Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
Size effect in reactivity of copper nanoparticles to
carbon tetrachloride degradation
Ya Hsuan Liou, Shang Lien Lo
, Chin Jung Lin
Research Center for Environmental Pollution Prevention and Control Technology, Graduate Institute of Environmental Engineering, National Taiwan University, Taipei 106, Taiwan
a r t i c l e
i n f o
Article history:
Received 11 August 2006 Received in revised form 13 December 2006 Accepted 8 January 2007 Available online 6 March 2007 Keywords:
Carbon tetrachloride Nanoparticle Size effect Zero valent metal
a b s t r a c t
Surface area-normalized rate constants (kSA) of reaction between metallic nanoparticles and reducible contaminants, such as chlorinated hydrocarbons, heavy metals, and nitrate, have been reported to be dramatically increased as compared to that of commercial metallic powder. However, kSAfor individual pollutants in previously published data vary by as much as 1–2 orders of magnitude and much of this variability is due to the effect of various sizes. The size dependence of the reactivity of nanoparticles is not yet fully understood; however, yielding nanoparticles with uniform size and without agglomeration during the period of reaction would demonstrate the effect of varying particle size. In this study, resin-supported zerovalent copper with average particle size of 7, 10, 18, 26, and 29, respectively, were successfully synthesized and evidenced no agglomeration during the reaction period of 10 h. The kSAof copper nanoparticles (kn,SA) was 110–120 times higher than that of powdered copper particles (kp,SA) when the copper particle size was about 10 nm. However, for diameters of 18–29 nm, the ratio of kn,SA/kp,SAwas around 10–20, indicating that the reactivity of small copper nanoparticles (10 nm) varies discontinu-ously. Thus, most variability in previous kSA is attributed to the presence of small nanoparticles.
&2007 Elsevier Ltd. All rights reserved.
1.
Introduction
Extensive researches over the past 15 years have demon-strated that chemical reduction of many substances in the environment can be coupled with oxidation of zerovalent metals (M0) (Gillham and O’Hannesin, 1994; Matheson
and Tratnyek, 1994; Orth and Gillham, 1996; Agrawal and Tratnyek, 1996). In the reductive dechlorination reaction by M0, for example, the anodic M0is oxidized into Mn+ions (n is valence), and chlorinated hydrocarbons, denoted as RCl, as electron acceptors are converted to hydrocarbons and chlor-ides under anaerobic conditions. The chemistry of the M0/RCl/H2O system is similar to that of the corrosion of M0 by RCl substituting oxygen as the oxidant. In this process, RCl
are mostly transformed into benign compounds such as hydrocarbons (RH) and chlorides (Cl
).
A surface-area-normalized kinetic model is most com-monly used for heterogeneous reactions between reducible contaminants and metallic particles. This model assumes that the degradation rate of a contaminant is first order with respect to both total metallic surface and contaminant concentration, expressed as follows:
d½Cw=dt ¼ kSAasrm½Cw ¼kSAra½Cw ¼kobs½Cw; (1) where kSA (L h1m2) is a surface area-normalized rate constant, as(m2g1) is the specific surface area of metal, rm (g L1) is the metal mass loading in the reactor, ra(m2L1of solution) is the surface area concentration of metal, kobsis the
0043-1354/$ - see front matter & 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2007.01.014
Corresponding author. Tel.: +886 2 23625373; fax: +886 2 23928821. E-mail address:[email protected] (S.L. Lo).
observed rate constant, and Cw is the concentration of a contaminant. The amount of available surface area is the most important factor that governs the reduction rate. Studies have demonstrated the linear relationship between kobsand rmfor carbon tetrachloride (Matheson and Tratnyek, 1994), 1,2-dibromo-3-chloropropane (Siantar et al., 1996), nitrobenzene (Agrawal and Tratnyek, 1996) and trichlor-oethene (Su and Puls, 1999). Additionally, reducing the diameter of the iron particles to vary the as values also proportionally accelerates the apparent rate of dechlorination reaction (Siantar et al., 1996).
