Introduction

Cyclic olefin copolymer (COC), a significant engineering plastic, is synthesized from cyclic olefin and α-olefin. Predominantly consisting of ethylene/norbornene (E/NB) copolymers, COCs are distinguished from conventional linear polyolefins by their superior optical clarity and enhanced transparency. These materials are celebrated for their elevated glass transition temperature (Tg), exceptional resistance to moisture, chemical inertness, biocompatibility, and robust dielectric properties, which have garnered widespread applications in the realms of optical devices, capacitor films, medical applications, and packaging [1, 2]. Notably, COC's performance parameters, particularly its Tg and thermal deformation temperature, are contingent upon the norbornene (NB) content, with higher NB content correlating to superior thermal properties [3]. Recently, Zhang et al. [4] produced COC foam with excellent thermal insulated performance, which show great potential in building fields, COCs with various norbornene content were fabricated into thermal insulated foam via supercritical CO2 foaming technique, the resultant COC foams exhibit higher porosity of 97% and closed cell content of ~ 99%. The highly-expanded COC foams have a thermal conductivity as low as 28.5 mW·m−1 k−1.

Nonetheless, the susceptibility of COCs to ignition from external fire sources poses a significant challenge, owing to the carbon and hydrogen constituents within their macromolecular structure. Upon combustion, COCs have the potential to release substantial quantities of heat, smoke, and potentially hazardous gases, leading to safety concerns regarding both human casualties and property damage. This inherent flammability restricts the use of COC materials in various critical applications. Consequently, there is an imperative need to enhance the flame retardancy of COCs. At present, there are few studies on the flame retardant properties of COC, and there are no specific application examples.

At present, there are two ways to improve flame retardant performance, additive type and reactive type. Additive flame retardants are incorporated into plastics during the processing of plastics, and are mostly used in thermoplastics; Reactive flame retardant is added in the reaction system of polymer polymerization reaction, participates in the reaction in the form of monomer, and is synthesized as a part of the polymer through chemical bonds. In this study, it is an additive type flame retardant. APP and PER are typical additive intumescent flame retardant. Zeolite socony mobil No. 5(ZSM-5) particles with and without boric modifications are synergist in this additive flame retardant. The microporous structure of ZSM-5 has a high specific surface area and pore capacity, which can adsorb organic molecules and gaseous substances in the combustion process, play a synergistic role as a synergistic agent and add flame retardant to enhance the flame retardant effect, and has good thermal stability, which can maintain structural integrity and activity at high temperature. This allows the ZSM-5 to function under high combustion conditions, reducing the flame temperature and smoke generation rate. Among all added flame retardants, halogen flame retardants present the advantages of high flame retardancy efficiency and without compromising mechanical properties. But halogen flame retardancy will release burned and toxic smoke leading into serious damage to the environment and human health [5]. Huo et. al. ultilized phosphorus-containing imidazolium to develop epoxy compounds with feature of superior thermal latency, heat resistance and fire safety [6, 7]. Nowadays, the flame retardant products are required to show green and environmental protection. The intumescent flame retardant (IFR) with low smoke, low toxicity and high flame retardant efficiency has attracted more attention [8].

Intumescent flame retardants (IFRs) are capable of delaying further combustion and the spread of fire through the formation of a protective carbonized layer. As per the fire tetrahedron theory, the cessation of flame propagation is contingent upon the disruption of one or more key elements necessary for combustion: heat, fuel, oxygen, and the chain reaction. IFRs effectively shield the polymer from additional combustion by creating a carbonized layer on the surface, which prevents oxygen diffusion and inhibits the mass transfer of combustible gases. Consequently, this delays the degradation of the polymer and curtail smoke production [9, 10]. Furthermore, the addition of synergistic agents can enhance the efficiency of the flame retardant without diminishing the material's physical and chemical attributes.

Synergistic effect in flame retardancy is defined as two or more components combined performance is greater than that the sum of each component individual used. The common synergists of IFR includes metal oxides [11, 12], organic montmorillonite [13], sepiolite [14, 15] and zeolite molecular sieve [16,17,18]. Bernardes et al. [19]. found that H-ZSM-5 could act as catalyzer to cause the reaction between ammonium polyphosphate (APP) and pentaerythritol (PER), and thus the high acid center concentration is conductive to the formation of expansion precursors. Wu et al. [20] studied the effect of boron modification on the acidity of ZSM-5, and the results showed that with the increase of boron content, the acid content of the weak acid site of the molecular sieve increased after boron modification.

