Environ. Sci. Technol. 2002, 36, 1337-1343 maintain the liquid state). In most of those studies, water Dechlorination of Lindane, Dieldrin, was employed only as an intermediate media, and the Tetrachloroethane, Trichloroethene, degradation was related to reagents such as zerovalent iron or other metals (1-6, 8, 9), a strong base (7), or oxidants (10) and PVC in Subcritical Water which were added to the water. The majority of investigations using pure water to degrade chlorinated organics have been performed using supercritical water (11, 12). For example, ALENA KUBAÄ TOVAÄ , the decomposition of poly(vinyl chloride) at temperatures ARNAUD J. M. LAGADEC, AND ranging from 400 to 600 °C was reported (13). However, the STEVEN B. HAWTHORNE* use of subcritical water may have advantages over super- Energy and Environmental Research Center, critical water (in addition to requiring lower temperatures University of North Dakota, Campus Box 9018, and pressures) in that dechlorination reactions may be Grand Forks, North Dakota 58202 enhanced. For example, the dechlorination of methylene chloride was much faster using subcritical water than when using supercritical water (14). Pure water has been used to dechlorinate aliphatic Recently, pure water at so-called subcritical conditions organics without the need for catalysts or other additives. has been used to extract and degrade high explosives and pesticides from highly contaminated soils (15, 16), and a Dehydrohalogenation (loss of HCl with the formation of a - preliminary study has demonstrated the dechlorination of double bond) occurred at temperatures as low as 105 200 polychlorinated dibenzodioxins on incinerator fly ash (17). °C for 1,1,2,2-tetrachloroethane, lindane (1,2,3,4,5,6- These results are apparently based on the ability of subcritical hexachlorocyclohexane, γ-isomer), and dieldrin water to enhance the solubility of nonpolar organics by as (1,2,3,4,10,10-hexachloro-6,7-epoxy-1,4,4a,5,6,7,8,8a-octahy- much as 5 orders-of-magnitude (18, 19), as well as promoting dro-endo, exo-1,4:5,8-dimethanonaphthalene). Complete some organic reactions (20). loss of the parent compounds was achieved in less than The present study investigates the use of subcritical water 1 h at 150, 200, and 300 °C for 1,1,2,2-tetrachloroethane, lindane, to dehalogenate aliphatic organochlorine compounds, in- and dieldrin, respectively. The initial dechlorination of cluding 1,1,2,2-tetrachloroethane and its degradation product lindane had an activation energy of 84 kJ mol-1 with an 1,1,2-trichloroethene, 1,9-dichlorononane, lindane (1,2,3,4,5,6- hexachlorocyclohexane, γ-isomer), dieldrin (1,2,3,4,10,10- Arrhenius pre-exponential factor of 1.5 × 106 s-1. hexachloro-6,7-epoxy-1,4,4a,5,6,7,8,8a-octahydro-endo, exo- Dehydrohalogenation of lindane formed trichlorobenzenes, 1,4:5,8-dimethanonaphthalene), and PVC. followed by subsequent hydrolysis and hydride/chloride exchange to form chlorophenols, lower chlorobenzenes, and Experimental Section phenol as the major final product. Reaction of poly(vinyl Degradation Experiments. Unless otherwise noted, all chloride) at 300 °C for 1 h formed aromatic hydrocarbons reactions were performed using a static (no flow) 4-mL cell ranging from benzene to anthracene and a char residue constructed from an Inconel 600 (21) threaded (npt) pipe with a ca. 1:1 carbon-to-hydrogen ratio (mol/mol). The residue fitting with end caps (63.5 mm long, 6.3 mm i.d.; Parker contained <1 wt % of chlorine compared to 57 wt % Hannifin Corporation, Columbus, OH). Inconel cells were chlorine in the original polymer. All compounds tested selected on the basis of their inertness in corrosion tests yielded chloride ion as the major product (at higher with subcritical and supercritical water (21). For each temperatures), indicating that complete dechlorination of experiment, 3 mL of water (HPLC grade; Fisher Scientific, Pittsburgh, PA), which had been purged with nitrogen for ∼2 some aliphatic organochlorines may be feasible. h to remove dissolved oxygen, was placed in the cell leaving about 1 mL of headspace when the vessel was capped. With this procedure, the internal pressure was governed by the Introduction steam/water equilibrium (i.e., pressures ranged from 1.3 to ° Organochlorine compounds represent one of the largest 170 bar for reactions performed over the 105 to 350 C groups of anthropogenic compounds found in the environ- temperature range used in this study (22)). Thus, all pressures ment. In addition to the concern for the environmental effects were substantially below the 428 bar rating of the Inconel of the parent compound (e.g., toxic effects of chlorinated reaction cells. [Safety note: It is imperative to have sufficient solvents and pesticides in groundwater), the presence of headspace in the vessel to ensure that interior pressure is organochlorines can inhibit the treatment and recycling of maintained by the steam/water equilibrium to avoid excessive waste streams. For example, poly(vinyl chloride) (PVC) is a pressures which can