@ vm PII: $0043-1354(98)00016-5 War. Res. Vol. 32, No. 9. pp. 2607-2614, 194 11998 Elsevier Science Lid All sights reserved Printed in Great Britain (0043135498 $19.00 + 0.00 IRON POWDER, GRAPHITE AND ACTIVATED CARBON AS CATALYSTS FOR THE OXIDATION OF 4-CHLOROPHENOL WITH HYDROGEN PEROXIDE IN AQUEOUS SOLUTION F, LUCKING*, H. KOSER, M. JANK and A. RITTER Institut fir Umweltschutztechnik, Martin-Luther-Universitit Halle-Wittenberg, Geuseer Stra, 15-06217 Merseburg, Germany (Fist received March 1997: accepted in revised form December 1997) Abstract—Solid materials as iron powder, graphite and activated carbon were tested for their catalytic properties for the oxidation of 4-chlorophenol in aqueous solution with hydrogen peroxide, Batch tests ‘were performed at 20°C, and continuous tests with granular activated carbon under ambient conditions (20°C). Iron powder was shown to act as a catalyst for the activation of hydrogen peroxide and owing. to this For the oxidation of 4-chlorophenol. The catalytic activity is based on iron ions leached from the iron powder which act as homogeneous catalysts. Thus, the process can be described as a Fenton reac- tion enabled by a dissolution of iron powder. The sume phenomenon was observed when iron impreg- nated activated carbon was applied as a catalyst. In contrast to iron powder, graphite and activated carbon were found to act as heterogeneous catalysts for the activation of hydrogen peroxide and the oxidation of 4-chlorophenol. During the oxidation of 4-chlorophenol with activated carbon as a cata- lyst the decomposition of H,0; is significantly slower than in the absence of 4-chlorophenol. This can be explained by the adsorption of 4-chlorophenol, duc to which the surface area of the activated carbon available for the decomposition of HO:, is reduced. At the same concenteation of catalytic material in batch tests, the oxidation of 4-chlorophenol proceeds much faster when iron powder instead of graphite or activated carbon is used. On the other hand, graphite and activated carbon are stable catalysts not affected by dissolution processes. Activated carbon can be utilized for a continuous process in’ fixed bed reactor. For this purpose the efficiency of 4-chlorophenol oxidation in granular activated carbon filled columns was investigated and 4 25% conversion of the 4-chlorophenol (1 gI”") was achieved at a retention time of 26 min, (1998 Elsevier Science Lid. All rights reserved key words—iron powder, graphite, activated carbon, hydrogen peroxide, 4chlorophenol, aqueous phase oxidation INTRODUCTION The chemical oxidation of toxic or refractory ‘organics is an emerging technique for the treatment of waste water (Kaiser, 1996). A catalytic process for the oxidation of organic compounds found in waste water is given by the use of Fenton's reagent (Feuerstein, 1982). The Fenton reaction is a catalytic provess for the generation of hydroxyl radicals from hydrogen per- oxide and is based on an electron transfer between 1,0; and iron ions acting as homogeneous cata- lysts. The hydroxyl radicals produced during this activation of hydrogen peroxide are a strong oxidiz- ing agent and able to oxidize organic compounds under ambient conditions. The advantage of Fenton's reagent is that no energy input is necess- ary to activate hydrogen peroxide. The disadvan- tage of Fenton's reagent is that the homogeneous ‘Author to whom all correspondence should be addressed. (Tel: +49-3461-462702 or (2701); Fax: +49-3461- 4462710, E-mail: manfred janka@ vt.uni-halle.de] catalyst, added as iron salt, cannot be retained in the process thus causing an additional water pol- lution, The oxidation properties of hydrogen peroxide and hydroxyl radicals depend on the pH of the sol ution, In an aqueous solution with a pH of 0 the redox potential is +1.77V for hydrogen peroxide and +2.8V for hydroxyl radicals, in an aqueous solution with a pH of 14 the redox potential is +0.88 V for hydrogen peroxide and +2.0 V for hy- Groxyl_ radicals (Feuerstein, 