9. Extractability and Chemical Forms of Radioactive Cesium in Designated Wastes Investigated in an On-Site Test

All reagents used in our study were of analytical grade unless specified otherwise.

We used MSW fly ash samples unaffected by the F1 accident to clarify the effect of extracting procedures on the extraction efficiency of various elements (including non-radioactive cesium) from the sample. In particular, we compared the use of 0.1 and 0.5 M hot oxalic acid extraction procedures on fly ash from a melting furnace treating sewage sludge from Western Japan. Since the sample was not contaminated with rad-Cs from the F1 accident, the extraction could be conducted in the university laboratory, and the concentration of various elements in the extracts could be determined using ICP-MS (model 7700 series, Agilent Technologies).

The three different extraction procedures were tested on the fly ash sample from the melting furnace. (1) One procedure was vacuum extraction (or continuous extraction), in which a 5 g fly ash sample placed on an i.d. 47 mm membrane filter holder (fitted with pore size 0.45 μm PTFE filter) was continuously eluted with hot oxalic acid (50 mL solution/40 min) supplied via a peristaltic pump for total 120 min. The filter holder was connected to a vacuum flask, and vacuum was applied periodically so that the extracting reagent (0.1 or 0.5 M hot oxalic acid) was passed through the filter while the fly ash sample was always soaked in the oxalic acid. The percolate was collected every 40 min, and was analyzed by ICP-MS. By using the vacuum extraction method, re-adsorption of once extracted components to the waste material, as well as temporary oversaturation of extracted metals in the solution phase, could be minimized. This procedure was also practicable in the on-site test. (2) Conventional batch extraction (2.5 g of fly ash in 75 mL of 0.1 or 0.5 M hot oxalic acid), with 200 rpm of shaking in a 90 °C hot bath, was conducted. (3) Also tested was a batch exchange extraction method, an intermediate method between continuous extraction and batch extraction. With this method, 2.5 g of fly ash was extracted with 25 mL of 0.1 M or 0.5 M hot oxalic acid with continuous shaking (200 rpm) in a hot bath for 1 h, centrifuged for solid–liquid separation (6,400 × g, 4 °C, 15 min), and the supernatant was removed and saved for the subsequent analysis by ICP-MS. Then 25 mL of 0.1 or 0.5 M hot oxalic acid was added to the solid waste again and the same extraction procedure repeated twice. The batch exchange method turned out to be laborious and was not practicable in the on-site test compared to the other two methods.

With hot oxalic acid extraction, a precipitate was sometimes formed in the extract after it was cooled to room temperature. In such cases, the extract was centrifuged, the precipitate was dissolved in dilute HNO₃ and the supernatant and the dissolved precipitate were analyzed separately by ICP-MS.

To determine the total element concentration in the fly ash, 100 mg of the fly ash sample, pulverized to pass a 150 mesh sieve, was digested by mineral acids in a pressurized Teflon vessel placed in a microwave oven model MARS5000 (CEM Co., US). The acids used were 5 mL of concentrated 68 % HNO₃, 1 mL of 70 % HClO₄ and 4 mL of 38 % HF. All the acids used were AA-100 grade from Tama Chemicals Corporation (Japan). The sample was acid decomposed under 70 psi pressure for 10 min. after 20 min. of ramp time. After the decomposition, 6 mL of 4.1 w/v percent boric acid was added to the digested sample to dissolve the fluoride precipitated with Ca, etc. as well as to neutralize HF left in the solution. With this procedure, a clear solution with no visible suspended particles was obtained. The solution was diluted and analyzed by ICP-MS. The results obtained were used to calculate the percentages of the elements extracted by 0.1 or 0.5 M oxalic acid from the fly ash per total amount of the elements in the original fly ash.

On-site analysis of rad-Cs was conducted using a portable Ge detector from Princeton Gamma Technology (P-type, relative efficiency 20 %), and a Ge detector from ORTEC (model Trans-Spec, p-type, relative efficiency 40 %), in 2013 and 2014, respectively. Ge detectors were pre-calibrated for absolute full-energy peak efficiency for the solid samples packed in commercially available U-8 vessels (i.d. 50 mm, height 68 mm) as well as for a 30 mL volume of aqueous solution in a 50 mL polypropylene bottle, and were used to quantitate rad-Cs in waste extracts and waste materials. A NaI detector (model AN-OSP-NAI, Hitachi-Aloka-Medical, pre-calibrated for determination of rad-Cs) was also used to quantitate Cs-134 and Cs-137 in waste materials.

