Sulfated Steelmaking Slags as Se(IV) Adsorbents: Effects of Preparation Conditions on Adsorption Performance

Abstract

Steelmaking slags treated with aqueous H₂SO₄ were characterized as Se(IV) adsorbents, and notable adsorption ability was shown to develop at H₂SO₄ concentrations of 0.18–0.20 mol/L. The enhanced adsorption performance of slags sulfated in the presence of H₂O₂ suggested that their Se(IV) adsorption sites contained Fe(III) and resembled those of schwertmannite. Langmuir analysis of Se(IV) adsorption isotherms was used to determine the corresponding adsorption capacities and adsorption constants. Thus, the obtained results allow one to simultaneously solve the problems of slag utilization and Se removal from wastewater or natural water, contributing to the establishment of a greener society.

1. Introduction

Selenium is an essential trace element that is commonly present in the active center of enzymes and thus plays an important role in the metabolism of living organisms (including humans).^1,2) However, as large intakes of Se may cause peripheral neuropathy,³⁾ the Se content of environmental waters is regulated not to exceed 0.1 mg/L in Japan. Se contained in environmental waters is mainly produced by the leaching of natural Se-containing ores and/or is supplied with Se-contaminated industrial waste.⁴⁾ Therefore, the removal of Se from wastewater and natural waters with Se levels exceeding the stipulated limits is an important task from the viewpoint of emission regulation and environmental protection.⁵⁾

Currently, Se separation and pre-concentration are mainly achieved by solvent extraction and solid-phase extraction. Although solvent extraction is a well-established method offering some sophisticated Se extraction reagents such as bismuthiol-II⁶⁾ and dithiocarbamates,^7,8) it employs harmful (and increasingly strictly regulated) organic solvents and has therefore lost ground to solid-phase extraction, an organic solvent–free technique featuring the benefits of high separation efficiency and operation simplicity. Consequently, numerous organic solid,¹¹⁾ inorganic solids^12,13,14) and organic–inorganic composites^15,16) have been developed as Se adsorbents. Among the inorganic solids, natural ores have been most widely studied as promising adsorbents for harmful element elimination, offering the advantages of high natural abundance and low cost.¹⁷⁾ In particular, schwertmannite is a well-known natural ore capable of adsorbing As and Se.^18,19,20)

In addition to natural ores, artificial inorganic solids have become important potential adsorbents for harmful elements. For example, steel slags, some of the major inorganic solids produced during steelmaking processes, are a promising artificial resource because of their stable production and well-known composition. Moreover, steel slags produced at industrial districts can be reused by other industries possibly located in the same district, which results in decreased waste generation. According to their genesis, steel slags can be categorized into blast furnace slags and steelmaking slags. In Japan, blast furnace slags are completely reused, while complete reuse or recycling of steelmaking slags has not been achieved despite having the potential to be employed as inorganic adsorbents and thus be. At this point, one should note that whereas the adsorption of heavy elements such as As(III, V), Cu(II), Cd(II), Ni(II), Pb(II), and Zn(II) on steelmaking slags has been studied,^21,22,23,24) that of Se remains unexplored.

Previously, we have developed adsorbents for harmful elements by functionalization of slags via chemical modification, as exemplified by the treatment of blast furnace slag with 3 mol/L aqueous NaOH to afford a Cs adsorbent.²⁴⁾ Herein, steel slags were treated with H₂SO₄ to afford Se(IV) adsorbents, with the best adsorption performance obtained using 0.18–0.20 mol/L H₂SO₄. The treated slags were characterized by several instrumental techniques, and their absorption characteristics (adsorption capacity and adsorption constants) were estimated by Langmuir analysis.

2. Experimental

2.1. Materials

H₂SO₄ and H₂O₂ were of supra-pure grade (Kanto Chemical), while other chemicals were of analytical reagent grade. A Se standard solution (1000 mg/L; Kanto Chemical) was used for atomic absorption analysis. Water (>18 MΩ cm) purified by an ultrapure water production unit composed of a Millipore Elix Advantage 5 system and an Organo PURELAB Ultra Ionic system was used for all experiments.

Steelmaking and blast furnace slags cordially donated by steelmakers were ground in a zirconia mortar, sieved to <75 μm using a stainless steel sieve, calcined at 350°C for 4 h, and stored in a desiccator filled with dehydrating agents. Table 1 summarizes the contents elements in steelmaking slag used.

Table 1. Contents of elements in various steel making slags.

