---

The Main Steps

The industrial-scale process can be articulated in multiple ways. It's important to note that the final plant's scale cannot be definitively determined based solely on laboratory results. However, we can identify the corresponding industrial processes required for high-yield recovery of textile waste. The necessary unit operations can be derived from:

1. Existing clothes recycling operations
2. Best practices in chemical and biochemical engineering

Initially, the textile waste must undergo preparatory steps to make it accessible to the spores' effect. This is most effectively accomplished through a two-step upstream process:

• Mechanical Disintegration: A physical breakdown of the fabric structure
• Chemical Pre-Treatment: Further preparation of the textile fibers

These steps ensure optimal conditions for the subsequent biological treatment, maximizing the efficiency of the recycling process.

It is rarely the case that a new laboratory solution is quickly adopted by industry and implemented on a tremendous scale. Most often, a pilot plant is constructed to prove that the investment is worthwhile and that the process can be carried out stably for years with a solid profit margin, low maintenance costs, and ideally no surprising changes to the setup.

For this purpose, a potential pilot plant with a conversion reactor of 100 L working volume is described here. This intermediate step serves as a critical bridge between laboratory-scale experiments and full industrial implementation, allowing for the assessment of process parameters, efficiency, and economic viability on a more representative scale before committing to a full-scale industrial plant. For an overview of such a pilot plant, see figure 1.

In the following sections, the aim is to elaborate on each main step and to point out the possibilities, the challenges, and the uncertainties. This should not be seen as a final concept but as a starting point for the later design of a plant.

---

Mechanical Disintegration

Clothes waste may arrive in various forms as substrate to our production stage, including post-consumer textiles, industrial scraps, and unsold items. To guarantee high-quality products and a stable process, homogeneity of the substrate is desired. Delivered substrates must be sorted to manually remove insoluble parts such as buttons, zippers, and other non-textile components. It must be mechanically disintegrated to destroy the fiber structure. This is best done with a shredder and a hammer mill. The target size of materials leaving the disintegration stage should be around 1 – 2 cm² for optimal processing.

To increase the implementation speed of our process, pretreated substance could be ordered from suppliers. Outsourcing the production decreases the complexity of the process; would in return, however, increase dependency on the suppliers. Examples of potential suppliers include ALTEX and Soex. The German company ALTEX, specializing in textile recycling, even sent us a sample of their typical product, disintegrated working clothes.

An advanced step would be to include sorting for PET (polyethylene terephthalate) / cellulose amounts. Due to the nature of our process, this is not necessary. Still, a similar component ratio of the substrate ensures a similar product result. The downstream process might benefit from similar compound ratios in the product.

---

Pre-Treatment

After disintegrating the textile waste, the fiber structure has indeed changed macroscopically but not microscopically (Cao et al., 2021). While this is not so important for PET, it is crucial for cellulose. To be attackable by enzymes, a further step must be taken to decrystallize the cellulose polymer. Literature research has shown that there are several methods to achieve this, whereas one treatment stood out due to its high enzymatic hydrolysis yields that could be subsequently observed. It consists of using 15% NaOH solution at 120 °C in an autoclave (Boondaeng et al., 2023).

The shredded textiles soak up the leach, which deprotonates the hydroxide groups of the cellulose, thereby increasing its overall charge. This weakens the intermolecular forces within the cellulose as equally charged particles repel each other. Compared to other common methods, the PET and cellulose fibers do not react too harshly to the alkaline treatment, so that the chemical structure is mainly preserved (Yokota, Nishimoto & Kondo, 2022). However, it was found that the degradation of PET fibers to their monomers occurs when treated with a highly concentrated base (Boondaeng et al., 2023). While it did not become clear to us to what degree this happens, it threatens our approach to use PETase to be needless. Former experiments have shown that the particle size and the choice of a catalyst are also important factors besides the temperature (Das et al., 2007). Since the feedstock is not pure PET pellets, the industrial transfer of the recycling process should not be completely geared to lab results in this case. Furthermore, added special chemical catalysts have supported the alkaline PET hydrolyzation (López‐Fonseca et al., 2009). These are often unattractive to use on a large scale. Therefore, we wanted to conduct our own experiment consisting of three steps: exact determination of the ALTEX sample’s composition, analysis of composition after the alkaline treatment, and determination of hydrolysis yield with commercial enzymes (Novozym 342 and Celluclast^®, kindly provided by Prof. Dr. Christiane Liers, International Institute Zittau).

