Carbon nanotubes (CNTs) and ammonium polyphosphate (APP) was used to improve the flame retardancy of linear low-density polyethylene/nylon-6 (LLDPE/PA6) blends. It was observed that APP or CNTs tended to be dispersed in the PA6 phase of the blends when all components were melt-blended together. CNTs dispersed in the PA6 phase caused the decrease of flame retardancy. Different processing methods were used to tailor the localization of APP and CNTs in the blends. The results showed that the localization of CNTs or APP strongly influenced the flame retardancy of blends. APP-incorporated CNTs had antagonism in blends with APP localized in the LLDPE phase and CNTs in the PA6 or LLDPE phases. A synergism between APP and CNTs was exhibited only in blend with the localization of APP in the PA6 phase and CNTs in the LLDPE phase. SEM observation showed that the residual char layer in blends with poor flame retardancy was either discontinuous or continuous but porous. A continuous and compact-residue char layer was observed in blends with excellent flame retardancy. Different morphologies of the residual char layer could be attributed to the difference of residual char mass and network structure.
The application of carbon nanotubes (CNTs) in flame-retarded polymers was first reported by Kashiwagi et al. in 2002, attracting significant recent research interest because it affords the enhancement of thermal stability and reduction in heat release rate in a polymer matrix through the use of low loadings [
The addition of CNTs into polymers primarily decreases their peak heat release rate (PHRR), which is a flame-retardant property obtained from a typical cone calorimeter test. CNTs did not exhibit consistent enhancement of flame retardancy when the limiting oxygen index (LOI) and standard UL 94 tests were used to evaluate the flame retardancy [
Intumescent flame-retardant additives (IFR), with characters of nonhalogen and low poison, have become a hot point in a flame retardant area [
Polymer blends are typically comprised of a two-phase structure. CNTs or IFR may localize in one of the two phases in polymer blends. Although the effect of IFR and CNT localization on blend flame retardancy has not been reported, some of the studies found that clay or IFR localization in polymer blends is critical in improving flame retardancy. Lu et al. [
The results that suitable localization of clay or IFR in polymer blends can improve their synergistic effects indicated that tailoring the localization of CNTs and IFR may be also used for optimizing their flame retardancy in polymer blends. Therefore, CNTs and ammonium polyphosphate (APP) were employed to improve the flame retardancy of LLDPE/PA6. The method [
Linear low-density polyethylene (LL6201XR) was supplied by ExxonMobil Corp. Polyamide-6 (PA6) (33500) was supplied by Xinhui Meida-DSM Nylon Chips Co., Ltd. Ammonium polyphosphate ((NH4PO3)
A corotating twin-screw extruder was used to melt-blend the blends, and an injection-molding machine was employed to shape the blends. Before processing, the components and blends were dried at 80°C for 24 h to remove any moisture. We followed the methods [
Four processing methods for controlling APP and CNT distribution in LLDPE/PA6 blends. Method 1: all the components were melt-blended at 230°C. The processing temperature was higher than the melt temperature of LLDPE and PA6. Method 2: the blends were prepared through two steps. LLDPE/PA6 blend was prepared at 230°C firstly, and then LLDPE/PA6 blend was blended with APP and CNTs at 160°C. In the second step, the processing temperature was lower than the melt temperature of PA6. Method 3: the blends were prepared through two steps. LLDPE/PA6/CNT blend was prepared at 230°C firstly, and then LLDPE/PA6/CNT blend was blended with APP at 160°C. In the second step, the processing temperature was lower than the melt temperature of PA6. Method 4: the blends were prepared through two steps. LLDPE/PA6/APP blend was prepared at 230°C firstly, and then LLDPE/PA6/APP blend was blended with CNTs at 160°C. In the second step, the processing temperature was lower than the melt temperature of PA6.
Formulation of blends.
Sample code | LLDPE | PA6 | APP | CNTs | Preparing method |
---|---|---|---|---|---|
Blends | 80 | 20 | 0 | 0 | |
Blends/CNTs | 80 | 20 | 0 | 0.5 | |
D1 | 60 | 15 | 25 | 0 | Method 1 |
M1 | 60 | 15 | 25 | 0.5 | Method 1 |
M2 | 60 | 15 | 25 | 0.5 | Method 2 |
M3 | 60 | 15 | 25 | 0.5 | Method 3 |
M4 | 60 | 15 | 25 | 0.5 | Method 4 |
The values of the LOI were tested in accordance with the standard oxygen index test (ASTM D2863-77). The vertical burning grade was evaluated in accordance with the UL 94 test (ASTM D635-77).