(mass%)
	Steel making slag No. 1*	No. 2	No. 3	No. 4
O*	39	39	41	40
Ca	26	28	33	27
Fe	18	18	14	14
Si	5.9	5.9	5.3	7.4
C	2.8	2.4	2.7	2.4
Al	2.6	1.4	0.65	1.5
Mn	2.4	2.4	1.1	4.6
Mg	2.0	1.9	1.1	1.6
P	0.81	0.94	1.1	1.4
Ti	0.28	0.26	0.093	0.56
Cr	0.18	0.14	0.19	0.21
V	0.091	0.20	0.066	0.12
S	0.051	0.031	0.067	0.026

The values were determined by XRF.

*  Content of oxygen is estimated by subtracting the contents of other elements from 100%.

2.2. Apparatus

The concentrations of Se in solutions were determined using a Shimadzu ICPS-7500 inductively coupled plasma emission spectrometer equipped with a Shimadzu HVG-1 type hydride generator. The Se measurement wavelength was set to 196.026 nm. A Horiba F-53 pH meter was used for pH measurements. X-ray fluorescence (XRF) spectroscopy and X-ray diffraction (XRD) analysis were performed using a ZSX Primus II instrument (Rigaku) and a RINT-2100 powder X-ray diffractometer (Rigaku), respectively.

2.3. Treatment of Slags with H₂SO₄

We modified the previous procedure²⁴⁾ to treat steelmaking slags with H₂SO₄. In brief, a 300-mL Erlenmeyer flask containing slag (5.0 g) was charged with aqueous H₂SO₄ (250 mL, 0.03–0.30 mol/L), and the mixture was stirred at room temperature for 24 h. The suspension was allowed to settle for several minutes, and the supernatant was discarded by decantation. The residue was treated with water (250 mL), and the mixture was vigorously stirred for 10 min. Subsequently, stirring was stopped, the suspension was allowed to settle for several minutes, and the supernatant was discarded as above. This operation was repeated until no sulfate ions could be detected in the supernatant by the BaSO₄ precipitation method. Thereafter, the treated slags were separated by filtration through No. 5C filter paper and oven-dried together with the filter at 60°C for 24 h. The dried slags were ground in a zirconia mortar and stored in a desiccator.

2.4. Se(IV) Adsorption Experiments

We also modified the previous method²⁴⁾ to quantify Se(IV) in a solution for the adsorption test. In brief, a 50-mL round-bottom centrifuge tube was sequentially charged with treated slag (0.10 g) and a solution of Se(IV) (40 mL, 5.0 × 10⁻⁵ mol/L). The suspension was shaken at 185 rpm for 42 h to establish an adsorption equilibrium and then centrifuged at 2500 rpm for 10 min. After pH measurement, the mixture was filtered through a 0.45-μm cellulose membrane filter, and the concentration of Se in the filtrate was determined as described in the “Apparatus” section.

3. Results and Discussion

3.1. Treatment of Slags with H₂SO₄

We have explored the treatment of steel slags with various agents to develop Se(IV) adsorbents. Although NaOH-treated blast furnace slags were capable of Cs adsorption,²⁴⁾ this treatment failed to impart Se(IV) adsorption ability to steelmaking slags and blast furnace slags. On the other hand, exploration of inorganic acids revealed that steelmaking slags treated with H₂SO₄ can adsorb Se(IV), while a treatment with HCl or HNO₃ resulted in slag dissolution.

The concentration of H₂SO₄ (c(H₂SO₄)) strongly influenced the development of Se(IV) adsorption ability, which appeared at c(H₂SO₄) = 0.18–0.20 mol/L for some slags (Fig. 1). Given that not all steelmaking slags developed this ability after treatment with H₂SO₄, we concluded that the emergence of Se(IV) adsorption ability largely depends on the inherent properties of raw steelmaking slags.

Fig. 1.

Effect of concentration of sulfuric acid in a treatment solution on Se(IV) adsorptivity of the treated slags and pH equilibrated for Se(IV) adsorption. Experimental conditions are as follows; the slag added is 0.1 g, volume of solution is 40 mL, initial concentration of Se(IV) is 5 × 10⁻⁵ mol/L, and shaking time is 42 h. (Online version in color.)

