Research on the Chlorine Removal and Upgrading of Waste Plastic Pyrolysis Oil Using Iron-Based Adsorbents

Abstract

The emergence of plastics as an essential item in modern society has led to the problem of accumulating plastic waste. Accordingly, research is being conducted around the world to reduce the production of new plastics and develop technologies to recycle waste plastics. Among the existing waste plastic recycling technologies, oil production is possible through pyrolysis, but the pyrolysis oil produced in this way has a wide carbon range (more than C5–C25), and a very high olefin content (the presence of aromatic compounds), and the resulting high calorific value of pyrolysis oil is limited in its application range. In the case of oil obtained by pyrolyzing waste plastic containing Cl, there is a concern about corrosion in the reactor. Accordingly, it is possible to diversify the range of use of pyrolysis oil produced by suppressing corrosion through Cl removal as well as oil upgrading through cracking. Therefore, this study used red mud mixed with a series of adsorbents for Cl removal and pyrolysis oil upgrade. The adsorbent was physically mixed with a binder (kaolin or methylcellulose) and activated carbon, and the results before and after the reaction were confirmed through basic characteristic analysis.

1. Introduction

Accumulation of plastic waste is one of the main causes of environmental pollution; therefore, reducing plastic waste through recycling efforts can help protect the natural ecosystem and reduce environmental pollution. Additionally, plastic manufacturing processes, such as raw material extraction, processing, and transportation, consume energy and emit greenhouse gases; however, proper recycling can help reduce energy consumption and mitigate emissions. Moreover, recycling plastics produced from petroleum can help replenish petroleum, which is a finite resource.

Oil production through thermal decomposition of plastic waste is an important technology for the production of raw chemical materials. The produced oil, which has a high calorific value, is used as boiler fuel to a limited extent. However, its application is limited by its wide carbon range (C5–C25 or higher) and very high olefin content (aromatic compound presence); therefore, technological upgrades are needed to broaden the application potential [1,2]. They can be converted into fuels and raw materials. When pyrolysis oil is produced, the chlorine content depends on the type of plastic waste. If the Cl is not removed, it can deactivate catalysts during the upgrading process and corrode the reactors, piping, and heat exchangers. Therefore, the removal of chlorine is very important for the utilization of pyrolysis oil [3].

Red mud (RM), a byproduct of the bauxite process in the aluminum industry, is a solid residue consisting mainly of Fe₂O₃ and Al₂O₃. RM can be used as a potential catalyst for biomass pyrolysis [4,5] or oil upgrading [6] because of its large surface area, resistance to sintering and poisoning, and low cost. Although RM is a solid waste, it can be applied in various processes, including water treatment (removal of nitrate, phosphate, and heavy metals) [7,8,9], building materials [10], SO₂ sorbents [11], H₂S adsorbents on high-temperature gases [12], and hydrodechlorination [13]. The catalytic properties possessed by the metal oxides contained in red mud lowered the activation energy of pyrolysis reactions, thereby accelerating the reaction rate [14], and the oil yield was improved using biomass pyrolysis [15]. Further, Liu et al. reported that the addition of RM during the pyrolysis of petroleum-contaminated soil significantly reduced the contents of saturates, aromatics, resins, and asphaltenes [16].

According to previous proximate analyses, raw pyrolysis material consists of 7.3 wt% ash and 92.7 wt% volatile matter, and the impurities in the ash include 0.03 wt% S and 0.17 wt% Cl; therefore, sulfur and chlorine may be present in the oil produced from pyrolysis. Because pyrolysis oil is mostly a heavy oil composed of olefinic hydrocarbons, it is important to upgrade light oil to paraffinic hydrocarbons. Therefore, in this study, we removed S and Cl from pyrolysis oil and upgraded the oil by adding red mud, which contains the iron oxide and other additives (activated carbon, kaolin, and methylcellulose) required to improve adsorption performance.

