excellent solubility for liquid crystalline epoxy, and the ultrasonic bath was employed after blending to improve dispersion. After that, liquid crystalline epoxy was poured into the system to synthesise liquid crystalline grafted clay. Finally, liquid crystalline epoxy as the surfactant for modification of clay was mixed with epoxy prepolymer with 'solution compounding'. During the curing process with the reasonable curing condition, liquid crystalline epoxy resin inside the clay galleries had much faster reaction rate than non-mesomorphic epoxy resin molecules. Consequently, it would seem to attempt properly to facilitate the reaction in the fabrication procedure for epoxy-clay nanocomposites based on the idea of increasing the rate of the intragallery reaction relative to the extragallery reaction. Furthermore, liquid crystalline with single epoxide group was incorporated into clay galleries and the clay was applicable in the contrastive analysis to prove that the influence of the intragallery reaction rate on exfoliated structure.. Meanwhile, mechanism of clay exfoliation, which is very different from the conventional ones was declared [19,34,35], and the thermal properties of the nanocomposites were also presented.

Instruments
Fourier transform infrared spectra (FT-IR) was measured on a Nicollet 510 FT-IR spectrometer (Nicollet Instruments, Madison) using samples in KBr pellets; Nuclear Magnetic Resonance (NMR) was tested by Varian WH-90PFT NMR Spectrometer (Varian Associates, Palo Alto, CA) with CDCl 3 as a solvent and tetramethylsilane (TMS) as an internal standard. The texture of 4-(Oxiran-2-ylmethoxy)-benzoic acid cholesterol ester and 4-(4-Oxiranyl-methoxyl)-benzoic acid cholesterol ester and clay modified by liquid crystalline epoxy was observed with a polarised optical microscopy (Leica DMRX Germany) equipped with a Linkam THMSE-600 (Linkam, Surrey, England) hot stage connected to a TP-92 temperature control unit at a heating and cooling rate of 3°C/min. X-ray diffraction measurements were performed with a nickel-filtered Cu-Ka radiation with a DMAX-3A Rigaku powder diffractometer. Sample was placed on the reflectivity stage and the recorded diffraction patterns ranged from 2°to 10°. The exfoliation state of clay in nanocomposites was observed with a transmission electron microscope. The sample specimens were cut using a Leica Ultracut UCT ultra-microtome. Microtomed thin sections were collected on 200 mesh copper grids and examined by a Japan JEOL JEM-1011 microscope operating at 200 kV in bright field mode. Distribution of clay in composites was characterised by using FT-IR imaging (Spotlight 300 infrared imaging system, PerkinElmer). Film sample was located in the measure stage and scanned infrared ray. Thermal properties were measured with a dynamic mechanical analysis (DMA) (NETZSCH DMA 200) and a three-point bending configuration at a heating rate of 3°C/min. Thermogravimetric analysis (TGA) were carried out on a NETZSCH TGA 209C system under a N 2 atmosphere at a heating rate of 5°C/min up to 600°C.

The synthesis of DOAC and SOAC
The synthesis routes of 4-(Oxiran-2-ylmethoxy)-benzoic acid cholesterol ester (SOAC) and 4-(4-Oxiranylmethoxyl)-benzoic acid cholesterol ester (DOAC) was presented in Figure 1. The synthesis procedure of SOAC and DOAC was described in Supporting Information. The NMR and polarised optical textures of SOAC and DOAC were shown in Supporting Information Figures S1-S4.