Figure 1 also shows the effect of c(H₂SO₄) on solution pH after 24 h equilibration, additionally presenting pH values obtained after 24 h equilibration of raw steelmaking slags in water as controls. The high pH values of the controls were ascribed to the presence of basic components such as CaO and MgO in steelmaking slags. With increasing c(H₂SO₄), the pH of the equilibrated solution rapidly decreased to approximately 4.0 but hardly decreased further at higher (>0.2 mol/L) concentrations, which was ascribed to the presence of buffer components (e.g., phosphates) in slags. Note that treated slags developed an ability to adsorb Se(IV) when the pH of the equilibrated solution reached ~4.0. However, at c(H₂SO₄) > 0.2 mol/L, the Se(IV) adsorption sites were destroyed, and the corresponding adsorption efficiency decreased.

Interestingly, in the case of slag No. 3, the efficiency of Se(IV) adsorption temporarily decreased at c(H₂SO₄) = 0.225 mol/L but recovered at approximately 0.25 mol/L. Similar behavior was observed for the treatment of steelmaking slag No. 1 with aqueous H₂SO₄ containing 12 wt% H₂O₂, which suggested the formation of two types of adsorption sites on steelmaking slags and further implied that the abundances of these sites depend on c(H₂SO₄) and H₂O₂ content. We mainly used slag No. 1 for the following investigations because of its abundant quantity compared to the other slags.

We examined the effect of H₂O₂ content on the efficiency of Se(IV) adsorption by steelmaking slags treated with H₂O₂ and H₂SO₄, as H₂O₂ is known to affect the structure of Cr(VI) adsorption sites on schwertmannite, a well-known As(III)-adsorbing ore.²⁰⁾ Figure 2 presents the results obtained for slag No. 1, revealing that the addition of H₂O₂ increased the efficiency of Se(IV) adsorption and widened the range of c(H₂SO₄) for the quantitative adsorption of Se(IV). This result suggests that once Fe in the slag is dissolved by H₂SO₄, the produced Fe(II) is oxidized by H₂O₂ to Fe(III), which affords Se(IV) adsorption sites similar to those of schwertmannite.20) Notably, slag No. 1 treated with aqueous H₂SO₄ containing 12 wt% H₂O₂ featured a bimodal profile similar to that of slag No. 3, as stated above (data not shown).

Fig. 2.

Effects of concentration of sulfuric acid and hydrogen dioxide in a treatment solution on Se(IV) adsorptivity of the treated slags. Experimental conditions are as follows; slag (No. 1) added is 0.1 g, volume of solution is 40 mL, initial concentration of Se(IV) is 5 × 10⁻⁵ mol/L, and shaking time is 42 h. (Online version in color.)

3.2. Effect of H₂SO₄ Treatment on Slag Composition

Figure 3 shows the effect of c(H₂SO₄) on slag elemental compositions determined by XRF, revealing that with increasing c(H₂SO₄), the S content of all slags increased, while the Ca content first decreased and then recovered. The slight decrease in Ca content at c(H₂SO₄) ≈ 0.1 M was ascribed to the dissolution of Ca from slags, and the subsequent recovery at high concentrations was attributed to the deposition of CaSO₄ on slags. On the other hand, the content of Fe monotonically decreased with increasing c(H₂SO₄) because of the simple dissolution of this element. The content changes of O, Ca, Fe, and S, the main constituents of slag, were not related to the Se(IV) adsorption profile, which suggested that these elements do not directly affect the adsorption of Se(IV). However, the profile of P, which is a minor component, was partially related to the Se(IV) adsorption profile, as shown in Fig. 4. More precisely, the decrease in Se(IV) absorption efficiency at c(H₂SO₄) ≈ 0.2 M was related to the decrease in P content, suggesting that P may participate in the formation of Se(IV) adsorption sites. We consider that hydroxyapatite is a possible P-containing phase capable of Se(IV) adsorption.^25,26)

Fig. 3.

Effect of concentration of sulfuric acid in a treatment solution on the contents of sulfur, iron, calcium, and oxygen in the treated slags. Each content of the element was determined by XRF. (Online version in color.)

Fig. 4.

Relationship between adsorptivity (close square) of Se(IV) and content of phosphorus (close circle) as a function of concentration of sulfuric acid. Experimental conditions are as follows; slag added is 0.1 g, volume of solution is 40 mL, initial concentration of Se(IV) is 5 × 10⁻⁵ mol/L, shaking time is 42 h. The dotted square in No. 2 clearly denotes a relationship between changes in both profiles. (Online version in color.)