Concentrations of metals in the extract obtained at the on-site test were determined using a portable voltammetry instrument (model PDV6000plus from Modern Water, UK) coupled with an autosampler. In 2014, Fe and Zn were determined. Determination of Fe was conducted using a glassy carbon working electrode, acetate buffer as a supporting electrolyte, deposition potential −1,600 mV and stripping step potential range of −1,100 to 50 mV. Determination of Zn was conducted at a mercury-film-coated glassy carbon working electrode, acetate buffer as a supporting electrolyte, deposition potential −1,300 mV and stripping step potential range of −1,200 to 50 mV.

The rad-Cs-contaminated fly ash samples from sewage treatment in the area affected by the F1 accident were extracted and analyzed for rad-Cs in the plant where they were produced (the on-site test). The information for the waste materials used for the on-site test is summarized in Table 9.2, together with extracting reagents tested on the samples. We used a distilled-water extraction [23] and a hot oxalic acid (0.1 M) extraction [24] that had previously been applied to the removal of rad-Cs from contaminated MSW. Also tested were 5 M HCl extraction and 0.5 M hot oxalic acid extraction in an attempt to achieve higher leaching of rad-Cs for some samples.

The extraction of rad-Cs containing waste materials was conducted using the vacuum extraction (continuous extraction) procedure in the on-site test conducted in 2014. This procedure was used because the results of three different extraction methods for uncontaminated waste showed that the batch extraction may be affected by loss of the extracted elements due to re-adsorption and/or precipitation processes as will be discussed in the Sect. 9.7. The extraction in 2013 was conducted by batch extraction.

The treatment of the extract by Fer precipitation technique was conducted as follows: The pH of the waste extract containing rad-Cs was adjusted to the desired values (shown later in Sect. 9.7) using NaOH and HCl. Potassium ferrocyanide (K₄[Fe(CN)₆], K-Fer hereafter) was used as a soluble Fer salt source. Since metal hydroxides (ineffective for Cs removal) rather than metal-Fer complexes could be formed in the solutions when pH values are in neutral to alkaline regions, a slightly acidic pH was used for Fer precipitation. After the addition of K-Fer, Ni-Fer or Fe(III)-Fer precipitate (prepared in a separate bottle) to the extract, the sample was centrifuged at 2,600 × g at 4 °C for 20 min. or stood still until the precipitate was settled. When more than a few liters of the extract were treated, 30 mg/L of cationic polymeric flocculant (polymethacrylic ester, C-303 or C-512 from MT AquaPolymer Inc., Japan) was added to facilitate the settling of fine precipitates.

Extractions of stable (non-radioactive) Ca, Fe, Zn, Ni, As and Cs from fly ash (collected from a melting furnace treating sewage sludge from Western Japan) with 0.1 and 0.5 M hot oxalic acid are compared in Fig. 9.1. Clearly, the percentages of metal extraction differed considerably depending on the concentration of oxalic acid as well as on the extraction procedures (continuous, batch and batch exchange). With the batch extraction procedure, we observed a high extraction percentage at an earlier time (40 min), followed by a decrease in the extraction percentage. This is probably explained by re-adsorption or re-precipitation (or co-precipitation) of once extracted elements. In contrast, the continuous extraction showed a gradual increase in extraction percentages with time. The extraction percentages by continuous extraction at 120 min of extraction time were similar to those obtained by batch extraction at 40 min for non-radioactive Cs and Fe. The batch exchange procedure method showed a gradual increase in the extraction percentages of the elements with time and was more similar to the continuous extraction than simple batch extraction. For both batch and batch exchange procedures, there could have been temporary oversaturation of the extracted elements in the solution after 40 min of extraction with hot oxalic acid. In the case of the batch exchange method, the oversaturated elements, if any, should have been precipitated during the centrifugation to exchange the supernatant. In simple batch extraction, unlike batch exchange extraction, supernatant was not exchanged, and the aliquot of the supernatant collected for analysis should contain the oversaturated element (if any). In any case, the batch exchange procedure was laborious compared to the continuous extraction and simple batch methods. Also, the centrifugation we conducted for solid-liquid separation in the batch exchange procedure every 40 min caused a temporary decrease in the extraction temperature, making it difficult to interpret the obtained results.

