The enhancement methods for the degradation of TCE by zero-valent metals

The enhancement methods for the degradation of TCE by

zero-valent metals

Shu-Fen Cheng

_{, Shian-Chee Wu}

Graduate Institute of Environmental Engineering, National Taiwan University, Taipei, Taiwan, ROC Received 25 August 1999; accepted 3 November 1999

Abstract

Batch tests were performed to compare the degradation rates of TCE on Fe0_{and Zn}0_{. Our results indicated that the}

degradating capability of Zn0_{to TCE was nearly 10 times higher than that of Fe}0_{. On the other hand, the degradation}

rates of Fe0 _{or Zn}0_{in conjunction with other metals for reduction of TCE was investigated. The selected metals were}

nickel (Ni0_{) and palladium (Pd}0_{) both of which have a strong enhancement e€ect. The reduction rates of Zn}0_/Pd0_and

Zn0_/Ni0_{for TCE were the fastest. Fe}0_{that had lost its surface activity could be activated again by the addition of Pd}0_or

Ni0_{. Ó 2000 Elsevier Science Ltd. All rights reserved.}

Keywords: Chlorinated organic compound; Zero-valent metal; Bimetallics; Chloroethylene

1. Introduction

In Taiwan, chlorinated organic compounds, such as PCE, TCE, CCl4, CHCl3and CH2Cl2, are widely used in

industry. They are mainly used as solvents in degreasing, washing, extraction, foaming, spraying and manufac-turing, etc. In recent years, the underground water of many sites in Taiwan has been heavily contaminated by chlorinated organic compounds. For example, the RCA site in Taoyuan is contaminated by PCE and TCE, the Philips site in Hsinchu is contaminated by PCE, and the An-Shuenn site in Tainan is contaminated by penta-chlorophenol (PCP). There is, therefore, considerable interest in the remediation technique of sites contami-nated by chloricontami-nated organic compounds. The reduction power of zero-valent iron to chlorinated organic com-pounds has been a focus of investigation in recent years among the techniques that are used in the remediation of sites contaminated by chlorinated organic compounds

(Matheson and Tratnyek, 1994; Gillham and Stephanie, 1994; Smyth et al., 1995; Orth and Gillham, 1996; Agrawal and Tratnyek, 1996; Weber, 1996; O'Hannesin and Gillham, 1998).

The research on the zero-valent iron technique in recent years has shown that zero-valent iron has many drawbacks in practical applications. Firstly, after a short period of reactions zero-valent iron is liable to form an oxide ®lm on the surface, which subsequently reduces the reaction activity (Wang and Zhang, 1997). Secondly, the retention of the surface activity of zero-valent iron is dicult to maintain. Once Fe0 _{is in contact with air,}

even under proper storage, its reactivity towards chlo-rinated organic compounds is inevitably reduced (Cheng and Wu, 1998). Thirdly, there is considerable variation in the reactivity towards chlorinated organic compounds of Fe0 _{of di€erent origins. The reaction rates can di€er}

by up to three orders of magnitude (Su and Puls, 1999). Matheson and Tratnyek (1994) proposed that if Fe0

received an HCl acid washing process prior to its use, this could increase the surface reaction activity. How-ever, according to the research of Su and Puls (1999) on the e€ects of the acid prewashing and our previous re-search results, the acid-washing process not only tends

*_{Corresponding author.}

E-mail address: [email protected] (S.-F. Cheng).

0045-6535/00/$ - see front matter Ó 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 5 - 6 5 3 5 ( 9 9 ) 0 0 5 3 0 - 5

to cause many ®ne particles of Fe0_{powder to be lost, but}

also causes the Fe0_{to have a faster oxidation rate in the}

processes of washing and drying. The above factors cause the zero-valent iron technique to be largely re-stricted in its application to in situ remediation.