2. Materials and Methods

2.1. Experimental Apparatus

Pyrolysis, chlorine removal, and upgrading of waste plastic were conducted in a reactor with a diameter of 3 inches and a height of 1.5 m. The bottom of the reactor was filled with 650 g of waste plastic flakes, and 130 g of red-mud-based molded adsorbent was placed at the top of the reactor. Figure 1a shows a schematic of the experimental device seen from the front, and Figure 1b is a photograph of the actual experimental device used for waste plastic pyrolysis. The oil vapor generated by the pyrolysis of plastic waste in the lower part of the reactor passes through the red mud layer, where chlorine removal and upgrading occur simultaneously. Subsequently, the oil is condensed and recovered in the condenser at the rear of the reactor, and the uncondensed oil vapor is discharged as gas.

A reactor containing waste plastic flakes and the adsorbent was installed in a rectangular electric furnace, and the temperature was maintained by heating the three sections separately. In this study, pyrolysis oil was produced using waste vinyl. The collected waste vinyl was crushed to a uniform size and fed into the pyrolysis reactor, and pyrolysis was conducted at 420 °C under oxygen-free conditions [17]. During this process, when polymer chains are broken by thermal energy, they decompose into low molecular weight substances such as gas, solids (residue, wax), and liquid (pyrolysis oil) [18,19,20]. The pyrolysis gas is converted to pyrolysis oil through a condensation process. In this study, waste vinyl was used, which may contain chlorine-containing substances in the raw material structure and may exhibit high chlorine content due to impurities not being completely removed during the pretreatment process.

Experiments conducted using the pyrolysis reactor shown in Figure 1 confirmed that wax components with boiling point of 450 °C or higher were removed in the pyrolysis reactor. Based on the results, subsequent experiments were conducted with the reaction temperature fixed at 420 °C. The lower and middle parts of the reactor were heated to 420 °C at a heating rate of 3.25 °C/min and maintained for 20 h for the pyrolysis of waste plastic flakes, and the upper part of the reactor was maintained at 350 °C to remove chlorine and upgrade the oil. The oil was subsequently condensed and recovered.

The adsorbent was physically mixed with red mud (KC Co., Ltd., Yeongwol, Republic of Korea) and additives (kaolin, methylcellulose, and activated carbon) in three different ratios (RM1~RM3), as shown in Figure 2. Kaolin and methylcellulose, which acted as binders to obtain spherical RM adsorbents, were physically mixed with methylcellulose for uniform mixing with activated carbon to increase the surface area. RM1, RM2, and RM3 were each molded into a 5 mm sphere and then calcined at 580 °C.

A pyrolysis raw material was used for composite waste vinyl chips, which were extruded by heat treatment at 200–250 °C after pulverizing, washing, and removing moisture from PE (polyethylene) and PP (polypropylene) series household waste vinyl. The pyrolysis oil obtained from the composite waste vinyl chips without red mud is denoted as CV, and the pyrolysis oil obtained using RM1 to RM3 is denoted as CV_RM1 to CV_RM3.

2.2. Characterization of Pyrolysis Oil and RM Adsorbents

The physical and chemical properties of the RM adsorbents were analyzed as follows. First, their chemical compositions were confirmed using X-ray fluorescence spectrometry (XRF, ZSX Primus, Rigaku); next, the acidity of the adsorbent was evaluated using NH₃-temperature programmed desorption (NH₃-TPD) with a BELCAT-M instrument. To remove the adsorbed water from the adsorbent, NH₃ molecules were adsorbed at 100 °C after pretreatment in a He flow. To accurately determine the acid site, physically adsorbed NH₃ molecules were removed with He, and then NH₃-TPD analysis was performed in the range of 100–800 °C with a ramping rate of 10 °C/min.

