Elimination of Odors Emitted from Hot-Melting of Recycle PS by Oxidative-Reductive Scrubbing

Hot-melt granulation of waste plastics emits fumes with unpleasant plastic-burning odor during the heating process. This study uses sodium hypochlorite (NaOCl) solution to scrub and oxidize the volatile organic compounds (VOCs) and the associated odors from polystyrene (PS) hot-melting exhaust. An alkaline hydrogen peroxide (H2O2) solution was followed to remove Cl2 emitted from the oxidative solution. Laboratory scrubbing bottles with a liquid volume of 600 mL were used with the hot melt gas injected into the system at a rate of 1 L/min at 25°C. Results indicate that by purging the test gas through a solution with an available chlorine of 300–370 mg/L at an adjusted pH (6.5–7.0) for oxidation of the absorbed odorous compounds, and then through a solution with pH > 12.0 and 0.35% H2O2 for absorption and reduction of Cl2 (< 11.2 ppm) in the exhaust gas from the oxidation liquid, around 90% of the VOCs in the range of 3.2–55.5 ppm (expressed as methane equivalent) was removed. Estimations indicate that it requires around 1.62, 0.0452, and 0.0827 kg NaOCl solution (12% available Cl2), H2O2 solution (35% H2O2), and sodium hydroxide solution (45% NaOH), respectively, for scrubbing 1,000 m of the exhaust gas with 10 ppm NMHC (expressed as methane equivalent). The chemicals cost a total of approximately USD 0.32. The study has developed a new, effective and economic process for reducing odorous compounds in the hot-melted gas. When commercialized the process is hoped to decrease a lot of odorous complaints.


INTRODUCTION
Sources of odor generated in plastic recycling processes mainly come from fumes produced by the hot (150 to 250°C) extrusion granulation process.Field experiences indicate that nylon recycling produces the smells of ammonia and burnt plastics and polypropylene (PP) plastic recycling produces mainly the smell of acrolein from oxidation of propylene on pyrolysis of PP.Based on a study by Tsai et al. (2009), melting of polyethylene (PE)/PP/polyvinyl chloride (PVC) waste plastics produces odor downwind of a plant's perimeter at concentrations of 100 to 229 (dimensionless, dilution times to odor threshold), which is higher than the regulation upper limit of 50 for industrial area set by the Environmental Protection Administration of Taiwan.
Related studies have been done by some researchers.Yamashita et al. (2009) showed that hot-melt virgin PS emits aromatic hydrocarbons and aldehydes as main organic compounds as shown in Table 1.All the emitted volatile organic compounds (VOCs) shown in Table 1 are harmful to the plant's workers and the neighboring inhabitants and the associated odor may raise public complaints.Other studies have also reported various VOC emission factors or compounds (Hoff and Jacobsson, 1981;Forrest et al., 1995;Barlow et al., 1996;Adams et al., 1999;Xiang et al., 2002).
As indicated in Fig. 1, for control of the emitted fumes to prevent them from venting to the operation area and the plant's perimeters, emitted fumes should be collected by two hoods, one set at above the gas vent hole located at approximately the middle point of the heated screw smelter and the other at the exit end of the smelter.The collected gas streams are then combined and conducted to a control system.In general, an electrostatic precipitator (ESP) is used to remove the mists in the collected gas and the mist removal efficiency depends strongly on the size and electrostatic conductivity of the mists.The odorous VOCs should further be controlled to eliminate possible public complaints when contact with the vent gas.Chemical scrubbing can be used for the control of waste gases with VOCs.Aqueous sodium hypochlorite was found to be among the most effective oxidants.A few studies focused on the hypochlorite oxidation of aqueous organics such as benzene, toluene, xylenes, phenolates, aldehydes, and ketones.Chungsiriporn et al. (2006) proposed the removal of toluene from waste air using a spray wet scrubber combining the absorption and oxidation reaction.Aqueous sodium hypochlorite (NaOCl) solution was used as the scrubbing liquid in the system.Mirafzal and Lozeva (1998) presented phase transfer catalyzed oxidation of alcohols with sodium hypochlorite in ethyl acetate media that resulted in good to excellent yield of aldehydes or ketones as oxidized products.
Cheng and Hsieh (2010) integrated chemical scrubber with sodium hypochlorite and surfactant to remove hydrocarbons in cooking oil fume.Results indicated that suitable operating parameters of NaOCl scrubber system at pH 6.5, 200 ppm of available chlorine and 11 L/m 3 of liquid/gas ratio were proposed.With the conditions, NMHC (non-methane hydrocarbons) in the cooking fume could be removed from19 ± 13 to 4 ± 2 mg/m 3 (as methane).The study didn't indicate, however, which components in the fume gas were removed or oxidized.Chang et al. (2011) used sodium hypochlorite (NaOCl) solution in series with alkaline hydrogen peroxide (H 2 O 2 ) solution in a two-stage scrubbing process to treat the associated odors and VOCs from PP hot-melting operations.VOCs in the hot melt exhaust consist mainly of acrolein, acetone, 2-butanone, and 4-methyl-2-pentanone. Results indicated that by purging the test gas through a solution with an available chlorine of 15-50 mg/L at an adjusted pH (9.3-10) for absorption and oxidation of the odorous compounds to a degree of around 90%.Following the operation, a solution of 0.35% H 2 O 2 with pH in the rang of 12.0-12.5 was used for absorption and reduction of Cl 2 (< 3.8 ppm) in the exhaust gas from the oxidation liquid.Estimations indicated that the chemicals cost a total of approximately USD 0.25 for treating 1,000 m 3 gas.
To the authors' knowledge, there is no report on the removal of VOCs or odorous compounds from waste polystyrene (PS) smelting and extruding processes.In the present study, chemical scrubbing technology using sodium hypochlorite as an oxidant for VOCs and odors and alkaline hydrogen peroxide as a reducing agent for the emitted chlorine from the oxidation liquid was tried to test its applicability to the odor control of the smelting operations for PS recovery.Emission characteristics, effects of pH and concentration of the available chlorine on the VOC and odor removal efficiencies, and if the fixed pH of 12.5 and hydrogen peroxide concentration of 0.35% could remove emitted chlorine in the exhaust gas from the oxidation liquid will be presented in the present study.