In this study, we have employed ZSM-5 with varying concentrations of boric acid as synergistic agents within the IFR system to enhance the flame retardancy performance of COC. Ammonium polyphosphate (APP) and pentaerythritol (PER) were also integrated into the COC composite. The modification with boric acid was characterized using scanning electron microscopy (SEM) coupled with energy-dispersive X-ray spectroscopy (EDS), while the boric acid content was quantified by X-ray photoelectron spectroscopy (XPS) and Fourier-transform infrared spectroscopy (FTIR). The flame-retardant performance of the ensuing COC composites was evaluated using the limiting oxygen index (LOI), vertical burning (UL-94), and cone calorimeter test (CCT). Our findings indicate that ZSM-5 modified with boron is an effective synergistic agent for IFR, successfully improving the flame retardancy of the COC composites.

Experimental section

Materials

The cyclic olefin copolymer (COC, CAS number 26007–43-2) was sourced from TOPAS Advanced Polymers GmbH. COC is a polymer composed of cyclic olefin and α-olefin, which is an important engineering plastic. At present, most of the COC structures are ethylene/norbornene (E/NB) copolymers, which is shown in Fig. 1. The zeolite socony mobil No. 5 (ZSM-5), characterized by a SiO2/Al2O3 ratio of 25, was acquired from Nankai University Catalyst Co., Ltd. Boric acid (H3BO3, CAS number 10043–35-3) and pentaerythritol (PER, CAS number 115–77-5) were supplied by Sinophosphate Chemical Reagent Co., Ltd. Additionally, ammonium polyphosphate (APP, CAS number 68333–79-9) was procured from Hangzhou Jelsi Flame Retardant Chemical Co., Ltd.

Fig. 1
figure 1

The structure of E/NB copolymer

Full size image

Preparation of B modified ZSM-5 (BZ5)

The specific preparation process of BZ5 is shown in Fig. 2. Initially, 15 g of ZSM-5 were introduced into 100 ml of deionized water and uniformly stirred using an ultrasonic bath set to a power level of 50%. Following a 10-min period of ultrasonication, 0.789 g of H3BO3 was added to the mixture, which was then subjected to ultrasonic treatment for an additional 30 min. The mixture was subsequently centrifuged and the resulting molecular sieve sample was dried overnight at a temperature of 110 degrees Celsius. The final product was designated as 5BZ5. The rest remained unchanged, and the samples obtained by changing the addition amount of B(OH)3 to 1.667 g, 2.647 g, 3.750 g, 5.000 g, 6.429 g and 15.000 g were recorded as 10BZ5, 15BZ5, 20BZ5, 25BZ5, 30BZ5 and 50BZ5, respectively.

Fig. 2
figure 2

Schematic diagram of the synthetic route of BZ5

Full size image

Fabrication of COC and COC composites

Firstly, COC was vacuum-dried at 50° C for 12 h, and APP, PER, ZSM-5 and a series of BZ5 samples were dried under the vacuum at 80 °C for 12 h. COC and COC composites were fabricated via a twin-screw extruder (SJZS-10 A, China) in the temperature ranged from 185 °C to 200 °C with a screw speed of 30 rpm. The formula composition and labeling of COC and COC composites are shown in Table 1.

Table 1 Composition of COC and composites
Full size table

Characterizations

Fourier transform infrared spectra (FTIR) (Spectrum Two, Perkin Elmer Instruments Co., Ltd., USA) were performed within the wavenumber range of 4000-400 cm−1 by Attenuated Total Reflectance (ATR) technique at room temperature.

X-ray photoelectron spectrometer (XPS) (AXIS SUPRA + , Shimadzu Corporation Co., Ltd., Japan) was used to perform elemental analysis of B/Z5.