An X-ray diffraction (XRD) analysis and Brunauer–Emmett–Teller (BET) method were performed to examine the structural and pore stability of the thermal decomposition reaction before and after the reaction. To confirm the bonding structure of the oxide material, it was scanned in the range of 10–80°, at a scanning rate of 5°/min, using an XRD (Multi-purpose X-ray diffractometer, Smartlab, Rigaku, Tokyo, Japan) analysis device equipped with a Cu Kβ X-ray beam. In addition, the specific porosity of the RM adsorbents was measured using the N₂ adsorption-desorption method with a Micromeritics Tristar II instrument and the BET equation. In addition, scanning electron microscopy with energy dispersive X-ray spectrometry (SEM-EDX, Mira 3, Tescan, Brno, Czech Republic) analysis of the RM adsorbents was performed before and after pyrolysis to observe the surface reactivity.

The characteristics of the pyrolysis oil were analyzed based on the degree of cracking caused by the RM adsorbent, impurities, and chlorine content in the pyrolysis oil. Basic density measurements (ASTM D4052-22) and high-temperature simulated distillation (SIMDIS, Ultra Sys, Thermo Electron) analysis were performed to confirm the degree of cracking. The boiling points of the hydrocarbons in the pyrolysis oil were measured, according to the number of carbons, using SIMDIS (ASTM D2887-22, Ultra Sys, Thermo Electron, Waltham, MA, USA), and GC-MS (capillary column Agilent 19091 J-413 HP-5 5% phenylmethylsiloxane, Agilent 5973, Agilent Technologies, Santa Clara, CA, USA). Finally, hydrocarbon types were determined.

The chlorine content of the pyrolysis oil was analyzed using a chlorine content analyzer (ESC3000, Thermo Fisher Scientific, Waltham, MA, USA) based on the Korea Petroleum Quality & Distribution Authority test method (KS M 2457:2003). Impurities in the pyrolysis oil (CV, CV_RM1–CV_RM3) were measured using inductively coupled plasma (ICP) atomic emission spectroscopy (OPTIMA 2100DV, Perkin Elmer, Waltham, MA, USA) and residual carbon (Micro method, ACR-M3, Tanaka, Singapore). The concentrations of Cd, Cr, Pb, and as were all below 1 ppm, and the residual carbon was <0.03 wt%, indicating little residual carbon.

3. Results

3.1. Characterization of RM Adsorbent

The chemical composition of the RM adsorbent according to the mixing ratio was confirmed using XRF (

Table 1. Chemical contents of RM adsorbents.

	RM1	RM2	RM3
Fe	49.4	47.9	48.2
Al	14.2	14.6	14.0
Si	7.9	10.2	8.0
Ti	8.6	8.4	8.4
Ca	9.8	9.9	9.4
Na	5.0	5.1	5.0
C	2.0	1.9	4.0
Other	3.1	2.0	3.0

). Since red mud accounts for about 83–95% of the adsorbent composition, iron oxide is predominant. Moreover, the Al and Si compositions increased in RM2, which contained a large amount of kaolin, the main components of which are alumina and silica. Finally, RM3, which contains activated carbon, contains approximately twice as much C as that of the other adsorbents. There are slight differences in concentrations of other materials, such as Ti, Ca, and Na, but this may occur depending on the process variables in which the red mud is generated [21].

The characteristic crystal structure of the red mud was confirmed to have peaks of iron oxide (H: hematite, S: sodalite, and G: goethite) [22,23,24]. RM2 contains the most kaolin, showing a high kaolinite peak (K) [22], whereas in RM3, the calcite (C) peak appears stable [25].

The BET analysis of RM adsorbents, as shown in

Table 2. Porous characteristics and acidity of RM adsorbents.

Adsorbents		RM1	RM2	RM3
BET	Surface Area (m2/g)	19.72	20.49	23.92
	Pore Volume (cm3/g)	0.03	0.03	0.04
	t-Plot Micropore Area (m2/g)	1.3508	1.2890	3.5703
	BJH Desorption cumulative volume (cm3/g)	0.03188	0.03344	0.03954
	Pore Size (nm)	6.13	6.14	6.52
Acidity (mmolNH3/g)	Weak (<300 °C)	0.01	0.03	0.01
	Strong (>500 °C)	0.11	0.12	0.37
	Total	0.12	0.15	0.39

, confirmed that the combination of kaolin and methylcellulose binder with activated carbon improved the specific surface area and pore volume of the adsorbents, of which RM3 showed the best overall textural properties and increased adsorption capacity owing to the addition of activated carbon. Although RM1 and RM2 had similar pore sizes, the overall properties of RM3 were better owing to the influence of the mixed activated carbon.