The oxalic acid concentration of the extract also affected the extraction efficiency of elements (in Fig. 9.1, only the results of Ca, Fe, Zn, Ni, As and Cs are shown). Briefly, in the case of continuous extraction, 0.5 M hot oxalic acid was more effective than 0.1 M hot oxalic acid in extracting non-radioactive Cs, Fe, Mg, Al, Mn, As, Se, Sr and Ba. On the other hand, for Zn, Ca, Co, Ni and Cu, 0.1 M hot oxalic acid was more effective than 0.5 M hot oxalic acid (in the case of continuous extraction). Oxalic acid dissolves Fe and some other metals preferentially through the ligand-promoted dissolution process [25], and the lower pH of 0.5 M oxalic acid compared to 0.1 M oxalic acid enhances the dissolution of elements in general. At the same time, with 0.5 M oxalic acid, metal oxalates having low solubility can be more easily formed than with 0.l M oxalic acid. We consider that the concentrations of Zn, Ca, Co, Ni and Cu in the 0.5 M oxalic acid extract were solubility controlled by the metal oxalates. Although the solubility values of metal oxalates freshly formed at high temperature (90 °C) may be hard to obtain, solubility determined for pure metal oxalates can be used as reference values. According to the literature [26], oxalates of Zn, Co, Ni and Cu have very low solubility at 20 °C. This information supports our inference that Zn, Co, Ni and Cu concentrations were solubility controlled. Since the concentration of Ca in the extract is high (between 67 and 600 mg/L in 0.5 M oxalic acid extract), it is no wonder that the Ca concentration was controlled by the solubility of calcium oxalate.

Based on the results above, we applied the continuous extraction method at the on-site test in 2014, in order to obtain more scientific estimates of the leaching of rad-Cs from the contaminated fly ash. We noted, however, that the behavior of non-radioactive Cs during the extraction was relatively simple, i.e. Cs was not precipitated with oxalic acid and therefore its extraction was higher in 0.5 M oxalic acid than in 0.1 M oxalic acid, and the highest extraction could be achieved with 40 min of batch extraction. From a practical point of view, by using 0.5 M hot oxalic acid in simple batch extraction mode, the majority of Cs that can be extracted with the reagent by the continuous method will be released from the solid phase, although there is a chance that the behavior of non-radioactive Cs we have investigated in this section and that of rad-Cs in fly ash samples are different, depending on the original chemical form of Cs.

The results are summarized in Table 9.3. The results indicate rad-Cs leaching from melting furnace fly ash (MF1, MF2 and MF3) is high. At least 40 % was extracted with distilled water, and the percentage of the rad-Cs extraction was close to 100 % when 0.5 M hot oxalic acid was used. Notably, the percentages of rad-Cs extraction were not much affected by the pH values of the extract. For example, distilled water extracts of MF3 had pH values around 7 (refer to Table 9.3), and rad-Cs extraction was higher than 60 %. The distilled water extracts of MF2 had lower pH values (around 1) compared to MF3, and the rad-Cs extraction was only 40 %, lower than the distilled water extracts of MF3 with pH value 7. The lower extraction percentages for MF2 are partly attributable to the lower solid-to-liquid ratio (1 g/2.5 mL) of extraction compared to 1 g/75 mL for MF3, implying that a thorough washing with larger volume of liquid leads to higher leaching.

As discussed in the preceding section, stable Cs extracted by distilled water from fly ash of the melting furnace treating sewage sludge from Western Japan was only 5 % of the total Cs contained in the ash. Interestingly, rad-Cs in the fly ash MF1, MF2 and MF3 from the similar melting furnace was more water soluble compared to the stable cesium in the other fly ash of similar origin.