Experimental
Schematics of the experimental systems are shown in Figs. 2 and 3. Fig. 2 shows a gas generation oven (Hipoint  Precision Oven, Type: OV-40, Taiwan) in which 800 g waste PS plastic chips was placed in an aluminum foil plate and the oven heated to around 160-200°C to melt the plastics.Compared the system shown in Fig. 2 with the field system, the screw extrusion operation shown in Fig. 1 gives the melted plastics a shear force that may crack the polymer chains and emit cracked hydrocarbons.In addition, in an extruder pelletizer, the exposure time of the heated plastics to air is reduced as compared with that in the oven shown in Fig. 2. Hence, setup shown in Fig. 2 could only be partially related to Fig. 1 on the electric heating of the plastic flakes.The emitted gas for test was only a simulated one.Experiences showed that it requires around 30 minutes to heat up the oven to the desired temperature.Fume gas generated after a heating time of 60-90 minutes and it was then drawn by a gas pump (Airchek Sampler, Type: 224-PCXR4, USA) operated at 5 L/min and the collected gas stored in two 20-L Teflon bags for each batch of the melted plastics.A gas sample in each bag was immediately drawn for THC (total hydrocarbon) analysis.Before the absorption test, the gas sample in the 20-L bag was transferred by a sampling pump to a 40-L bag followed by introducing 20 L of the ambient air in the laboratory to the bag.The diluted gas was ready for a scrubbing test.
The scrubbing system shown in Fig. 3 consists of a gas pump and two 1000-mL scrubbing bottles.The first bottle (oxidation scrubber) was filled with 600 mL of aqueous sodium hypochlorite solution for absorbing and oxidizing the chemicals from the fume gas at a fixed influent flow rate of 1 L/min.The second one (reduction scrubber) contained 600 mL of an aqueous hydrogen peroxide and sodium hydroxide solution for absorbing and reducing the chlorine gas stripped from the oxidation solution.The associated reductive chemical reaction is: (1) The gas and the scrubbing liquors were all kept at 23-27°C.
In the course of reaction, gas samples were collected by sampling bags for analysis of THC (total hydrocarbons), NMHC (non-methane hydrocarbons), Cl 2 , and odor intensity.Scrubbing liquors were also sampled for analysis of pH and residual chlorine.After each test, initial concentration of sodium hypochlorite or pH was adjusted to test its gas cleaning effect.Hydrogen peroxide solution in the second bottle was kept at an initial concentration of 0.35% and adjusted to pH 12.5 by 25 wt% NaOH solution.Around 34 mL of the NaOH solution or 10 g pure NaOH was required for 1 L of the hydrogen peroxide solution.