Differences in the adsorbent mixing ratios were also observed in the NH₃-TPD results. Acid sites are largely divided into weak acid sites at temperatures below 300 °C, medium acid sites at temperatures between 300 and 500 °C, and strong acid sites at temperatures above 500 °C [26,27]. The acid sites of the RM adsorbent in the present study only exhibited weak and strong acid sites, which are listed in Table 2. As the proportion of kaolin in the adsorbent increased, the proportion of weak acid points tended to increase, and when activated carbon was added, strong acid points appeared relatively high.

3.2. Upgrading Effect of Red Mud on Pyrolysis Oil

The thermal decomposition results obtained from 650 g of composite waste vinyl chips showed no wax products, and since the pyrolysis reaction of raw materials occurs under the adsorbent layer, no difference was observed in the amount of char recovered from the lower part of the reactor regardless of the type of adsorbent used. The oil yield was calculated by comparing the weight of the pyrolysis oil to that of the composite waste vinyl chips. Approximately 520 g of pyrolysis oil was obtained when the RM adsorbent was not used (CV). However, when the pyrolysis reaction was performed using the RM adsorbent series, approximately 45–55% of the pyrolysis oil was obtained of the weight of the feedstock (Figure 3). This is because RM adsorbents affect catalytic cracking in the pyrolysis reaction, causing pyrolyzed hydrocarbons from composite waste vinyl chips to light and heavy hydrocarbons, and light hydrocarbons of C1–C4 are vented into gas, resulting in differences depending on the mixing ratio of RM adsorbents. Differences in the degree of cracking occurrence were confirmed in the density of each pyrolysis oil, with a similar trend to that of the oil yield, wherein CV (776.5 kg/m³ > CV_RM1 (767.5 kg/m³) ≒ CV_RM3 (766.9 kg/m³) > CV_RM2 (764 kg/m³).

The SIMDIS analysis results, depicted in Figure 4, confirmed that CV samples with high densities tended to volatilize in the overall high-temperature range. The naphtha content below boiling point (200 °C) was low at about 40%, and that above 300 °C was high at about 20%. However, as the oil vapor generated from the thermal decomposition of the composite waste vinyl chips passed through the RM adsorbent layer, the boiling point distribution significantly decreased. Owing to the cracking of the pyrolysis oil, the naphtha content increased to 52–65%, and the Bunker Oil content decreased to less than 10%. More specifically, CV_RM1 and CV_RM3, which have similar kaolin contents, showed similar boiling point graphs, whereas cracking progressed more actively as the kaolin content increased to 10%, and as a result, the pyrolysis oil of CV_RM2 had a very high naphtha ratio of 65%. Accordingly, we found that the effect on cracking was largely due to the weak acid sites rather than the strong acid sites or the surface area and overall acidity of the adsorbent.

GC-MS analysis (Figure 5a) was performed to confirm the detailed properties of these pyrolysis oils. Pyrolysis oil obtained from composite waste vinyl-based chips contained a wide range of 40 to 50 hydrocarbon components from C5 to C20. In the composition of pyrolysis oil (Figure 5b), hydrocarbons with C5–C12 carbon atoms were most distributed, and paraffin was the most distributed among hydrocarbons. The pyrolysis oil CV without using RM adsorbent exists at 42.03% and 33.05% of light to middle oil, respectively, and the proportion of heavy oil above C19 is very high at 24.92%. The GC-MS results of CV_RM1, CV_RM2, and CV_RM3, the C5–C12 of carbon number increased to more than 70% owing to cracking. In particular, CV_RM2, which has the highest kaolin content, has the highest C5–C12 hydrocarbon in the light naphtha range at 78.57%. It can be confirmed once again that cracking actively occurred due to the weak acid point present in the RM2 adsorbent. In terms of the structure of hydrocarbons, the presence of RM adsorbent reduced the paraffin content from 59.58% to 43–48%, and overall cyclic hydrocarbon increased, but olefin increased only in the result of CV_RM1. It can be seen that the chain is broken due to simple pyrolysis, the adsorbent cracks are due to light hydrocarbon, and double bonds or cyclization occur in the internal pores.