3.3. Characterization of Treated Slags

The crystal structures of treated slags were characterized by XRD, with typical patterns of H₂SO₄-treated slag No. 1 shown in Fig. 5. In view of its amorphous structure, the untreated slag showed only a weak unassigned diffraction peak at 2θ = 18.1°, while the treated slag additionally showed four intense peaks. This result reveals that treatment with H2SO4 resulted in dissolution followed by recrystallization. The peaks of treated slag were assigned as follows: 2θ = 11.6° (CaSO₄), 20.7° (CaSO₄·H₂O), 23.3° (hydroxyapatite), 29.1° and 40.0° (schwertmannite), 43.3° and 43.6° (hydroxyapatite). The addition of H₂O₂ to H₂SO₄ hardly affected the XRD patterns of treated slags. The presence of hydroxyapatite supported the participation of P in the formation of Se(IV) adsorption sites, as mentioned above.

Fig. 5.

XRD patterns of steelmaking slags treated with sulfuric acid. Possible assignments are as follows: I, CaSO₄; II, CaSO₄·H₂O; III, hydroxyapatite; IV, schwertmannite. (Online version in color.)

The particle morphologies and surface structures of treated slags were observed by scanning electron microscopy (SEM). As shown in Fig. 6, untreated steelmaking slag comprised crushed particles, while treated slag particles had plate-/rod-like shapes and smooth flat surfaces, which confirms the occurrence of dissolution and recrystallization during H₂SO₄ treatment. The particle sizes of the treated slags observed in Fig. 6 are much smaller than 75 μm which is the sieve opening used, confirming dissolution of the raw slags. The addition of H₂O₂ had no effects on slag surface structure. Considering the results presented in Figs. 4, 5, 6, we concluded that treated slags contain CaSO₄ as the main component, while schwertmannite and hydroxyapatite are minor components largely determining slag adsorption performance.

Fig. 6.

SEM images of steelmaking slags No. 1 treated with sulfuric acid and sulfuric acid with hydrogen peroxide.

The pore size distribution of slag treated with 0.18 mol/L H₂SO₄ was measured by the Brunauer–Emmett–Teller method. Slag No. 1 was shown to have a specific surface area of 30.4 m²/g and contain micropores with an average diameter of 2.4 nm.

3.4. Effect of Equilibrated Solution pH on Se(IV) Adsorption

Generally, the adsorption behavior of inorganic ions is governed by the pH of the equilibrated solution, as in aqueous media, these Lewis-acidic ions compete with protons for adsorption sites. Figure 7 shows the effect of pH on the efficiency of Se(IV) adsorption. pH in the acidic range was adjusted using an acetate buffer, while pH in the basic range was adjusted using an ammonia buffer. In view of the strong buffering action of slags, the equilibrium pH could not be controlled in the ranges of 5–8 and 9–10, reaching a value of 4.0 when treated slags were added to water without a buffering component. Whereas treated slags showed a 100% Se(IV) adsorption efficiency under acidic conditions (pH < 5.0), the adsorption efficiency of Se(IV) under alkaline conditions was significantly affected by the pH of the equilibrated solution. As the equilibrated solution pH could not be precisely controlled because of the buffering action of slags, further quantitative studies could not be performed in the basic pH range.

Fig. 7.

Effect of solution pH equilibrated on adsorption of selenium with steel making slag No. 1 treated with 0.16 mol/L H₂SO₄ containing 6% H₂O₂. Experimental conditions are as follows; adsorbent is 0.1 g, volume is 40 mL, concentration of Se is 5 × 10⁻⁵ mol/L, and shaking time is 42 h.

As mentioned in the section on treated slag characterization, XRD analysis revealed the existence of a schwertmannite phase, which adsorbs Se(IV) as well as As(III). Considering the reported mechanism of As adsorption on schwertmannite, Se(IV) was assumed to be incorporated into schwertmannite as a result of ion exchange with sulfate ions in Fe(III) hydroxide clusters. As ion exchange does not involve the release of protons from the adsorbent, the efficiency of Se(IV) adsorption under acidic conditions was independent of solution pH.

3.5. Adsorption Characteristics

The characteristics of Se(IV) adsorption by treated slag were elucidated by analysis of the corresponding adsorption isotherms. The adsorption isotherms of slags treated with 0.18 mol/L H₂SO₄ containing H₂O₂ and the Langmuir analyses of these isotherms are depicted in Figs. 8(a) and 8(b), respectively. Langmuir analysis was adopted to estimate the quantity of saturated adsorption sites and adsorption constants. Even though Freundlich analysis afforded slightly better correlation coefficients than Langmuir analysis, Langmuir adsorption was a more suitable model to follow the ion exchange of Se(IV) at the surface of treated slags. Because Freundlich model considers a multi-layer adsorption, it is not suitable to analyze ion exchange adsorption with Freundlich model.