Dimensions of metallic particles in the range 1–100 nm possess the advantages of larger specific surface area and higher surface reactivity over microscale powder to increase the chemical reduction rate of reducible contaminants. For example, the reduction rates for chlorinated methanes (Lien and Zhang, 1999; Feng and Lim, 2005), chlorinated ethenes (Wang and Zhang, 1997;Li et al., 2003), chlorinated benzenes (Lowry and Johnson, 2004), arsenic (Kanel et al., 2005), and nitrate (Liou et al., 2006) by iron nanoparticles on a mass basis were more than 2 orders of magnitude higher than those of commercial iron powder. Nanoparticles can be anchored onto a solid matrix for treatment of water, wastewater, and gaseous steams (Meyer et al., 2004). Additionally, direct subsurface injection of iron nanoparticles to effectively degrade chlorinated organic compounds has been demon-strated (Cantrell et al., 1997). The technology provides enormous flexibility for in situ or ex situ remediation. Values of kSA are most often used to compare reactivity of nano-particles to that of microscale powder, and are also critical parameters for the design of full-scale remediation opera-tions. With the abundance of kinetic data now available for many compounds, kSAvarying by 1–2 orders of magnitude for individual compounds can be observed. The kSAvalue for trichloroethene (C2HCl3) reduced by iron nanoparticles as reported by Wang and Zhang (1997) is 3 103L h1m2, nearly similar to the results of Liu et al. (2005)
(kSA¼2 103L h1m2), though not as high as that reported by He and Zhao (kSA¼2 102L h1m2). Moreover, the differences in kSA between the results of Lien and Zhang (1999)andFeng and Lim (2005)for carbon tetrachloride (CCl4) and chloroform (CHCl3) vary by 1 and 2 orders of magnitude, respectively. In this case, the utility of kSAvalues in the design of a treatment system would be questionable.
Iron nanoparticles were typically prepared by borohydride reduction of an aqueous iron salt (Wang and Zhang, 1997;
Choe et al., 2001;Ponder et al., 2000;Schrick et al., 2002;Liao et al., 2003). However, due to the extremely high reactivity, the initially formed nanoaprticles tended to be either oxidized quickly by surrounding media or to agglomerate during the reaction period, resulting in a rapid change in the number of reactive surface sites. Unfortunately, as determined by N2 isothermal adsorption (the BET surface area) presents the total surface area including both exposed surface of redox-active metal and redox-inredox-active metal oxide. This may not provide a true representation of the actual number of sites on the iron surface.
A more extreme source of variability in kSAresults from the effect of various sizes. Iron nanoparticles prepared by a solution method with NaBH4 as a reductant have sizes
between 1 and 100 nm with an average diameter of 57716 nm (Cao et al., 2005). Applying a stabilizer to modify the solution method has been developed for the formation of much smaller nanoparticles. Li et al. (2003) used cetyltri-methylammonium bromide as a stabilizer in the water-in-oil microemulsion system to obtain the average particle size of about 10 nm. The results of their research indicated that the initial dechlorination rate for microemulsion nanoparticles was 2.6 times higher than that for the solution nanoparticles.
He and Zhao (2005) applied a water-soluble starch as a stabilizer to prepare starched-iron nanoparticles. The mean particle size was 14712 nm with much of the particles less than 10 nm. The starched nanoparticles exhibited values of kSA10 times higher than those of the solution nanoparticles when used for dechlorination of C2HCl3. Additionally, our previous study (Liou et al., 2006), differing iron precursor concentrations to obtain various nanoparticle size distribu-tion, indicated that the reactivity of iron nanoparticles with diameters of 9–10 nm for the denitrification of aqueous nitrate was higher than that of iron nanoparticles with diameters of 20–60 nm and of 20–70 nm. The most likely cause for the high chemical reactivity of small iron nano-particles is the rise in adsorbed atomic hydrogen (Hads) production on its surface. The peculiarity of these nanopar-ticles is due to their intrinsic heterogeneities: presence of different facets and presence of low coordinated edge sites (Henry, 2000). Thus, particle size is an important factor in controlling its reactivity.
In this study, the use of a support medium to stabilize and isolate a metal particle to clarify relations between size of copper nanoparticles and the reactivity to CCl4dechlorination avoids the effect of agglomeration. The use of copper as a reductant was preferred due to the following: (1) The reliable in situ measurement of the surface area of resin-supported copper particles by temperature-programmed reduction (TPR) has been demonstrated (Grift et al., 1991), and (2) nanoscale zero valent copper exhibits relative stability in water, avoiding reaction with surrounding media. Nanoparticles with uni-form size, and absence of agglomeration during the period of reaction would demonstrate the effect of varying particle size on its reactivity to the dechlorination of CCl4.