It was confirmed that the pore stability influenced pyrolysis through the pore analysis of RM adsorbents after the reaction. Considering the porous characteristics of the RM adsorbents after the pyrolysis reaction (

Table 3. Porous characteristics of RM adsorbents after pyrolysis.

	CV_RM1	CV_RM2	CV_RM3
Surface Area (m2/g)	5.8026	5.5067	8.6391
Pore Volume (cm3/g)	0.0062	0.0059	0.0113
t-Plot Micropore Area (m2/g)	0	0	0
BJH Desorption cumulative volume (cm3/g)	0.00622	0.00592	0.01128
Pore Size (nm)	4.9576	5.5795	6.3015

), CV_RM2, which obtained oil cracked with less hydrocarbon, decreased surface area by 73% and pore volume by 82% after the reaction, showing the largest decrease in the overall porous property. On the other hand, in the case of CV_RM3, which has hydrocarbons of C19 or more in pyrolysis oil, the reduction rates of the surface area and pore volume were 64% and 71%, respectively, showing a big difference from the other two adsorbents.

The difference between before and after the reaction (Figure 6) also appeared in the structural characteristics. In the case of RM1 and RM3, in which less kaolin was mixed, relatively large structural changes did not occur depending on the pyrolysis reaction. On the other hand, the decrease in (K) or (C) peak after RM2 was used in the pyrolysis reaction indicates that structural deactivation occurred due to a lot of cracking [28,29,30,31,32].

3.3. Chlorine Removal Effect of Red Mud in the Production of Pyrolysis Oil

The chlorine content of pyrolysis oil was measured to observe the effect of the RM adsorbent on chlorine removal performance. The chlorine removal rate was calculated using Equation (1) below, based on the chlorine content of the CV with no adsorbents.

$$Clremovalrate\left[\%\right]={\displaystyle \frac{ClinpyrolysisoilwithoutRMadsorbent-ClinpyrolysisoilwithRMadsorbent}{ClinpyrolysisoilwithoutRMadsorbent}}$$

(1)

The measured chlorine content (%) and calculated chlorine removal rates of the pyrolysis oil samples are summarized in Figure 7. The CV oil contained 115 ppm Cl, and the chlorine content was reduced by more than 45% as it passed through the RM adsorbent layer. The chlorine content of CV_RM3 was 28 ppm, and the chlorine removal rate was the highest at 75.65%. In the case of RM3, the activated carbon in the adsorbent improved the adsorption capacity by improving the surface area, which seems to have directly affected the Cl removal effect.

