Using Biogenic Manganese Oxides

0 downloads 0 Views 464KB Size Report
Oct 7, 2009 - Data were analyzed with CASAXPS software (Ver. 2.1.1). Backgrounds were drawn using the. Shirley method.16). 3. Results and Discussion.
Materials Transactions, Vol. 50, No. 11 (2009) pp. 2643 to 2648 #2009 The Mining and Materials Processing Institute of Japan

Selective Sorption of Co2þ over Ni2þ Using Biogenic Manganese Oxides Keiko Sasaki, Takuya Kaseyama and Tsuyoshi Hirajima Department of Earth Resources Engineering, Kyushu University, Fukuoka 819-0395, Japan Preferential sorption of Co2þ ions over Ni2þ ions was achieved using biogenic Mn oxides produced by the Mn-oxidizing fungus Paraconiothyrium sp. WL-2 strain with a maximum selectivity coefficient (Co ) of 18. The selective sorption was based on different sorption mechanism for Co2þ and Ni2þ and unique properties of biogenic Mn oxides. The octahedral Co2þ ions occupy vacancies of central metal sites and edge sites in the octahedral Mn oxide unit structures of biogenic Mn oxides, where they are immobilized by oxidation to CoOOH by Mn(III). In contrast, Ni2þ ions are sorbed primarily on layer edges at circumneutral pHs without oxidation. Selective sorption of Co2þ over Ni2þ on the biogenic Mn oxides results from more vacant sites, higher Mn(III) contents, and larger specific surface areas compared to synthetic Mn oxides. [doi:10.2320/matertrans.M-M2009821] (Received April 10, 2009; Accepted August 4, 2009; Published October 7, 2009) Keywords: selective sorption, cobalt ions, biogenic manganese oxides, selectivity coefficient

1.

Introduction

Manganese oxide deposits are often found in natural biofilms at neutral pHs in hot springs, mining sites, and hydrothermal vents. Most of these oxides are produced by microbiologically induced processes. These processes have recently been applied to treat Mn-bearing drainage using natural attenuation in wetlands.1,2) The oxidation products are mainly insoluble Mn(III, IV) oxides of the birnessite family, composed of edge-sharing MnO6 octahedral with high negative structural charge due principally to vacant cation layer sites, leading to high cation sorption capacities.3–6) Natural Mn oxides often are poorly crystalline solids with high specific surface areas,7) and few structural investigations of them have been performed.8) Their unique properties should cause various interactions with major cations, heavy metals and rare earth elements in the environment including sorption, oxidation/reduction reactions, and precipitation. Dong et al. have found that biogenic Mn oxides accumulated Pb2þ preferentially over Cd2þ .9) This observation was explained on the basis of the different ionic radii of the cations and their ability to occupy vacant sites on the surface of biogenic Mn oxides. It is well known that manganese nodules are rich in several rare metals especially cobalt.10) Detailed studies of natural accumulation would provide information to establish separation and recovery techniques. Nickel and cobalt occur in association with manganese in ocean nodules,11) nickel laterite,12) and manganiferous deposits.13) In the present work, selective sorption of Co2þ over Ni2þ using biogenic Mn oxides was investigated at ambient temperatures and circumneutral pHs. 2.

Experimental

Biogenic Mn oxides were produced by a Mn-oxidizing fungus, Paraconiothyrium sp. WL-2 strain14) and characterized as poorly crystalline birnessite with a specific surface area of 81.4 m2 g1 . The details of the characterization were reported elsewhere.15) Synthetic Mn oxide was produced by oxidation of MnSO4 solution with ammonium peroxodisulfate with heating to boiling. The product was ramsdellite

