Comments on “Removal of thiobencarb in aqueous solution by zero valent iron” by Md. Nurul Amin et al. [Chemosphere 70 (2008) 511–515]. 1 2 3 4 5 6 C. Noubactep Angewandte Geologie, Universität Göttingen, Goldschmidtstraße 3, D - 37077 Göttingen, Germany. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 e-mail: cnoubac@gwdg.de; Tel. +49 551 39 3191, Fax: +49 551 399379 Introduction In a recent article entitled “Removal of thiobencarb in aqueous solution by zero valent iron”, Nurul Amin and his coworkers (2008) reported the development of a cost-effective method for the purification of thiobencarb-contaminated water (agricultural wastewaters) using powdered elemental iron (zerovalent iron or Fe0). For this purpose 10 mL of a 10 mg L-1 thiobencarb (TB) solution was allowed to react with 0.05 to 2 g powdered Fe0 [d ≤ 150 μm, specific surface area (SSA) = 7.0 m2 g-1] at 25 °C for 1 to 24 h. Some experiments were conducted with elemental magnesium and elemental zinc. As result 10 mg L-1 TB could be almost completely removed from 10 mL solution by 0.1 g Fe0 (mass loading: 10 g L-1). TB removal was accompanied by chloride ion release to the aqueous solution suggesting that reduction and adsorption are the two processes responsible for TB removal. The study of Nurul Amin et al. (2008) is very informative to researchers interested in the field of iron technology as it relates the state-of-the-art of the technology. However, the article contains areas where improvements could be made that will be discussed below. Removal Efficiency with Fe0, Mg0, and Zn0 Nurul Amin et al. (2008) presented the results of TB removal by Fe0, Mg0 and Zn0 as summarised in Table 1 without any further discussion. Assuming that Mg0 and Zn0 were of 1 comparative particle size and purity like Fe0, these results suggest that TB reduction by Fe0 (direct reduction, see below) may not be the major removal mechanism in the Fe 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 0-H2O system. In fact, on the reduction-potential scale, Mg0 and Zn0 are more susceptible to oxidation than Fe0 (Table 1). Data from the synthetic organic chemistry support this conclusion. Elemental metals are traditionally used in the synthetic organic chemistry as reductive agents, for example for the transformation of aromatic nitro compounds to their corresponding aromatic amines. The most classic and practical reductants are zinc (Zn0; E0 = -0.76 V), tin (Sn0; E0 = -0.14 V), or iron (Fe0; E0 = -0.44 V) in the presence of an acid (Boix and Poliakoff, 1999; Wang et al., 2003). The reactions are performed in organic solvents or in the presence of acids, which pose waste-handling problems (Wang et al., 2003). Quantitative aqueous reduction reactions are only possible at elevated temperatures. For example Boix and Poliakoff (1999) reported a selective reduction of nitroarenes to anilines using Zn0 in water at 250 °C (near-critical water) in high yields, but only 10% yield of aniline was observed from the reduction of nitrobenzene using Fe0 under the reaction conditions. The reactivity of Zn0 is superior to that Fe0, while metallic iron is non-toxic, cheap and commercially available. To perform the reduction of nitro compounds to amines with iron powder, the reaction should be carried out at higher temperature (> 250 °C) owing to the low reactivity of iron metal (Wang et al., 2003). Clearly, aqueous reduction reactions that are quantitative with Zn0 at a given temperature should exhibit a lower yield with Fe0 under the same conditions. Under the experimental conditions of Nurul Amin et al. (2008), the yield of TB removal by Zn0 was 13%. This result suggests that the observed quantitative TB removal (100%) in the Fe0-H2O system is due to at 2 least 90% of removal agents (including reductants) other than the available amount of Fe0 at the surface of the used Fe 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 0 materials. These processes are inherent to iron corrosion. From open literature on metal corrosion, it is well-known that the corrosion behaviour of a metal is not a direct reflect of its standard electrode potential (E0 value; Bojic et al., 2007 and references therein). The protectiveness