Preparation of organophilic clay
20 g pristine clay was dispersed into 3000 mL of distilled water to form a suspension. Protonated form of p-aminobenzoic acid was prepared by slowly dropping 200 ml 0.2 N HCl solution (40 mmol) into 5.48 g (40 mmol) p-hydroxybenzonic acid in 100 ml ethanol under a magnetic stirring. The solution was then dropped into the clay/water suspension and stirred vigorously for 8 h at 80°C. A white precipitate formed and was isolated by filtration and washed several times with distilled water/ethanol until no chloride was detected in the filtration by one drop of 0.1 N AgNO 3 solution. The modified clay (M-clay) was then dried under vacuum oven at 50°C for 24 h, ground with a mortar and a pestle, and the <44 μm fraction was collected. The cationic exchange ratio can be Kjeldahl method for nitrogen determination and the result was 89.4% in this study. The experimental pattern and details are presented in the part one of supporting information.
2.5. Fabrication of liquid crystalline grafted clay and epoxy/clay nanocomposites 10 g M-clay was dispersed in 400 mL CHCl 3 and ultrasoned in a 100 W bath sonicator for 1 h and 28.9 g DOAC (50 mmol) was also dissolved in 80 mL CHCl 3 . Excess of DOAC was added to guarantee reaction between single epoxide group and carboxyl group of M-clay. The two systems were then mixed together and sonicated for another hour. Tetrabutylammoniumbromide (5 wt%) was added into the solution as catalyst before the solution was refluxed at 80°C for 6 h. The liquid crystalline-functionalised clay (DOACM-clay) was washed with CHCl 3 three times, incanus precipitate was obtained and then the wet product was added to 80 mL of CHCl 3 to form a clay/CHCl 3 slurry. The CHCl 3 / DOACM-clay slurry was directly added into desired amount of epoxy. Afterward, CHCl 3 was evaporated by rotary evaporator at 50°C for 2 h and then a stoichiometric amount of PDA (30 g/100 g of epoxy resin) of the curing agent was added, and the mixture was stirred and degassed under vacuum at 70°C for 60 min. Finally the mixture was cured at 140°C for 2 h and postcured at 180°C for 5 h.
The synthesis method of SOACM-clay and nanocomposites were similar to the one be used for DOACM-clay and DOACM-clay/epoxy nanocomposites.

Process of clay exfoliation
It is well known that the pristine clay is hydrophilic silicate and can suspend steadily in water in which clay is dispersed as isolated sheets or small domains consisting of a few sheets. Therefore, the energy of organic/inorganic interface between clay platelets and the polymer matrix should be reduced from the modification of clay surfaces with the hydrophobic molecules using a cationic exchange process so that various dispersion morphologies may be achieved in the polymer-based composites [36]. Moreover, organophilic functionalised surfactants further imparts functional groups to graft liquid crystalline. In our work, p-aminobenzoic acid, containing amine groups and carboxyl group, was chosen as the surfactant for modification of sodium clay. The carboxyl group can be further react with liquid crystalline epoxy to fabricate liquid crystalline grafted clay. A schematic illustration of the grafting reaction and preparation process of nanocomposites is presented in Figure 2 and the uniform preparation methods were adopted for the synthesis of DOACM-clay and SOACM-clay as well as fabrication of nanocomposites. The purpose for dispersing M-clay in trichloromethane by ultrasoned in initial step is to react more easily between liquid crystalline epoxy and M-clay. Moreover, because trichloromethane has excellent solubility for liquid crystalline epoxy, it would be an ideal way to obtain stable suspension of liquid crystalline grafted clay in which clay intragallery are filled with plenty of solvent ( Figure S5). Another one is to facilitate a better dispersion of modified clay in epoxy prepolymer in the following step, because the latter is soluble in trichloromethane.