Materials
Waste PS plastics (mostly containers for instant foods) used in this study were collected from community recycling stations.The collected plastic material was cut into pieces of approximately 1 × 1 inch in size and water washed.The recycled material contains metal oxide pigments which may catalyze the cracking of polymer chains and oxidation of the cracked hydrocarbons.Actually, some virgin PS pellets are manufactured by blending pigmented and non-pigmented PS pellets before being fed into pellet machines.Therefore, properties of the simulated emitted gas are believed to be similar to those emitted from commercial machines.
All chemicals (aqueous sodium hypochlorite solution with 12 wt% available chlorine, 35 wt% aqueous H 2 O 2 solution, and sodium hydroxide) are reagent grades.

Analytical
THC and NMHC concentrations in gas samples were analyzed with a portable flame ionization detector (FID, Thermo, TVA-1000B, USA).Compositions of VOCs in gas samples were analyzed by a gas chromatography (6890N Series GC System, Agilent, USA) and a mass spectrometry (5973 Network Mass Selective Detector, Agilent, USA).
A chlorine analyzer (ToxiRAE II, 045-0516-000, USA) was used for the analysis of gaseous chlorine concentrations.Colorimetry method (USEPA, 1983) was used for the detection of residual chlorine in the scubbering liquor.A pocker colorimeter (Cat No.58700-00, HACH Co., Japan) was used for the color determination.
The "triangle odor bag method" used in Japan since 1972 was used to evaluate the odor intensity of the gas samples (Iwasaki, 2003).The method is an air dilution approach using statistical methods in which "odor index" is measured.The odor intensity or concentration is expressed as dilution times (dimensionless) to threshold.

