Comments on: “Comparison of reductive dechlorination of p-chlorophenol using Fe and nanosized Fe 0 0.” By R. Cheng et al. J. Hazard. Mater. 144 (2007) 334. C. Noubactep Centre of Geosciences, University of Göttingen, Goldschmidtstraße 3, D - 37077 Göttingen, Germany. e-mail: cnoubac@gwdg.de; Tel. +49 551 39 3191, FAX: +49 551 399379 The article of Cheng and colleagues [1] was intrinsically interesting as its intended to compare the efficiency of conventional Fe (mm to μm size) 0 and nanosized Fe for aqueous reductive dechlorination of 0 p-chlorophenol. Such investigations are necessary to gain more details on the expected reactivity of nanosized Fe . In fact, the introduction of nanosized Fe in groundwater remediation was not univocal [2]. Therefore, such comparative works from the same research group should be very helpful. 0 0 Unfortunately, due to the lack of a unified procedure for conducting contaminant removal experiments in investigating processes in Fe0–H2O systems (here Fe0, 4-CP, H2O, O2), almost any researcher uses a different experimental procedure. This deficiency has accounted for controversial results in other branches of science [3,4]. For example, Büchler et al. [3] have found out that the strong influence of the hydrodynamic conditions due to mixing operations explains many contradictory literature results on the process of the formation, growth, and dissolution of the passive film on iron in neutral and alkaline solutions. Despite the failure of a unified procedure, used experimental conditions should be rationalized either by the objective of the study or by field situations to be mimicked. These factors include [5,6]: elemental composition of used Fe0 materials, Fe0 pre-treatment (e.g., acid wash), Fe0 particle size (mm, μm, nm), buffer application, the molar ratio of Fe0 to contaminant (Fe0 mass loading and initial contaminant concentration), volume of the bottles used in the experiment, volume of model solution added, mixing operations (bubbling, shaking, stirring), geometry of the reaction vessel, experimental duration or reaction time. 1 The experimental conditions of Cheng et al. [1] clearly show that the set up is not appropriate to adequately achieve the goal of the study (Tab. 1). The molar ratios Fe0/4-Chlorophenol are too different for conventional Fe0 (R20 and R40) and synthesized nanosized Fe 0 (R50 and R100) respectively. R-values for experiments with conventional Fe0 clearly indicate high Fe0 excess (750 to 3000) whereas in experiments with nanosized Fe , 0 Fe0 excess were clearly lesser (7 to 14). This huge difference in the respective amount of conventional and nanosized Fe0 is not justified by the authors and complicates an objective discussion of the results, since no experiment with stoichiometric Fe0 amount was performed to serve as reference. An approach for a better interpretation of experimental results is presented below. Another critical point of Cheng et al[1] work’s is the used shaking intensity of 150 rpm. It is difficult to imagine which real world situation is mimicked by such a high shaking intensity. While mixing operations are generally considered as an important tool to facilitate the transport of contaminant to the Fe0 surface, and therefore, speed up the reduction kinetics, their role in enhancing Fe0 oxidation and hindering the formation of oxide film on Fe0 is often ignored. Therefore mixing intensities above natural turbulences should be avoided. In comparing the reactivity of Fe0 materials for the removal of a given contaminant from aqueous solutions, two basic principles are available for the rationale selection of the masses to be used: (i) the reaction stoichiometry, and (ii) the Fe0 specific surface area. Whereby, the Fe0 specific surface area should be coupled to the reaction stoichiometry. On the basis of the reaction stoichiometry, the used Fe0 mass is ideally equal to the equivalent amount of contaminant to be removed. However, data from the synthetic chemistry have shown that an Fe0 excess is always necessary. Because Fe0 materials usually content > 90 % Fe, the difference in the used masses should be minimal. This approach does not directly take into account obvious effects of particle size on the kinetics of Fe0 oxidation (increased oxidation rate with decreasing particle size). However, by using stoichiometric amounts of Fe0 (mm, μm and nm) the effects of particle size can be properly discussed. Cheng et al. [1] 2 have used various Fe0 excess for both material classes (Tab. 1). Because the variations for both material classes vary by two orders of magnitude (3000:15 = 200), the obtained results can not be properly discussed and have mostly a qualitative character. Clearly, if the authors have had used the same Fe0 excess for conventional and nanosized Fe0 (e.g factor 7 for conventional Fe0 or 100 mg in 150 mL), the reactivity comparison would have been eased. The reactivity comparison on the basis of the available surface area (S) is based on the determination of the specific surface area (SSA in m2/g). Assuming a spherical particle, SSA can be calculated by the following equation:[7] Surface Area πd2 6 SSA≅ S = --------------------- = ------------ = ---- (Eq. 