Oxidation of Phenol in Acidic Aqueous Suspensions of Manganese Oxides

-Phenol (benzenol) oxidation by three synthetic manganese oxides (buserite, manganite, and feitknechtite) has been studied in aerated, aqueous, acidified suspensions. The rate of reaction was pH dependent. Oxidation was greatly enhanced below pH 4, when diphenoquinone and p-benzoquinone were identified as the first products. Initial reaction rate was correlated with standard reduction potential (E ~ of the oxides following the order: feitknechtite > manganite > buserite. A more gradual process of phenol oxidation after the initial reaction was influenced by electrochemical properties of the solution. High soluble manganese activity and increase in pH adversely affected reaction rates. Thus, the reactivity of the oxides was related to their stability and possibly the ability to readsorb Mn(II), following the order: buserite > manganite > feitknechtite. The results indicate that thermodynamic and electrochemical data for oxides and phenols are useful in predicting under which conditions phenols can be oxidized by a given system. Key Words--Buserite, Feitknechtite, Manganese oxide, Manganite, Oxidation, Phenol, Quinones, Reduction potential. I N T R O D U C T I O N Manganese oxides can act as electron-transfer agents in reactions that involve phenolic compounds and molecular oxygen (Ono et al., 1977; Stone and Morgan, 1984; McBride, 1987, 1989a, 1989b; Stone, 1987; Kung and McBride, 1988; Ulrich and Stone, 1989). The manganese +2, +3, and +4 oxidation states appear to be involved in electron-transfer reactions, but the factors influencing the reactions are still poorly understood. Phenol disappearance from acidified water samples in the presence of manganese oxide has been observed by Thielemann (1971). The extent of decrease in phenol concentration was greatest at low pH. No change in phenol concentration was observed for samples at pH > 6 over a period of days. Chen et al. (1991) investigated low phenol recoveries by EPA Method 625 in water samples containing manganese, and found phenol was oxidized to p-benzoquinone in acidified water samples containing manganese oxides. No detailed systematic study of conditions under which phenol oxidation by manganese oxides occurs has been conducted. This study investigated some properties of oxides (stability, structure, reduction potential, manganese oxidation state, surface area), and solutions (phenol and manganese(II) concentrations, pH, presence of complexing agent) that influence the kinetics of oxidation of unsubstituted phenol by manganese oxides. MATERIALS A N D METHODS Buserite in Na-saturated form (Na4Mn~4027"9H20) was synthesized according to the procedure of McCorresponding author. Copyright 9 1992, The Clay Minerals Society Kenzie (1981). Solutions of 0.8 M MnSO 4 in 1 liter water and 11 M NaOH in 3 liters water were cooled to 5~ and mixed. Oxygen was bubbled through the mixture for 5 hours. Buserite in Mg-saturated form was prepared by washing Na-buserite with 0.1 M Mg(NO3)2 (Giovanoli et al., 1975). This form was chosen for experiments because it is more stable than other non-transition metal buserites (Giovanoli et al., 1975), and it does not collapse to the 0.7 nm phase until the relative humidity is less than 10% (Tejedor-Tejedor and Paterson, 1979). The oxide was air-dried and ground gently in a mortar. Freeze-drying was avoided because it resulted in removal of interlayer water. X-ray powder diffraction (XRD) showed characteristic d-values of 0.959, 0.479, and 0.320 nm. Manganite (3,-MnOOH) was synthesized by the method of McKenzie (1971). Manganese(II) hydroxide was precipitated by adding 1 liter of 0.75 M N a O H to 2 liters of 0.35 M MnSO4' H20, and washed by decantation keeping the flask stoppered. It was oxidized to a mixture of hausmannite and manganite by slowly adding 90 ml of 30% hydrogen peroxide. The mixture was boiled while bubbling with air for 35 days, at which time X R D showed d-values of manganite: 0,342, 0.264, 0.226, 0.220, 0.170, 0.167, and 0.165 nm. The infrared spectra showed strong sharp bands characteristic of manganite at 1087, 1119 and 1149 cm ~. The oxide was washed and sonicated repeatedly with distilled water to remove the excess salts, and then freeze-dried. Feitknechtite (B-MnOOH) was synthesized according to the method of Hem et al. (1982) by raising the pH of 900 ml of 0.0098 M Mn(C104)2 to 9.00, and holding it at that pH by addition of carbonate-free 0.2 M NaOH until the reaction was complete. The temperature of the solution was held between 0 ~ and 2~ 157 158 Ukrainczyk and McBride Clays and Clay Minerals Table 1. Properties of manganese oxides. Average Mn oxidation Oxide state % Mn Formula E~ Mg-buserite 3.55 44 MgMnloO18.7 1.25 t'2 Manganite 3.01 60 MnO~.s 1.50 ~ Feitknechtite 3.02 57 MnO~5 1.653 Bricker (1965). 2 Value for birnessite is reported, since no value for buserite was available. 3 Hem et al. (1982). using an ice bath and air was continuously bubbled through the stirred solution. The pH was then adjusted to 4.2 with HCIO4 and soluble Mn 2+ to 168 mg/liter using a Mn(C104)2 solution, and the suspension was stored in a refrigerator at 5~ Following 4 days of aging, the oxide was separated from the suspension, washed, and sonicated several times with deionized distilled water, and freeze-dried. Powder X R D showed the oxide to be feitknechtite with d-spacings of 0.463, 0.253, 0.235, and 0.196 nm. The oxidation state of Mn (Table 1) was determined by dissolving the oxide in a ferrous sulfate/sulfuric acid mixture and back-titrating a fraction of the solution with potassium dichromate in the presence of diphenylaminosulfonic acid as indicator. Another fraction of the solution was used for determining dissolved Mn by atomic absorption spectroscopy (AAS). This procedure avoids any need for drying the solid (Hem et al., 1982). Dichromate titration was found superior to the commonly used permanganate titration, giving a much sharper end point, and reproducible results using commercial MnO2 as a standard. The surface area of oxides measured by BET analysis o f N 2 adsorption was 30 m2/g for Mg-buserite, 8.5 m~/g for manganite, and 100 m2/g for feitknechtite. These values are probably lower than actual particle surface area in aqueous solution due to aggregation and sedimentat ion during drying (Stone, 1987). For buserite this value almost certainly corresponds only to external surface area, and does not measure the interlayer surface. A low BET surface area of manganite can be attributed to its needle-like particles collapsing together upon drying. Qualitatively, it was observed that manganite disperses into very fine particles in aqueous solutions, so it is likely to have a much larger surface area in suspension. In oxidation experiments, a weighed quantity of oxide was combined with water and the desired amount of acid (HC1, HNO3, or CH3COOH) for 15 minutes (unless otherwise indicated) in ground-glass stoppered Erlenmeyer flasks prior to phenol addition. The mixture was stirred with a magnetic stirrer at 22 ~ _+ I~ Aliquots of suspension were withdrawn during reaction and filtered through 0.2 um Nucleopore membrane r -