A cone calorimeter (Fire Testing Technology Ltd, United Kingdom) was employed to test the flammability of samples (
The blend of LLDPE/PA6/CNTs (80/20/0.5) was cryogenically cut at -120°C to obtain the slice with the thickness of 70 nm. The slice was examined by TEM (FEI TECNAI-G20) to observe the dispersion of CNTs at an acceleration voltage of 300 kV.
The thermal degradation of the samples was characterized by thermogravimetric analysis (NETZSCH, STA409PC). The test was implemented using a heating rate of 20°C/min in nitrogen.
The morphology of the fractured specimen surface and char residue was observed by SEM (JEOL 6301F). The samples were immersed in liquid nitrogen for 2 h and then broken quickly to obtain the fractured specimen. Char residue was obtained from the samples after vertical flammability tests. The samples were coated with a gold layer before examination. The elemental compositions of fractured specimen surface were measured using an energy-dispersive spectrometry (EDS) analyzer.
The rheological performance of the samples was measured using an ARES rheometer (AR 1500ex) in dynamic mode, on a parallel-plate geometry with a diameter of 25 mm and a gap of ~1 mm. The measurement temperature was 225°C and frequencies ranged from 100 to 0.01 rad/s.
The blend of LLDPE/PA6/CNTs (80/20/0.5) was melt-blended at 230°C, and the localization of CNTs in the blend was investigated by TEM (Figure
TEM of (a) LLDPE/PA6/CNTs prepared above the melt temperature of LLDPE and PA6 and (b) enlarged view.
Samples of LLDPE/PA6/APP (60/15/25) were prepared by two processing methods, in which LLDPE/PA6 blends were melt-blended with CNTs at 230°C or 160°C, respectively. The samples were characterized by SEM-EDS to investigate the dispersion of APP in the blends. The SEM-EDS results are shown in Figure
SEM-EDS of LLDPE/PA6/APP prepared by different methods. (a) APP blended with LLDPE/PA6 at 160°C, (b) APP blended with LLDPE/PA6 at 230°C.
The melt temperature of PA6 is about 220°C. Therefore, the PA6 phase was melted when APP was processed with LLDPE/PA6 at 230°C. Correspondingly, the PA6 phase did not melt when APP was processed with LLDPE/PA6 at 160°C. Therefore, SEM-EDS results indicated that APP was coated by the melted PA6 phase when APP were melt-blended with LLDPE/PA6 at 230°C. While APP were melt-blended with LLDPE/PA6 at 160°C, APP was obligated to localize at the LLDPE phase due to that the PA6 phase was not melted.
The spontaneous dispersion of APP or CNTs in the PA6 phase of the LLDPE/PA6 blends prepared at 230°C indicated that the localization of APP and CNTs in the blends can be tailored by the methods employed in this paper. The blends with the localization of APP and CNTs in the PA6 phase can be obtained when all the components were melt-blended at 230°C (method 1). For method 2, when LLDPE/PA6 blend was blended with APP and CNTs at 160°C, APP and CNTs were obligated to localize at the LLDPE phase due to that the PA6 phase was not melted. For method 3, CNTs were localized in the PA6 phase due to that CNTs were melt-blended with LLDPE and PA6 at 230°C in the first step. Meanwhile, APP was obligated to localize in the LLDPE phase because APP was melt-blended with LLDPE/PA6/CNTs at 160°C in the second step. In the method 4, LLDPE/PA6/APP blend was prepared at 230°C firstly, and then LLDPE/PA6/APP blend was blended with CNTs at 160°C. Therefore, the blends with the dispersion of APP in the PA6 phase and CNTs in the LLDPE phase can be prepared.
The flame retardancy of the blends was investigated by LOI and horizontal burning rating (UL 94) tests, as shown in Table
Flammability characteristics of blends.