of the oxide layers formed on the individual metals is a key factor in this regard. Iron has been shown to exhibit the tendency of forming oxide layers which are not adherent to metallic iron. Therefore, generated oxide layers constantly flake off and expose fresh iron surface for attack (Dickerson et al., 1979; Campbell, 1990). Consequently, processes responsible for increased TB removal in the Fe0-H2O system are coupled with progressive iron corrosion. In this regard TB may be (i) reduced by Fe0 (direct reduction) or secondary reductants like FeII, H/H2 (indirect reduction), (ii) adsorbed onto already generated corrosion products (adsorption) or (iii) co-precipitated with nascent or recrystallizing corrosion products (co-precipitation). Note that, if TB is reduced in the aqueous phase, the products of the reaction may equally adsorb onto or co-precipitate with corrosion products. In considering solely direct reduction and adsorption, Nurul Amin et al. (2008) have overseen indirect reduction and co-precipitation. Contaminant co-precipitation through corrosion products has been shown to be the primary removal mechanism in Fe0-H2O systems (Noubactep, 2007). Note that chloride ions from TB reduction may co-precipitate with iron corrosion products. Therefore, regarding the absence of Cl- in the aqueous solution as indication for TB adsorption is questionable. Rationale for Experimental Conditions First of all, the used Fe0 mass loadings varied from 5 to 200 g L-1, corresponding to available surface areas varying from 0.35 to 14 m2 (in 10 mL solution). As the cross-section of individual TB molecules is not given, the reader can not imagine whether the available 3 surface area in each reaction bottle is sufficient to realize “stoichiometric” degradation (monolayer coverage of the Fe 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 0 material). Ebie et al. (2001) have shown effective activated carbon adsorption of TB in pores of size below 15 Å. Considering TB as a rigid sphere with a radius of 7.5 Å (one half of 15 Å), the total surface occupied by the amount of TB in 10 mL solution is calculated as 0.413 m2 (Jia et al., 2007). This surface is purchased by 0.06 g of the used Fe0 material (SSA = 7 m2 g-1). Therefore, using a mass loading of 6 g L-1 would have been sufficient for a “stoichiometric” TB degradation. These calculations show clearly that in the majority of reported experiments, Nurul Amin et al. (2008) used large excess of Fe0. On the other site, the molar ratio Fe0/TB varies from 2279 to 91150 (assuming 92% Fe in the used Fe0 material). Therefore, the conclusion of Nurul Amin et al. (2008) that “1 kg of iron powder can be applied to the treatment of 100 L TB-contaminated water with 10 μg mL-1 concentration” (1g TB per kg Fe0 or 1 mg g-1) may be speculated if only the numbers of treatment batch systems are increased. The possibility of large scale plant is not supported by any scientific criterion. Furthermore the presented “simple, easy handling and convenient method with iron metal” is not compared with available technologies for TB removal such as the use of activated carbon adsorption. A good decolourising activated carbon should fulfil at least 200 mg g-1 removal capacity for methylene for example (Attia et al., 2008 and references therein). Therefore, although the present work is very informative to researchers interested in the field of iron remediation technology, it is questionable whether the proposed method is really affordable (1 mg g-1). The second point concerns the chosen experimental duration (treatment time). Most of the experiments have been performed with a Fe0 mass loading of 10 g L-1 for 1 to 24 h. The mixing intensity is not specified. It is obvious that the extent of TB removal dictates the treatment time. Figure 1 summarizes the effect of iron surface area (m2) on the TB removal efficiency. It can be seen that for Fe0 surface area ≥ 0.7 m2 (0.1 g Fe0) the removal efficiency 4 is larger than 97%. Therefore, it is questionable why results with essentially larger Fe0 surface areas (up to 14 m 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 2) are reported. Furthermore, for a purposeful discussion of