Fig. 8.

a) Adsorption isotherms of Se(IV) on steel making slag No. 1 treated with 0.18 mol/L sulfuric acid with hydrogen peroxide and b) their plots of Langmuir analyses. Experimental conditions are as follows; adsorbent is 0.1 g, volume is 40 mL, concentration of Se is 5 × 10⁻⁵ mol/L, shaking time is 42 h, temperature is 25°C, and pH is 3.5. (Online version in color.)

Langmuir equation is expressed with Eq. (1)   

$$C\text{/q}=\text{C/}{Q}_{\text{max}}+K\text{/}{Q}_{\text{max}}$$(1)

where C is equilibrated concentration of Se(IV) in a solution [mol/L], q is amount of Se(IV) adsorbed on slags [mol/kg], Q_max is saturated amount of Se(IV) on slags [mol/kg], K is equilibrium constant of adsorption of Se(IV). Figure 8(b) depicts plots of C/q vs. C. Q_max and K were estimated from slops and intercepts of the linear curves in Fig. 8(b). Table 2 summarizes the results obtained for slag No. 1 treated using 0.18 mol/L H₂SO₄ with varying contents of H₂O₂. Interestingly, as shown in Fig. 8, the adsorption isotherm of the slag treated with H₂SO₄ containing 3 wt% H₂O₂ was positioned lower than that of the slag treated with H₂SO₄ without H₂O₂. On the other hand, the monotonic increase of adsorption constants (Table 2) with increasing H₂O₂ content indicated that H₂O₂ facilitates the formation of stronger adsorption sites to increase the average adsorption constant. As already explained above, H₂O₂ oxidized dissolved Fe(II) to Fe(III), which participated in the formation of Se(IV) adsorption sites. The minimal adsorption capacity was observed for the slag treated with H₂SO₄ containing 3 wt% H₂O₂, in agreement with the order of isotherms presented in Fig. 8(b). Probably, the formation of Fe(III)-containing adsorption sites might compete with that of Fe(III)-free sites.

Table 2. Adsorption constants and amounts of saturated adsorption sites of steel making slag No. 1 analyzed by Langmuir equation.

Treatment condition	Qmax/mol/kg	K/L/mol
0.18 mol/L H2SO4	0.07	1.7 × 104
0.18 mol/L H2SO4 + 3% H2O2	0.05	2.1 × 104
0.18 mol/L H2SO4 + 6% H2O2	0.10	5.3 × 104
0.18 mol/L H2SO4 + 3% H2O2	0.17	4.1 × 104

Data from Fig. 8 were calculated.

4. Conclusion

Adsorbents for Se(IV) were produced by the treatment of steelmaking slags with H₂SO₄, and notable Se adsorption was observed only for H₂SO₄ concentrations of 0.18–0.20 mol/L. The addition of H₂O₂ to H₂SO₄ increased the efficiency of Se(IV) adsorption. The XRD patterns of treated slags revealed the formation of schwertmannite and hydroxyapatite phases providing Se(IV) adsorption sites, and the characteristics of Se(IV) adsorption on treated slags (adsorption capacity and adsorption constant) were determined using the Langmuir model to elucidate probable adsorption sites.