In recent years, a great deal of research has concen-trated on the improvement of the zero-valent iron technique. This research has included the use of ultra-sonic oscillation to remove the oxide ®lm on the Fe0

surface (Ruiz et al., 1998), the use of external voltage to maintain the surface activity of Fe0 _{(Cheng and Wu,}

1998), the use of the bimetallics technique (Appleton, 1996; Siantar et al., 1996), the use of palladized iron (Muftikian et al., 1995) and the use of synthesized nanoscale palladized iron (Wang and Zhang, 1997), etc. Matheson and Tratnyek (1994) proposed three pathways for the reduction dechlorination reactions of Fe0 _{to chlorinated organic compounds: (1) to direct}

electrons transfer from iron metal at the metal surface; (2) to catalyze hydrogenolysis by the H2 that is formed

by reduction of H2O during anaerobic corrosion; (3) to

reduction by the Fe‡2_{, which results from corrosion of}

the metal. From the viewpoint of electrochemical theo-ry, the ®rst pathway means that a metal with a stronger reducing power is advantageous to the dechlorination reactions of the chlorinated organic compounds. Ac-cording to electrochemical corrosion theory, in an oxy-gen-free Zn±H2O or Fe±H2O system containing H‡, the

surface of the metal (such as Zn0_{) will generate the}

fol-lowing reaction: Zn0_{‡ 2H}‡_{! Zn}‡2_{‡ H}

2"

According to electrolysis chemical theory, the reduction power of a cathode in an electrolysis system is mainly derived from the function of hydrogen atoms. Once the hydrogen atoms, through the reaction of H ‡ H ! H2",

combine together and form a hydrogen gas in a bubble form, they no longer participate in the reduction reac-tion. Furthermore, the accumulation of H2 gas bubbles

on the metal surface will hinder the progress of the reaction (Matheson and Tratnyek, 1994; Ballapragada et al., 1997). Due to the di€erent characteristics of metals, the formation rate of hydrogen gas on the sur-face of each metal varies; this is called the hydrogen overvoltage of the metal. The second pathway means that a metal with a higher hydrogen overvoltage is ad-vantageous to the reduction of the chlorinated organic compounds. The reduction potential of Zn0 _{is stronger}

than that of Fe0_{, and Zn}0 _{is a metal with a high}

hy-drogen overvoltage, second only to that of mercury. Hence, from the above two points of view, Zn0 _should

be the better selection in degradation of chlorinated organic compounds.

Recent studies pointed out that where zero-valent iron was used in conjunction with another metal for the

degradation of chlorinated organic compounds, the second metal primarily has the following functions: (1) as a catalyst (Wang and Zhang, 1997), (2) preventing the formation of the oxide ®lm on the surface of Fe0_(Wang

and Zhang, 1997), and (3) inducing Fe0_{to release}

elec-trons at a faster rate due to the di€erence of electric potentials between Fe0 _{and the second metal to reduce}

the chlorinated organic compounds (Gavaskar et al., 1998). Appleton (1996) proposed that nickel has the e€ect of accelerating the degradation of TCE by Fe0_.

Wang and Zhang (1997) synthesized nanoscale palla-dized iron to degradation of TCE and PCBs. Muftikian et al. (1995) used palladized iron to reduction of CCl4

and TCE. All these studies showed a signi®cant pro-motional e€ect.

The objectives of our study were to: (1) compare the degradation rates of di€erent Fe0 _{and Zn}0 _{to TCE in}

order to further evaluate the feasibility of using Zn0 _to

replace Fe0_{; (2) investigate the promotional e€ects of a}

combination Fe0_{or Zn}0_{with a second metal Ni}0_{or Pd}0

for degradation of TCE.

2. Experimental

2.1. Chemicals and materials

Chlorinated compound, trichloroethylene (TCE) was obtained from Merck (99.5+%, GR grade). TCE stock solution was prepared by weighting 0.0695 g of pure TCE solvent to dissolve in 100 ml of methanol. Meth-anol was obtained from Acros (99.8+%, PA grade). TCE aqueous solutions were made by diluting the stock so-lution with Milli-Q water. The Milli-Q water was spar-ged with Argon gas.

Two kinds of iron were used: one was obtained from Riedel-deHaen, powdered Fe0 _{(99+%, RG, made}

by reduction), the other was obtained from Aldrich, granular Fe0 _{(10±40 mesh, 99.999%, stored under}

ni-trogen). Zincs were obtained from three di€erent companies: Aldrich granular Zn0 _{(ÿ10 ‡ 50 mesh,}

99.8%, ACS reagent), Acros granular Zn0 _{(30 mesh,}

99.7%, PA grade), and Hanawa powdered Zn0 _(90.0%,

guaranteed reagent). Nickel was obtained from Aldrich powdered Ni0 _{(ÿ100 mesh, 99.99%). Palladium was}

obtained from Acros powdered Pd0 _{(99.99%, ACS}

reagent).