The fire behavior for all samples were evaluated by limiting oxygen index (LOI), vertical burning (UL-94) and cone calorimeter test (CCT). The LOI was tested with a Digital Oxygen Index Tester (5801A, Suzhou Yangyi Vouch Testing Technology Co., Ltd., China) with dimensions of 130 × 6.5 × 3 mm3 according to ASTM D2863-97. The UL-94 test was performed on a Horizontal-vertical Burning Tester (5402, Suzhou Yangyi Vouch Testing Technology Co., Ltd., China) with the dimension of 130 × 13 × 3 mm3 according to ASTM D3801. The CCT was carried out on cone calorimeter (6810, Suzhou Yangyi Vouch Testing Technology Co., Ltd., China) with dimensions of 100 × 100 × 3 mm3 according to ISO 5660 with a heat flux of 35 kW/m2.

X-ray diffraction (XRD) was carried out using a powder diffractometer (D8 ADVANCEA25, Bruker AXS Co., Ltd., Germany) at Cu Kα radiation with the wavelength of 1.54 Å (λ = 1.54 Å) at room temperature. All the samples were dried in vacuum at 130 °C for two hours before XRD test to remove moisture.

Morphology was observed by Scanning Electron Microscopy (SEM) (ZEISS EVO18, Carl Zeiss AG Co., Ltd., Germany) equipped with energy dispersive X-ray spectroscopy (EDS) at 15.00 kV.

To investigate the thermal stability of samples, thermogravimetric analysis (TGA) was performed on TA Instrument (Pyris 1, Perkin Elmer Instruments Co., Ltd., USA), all of the specimens were heated from 50 to 790 °C at a heating rate of 10 °C/min under N2 atmosphere with flow rate of 50 mL/min.

Temperature programmed desorption of NH3 (NH3-TPD) was tested on a chemisorbed instrument (Model: AutoChem1II2920, USA). All samples were heated at a heating rate of 10 ℃/min to 300 ℃ under a helium flow rate of 30 ml/min, pre-treated for 1 h, then cooled to 120 ℃, and a mixture of NH3/He (volume ratio = 15/85) was introduced for adsorption. For gas desorption, after helium purge, all samples were heated from 120° C to 650° C at a rate of 10° C /min.

Results and discussions

Characterization of boric acid modified ZSM-5 samples

To examine the morphology of the boric acid-modified ZSM-5 molecular sieve, scanning electron microscopy (SEM) coupled with energy-dispersive X-ray spectroscopy (EDS) was conducted. As depicted in Fig. 3a, the modified ZSM-5 exhibits particles approximately 2 µm in size, a characteristic retained from the unmodified ZSM-5, suggesting that the boric acid modification does not alter the microscopic structure. EDS analysis confirmed the uniform dispersion of boron, aluminum, and silicon elements across the ZSM-5 molecular sieve.

Fig. 3
figure 3

SEM (a) of 25BZ5 and its element distribution (b) (c) (d)

Full size image

To quantitatively assess the boric acid content within the ZSM-5 molecular sieve, a dual analytical approach involving Fourier-transform infrared (FTIR) spectroscopy and X-ray photoelectron spectroscopy (XPS) was employed, with the results depicted in Fig. 4. The full-reflection infrared spectra in Fig. 4a highlight the presence of a pronounced absorption peak at 1393 cm−1, which is indicative of the B-O interaction [21]. An augmentation in the intensity of this peak is observed with increasing boric acid content, as illustrated in Fig. 4b. To further elucidate the boric acid content, a series of ZSM-5 samples with varying boric acid loadings were subjected to XPS analysis, with the corresponding spectra presented in Fig. 4c. The distinct peak at 193.4 eV, associated with the B1s orbital, is prominently visible in the XPS spectra, as shown in Fig. 3d (b). The escalation in peak intensity across the spectra correlates with an incremental rise in boric acid content. Accordingly, the loading of boric acid were 2.17%, 7.48%, 7.11%, 8.95%, 9.23%, 12.79%, 16.64%, respectively.