As cellulose is also part of wood, we asked the Institute of Wood Technology Dresden, to give us advice regarding the problem of how to turn cellulose in something less resistant. It turned out that they even have a steam explosion plant – for an example, see the “Schematic of the steam explosion system (“steam gun”)” (Pielhop et al., 2016). However, they digest wood chips only hydrothermally under high temperatures and pressures, with up to 240 °C and 30 bar, respectively, while we wanted to perform experiments with highly concentrated sodium hydroxide. Nevertheless, they were interested in our project and even offered us to perform experiments, including analyses with the steam explosion plant, setting the parameters as we desire.

That would have been an interesting cooperation because of the combined power of chemical and physical forces that come together in the steam explosion process. The cellulose structure gets disrupted more efficiently, leading to an even more damaged polymer (Auxenfans et al., 2017). Unfortunately, we could not manage to conduct any experiments in time except soaking the ALTEX textile sample in NaOH (see Fig. 2). Therefore, we concluded to keep the representation of the pre-treatment step rather simple. Thus, we ended up planning a heated vessel containing NaOH solution without considering that PET is fully degraded to its monomers. To prove a significant PET degradation on a large scale, so called down-scaling experiments would be necessary.

For potential implementation later, a comparison of the degree of efficiency between an autoclave and a steam explosion machine would be necessary as well. Since, regarding economic and ecological factors, the most crucial part of the pre-treatment step is presumably which temperature and pressure suffice for an optimal decrystallization of the cellulose. Speaking of economic and ecological factors, it should also be kept in mind that the high concentrated NaOH solution might be reusable. Since the periodical consumption of such big amounts is resource-intensive, it also leads to higher costs.

---

Biochemical Degradation

The heart of the recycling plant is the conversion of the fibers into glucose, terephthalic acid and ethylene glycol. As the solution of the pre-treatment is very alkaline, the pH value must be adjusted to the optimum for the enzymes on the spore crust. An appropriate port to add the pH corrective must be installed. To make PET more accessible for PETase, the reactor must be heated up to 70 °C, as this is the glass-transition temperature of PET (Linseis, (n.d.)). At this temperature, PET becomes more amorphous and thereby less densely packed. Hence, the resilient nature of the spores led us to choose them as an anchor for the enzymes in the first place. In the end, they serve as a chemical catalyst, surviving harsh environmental conditions with no change of concentration over time in a batch process. Future approaches could even test if merging the alkaline pre-treatment and the enzymatic conversion is possible, although it depends, of course, on the enzymes used. Furthermore, with the technology to install foreign enzymes in the spore crust of Bacillus subtilis, several combinations are imaginable. For instance:

1. One stem with PETase and all three cellulases
2. One stem with PETase and one stem with all three cellulases
3. Four stems with one enzyme each, including PETase and the three different cellulases

The reactor layout should be aligned with the chosen spore-enzyme combination. For example, in case 2 and 3, it could be an option to pursue a cascade concept, as the PET and cellulose degradation can be separated. This could have advantages in terms of preventing undesired interactions and thereby earlier elimination of PET monomers. But as our project focuses mainly on the proof of concept, we act on the assumption that one batch reactor suffices in order to get a notion of how the whole industrial process could look.

---

Separation and Filtration of Products

As the spores are only the catalyst and not part of the desired product, they must be removed. This is possible by using conventional filtration technologies like tangential flow filtration. It is efficient for processing large volumes of liquid and works under lower pressure than other filtration methods. Therefore, it is perfect for our purpose of separating the spores from the remaining solution on an industrial scale. The lower pressure even helps preserve the filtration membrane, so the time between failures is relatively longer, making it more ecological and economical. However, the energy needed is higher than traditional membrane filtration (Rocker Scientific Co., Ltd. (n.d.)).

Fortunately, solutions to realize large-scale spore separation already exist. Managing the separation of the three-component mix – glucose, terephthalic acid and ethylene glycol – is the actual challenge. As our aim is to split off one substance after the other, the first step was to look for different physical properties between them. Terephthalic acid can be precipitated under acidic conditions, while such conditions do not influence the solubility of glucose or ethylene glycol (Das et al., 2007; López‐Fonseca et al., 2009). The precipitate could be harvested by centrifugation or by using a gravity separator. For this reason, we chose to represent the TPA filtration step with the general filtration symbol, keeping both possibilities in mind. The differences in terms of physical properties between glucose and glycol are, however, not so significant. Although distillation would be conceivable due to the different boiling points, the water fraction of the solution would still be too high. That means the energy needed for the distillation column would be mainly spent on vaporizing water, which does not sound like an economical process at all.