2.2. Experimental methods

Batch tests were conducted to investigate the degra-dation of TCE. TCE aqueous solutions were prepared by diluting the stock solution with Ar-sparged Milli-Q water to 3 mg TCE lÿ1_.

2.3. Single metal tests

Three grams of each metal (Fe0_{, Zn}0_{, Ni}0_{) were}

added to each brown serum vial (with measured internal volumes of 15 ml). Each vial was ®lled with TCE aqueous solution with no headspace, and was then sealed immediately with aluminum crimp caps with Te¯on faced septum. For each test, 10 vials containing TCE and metal and 10 controls containing TCE only were prepared for di€erent reaction times. All the vials were put on a shaker (oscillated frequency 130 rpm, at 25°C). At each sampling time, 5 ml of subsamples were transferred via a syringe from the sample to the other clean vials, and then were sealed immediately with alu-minum crimp caps with Te¯on faced septum. Before analysis, all the subsamples were stored in the oven (set at 25°C) for more than 0.5 h to let the TCE reach the equilibrium between the headspace and the aqueous phase.

2.4. Bimetallics tests

For the combination-with-palladium tests, 50 mg of palladium and 3 g of tested metal were added to each vial. For the combination with nickel tests, 3 g of nickel and 3 g of tested metal were added to each vial. The remaining steps were the same as for the single metal tests.

2.5. Analytical methods

The concentrations of TCE and its chlorinated products were determined by using the gas chromatog-raphy headspace equilibration method. For each sam-ple, 5 ll of the headspace gas was taken by using a glass gas syringe, and then was injected into the chromato-graph.

Analyses for TCE and its chlorinated products were conducted using a 5890II Hewlett Packard gas chro-matograph equipped with a 30 m  0:53 mm …ID† 3:0 lm (thickness), DB624 analytical column (J & W) and an electron capture detector (ECD). The tempera-ture was set as follows: oven temperatempera-ture: 60°C, injec-tion port temperature: 220°C, detector temperature: 250°C. Nitrogen was used as the carrier gas at a ¯ow rate of about 4.5 ml/min. The method detection limit (MDL) for TCE was 0.005 mg/l.

3. Results and discussion

3.1. The e€ects of Fe0_{characteristics}

Many research results lead to an unanimous con-clusion that the degradation rate of Fe0 _{to the}

chlori-nated organic compounds is in¯uenced by the

magnitude of the ``clean speci®c surface area'' of zero-valent iron (Matheson and Tratnyek, 1994; Weber, 1996). In the past, most studies emphasized the study of the e€ect of the speci®c surface area of Fe0_{. However,}

according to this research and recent research results (Cheng and Wu, 1998; Su and Puls, 1999), the in¯uence of ``cleanness'' on the reaction rate is much greater than the in¯uence of the magnitude of the speci®c surface area of Fe0_{. The source, quality, purity and freshness of}

Fe0_{have a signi®cant in¯uence on the reaction rate. This}

research used two types of Fe0 _{of di€erent sources,}

dif-ferent particle sizes, and di€erent unsealed time. One was a granular iron of 10±40 mesh, produced by Ald-rich, with a purity of 99.999%, stored in N2, and never

unsealed prior to its use. The other was a powdered iron, produced by Riedel-deHaen, with a purity of 99%, made by reduction, and unsealed for a few months under proper sealing and storage and showing no oxidation. The research results showed that the Riedel-deHaen powdered Fe0 _{has almost no promotional e€ect on the}

degradation of TCE after nearly 250 h of reaction, as shown in Fig. 1. On the other hand, the Aldrich granular Fe0_{has a signi®cant e€ect. Under the system used in our}

research, the degradation reaction of TCE indicated a half-life of 239 h, about 10 days.

Although the magnitude of the speci®c surface area of Fe0 _{is proportional to the degradation rate of TCE,}

the purity and the freshness of Fe0_{a€ect whether or not}

the degradation of TCE proceeds. In our past study, the Riedel-deHaen powdered Fe0 _{that was used in this}

re-search had been used in experiments on the reduction of CCl4 and CHCl3 (Cheng and Wu, 1998). During the

initial stage of unsealing, it has a conspicuous promo-tional e€ect on the degradation of CCl4 and CHCl3.