2.
Experimental section
2.1. Chemicals
All aqueous solutions were made in water purified with a Milli-Q system (18.2 MO cm1). High-pressure liquid chroma-tography grade carbon tetrachloride was obtained from Aldrich (USA). Copper (P) nitrate trihydrate was from Alfa (USA). Copper powder was obtained from Riedel-de Haen (approx. 0.04 mm, 499.5%, GR grade, USA).
2.2. Preparation of copper nanoparticles coated resin
Firstly, the resin was exchanged with desired concentration of Cu2+from an aqueous solution of Cu(NO3)23H2O, followed by thorough washing to remove excess physisorbed copper ions. After washing, the fresh samples were dried in air at 120 1C for
12 h. The dried samples were reduced from ambient to 300 1C at 10 1C min1in a flow of H2/N2(20 vol%, 100 mL min1) and keeping it at 300 1C for 3 h, then cooled down to room temperature in the reducing gas atmosphere. According the above description, five samples were prepared at various copper loading including 5, 10, 20, 30, 50 mg-Cu g-resin1.
2.3. Characterization of copper nanoparticles coated resin
The oxidation states of copper on the resin were identified by electron spectroscopy for chemical analysis (ESCA). The ESCA measurements were perform by using a Vacuum Generators ECSALAB MKP photoelectron spectrometer (East Grinsted, UK) with an Al Ka X-ray source (1486.6 eV) and a hemisphe-rical 150 mm mean radius electron analyzer with a take-off angle of 901. During the data acquisition, the pressure in the sample chamber did not exceed 9.0 109Pa. Hitachi H-7100 transmission electron microscopy (TEM) was used to char-acterize the size distribution of the metal particles.
2.4. Surface area and average particle size of attached copper particles
Two experiments were performed by TPR studies with the apparatus similar to that described previously (Bond and Namijo, 1989). TPR apparatus of standard design was allowed flows of (1) pure helium gas, (2) 10 vol% H2in Ar, (3) pure N2O to be passed sequentially through the reactor and the thermal conductivity detector (TCD). Before these experiments, the fresh sample of Cu2+on the resin was dried in air at 120 1C for 18 h. Then, it was heated in a flow of helium (50 mL min1) from room temperature to 250 1C at 10 1C min1. Dwelling for 30 min at 250 1C to remove the volatiles and then cooled down to room temperature. In the first TPR, a flow of H2/Ar (10 vol%, Hoehloos) was used as reducing gas. The oven temperature was programmed from ambient to 350 1C at 10 1C min1and held for 60 min. Subsequently, the second TPR was to estimate the specific surface area and the particle size of Cu0on the resin. In that, pure N2O (99.998 vol%, Matheson) was used as oxidizing gas, helium gas (99.99 vol%, Hoekloos) as flushing gas and H2/Ar (10 vol%, Hoehloos) as reducing gas. After the first TPR experiment, these gases were passed sequentially through the oven and detector. A sample of Cu2+ on the resin was reduced by the first TPR experiment, and then cooled down to room temperature. The flow of H2/Ar was then replaced by helium gas (60 mL min1) to purge the reduced sample for 10 min. Exposing the reduced sample to a flow of pure N2O (60 mL min1) at 100 1C for 120 s, and then cooled down to room temperature in a flow of helium gas (60 mL min1). Finally, a flow of H
2/Ar (10 vol%, 60 mL min1) was introduced to reduce CuO2on the Cu0surface, named the second TPR.
2.5. Reactor system
The degradation and sorption of CCl4 by Cu0/resin were measured in the closed batch system with zero headspace. In these systems, 0.2 g Cu0/resin and 14.8(70.05) mL Ar-purged unbuffered Milli-Q water were added into 15 mL amber serum vials (Supelco). A 100 mL aliquot of CCl4 (700 mg L1)
was then added under the water level. Immediately after CCl4 addition, the vials were capped with Teflon silicone septa and aluminum seals and then mixed on a rotary shaker (50 rpm) at room temperature (2371 1C) in the dark.