XRD is an important measure of the degree of dispersion and exfoliation of clay platelets. XRD analysis was used to monitor change of morphology of clay during each stage of processing. The XRD diagrams of pristine clay, M-clay, liquid crystalline grafted clay in trichloromethane, uncured 5 wt% liquid crystalline modified clay/epoxy mixture, as well as nanocomposites are illustrated in Figure 3. Compared with the pristine clay which exhibits a relative strong peak at about 2θ = 7.5°, representing the interlayer space of about 1.2 nm (Figure 3(a), curve 1), XRD pattern of M-clay (Figure 3(a), curve 2) displays a single and large peak at around 6.5°, corresponding to an interlayer space of 1.3 nm. The increase from 1.2 to 1.3 nm mainly stem from the p-aminobenzoic acid grafted onto the clay layer surface via ion exchange. The XRD curves of liquid crystalline modified clay are presented in part b and c of Figure 3. For DOACM-clay in trichloromethane are presented in Figure 3(b) (curve 1), the change of (0 0 1) diffraction is significant, which indicates that clay has been destroyed. The phenomenon can be assumed that spaces of DOACM-clay layers were swollen by the absorbed trichloromethane. Curve 2 of Figure 3(b) depicts XRD pattern of uncured epoxy/DOACM-clay (loading 5 wt%) mixture.  Trichloromethane filled with clay intragallery is a good solvent for DGEBA and DOAC adopted to enhance compatibility between organoclay and epoxy. When the trichloromethane was removed under vacuum, epoxy resin remained inside the platelets. So the curve 2 of mixture before adding curing agent exhibits a weak, broad peak. However, there are still a small part of platelets with a parallel structure with an enlarged d-spacing. After adding curing agent, DOAC of clay intragallery has much faster reaction with agent at reasonable curing temperature in the liquid crystalline range because of self-oriented alignment of liquid crystalline epoxy resin molecules resulted from rigidity of mesogen component. The conclusion was confirmed in our own recent work [33]. Consequently, before clay extragallery viscous force derived from non-mesomorphic epoxy resin molecules came into existence, the DOACM-clay platelets had been further delaminated in epoxy matrix via the elastic force offered by curing reaction of liquid crystalline epoxy of intragallery. Finally, the evidence that exfoliated epoxy/clay nanocomposites obtained was observed by XRD with absence of (0 0 1) diffraction peak.
As a comparison benchmark, The XRD diagrams of SOACM-clay and epoxy nanocomposites are depicted in Figure 3(c). The results show that the (0 0 1) diffraction peak of curve 1 and 2 are similar to condition of DOACM-clay. However, weak and broad diffraction peak of epoxy nanocomposites (curve 3) can still be observed. The liquid crystalline grafted SOACM-clay have failed to achieve exfoliation structure in the final epoxy matrix was mainly described to cross-link reaction of the liquid crystalline during curing process was not taken place in clay intragallery because of SOACMclay does not contain crosslinkable activity group, based on the reason, elastic driving force, which is responsible for exfoliation, can be not generate in clay intragallery to facilitate exfoliation further. Through the contrastive analysis in the fabrication process, we believe that the liquid crystalline epoxy modified clay, which has accelerated curing relative to ordinary epoxy, is an efficient mean to achieve exfoliated structure nanocomposites.
The evidence for the reaction between modifier and clay platelets comes from FTIR analysis which can be seen in Figure 4 Figure 4, the spectrum of DOACM-clay displays almost the same pattern as that of pristine clay in the range of 1300 cm −1 and 500 cm −1 , indicating that the main structure of the silicate layer was unchanged. In addition, the appearance of the absorption in the region of 3096-3168 cm −1 assigned to the ionic bonded N-H stretching vibration of N + H 3 indicated the evidence that the interlayer Na + was replaced by protonated p-aminobenzoic acid by cationic exchange. Besides, the appearance of the absorptions at 1722 cm −1 for C=O stretching and that a new band of -CH stretching at 2852 cm −1 , which confirms the grafting of liquid crystalline on the clay surface.
The phase behaviour of the liquid crystalline grafted clay sample is shown in Figure 5. The similar texture of DOACM-clay and SOACM-clay was found to appear respectively from 134°C to 186°C and from 131°C to 177°C in the heating process by an optical polarising microscope. It can be seen that the textures observed is not typical pattern of liquid crystalline and the liquid crystalline temperature range is changed relative to liquid crystalline monomer due to clay destroy interaction force between liquid crystalline molecules.