Meanwhile, the powdered Fe0 _{(325 mesh, purity 97%,}

hydrogen reduced, Cat. No. 20930-9) was obtained from Aldrich was used in the same tests; after more than 100 h of test, there were no indication of any promotional

Fig. 1. TCE degradation by using Fe0 _{and Ni}0_{. Conditions:} batch tests were performed by using 3 g of Fe0_{or Ni}0_{in serum} vials (15 ml by volume) in contact with 3 mg TCE lÿ1_solution at 25o_{C, and mixed with the shaker at 130 rpm.}

e€ect by the Aldrich Fe0_{on the degradation of CCl4. Su}

and Puls (1999) also used the same Fe0 _{in the}

degrada-tion tests of TCE. Their results indicated that the reac-tion rate thereof can di€er from that of Fe0 _{from other}

manufacturers by up to three orders of magnitude. Matheson and Tratnyek (1994) proposed that sub-jecting Fe0 _{to an acid washing of HCl aqueous prior to}

reaction can remove the oxide ®lm on the surface of iron, thereby increasing the available reactive sites and increasing the rate of the degradation. However, our research indicated that iron washed by aqueous HCl was oxidized into brown oxide at a high speed upon coming into contact with air during the washing, rinsing and drying processes. The acid washing of aqueous HCl in this way not only causes many ®ne Fe0 _{particles to be}

lost, but also accelerated the oxidation rate of Fe0_{. The}

research on the HCl pretreatment e€ects done by Su and Puls (1999) also indicated that the acid pretreatment processes may have generated more non-reactive sites relative to the reactive sites, and caused a decrease in the reaction rate constant. Therefore, acid prewashing may not be an e€ective and convenient method for improving the drawbacks of Fe0_{unless the contact of iron surfaces}

with air can be completely avoided. 3.2. TCE degradation by Zn0

With electrochemical theory, not only can the de-gradation mechanism proposed by Matheson and Tratnyek be explained, but also Zn0 _{can be considered}

to be the best metal in reduction of chlorinated organic compounds. For the ®rst pathway in which the metal is directly used for reduction of the chlorinated organic compounds (Matheson and Tratnyek, 1994; Vogel et al., 1987; Criddle and McCarty, 1991; Gold et al., 1997; Roberts et al., 1996), the standard reduction potential of iron is ÿ0:44 V, and the standard reduction potential of zinc is ÿ0:763 V. Conspicuously, zinc more easily re-leases an electron to reduce the chlorinated organic compounds than iron does. The second pathway can be explained in greater detail by using the electrochemical theory. Metals with a higher hydrogen voltage are less liable to form hydrogen gas in the system. Therefore, most hydrogen exists in the atomic state, thereby gen-erating a strong reduction potential. In the second pathway, H2designates hydrogen in its atomic state, and

a metal with a high hydrogen overvoltage shall be the best selection of a catalyst. Zn0 _{has a hydrogen}

over-voltage of 0.7 V, which is a metal with the second highest hydrogen overvoltage, second only to mercury. For the above two reaction pathways, Zn0_{not only can be used}

as a strong reducing agent, but also as a good catalyst. This research used Zn0 _{from di€erent sources, with}

di€erent particle sizes and di€erent purities, to carry out the degradation tests on TCE, in order to study the feasibility of using Zn0 _{for remediation of sites}

con-taminated by chlorinated organic compounds. The re-search results indicated that the degradation reaction of Zn0 _{to TCE approximates to a ®rst-order degradation}

reaction model …R2_{> 0:99†. The reaction rate thereof}

can be shown by the following equation: Ct

C0ˆ e

ÿkt_{; …1†}

where C0 is the initial TCE concentration (mg/l), Ct the

TCE concentration (mg/l) at a reaction time t (h), and k is the degradation rate constant (hÿ1_{). The research}

re-sults are shown in Fig. 2. The Hanawa powdered Zn0

has a degradation rate constant, k, of 0.0278 (hÿ1_{) and a}

half-life, t1=2, of 26.8 h. The Aldrich granular Zn0has a

degradation rate constant, k, of 0.013 (hÿ1_{) and a}

half-life of 56.8 h. The Acros granular Zn0 _{forms a}

degra-dation reaction to TCE less conspicuous than that of the previous two Zn0_{. Among the three kinds of Zn}0_{, the}

Aldrich granular Zn0 _{reagent has been unsealed for}

more than one year, and yet the degradation rate thereof to TCE was quite fast in comparison with other Zn0_.