(-MnO2 ) with a specific surface area of 11.5 m2 g1 . Average oxidation states in biogenic and synthetic Mn oxides were estimated by determination of Mn(II), Mn(IV) and total Mn.15) The Mn(II) contents were determined by atomic absorption spectrometry (Solar AAS Thermo Elemental) after extraction with 50 mM CuSO4 . After the required mass of Mn oxides were dissolved in oxalate to reduce Mn(IV), the residual oxalate was determined by titration with KMnO4 . Sorption experiments using both materials were conducted as follows: 0.020 g of freeze-dried biogenic Mn oxide or 0.015 g of freeze-dried synthetic Mn oxide powder was suspended in 0.150 dm3 of 0.01–1.7 mmoldm3 Co(NO3 )2  6H2 O or 0.03–2.11 mmoldm3 NiCl2 6H2 O in 0.500 dm3 in Erlenmeyer flasks capped with rubber stoppers. Under these conditions the solid Mn concentration corresponds to 1 mmoldm3 Mn for both the biogenic and synthetic Mn oxides. KNO3 was added in each sample to provide a constant ionic strength of 100 mmoldm3 . The initial pH of the Co2þ and Ni2þ solutions, buffered using PIPES (piperazine-N, N0 -bis(2-ethane sulfonic acid)), was 6.5. Each experiment was conducted in duplicate. The flasks were installed on a reciprocating shaker at 25 C and 100 rpm for 300 h. At intervals, the supernatants were sampled and filtered by a 0.2-mm membrane filter for determination of the Co2þ , Ni2þ and Mn2þ concentrations in the solution using inductively coupled plasma-mass spectrometry (ICPMS, Agilent 7500c). The detection limits by ICP-MS are 0.06 mgdm3 for Mn, and 0.01 mgdm3 for Co and Ni. Sorption experiments of Co2þ and Ni2þ ions in a binary system using biogenic and synthetic Mn oxides were also conducted. For these experiments the initial concentration of Co2þ ions was fixed at 0.3 mmoldm3 while the initial concentrations of Ni2þ were varied from 0.1 to 1.0 mmol dm3 . The ratio of sorbent mass to solution volume was the same as in a single-solute experiment. Each experiment was conducted in duplicate. The solid residues were filtered using a 0.45-mm membrane filter, and then lyophilized overnight. To investigate the chemical states of Mn, Co and Ni in the precipitates, XP-spectra for the biogenic and synthetic Mn oxides after Co-Ni sorption were collected using a PHI 5800 ESCA. After evacuating to less than 5:0  107 Pa for more

2644

K. Sasaki, T. Kaseyama and T. Hirajima

Table 1 Molar ratios of Mn(II), Mn(III) and Mn(IV) and average oxidation states (AOS) of Mn in the biogenic and synthetic Mn oxides.

3.09

synthetic

0.014

0.296

0.690

3.68

than 30 min, the sample was transferred into the analysis chamber at less than 2:0  109 Pa, and then irradiated with monochromatic Al K X-rays using a neutralizer. The pass energy was 93.90 eV, the step energy was 0.800 eV and the step was 20 ms. The binding energy, EB , was calibrated with EB [C 1s] ¼ 284:8 eV. Data were analyzed with CASAXPS software (Ver. 2.1.1). Backgrounds were drawn using the Shirley method.16) 3.

