Normal-hexane treatment on PET-based waste fiber depolymerization process

2.1 Depolymerization and normal-hexane treatment process method

To enhance the purity of TPA extracted from waste quilt batting collected from landfills, a normal hexane treatment was incorporated into the depolymerization process. The mother liquor resulting from the hydrolysis of quilt batting was kindly provided by Kolon Industries Inc. The detailed workflow of the hydrolysis process is illustrated in Figure 1. The impurity removal process is divided into four stages. The first stage involves using a bag filter to remove undissolved substances. The second stage eliminates small undissolved particles and adhesive components used in the fiber manufacturing process. The third stage removes micro-sized undissolved impurities and adhesives. Finally, the fourth stage employs an activated carbon process to remove impurities such as oil, amines, and surfactants. The activated carbon used in this process was manufactured through the process shown in Figure 1, and the manufactured activated carbon was used in the relevant stage during the process in the yellow box. A normal hexane process was introduced after the third stage of undissolved impurity treatment and before the fourth stage of activated carbon treatment.

Figure 1

Workflow of pitch-based activated carbon manufacturing process and PET-based feedstock hydrolysis process and impurity removal step.

In the standard hexane treatment process, a squibb-shaped separating funnel was filled with 300 mL of mother liquor, followed by the addition of normal hexane at a volume ratio equivalent to 10% of the mother liquor. After thorough mixing, the mixture was allowed to settle until complete layer separation was observed, with a distinct yellow mother liquor layer at the bottom and a transparent normal hexane layer at the top. The lower mother liquor layer was then collected and separated. To evaluate the foaming tendency caused by the surfactant properties of the mixture, this normal hexane treatment process was repeated for 1–5 cycles. Visual observation of the foam generation revealed minimal foaming and rapid dissipation after four and five cycles. Therefore, the experiment proceeded with four cycles as the standard. The recovered TPA (rTPA) produced under this process was labeled according to the number of normal hexane treatment cycles and activated carbon treatment time: rTPA-NH4c-1AC, rTPA-NH4c-3AC, rTPA-NH4c-6AC, rTPA-NH4c-12AC, rTPA-NH4c-24AC, and rTPA-NH4c-48AC.

2.2 Activated carbon manufacturing and analysis

The activated carbon used for impurity removal was prepared by refining PFO to produce pitch. The pitch was stabilized under air atmosphere at a heating rate of 5°C·min⁻¹, operating temperature of 320°C, and operating time of 3 h. Subsequently, carbonization was performed under high-purity N₂ gas atmosphere at a heating rate of 5°C·min⁻¹, operating temperature of 880°C, and operating time of 1 h. Activation was then carried out as a continuous process under steam and a high-purity N₂ gas atmosphere at 880°C for 3 h. The specific surface area and micropore volume of the prepared activated carbon were analyzed using ASAP2460 (Micromeritics). Scanning electron microscopy (SEM) was performed using NX2000 (Hitachi) to examine the surface and pore structure of the activated carbon and 3D CT analysis was conducted using Skyscan 2214 (Bruker).

2.3 Activated carbon manufacturing and analysis methods

To compare the purity of TPA produced with and without normal hexane treatment, TPA samples were prepared by subjecting the provided mother liquor 300 mL to 10 wt% activated carbon treatment for 1, 3, 6, 12, 24, and 48 h. These samples were designated as rTPA-1AC, rTPA-3AC, rTPA-6AC, rTPA-12AC, rTPA-24AC, and rTPA-48AC, respectively. Nuclear magnetic resonance (NMR) analysis was performed using a 400-MR DD2 (Agilent) with dimethyl sulfoxide as the solvent for 1H proton spectrum analysis. Fourier transform infrared spectrometer (FT-IR) analysis was conducted using an FT-IR 6100 (Jasco) with attenuated total reflectance mode in the range of 400–5,000 cm⁻¹. UV analysis was performed using a UV-1900i (Shimadzu) to measure the alkali transmittance at 340 nm. The test solution was prepared by dissolving 0.1 g of the sample in a 250 mL volumetric flask with 3 mL of ammonia solution and 7 mL of purified water, followed by complete dissolution and standing at room temperature. Distilled water was then added to make up the volume to 250 mL. The blank test solution was prepared in the same way without the sample. The test was conducted by referring to the ASTM D 8130-17 and ASTM D 7884-20 standards.