1982: Weichgrebe, 1992), Due to this, acidic conditions are applied for oxidative treatment of waste waters with hydrogen peroxide. Recently, attempts have been made to avoid the additional water pollution due to metal ions by replacing the homogeneous catalysts by hetero- geneous catalysts. For this purpose, supported metal catalysts can be prepared (Al-Hayek and Doré, 1990; Fajerweg and Debellefontaine, 1996) ‘These catalysts provide catalytic setivity for the for~ mation of hydroxyl radicals from hydrogen per- 26072608 oxide but simultaneously the metals are leached from the support material The metal ions leached from the solid material act as homogeneous catalysts for the activation of hydrogen peroxide which makes it difficult to differ- entiate between homogeneous and heterogeneous catalytic effects. Another problem caused by the leaching of the metal ions from the catalysts is cata- Iyst deactivation, In view of these problems, the present study will provide further information about the catalytic properties of different solid materials for the oxi- dation of a chlorophenolic model compound with hydrogen peroxide, ‘The catalyst materials selected for this study were: iron powder, because it has been reported that the amount of iron ions present in a waste water can be reduced by 50% compared to a Fenton reac- tion when the iron salts are replaced by iron pow= der (Wada ef al., 1993). activated carbon, because of its ability to pro- duce free radicals during the decomposition of hydrogen peroxide (Ono ef al., 1977) and - graphite, because its physical properties are similar to those of metals (thermal and electrical conductivity) and, in many respects. it is also re- lated to activated carbon (Kienle and Bader, 1980), EXPERIMENTAL For all experiments reported in this study 1 g1' 4 chlorophenol (Merck-Schuchardt > 98%} dissolved in deio- nized water was used as a synthetic model waste water Sulfuric acid (96%) for pH adjustment and hydrogen per oxide (35%) were obtained from Conlag Leipzig. The iron ponder (95% Fe, Isocommerz VE Ausen- und Binnenhandelsbettied BT Herzbere’E.) used was. applied at a particle size of 70-100 pm. The particle chameter of graphite (99.8% C, Labotchemie Apolda GmbH) was 100m. The activated carbons used sere F300 (Chemviron Carbon) (BET surface 1000: g"!; Ash con- Parco GCW. (Norit) (BET surface Ash content < 8%) made trom hurd coal, ra (Norit) (BET surface 980m" g's Ash content < 6%) made from peat and REZ I (Nori) (BET surface 800mg"; Ash content 1S 20%} made fram peat followed by an impregnation with iron hydroxide For butch experiments the activated cathons were milled and sieved 10 a particle size of = $0 jm, The activated ear- bons were ssushed with hydrochloric acid (20% HCI 100°C; 2) and the concentration of the metals to be lea ched (Al, Cu Fe. Ni, Co, Sn. Mn and! Mg) in the extesct was determined with an ICP spectromeisr (Philips PU 7000), In all extracts, iron was found 0 be the predomi- nant metal. The determined content of igon leached from the carbons way 491mg e! for F-300, 4.Stmeg” lor ROW, 0.6 meg | for GCW and 90.2 mee! for RFZ. Batch experiments were performed in brown hiss bot Hes with magnetic stirrer ina thermoconstant bath at 30°C. In al tests, the applied concentration of catalyst ma terial, iron powder. activated catbon snd graphite) was Tgt'. Hydrogen peroxide and catalytic material were dosed simultaneously to start the reaction, The amount of hydrogen peroxide (15 ml! (35%). 