2.6. Sample analysis
Each vial was analyzed by liquid–liquid and liquid–solid extraction using n-hexane as a solvent to determine aqueous phase and total CCl4concentration according the sampling method described in our previous study (Lin et al., 2005). The CCl4 concentrations were measured using a HP5890 GC equipped with a DB-624 capillary column and an electron capture detector operated in the splitless mode. The total CCl4 mass (mg vial1) come from the sum of both the measured masses in the liquid–liquid and the liquid–solid extractions. Control total (mass per vial) concentrations were stable over the reaction period. The extraction recoveries for CCl4 from aqueous and sorbed phases in the blank resin vials (0.2 g-resin vial1) were ranged from 91% to 103%, that reasonable assumed this method can extract total CCl4mass in vial. All experiments were duplicated or triplicated.
3.
Results and discussion
3.1. Characterization of copper nanoparticles coated resin
The oxidation states of the resulting copper on the resin (5 mg-Cu g-resin1) were characterized by ESCA. The XPS spectra of both the fresh prepared sample and the used one are presented inFig. 1(a) and (b), respectively. These spectra consist of the Cu2p3/2main photoelectron peak (centered at 932.8–935.1 eV). The Cu2p3/2 peaks at binding energy of 932.8 eV found in the XPS spectra of fresh prepared Cu/resin (Fig. 1(a)) indicated nearly all copper in Cu/resin is present as Cu0. This preparation process of Cu/resin was proven to be high enough to ensure nearly complete reduction to Cu0. After the dechlorination reaction of CCl4for 10 h, three Cu2p3/ 2peaks were found in the XPS spectra at binding energy of 932.8, 933.7 and 935.1 eV, corresponding to Cu0, CuO and CuCl2 with 4, 10, 86% of total copper loading by mole (the deconvolution results are shown inFig. 1(b)). Thus, the XPS spectra of Cu/resin indicates the relative abundance of Cu (II) species on its surface after dechlorination reaction.
The morphology and size of the resulting Cu0on the resin (10 mg-Cu g-resin1) was viewed with TEM.Fig. 2(a)showed the copper particles were nearly spherical in shape and uniform in size with the average particle size about 10 nm. After the dechlorination reaction of CCl4, copper particles didn’t significant aggregate and presented similar particle size distribution (Fig. 2(b)).
According to the set of experiments of Cu0 surface oxidation in N2O flow at various temperatures by Evans et al. (1983), the surface area of Cu0particles determined by TPR was close to its BET surface area when the cross-sectional area of the active adsorptive atom was assumed to be 1.4 1019Cu-atoms m2. As described byGrift et al. (1991), the determination of the surface area of attached copper particles (SCu, m2-Cu g-resin1) is based on the measurement
938 936 934 932 930 928 0 200 400 600 800 1000 1200 1400 1600 Cu(OH)2 Cu0 Cu 2p3/2 CPS Binding Energy, eV Peak (ev) FWHM (ev) Area (%) Cu0 932.8 1.75 94 Cu(OH)2 935.1 1.3 6 940 938 936 934 932 930 0 200 400 600 800 1000 1200 1400 1600 Cu0 CuO CuCl2 Cu 2P3/2 CPS Binding Energy, ev Peak (ev) FWHM (ev) Area (%) Cu0 932.8 1.48 4 CuO 1.80 10 CuCl2 935.1 3.44 Centre Centre 933.7 86
Fig. 1 – XPS spectra of Cu2p3/2for (a) before; (b) after the dechlorination reaction for 10 h.
Fig. 2 – The TEM images of resin (10 mg-Cu g-resin1) for (a) before; (b) after the dechlorination reaction for 10 h. The metallic particles appear as spots with high contrast.
of hydrogen consumption after surface oxidation of the copper by N2O.
Step 1:
2Cu þ N2O ! Cu2O þ N2ðcopper surface atoms onlyÞ: (2) Step 2:
Cu2O þ H2!2Cu þ H2O: (3)
Compared to the area of the TCD signals of 1 mL H2 (40.9 mmol at 1 atm, 25 1C) passing through the reactor of TPR, the H2 consumption of various copper ions loading samples in the first and second TPR would be obtained; the numbers of total and surface copper atoms were calculated by multiplying a factor that expresses the stoichiometry of Eqs. (2) and (3). Copper surface area (SCu) on the resin would be readily calculated with above related values.