The acid value analysis was carried out according to the ASTM D 8031-20 standard. Briefly, 1 g of the sample was placed in a 250 mL flask and mixed with 20 mL of dimethyl sulfoxide. The mixture was stirred until the sample was completely dissolved. Then, 20 mL of purified water and 0.1 mL of phenolphthalein indicator solution (0.1 g of phenolphthalein powder dissolved in 100 mL of ethanol) were added. The solution was titrated with standard 0.5 M NaOH solution using a stirrer until a pink endpoint persisted for about 15 s. The volume of 0.5 M NaOH used was recorded. The acid value was calculated using the following equation (Eq. 1):

where A is the volume (mL) of the NaOH standard solution used for sample titration, B is the volume (mL) of the NaOH standard solution used for blank titration, C is the concentration (mol·L⁻¹) of the NaOH standard solution, f is the concentration correction factor of the NaOH standard solution, M is the molar mass of KOH (56.11 g·mol⁻¹), and W is the sample weight (g).

3.1 Activated carbon performance analysis

The activated carbon used after the normal hexane treatment was derived from petroleum refining byproducts. This pitch-based activated carbon was manufactured through a series of processes, including crude purification, stabilization, carbonization, and activation (11,12). The specific surface area and micropore volume of the activated carbon were determined by BET analysis to be 1,974 and 1,222 m²·g⁻¹, respectively. The N₂ gas adsorption–desorption isotherm and pore size distribution are shown in Figure 2. Additionally, SEM and 3D CT analyses were conducted to visualize the pore distribution, revealing a widespread distribution of micropores and mesopores both on the surface and within the interior of the pitch-based activated carbon (Figure 3).

Figure 2

(a) N₂ adsorption–desorption isotherms and (b) BJH adsorption–desorption pore characteristics of the referenced activated carbon.

Figure 3

Analysis of referenced activated carbon: (a) SEM 1.0 K, (b) SEM 10.0 K, (c) 3D CT external analysis, and (d) 3D CT internal analysis.

3.2 TPA purity and characterization

Impurities were removed from each sample through normal hexane and conventional activated carbon treatment processes tailored to the specific preparation conditions. Following impurity removal, rTPA was precipitated and separated using a 35% HCl solution at pH 2. The NaCl formed during this process was dissolved using distilled water and then removed and separated using a 0.45 μm membrane filter. The purity of the obtained rTPA was analyzed using NMR, yielding the results shown in Figure 4. After removing the H₂O peak at 3.3 ppm and the dimethylsulfoxy peak at 2.5 ppm, the area integrals of the remaining peaks were applied to Eq. 2 to calculate the P _sample:

Here, I represents the integrated area, H denotes the number of protons, m signifies mass, M stands for the molecular weight, and p indicates purity. The TPA purity ranged from 98.28% to 99.60% when only activated carbon treatment was applied, reaching its maximum at 48 h. However, the purity was consistently higher (99.67–99.81%) when the n-hexane treatment was introduced, demonstrating its superior performance. Furthermore, four cycles of n-hexane treatment reduced the impurity peaks corresponding to 0.0–1.5 and 4.5–5.75 ppm, resulting in the highest overall purity.

Figure 4

NMR analysis of samples processed with activated carbon treatment is only listed in (a) rTPA-1AC, (b) rTPA-3AC, (c) rTPA-6AC, (d) rTPA-12AC, (e) rTPA-24AC, and (f) rTPA-48AC. Samples processed with an additional normal hexane treatment are listed in (g) rTPA-NH4c-1AC, (h) rTPA-NH4c-3AC, (i) rTPA-NH4c-6AC, (j) rTPA-NH4c-12AC, (k) rTPA-NH4c-24AC, and (l) rTPA-NH4c-48AC.