5.31"! H30s) was LS times the stoichiometric umount (344g HsOsje 4 hlorophenol) required far the complete oxidation of 4 F. Licking er al chlorophenol 10 COz, HO and HCL. The batch exper- iments were run for 6h (iton powder), 96 h (graphite) and 816 h (activated carbons) respectively. ‘Additional continuous tests were run under ambient conditions (20°C) in glass columns filled with granular activated carbon. The columns with a diameter of 10mm. were fed using peristaltic pumps. The model waste water applied in the continuous tests was deionized water with Ig 4chloropheno! and 15 ml! HO, (35%) adjusted to pH = 3. The columns were filed with 2g of each cur bon type (bed height = 5.5 cm). The particle size of the activated carbons was 05-1 mm for F-300: 0.8mm for ROW; 0.5-1 mm for GCW and I mm for RFZ. The degradation of 4-chlorophenol was detected by measurement of dissolved organic carbon (DOC) and chloride released from the 4chlorophenol, The DOC (gr 4chlorophenol = $60mgI" DOC) (graphite and activated carbons with a 0.45 jam filter, Carl Roth GmbH) was analyzed with the TOC-analyzer (Shimadzu 5030) according to the standard method DIN 38409 Part 3 (German standard methods for the examination of water. waste water and sludge, general measures of effect and substances (group H); determination of total organic car- bon (TOC) (H3) June 1983). Chloride (maximum 277 mg for the chosen concentration of 4-chloraptenol) was measured with the Chloride Analyzer (Corning 926) according to DIN. 38405 Part 1 (German standard ‘methods for the examination of water, waste water and sludge, general measures of anions (group D); dete:mi- nation of chloride ions (DI-2) December 1985). H,0; com: centration was analyzed according to DIN 38409 Part 15, (German standard methods for the examination of water waste water and sludge, general measures of effect and substances (group H); determination of hydrogen peroxide and its adducts (H 15) June 1987) Iron ions leached during the experiments were analyzed with atom absorption spectroscopy (Varian, SpectrAA- 20), 1.0, ised ion {mac cone] init cone} Bo Bo cr S OC removed AL \ | | | Time (h} Fig. 1. Efficiency of iron powder as catalyst for the oxidation of 4-chlorophenol (I gi"! iron powder: 1 pI" 4-chlorophenol: 5.3 g1"! HO 30°C),Catalytic aqueous phase oxidation —S H,0, [of initial concentration) % 2609 CT [of maximum concentration} — __ Ee [of initial concentration] Fig. 2. Efficiency of I mg! Fe Ig! RESULTS. Iron powder Figure | shows the pH, the amount of iron ions leached from the iron powder, the H;02 concen- tration, the concentration of chloride released and the DOC concentration during the oxidation of 4- chlorophenol as a function of the reaction time. ‘The iron ion concentration in the liquid phase increases over the test period. The leaching process is accelerated with decreasing pH. With increasing iron concentration the HO; activation and the degradation of 4-chlorophenol are accelerated. Chloride ions are released during the degradation 20 40 Time (] Fig. 3. Efficiency of graphite as catalyst for 4chlorophenol: 53.21 ts catalyst for the oxidation of 4-chlorophenol (2.714 mel 4-chlorophenol, 53 g1' H,O5: 30 C: initial pH = 3), ae 1,0, [of incom oo oo a oxidation of 4cblorophenot (1 g1" graph N03, 30°C: initial pH = 3) 100 150 Time [h] FeSO.: of 4-chlorophenol and the DOC of the solution is reduced. {In the tests with initial pH of 3 and 5 a decrease in pH was observed. This can be explained by the formation of reaction products such as HCI and or- ganic acids At initial pH<3 the leaching of iron, the de- composition of H;02 and the release of chloride ions from 4-chlorophenol occur immediately after having started the experiments. At an initial pH of 5 the leaching of iron ions and the reaction of H,0; and 4-chlorophenol are delayed. The delay is correlated with a decrease in pH during the test. As the amount of iron ions leached from the iron CY [of maximum concentration] {of initial concentration} 60 802610 powder increases, the H02 is decomposed and the 4-chlorophenol is degraded. In all tests the chloride of 4-chlorophenol is com- pletely transformed into chloride ions after 5h, while only 60% of the DOC are removed. In spite of HO, still being present, at least in the exper- iment with initial pH of 5, no further removal of DOC can be observed. This can be explained by the formation of organic reaction products which are not further oxidized (Sedlak and Andren, 1991), Graphite During the tests with graphite and an initial