the reaction mechanism experiments with less Fe0 or smaller mixing intensities could have been advantageous to access the region of sharply removal efficiencies (first 3 h). The monitoring of the chloride ion generation in such systems could also enable a better discussion of the extent of TB reduction as relative little amount of corrosion products is available to interfere with generated Cl-. Under shaken conditions iron corrosion, and thus corrosion products precipitation is accelerated. Cl-, TB, and TB reduction products are certainly entrapped in the matrix of co-precipitating iron oxides. Therefore, the absence of Cl- in the solution is by no means a hint for contaminant adsorption, as suggested by Nurul Amin et al. (2008). Since the iron mass loading, the mixing intensity and the treatment time are not independent variables, the mixing intensity should always be specified to enable comparison with other published data. Conclusions Nurul Amin et al. (2008) certainly showed the capacity of Fe0 to quantitatively remove TB from aqueous solutions (e.g., agricultural wastewaters in general). However, their experimental conditions may not be pertinent to enable traceable conclusions. Because the formulated criticisms are valid for the majority of works dealing with the process of contaminant removal in Fe0-H2O systems, an unified procedure for the investigation of processes in Fe0-H2O systems is needed. The discussion above also shows that the well-established premise that contaminants are removed from the Fe0-H2O systems under field and laboratory conditions (T ≤ 35 °C) by direct reduction is partly a contradiction of good data from synthetic organic chemistry. 5 Therefore, a close investigation of the processes of contaminant removal in Fe0-H2O systems is necessary to access the long-term stability of removed contaminants. 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 References Attia, A.A., Girgis, B.S., Fathy, N.A., 2008. Removal of methylene blue by carbons derived from peach stones by H3PO4 activation: Batch and column studies. Dyes Pigments 76, 282– 289. Boix, C., Poliakoff, M., 1999. Selective reductions of nitroarenes to anilines using metallic zinc in near-critical water. J. Chem. Soc. Perkin Trans. 1, 1487–1490. Bojic, A.L., Purenovic, M., Bojic, D., Andjelkovic T., 2007. Dehalogenation of trihalomethanes by a micro-alloyed aluminium composite under flow conditions. Water SA 33, 297–304. Campbell, J.A., 1990. General Chemistry. 2nd Ed., VCH Weinheim, 1223 pp. (in German) Dickerson, R.E., Gray, H.B., Haight Jr., G.P., 1979. Chemical Principles. 3rd Ed., Benjamin/Cummings Inc. London, Amsterdam, 944 pp. Ebie, K., Li, F., Azuma, Y., Yuasa, A., Hagishita, T., 2001. Pore distribution effect of activated carbon in adsorbing organic micropollutants from natural water. Water Res. 35, 167–179. Jia, Y., Aagaard, P., Breedveld, G.D., 2007. Sorption of triazoles to soil and iron minerals. Chemosphere 67, 250–258. Noubactep, C., 2007. Processes of contaminant removal in "Fe0-H2O" systems revisited: The importance of co-precipitation. Open Environ. J. 1, 9–13. Nurul Amin, Md., Kaneco, S., Kato, T., Katsumata, H., Suzuki, T., Ohta, K., 2008. Removal of thiobencarb in aqueous solution by zero valent iron. Chemosphere 70, 511–515. 6 153 154 155 156 Wang, L., Li, P., Wu, Z., Yan, J., Wang, M., Ding, Y., 2003. Reduction of nitroarenes to aromatic amines with nanosized activated metallic iron powder in water. Synthesis 13, 2001– 2004. 7 156 157 158 159 Table 1: Characteristics of elemental metals and their percent removal of thiobencarb (P) from aqueous solution. n.s.= not specified. Material E° d S P (V) (μm) (m2 g-1) (%) Mg0 -2.2 n.s. n.s. 60 Zn0 -0.76 n.s. n.s. 13 Fe0 -0.44 < 150 7.0 100 160 161 8 161 0 3 6 9 12 15 0 20 40 60 80 100 Exp. conditions [TB]0 = 10 mg L -1 V = 10 mL T = 25 °C pH = 6 0 < mFe(0) (g) < 2 6 hours 24 hours TB re m ov al / [% ] Fe0 surface / [m2] 162 163 164 165 Figure 1: The effect of iron surface area (m2) on the thiobencarb (TB) removal efficiency from the aqueous solution. Data from Table 1 in Nurul Amin et al. (2008). The lines are not fitting functions, they simply connect points to facilitate visualization. 9