During the pyrolysis of waste vinyl-based chips, chlorine present in the raw material is released as hydrogen chloride (HCl) gas, which is subsequently captured by various metal oxides contained in the red mud (RM) adsorbent, including Fe₂O₃, CaO, and MgO. The adsorption process involves the conversion of these oxides into chloride compounds, while oxygen atoms are simultaneously reduced to water (H₂O) [33,34]. SEM-EDX analysis conducted to examine the elemental distribution on RM adsorbent surfaces before and after the pyrolysis reaction confirmed that oxygen content decreased while chlorine content increased on all adsorbent surfaces; however, significant variations were observed in the rates of surface oxygen reduction and chlorine accumulation among different adsorbents (Figure 8). RM3, which achieved the highest chlorine removal rate, exhibited distinct characteristics where oxygen remained present on the surface after the reaction and the rate of chlorine accumulation was the lowest among all tested adsorbents. In contrast, RM2 demonstrated the poorest chlorine removal performance, with no oxygen detected on the surface after the reaction and a high rate of chlorine accumulation observed. Post-reaction pore analysis (Table 3) revealed that RM2 exhibited the greatest decrease in pore characteristics, indicating substantial structural degradation, while RM3 maintained relatively high structural stability with minimal changes in pore properties. During the thermal decomposition process, oil vapor passes through the internal pores following initial surface reactions, facilitating chlorine adsorption onto metal oxides and promoting catalytic cracking reactions. RM2 exhibited a high cracking occurrence rate; however, due to its small internal pores, surface reactions were predominantly enhanced, resulting in rapid reduction of surface oxides and oxygen, increased chlorine adsorption concentrated on the surface, and accelerated surface deactivation. Conversely, RM3 demonstrated superior performance due to its structural characteristics, where minimal pore structure changes maintained reaction accessibility, chlorine adsorption occurred throughout the internal structure rather than merely on the surface, and the highest chlorine removal rate was achieved through distributed reaction sites. Despite high chlorine removal efficiency, oxygen remained on the RM3 surface with reduced chlorine accumulation, indicating sustained catalytic activity. The superior performance of RM3 can be attributed to the beneficial effects of activated carbon, which provides enhanced surface area for increased availability of reaction sites, improved structural stability to prevent pore collapse and maintain internal reaction pathways, and reduced surface deactivation to preserve catalytic activity throughout the reaction process. This study demonstrates that effective chlorine removal in waste vinyl pyrolysis is critically dependent on the adsorbent’s pore structure and stability, where RM3’s combination of adequate pore accessibility and structural integrity enables distributed chlorine adsorption throughout the material, resulting in superior performance compared to RM2’s surface-limited reaction mechanism.

4. Conclusions

This research compares the properties of the pyrolysis oil of composite waste vinyl chips according to the type of adsorbent mixed with red mud. The cracking behavior was significantly influenced by the presence or absence of an adsorbent. The pyrolysis oil obtained from the pyrolysis reaction conducted without an adsorbent contained approximately 58% hydrocarbons with a carbon number of C13 or higher, mostly in the form of paraffin. In contrast, when pyrolysis oil was collected after passing through the adsorbent layer, the number of light hydrocarbons discharged as gas increased, and the composition of the pyrolysis oil showed a high content of hydrocarbons in the C5–C12 range, exceeding 70%. As the oil vapor generated from the thermal decomposition of the composite waste vinyl chips passed through the adsorbent layer, cracking occurred due to the pore characteristics and acid sites of the adsorbent. The strength of the acid sites of the adsorbent varies depending on the kaolin content, and the influence of weak acid sites was significant under the thermal decomposition and Cl adsorption conditions below 500 °C. The observed trend showed that more weak acid sites produced lighter hydrocarbons. Moreover, the amount of oil obtained was small compared to the weight of the raw materials supplied, indicating that a large number of hydrocarbons (C5–C12) were present in the obtained oil. Although cracking occurred, cyclization also occurred due to the influence of the adsorbent.

Regarding the Cl removal performance of the RM adsorbent, the Cl content of the pyrolysis oil obtained without the adsorbent was 115 ppm, while the adsorbent demonstrated a Cl removal efficiency of more than 45%. Significant differences were observed in the Cl removal performance of each adsorbent based on the degree of cracking. When the pyrolysis oil composition consists of lighter hydrocarbons, the adsorbent experiences a decrease in surface area for Cl adsorption due to the reduction of internal pores, resulting in increased surface dependence; consequently, it was found that the structural stability of the adsorbent is important to achieve both upgrading and Cl removal effects to enable waste plastics to be utilized as alternative resources.

Future studies will investigate the individual characteristics of each additive to better understand their specific effects on the pyrolysis process and chlorine removal performance.