Toxicity Evaluation of Fly Ash by Microtox ®

Fly ash samples from a cooling tower were extracted after incinerating plastic solid waste (PSW) and organic liquid waste (OLW) by n-Hexane or dichloromethane/n-Hexane mixtures to evaluate the toxicity. The metal and PCDD/Fs were analyzed by ICP OES and HRGC/HRMS, respectively. The toxicity of the extracted fly ash was evaluated by Microtox. The results showed the environmental risk factor (ERF) of Hg in PSW fly ash was the highest compared to other metals, by more than 60%. Additionally, the acute toxicity tests of the fly ash showed that dichloromethane/n-hexane extracts were all very toxic, except for the PSW-1 obtained through Soxhlet extraction following the column clean-up procedure. The n-Hexane extracts for OLW-1 obtained through Soxhlet extraction following the column clean-up procedure were extremely toxic. There were no significant relationships among the concentrations of the regulated heavy metals (As, Cd, Cr, Cu, Hg, Ni, Pb, Zn), PCDD/Fs concentration and the TU values in the toxicity test. Furthermore, the results of the statistical analysis showed that there were significant differences in the results of the Microtox test with regard to the solvents, solutes and various extraction methods. However, it is remains a complicated process to differentiate between the various compounds in order to produce accurate results with regard to acute toxicity in the fly ash, and thus this issue warrants further investigation.


INTRODUCTION
Incineration is still the main treatment method for wastes in Taiwan.More than 1.2 millions of ash were produced in 2011 and nearly 60% of them will be landfilled as final treatment (EPA, Taiwan).It was unknown how many ashes are classified as hazardous wastes because of the complicated measurements.Many research efforts are focused on the investigation of the ash' physical structure (Ontiveros et al., 1989;Álvarez-Ayuso et al., 2008) and chemical composition (Criado et al., 2004), the reuse of the fly ash (Mangialardi, 2001), or the detection of metals concentration (Karuppiah and Gupta, 1997;Prokop et al., 2003;Kalderis et al., 2008) or PCDD/F concentration in the fly ash (Lin et al., 2011;Chin et al., 2012).Chemical analysis assists in determining the anthropogenic concentration and provides estimates of their distribution (Chiu et al., 2011;Huang et al., 2011).However, the chemical data alone provide no direction as to the potential effects on the environment.
Toxicity bioassay has been widely applied in water (Gutiérrez et al., 2002;Sarmiento et al., 2011), sediment (Jacobs et al., 1993;Niemirycz et al., 2007) or soil (Maxam et al., 2000;Loureiro et al., 2005).Algae (Hörnström, 1990), daphnia (Guilhermino et al., 2000) or fish (Lammer et al., 2009) have been used to process many toxicity tests; however, these tests are time and cost consuming and some test organisms require further culturing.Table 1 compared the advantages of various test organisms, bioassay based on bacteria (such as Microtox ® ) implied a simple procedure, short testing time and high convenience.Additionally, Microtox ® is distributed in many countries and is recommended by standards in America, Germany and Poland (standard methods, 1995, ISO/DIS 11348, DIN 38412) and performed as a rapid, economical monitoring tool for toxicity of environmental contaminants.It was applied in metal plating wastewater (Choi and Meier, 2001) (Lin and Chao, 2002).However, limited studies were found to discuss the biotoxicity (Chakraborty and Mukherjee, 2009;Chou et al., 2009) for the fly ash.In this study, Microtox ® was applied to evaluate the acute toxicity of extracted fly ash and the feasibility to be a fastprescreening tool.