Microstructure of epoxy/clay nanocomposites
TEM is a very useful tool for providing evidence of the detailed features about the structure of clay nanocomposites. TEM morphologies of epoxy nanocomposites with 5 wt% organoclay loading are presented in Figure 6. From the TEM images of the samples with DOACM-clay and SOACM-clay, it is found that the number and average size of agglomerations which are dispersed throughout the bulk of the sample are distinguished significantly. In the case of the epoxy system with DOACM-clay, which are displayed in Figure 6(a) obtained from the low magnification TEM, most of the clay layers are well uniformly and homogeneous disperse in epoxy matrix and no large aggregates, whereas several agglomeration regions can be seen in Figure 6 (b) for the sample with SOACM-clay. Part c of Figure 6 shown here are representative of the sample with DOACM-clay are made at the high magnification TEM. It can be observed clearly that clay platelets have been exfoliated disorderly in epoxy in the form of single layers or very thin tactoids that consist of 2-3 clay layers. On the other hand, in respect of TEM analysis at higher magnification of the nanocomposites with SOACM-clay can be exhibited in Figure 6(d). The clay platelets appear to be somewhat thicker than that of system with DOACM-clay. In addition, it is worth mentioning that these agglomerations observed at the lower magnification are comprised of ordered structure of clay with extended d-spacing. It is supposed from the above discussion that the incorporation of the liquid crystalline epoxy improves significantly the nanostructure as identified by TEM. Through the contrastive analysis, it can be summarised that the curing stage which makes the intragallery polymerisation at a faster rate compared to extragallery polymerisation is the crucial important step except for the state of dispersion and exfoliation produced by the epoxy prepolymer prior to curing reaction.
Spectrum spotlight FT-IR imaging system is an effective way to obtain information about distribution of component in blending system, where we can distinguish between the chemical composition of the visible image using chemimap analysis of FT-IR imaging system. In order to investigate clay distribution in nanocomposites system, the total IR absorption of epoxy/clay nanocomposites was tested by FT-IR imaging transmission mode. Subsequently, the exclusive IR absorption of clay at 1044 cm −1 was selected to represent for clay in the total absorption of nanocomposites and marked as magenta. In regard to epoxy matrix, the distinct absorption at 1580 cm −1 of benzene ring, not observed in clay, was made to represent for epoxy matrix in the total absorption of nanocomposites and marked as black. The FTIR images of epoxy/clay nanocomposites with different content DOACM-clay are displayed in Figure 7. It is clearly seen from part a and b of Figure 7 that the magenta areas represented for clay as dispersion phase are uniformly scattered in the contiguous black areas represented for epoxy matrix. On the other hand, as shown in the FTIR images, reducing the clay content to 1 wt% does not result in a significantly better dispersion of the clay. The results described in FTIR image analysis can  confirm further that DOACM-clay was uniformly and homogeneously dispersed in epoxy matrix equalling to TEM observation at lower magnification.

Thermal phase transition
DMA is an efficient method for providing significant information about the microstructure of thermoset networks, as well as information on molecular motion, relaxation processes and thermal transitions. In this method the loss factor, tanδ, is defined as the ratio of the loss modulus to the storage modulus, which is very sensitive to the glass transition temperature (Tg) of a thermoset network. As the temperature is increased, tanδ reaches a maximum in the transition region because of the initiation of micro-Brownian motion of the molecular chain segments and their stress relaxation [37]. Consequently, the tanδ peak values of the results from the DMA test predicate Tg of the composites. The temperature dependence of tanδ for the neat epoxy and epoxy nanocomposites with various clay contents are depicted in Figure 8. It can be seen that the glass transition temperature of the former increased with the incorporation of DOACM-clay into the epoxy resin. When containing 7 wt% DOACM-clay, the Tg of nanocomposites is about 24°C higher than neat epoxy polymer. According to the previously reported investigation [38,39], the effect of clay on the Tg of epoxy resin depend primarily on the dispersion and extent of exfoliation of clay as well as the phase interfacial action. Therefore, the significant increase in Tg value of nanocomposites reported here can be associated with the high exfoliation of DOACM-clay in epoxy matrix as well as the strong interfacial interaction resulted from liquid crystalline epoxy involved in curing reaction restricted in molecular motion and reduced in free volume or higher degree of cross-linking. Meanwhile, LC domains in nanocomposites derived from orientational order locked in networks during curing reaction can prevent the movement of the segments and thus results in a positive effect on the value of Tg.