occur with a completely full cell (23). ° major source of chlorine for chlorinated dioxin formation At temperatures above 275 C, the density of water drops so during the incineration of municipal wastes. A number of that the cell is full of liquid water, which can cause pressures studies have investigated degradation of chlorinated pol- above the gas/liquid equilibrium (22). Therefore, the pro- < lutants via reductive dehalogenation to transform chlorinated portion of water added to the cell should be 75% at higher organics into their non-chlorinated analogues (1-10). Recent temperatures to ensure that adequate headspace remains). degradation experiments involving chlorinated aliphatic (7) Initial reactions were performed in water spiked with 10 and aromatic compounds (8-10) have been performed in µL of an acetone solution of tetrachloroethane (final con- supercritical water (i.e., at temperatures and pressures higher centration in water of 1 mM) and lindane (final concentration than the critical point of 374 °C and 221 bar) and subcritical of 0.57 mM). To aid in determining reaction products and water (defined as hot water with sufficient pressure to chlorine mass balance, experiments were also performed with higher concentrations (i.e., near their saturation con- * Corresponding author phone: (701)777-5256; fax: (701)777-5181; centrations at ambient conditions). Neat compounds were e-mail: [email protected]. spiked in the water to obtain final concentrations of 16 mM 10.1021/es011186k CCC: $22.00 2002 American Chemical Society VOL. 36, NO. 6, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 1337 Published on Web 02/07/2002 for tetrachloroethane, 8.9 mM for lindane, and 19 mM for min-1 ramp to 320 °C. Quantitations were based on the trichloroethene. Degradation experiments were also per- relative response factors of individual standards to the formed on 1,9-dichlorononane and dieldrin, with both at internal standard. Similarly, the identification and quanti- concentrations of 9 mM. PVC studies used 100 mg of powder fication of degradation products were based on comparisons (number average molecular weight, Mn, ca. 35 000; weight of mass spectra and retention data of pure standards with average molecular weight, Mw, ca. 62 000; supplied by Aldrich, those of the sample species, unless otherwise noted. Milwaukee, WI) in 3 mL of water for each test. Solid-phase microextraction (SPME) with 7 µm poly- Reaction cells were heated in a GC oven (Hewlett-Packard (dimethylsiloxane) (PDMS) fibers was used (24) in an attempt model 5890). Because the heating rate of the interior solution to identify organics which were too volatile to be determined was not initially known, a cell was modified by replacing the in the methylene chloride extracts. The analytes were directly 1 1 1 cap with a /2 in. npt × /8 in. tubing fitting to allow a /8 recovered with SPME from the product water and the diameter thermocouple to be placed directly into the water headspace above the water followed by direct thermal inside the cell during the heating step. This modified cell desorption in the injection port of the GC/MS. was used to determine the optimal heating procedure for Chloride Ion Determination. The determination of the subsequent reaction cell studies. chloride ion content in the product waters from the reactions When the GC oven was heated from room temperature using the higher concentrations of test compounds was after placing the cell in the oven, the heating rate of the performed with a chloride ion selective electrode (Cole- water inside the cell was quite slow (e.g., 15 min was required Parmer, Vernon Hills, IL). All determinations were performed for the water temperature to reach 240 °C with an oven in triplicate or quadruplicate. setpoint of 250 °C). Because more rapid heating of the reaction water was desired, the procedure was modified so that the Results and Discussion oven was first preheated to 50 °C above the desired temperature, and then the cell was quickly inserted into the Lindane. Lindane was selected as a test compound for two oven through the oven door. The GC oven temperature was reasons. First, it is a widespread toxic pollutant (25, 26). maintained at 50 °C above the desired temperature until the Second, our results obtained in pure subcritical water could cell interior temperature reached ca. 10 °C below the desired be compared with the data previously obtained in hot water temperature, and then the oven temperature was lowered to under alkaline conditions (7). the desired reaction temperature. With this procedure, the As shown in Table 1, the rate of lindane removal rapidly interior cell temperature monitored by the thermocouple increased with the increase in temperature from 125 to 200 reached within 10 °C of the desired temperature in 3-6 min °C. Using the data presented in Table 1, linear regressions (for 150 and 300 °C reaction temperatures, respectively) and were obtained for the logarithm of lindane concentration never exceeded the desired temperature by more than 2 °C.