The extractability of rad-Cs was lower for incinerator fly ash (IF1 and IF2) compared to fly ash from the melting furnace (MF1, MF2 and MF3). The ashes collected in different periods, IF1 and IF2, also exhibited quite different leaching of rad-Cs, Fe and Zn. When extracted with distilled water, rad-Cs extraction percentages were only ca. 1 % for IF1, whereas it was 10 % or higher for IF2. The IF2 ash also had higher amount of extractable Fe and Zn compared to IF1. The IF1 ash was obtained when sewage sludge was incinerated together with soil, while IF2 ash was obtained when only sewage sludge was incinerated. The other researchers [9] pointed out that the incineration with soil is likely to inhibit the leaching of rad-Cs from the fly ash, and the results we obtained are in accordance with their observation. As for IF1, we also used 5 M HCl extraction in an effort to facilitate the leaching of rad-Cs, but the percentage of rad-Cs extracted was only 20 % with this method. Kozai et al. [11] reported that leaching of rad-Cs from sewage ash can be enhanced by pulverizing the ash and using hot HCl extraction. Out of concern for safety, we could not conduct the pulverizing of already fine-grained, powder-like incinerator fly ash nor use hot HCl during our on-site test. In contrast, more than 60 % of rad-Cs in IF2 was extracted with 0.5 M hot oxalic acid (the percentage was 10 % for IF1). We cannot rule out the possibility that the continuous extracting procedure we used for IF2 (IF1 was extracted by batch method) was more effective in leaching rad-Cs from the sample. However, as we discussed in the section above, the leaching behavior of Cs is relatively simple because of the high solubility of Cs in water, and it is unlikely that the difference in extraction procedure (continuous or batch) led to a drastic difference in the extraction rate with the same hot oxalic acid. We therefore conclude that the incinerating of sewage sludge and soil together in the fluidized bed incinerator led to the drastic reduction of rad-Cs leaching from fly ash.

The bottom ash IB1, which in fact is the soil incinerated in the fluidized bed incinerator, exhibited a very low leaching of rad-Cs (0 % with distilled water and 3 % with 0.1 M hot oxalic acid). Considering the reduction in leaching of rad-Cs from fly ash, it is obvious that rad-Cs leaching was inhibited in the presence of soil. It should be noted that the soil itself did not contain much rad-Cs before it was incinerated. The rad-Cs contained in the bottom ash originated from contaminated sewage sludge.

Table 9.4 shows results of Fer coprecipitation conducted on extracts of rad-Cs contaminated fly ash. The extract of molten fly ash showed c.a. 93 % removal of rad-Cs when Fe(III)-ferrocyanide or Ni(II) –ferrocyanide was added and the supernatant was analyzed. Removal of rad-Cs, when we tried the in situ formation of insoluble Fer solids by adding Fe(III) salt and K-Fer salt, was only 80 % because the very fine precipitate formed could not be completely removed, even though we added a coagulant. Optimal solid-liquid separation for the in situ Fer solid formation method needs to be pursued further.

As for the extracts of incinerator ash IF2, only one oxalic acid extract sample contained an easily detectable amount of rad-Cs (235 Bq/L). We therefore focused on the removal of rad-Cs from this sample. After some trial and error, we conducted coprecipitation with Ni(II)-ferrocyanide at pH7. The resultant removal of rad-Cs was only 59 %.

To investigate the cause of this low removal of rad-Cs, in the 2014 on-site test, the 0.5 M oxalic acid extracts of fly ashes from the incinerator and melting furnace (both pre-filtered by a pore size 0.45 μm membrane filter) were ultra-filtered using molecular weight (MW) 3,000 Sartorius centrifugal ultrafiltration kit. The extract of fly ash from the melting furnace left no residue on the ultrafilter, indicating that there are probably no colloidal particles in the extract of the melting furnace. On the other hand, the extract of the incinerator ash was retained on the ultra-filter, and by analyzing the rad-Cs content of the retained liquid, c.a. 30 % of rad-Cs in the extract was estimated to be colloidal (MW >3,000).

These results explain why rad-Cs removal from the 0.5 M oxalic acid extract of the incinerator fly ash by Fer coprecipitation was low (58 %). Only rad-Cs cations can be removed by the Fer precipitation technique, and if the rad-Cs in colloidal form in the incinerator fly ash extract had been in cationic form, the removal of rad-Cs would have increased to c.a. 90 %, i.e. the removal that should be attained with the Fer method under normal circumstances. As for the extract of incinerator fly ash, we also observed the low removal of rad-Cs in the on-site test in 2013 and attributed the results to the lack of time to optimize the coprecipitation conditions for the sample [6]. However, based on the recent findings, the low removal of rad-Cs in the 2013 test was also likely to have been the effect of rad-Cs being present in colloidal form. Although the conversion of the colloidal rad-Cs to rad-Cs cations may not be easy, by selecting and using an appropriate coagulant, colloidal rad-Cs can be removed by coagulation and precipitation. The origin of the colloidal rad-Cs in the extract needs to be investigated in the future research.