Results and Discussion

Molar ratios of Mn(II), Mn(III) and Mn(IV), and average oxidation states (AOS) of biogenic and synthetic Mn oxides are summarized in Table 1. The proportions of constituents in the two solids are quite different. The Mn(II) species are ten times greater in biogenic Mn oxides than synthetic Mn oxides. The dominant Mn species are Mn(III) in biogenic Mn oxides and Mn(IV) in synthetic Mn oxides, resulting in 3.17 and 3.70 of AOS, respectively. A difference in AOS would affect the reactivity as an oxidant.16) Figure 1 shows the time courses of Co2þ and Mn2þ ions during sorption of Co2þ ions on biogenic Mn oxides at 25 C. The Co2þ ions were well sorbed on the biogenic Mn oxides in Fig. 1(a). About 200 h were required to achieve equilibrium when the initial Co2þ concentration was 1.74 mmoldm3 . In experiments with higher initial concentrations of Co2þ , data scatter was probably caused by the small amount of sorbent (0.020 g) relative to solute. Initial release of Mn2þ was caused by ion-exchange with Co2þ ions, indicating that the selectivity of ion-exchange is in the order of Co2þ > Mn2þ , followed by gradual re-sorption of Mn2þ in Fig. 1(b). The maximum release of Mn was from 0.10 to 0.12 mmoldm3 in the presence of 0.197–1.74 mmoldm3 Co2þ ions. The mass of biogenic Mn oxides used here includes Mn(II) corresponding to around 0.12 mmoldm3 of Cu-exchangeable Mn(II), based on around 5 mass% Mn(II) contents as previously reported.7) The re-sorption rate Mn2þ was faster with lower initial Co2þ concentrations. In the absence of Co2þ Mn release was still observed (about 0.04 mmoldm3 ), but the re-sorption was much slower than when Co2þ was present. It has been reported that Co2þ ions are oxidized by biogenic and synthetic Mn oxides to form CoOOH and that Mn(III) species are more effective oxidants of Co2þ than Mn(IV) species in the biogenic Mn oxides.7) These observations indicate that released Mn(II) ions are immobilized by co-precipitation with CoOOH.17) In contrast, Co2þ ions were less sorbed on the synthetic Mn oxides and no release of Mn ions were observed (Fig. 2). The results are consistent with the lack of Cu-exchangeable Mn(II) and Mn(III) in the synthetic Mn oxides.7) Nickel ions were much less sorbed than Co2þ ions on the biogenic and synthetic Mn oxides as shown in Figs. 3(a) and 4(a). These results are consistent with previous reports by

2.0 1.5 1.0

2+

AOS

0.196

Co concentration / mM

Mn(IV)

0.694

0.5 0.0

0

50

100

150

200

250

300

Time /h 0.14

(b)

2+

Mn(III)

0.11

(a)

Mn concentration / mM

Mn(II) biogenic

2.5

0.12 0.10 0.08 0.06 0.04 0.02 0.00

0

50

100

150

200

250

300

Time /h Fig. 1 Sorption behavior of Co2þ ions on the biogenic Mn oxide at 25 C (a) and changes in released Mn2þ ions with time (b) (n ¼ 2). Symbols: : initial [Co2þ ] = 0 mmoldm3 , : 0.0539 mmoldm3 , : 0.197 mmol dm3 , : 0.314 mmoldm3 , : 0.670 mmoldm3 , : 0.957 mmol dm3 , : 1.74 mmoldm3 . The error bar indicates 1.

Tani et al.18) Very little sorbed Ni2þ was detectable by ICPMS with the synthetic Mn oxides (Fig. 4(a)). The required times to reach equilibrium were shorter than in case of Co2þ in both materials (Figs. 1(a), 2(a), 3(a), 4(a)). It took less than 50 h to attain equilibrium with the biogenic Mn oxide even when the initial Ni2þ concentration was 2.3 mmoldm3 (Fig. 3(a)). Fewer Mn2þ ions were released (Fig. 3(b) and Fig. 1(b)) and the response was slower, depending on the loading of Ni2þ ions (Fig. 3(b)). The selectivity for ion-exchange is in the order of Co2þ > Mn2þ > Ni2þ (Figs. 1(b), 3(b)). Sorption data for Co2þ and Ni2þ on the biogenic and synthetic Mn oxides system were well described by the linearized Freundlich isotherm as shown in Fig. 5. log Q ¼ log F þ n log Ce

ð1Þ

in which Q is the sorbed quantity of the solute per unit mass of sorbent as Mn-kg (mmol kg1 -Mn), Ce is the equilibrium concentration of solute (mmol dm3 ), and F (dm3 kg1 ) and n (dimensionless) are constants. In preliminary experiments, it was confirmed that cellular tissues were not responsible for sorption of Co2þ and Ni2þ on the biogenic Mn oxides. The Q values were normalized for mass of Mn(III, IV) excluding Mn(II) in Mn oxides in Fig. 5. It can be seen that Q values for Co2þ and Ni2þ sorption to biogenic Mn oxide were in an order of magnitude greater than those to the synthetic Mn oxide. The difference is primarily caused by the large