Nsurf Cu¼ ðAsurf Cu=A1 mL H2Þ40:9 10 6N
sðmoleÞ: (4)
SCu¼ ðNsurf;CuNavÞ=ð1:4 1019WresinÞ ðm2Cu g sample1Þ; (5) where Nsurf Cuis the number of surface copper atoms; Asurf Cu is the area of the TCD signal for the second TPR; A1 mL H2is the area of the TCD signal for 1 mL H2; Ns( ¼ 2) is the stoichio-metry of Eq. (5); Navis Avogadro constant (6.023 1023mole1); Wresinis the mass (g) of the resin in the reactor of TPR.
The dispersion (D) of copper was defined as the ratio of the number of copper surface atoms and total number of copper atoms (Eq. (8)). In other words, D is the fraction of the total of copper accessible to the reactant molecules. AsGrift et al. (1991)described, the measurement of dispersion of Cu on the resin was based on the amount of the H2 consumption of the first and second TPR, named Y and X, respectively. The reactions corresponding to the first and second TPR, respec-tively, are Eqs. (6) and (7).
PolymerðSO3Þ2þH2!Cu0þPolymerðSO3Þ2;
hydrogen consumption ¼ Y; (6)
Cu2O þ H2!2 Cu0þH2O; hydrogen consumption ¼ X: (7) The dispersion of Cu on resin was calculated by
D ¼ Nsurf Cu=Ntot Cu¼2X=Y; Dp1; (8) where Ntot,Cuis total number of copper atoms. Assuming the copper particles in spherical shape, Boundart and Djega-Mariadassou (1984) proposed an approximate conversion between dispersion D and average particle size dAV(nm).
dAV¼0:9=dispersion ðnmÞ: (9)
Figs. 3(a) and (b)show the profiles of the first and second TPR, respectively. The values of Xiand Yi(i ¼ 5, 10, 20, 30, 50) of these samples were normalized to the area of the TPR signal for 1 mL H2. To establish the same base of the TPR profiles for various Cu loadings, samples of different weights were added to the reactor of TPR, 0.4 g for the 5 mg-Cu g-resin1, 0.3 g for the 10 mg-Cu g-resin1, and 0.2 g for the 20, 30 and 50 mg-Cu g-resin1. Significant signals for initially consumed H2were appeared at 250–290 1C in the first TPR, but at 150–180 1C in the second TPR. The difference of activated energy between sorbed Cu ions on the resin and Cu2O reduced by H2may account for this observation.Table 1lists
the resulting values of SCu, D and dAVof Cu on the resin. As Cu loading on the resin increases, the SCuincreased from 0.50 to 0.68 Cu-m2 g-resin1 for Cu loading between 5 mg-Cu g-resin1 and 10 mg-Cu g-resin1. However, it holds at the amount of approximate 0.55 Cu-m2 g-resin1 for 20 mg-Cu g-resin1up to 50 mg-Cu g-resin1. Consequently, increasing the Cu loading on resin served initially to increase the number density rather than the size of particles on the resin. Then, a further increase led to agglomeration of Cu particles at the temperature of reduction, simultaneously indicated by an increase in the average particle size. The prepared resin-supported copper were characterized by ESCA, TEM, and TPR, indicating the values of exposed surface area of Cu0on the resin without significant change during the period of reaction.
3.2. Dechlorination of CCl4
A conceptual model of the nanoscale zero valent copper-resin-CCl4-water system incorporating concurrent the degra-dation reaction by Cu0and sorption/desorption by the resin is similar to that proposed by Burris et al. (1998). The disap-pearance of CCl4 from aqueous solutions may be due to degradation reactions by zero valent copper or sorption onto the resin.