Fig. 4
figure 4

FTIR spectra of ZSM-5, 5BZ5, 10BZ5, 15BZ5, 20BZ5, 25BZ5, 30BZ5, and 50BZ5 (a) locally amplified FTIR spectra (b) XPS spectra (c) and B1s spectra of 5BZ5, 10BZ5, 15BZ5, 20BZ5, 25BZ5, 30BZ5, and 50BZ5

Full size image

Distribution of flame retardants and BZ5

To evaluate the dispersion of the modified ZSM particles within the COCs/25BZ5 composite, an examination was conducted using a scanning electron microscope (SEM) equipped with energy-dispersive X-ray spectroscopy (EDS), with the findings presented in Fig. 5. The elemental distribution within the composite is indicative of the uniform dispersion of the additive particles. Specifically, the presence of phosphorus (P) is attributed to the addition of ammonium polyphosphate (APP), oxygen (O) is associated with pentaerythritol (PER), and silicon (Si) is a signature element of the ZSM-5 molecular sieve. The SEM–EDS analysis confirms the uniform and homogeneous distribution of these particles throughout the COCs/25BZ5 composite matrix.

Fig. 5
figure 5

SEM of COCs/25BZ5 (a) and its element distribution (b) (c) (d)

Full size image

Flame retardancy of COC and COC composites

The limited oxygen index (LOI) and UL-94 vertical burning tests were employed to evaluate the flame retardancy of the pristine COC and its composites, with the specific data presented in Table 2. The neat COC exhibited an LOI value of merely 15.3% and did not meet the requirements of the UL-94 test, underscoring the necessity for flame retardant enhancement. Upon the introduction of the intumescent flame retardant (IFR) to the COC composite, the LOI value progressively elevated to 20.5%. Furthermore, the incorporation of unmodified ZSM-5 led to a slight increase in the LOI value to 23.8%, achieving a V-2 rating in the UL-94 test. Notably, after two successive 10-s combustion tests, the flame was extinguished within 60 s, and the sample was observed to ignite medical cotton placed 30 cm below. To further incorporated boric acid-modified ZSM-5, ranging from 2.2% in the 5BZ5 sample to 16.6% in the 25BZ5 sample, resulted in an initial increasement regarding of the LOI value of 28.5%, followed by a subsequent decline to 24%. This trend suggests an optimal boric acid modification content for ZSM-5, which significantly enhances the flame retardancy of the COC composites. Despite improvements in ignition time during the UL-94 test, the COCs/25BZ5 sample was not upgraded to a V-0 rating due to the pronounced dripping phenomenon, which consistently ignited the cotton pad placed in proximity.

Table 2 LOI, UL-94 and CCT data of COC and COC composites
Full size table

The cone calorimetric test (CCT) was employed to simulate and analyze the combustion behavior of the COC materials under conditions that mimic real-world fires, based on oxygen consumption. The CCT provides critical insights into parameters such as the heat release rate (HRR), mass loss rate (MLR), and total heat release rate (THR), as depicted in Fig. 6. Boron-based flame retardants mainly achieve flame retardant effects through two ways: 1)The flame retardant melts during combustion and absorbs part of the heat. 2) The flame retardant covers the surface of the polymer to isolate the flammable from contact with oxygen. The flammability of COCs was effectively reduced by the addition of modified boric acids, which can promote the formation of a stable carbon layer on their surface when exposed to heat or flames. This carbon layer acts as a physical barrier that isolates the underlying polymer from heat and oxygen, slowing down the combustion process. At the same time, boron compounds can release gases such as water vapor and carbon dioxide during thermal decomposition. These gases dilute the concentration of flammable gases in the vicinity of the flame, reducing the likelihood of combustion and flame spread. HRR is used to assess fire intensity and spread rate, and effective fire retardant systems often have low HRR values. Unlike the HRR behavior of neat COC, the HRR curves of all composites with boric-modified IFR additives exhibit a special "M" behavior (Table 2), which is resulted from the rupture of the carbon layer formation in the early stage leads to the rapid release of heat in the second stage. The beginning time of HRR in COCs/25BZ5 and COCs/50BZ5 materials increased dramatically due to the catalytic action in the decomposition of the base material. Compared with neat COC, the heat release time of all composites dramatically increased, which greatly reduces the fire risk of the material. Peak heat release rate (pHRR) can be obtained from the heat release rate curve (Fig. 6a, Table 2). The pHRR value decreased from 1368.71 kW/m2 for neat COC to 331.61 kW/m2 for COCs/25BZ5, 75% lower than that for neat COC. The mass variation can be used to confirm the decomposition of COCs/25BZ5 and COCs/50BZ5, as shown in Fig. 6b. From Table 2, one can see the THR reduced from 152.95 MJ/m2 for neat COC to 127.68 MJ/m2 for COCs/ZSM- 5 composite, which indicated that the heat release can be absorbed with the addition of unmodified ZSM-5. This heat absorption lowers the temperature of the polymer and suppresses flame propagation [22]. HRR and MLR of all BZ5-containing COC samples were lower than that of neat COC. TSR is the total smoke rate demonstrating the total amount of accumulated smoke produced per unit area of the burned sample. With boric acid content increase from 0 to 25% in flame retardant synergist, the TSR of COC composite increased from 2443.63 m2/m2 to 3957.27 m2/m2, which was about 52.8% higher than that of neat COC, which illustrate boric acid modification increase chances to form smoke. The increasement of TSR is resulted from more flammable substances content [23]. Moreover, boron based flame retardants can promote the carbonization to generate a stable carbon layer, which could be observed from SEM images (Fig. 7). Av-EHC stands for the effective heat calorific value of combustion. With boric acid content increase from 0 to 25% in flame retardant synergist, the Av-EHC increased from 32.40 MJ/Kg to 48.10 MJ/Kg, which was about 48.5% higher than that of neat COC. The increasement of Av-EHC demonstrated absorbing heat energy from the surrounding environment, which could decrease surrounding temperature by heat absorption to prevent flame propagation [24]. In this study, we selected 1 wt.% synergist as a control due to the flame retardant performance can be apparently affected by such amount of synergist in our previous work [25]. Therefore, we select 1 wt.% synergist as a control in order to investigate boric acid content effect.