Selective Sorption of Co2þ over Ni2þ Using Biogenic Manganese Oxides

2645

0.14

2.5

(a)

(b) concentration / mM

0.12

Co concentration / mM

2.0 1.5

Mn

2+

2+

1.0 0.5

0.10 0.08 0.06 0.04 0.02

0.0

0

50

100

150

200

0.00

250

0

50

100

Time / h

150

200

250

Time / h

Fig. 2 Sorption behavior of Co2þ ions on the synthetic Mn oxide at 25 C (a) and changes in released Mn2þ ions with time (b) (n ¼ 2). Symbols: : initial [Co2þ ] = 0.0922 mmoldm3 , : 0.184 mmoldm3 , : 0.354 mmoldm3 , : 0.675 mmoldm3 , : 1.11 mmoldm3 , : 2.19 mmoldm3 . The error bar indicates 1. 0.14

2.5

concentration / mM

(b)

2.0 1.5 1.0

Mn

Ni

2+

2+

concentration / mM

(a)

0.5

0.12 0.10 0.08 0.06 0.04 0.02

0.0

0

50

100

150

200

250

0.00

300

0

50

100

150

200

250

300

Time /h

Time /h

Fig. 3 Sorption behavior of Ni2þ ions on the biogenic Mn oxide at 25 C (a) and changes in released Mn2þ ions with time (b) (n ¼ 2). Symbols: : initial [Ni2þ ] = 0 mmoldm3 , : 0.0277 mmoldm3 , : 0.108 mmoldm3 , : 0.280 mmoldm3 , : 0.552 mmoldm3 , : 0.922 mmoldm3 , : 1.36 mmoldm3 , : 2.31 mmoldm3 . The error bar indicates 1. 0.14

2.5

(b) Mn concentration / mM

2.0 1.5

2+

1.0

2+

Ni concentration / mM

(a)

0.5

0.12 0.10 0.08 0.06 0.04 0.02

0.0

0

20

40

60

80

100

120

140

Time /h

0.00

0

20

40

60

80

100

120

140

Time /h

Fig. 4 Sorption behavior of Ni2þ ions on the synthetic Mn oxide at 25 C (a) and changes in released Mn2þ ions with time (b) (n ¼ 2). Symbols: : initial [Ni2þ ] = 0.0278 mmoldm3 , : 0.107 mmoldm3 , : 0.272 mmoldm3 , : 0.549 mmoldm3 , : 3 3 0.857 mmoldm , : 1.27 mmoldm . The error bar indicates 1.

specific surface area of the biogenic oxides. In addition, the Q values were in an order of magnitude greater for Co2þ than for Ni2þ in both Mn oxides. This indicates there is specific sorption of Co2þ rather than physical sorption. The sorption data of Co2þ and Ni2þ ions on the biogenic and synthetic Mn oxides in Figs. 1, 2, 3 and 4 were also

fit with the linearized Langmuir equation, as shown in Fig. 6. Ce Q1 ¼ 1=ðQmax LÞ þ Ce =Qmax ;

ð2Þ

in which Qmax is sorption capacity of the solid and L is the Langmuir constant corresponding to the affinity of com-

2646

K. Sasaki, T. Kaseyama and T. Hirajima 4.5

Ni 2+ logQ=3.33+0.544*logCe Ni

3.5

-1

3.0

logQ=4.10+0.506*logCe

2.5

logQ=2.70+0.480*logCe

1.5 1.0

-5

-4

-3

-2

0.3 0.2 0.1

-1

0.0

0

1

-3

log [Ce / mmol dm ] 2þ

(a)

Co

2+

logQ=3.31+0.445*logCe

2.0



Ni

2+

Fig. 5 Sorption isotherms of Co and Ni on biogenic and synthetic Mn oxides fitted with the linearized Freundlich model. Symbols: & : Co2þ ; & : Ni2þ . Solid and open symbols indicate biogenic and synthetic Mn oxides.

concentration / mM

log [Q / mmol kg -Mn(III,IV)]