Therefore, the activity of Zn0 _{lasted longer and the}

de-creasing rate of the surface activity was slower than that of iron. In other words, the storage method of Zn0 _is

easier than Fe0 _{when used in a remediation technique.}

Table 1 lists the half-lives for TCE degradation by Fe0_,

Zn0_{, Ni}0 _{and bimetallics.}

In the degradation tests of Zn0_{to TCE, this research}

also investigated the in¯uence of the amount of Zn0

added on the reaction rate. Three grams and 5 g of Aldrich granular Zn0_{, respectively, were separately}

added into 15 ml of aqueous TCE. The results indicated that the run with 3 g of Zn0 _{had a degradation rate}

constant, k, of 0.013 (hÿ1_{) and a half-life of 56.8 h; while}

the run with 5 g of Zn0_{had a degradation rate constant,}

k, of 0.0255 (hÿ1_{) and a half-life of 35.7 h. The}

rela-Fig. 2. TCE degradation by using Zn0_{. Conditions: batch tests} were performed by using di€erent mass and sources of Zn0_in serum vials (15 ml by volume) in contact with 3 mg TCE lÿ1 solution at 25o_{C, 130 rpm. Curve Zn(A): Acros, granular, 3 g.} Curve Zn(B) and Zn(C): Aldrich, granular. Zn(B): 3 g; Zn(C): 5 g. Curve Zn(D): Hanawa, powdered, 3 g.

tionship between the degradation rate of TCE and the amount of addition of Zn0_{, within the scope of this}

study, seemed to be in a proportional relationship. The comparison results in the system of this research indicated that the degradating rate of Zn0 _{to TCE was}

much faster than that of Fe0 _{under the same reaction}

conditions. There can be a di€erence of nearly 10 times between TCE half-lives. The research results also indi-cated that the method of using Zn0 _{to decompose TCE}

was quite an e€ective method. In the future, studies can be carried out to further investigate the degradation process of Zn0 _{on chlorinated organic compounds in}

order to evaluate the feasibility of using it to replace Fe0_.

4. The promotional e€ects of bimetallics on TCE degra-dation

4.1. The promotional e€ects of Pd0 _{and Ni}0_on

degrada-tion of TCE by Fe0

This research ®rst investigated the e€ects of com-bining Fe0 _{with a second metal of Ni}0 _{or Pd}0 _{on the}

reaction rate of decomposing TCE. The research results indicated that Ni0 _{and Pd}0 _{all have a rather strong}

promotional e€ect on the degradation reaction of TCE

by Fe0_{. The degradation reaction of TCE by using}

Riedel-deHaen powdered iron singly was not very sig-ni®cant. However, the TCE conspicuously decomposed in a linear attenuation model when Aldrich powdered Ni0 _{or Acros powdered Pd}0 _{was added to the reaction}

system. The degradation rate of TCE by combining Riedel-deHaen powdered Fe0_{with 3 g of Ni}0_{had a}

half-life of 14.3 h; while by combining with 50 mg of Pd0_the

half-life was 32.8 h. The half-life of TCE was 239 h for Aldrich granular Fe0 _{alone; it was 3:9±5 h for}

combi-nation of the Aldrich granular Fe0 _{with Ni}0_{, and 1.3 h}

for the combination of the Aldrich granular Fe0 _with

Pd0_{. Figs. 3 and 4 show the variation of TCE}

concen-tration to the reaction time in a degradation reaction system of TCE where Fe0_{in conjunction with Ni}0_{or Pd}0

was used. The degradation process of granular iron in conjunction with Ni0_{approximates a linear slow}

atten-uation model at the initial stage. After a short period of reaction, however, the degradation reaction model

Fig. 3. TCE degradation by using powdered Fe0_combination with Pd0 _{and Ni}0_{. Conditions: batch tests were performed in} serum vials (15 ml by volume) in contact with 3 mg TCE lÿ1_at 25o_{C, 130 rpm. Curve Fe(P)/Pd used 3 g of Fe}0_{and 50 mg of} Pd0_{. Curve Fe(P)/Ni used 3 g of Fe}0_{and 3 g of Ni}0_.