Fig. 6
figure 6

HRR (a), Mass (b), TSR (c) and THR (d) of COC and its composites

Full size image
Fig. 7
figure 7

Char residue side, front and SEM photographs of COCs (a), COCs/5BZ5 (b), COCs/25BZ5 (c) and COCs/50BZ5 (d)

Full size image

Analysis of char residue

To gain a deeper insight into the alterations of the composite materials subjected to the CCT test, scanning electron microscopy (SEM) was utilized to examine the carbon residue of the COC composite, as depicted in Fig. 7. The side and front SEM images of the carbon residue post-CCT test are presented in Fig. 7 a1-d1 and a2-d2, respectively. Observations from the SEM analysis revealed that both the COCs and COCs/25BZ5 composites produced a smooth and intact carbon residue, devoid of apparent cracks in the frontal SEM images. In contrast, the carbon layer of the COCs composite appeared fragmented, indicating a lack of structural integrity. Conversely, the carbon layer in the composite material that included BZ5 as a synergist was more uniformly continuous and densely packed. This finding suggests that the inclusion of BZ5 significantly contributes to the formation of a robust carbon layer. The modified boric acids effectively diminish the flammability of COCs by catalyzing the creation of a stable carbon layer on the surface when confronted with heat or flames. This carbon layer serves as a physical barrier, shielding the underlying polymer from heat and oxygen, and thereby retarding the combustion process.

To elucidate the chemical composition of the carbon residue, Fourier-transform infrared (FTIR) spectroscopy was conducted on the residue post-cone calorimetric testing, with the results depicted in Fig. 8. The distinct spectral peaks observed at 1750 cm−1, 1607 cm−1, 1132 cm−1, and 998 cm−1 are respectively ascribed to the carbonyl (C = O), carbon–carbon double bond (C = C), ether (C–O–C), and phosphorus-oxygen-carbon (P-O-C) functional groups within the aromatic compound structure [26, 27]. These observations suggest that upon exposure to high temperatures, ammonium polyphosphate (APP) undergoes decomposition to form phosphoric acid, pyrophosphate, or polyphosphate, which subsequently becomes cross-linked with pentaerythritol (PER). Concurrently, the small olefin molecules that are generated through the process of pyrolysis are transformed into aromatic compounds under the catalytic influence of both ZSM-5 and BZ5 [28]. This cross-linked carbon layer can effectively slow down the transfer of heat and combustible materials between the gas phase and the condensed phase, thus inhibiting the combustion reaction of the substrate. The characteristic peak at 752 cm−1 corresponds to the stretching vibration of Si–O in ZSM-5 molecular sieve [29], which indicates that ZSM-5 molecular sieve is finally embedded in the carbon layer structure, which is conducive to improving the toughness of the carbon layer. The weak absorption peak at 1403 cm−1 corresponds to B-O-C [30], indicating that boric acid loaded on ZSM-5 molecular siolites also chemically reacted with adsorbed flame retardants.