pH of 3 a final iron concentration of 0.5mgi' was measured in the aqueous solution, In order to decide whether this iron concen- tration has already an appreciable effect on the reaction, the oxidative degradation of 4-chlorophe- nol with added FeSO, (I mg I"! Fe’) was investi- gated. The results are shown in Fig, 2 Compared to the oxidative degradation of 4- chlorophenol with graphite as catalyst (Fig. 3) the oxidation proceeds significantly slower when I mgt! Fe>” is applied as a homogeneous catalyst. With I mgt! Fe** being twice the concentration of the iron measured in the liquid phase during the experiments with graphite, a participation of homo- geneous catalytic effects can not be excluded but is dominated by heterogeneous effects when graphite is used as a catalyst Graphite acts as a heterogeneous catalyst for the activation of hydrogen peroxide. HO, is decom- posed while 4-chlorophenol is oxidized. The removal of the DOC and an increase in the chloride concentration (Fig. 3) furnish proof of this effect Additional tests were run with graphite to test adsorption (no H;02 in the solution) and decompo 100: 80 H,0, [%of initial concentration] & Fig. 4. Efficiency of F. Lacking etal sition of hydrogen peroxide (no 4-chlorophenol in the solution). The adsorption of 4-chlorophenol on graphite (1 gt 4-chlorophenol, 1 gt graphite and pH = 3) was determined via DOC taking 9 mg 4-chlorophenol/g graphite, it was considered to be negligible for the interpretation of the results. The test of the H302 decomposition (5.3 1°! H:0>, Lgl graphite and pH = 3) showed that the H:O2 decomposition is independent of the presence of 4 chlorophenol in the solution, ‘Compared to the iron catalyzed reaction (Fig. 1) the graphite reaction is slow. An 80% chloride release from 4-chlorophenol with graphite as cataly- tic material is achieved after 96 h, compared to 1 to 3h when iron powder is used. In the presence of graphite only 25% of the initial HyOy are decom- posed after 96, whereas iron powder causes under the same conditions a complete decomposition of HO) after 6h. An advantage of graphite in com- ison to iron powder is that it is not affected by leaching or dissolution processes. Activated carbon-batch experiments The results obtained from batch tests with differ- ‘ent activated carbons are shown in Figs 4-7, In Fig. 4 the decomposition of hydrogen peroxide is shown with no organic compound present in the aqueous phase. Hydrogen peroxide is decomposed by activated carbon, The decomposition rate depends on the nature of the activated carbon used (Fig. 4), During the decomposition of hydrogen peroxide no change in pH was observed. The iron found in the aqueous solution after a reaction time of 4h was less than 0.5 mg" for all activated car- bons used (RFZ included). Due to this low concen- tration of iron ions a homogeneous reaction participating in the decomposition of hydrogen per- Time (h) decomposition by different activated carbons (Ig I"! activated carbon: 53g H:0z: 30°C: pH = 2),Catalytic aqueous phase oxidation H,0, [%of initial concentration) 2611 Time (8) Fig. 5. Eficiency of HO, decomposition by different activated carbons during the oxidation of 4-chlor- phenol (1g activated carbon: 11" 4-chlorophenol: 53g!" HaO3: 30°C: inital pH ~ 3) oxide can be excluded (Fig. 3). The decomposition of hydrogen peroxide was similar for ROW, GCW and RFZ. In contrast, the decomposition of hydro- gen peroxide proceeds considerably slower when F- 300 is used, Addition of 4-chlorophenol to the aqueous phase significantly reduces the decomposition of HO: (Fig. 5). For a 40% decomposition of H502 the reaction time is increased by about 60 times when F-300 is used, by 490 times when ROW is used, by 440 times when GCW jis used, and by 57 times when RFZ is used. This phenomenon is a conse- quence of the 4chlorophenol adsorption which reduces the surface area of the activated carbon available for the decomposition of H:0: During the oxidation of 4-chlorophenol with different