Charting a path to catalytic upcycling of plastic micro/nano ﬁ ber pollution from textiles to produce carbon nanomaterials and turquoise hydrogen †

Washing synthetic textile ﬁ bers releases micro/nano plastics, endan-gering the environment. As new ﬁ lters and associated regulations are developed to prevent ﬁ ber release from washing machines, there emerges a need to manage the collected waste, for which the only current options are combustion or land ﬁ ll. Herein we show for the ﬁ rst time the application of a catalytic pyrolysis approach to upcycle textile derived ﬁ brous micro/nano plastics waste, with the aim of keeping carbon in the solid phase and preventing its release as a greenhouse gas. Herein, we demonstrate the co-production of hydrogen and carbon nanomaterials from the two most prevalent global textile micro ﬁ ber wastes: cotton and polyester. Our results pave a way forward to a realistic process design for upcycling mixed micro/nano ﬁ ber waste collected from laundering, drying, vacuuming, and environmental cleanup.


Introduction
From 2 million metric tonnes in 1950 to 390.7 million in 2021, plastic manufacturing has grown dramatically (4%) and is predicted to double within 20 years. 1,2China is the world's largest producer of plastics, accounting for 32% of global production in 2021. 2 Weathering and aging turn most discarded and mismanaged plastics into micro/nano plastics.Micro/nano plastics have gained increased attention in recent years as they are present in land, air, and water in alarmingly large quantities. 3Due to their small size and large surface-to-volume ratio, micro/nano plastics can also sorb and accumulate pollutants which they can then transfer to organisms, causing toxicity across the food chain.In addition to this, the additives present in plastics (plasticizers, UV blockers, etc.) can also release toxins and interact with adjacent pollutants. 3ynthetic textiles contribute heavily to microplastics release, making up 35% of primary microplastics released to the environment. 4The natural textile cotton and the synthetic textile polyester (PET) make up most of the fabrics used in clothing.While alternative methods like the Lyocell spinning process can also be used for recycling cotton waste, 5,6 it is essential to evaluate the suitability and feasibility of the produced micro-bers for large-scale production of Lyocell in terms of yield and purity.Moreover, the presence of polyester contamination presents additional obstacles in extracting cellulose for ber spinning processes.In this context, pyrolysis emerges as a more robust approach capable of effectively handling anticipated contaminants and facilitating the utilization of cotton, polyester, and other materials.It should be noted that recycling and pyrolysis techniques are not mutually exclusive, as textiles cannot be recycled indenitely.Micro/nanobers from garment nishing, laundering, drying, and even wearing are oen mixed synthetic and natural bers, with no central collection or processing mechanism, making pollution abatement difficult.8][9] In terms of environmental damage, the present research indicates that the two types of bers do not behave the same.However, because it is not possible to separate collected micro/nanobers, it is crucial to understand the behaviour of both types of bers under the same thermochemical upcycling technique, to utilise thermochemical conversion as a means to process a thoroughly mixed feedstock of natural and synthetic bres.While reduction and substitution of plastics use are necessary to prevent pollution, there are also signicant efforts underway to capture the generated waste, such as by tting washing machines with lters.California, France, and Australia are already ghting microbre pollution.Aer January 2029, all residential, commercial, and state-owned washing machines in California will be required to have a built-in microbre ltering system (mesh size no higher than 100 mm). 10 The National Plastics Plan of the Australian government mandated microber lters in commercial and home washers by July 2030.2][13] In December 2019, the European Commission announced the European Green Deal (EGD) to make Europe the rst climateneutral continent by 2050. 11,12In 2015 all United Nations Member States agreed to the 2030 Agenda for Sustainable Development, which is directed at achieving peace and prosperity for people and the planet.The agenda presented seventeen Sustainable Development Goals (SDGs), connecting the mission to eradicate poverty and other deprivations with targeted actions and initiatives promoting health and education, decreasing inequality, stimulating economic development, addressing climate change and preserving our seas and forests. 14As legislative efforts are underway and debates are sparked on how to manage and collect this emerging micro/ nano waste, it is also necessary to consider how it will be ensured that the collected plastics are not released back to the environment (through landlls for example) or contribute to climate change through their uncontrolled combustion. 13erein, we provide proof of concept for the catalytic pyrolysis upcycling of micro/nano plastic waste into energy carriers (such as H 2 ) and carbon nanotubes (as well as other nanostructured carbon products).We show that an Fe-Ni catalyst can convert the two most dominant fabrics used globally, cotton (natural ber) and polyester (synthetic/polymer ber) 15 to hydrogen and solid carbon nanomaterials.With suitable catalysts, thermal conversion technologies can be optimised to produce turquois hydrogen and tailor the structure of the carbon products towards need.Ni-based catalysts present high reactivity for C-O and C-H bond cleavage, making them effective for polymer cracking and reforming reactions. 16One of the targets of this research work was to produce as a secondary product carbon nanomaterials, such as carbon nanotubes that have attracted much interest since their discovery due to their exceptional chemical and physical properties.Other types of carbon materials can also potentially be targeted with this technology, such as graphite, graphene, carbon bers, and amorphous carbon. 13is approach would transform harmful waste materials into useful product, while keeping carbon in the solid phase (avoiding greenhouse gas emissions).The hydrogen produced by pyrolysis of textile micro/nano bers may be classied as turquoise hydrogen since it is produced by thermal breakdown of plastics/biomass while producing minimal emissions. 13,17In addition, hydrogen has numerous industrial applications, is considered a clean fuel, is central to future energy scenarios and can be utilised to provide some of the energy necessary in the pyrolysis via a low emission fuel. 18The global scope of the problem and the urgent need for progress necessitate ambitious collaborative efforts among fundamental scientic research, engineering development, industries, governments, and computational studies to accelerate research and develop and optimize promising processes leading to viable circular economy solutions. 18