dCw=dt ¼ kaCwþasCwadCs; (10) where Cw and Cs are the aqueous, and sorbed-phase CCl4 concentrations mg per vial; as and ad are the soprtion and desorption rate constants, respectively. The loss of total mass of CCl4 (CT) in a closed reactor was attributed to the degradation reaction by Cu0and fitted with a first-order rate equation of the form
dCT=dt ¼ kobsCw¼kSAasrmCw: (11) Eqs. (10) and (11) state the decline in the concentration of target contaminants in the batch system was caused only by reduction reaction at the Cu0surface; however, the apparent concentration in the aqueous phase was reduced by the combination of sorption and reduction reaction. Regardless of the mechanism of sorption, Eq. (10) or (11) can be simplified as Eq. (1).Fig. 4(a)is the result of controlled experiment with 91–103% recovery of total CCl4mass. Fig. 4(b)–(f)separately present experimental results (the residual aqueous and total CCl4mass in the vial) for the reaction of 0.2 g of 5, 10, 20, 30, 50 mg-Cu g-resin1with 1.05 mg vial1 initial mass of CCl4. Additionally,Fig. 4(g)presents the residual aqueous CCl4mass in the vial containing 0.3 g powdered copper over the period of reaction. As presented inFig. 4(b)–(f), both total and aqueous CCl4mass declined substantially. The rate of loss of total CCl4 mass obtained using 5 mg-Cu g-resin1 resembled that of 10 mg-Cu g-resin1, and both losses were significantly faster than those obtained using 20, 30, and 50 mg-Cu g-resin1. The first-order observed rate constants (kobs) of the dechlorination of CCl4in contact with Cu g-resin1were obtained from the slope of a plot of total CCl4loss rate (dCT/dt) versus CCl4 concentration in the aqueous phase, Cw in mg-CCl4vial1, using linear least-squares analysis. The kobs values were 1.64470.174, 2.04670.132, 0.34270.042, 0.20970.011 and 0.26470.03 h1 for 5, 10, 20, 30, and 50 mg-Cu g-resin1, respectively (shown in Table 1). The values of kSA were
obtained by normalizing kobs according to the SCu and the mass concentration of the sample (0.2 g Cu0/resin per 15 mL amber serum vial). As indicated in Table 1, the kSA of nanoscale Cu particles (named kn,SA) was 110–120 times higher than that (named kp,SA) of commercial powdered Cu
particles when the Cu0 particle size was about 10 nm. However, for diameters of 20–30 nm, the ratio of kn,SA/kp,SA was around 10–20. This indicates the size dependence of the reactivity of small copper nanoparticles varies discontinuously.
Table 1 – The specific surface area, dispersion and average diameter of particles, and its reactivity to CCl4
Sample (mg-Cu g-resin1) Wsample (g) H2consumption (mole g-resin1) SCua(m2-Cu g-resin1) Dispersion (D) dAV (nm) kobvb(h1) kSA102b (L m2h1) kn;SA kp;SA 5 0.4 11.51 0.50 0.127 7 1.64470.174 24.9072.64 125 10 0.3 15.86 0.68 0.089 10 2.04670.132 22.4971.45 112 20 0.2 13.04 0.56 0.051 18 0.34270.042 4.5770.57 23 30 0.2 12.02 0.52 0.034 26 0.20970.011 3.0370.16 15 50 0.2 12.28 0.53 0.031 29 0.26470.03 3.7570.42 14 Powdered Cu 0.3 23.87 1.03 — — 0.02870.002 0.2070.02 1 a The S
Cuof these samples are given in square meter per gram resin, Cu-m2g-resin1, but in square meter per gram Cu, Cu-m2g-Cu1, for
commercial powdered Cu.
b Average795% confidence level.
0 100 200 300 Intensity, a.u. Y50=391, 50mg-Cu/g-resin, 0.2g Y40=347, 40mg-Cu/g-resin, 0.2g Y10=177, 10mg-Cu/g-resin, 0.4g Y20=248, 20mg-Cu/g-resin, 0.3g Y30=262, 30mg-Cu/g-resin, 0.2g Isothermal Temperature, °C Intensity, a.u. Cu powder 0.3g, Xp=18 50mg-Cu/g-resin 0.2g, X50=6 40mg-Cu/g-resin 0.2g, X40=6 30mg-Cu/g-resin 0.2g, X30=7 20mg-Cu/g-resin 0.3g, X20=12 10mg-Cu/g-resin 0.4g X10=11 Isothermal 300 200
a
b
Temperature, °CFig. 3 – Consumption of H2during temperature programmed reduction of nano-Cu0/resin: (a) after complete oxidation (the first TPR); (b) after surface oxidation (the second TPR).
4.