activated carbons as catalysts the pH CI {%of maximum concentration] 0 200 Fig. 6. Bfiiciency of chloride release from 4-chlorophenol during the oxidation with different a decreases from 3 to 2. When RFZ is applied the de- composition of hydrogen peroxide proceeds faster compared to the use of the other activated carbons (Fig. 5). This can be explained by a release of iron ions from the activated carbon RFZ. When REZ was used 56mgI-' iron were detected in the aqu- cous phase after 800h while the concentration of iron ions in the tests with the other activated car- bons was less than 2 mg” ‘The chloride release during the degradation of 4- chlorophenol is shown in Fig. 6. In the test with RFZ the chloride release from 4-chlorophenol pro- ceeds faster than in the tests with other activated carbons. This is correlated with the decomposition of H:0:, During the first 300 h of the experiments chloride release is faster with ROW than with GCW or F- 400 ‘Time [h} 600 800 ated carbons a5 catalyst (Ig activated carbon: IgI” 4chlorophenol, $.3g" HyO3, 30°C. inital pH = 3)2612 [2% of initial concentration) DOC reduced by oxidation Fig. 7. Efficiency of oxidative DOC removal during vated carbons as catalyst (1 gl 300, After 500 h the chloride found in the aqueous phase is almost the same for all activated carbons, With iron powder a complete chloride release could be achioved after Sh while only 80% were released after 800 h when activated carbon was used, The amount of dissolved organic carbon removed from the aqueous phase by oxidation was deter- mined as DOC reduction by oxidation in oxidation experiments — DOC adsorption experiments, The DOC reduction in the adsorption exper: iments was determined under the same conditions as in the oxidation experiments (no addition of H,0.). The amount of DOC removed in. the adsorption tests wats REZ: 39.4% of initial DOC GEW: 41.2% of initial DOC ROW. 48.7% of initial DOC F300: 44.1% of initial DOC DOC reduction reduction in Compared to the other activated carbons the oxi= dative degradation of DOC proceeds faster when RFZ is used (Fi. 7), This corresponds to the release of chloride and the HO decomposition and ‘can be explained by a Fenton reaction initiated by the release of iron ions from the iron impregnated activated carbon RFZ, During the experimental Table 1 Etuent data of BOC. C1 HO: and disolved iron at 324 FP. Litcking er al 800 400 Time (h] 600 the oxidation of 4chlorophenol with different acti- activated carbon: 5.3 1" HyOx; 30°C: initial pH = 3), decay the oxidative reduction of DOC reaches about 25% of the initial DOC when the carbons ROW, GCW or F-300 are used as catalysts, Compared to the iron powder experiments the oxidation of 4-chlorophenol proceeds about 200 times slower with activated carbon under the given conditions, Activated carbon-flow experiments In continuous tests a steady state for the DOC in the effluent of the columns was reached after 20 days with the activated carbons F-300, GCW and ROW. In the continuous test with RFZ the DOC in the effluent is increasing gradually and after 160days a steady state was not reached. It may be supposed that this is a result of a stil ongoing leaching of the surface iron. The DOC, H302, chloride and iron ion eoncen- trations in the effluent of the columns after 32 and 160 days are shown in Table When RFZ is used large concentrations of iron are leached from the activated carbon. The iton ions in the aqueous phase act as homogeneous cata- Iysts and are responsible for the faster decompo- sition of hydrogen peroxide and the higher grade of 4-chlorophenol oxidation achieved as with carbons which were not impregnated with iron, The amount ood (Peed solution: 1g1°' 4ehlorophenok 3g" HO, 1H 3/20 retention time ~ 26 mins flow rate 351 mith) ‘Specie HO, DOC removal Chloride teased 1.03 consumption Dissolved iron consumption of fe) of may of feed) (me) Sampling 3 oo ® 160 2 10 2 ro Rez 4” » os 3s st “4 ry Gow 1s 0 x % a6 04s Row B r o 9 1% 8 on P00 Ws 5 6 2s 2 7 046 tk