Introduction

Phthalocyanine (Pc) polymers are a unique class of high temperature thermoset materials, which display good mechanical properties, superior flame and abrasion resistances, outstanding hydrolytic and oxidative stabilities, high char yield and chemical inertness, excellent thermal and thermo-oxidative stabilities [1–8]. Pc monomers can be easily melted and readily processed into shaped forms in the presence of curing additive to get high temperature Pc polymers [8]. Thus they are good polymer matrices which can be used to prepare advanced composites. Such composites have been showing a variety of potential uses for adhesive [1], electronic [2–4], and structural applications [5–7].

Carbon nanotubes (CNTs), due to their high strength (100 times stronger than steel) and modulus (about 1 TPa), high thermal conductivity (about twice as high as diamond), excellent electrical conductivity (1000 times higher than copper), and thermal stability (2800 °C in vacuum) [9–13], have been considered as one of the most promising fillers for preparing polymer nanocomposites with multifunctional properties. However, large-scale use of CNTs is hindered due to their poor dispersion states in matrices because of Van Der Walls and π–π interactions [14, 15]. One of challenges in the preparation of CNTs-based nanocomposites lies in CNTs’ uniform dispersion. At present, two different approaches are usually used to achieve good dispersion of CNTs: chemical methods and physical methods. Chemical methods, on the one hand, use surfactants [16] or functionalization [17] to change CNT surface energy, reduce the tendency to agglomerate in the matrices and thus lead to large improvements in the physical properties of the resultant nanocomposites. On the other hand, defects are preferentially observed at the open ends of CNTs in these methods [18]. However, physical dispersion methods such as sonication, ball milling, grinding and high speed shearing, can maintain CNTs’ nature merits and possess the potential large-scale industrial production. Among them, melt-mixing is most economically and environmentally viable since it does not require solvents, but low shear force can not provide suitable dispersion of CNTs [19, 20]. To achieve the target of uniform dispersion, secondary blend was used to enable uniform dispersion of CNT bundles in the polymer matrix through masterbatch dilution [21, 22]. During the first blend of masterbatch, high shear force can provide suitable CNT dispersion because CNT disentanglement is facilitated. More importantly, first blend of masterbatch can promote the compatibility between CNTs and polymer matrices. Thus, masterbatch technique is a simple and effective method to incorporate CNTs into polymer hosts.

To expand various advantages and applications of Pc and CNT materials, in the current work, CNTs were firstly coated by Pc monomer 4,4'-bis(3,4-dicyanophenoxy) biphenyl (BPH) to prepare BPH-treated CNTs: BPH/CNTs masterbatch. Then, the effect of blend method melt-mixing and masterbatch on the properties of Pc/CNTs nanocomposites were investigated. Consequently, Pc/CNTs nanocomposites prepared by masterbatch did show better CNT dispersion states and physical properties. Finally, the effect of CNT loading on the electrical, dielectric, mechanical, thermal properties of Pc/CNTs nanocomposites were investigated and discussed.

Experimental

Materials

Multi-walled carbon nanotubes (CNTs) (diameter: 10–30 nm, length: 3–10 μm, purity: >95 %), synthesized via chemical vapor deposition, was supplied by Chengdu Organic Chemicals CO. LTD. Chinese Academy of Science. Phthalocyanine monomer, 4,4'-bis (3,4-dicyanophenoxy) biphenyl (BPH), was synthesized in our laboratory. The synthetic procedure and characterization were reported previously [6]. 4,4'-Diaminediphenyl sulfone (DDS, as curing agent) was purchased from Sichuan Dongcai chemicals.

Preparation of the BPH/CNTs masterbatch

The preparation process of the BPH/CNTs masterbatch was carried out as follows. First, BPH monomer (190 g) was melted at 250 °C in a beaker with a mechanical stirrer for 10 min. Then, CNTs (10 g) was added slowly but stirred vigorously for 15 min, producing a dark green BPH/CNTs mixture. Then the mixture was allowed to cool to room temperature, producing a dark green solid. Finally, the solid was smashed in a high-speed disintegrator for 5 min; the BPH/CNTs masterbatch powder (CNTs content: 5 wt%) was thus obtained.