Results and discussion
In order to model the textile ber waste generated from washing and drying cotton and polyester fabrics, a milling method was used resulting in the length and width distributions shown in Fig. 1A and B. The bers produced from polyester fabrics with the 40-mesh screen were between 200 mm (0.20 mm) and 1-2 mm in length whereas the cotton fabrics produced bers >300 mm (0.3 mm) and max 4-7 mm in length.In terms of width, cotton bers are thicker (∼20 mm), whereas polyester bers have a slightly thinner width (∼18 mm).Fig. 1E-J shows optical and SEM images, respectively, of the bers where the presence of nes (particles in between 0.03 to 0.05 mm in size) can be appreciated.Microbres are dened as threadlike residues with widths of approximately 6 to 175 mm, diameters of approximately 28 mm on average, and lengths of approximately 250 to 6250 mm, 19,20 depending on the source of the ber.Therefore, the particles selected can be considered as a representative sample.Due to their small size, high surface area and relatively predictable composition (formed mainly of cotton and polyester), laundry derived micro/nano textile bers were hypothesised to form a suitable starting material for devising a thermochemical catalytic upcycling procedure aimed at making highly valuable carbon nanomaterials.Moreover, the higher surface area compared to other plastic wastes was expected to increase reactivity, facilitate better heat and mass transfer as well as have favourable interaction with catalyst. 18olyester and cotton microbers were catalytically pyrolysed using a Ni-Fe/g-Al 2 O 3 catalyst (details in ESI †) over a broad temperature range and showed distinct gas evolution features (Fig. 2).Both cotton and polyester released hydrocarbons as well as hydrogen during pyrolysis (Fig. 2).H 2 production was at its peak at 500 °C for both samples but displayed a different onset temperature for cotton and PET (Fig. 2A1 and B1).H 2 production occurred aer the observation of other gas products of pyrolysis, which is indicative of secondary hydrocarbon cracking reactions. 21Furthermore, the hydrogen production peaks were quantied (details in ESI †), which facilitated iden-tication of temperature ranges maximising hydrogen production (Tables 3S and 4S †).Notably, the results consistently indicate that the highest production peak for both feedstocks was observed at a temperature of 500 °C.This nding emphasizes the signicance of this specic temperature as the optimal condition for maximizing hydrogen production across different feedstocks.Signicant insights can be gleaned from the observed CO and CO 2 production peaks depicted in Fig. 2. The data clearly indicate that the current process is not yet optimized.Consequently, the subsequent phase of this research will entail conducting a comprehensive parametric study of the process and to optimize the catalyst composition, thereby reducing the emissions of CO and CO 2 while simultaneously enhancing the overall hydrogen production.By ne-tuning the catalyst composition, it is anticipated that the undesired emissions will be minimized, leading to a more efficient and environmentally friendly hydrogen production process.In an ideal application only H 2 would be produced in the gas phase and emission of polluting and greenhouse gases would be avoided by adjusting the process temperature, temperature ramp rate, number of heating zones used in pyrolysis as well as catalyst structure and composition to promote secondary cracking reactions.Fig. 2 is important for illustrating catalytic pyrolysis of a well dened solid feedstock representative of laundry derived waste and forms a rst step in catalyst and process optimisation.As the monomer of cotton (C 10 H 20 O 10 ) contains a larger quantity of hydrogen than the monomer of PET (C 10 H 8 O 4 ), the observation of a larger hydrogen production peak when cotton is used as a feedstock is consistent with the higher H/C ratio in the chemical formulae of starting feed (Fig. 2C and D). 22While cotton is mainly composed of cellulose, PET is derived from terephthalic acid (TPA) and ethylene glycol (EG), and the high concentration of aromatic monomers on the backbone reduces chain mobility. 15ig. 3A and C presents TGA and DSC during temperature programmed oxidation (TPO) of the solid products of pyrolysis (mixed with catalyst), to determine the thermal stability and chemical structure of the carbon produced.By looking at the mass loss peaks (TGA) and exothermic/endothermic features (DSC), it is possible to perform carbon nanomaterial identication on the sample.This is a common method to characterise the purity of carbon nanomaterials, e.g., carbon nanotubes. 16rom Fig. 3A and C, three distinct mass loss regions are  observed: (1) under 100 °C, moisture present in the sample; (2) between 350 and 450 °C, combustion of amorphous carbon; 16,23 and (3) from 450 until 700 °C, associated with lamentous carbon. 16,23Comparing TGA and DSC curves of the post-reaction samples with the TGA and DSC curves of the fresh samples (Fig. 2S †) a catalytic transformation is conrmed along with the presence of multiple types of carbon on the post-reaction samples.Furthermore, the solid carbon product yield was calculated (further information in ESI †) as 43% for cotton and 28% for PET.Table 1 displays the analysis of the post pyrolysis carbon products, based on their oxidation temperatures.It can be observed that cotton pyrolysis results in a major percentage of lamentous carbon.
