Thermogravimetric analysis of face mask waste: kinetic analysis via iso-conversional methods.

. The surge of face mask waste in response to the global pandemic has proven to be a liability to the environment. Microfibers from plastic constituents of the face mask would cause microplastic pollution in the water bodies. Fortunately, these waste


Introduction
The World Health Organisation (WHO) has made wearing a face mask a mandatory requirement, as a social distancing effort to combat the spread of the novel coronavirus disease 2019 . In response, the face mask production has rapidly increased to meet the demand globally. It is estimated that, face mask consumption could reach 129 billion per month [1]. These disposable face masks typically consist of 3-layered composites filter; the middle melt blown filter layer separating the inner and outer layers of non-woven polypropylene (PP) [2].
These face masks are often classified as single-use plastic waste made from petrochemical polymers such as PP and polyethylene (PE) [3]. The consequence of the accumulation of these plastic wastes is the microplastic pollution. Based on conducted studies, microfibers (183 to 1247 particles piece −1 ) from face mask were being detected in the water bodies [4].
There are several methods to reduce these COVID-19 related wastes, i.e., reprocess, and recycling. In reprocessing strategies, there are studies on the sterilisation of used of face masks through supercritical CO2 sterilisation [5] and photoactive antiviral self-sterilisation face masks [6]. In recycling strategies, studies have ventured in repurposing the face masks into sound porous absorbers [7].
Following this, there are studies that look into converting these wastes into renewable energy source. According to Nawaz and Kumar [8], pyrolysis is the preferred thermochemical conversion method as it is able to remove potential pathogens, while converting these wastes into valuable green fuels. However, in their studies, only low to moderate heating rates (10 °C min -1 to 100 °C min -1 ) were utilised to study the pyrolytic behaviour of the face mask. Whereas the pyrolysis reactions, it is usually conducted in 2 modes, slow (<60 °C min -1 ), and fast pyrolysis (>60 to 1200 °C min -1 ) [9].
Therefore, this work focuses on the study of the pyrolysis of the three-layer samples of face mask via thermogravimetry. The thermogravimetric data is fit into an iso-conversional kinetic model, the Flynn-Wall-Ozawa (FWO), and Starink method.

Sample Preparation and Characterisation
Face mask samples were collected from the brand Pomerol. The face mask components were divided into the outer layer, middle layer, inner layer, cotton straps, and metal strip. The components were individually weighed and recorded in Table 1. In this research project, only the plastic face mask layers were considered. The face mask layers were cut and sieved into 1 mm particle size.

Thermogravimetric analysis (TGA)
Face mask pyrolysis was conducted with the Seiko EXSTAR TG/DTA 6300 thermogravimetric analyser, with varying heating rates of 10, 20, 50, and 100 °C min -1 . The inert gas (N2) supply was injected with a constant flow rate of 100 mL min -1 . The samples were subjected to non-isothermal heating from room temperature to 900 °C.

Kinetic model
Iso-conversional methods such as Flynn-Wall-Ozawa (FWO) and Starink method are useful tools applied to compute the activation energies (Ea) of the pyrolysis process. The model assumed that solid fuel follows a single step devolatilization reaction [10].

Flynn-Wall-Ozawa (FWO)
and are the heating rates, pre-exponential factor, activation energy, the integral form of the reaction model, universal gas constant and the temperature. The Ea can be determined by plotting ln vs 1/T, at each of the conversion rate, at varying [11].

Starink method
• The Starink method was employed to compare and validate the kinetic parameters determined from the FWO method. Ea is obtained from the slope of plot ( ) vs 1/T [11].

TGA results
From Figure 1, the TG and DTG curves for the pyrolysis of face mask are observed to show only one degradation peak, at lower heating rate of 10 °C min -1 , the degradation temperature range is determined to be 218 to 424 °C. A literature search was found that studied the thermal degradation range within 312 to 471°C for the face mask, which is comparatively similar to the current study [8]. At 100 °C min -1 , the degradation temperature range increased to 380 to 550 °C. This shift of temperature range is caused by the phenomenon called thermal lag, where temperature gradient is being formed at the surface to the center of the particle at high heating rates [12]. The start and end temperature of the major degradation of face mask (300 -420°C) were positioned between those of PP and PE (275 to 500 °C), which were the major components in the production of face masks [13]. Moreover, at all heating rates, the face mask achieved complete conversion into volatiles and non-condensable gases within the range of 364 °C and 540 °C at 10 °C min -1 and 100 °C min -1 respectively. The maximum degradation rates of the face mask waste samples were 10.65, 24.15, 83.82, 172.51 wt.%/min, for the heating rates of 10, 20, 50, and 100 °C min -1 respectively.

Kinetic analysis
Following this, the TGA data was fitted into two kinetic models, i.e., FWO and Starink method to determine the kinetic parameters of the pyrolysis of face mask wastes. The Arrhenius plots are shown in Figure 2. Based on the summary of the results in Table 2, both models showed great fitting with the coefficient of determination (R 2 ) ranging from 0.9722 to 0.9990 for FWO method, and 0.9954 to 0.9988 for Starink method.
Kinetic parameters such as activation energy (Ea) and pre-exponential factor (A) are important as Ea measures the energy required for the pyrolysis process to occur, enhancements of the process can also be reflected in the process i.e. catalyst, pre-treatment. The lower the Ea, the less energy is required [11]. The A refers to the frequency of collisions of particles within the system, where the higher the value, the quicker the reaction [14]. The Ea obtained from the results are 41.31 kJ mol -1 and 10.43 kJ mol -1 for FWO and Starink method, respectively. The A computed from the y-intercept of the Arrhenius plots are 0.9965 and 0.9901, respectively for FWO and Starink method. The discrepancies of both kinetic methods arise from its derivation methods and assumptions. The FWO method linearises the temperature integral using Doyle's approximation method [15]. While Starink method is an optimised expression of both FWO and Kissinger-Akahira-Sunose (KAS) method [15]. Further analyses by comparing the Ea and the , A and were carried out, and illustrated in Figure 3. It was observed that FWO method showed more fluctuations in the activation energies, while Starink method showed a strong resistance to change at each conversion stage. Both methods showed similar trend for the preexponential factors determined at each conversion stage.

Conclusion
In summary, the thermal decomposition of face mask waste can be observed as a single peak DTG curve ranging from 218 to 424 °C at the lower heating rate of 10 °C min -1 , with a maximum degradation rate of 10.65 wt.% min -1 . Further, when the sample was subjected to a higher heating rate of 100 °C min -1 , the maximum degradation rate was enhanced to 172.51 wt.% min -1 . From the kinetic analysis, Starink method proves to be a much reliable model for the pyrolysis process, yielding higher average R 2 value (0.9982) than FWO (0.9949) method. The kinetic parameters obtained from Starink method was 10.43 kJ mol -1 and 0.9901 s -1 for Ea and A respectively. Future work for this study includes the determination of the reaction mechanism via model fitting methods, such as Criado method or Coats Redfern method. Furthermore, the effects of catalyst, as well as the co-pyrolysis technique with biomass could also be explored.