Samples | Flammability | ||
---|---|---|---|
LOI (%) | UL 94 rating | ||
Blends | 19.8 | No rating | |
Blends/CNTs | 20.4 | No rating | |
D1 | 25.5 | No rating | |
M1 | 24.8 | No rating | -1.3 |
M2 | 25.5 | No rating | -0.6 |
M3 | 24.8 | No rating | -1.3 |
M4 | 28.3 | V-1 | 2.2 |
The flame retardancy of blends with different CNTs and APP localization is also shown in Table
The comparison of
The
An intumescent flame-retardant primarily plays a role in the condensed phase by forming porous carbonaceous char, which helps to provide effective fire retardation [
The photos of the aspect of the crust of blends after a cone calorimeter test.
For blends with continuous residual char, the flame-retardant properties of M1 and D1 were lower than that of M4. In order to understand this, SEM was used to investigate the microstructure of the residual char, as shown in Figure
SEM of intumescent char residue for LLDPE/PA6/APP and LLDPE/PA6/APP/CNTs after vertical flammability tests.
TGA and cone calorimetry were used to investigate the residual char mass of LLDPE/PA6/APP/CNT blends. The residual char mass curves from TGA and cone calorimetry are corresponding to oxygen-free thermal degradation and oxygen combustion, respectively. The curves of residual char mass obtained from the cone calorimeter test are shown in Figure
Residual mass curves of LLDPE/PA6/APP and LLDPE/PA6/APP/CNTs prepared by different processing methods.
The TGA curves for LLDPE/PA6/APP and LLDPE/PA6/APP/CNTs are shown in Figure
TGA curves of LLDPE/PA6/APP and LLDPE/PA6/APP/CNTs prepared by different processing methods.
The second step began at approximately 360°C, and degradation curves of different blends were overlapped, indicating similar thermal degradation process. As above, the thermal stability shows that the APP and CNT dispersion in blends had a little influence on the degradation of the LLDPE and PA6 at this step. However, APP and CNT dispersion in blends did have a major impact on char residue mass. The char residue mass in blends with APP dispersed in the PA6 phase (M1, M4) was higher than the blends with APP dispersed in the LLDPE phase (M2, M3). The comparison between M1 and M3 or between M2 and M4 shows that the CNT localization in the LLDPE phase is more advantageous to increasing residual char mass than in the PA6 phase.
The results of TGA and cone calorimetry show that APP localization in the PA6 phase or CNTs in the LLDPE phase are both beneficial in promoting charring. Compared to blends with APP dispersed in the PA6 phase, a lower residual char mass was exhibited in blends with APP dispersed in the LLDPE phase regardless of where the CNTs were localized. These results indicate that APP localization in the PA6 phase was more conductive to improve residual char mass than the localization of CNTs in the LLDPE phase. The reaction of APP and PA6 produced the intumescent char. For blends with APP localized in the PA6 phase, APP and PA6 can contact each other, benefitting their reaction and causing the improvement of char residue mass. For blends with APP dispersed in the LLDPE phase, APP and PA6 were insulated from contact by the LLDPE. The isolation between APP and PA6 in blends prevented their interaction, causing a low char residue mass.
CNTs can form a network structure to enhance the thermal stability of polymer matrix, causing the reduction of heat release rate and the improvement of flame retardancy [
The relationship between storage modulus (
Storage modulus (
The results of Figure
The effect of CNT and APP localization in the LLDPE/PA6/APP/CNT blends on complex viscosities (
Complex viscosity (
High viscosity of blends can hinder the penetration of O2 and combustible gas [
The cone calorimeter was employed to investigate the combustion process of LLDPE/PA6/APP and LLDPE/PA6/APP/CNTs, as shown in Figure
Heat release rate curves (a) and smoke production rate curves (b) of LLDPE/PA6/APP and LLDPE/PA6/APP/CNTs prepared by different processing methods.
Cone calorimeter data of blends.