Raman and FT-IR measurements validate the TGA-DSC results.FT-IR (Fig. 3B) shows that thermal treatment produced clear destruction of the C]O and C-H chemical functionalities of both cotton and PET, indicative of a high degree of carbonisation.
Raman spectroscopy (Fig. 3D) also shows notable differences in cotton and PET aer catalytic pyrolysis.The appearance of two characteristic peaks at around 1350 and 1580 in the post-pyrolysis samples are indicative of sp 2 bonded carbon. 24The Raman spectra of disordered graphite exhibits two modes, the G peak at 1580-1600 cm −1 and the D peak at 1350 cm −1 , which are oen attributed to phonons with E 2g and A 1g symmetry, respectively. 25The existence of the D peak band can indicate the presence of aromatic compounds with a ring size greater than six fused rings. 26G band is commonly referred to as the "graphite band" because it involves the in-plane bondstretching motion of pairs of C sp 2 atoms (E 2g ), whereas D band is commonly referred to as the "defect band" because it provides information about the morphological disorder and defects that are characteristic of disordered graphite (A 1g ). 24As it can be appreciated in Fig. 3D, the shape and position of these bands differ slightly across the different cotton and PET postreaction samples indicating structural differences between the carbonaceous structures produced by catalytic pyrolysis of the fresh samples.The D band is shied towards higher Raman shi values in the cotton post-reaction sample, while the G band is quite stable in the same value for both post-reaction samples (cotton and PET).The D band is a double-resonant process in Raman.When the number of defects rise, the D  band usually moves to higher frequencies. 24The ratio of the intensities of these bands (D/G) is a crucial parameter for distinguishing the defect concentration of carbon generated during thermal decomposition caused by catalytic pyrolysis processes. 27The conversion of nano-crystalline graphite to amorphous carbon increases the D/G ratio. 279][30] The D/G ratio varies widely depending on the type of CNT material.The D/G ratio of many multiwall CNT-based materials is between 0.27 and 0.90. 31,32ingle-walled and double-walled CNTs (SWCNTs and DWCNTs, respectively) have higher D/G ratio, for 532 nm laser excitation, from 0.09 as-is to 0.01 puried, and from 0.12 as-is to 0.04 puried (for 785 nm laser excitation).4][35] The D/G ratio value obtained for the cotton and PET post-reaction samples were 0.65 and 0.75 respectively.This shows that carbon produced by the utilization of cotton as feedstock is more graphitised. 36SEM-EDX (Fig. 4) was used to characterise the growth/deposition of carbon nanostructures on the catalyst and showed that carbon was co-located with Ni and Fe, indicating a gas phase cracking reaction on the catalyst rather than a solid-to-solid transformation. 21Further investigation via TEM-EDX revealed that catalyst sintering has occurred during reaction and identied stabilization of nano-alloys of Ni and Fe as a promising path forward to produce size-controlled carbon nanotubes.
Our pyrolysis proof of concept experiments, together with characterisation show that it is possible to produce hydrogen via chemical processes from both cotton and polyester micro-ber waste, modelling the majority feedstock of waste that can be collected from laundering, drying, and vacuuming with the use of lters.We show that lamentous carbon and hydrogen can be produced from both cotton and polyester, indicating a process could be developed for catalytic upcycling of mixed microber waste.Improvements in catalyst formulation and especially stability are key, as we have observed some agglomeration of NiFe nanoparticles (Fig. 1S †), the proposed active phase for the growth of lamentous carbon.Process improvements can include a dual reactor conguration where micro-ber waste can be heated to a lower temperature while the catalytic pyrolysis would be performed above 500 °C, which can achieve greater selectivity over products due to the separation of the gasication and cracking steps.The production of hydrogen (which can be called turquoise hydrogen due to the carbon remaining in the solid phase) through this process is signicant for offering a low emission source of heat to drive the thermochemical upcycling process.Based on the chemical formula of polyester, the heat of pyrolysis 22 and the lower heating value of hydrogen gas, it is estimated that up to 19% of the energy requirement for pyrolysis of polyester can be provided by the H 2 released from reaction (details in ESI †).While the process will require external energy input, it can still serve as a low emission alternative for the synthesis of carbon nanotubes which have a vastly growing demand, and the state-of-the-art processes for their production all rely on fossil fuel derivatives as reactant.Moreover, upon shiing the mix of textiles used in society to depend less on fossil derivatives and more on biobased alternatives, the resulting changes in brous feedstock can further lower the carbon footprint of bers derived carbon nanotubes.