Effects of Initial Available Chlorine Concentration on the NMHC Removal
Table 2 shows experimental data for the oxidation scrubbing liquid and those of the calculated specific Cl 2 consumption for the 8 test runs.As shown in the table, initial available chlorine concentrations were varied from 44 to 375 mg/L with adjustment of pH and the initial pH varying from 6.5 to 7.0.NMHC in the influent gas varied from 3.13 to 25.3 ppm (expressed as methane equivalent).Table 3 shows experimental data for the reduction scrubbing liquid and data on the calculated fractional Cl 2 purged from the oxidation scrubbing liquid.
In Tables 2 and 3, the calculated data on the chlorine consumption need to be explained.Chlorine may be consumed due to the oxidation of the influent NMHC and to being purged from the oxidation liquid to the effluent gas.The mass of Cl 2 consumption can be estimated as follows: where Δ[Cl 2 ] L is the difference of available Cl 2 concentrations in the oxidation liquid at the initial and end of the absorption test and 0.60 L is the oxidation liquid volume.
Chlorine being purged from the oxidation liquid may account for a significant part of the consumption.The purged mass can be estimated as follows: Mass of Cl 2 purged (mg) = [Cl 2 ] G (ppm) × [71 (mg/m 3 )/ (24.5 ppm)] × V (m 3 ) (3) where [Cl 2 ] G is the average Cl 2 concentration in the effluent gas from the oxidation liquid at the initial and end of the absorption test, V is the total test gas volume (evaluated at 25 ± 2°C) introduced to the scrubbing system in the test run.The quotient "71 (mg/m 3 )/(24.5ppm)" converts volumetric concentration (ppm) to mass-volume concentration (mg/m 3 ) for gaseous chlorine at the condition of atmospheric pressure and 25°C.
As shown in Table 3, the average Cl 2 concentration [Cl 2 ] G in the effluent gas from the oxidation liquid is equivalent to that in the influent gas to the reduction liquid.[Cl 2 ] G ranged from 2.3-11.4ppm as shown in the table.The values were much higher than those calculated from partition of chlorine in gas and liquid water as described by the Henry's law: where P (bar) is the equilibrium partial pressure of chlorine in the gas phase, C (mol/L) is the equilibrium mol concentration of chlorine in the aqueous phase, and H (mol/L/bar) is the Henry's law constant.At 25°C, H is 0.095 mol/L/bar as given by Lide and Frederikse (1995) and 0.061 mol/L/bar by Wagman et al. (1982).
For estimation of P in the gas phase in equilibrium with the available chlorine concentration in the aqueous phase, consider the following equations: HOCl ⇋ H + +OCl - The equilibrium constants of the above equations at 25°C are: H = C/P = 0.061 or 0.095 M/bar (8) [1] Total test gas volume (evaluated at 25 ± 2°C) introduced to the scrubbing system in the test run.
[2] Available Cl 2 in the oxidation liquid before the absorption test.
[3] Available Cl 2 in the oxidation liquid after the absorption test.
[4] pH of the oxidation liquid before the absorption test.
[5] pH of the oxidation liquid after the absorption test.
[6] Time-averaged NMHC concentration of the influent gas to the oxidation scrubbing liquid.
[7] Time-averaged NMHC concentration of the effluent gas from the oxidation scrubbing liquid.
[8] NMHC removed (%) = (C in -C out )/C in × 100%.[2] Average Cl 2 concentration in the effluent gas from the oxidation liquid at the initial and end of the oxidation test.
[3] Average Cl 2 concentration in the effluent gas from the reduction liquid at the initial and end of the reduction test.
[4] pH of the reduction liquid before the absorption test.
[5] pH of the reduction liquid after the absorption test.
[6] Mass of Cl 2 being purged from the oxidation liquid, as calculated by Eq. ( 3).
[7] Mass of Cl 2 consumed in the in the test run, as calculated by Eq. ( 2).
where K H is the hydrolysis constant of chlorine as given by Greenwood and Earnshaw (1997), and K A is the ionization constant of HOCl as given by Morris (1966).According to the equations, P were estimated to be 3.00 × 10 -9 bar (3.00 × 10 -3 ppm) and 3.79 × 10 -8 bar (3.79 × 10 -2 ppm) with a total available chlorine concentration ([Cl 2(aq) ] + [HOCl] + [OCl -]) of 213 mg/L (10 -3 M) at liquid pH of 7.0 and 6.0, respectively, at 25°C.With a total available chlorine concentration of 2,130 mg/L (10 -2 M) at pH 6, P was estimated to be 3.79 × 10 -6 bar (3.79 ppm) at 25°C.The calculated P values imply that chlorine in the vented gas from the oxidation liquid with an available chlorine concentration of as high as 2,130 mg/L at pH 6 is around 3.79 ppm.The result indicates that chlorine loss from the vented gas should be rather limited.The much higher [Cl 2 ] G (2.3-11.4ppm) in the vented gas from the oxidation liquor might be due to the mist entrained by the bubbles.
According to data listed in Tables 2 and 3, Fig. 4 shows variations of NMHC removal efficiency with the initial available chlorine concentration ([Cl 2 ] o ) of the oxidation scrubbing solution.Results show optimal initial available chlorine concentrations of around 300-370 mg/L for achieving a NMHC removal of around 90%.As shown in Table 3, expect Run.No. 5, the fractional loss of Cl 2 from the purged gas was less than 10.4% and oxidation of the influent NMHC with the [Cl 2 ] o kept in the range of 300-370 mg/L is adequate for practical applications.Data shown in Table 3 also show that conditions with pH in the range of 12.0-12.3and 0.35% H 2 O 2 in the reduction liquid, the effluent gas from the liquid had no detectable chlorine and other odors.Table 4 shows calculated data for obtaining the mass consumption Cl 2 for the oxidation of unit mass of the NMHC.In Table 4, the total mass of NMHC removed from the gas in a test run can be estimated by the following equation: where, Δ[NMHC] G is the average of the difference of NMHC concentrations in the influent and effluent gas at the initial 0.154 5.08 [1] NMHC removed (mg) = Gas NMHC removed (mg/m 3 ) × total test gas volume introduced to the scrubbing system in the test run (m 3 , evaluated at 25 ± 2°C).
[3] Data of Run No. 4 and 8 not included for obtaining the average value.and end of the absorption test.The quotient "16 (mg/m 3 )/ (24.5 ppm)" converts volumetric concentration (ppm) of gaseous NMHC (expressed as methane equivalent) to massvolume concentration (mg/m 3 ) for the gas at the condition of atmospheric pressure and 25°C.The total mass of Cl 2 consumption for the oxidation of the removed NMHC from the gas in a test run can be estimated by subtracting the mass of Cl 2 purged from the oxidation solution by the mass of Cl 2 consumed in a run.By the data, Fig. 5 shows a correlation between the mass of Cl 2 consumption and mass of NMHC removal for the six runs.It seems that Cl 2 consumption decreased approximately with increasing mass of NMHC removal.On an average, as shown in Table 4, 5.08 mg Cl 2 was consumed for the removal of 0.154 mg NMHC, or 33 mg Cl 2 consumed for the removal of 1.0 mg NMHC.