Preparation of the Pc/CNTs nanocomposites

For a comparison, Pc/CNTs nanocomposite containing 2 wt% CNT content was prepared by direct melt-mixing. Firstly, BPH monomer (190 g) was melted at 250 °C in a beaker with a mechanical stirrer for 10 min. Then, DDS was added and kept for 5 min. Next, CNTs (10 g) were added slowly but stirred vigorously for 15 min, producing a dark green BPH/CNTs mixture. Finally, the mixture was poured into a preheated mold (250 °C) and cured by a sequential mode of temperature program at 250 °C, 280 °C, 300 °C, 350 °C, and 370 °C for 4 h, respectively. After being cured and cooled to room temperature, Pc/CNTs nanocomposite was finally obtained.

Pc/CNTs nanocomposites with different CNTs (0.5 wt%; 1 wt%; 2 wt%; 3 wt%; 5 wt%) loading were prepared by diluting 5 wt% BPH/CNTs masterbatch through adding weight-measured BPH. And the next procedures were carried out as identically as melt-mixing method described above.

Characterizations

The morphology of the fracture surfaces of the Pc/CNTs nanocomposites was observed with scanning electron microscope (JEOL JSM-5900LV). Resistivity measurements were performed using the conventional four-probe method. Dielectric measurements for the Pc/CNTs nanocomposites were performed by using a dielectric analyzer (DEA 2970, TA Instruments) as reported before [23, 28]. Flexural tests (three-point bending mode, crosshead: 5 mm/min) of the samples were performed with a SANS CMT6104 series desktop electromechanical universal testing machine (Shenzhen, China) at room temperature. The final results were calculated as averages of five specimens for each Pc/CNTs nanocomposite. Thermogravimeric analysis (TGA) was carried out under both nitrogen and air atmosphere at a heating rate of 10 °C/min using TA Q50 series analyzer system combination with data processing station.

Results and discussions

Effect of blending method on the properties of Pc/CNTs nanocomposites

The 2 wt% CNTs-filled Pc nanocomposites were prepared by melt-mixing and masterbatch methods and their physical properties were summarized in Table 1. With the same CNTs content, Pc/CNTs nanocomposites prepared by masterbatch method show better dielectric and mechanical properties than those prepared by direct melt-mixing method. For example, dielectric constants of Pc/CNTs nanocomposites prepared by masterbatch method are 2–4 times larger than those of samples prepared by melt-mixing method. The flexural strength of Pc/CNTs nanocomposites prepared by masterbatch is also 24.06 MPa higher than those of Pc/CNTs nanocomposites prepared by melt-mixing. The difference of dielectric and mechanical properties can reflect the different dispersion and adhesion state of CNTs in those two systems, which is confirmed by their SEM images. It can be seen in Fig. 1(a) and (b) that CNTs in melt-mixing systems aggregate heavily with sword-like bundles or entanglements; and many CNTs are pulled-out with sharp ends. However, CNTs in the masterbatch system exhibit relatively better distributions and most of CNTs are broken rather than pulled out from the matrix. As a result, better dispersion and adhesion of masterbatch contribute more to dielectric and mechanical performances of resulted nanocomposites. It is also interesting to note that Pc/CNTs nanocomposites prepared by direct melt-mixing method show higher thermal stabilities than those by masterbatch method. Wu et al. [9] reported no obvious thermal increments in the polylactide/CNT composites and the authors attributed those to the crosslinking reduction of polycarbonate at the presence of CNTs and the additional components or impurities such as compatibilizer or remaining catalyst from the CNT synthesis. This explanation may be true to BPH/CNTs composites since well-dispersed CNTs may retard the curing degree of BPH matrix itself and CNTs were not chemically treated in this study. Schartel et al. [10] even observed a remarkable decrease in the polycarbonate/CNT composites and the authors thought bare and aggregated CNTs could take up heat and obstruct transport of degradation products more effectively. In this sense, these results also manifest that the dispersion states of CNTs in the masterbatch are better than those in the direct melt-mixing, which agrees well with the SEM observations. Therefore, in the following study, masterbatch method was chosen to prepare Pc/CNTs nanocomposites and their physical properties as a function of CNTs loading were investigated.