0.4

2+

Co 2+ Co

concentration / mM

2+

biogenic 4.0 synthetic

-3

(b)

1.0 0.8 0.6 0.4 0.2 0.0

1.0x10

2+

(a) Biogenic

Co

2+

0.6

-4

Ce /Q = 1.5*10 + 3.0*10 *Ce 2

(r = 0.852)

-1

Ce·Q / kg·dm

-3

-4

concentration / mM

Ni

0.8

-5

0.2

-5

Ce /Q = 1.9*10 + 8.1*10 *Ce 2

4x10

0.5

1.0 1.5 -3 Ce / mmol·dm

2.0

2.5

-3

(b) Synthetic

Co Ni

2+

2+

-3

3

-1

Ce·Q / kg·dm

0.04 0.00

(r = 0.746)

0.0 0.0

0.08

Mn

2+

0.4

(c) 0.12

-4

-3

Ce /Q = 3.0*10 + 2.0*10 *Ce 2

(r = 0.963)

0

50

100

150

200

250

300

Time / h Fig. 7 Time courses of (a) Co2þ , (b) Ni2þ , (c) Mn2þ concentrations during sorption of Co2þ and Ni2þ on the biogenic Mn oxide. Symbols: : initial [Co2þ ] and [Ni2þ ] are 0.31 mmoldm3 and 0 mmoldm3 , : initial [Co2þ ] and [Ni2þ ] are 0.31 mmoldm3 and 0.10 mmoldm3 , : initial [Co2þ ] and [Ni2þ ] are 0.35 mmoldm3 and 0.35 mmoldm3 , : initial [Co2þ ] and [Ni2þ ] are 0.34 mmoldm3 and 1.00 mmoldm3 . The error bar indicates 1.

2

-4

-4

Ce /Q = 2.0*10 + 2.1*10 *Ce

1

2

(r = 0.879)

0 0.0

0.5

1.0

1.5

Ce / mmol·dm

2.0

2.5

-3

Fig. 6 Sorption isotherm of Co2þ and Ni2þ on biogenic and synthetic Mn oxides fitted with the linearlized Langmuir model. Symbols are the same as in Fig. 5.

pounds to the solid.19) As can be seen from the correlation coefficients, larger scatter was observed for the biogenic Mn oxides compared to the synthetic ones, and sorption of Ni2þ was better described by the linearized Langmuir equation than was sorption of Co2þ . These results suggest that there is poorer uniformity in sorption sites on biogenic Mn oxides than on synthetic ones, and that Ni2þ ions are more preferably sorbed on the limited and uniform sites than are Co2þ ions.

Time courses of Co2þ , Ni2þ , and Mn2þ for competitive sorption in the presence of 0.3 mmoldm3 Co2þ and 0.1– 1.0 mmoldm3 Ni2þ onto the biogenic Mn oxides are shown in Fig. 7. The Mn2þ ions (0.10–0.12 mmoldm3 ) were released within 24 h primarily due to ion-exchange with Co2þ ions in biogenic Mn oxides (Fig. 8(a), (c)), in a manner similar to the data shown in Fig. 1. After that Mn2þ ions were immobilized by co-precipitation with CoOOH. While nickel ions were less sorbed, they inhibited sorption of Co2þ ions (Fig. 7(a), (b)). This trend was observed also in synthetic Mn oxides in Fig. 8(a), (b). Significantly greater selective sorption of Co2þ over Ni2þ was observed on the biogenic Mn oxides. The selectivity coefficient (1 ) of component 1 in the presence of component 2 is expressed in the following equations, where X and C indicate the mole fractions in sorbed and aqueous phases, respectively. 1 ¼ ðX1 =C1 Þ=ðX2 =C2 Þ

ð3Þ

The selectivity coefficients of Co2þ (Co ) and Ni2þ (Ni )

Selective Sorption of Co2þ over Ni2þ Using Biogenic Manganese Oxides

240

2+

(a) Intensity / cps

Co concentration / mM

0.4 0.3 0.2

(a) biogenic

Mn 3s

200 180 160

0.1

95 450 Intensity / cps

concentration / mM 2+

220

140

0.0

Ni

2647

(b)

1.0 0.8

90

85

80

(b) synthetic

Mn 3s

400 350 300

0.6

95

0.4

90

85

80

Binding energy / eV

0.2

Fig. 9 XP-spectra of the Mn 3s region for (a) the biogenic Mn oxide and (b) the synthetic Mn oxide after Co-Ni sorption.