Based on NMR analysis, the energy consumption of the lab-scale stirring process (MS300HS, Misung Scientific) was 25 W, and the mixing and separation process using a Funnel Shaker (C-SKR, Changshin Science) was 200 W. Applying these values and process times to Eq. 3, the energy consumption for the activated carbon treatment alone was 1.20 kW·h, while the introduction of the normal hexane treatment increased energy consumption to 1.33 kW·h. However, considering TPA purity, the energy consumption of 99.60% pure rTPA-48AC was 1.20 kW·h, whereas the energy consumption of 99.67% pure rTPA-NH4c-1AC was significantly lower at 158.40 W·h due to a substantial reduction in activated carbon treatment time. The energy consumption of rTPA-48AC was approximately 7.6 times higher than that of rTPA-NH4c-1AC, suggesting that the introduction of the normal hexane treatment process is suitable. Additionally, the normal hexane can be recycled and reused in the process (13):

Here, E represents electrical energy, measured in joules (J), P denotes power, quantified in watts (W), and t signifies time, expressed in seconds (s). This equation elucidates that the electrical energy consumed or produced is directly proportional to both the power utilized or generated and the duration for which this power is sustained.

To conduct an accurate qualitative analysis, we performed FT-IR spectroscopy on three samples: reference sample, rTPA-48AC (produced by treating the mother liquor with activated carbon only), and rTPA-NH4c-48AC (treated with normal hexane before activated carbon treatment, showing the fewest impurity peaks). As shown in Figure 5 and Table 1, the peaks of the reference sample and rTPA-NH4c-48AC were largely consistent. However, in the FT-IR spectrum of rTPA-48AC, a unique peak appeared at 3,400 cm⁻¹. This peak was identified as a primary amine, likely formed as a byproduct during the NaOH hydrolysis of amide-containing spinning oils present in the waste fibers. Ideally, such impurities should be completely removed or adsorbed by the activated carbon treatment. However, in the sample without normal hexane pretreatment, it is presumed that the pores of the activated carbon were blocked or clogged by oils, fatty acids, and hydrocarbons, hindering the adsorption of the amine group. Normal hexane treatment dissolves or separates surfactants, hydrocarbons, oils, and fatty acids, either directly or through the action of surfactants present in the mother liquor (14). Surfactants form micelles due to their dipole moment, with the hydrophobic parts forming the core and the hydrophilic parts adsorbing to the water interface. In this structure, the hydrophobic parts react with normal hexane, leading to phase separation. Therefore, our results confirm that surfactants, spinning oils, fatty acids, and oil components in the waste fiber hydrolysate and the wastewater from the fiber heat-setting process were not completely removed by activated carbon treatment alone. The normal hexane treatment process effectively reduced the impact of impurities on activated carbon, leading to the complete removal of the primary amine component.

Figure 5

FT-IR results of the reference sample, rTPA-48AC, and rTPA-NH4c-48AC.

To assess impurity levels, the alkali transmittance of rTPA at 340 nm was compared using a UV spectrophotometer. The transmittance was analyzed for the reference sample, rTPA-48AC, and rTPA-NH4c-48AC. As shown in Figure 6, the reference sample exhibited a transmittance of 86.429%, while rTPA-48AC and rTPA-NH4c-48AC showed 79.062% and 82.746%, respectively. In UV analysis, dissolution is typically achieved using NaOH solution or aqueous ammonia. This study utilized aqueous ammonia, potentially impacting solubility. To address this, a comparative analysis with reference samples was conducted under identical test conditions for validation. Although all samples displayed lower transmittance values compared to the reference, the rTPA-NH4c-48AC sample demonstrated a transmittance closer to that of the reference, suggesting a higher purity level. The rTPA-NH4c-48AC samples treated with normal hexane exhibited values similar to the reference sample (Table 2). However, assuming an error range of ±1 for the reference sample, rTPA-48AC also displayed approximate values, suggesting a minimal difference in purity. Nevertheless, in the pursuit of high-purity TPA, an error range of ±1 can significantly impact the attainment of high purity.

Figure 6

Analysis of alkali transmittance (340 nm, %) of the reference sample, rTPA-48AC, and rTPA-NH4c-48AC via UV spectroscopy.