Sample code | Peak HRR (kW/m2) | THR (MJ/m2) | MAHRE (kW/m2) | Time to ignition (s) |
---|---|---|---|---|
D1 | 222.15 | 114.78 | 132.37 | 84 |
M1 | 242.00 | 110.92 | 144.65 | 116 |
M2 | 245.29 | 111.99 | 160.15 | 117 |
M3 | 284.77 | 122.70 | 182.32 | 128 |
M4 | 206.80 | 116.26 | 123.97 | 158 |
In Figure
For the LLDPE/PA6/APP blend (D1), dispersed PA6 particles acted as “flame-retardant particles” because of the flame-retardant properties of dispersed APP in the PA6 phase. In the first combustion process, the decomposition of APP and interaction of APP and PA6 transformed “flame-retardant particles” into charred particles. NH3 produced by the decomposition of APP expanded charred particles until they joined each other to form a continuous residual char. With the development of the combustion process (the second process), the accumulation of pyrolysis gas in the interior of the material caused the rupture of the superficial intumescent char, resulting in the formation of residual char with abundant voids, as shown in Figure
When APP and CNTs were both localized in the PA6 phase (M1), the first or second PHRR and SPR values were higher or lower than that of LLDPE/PA6/APP, respectively. In this blend, dispersed PA6 particles also acted as “flame-retardant particles” due to their localization. CNTs sharply increased the melt viscosity of the PA6 particles due to the formation of a network structure in the PA6 phase. In the first combustion process, “flame-retardant particles” were transformed into charred particles. However, high melt viscosity of the PA6 phase restrained the coalescence of intumescent charred particles to form a continuous char layer, resulting in a high PHRR value in the first combustion process. The morphology of spherical cokes formed in the residual char also indicates the inhibition of CNTs in aggregating into charred particles. Spherical cokes cannot provide effective protection for the matrix during the combustion, resulting in poor flame retardancy exhibited in this blend.
Good flame retardancy was exhibited in the blend with APP localized in the PA6 phase and with CNTs in the LLDPE phase (M4). The results attributed the network structure of CNTs formed in the continuous LLDPE phase and the localization of APP in the PA6 phase. In the initial combustion process, interaction of APP and PA6 formed charred particles. Network structure increased the melt viscosity of the LLDPE and delayed the aggregation of charred particles, resulting in its initial PHRR values being higher than that of LLDPE/PA6/APP in its initial stage. As combustion developed, CNTs would perform three beneficial roles. The localization of the CNTs in LLDPE improved the residual char mass, causing a thick char layer. Meanwhile, the network structure of the CNTs reinforced that char layer. The high strength of the intumescent char layer was difficult to rupture during combustion, causing the formation of a tight and compact intumescent char layer. Moreover, CNTs can enhance the antidripping property due to that CNTs improved melt viscosity. As a result, the synergistic effect between CNTs and APP was exhibited.
For blends where APP was dispersed in the LLDPE phase (M2 and M3), the first or second PHRR values were both higher than that of LLDPE/PA6/APP, indicating an antagonistic effect between the CNTs and APP with regard to flame retardancy. The results of TGA and the cone calorimeter tests both indicate that the localization of APP in blends played a key role in the continuity of the residual char layer. The separation of APP and PA6 did not benefit their interaction, causing low residual char and the formation of a discontinuous char layer. Clearly, this kind of char layer provided poor protection for the substrate, causing poor flame retardancy.
The influence of localization of CNTs and APP in LLDPE/PA6/APP/CNT blends on flame retardancy was investigated. The different localization of CNTs and APP in blends formed different morphologies of a residual char layer, causing remarkable differences in flame retardancy. For blends where APP was localized in the LLDPE phase and CNTs were localized in the PA6 or LLDPE phases, APP localization in the LLDPE phase contributed to a decrease in the interaction of APP and PA6, causing a low residual char mass and a discontinuous char layer to appear, resulting in poor flame retardancy. For blends with APP and CNT localization in the PA6 phase, even though the APP localization in the PA6 phase promoted the formation of a continuous residual char layer, the CNT network structure in the PA6 phase restrained the swell and coalescence of the char layer, causing a continuous residual char layer to appear that consisted of microscopic spherical cokes. As a result, poor flame retardancy was exhibited in this blend. For blends where APP and CNTs were localized in the LLDPE phase, the localization of APP in the PA6 phase benefitted the formation and appearance of a continuous char layer. In addition to reinforcing the residual char layer, the CNT network structure in the LLDPE phase showed poor influence on the swell and coalescence of the char layer. As a result, a tight and compact intumescent char layer was formed, causing good flame retardancy.
The data used to support the findings of this study are available from the corresponding author upon request.
The authors declare that there is no conflict of interest regarding the publication of this paper.
The authors gratefully acknowledge the financial support of this work by the National Natural Science Foundation of China (Contract Number: 51673059), the Natural Science Foundation of Education Department of Henan Province (Contract Number: 17A150009), the Natural Science Foundation of Henan Province (Contract Number: 112300410208), and the Project National United Engineering Laboratory for Advanced Bearing Tribology of Henan University of Science and Technology (Contract Number: 201813).