Conclusions
Herein, we have demonstrated a controlled proof of concept for the catalytic pyrolysis behaviour of the two most common microber wastes released from laundering fabrics.A process temperature of 500 °C has been shown to be necessary for H 2 production from both cotton and polyester.We showed via FT-IR, Raman, TGA-DSC, SEM-EDX, TEM-EDX that both cotton and PET yielded lamentous carbons, with cotton microbers generating more graphitised carbonaceous material.This is a novel approach to upcycle the most difficult type of waste (microbers) among 92 million tonnes of waste generated annually by the fashion industry, most of which is incinerated, landlled, or exported to developing countries. 37Based on this foundational proof of concept, the design of stable catalysts with control over carbon product selectivity can enable the integration of microber waste into a circular economy, reducing the impact of fast fashion on the climate as well as land and water ecosystems.

Fig. 1 (
Fig. 1 (A) Fiber length distribution of microfibers.The average fiber length 1-2 mm for PET and 4-7 mm for cotton (B) fiber width distribution of microfibers.The average fiber with is around 18 mm for PET and 20 mm for cotton (C) chemical formula and structure of cotton and (D) chemical formula and structure of PET.Comparing (C) and (D) cotton have more hydrogen that can be realised upon (E) and (F) optical microscope images of textiles (E) cotton and (F) PET.(G) and (H) Optical microscope pictures of fines of microfibers (G) cotton and (H) PET.(I) and (J) TEM images of microfibers (I) cotton and (J) PET showing cut/ripped ends.

Fig. 2
Fig. 2 Gas evolution profile for catalytic pyrolysis experiment (with zoom of H 2 evolution).(A) PET and (B) cotton.

Fig. 3
Fig. 3 TGA and DSC curves of combustion of post-reaction samples in air atmosphere (5 °C min −1 ) (A) PET and (C) cotton.Two oxidation peaks are shown in both curves, indicating the production of two different types of carbon (B) FT-IR patterns.The disappearance of C-O and C-H shows a certain degree of carbonization (D) Raman spectra of pre/post reaction cotton and PET samples (532 nm laser was used to conduct post-reaction measurements).D/G ratio indicates higher degree of graphitization when the value approaches 1.

Fig. 4
Fig. 4 SEM-EDX images.(A, C, E, G, I) Cotton post-reaction and (B, D, F, H, J) PET post-reaction.Carbon growth over the catalyst particle can be appreciated.

Table 1
Distribution of carbon products of cotton and PET pyrolysis, determined via TGA-DSC