Table 1 Electrical, dielectric, mechanical and thermal properties of 2 wt% CNTs-filled Pc nanocomposite prepared by Melt-mixing and Masterbatch methods
Full size table
Fig. 1
figure 1

SEM images of 2 wt% CNTs-filled Pc nanocomposites prepared by a masterbatch method, b melt-mixing method and c enlarged black area of b

Full size image

Electrical and dielectric properties of Pc/CNTs nanocomposites

Figure 2 is a plot of the electrical resistance versus CNT loading of the Pc/CNTs nanocomposites measured at room temperature. The electrical resistance decreases with the increase in the amount of the CNT loading in the nanocomposites. A sharp decrease of the electrical resistance was observed between 0.5 and 1 wt%. When the weight percentage of the CNTs increased to 1 wt%, electrical resistance of the nanocomposite dramatically decreased, being 3 orders of magnitude smaller than that of the pure Pc matrix. This may be ascribed primarily to the large aspect ratio of the conductive CNTs. Similar electrical behavior was also reported in other CNTs-based nanocomposites and this can be attributed to the fact that a percolation network of CNTs was formed within the composite systems [24–27]. In our case, the decrease in the electrical resistance is still not large. This may be due to that thermoset BPH monomer itself can be cured and net-like macromolecules were formed after heat treatment while good thermal conductivities of CNTs may further promote BPH curing process. As a result, CNTs or CNTs network is inevitably separated by the polymer chains, leading to small decrease of electrical resistance values.

Fig. 2
figure 2

Resistivity of Pc/CNTs nanocomposite as a function of CNT content

Full size image

Figure 3 shows the dielectric constant of Pc/CNTs nanocomposite as a function of CNT loading at 100 Hz. It can be seen that dielectric constant of 0.5 wt% CNTs-filled Pc nanocomposite almost shows the similar value as that of Pc matrix. However, incorporation of 1 wt% CNTs significantly increases the dielectric constant, from 3.44 to 14.57, leading to 320 % increment. The dramatic increase at 1 wt% is a critical transition for dielectric constant, which agrees well with the electrical resistance observation. From that point on, dielectric constant of Pc/CNTs nanocomposite increases steadily as CNT content goes. For 5 wt% CNTs-filled Pc nanocomposite, a high dielectric constant of 472.72 is obtained. The increase in the dielectric constant can be attributed to polarization effect among the high aspect ratio CNTs [23]. In this system, many conducting CNTs are isolated by thin dielectric layers; the high dielectric constant values originate in the significant blocking of charge carriers at internal interfaces (namely the MWS effect) under an electric field, because of the formation of a large number of micro-capacitors [25].

Fig. 3
figure 3

Dielectric constant of Pc/CNTs nanocomposite as a function of CNT content at 100 Hz

Full size image

The dielectric transition at 1 wt% CNTs loading can be also confirmed by their frequency dependence. Figure 4 shows the dielectric constant and loss of the Pc/CNTs nanocomposites as a function of frequency for different CNT contents at room temperature. Consequently, dielectric constant increase with the increasing CNT content at the specified frequency. However, the dependence of dielectric constant on the frequency is distinguished with different CNT content. Namely, below the 1 wt% CNTs content, the dielectric constant and loss of Pc/CNTs nanocomposites show little frequency dependence. However, when CNT loading exceeds 1 wt%, the dielectric relaxation becomes evident, which results in a decrease of the dielectric constant. These results are consistent with the observations in Figs. 2 and 3. To conclude, the dramatic electrical and dielectric transition happens when CNT content is about 1 wt%.

Fig. 4
figure 4

Dielectric properties of Pc/CNTs nanocomposite as a function of frequency: a dielectric constant and b dielectric loss.

Full size image

Mechanical and morphological properties of Pc/CNTs nanocomposites

The flexural test can measure the deformation energy or fracture strength that bears the bending moment for the tested materials. Figure 5 shows the effect of CNT content on the flexural properties of Pc/CNTs nanocomposites. It can be seen that the increase in the flexural modulus is very encouraging with small CNT addition. For example, flexural modulus shows an increase from 3852 MPa in the Pc matrix to 3964 MPa in the 0.5 wt% CNTs-filled Pc nanocomposite.

Fig. 5
figure 5

Mechanical properties of Pc/CNT nanocomposite as a function of CNT content: flexural modulus and flexural strength.

Full size image

Figure 5 also depicts the flexural strength of Pc/CNTs nanocomposites. As CNT content increases, the flexural strength of the Pc/CNTs nanocomposite increases firstly, and then becomes stable, and decrease afterwards. It is interesting to note that the maximized flexural strength of 111.4 MPa is obtained in the 1 wt% CNTs-filled Pc nanocomposite, which is increased by 51.4 % in comparison with Pc host (73.7 MPa). And then with higher CNT loading (1–3 wt%), the flexural strength of the nanocomposites slightly decrease. These mechanical improvements can be due to the fact that uniform-dispersed CNTs restrict the mobility of polymer chains of Pc, while the processing method of masterbatch can promote compatibility and adhesion between CNTs and Pc matrices. High strength and aspect ratio of CNTs also contributes to the mechanical improvement. Moreover, a gradual formation of CNT network can further restrict polymer chain and result in the optical mechanical performances. Comprehensively evaluating, Pc/CNTs nanocomposites with 1–3 wt% CNT content showed the best mechanical performances. However, the decrease of flexural strength in the 5 wt% CNTs-filled nanocomposite can be due to the aggregate of CNTs and also the decreased curing degree of Pc in the presence of aggregated CNTs.