Conclusions
In this work, the size dependence of reactivity toward carbon tetrachloride was studied via synthesizing copper nanoparti-cles with uniform size dispersed on a cation resin. The kn,SA was sharply increased by a factor of 110–120 when Cu0 particle size was close to 10 nm. However, for diameters of 20–30 nm, the ratio of kn,SA/kp,SAwas around 10–20, indicating the reactivity of small Cu particles (10 nm) varies discon-tinuously. Thus, most variability in previous kSAis attributed to the presence of small nanoparticles.
Acknowledgement
The authors would like to thank the National Science Council of the Republic of China for financially supporting this research under Contract No. NSC 92-2211-E-002-006.
RE F ER E NC E S
Agrawal, A., Tratnyek, P.G., 1996. Reduction of nitro aromatic compounds by zero-valent iron metal. Environ. Sci. Technol. 30, 153–160.
Bond, G.C., Namijo, S.N., 1989. An improved procedure for estimating the metal surface area of supported copper catalysts. J. Catal. 118, 507–510.
Boundart, M., Djega-Mariadassou, G., 1984. Kinetics of Hetero-geneous Catalytic Reactions. Princeton University Press, Princeton, NJ, pp. 20–26.
Burris, D.R., Allen-King, R.M., Manoranjan, V.S., Campbell, T.J., Loraine, G.A., Deng, B., 1998. Chlorination ethene reduction by
cast iron: sorption and mass transfer. J. Environ. Eng. 10, 1012–1019.
Cantrell, K.J., Kaplan, D.I., Gilmore, T.J., 1997. Injection of colloidal Fe0particles in sand with shear-thinning fluids. J. Environ. Eng., ASCE, 786–791.
Cao, J., Elliott, D., Zhang, W.X., 2005. Perchlorate reduction by nanoscale iron particles. J. Nanopart. Res. 7, 499–506. Choe, S., Lee, S.H., Chang, Y.Y., Hwang, K.Y., Khim, J., 2001. Rapid
reductive destruction of hazardous organic compounds by nanoscale Fe0. Chemosphere 42, 367–372.
Evans, J.W., Wainwright, M.S., Bridgewater, A.J., Young, D., 1983. On the determination of copper surface area by reaction with nitrous oxide. J. Appl. Catal. 7, 75–83.
Feng, J., Lim, T.T., 2005. Pathways and kinetics of carbon tetrachloride and chloroform reductions by nano-scale Fe and Fe/Ni particles: comparison with commercial micro-scale Fe and Zn. Chemosphere 59, 1267–1277.
Gillham, R.W., O’Hannesin, S.F., 1994. Enhanced degradation of halogenated aliphatics by zero-valent iron. Ground Water 32, 958–967.
Grift, C.J.G., Wielers, A.F.H., Joghi, B.P.J., Beijnum, J., Boer, M., Versluus-Helder, M., Gues, J.W., 1991. Effect of the reduction treatment on the structure and reactivity of silica-supported copper particles. J. Catal. 131, 178–189.
He, F., Zhao, D., 2005. Preparation and characterization of a new class of starch-stabilized bimetallic nanoparticles for degra-dation of chlorinated hydrocarbons in water. Environ. Sci. Technol. 39, 3314–3320.
Henry, C.R., 2000. Catalytic activity of supported nanometer-sized metal clusters. Appl. Surf. Sci. 164, 252–259.
Kanel, S.R., Manning, B., Charlet, L., Choi, H., 2005. Removal of arsenic(III) from groundwater by nanoscale zero-valent iron. Environ. Sci. Technol. 39, 1291–1298.
Li, F., Vipulanandan, C., Mohanty, K.K., 2003. Microemulsion and solution approaches to nanoparticles iron production for degradation of trichloroethylene. J. Colloids Surf. A: Physico-chem. Eng. Aspects 223, 103–112.