References

• 1)   J. T.  Rotruck,  A. L.  Pope,  H. E.  Ganther,  A. B.  Swanson,  D. G.  Hafeman and  W. G.  Hoekstra: Science, 179 (1973), 588. https://doi.org/10.1126/science.179.4073.588
• 2)   C. K.  Chow and  C. J.  Chen: J. Nutr., 110 (1980), 2460. https://doi.org/10.1016/j.arcmed.2016.12.010
• 3)   L.  Letavayová,  V.  Vlčková and  J.  Brozmanová: Toxicology, 227 (2006), 1. https://doi.org/10.1080/00034983.1978.11719356
• 4)   C. B.  Tabelin,  A. H. M.  Basri,  T.  Igarashi and  T.  Yoneda: Water Air Soil Pollut., 223 (2012), 4153. https://doi.org/10.1007/s11270-012-1181-x
• 5)   A.  Król,  K.  Mizerna and  M.  Bożym: J. Hazard. Mater., 384 (2020), 121502. https://doi.org/10.1016/j.jhazmat.2019.121502
• 6)   H.  Yoshida,  M.  Taga and  S.  Hikime: Bunseki Kagaku, 14 (1965), 1109 (in Japanese). https://doi.org/10.2116/bunsekikagaku.14.1109
• 7)   J.-M.  Lo,  C.-C.  Lin and  S.-J.  Yeh: Anal. Chim. Acta, 272 (1993), 169.
• 8)   A. K.  Malik and  A. L. J.  Rao: Talanta, 37 (1990), 1205.
• 9)   D.  Das,  U.  Gupta and  A. K.  Das: TrAC Trends Anal. Chem., 38 (2012), 163. https://doi.org/10.1016/j.trac.2011.01.020
• 10)   C. H.  Latorre,  J. B.  Garcia,  S.  García Martín and  R. M.  Peña Crecente: Anal. Chim. Acta, 804 (2013), 37. https://doi.org/10.1016/j.trac.2011.01.020
• 11)   K.  Hagiwara,  Y.  Koike,  M.  Aizawa and  T.  Nakamura: Anal. Sci., 34 (2018), 1309. https://doi.org/10.2116/analsci.18P217
• 12)   Y.  Zhang,  B.  Chen,  S.  Wu,  M.  He and  B.  Hu: Talanta, 154 (2016), 474. https://doi.org/10.1016/j.talanta.2016.04.003
• 13)   K.  Kocot,  R.  Leardi,  B.  Walczak and  R.  Sitko: Talanta, 134 (2015), 360. https://doi.org/10.1016/j.talanta.2014.11.036
• 14)   H.  Abdolmohammad-Zadeh,  A.  Jouyban,  R.  Amini and  G.  Sadeghi: Microchim. Acta, 180 (2013), 619. https://doi.org/10.1007/s00604-013-0967-x
• 15)   K.  Huang,  K.  Xu,  W.  Zhu,  L.  Yang,  X.  Hou and  C.  Zheng: Anal. Chem., 88 (2016), 789. https://doi.org/10.1021/acs.analchem.5b00511
• 16)   M.  Llaver,  E. A.  Coronado and  R. G.  Wuilloud: Spectrochim. Acta B, 138 (2017), 23. https://doi.org/10.1016/j.sab.2017.10.003
• 17)   R. K.  Behera,  K.  Rout,  B.  Nayak and  N. N.  Das: Adsorpt. Sci. Technol., 30 (2012), 867. https://doi.org/10.1260/0263-6174.30.10.867
• 18)   K.  Khamphila,  R.  Kodama,  T.  Sato and  T.  Otake: J. Miner. Mater. Charact. Eng., 5 (2017), 90. https://doi.org/10.4236/jmmce.2017.52008
• 19)   M.  Takada,  K.  Fukushi,  T.  Sato and  T.  Yoneda: Nendo Kagaku (J. Clay Sci. Soc. Jpn.), 47 (2008), 255 (in Japanese). https://doi.org/10.11362/jcssjnendokagaku.47.4_255
• 20)   Z.  Zhang,  G.  Guo,  X.  Li,  Q.  Zhao,  X.  Bi,  K.  Wu and  H.  Chen: J. Hazard. Mater., 367 (2019), 520. https://doi.org/10.1016/j.jhazmat.2018.12.116
• 21)   Y.  Xue,  H.  Hou and  S.  Zhu: J. Hazard. Mater., 162 (2009), 391. https://doi.org/10.1016/j.jhazmat.2008.05.072
• 22)   V. K.  Jha,  Y.  Kameshima,  A.  Nakajima and  K.  Okada: J. Hazard. Mater., 114 (2004), 139. https://doi.org/10.1016/j.jhazmat.2004.08.004
• 23)   C.  Oh,  S.  Rhee,  M.  Oh and  J.  Park: J. Hazard. Mater., 213–214 (2012), 147. https://doi.org/10.1016/j.jhazmat.2012.01.074
• 24)   N.  Uehara,  H.  Tairako,  T.  Yamaguchi,  K.  Hanada and  K.  Fujimoto: Bunseki Kagaku, 66 (2017), 809 (in Japanese). https://doi.org/10.2116/bunsekikagaku.66.809
• 25)   M.  Duc,  G.  Lefevre,  M.  Fedoroff,  J.  Jeanjean,  J. C.  Rouchaud,  F.  Monteil-Rivera,  J.  Dumonceau and  S.  Milonjic: J. Environ. Radioact., 70 (2003), 61.
• 26)   S.  Kongsri,  K.  Janpradit,  K.  Buapa,  S.  Techawongstien and  S.  Chanthai: Chem. Eng. J., 215–216 (2013), 522. https://doi.org/10.1016/j.cej.2012.11.054