In addition, the tanδ of 5 wt% SOACM-clay/epoxy are compared with that of the 5 wt% DOACM-clay/ epoxy in part b of Figure 8. It can be seen that 5 wt% SOACM-clay/epoxy has lower Tg than that 5 wt% SOACM-clay/epoxy. This phenomenon can be explained by the intercalated and sometimes agglomeration structure formed in the nanocomposites system. Therefore, the inhibited effect of clay platelets to epoxy matrix molecules is limited to a certain extent.

Thermal stability
The thermal stability of polymers, which is related to molecular structure and molecular weight, play an important role in evaluating their working temperature limits and environmental conditions. The TGA curves of neat epoxy and nanocomposites with different content DOACM-clay at a heating rate of 5°C/min in N 2 are shown in Figure 9(a). The results indicate that the temperatures at 5% weight loss (decomposition temperatures T 5% ) of neat epoxy is 312°C. While the T 5% of epoxy nanocomposites with 1, 3, 5 and 7 wt% DOACM-clay are 328°C, 340°C, 367°C and 360°C, respectively. The reason for the increased decomposition temperatures of epoxy in the presence of clay is primarily attributed to excellent dispersion of clay in polymer matrix and strong interfacial interaction because clay is well known for its excellent thermal stability to provide a retarding effect on the thermal degradation of the organic component of composites. However, the somewhat decline in epoxy nanocomposites with 7 wt% loading DOACM-clay can be believed that the crosslink density was influenced by clay aggregation derived from the higher clay concentration was dispersed uncomfortably in epoxy matrix due to system viscosity.
The peaks in the first derivatives of neat epoxy and nanocomposites with different content DOACMclay exhibit that the samples possess almost the same temperatures at maximum decomposition rate (T rm ), it is demonstrated that the thermal decomposition behaviour of the samples mainly rests with epoxy matrix. Horowits-Metzger integral kinetic method is applied to further examine the decomposition process [40].
where ΔT is defined as ΔT = T−T rm , E a is activation energy of thermal decomposition, R is universal gas constant, k is the intercept of linear equation and w is residual weight percentage. On the basis of the slope of the plot of ln(−lnw) versus ΔT (Figure 9(b)), the E a of neat epoxy and nanocomposites with different content DOACM-clay are calculated to be 53, 65, 78, 89 and 85 kJ/mol, and the increased E a can be ascribed to the stable thermal stability of clay. Meanwhile, the increased E a values are also fit with the increased decomposition temperature, this phenomenon indicates that the evenly dispersed clay can enhance the thermal stability of epoxy.

Conclusion
The liquid crystalline epoxy grafted clay was employed to develop highly and disorderly exfoliated epoxy/clay nanocomposites using 'solution compounding' process, which involved a new exfoliation mechanism via XRD tracing test. By comparison, the crosslinkable liquid crystalline epoxy incorporated intragallery platelets played a crucial role in facilitating exfoliation of clay due to accelerated curing reaction resulted from selforiented alignment of liquid crystalline epoxy resin molecules relative to extragallery epoxy. The high exfoliation and well-dispersion of the clay in epoxy matrix, which leads to significant enhancement in Tg and excellent thermal property of the resulting nanocomposites, was confirmed by TEM and FT-IR images system.