MATERIALS AND METHODS
Fly ash samples were taken from a laboratory waste incinerator in the southern part of Taiwan.The incinerator is equipped with the followings: first cooling tower, secondary cooling tower, bag filter and scrubber which all serve as air pollution control devices.The fly ash from the first cooling tower and secondary cooling tower were collected by two different feeding wastes: Plastic solid waste (PSW) and organic liquid waste (OLW) following Taiwan's National Institute of Environmental Analysis (NIEA) method R118.02B.The ash sample was simplified as PSW-1 to represent fly ash from the first cooling tower by feeding plastic solid waste; PSW-2 to represent fly ash from the secondary cooling tower by feeding plastic solid waste; OLW-1 to represent fly ash from the first cooling tower by feeding organic liquid waste and OLW-2 to represent fly ash from the secondary cooling tower by feeding organic liquid waste.
Two kinds of solvents: n-hexane (Hx) and dichloromethane/n-hexane (DCM-Hx) via two kinds of extraction methods: sonication and Soxhlet were chosen separately.Two grams of fly ash were extracted with 300 mL solvents.These extracted samples were divided into two parts: one was evaporated to near dryness and dissolved in dimethyl sulfoxide (DMSO).The other one was also evaporated to near dryness and then transferred to the CAPE-coupled carbon-acid silica column for cleanup.The cleanup procedure using the CAPE-coupled carbon-acid silica column was the one previously described in detail (Chen et al., 2007;Lee et al., 2009).

Metal Analysis
Inductively coupled plasmaoptical emission spectrometry (ICP-OES, VISTA-MPX, Varian) was used to determine the presence of Ag, As, Ba, Cd, Cr, Cu, Hg, Mn, Na, Ni, Pb, Se, and Zn following U.S. EPA Method 200.7.All the dehydrated solid specimens were pretreated with microwaveassisted acid digestion, following NIEA method R317.10C (equivalent to U.S. EPA Method 3015A).The specimens were digested using a 400-W microwave MARS/MARS Xpress CEM microwave, with an 800 psi limit at 200°C for 15 minutes.Additionally, the toxicity characteristic leaching procedure (TCLP) was performed to examine the leachability of the ash samples following NIEA method R201.14C (equivalent to U.S. EPA SW846 Method 1311).

PCDD/F Analysis
Standard solutions of PCDD/Fs (1613-LCS, labeled compounds stock solution; 1613-ISS, internal standard spiking solution; 1613-CSS, cleanup standard spiking solution; 1613CVS, EPA Method 1613 calibration and verification solution) were purchased from Wellington Laboratories (Ontario, Canada).Silica gel (100-200 mesh) was obtained from Fisher (Leicestershire, England).For the HRGC/HRMS method, all samples were added with the different PCDD/F internal standards before extraction and the labeled cleanup standards for PCDD/F analysis were added before the CAPE-coupled carbon-acid silica column cleanup (Mi et al., 2012).The extract was analyzed using a HRGC/HRMS (HP6890/JEOL JMS-700) equipped with a DB-5MS 60 m column.The quality of QA/QC met the criteria of Taiwanese EPA Method.

Microtox Analysis
For microtox tests, extracts were placed in 10-ml graduated concentrator tubes, and solvents was removed from the extracts in a 25°C temperature water bath by introducing a constant stream of nitrogen over the solvent surface.Once the volume was reduced to near dryness and exchanged by 1 mL DMSO (HPLC-grade).The DMSO extracts were placed in borosilicate glass vials and stored at 4°C for microtox analysis.
Microtox Model 500 analyzer (SDIx, USA) was used for analysis and all detailed procedures were followed in the Microtox Users Manual.Freeze-dried luminescent bacteria, Vibrio fischeri were reconstituted and exposed in duplicate to a series of four diluted DMSO extracts, osmotically adjusted to a salt content of 2% and using one saline water control.The solvent vehicle, DMSO, at volumes not exceeding 2.5% of the total assay volume did not significantly affect light emission (Kahru et al., 1996).The resulting decrease in bioluminescence was measured after 5 and 15 minutes at a constant temperature of 15°C.Fifteen minute data are reported in this study.All Microtox data were recorded and analyzed by on-line software, and results are expressed as the Effective Concentration 50% (EC 50 ) in percent of extract per 2 g ash sample.
Toxicity Unit (TU) was recommended and identified as follows: TU = (1/EC 50 ) × 100% (1) TU is unitless.The high TU value indicates high toxicity.However, TU is relative toxicity classified into four categories as Table 2.
To ensure reliability of the Microtox method and reagents, a toxicity test using an aqueous solution of phenol (100 mg/L) was done to confirm each day prior to beginning sample tests.This check was then compared with the Microtox quality assurance product data accompanying the reagent.Procedural blanks were also prepared from cleaned filters and cartridges extracted as described above and were tested to determine if any toxicity was being contributed by the residual extracts and glassware.In this study no toxicity was detected in these blanks.