GC-MSD Examination
GC-MSD examination results (Table 5) indicate that acrolein, acetone, benzene, ethyl benzene, styrene and αmethyl styrene were detected in the untreated test gas and some of them could be removed by the oxidation liquid.
Benzene and ethyl benzene were not effectively removed due possibly to that both have a lower water solubility or higher Henry's law constant (dimensionless Henry's law constant m = y/x = gaseous concentration (mg/m 3 )/aqueous concentration (mg/m 3 ) = 0.23 at 25°C) than styrene (m = 0.109) (Eastem Research Group, 1997).
Table 5 also shows that the influent styrene was 1.06 ppm (4.5 mg/m 3 at 25°C) and the total mass of the influent styrene for 30 minutes was 0.135 mg (4.5 mg/m 3 × 1 L/min × 0.001 m 3 /L × 30 min = 0.135 mg).According to the Henry's law, the saturated concentration of styrene in water was around 0.041 mg/L when equilibrium with a styrene concentration of 4.5 mg/m 3 in the gas phase.The oxidation liquid had only a maximum capacity of absorbing 0.0246 mg of styrene (0.041 mg/L × 0.6 L = 0.0246 mg).Accordingly, most of the influent styrene (0.135 mg) had been oxidized rather than absorbed.The most possible one of the oxidation products of styrene by hypochlorous acid may be styrene oxide (Liu et al., 2002).

Olfactory Test
According to the GC-MSD data (Table 5) for the influent and effluent gases to and from the oxidation liquid, the odor threshold of styrene is 35 ppb and the odor concentration of the influent gas was estimated to be 30 which could be easily detected by olfactory sense.The indicator of odor in this work is styrene Take Run 8 as an example, odor intensities of the influent gas could be reduced from 3,090 (expressed as dilutions to threshold) to 55. Pungent burnt plastic odor in the test gas was almost completely removed.The odor intensity (55) of the scrubbed gas meets the regulation of less than 1,000 for exhaust gas from an emission pipe with a height of lower than 18 m above ground, as set by the EPA of Taiwan.