Fig. 4. TCE degradation by using granular Fe0 _combination with Pd0 _{and Ni}0_{. Conditions: batch tests were performed in} serum vials (15 ml by volume) in contact with 3 mg TCE lÿ1_at 25o_{C, 130 rpm. Curve Fe(G)/Pd used 3 g of Fe}0_{and 50 mg of} Pd0_{. Curve Fe(G)/Ni used 3 g of Fe}0_{and 3 g of Ni}0_. Table 1

The half-lives (t1=2, for 3 mg TCE lÿ1, 15 ml) for TCE degra-dation by Fe0_{, Zn}0_{, Ni}0_{and bimetallics}

Materials t1=2(h) Riedel powdered Fe0_{(3 g) 2589} Aldrich granular Fe0_{(3 g) 239} Hanawa powdered Zn0_{(3 g) 26.8} Acros granular Zn0_{(3 g) 716} Aldrich granular Zn0_{(3 g) 56.8} Aldrich granular Zn0_{(5 g) 35.7} Aldrich powdered Ni0_{(3 g) 1226}

Riedel powdered Fe0_{(3 g)/Aldrich powdered}

Ni0_{(3 g)} 14.3

Aldrich granular Fe0_{(3 g)/Aldrich powdered}

Ni0_{(3 g)} 3.9±5.0

Riedel powdered Fe0_{(3 g)/Acros powdered}

Pd0_{(50 mg)} 32.8

Aldrich granular Fe0_{(3 g)/Acros powdered}

Pd0_{(50 mg)} 1.3

Riedel powdered Fe0_{(3 g)/Aldrich granular}

Zn0_{(3 g)} 84.1

Hanawa powdered Zn0_{(3 g)/Aldrich}

pow-dered Ni0_{(3 g)} 0.86

Aldrich granular Zn0_{(3 g)/Aldrich powdered}

Ni0_{(3 g)} 1.69

Acros granular Zn0_{(3 g)/Aldrich powdered}

Ni0_{(3 g)} 0.98

Acros granular Zn0_{(3 g)/Acros powdered}

approximates a logarithmic rapid attenuation model. It can be assumed that the reaction rate at the initial stage was predominantly controlled by the transport mecha-nism (Burris et al., 1995; Scherer et al., 1997).

The addition of Pd0 _{or Ni}0 _{does indeed have a}

sig-ni®cant promotional e€ect on the degradation of TCE by Fe0_{. In particular, the addition of Pd}0 _{or Ni}0 _{has a}

reviving e€ect on iron that has lost the surface activities thereof.

4.2. The promotional e€ects of Pd0 _{and Ni}0_on

degrada-tion of TCE by Zn0

This research also tried to combine Zn0_{with a second}

metal of Ni0 _{or Pd}0_{in the degradation reaction of TCE}

in order to understand the promotional e€ect of Ni0 _or

Pd0 _{on the degradation of TCE by Zn}0_{. The research}

results indicated that the addition of Ni0 _{or Pd}0 _rapidly

increased the degradation reaction rate of TCE by Zn0_.

When Ni0_{was added, the half-life of the reaction system}

could be reduced to about 3% of that of the reaction system where Zn0 _{was used singly. The half-life by the}

Hanawa powdered Zn0_{was 26.8 h, which fell to 0.86 h}

after combining with Ni0_{. The half-life by the Aldrich}

granular Zn0 _{was 56.8 h, which fell to 1.69 h after}

combining with Ni0_{. The e€ects were even more}

signif-icant for the Acros granular Zn0 _{that had a weaker}

re-activity. When the Acros Zn0_{was used singly, the}

non-decomposed TCE concentration remained larger than 80% after a reaction time of 237 h. The half-life fell to less than 1 h when Ni0 _{was added. The reaction rate}

increased even further when Pd0_{was added. The half-life}

of a degradation reaction where the Aldrich granular Zn0_{was used in conjunction with Pd}0_{was 0.46 h. Fig. 5}

shows the degradation reaction curve of TCE when Zn0

was used in conjunction with Ni0_{, and Fig. 6 for Pd}0_.

Fig. 5 clearly indicated that the degradation reaction of

TCE exhibited a stagnation phenomenon after the re-action at the initial stage of the rere-action, then rapidly attenuated exponentially. It was inferred that the reac-tion rate was predominantly controlled by the transport mechanism during the initial stage of the reaction (Burris et al., 1995; Scherer et al., 1997).

The bimetallics formed by Zn0_{or Fe}0_{in conjunction}

with Ni0 _{or Pd}0 _{did indeed have a tremendous}

promo-tional e€ect on the degradation reactions of TCE. Moreover, another characteristic of the bimetallics to the degradation reaction of TCE was that from our analysis result there was almost no formation of any chlorinated organic intermediate during the entire re-action process.