Statistical Analysis
Analysis of variance conducting by SPSS 11.0 was applied in this study to investigate the effect of experimental parameters on the analysis results.The test was performed to compare the variation of TU values in various solvent extracts, extraction methods and clean-up over four various ash samples.

Environmental Risk Factor
To evaluate the environmental risk factor (ERF) for the regulated metals in the ash, ERF (Sarmiento et al., 2011) was defined as: Table 2. Category of TU (Kahru et al., 2000).

TU Toxicity < 1
Non toxic 1-10 Toxic 10-100 Very Toxic > 100 Extremely Toxic C n is the toxic element concentration; C SQV is the highest concentration of the studied element non-associated with biological effects defined by Sarmiento et al. (2011).It's hard to define the highest concentration of the studied element not associated with biological effects in ash samples, so the regulated concentration in soil was adopted in this study to represent C SQV .Fig. 1 showed the environmental risk factors for the ash.Incinerating PSW led to high Hg ERF warning about a potential adverse biological effect associated with Hg and warrants further investigation in the future.

Solvent Blank
To understand the effects of various solvents on TU, a solvent blank was tested.Hx, DCM-Hx and DMSO were tested following the procedure of extracted samples to compare the TU by individual solvent.Table 3 showed no toxic for DMSO, toxic for n-Hexane (Hx) and very toxic for dichloromethane/n-Hexane (DCM-Hx).Accordingly, all extracts were transferred into DMSO to avoid the interference of extracting the solvent's toxicity.Furthermore, the water extracted samples were also tested to simulate the rainfall condition, however, no any toxic result was found.

Toxicity Test
Tables 4 and 5 summarizes the acute toxicity of the ash samples based on Table 1.To understand the effects of various solvent extracts on TU, n-Hexane and dichloromethane/n-Hexane solvents were chosen to extract ash samples.These two kinds of solvents were always used to study the organic compounds in samples.Table 4 showed that the toxicity of  concentrated samples (without cleanup) was evidently higher than that of the cleanup samples by sonication extraction based on n-Hexane.However, the result for soxhlet extraction is different.The toxicity of cleanup samples was evidently higher than that of concentrated samples by the soxhlet extraction based on n-Hexane solvent.
The result of Table 5 showed the toxicity for all samples were very toxic, except PSW-1 for cleanup samples by soxhlet extraction based on dichloromethane/n-Hexane solvent.The toxicity of these four dichloromethane/n-Hexane extracted samples showed no obvious difference.To further investigate the results, the regulated metal concentration and PCDD/Fs was compared with the results of toxicity.
Fig. 2 showed the acute toxicity mean value (TU%) for the concentrated samples and the corresponding PCDD/Fs concentration.Only n-Hexane sonicated extraction sample revealed to have a similar trend with PCDD/Fs concentration.
Fig. 3 showed the acute toxicity mean value (TU%) for the clean-up samples and the corresponding PCDD/Fs concentration.No similar trend was found in Fig. 3. Nor in the metal concentration and acute toxicity, where no obvious correlation was found.
The analysis of variance (ANOVA) was applied to interpret the effect of the solvent, extraction method and cleanup process on acute toxicity.The result was reported in Table 6.The p-values for the three factors: solvent, solute and extraction were all smaller than 0.01.This implies that the different levels of the three factors affected the observed amount of acute toxicity.The mean effect of clean-up on the observed amount of acute toxicity is not statistically different (p-value > 0.3) whether the sample is clean-up or not.It is also seen from the    interactions have an effect on the acute toxicity.Finally, all three-factor interactions and the four-factor interactions are also valid from the p-values reported in the table.It would be warranted as a future research topic to understand why interactions affect the acute toxicity.

Table 3 .
The toxicity of solvent blank.

Table 4 .
table that, except (1) cleanup *extraction and (2) solute *extraction, almost all two-factor Acute toxicity of ash sample extracted by n-Hexane.

Table 5 .
Acute toxicity of ash sample extracted by dichloromethane/n-Hexane.