Therefore, we conferred with Prof. Dr. Markus Schubert from the Department of Chemical Engineering to help us approach the problem. He proposed two ideas, which might be costly to implement but at least more successful: enzymatic purification and purification by membranes. While enzymes are highly specific, the one that matches the requirements of the process must be found first and tested. They are often expensive but usually do not require any consecutive purification steps.

On the other hand, membranes offer a versatile and adaptable solution, with a range of pore sizes and material compositions that can be tailored to specific separation needs. The challenge lies in selecting or developing a membrane that can effectively differentiate between glucose and ethylene glycol based on their molecular size or other physicochemical properties. Using membranes for separation is advantageous due to their ability to handle continuous processing, making them suitable for industrial applications. However, membranes can be expensive, especially when using advanced materials designed for high specificity and durability. The cost is mitigated by their reusability and the potential for regeneration, which extends their operational lifespan.

There is yet another approach beyond purification technologies. As we are dealing with glucose, a consecutive fermentation with a strain that produces ethylene glycol, thereby enriching the alcohol fraction, is conceivable. In this case, the strain would likely need to tolerate certain ethylene glycol concentrations due to its toxic effect. In times of consistently improving synthetic biology technologies, this is not only an idea but a real opportunity to build such strains. Since there are many possibilities for separating glucose and ethylene glycol, we decided to simply represent both fractions together as the result of the TPA filtration. Considering the possibilities, which technology to use only for the last step could already cover a whole new project.

---

Cultivation of Spores

Working with recombinant strains on an industrial level comes with high safety criteria. An entire plant that is planned according to these standards would have an immensely higher cost than a typical chemical plant. Therefore, it is easier to cultivate the chosen organism independently of the main plant.

The seed train could be managed with a focus on ideal growth and high-quality spore generation in a closely monitored environment. For a typical seed train, the initial cells are first grown in precultures of increasing volume. The scale of the bioreactor for the main culture should match the 100 L conversion tank. Since the biomass is not wasted after each cycle but reused instead, a 10 L bioreactor for the main culture should be suitable. Because we are not trying to generate a separate product during the main cultivation, the process should be optimized for maximum biomass generation. Thus, a fed-batch reactor would fit the task best. The optimal growth of B. subtilis occurs aerobically at 37 °C at a pH of 7.0 with glucose as the primary substrate (Luo et al., 2010). The main culture is not the end of the spore generation process.

After the cells are successfully enriched in the main culture, the spore generation process can start. It is the essential part of the spore cultivation. The safe transition from growing cells to spores is also vital for product safety. If there is a guarantee that only spores remain, their storage and feeding to the conversion step become a lot easier. To start the starvation of the cells, the broth from the main reactor is transferred into a barrel. This barrel should have conditions similar to the starvation with the DSM medium used in the lab during spore preparation (see Experiments ).

After the sporulation is complete, wet storage of the spore suspension under sterile and cool conditions allows for a theoretical lifetime of the spores for many years. Dry storage of the spores could be an option as well, but it would involve another separation step, which increases costs further. As a part of this regimen, batches with inactive spores could be identified before entering the main conversion step. High-activity spores, on the other hand, could be prioritized. The impact of cultivation parameters like aeration, agitation, substrates, and pH on the spores' quality must still be tested.

Vision and Potential Customers

Based on the expert talk with Prof. Dr. Markus Schubert, we identified two options for product management. He indicated that there is a market for TPA, EG, and glucose, suggesting that our initial approach could involve selling these products to customers in the plastic and polyester production sectors, as well as for other chemical applications following separation. According to a report from Emergen Research, the global terephthalic acid market size was valued at USD 58.24 billion in 2021 and is expected to register a revenue compound annual growth rate of 4.1% until 2032, reflecting its ongoing demand in the industry (Emergen Research, 2024). The second option for our products involves further processing within the facility. The complicated separation stage could be avoided this way. Instead, the products could become substrates for another conversion step to produce higher-value chemicals.

---

References