0.0 1600 Intensity / cps

0.12 0.10 0.08 0.06

1400 1200 1000

0.04

810

0.02 0.00

Co 2p

(a) biogenic

(c)

2200

0

50

100

150

200

250

300

Time / h Fig. 8 Time courses of (a) Co2þ , (b) Ni2þ , (c) Mn2þ concentrations during sorption of Co2þ and Ni2þ on the synthetic Mn oxide. Symbols: : initial [Co2þ ] and [Ni2þ ] are 0.31 mmoldm3 and 0 mmoldm3 , : initial [Co2þ ] and [Ni2þ ] are 0.31 mmoldm3 and 0.10 mmoldm3 , : initial [Co2þ ] and [Ni2þ ] are 0.35 mmoldm3 and 0.35 mmoldm3 , : initial [Co2þ ] and [Ni2þ ] are 0.34 mmoldm3 and 1.00 mmoldm3 . The error bar indicates 1. Table 2 Selectivity coefficients of Co2þ and Ni2þ ions in the binary system. Biogenic 2þ

Synthetic

initial[Co ]/mM

0.31

0.35

0.34

0.31

0.35

0.34

initial[Ni2þ ]/mM

0.1

0.35

1.0

0.1

0.35

1.0

Co

7.92

7.47

17.7

5.65

2.23

2.65

Ni

0.126

0.134

0.056

0.177

0.449

0.378

based on the results in Figs. 7 and 8 are summarized in Table 2. It can be seen that biogenic Mn oxides showed greater Co values than synthetic ones, and that the greatest Co value was obtained with the greatest co-existing Ni2þ concentration. Release of Mn2þ ions was almost independent of coexisting Ni2þ , and primarily dependent on loaded mass of Co2þ ions (Fig. 1(b), Fig. 7(c)). This is reasonable considering that

Intensity / cps

2+

Mn concentration / mM

0.14

800

790

780

(b) synthetic

770

760

Co 2p

2000 1800 1600 810

800

790

780

770

760

Binding energy / eV Fig. 10 XP-spectra of the Co 2p-Mn 2s region for (a) the biogenic Mn oxide and (b) the synthetic Mn oxide after Co-Ni sorption.

selectivity of ion exchange with Mn2þ in the biogenic Mn oxide is greater for Co2þ over Ni2þ . However, as shown in Fig. 8, the selective sorption as above was not clearly seen with the synthetic Mn oxides. The high selectivity is one of the unique characteristics of the biogenic Mn oxides. XP-spectra of the Mn 3s, Co 2p, and Ni 2p regions for the biogenic Mn oxide and the synthetic Mn oxide after Co-Ni sorption in a binary system were collected. Based on the analysis of Mn 3s spectra for both Mn oxides, the splitting of Mn 3s and the satellite was 4.7 eV for both (Fig. 9), suggesting that the surfaces of the Mn oxides consist of mostly Mn(IV). Figure 10 shows there are two distinct peaks at EB [Co 2p3=2 ] ¼ 780:8 eV and EB [Co 2p1=2 ] ¼ 795:9 eV without satellite peaks of Co 2p, that is, Co(III) is the dominant species on both surfaces. Redox reactions between Co2þ and Mn(III, IV) on the solid surfaces generated Mn(II) and/or Mn(III). However, in the Ni 2p spectra for