5. Conclusions

The results of this research indicated that factors such as the source characteristics, the purity, etc. of Fe0_,

had a strong in¯uence on the degradation rate of TCE. For Fe0 _{of di€erent sources, the reaction rates thereof}

could vary by up to three orders of magnitude. Fur-thermore, the storage of Fe0 _{was another troublesome}

factor. Therefore, care must be taken in selecting zero-valent iron in the remediation technique.

In both electrochemical theory and the ®ndings in the actual experiments, Zn0 _{appeared far more suitable for}

the degradation of the chlorinated organic compounds than Fe0 _{did. Zn}0 _{had a faster degradating rate to the}

chlorinated organic compounds. The half-life of TCE with Zn0 _{was only one-tenth to that with Fe}0_{. The}

storage of Zn0 _{was also easier than that of Fe}0_{. Even}

though a long period had elapsed since it was unsealed, a strong reactivity was still retained. Moreover, Zn0_{is an}

indispensable trace element required by the human body. The tolerable concentration thereof in the

drink-Fig. 5. TCE degradation by using di€erent Zn0 _combination with Ni0_{. Conditions: batch tests were performed by using 3 g} of Zn0 _{and 3 g of Ni}0 _{in serum vials (15 ml by volume) in} contact with 3 mg TCE lÿ1_{at 25}o_{C, 130 rpm. Zn(A): Hanawa} powdered Zn, Zn(B): Aldrich granular Zn, Zn(C): Acros granular Zn.

Fig. 6. TCE degradation by using Zn0_{combination with Pd}0_. Conditions: batch tests were performed by using 3 g of Zn0_and 50 mg of Pd0_{in serum vials (15 ml by volume) in contact with} 3 mg TCE lÿ1 _{at 25}o_{C, 130 rpm. Zn(G) was obtained from} Aldrich granular Zn.

ing water is rather high (5 mg/l). Therefore, the evalu-ation of the use of Zn0_{replacing Fe}0_{in the techniques of}

remediation of groundwater contaminated by chlori-nated organic compounds deserves further study.

The combinations of Zn0 _{or Fe}0 _{with Ni}0_{or Pd}0 _did

indeed have strong promotional e€ects on the degrada-tion reacdegrada-tions of TCE. The half-life of Fe0_/Ni0_{could be}

reduced to 3.9±5 h; the half-life of Fe0_/Pd0 _{could be}

re-duced to 1.3 h, the half-life of Zn0_/Ni0_{could be reduced}

to 0.86 h, and the half-life of Zn0_/Pd0_{could be reduced}

to 0.46 h. The other characteristics of the bimetallics were that Fe0 _{with inactive surfaces could be e€ectively}

revived and have an e€ective degradation on TCE. The application of the bimetallics technique enabled a sub-stantially complete dechlorination reaction that was generally free of the formation of any chlorinated or-ganic intermediates.

Although the combination use of Pd0 _{gave an}

opti-mum result, Pd0_{is a precious metal and its use seems not}

feasible in economic terms. Ni0 _{would be a more}

ap-propriate choice. The research results indicated that the combination of Zn0 _{and Ni}0 _{had a promotional e€ect}

nearly the same as that of the combination of Zn0 _and

Pd0_{. It should be possible for the promotional e€ects}

thereof on the degradation rate of the chlorinated or-ganic compounds to be raised further if further studies can be carried out on the combination ratios, the com-bination forms, and the control of reaction conditions. Furthermore, there is no speci®c regulation on the tol-erable concentration of nickel in the standard of water qualities. The feasibility of this technique can be further evaluated by exploring other areas including the residual concentration of nickel in the aqueous solution after reaction, the in¯uence of nickel on the human body, and the tolerable concentration of nickel in the drinking water.

Acknowledgements

The authors gratefully acknowledge the ®nancial support provided by the National Science Council, ROC (Contract No. NSC 88-2218-E-002-035) for this research work.

References

Agrawal, A., Tratnyek, P.G., 1996. Reduction of nitro aromatic compounds by zero-valent iron metal. Environ. Sci. Tech-nol. 30, 153±160.

Appleton, E.L., 1996. A nickel-iron wall against contaminated groundwater. Environ. Sci. Technol. 30, 536±539. Ballapragada, B.S., Stensel, H.D., Puhakka, J.A., Ferguson,

J.F., 1997. E€ect of hydrogen on reductive dechlorination of chlorinated ethenes. Environ. Sci. Technol. 31, 1728±1734.