g
0 0.02 0.04 0.06 0.08 0.1 0.12 0 2 4 6 Reaction time, hr Mass of CCl 4 , mg/vial 0 0.02 0.04 0.06 0.08 0.1 0.12 0 2 5 Reaction time, hr Mass of CCl 4 , mg/vial 0 0.02 0.04 0.06 0.08 0.1 0.12 0 2 4 6 Reaction time, hr Mass of CCl 4 , mg/vial 0 0.02 0.04 0.06 0.08 0.1 0.12 0 2 4 5 6 Reaction time, hr Mass of CCl 4 , mg/vial 0 0.02 0.04 0.06 0.08 0.1 0.12 0 2 4 6 Reaction time, hr Mass of CCl 4 , mg/vial 0 0.02 0.04 0.06 0.08 0.1 0.12 0 3 6 Reaction time, hr Mass of CCl 4 , mg/vial 0 0.02 0.04 0.06 0.08 0.1 0.12 0 3 4 5 6 Reaction time, hr Mass of CCl 4 , mg/viala
b
1 3 5 1 3 4 6c
1 3 5d
1 3e
f
1 3 5 1 2 4 5 1 2Fig. 4 – The residual aqueous and total CCl4mass with 95% confidence interval for (a) control with resin only, (b) 5 mg-Cu g-resin1; (c) 10 mg-Cu g-resin1; (d) 20 mg-Cu g-resin1; (e) 30 mg-Cu g-resin1; (f) 50 mg-Cu g-resin1; (g) Cu powder. (J) referred to the residual aqueous CCl4mass; (K) referred to the residual total CCl4mass.
Liao, C.H., Kang, S.F., Hsu, Y.W., 2003. Characteristics of reductive removal of nitrate by suspended zero-valent iron powder. J. Chinese Inst. Environ. Eng. 13, 251–261.
Lien, H.L., Zhang, W.X.J., 1999. Transformation of chlorinated methanes by nanoscale iron particles. Environ. Eng., ASCE 9, 1042–1047.
Lin, C.J., Lo, S.L., Liou, Y.H., 2005. Degradation of aqueous carbon tetrachloride by nanoscale zerovalent copper on a cation resin. Chemosphere 59, 1299–1307.
Liou, Y.H., Lo, S.L., Kuan, W.H., Lin, C.J., Weng, S.C., 2006. Effect of precursor concentration on the characteristics of nanoscale zerovalent iron and its reactivity of nitrate. Water Res. 40, 2485–2492.
Liu, Y., Majetich, S.A., Tilton, R.D., Sholl, D.S., Lowry, G.V., 2005. TCE dechlorination rates, pathways, and efficiency of nanoscale iron particles with different properties. Environ. Sci. Technol. 39, 1338–1345.
Lowry, G.V., Johnson, K.M., 2004. Congener-specific dechlorination of dissolved PCBs by microscale and nanoscale zerovalent iron in a water/methanol solution. Environ. Sci. Technol. 38, 5208–5216.
Matheson, L.J., Tratnyek, P.G., 1994. Reductive dehalogenation of chlorinated methanes by iron metal. Environ. Sci. Technol. 28, 2045–2053.
Meyer, D.E., Wood, K., Bachas, L.G., Bhattacharyya, D., 2004. Degradation of chlorinated organics by membrane-immobi-lized nanosized metals. Environ. Progr. 23, 232–242.
Orth, W.S., Gillham, R.W., 1996. Dechlorination of trichloro-ethene in aqueous solution using Fe0. Environ. Sci. Technol. 30, 66–71.
Ponder, S.M., Darab, J.G., Mallouk, T.E., 2000. Remediation of Cr(VI) and Pb(II) aqueous solutions using supported nanoscale zero-valent iron. Environ. Sci. Technol. 34, 2564–2569.
Schrick, B., Blough, J.L., Jones, A.D., Mallouk, T.E., 2002. Hydrode-chlorination of trichloroethylene to hydrocarbons using bimetallic nickel-iron nanoparticles. Chem. Mater. 14, 5140–5147.
Siantar, D.P., Scheier, C.G., Chou, C.S., Reinhard, M., 1996. Treatment of 1, 2-Dibromo-3-chloropropane and nitrate-con-taminated water with zero-valent iron or hydrogen/palladium catalysts. Water Res. 30, 2315–2322.
Su, C., Puls, R.W., 1999. Kinetics of trichloroethene reduction by zero-valent iron and tin: pretreatment effect apparent activa-tion energy and intermediate products. Environ. Sci. Technol. 33, 163–168.
Wang, C.B., Zhang, W.X., 1997. Synthesizing nanoscale iron partical for rapid and complete dechlorination of TCE and PCBs. Environ. Sci. Technol. 31, 2154–2156.