Fig. 8
figure 8

FTIR spectra of char residue for COCs, COCs/ZSM-5, COCs/5BZ5, COCs/25BZ5 and COCs/50BZ5

Full size image

Flame-retardant mechanism

To elucidate the synergistic mechanism between the acid strength and content of BZ5 and its interaction with intumescent flame retardants (IFR), NH3-temperature programmed desorption (NH3-TPD) analyses were performed on both ZSM-5 and BZ5 samples. As illustrated in Fig. 9, the ammonia desorption profiles exhibit two distinct peaks; the peak within the range of 150–400 ℃ is attributed to weak acid sites, while the peak between 400–600 ℃ is indicative of strong acid sites [31]. All BZ5 samples showed an increase in acid content compared to ZSM-5. Among them, 5BZ5 and 10BZ5 had only weak acid sites but no strong acid sites, indicating that a small amount of boric acid modification introduced a new weak acid site to ZSM-5, and covered or eliminated the strong acid site [20]. When more boric acid was used to modify ZSM-5, the acidity concentration of the weak acid site of BZ5 sample gradually increased from 15BZ5, and the strong acid site reappeared and strengthened. From the point of view of corresponding temperature of acid site, the corresponding temperature of weak acid acidic site of BZ5 sample first increased and then decreased with the increase of boric acid modification amount, and the highest temperature was 299.2 ℃ of 20BZ5 sample. It is particularly significant that the 25BZ5 sample aligns with a weak acid site temperature of 273.2℃, which coincides with the initial decomposition temperature of ammonium polyphosphate (APP), potentially contributing to the superior flame retardant performance observed in the COCs/25BZ5 composite. Furthermore, while the 50BZ5 sample possesses the highest number of acidic sites, no corresponding improvement in flame retardancy was observed. This observation suggests that the synergistic efficacy between the molecular sieve and IFR is not solely dictated by the quantity of acid present, but may also be influenced by the temperature associated with the acidic sites.

Fig. 9
figure 9

NH3-TPD of ZSM-5, 5BZ5, 10BZ5, 15BZ5, 20BZ5, 25BZ5, 30BZ5, and 50BZ5

Full size image

Conclusions

By modifying ZSM-5 with different boric acid content, a series of BZ5 supported by different boric acid content were obtained, and then COC flame retardant composites were prepared with IFR and BZ5. The surface morphology, distribution of flame retardants, flame retardancy and morphology and structure of carbon residue of the composite were studied.

Energy-dispersive X-ray spectroscopy (EDS) analysis confirmed the homogeneous dispersion of ammonium polyphosphate (APP), pentaerythritol (PER), and BZ5 within the composite matrix. Limited oxygen index (LOI) testing revealed that the LOI value of the composites initially increased with the BZ5 boric acid loading, peaking at 28.5% for the COCs/25BZ5 formulation, a value indicative of significantly improved flammability resistance. Cone calorimetry (CCT) assessments indicated that the peak heat release rate (pHRR), total heat release (THR), and total smoke release (TSR) of the BZ5-synergized composites were superior to those of unmodified ZSM-5, with the exception of the COCs/25BZ5 composite. This suggests that an optimal boric acid modification level is crucial for enhancing flame retardancy. Scanning electron microscopy (SEM) and Fourier-transform infrared spectroscopy (FTIR) were employed to examine the microstructure and chemical composition of the carbon residue post-CCT, revealing that BZ5 significantly contributes to the formation of a dense and continuous carbon layer. This layer, formed through catalytic esterification crosslinking and pyrolysis of olefin molecules to produce aromatic compounds, augments the carbon layer's toughness. NH3-TPD analysis of the BZ5 samples highlighted the relationship between the synergistic effect of BZ5 and IFR, and the temperature corresponding to the weak acid sites, providing insights into the role of acid site strength in the flame retardancy mechanism.