Burris, D.R., Campbell, T.J., Manoranjan, V.S., 1995. Sorption of trichloroethylene and tetrachloroethylene in a batch reactive metallic iron-water system. Environ. Sci. Technol. 29, 2850±2855.

Cheng, S.F., Wu, S.C., 1998. Enhancing chlorinated methanes degradation by modifying the Fe0 reduction system. In: Physical, Chemical, and Thermal Technologies, vol. C1-5. Battelle Press, USA, pp. 299±304.

Criddle, C.S., McCarty, P.L., 1991. Electrolytic model system for reductive dehalogenation in aqueous environments. Environ. Sci. Technol. 25, 973±978.

Gavaskar, A.R., Sass, B.M., Drescher, E., Cumming, L., Giammar, D., Gupta, N., 1998. Enhancing the reactivity of permeable barrier media. In: Designing and Applying Treatment Technologies, vol. C1-6. Battelle Press, USA, pp. 91±96.

Gillham, R.W., Stephanie, F.H., 1994. Enhanced degradation of halogenated aliphatics by zero-valent iron. Ground Water 32, 958±967.

Gold, G., Angst, W., Holliger, C., Schwarzenbach, R.P., 1997. Corrinoid-mediated reduction of tetrachloroethene trichlo-roethene and trichloro¯uotrichlo-roethene in homogeneous aque-ous solution: reaction kinetics and reaction mechanisms. Environ. Sci. Technol. 31, 253±260.

Matheson, L.J., Tratnyek, P.G., 1994. Reduction dehalogen-ation of chlorinated methanes by iron metal. Environ. Sci. Technol. 28, 2045±2053.

Muftikian, R., Fernando, Q., Korte, N., 1995. A method for the rapid dechlorination of low molecular weight chlorinat-ed hydrocarbons in water. Wat. Res. 29, 2434±2439. O'Hannesin, S.F., Gillham, R.W., 1998. Long-term

perfor-mance of an in situ ``iron wall'' for remediation of VOCs. Ground Water 36, 164±170.

Orth, W.S., Gillham, R.W., 1996. Dechlorination of trichloro-ethene in aqueous solution using Fe…0†

. Environ. Sci. Technol. 30, 66±71.

Roberts, A.L., Totten, L.A., Arnold, W.A., Burris, D.R., Campbell, T.J., 1996. Reductive elimination of chlorinated ethylenes by zero-valent metals. Environ. Sci. Technol. 30, 2654±2659.

Ruiz, N.E., Reinhart, D.R., Clausen, C.A., Geiger, C.L., Lau, N., 1998. Enhanced zero-valent iron degradation of chlo-rinated solvents using ultrasonic energy. In: Designing and Applying Treatment Technologies, vol. C1-6. Battelle Press, USA, pp. 71±76.

Scherer, M.M., Westall, J.C., Ziomek-Moroz, M., Tratnyek, P.G., 1997. Kinetics of carbon tetrachloride reduction at an oxide-free iron electrode. Environ. Sci. Technol. 31, 2385±2391.

Siantar, D.P., Schreier, C.G., Chou, C.S., Reinhard, M., 1996. Treatment of 1,2-dibromo-3-chloropropane and nitrate-contaminated water with zero-valent iron or hydrogen/ palladium catalysis. Wat. Res. 30, 2315±2322.

Smyth, D., Cherry, J., Jowett, R., 1995. Treat groundwater in place ± in situ funnel-and-gate system corrals water for treatment. Soil Groundwater Clean up, December, 36±43.

Su, C., Puls, R.W., 1999. Kinetics of trichloroethene reduction by zerovalent iron and tin: pretreatment e€ect apparent activation energy, and intermediate products. Environ. Sci. Technol. 33, 163±168.

Vogel, M.T., Criddle, C.S., McCarty, P.L., 1987. Transforma-tions of halogenated aliphatic compounds. Environ. Sci. Technol. 21, 722±736.

Wang, C.B., Zhang, W.X., 1997. Synthesizing nanoscale iron particles for rapid and complete dechlorination of TCE and PCBs. Environ. Sci. Technol. 31, 2154±2156.

Weber, E.J., 1996. Iron-mediated reductive transformations: investigation of reaction mechanism. Environ. Sci. Technol. 30, 716±719.