1) Mass ρ[(π/6)d3] ρd Where ρ is the iron density. For two Fe0 materials of different particle diameter (d1, d2), the surface area ratio can be deduced from Eq. 1 by the following equation: S1 d2 ------ = ---- (Eq. 2) S2 d1 Equation 2 suggests that, if the specific surface area of one material (S1 or S2) is known, then the used mass of the other material can be calculated from its SSA-value. The comparison on the SSA basis is the most rigorous approach as heterogeneous processes are surface controlled. However, the materials have to be hold in suspension to assure the direct availability of the total surface. For this purpose, the use of vigorous mixing operations (above natural turbulences) is operationally justified. Because Cheng et al. [1] (d2/d1 ≤ 2700) have not reported any SSA-value, the appropriated amounts of individual Fe0 materials can not be further discussed here. Ideally, one should start with the mass of nanosized Fe0 yielding the monolayer Fe0 surface coverage. If the surface (cross section) of the contaminant molecule is not available, one should start with the mass of nanosized Fe0 yielding stoichiometric contaminant reduction and deduced the mass corresponding mass of 3 conventional Fe0. Then the masses can be increased in the same proportion depending on the observed removal goals. The above remarks corroborate the necessity of a unified experimental procedure for the investigation of the processes of contaminant removal in Fe0–H2O systems. A particular attention should be paid to the reaction stoichiometry (molar ratio Fe0/contaminant). Generally, the Fe0 mass loading (ρm in g/L) is given and considered for example in the expression of kobs to account for the available surface area [8]. But ρm alone is meaningless, even though Fe0 is in large excess in the large majority of works [9,10]. In conclusion, when two (or more) Fe0 materials to be employed in an investigation possess different surface areas, differences in surface area loading need to be accounted for in order to make accurate reactivity comparisons. For each Fe0 material, the area loading (in m2 surface/L) is found by multiplying the B.E.T. surface area (in m2/g) times the employed mass loading (ρm in g/L). Clearly, comparisons have to be performed on the basis of surface area loadings as the surface area (particle size) is not the only reactivity determining factor. References [1] R. Cheng, J.-L. Wang, W.-X. Zhang, Comparison of reductive dechlorination of p- chlorophenol using Fe and nanosized Fe0 0, J. Hazard. Mater. 144 (2007) 334–339. [2] R.W. Gillham, Discussion of “nano-scale iron for dehalogenation”. by Evan K. Nyer and David B. Vance (2001), Ground Water Monit. Remed. 21 (2), 41–54, Ground Water Monit. Remed 23 (2003), 6–8. [3] M. Büchler, P. Schmuki, H. Böhni, Iron passivity in borate buffer, J. Electrochem. Soc. 145 (1998) 609–614. [4] R.D. Vidic, R.D. Suidan, Role of dissolved oxygen on the adsorptive capacity of activated carbon for synthetic and natural organic matter, Environ. Sci. Technol. 25 (1991), 1612– 1618. 4 [5] S. Choe, Y.Y. Chang, K.Y. Hwang, J. Khim, Kinetics of reductive denitrification by nanoscale zero-valent iron, Chemosphere 41 (2000) 1307–1311. [6] H. Song, E.R. Carraway, Reduction of Chlorinated Methanes by Nano-Sized Zero-Valent Iron. Kinetics, Pathways, and Effect of Reaction Conditions, Environ. Eng. Sci. 23 (2006), 272–284. [7] C. Macé, S. Desrocher, F. Gheorghiu, A. Kane, M. Pupeza, M. Cernik, P. Kvapil, R. Venkatakrishnan, W.-X. Zhang, Nanotechnology and groundwater remediation: A step forward in technology understanding, Remed. J. 16 (2006), 23–33. [8] D.M. Cwiertny, A.L. Roberts On the nonlinear relationship between kobs and reductant mass loading in iron batch systems, Environ. Sci. Technol. 39 (2005) 8948–8957. [9] C. Noubactep, About the operation of reactive walls: the origin of the premise that quantitative contaminant reduction occurs at the surface of elemental iron materials, (in German) wlb Wasser, Luft und Boden 50 (3/4), TerraTech 3-4 (2007), TT10–14. [10] C. Noubactep, A. Schöner, G. Meinrath, Mechanism of uranium (VI) fixation by elemental iron, J. Hazard. Mater. B132 (2006), 202–212. 5 Table 1: Summary of the experimental conditions of Cheng et al. [1]. ZVI refers to the Fe0 material; m1ZVI is the mass of Fe0 material containing the equivalent amount of Fe0 for stoichiometric reduction of 4-Chlorophenol; mZVI is used amount of Fe0 material; The ratio mZVI/m1ZVI gives the excess factor (7 to 1500); ρm (g/L) is the Fe0 mass loading, Ri is the molar Fe0/comtaminant ratio at i (mg/L) initial concentration. 4-Chlorophenol (4- CP): 128.6 g/mol. 4-Chlorophenol Iron material (Fe0) [4-CP]0 [4-CP]0 V n m1ZVI mZVI mZVI/m1ZVI (mg/L) mM (L) (mM) (mg) (mg) (-) 20 0.156 0.15 0.0233 1.33 2000 1500 40 0.311 0.15 0.0467 2.66 2000 750 50 0.389 0.01 0.0039 0.22 3 14 80 0.622 0.01 0.0062 0.35 3 9 100 0.778 0.01 0.0078 0.44 3 7 common iron particles (conventional Fe0) mZVI mFe0 nFe0 ρm nFe0 R20 R40 (g) (g) (mol) (g/L) (mM) 2 1.96 0.035 13.3 35 1500 750 4 3.92 0.070 26.7 70 3001 1500 synthesized iron particles (nanosized Fe0) mFe0 V nFe0 ρm nFe0 R50 R100 (g) (L) (mol) (g/L) (mM) 0.003 0.01 0.000054 0.3 0.054 14 7 6