Mechanical properties of the nanofiller/polymer system are related to the dispersion, orientation, adhesion, interfacial interaction and stabilization of the nanofillers in/with the polymer matrices [13, 18, 28]. Hence, fracture surface images of the nanocomposite are given to further evaluate their mechanical properties, as is shown in Fig. 6. The surface of neat Pc host is quite smooth, showing a typical brittle fracture (Fig. 6a). However, incorporation of CNTs significantly changes the surface of Pc (Fig. 6b-c), increasing the surface area on the fracture surface. From the Fig. 6b-c, it can be also observed CNTs are uniformly dispersed in both 1 and 3 wt% CNTs-filled Pc nanocomposites. In addition, pull-out phenomena can been observed in the nanocomposites, but most of CNTs are well embedded in the Pc matrix, indicating that CNTs have good adhesion to Pc host due to merits of the masterbatch processing method. These results confirmed the enhancement of the mechanical performances investigated, especially for 1–3 wt% CNTs-filled Pc nanocomposites.

Fig. 6
figure 6

SEM images of fractured surface of the Pc/CNTs nanocomposites containing: a 0 wt% CNT (neat Pc matrix), b 1 wt% CNT, and c 3 wt% CNT

Full size image
Fig. 7
figure 7

Thermal stabilities of Pc/CNTs nanocomposites at different atmosphere: a nitrogen and b air

Full size image

Thermal stabilities of Pc/CNTs nanocomposites

To show the effect of CNTs on the thermal properties of Pc composites, the main TGA results (Fig. 7) performed in both nitrogen and air atmosphere were summarized in Table 2, in which the initial decomposition temperatures at weight loss of 5 %, 10 %, 30 % (T d5% , T d10% , T d30% ) and maximal decomposition temperature (T max ) were displayed. For the T d5% , T d10% , T d30% and T max , they are all higher than 535 °C, indicating the excellent thermal and thermo-oxidative stabilities of Pc/CNTs nanocomposites. Compared with Pc matrix, however, no improvement in thermal and thermo-oxidative stabilities was observed in the composites. Wu et al. [9] reported no obvious thermal increments in the polylactide/CNTs nanocomposites and Schartel et al. [10, 29] also observed a remarkable decrease in the polycarbonate/CNTs composites.

Table 2 Thermal stabilities of Pc/CNTs nanocomposites under nitrogen and air atmosphere
Full size table

Conclusions

To prepare good Pc/CNTs nanocomposites, the effect of different blending method melt-mixing and masterbatch were firstly studied. Consequently, CNTs in melt-mixing systems aggregate heavily with sword-like bundles or entanglements; and many CNTs are pulled-out with sharp ends. Masterbatch nanocomposites had good CNT dispersion state and more obviously enhanced physical properties. Then, Pc/CNTs nanocomposites containing different CNT content were prepared by masterbatch route; and the feasibility of using Pc/CNTs nanocomposites was investigated by evaluating their electrical, dielectric, mechanical, morphological and thermal properties as a function of CNT loading. Consequently, the dramatic electrical and dielectric transition happened when CNTs was about 1 wt%. For the 1 wt% CNTs-filled Pc nanocomposites, a 51.4 % increase in flexural strength was obtained and flexural modulus was also improved from 3851.7 MPa to 3973.3 MPa. These improvements were mainly due to the fact that masterbatch method can promote good dispersion and strong adhesion of CNTs in/to the Pc host. All Pc/CNTs nanocomposites showed high thermal and thermo-oxidative stabilities up to 535 °C. Pc/CNTs nanocomposites with multifunctional properties can find uses under some critical circumstances with requirements of high strength and temperatures.

In addition, large-scale production of CNTs using iron–Pc polymers as only carbon source at ambient pressure in nitrogen atmosphere has been reported in our lab [2]. However, the mechanism remained unclear. This work investigates the presence of CNTs on the properties of Pc polymers and may provide us some useful information on discovering the mechanism behind the production of CNTs from iron–Pc polymer.