Estimation of Chemical Requirements and Costs
Solutions of sodium hypochlorite solution, hydrogen peroxide, and sodium hydroxide are needed to add to the solutions of the 2-stage chemical scrubbers used in the present study.
Take an emitted gas with 10 ppm NMHC and a NMHC removal of 90% as an example, the volumetric Cl 2 consumption can be estimated to be 1.62 kg 12% chlorine bleaching solution for treating 1,000 m 3 of the gas.According to data shown in Table 4, effluent gas from the oxidation liquid (influent gas to the reduction liquid) has a maximum chlorine of 11.4 ppm, the required chemicals for reduction of the chlorine introduced to the reduction liquid and neutralization of the reduced product (HCl) can be calculated by following the stoichiometry of Eq. (1).Taking 1,000 Nm 3 of the influent gas with 11.4 ppm chlorine to the reduction liquid as a base, mass of Cl 2 = 4.65 × 10 -4 kmol, mass of 35% H 2 O 2 = 0.0452 kg, and mass of 45% NaOH = 0.0372 kg.
Table 6 lists the required amount of chemicals for treating 1,000 m 3 of the emitted melting gas.Based on the local costs for the chemicals, around USD 0.32 are required.The cost analysis in Table 6 addressed only on the required chemical costs for treating 10 3 m 3 of the melting gas.Fixed cost for an application of the technique to treat a full-scale effluent gas should consider sizing of scrubbers for absorbing the pollutants from the gas.The cost depends on the gas flow rate and the required instruments for process control.

Secondary Pollution Control
Replacement of the scrubbing liquors may be once per day or a few days, depending on the pollutant contents in the gas or liquors.The replacement can also be continuously following the continuous operation mode of the hotmelting operations.The chemicals employed in this study including NaOCl, NaOH and H 2 O 2 are all strong oxidizing

CONCLUSIONS
This study used a small-scale two-stage scrubbing method in a laboratory setting to remove odors and VOCs from PS melt exhaust gas.The following conclusions can be drawn: 1.By the oxidation solution with an available chlorine of 300-370 mg/L at an unadjusted pH (6.5 to 7.0), and reduction solution with pH > 12.0 and 0.35 % H 2 O 2 , around 90% of the hot-melt VOCs in the range of 3.2-55.5ppm (expressed as methane equivalent) could be removed.Effluent gas from the system had no detectable chlorine and other odors.2. VOCs in the hot melt exhaust consist mainly of acrolein, acetone, benzene, ethyl benzene, styrene and α-methyl styrene.The two-stage chemical scrubbing process removed 82-100% of the detected VOCs except benzene ethyl benzene.3. It requires around 1.62, 0.0452, and 0.0827 kg NaOCl solution (12% available Cl 2 ), H 2 O 2 solution (35% H 2 O 2 ), and NaOH solution (45% NaOH), respectively, for scrubbing 1,000 m 3 of the exhaust gas with 10 ppm NMHC (expressed as methane equivalent).The chemicals cost a total of approximately USD 0.32.

Fig. 1 .
Fig. 1.Schematic diagram of a plastic smelting and extrusion system, the length of the main smelter (2) is around 4-5 meters.

Fig. 2 .
Fig. 2. Schematic diagram of the waste gas generation and collection facility.

Fig. 5 .
Fig. 5. Correlation between the mass of Cl 2 consumption and mass of NMHC removal for the six runs.

Table 2 .
Experimental data for the oxidation scrubbing liquid and data on the NMHC removed.

Table 3 .
Experimental data for the reduction scrubbing liquid and data on the calculated fractional Cl 2 purged from the oxidation scrubbing liquid.

Table 4 .
Calculated data for the mass of Cl 2 consumption and mass of NMHC removal in each run.

Table 5 .
GC-MSD data for the influent and effluent gases to and from the oxidation liquid.

Table 6 .
Requirement of chemicals and the costs for treating 1,000 m 3 emitted melting gas.Release of these chemical solutions into wastewater treatment plant or aqueous environment may lead to an adverse effect on the biological process of wastewater treatment plant or the ecosystem.Therefore, these chemicals should be appropriately treated before they are discharged into aqueous system.It is suggested that the effluent scrubbing solutions from the oxidation scrubber (mainly HOCl) and the reduction one (mainly H 2 O 2 + NaOH) should be introduced into a mixing chamber and reacted for a definite time for reducing the possible adverse effects.The associated reaction in the mixing chamber is HOCl + H 2 O 2 + NaOH → NaCl + 2H 2 O + O 2 .A reducing agent (e.g, NaHSO 3 ) should be added to the effluent from the chamber to reduce either excess free chlorine (OCl -and HOCl) or H 2 O 2 to either chloride (OCl - + NaHSO 3 → NaCl + HSO 4 -) or H 2 O (H 2 O 2 + NaHSO 3 → H 2 O + NaHSO 4 ).