2648

K. Sasaki, T. Kaseyama and T. Hirajima

Intensity / cps

1350 1300

(a) biogenic

Ni 2p

1250 1200 1150 1100 890

880

870

860

850

Intensity / cps

2200 2100

(b) synthetic

Ni 2p

2000 1900 1800 1700

890

880

870

860

850

the biogenic Mn oxides. Nickel ions do not compete with Co2þ for edge sites. In addition, the Co2þ ions taken up would be preferentially oxidized by the nearest Mn(III) and/ or Mn(IV) and immobilized as Co(III) compounds such as CoOOH with the release of Mn2þ ions. Therefore, in the biogenic Mn oxides, competitive sorption of Mn2þ and Ni2þ ions occurred at the vacant sites. It has been speculated based on XANES and other structural analysis techniques21) that there are many more vacant sites and edge sites in biogenic Mn oxides than in synthetic Mn oxides. In addition, biogenic Mn oxides are usually poorly crystalline with non-uniform morphologies but highly uniform pore sizes of around 4 nm.7) The unique properties of biogenic Mn oxides suggest a potential to separate Co from Ni in leachates of natural resources. Acknowledgement

Binding energy / eV Fig. 11 XP-spectra of the Ni 2p region for (a) the biogenic Mn oxide and (b) the synthetic Mn oxide after Co-Ni sorption.

Financial support was provided by Japan Society for Promotion of Science (JSPS Grant-in-Aid for Scientific Research No. 21656230). The authors thank Dr. Mario Villalobos at Institute of Geografia Ciudad University and Dr. Robert Bowman at New Mexico Institute of Mining and Technology for valuable discussion. REFERENCES

Fig. 12 Sorption mechanism onto the biogenic Mn oxide. (1) ion-exchange with Mn2þ in vacant site in the Mn oxide layer, (2) ion-exchange with Mn2þ at the layer edge, (3) occupation in the vacant sites in the Mn oxide layer, (4) coordination to oxygen atoms at the layer edge with doublesharing surface complex, (5) coordination to oxygen atoms above the vacant sites in the Mn oxide layer with triple-sharing surface complex.

both materials there are satellite peaks around 861.5 eV and 879.5 eV as well as EB [Ni 2p3=2 ] ¼ 856:0 eV and EB [Ni 2p1=2 ] ¼ 873:8 eV (Fig. 11), indicating Ni(II) is the dominant species.20) Nickel ions are simply sorbed on Mn oxides without any redox reactions. Manceau et al.,8) using analysis of XANES spectra, reported Ni2þ ions have a strong preference for isomorphic substitution for Mn in the manganese layer at circumneutral pH. Released Mn2þ ions are more competitive with Ni2þ ions to occupy the same sites on the biogenic Mn oxides, compared with Co2þ ions (Figs. 1(b), 3(b)). The selective sorption of Co2þ over Ni2þ is based on the different sorption mechanisms of Co2þ and Ni2þ . Ion exchange with labile Mn2þ ions were observed for both Co2þ and Ni2þ (Fig. 12(1), (2)), with the preference in the order of Co2þ > Mn2þ > Ni2þ . The Co2þ ions are taken up in vacant sites (Fig. 12(3)) and edge sites (Fig. 12(4), (5)) of

1) S. Groudev, P. Georgiev, I. Spasova and M. Nicolova: Hydrometallurgy 94 (2008) 93–99. 2) B. Lesley, H. Daniel and Y. Paul: Sci. Total Environm. 394 (2008) 1–8. 3) V. A. Drits, E. Silvester, A. I. Gorshkov and A. Manceau: Am. Mineral. 82 (1997) 946–961. 4) E. Silvester, A. Manceau and V. A. Drits: Am. Mineral. 82 (1997) 962– 978. 5) B. Lanson, V. A. Drits, E. J. Silvester and A. Manceau: Am. Mineral. 85 (2000) 826–838. 6) H. Bilinski, R. Giovanoli, A. Usui and D. Hanzel: Am. Mineral. 87 (2002) 580–591. 7) K. Sasaki, M. Matsuda, T. Urata and T. Hirajima: Mater. Trans. 49 (2008) 605–611. 8) A. Manceau, M. Lanson and N. Geoffroy: Geochim. Cosmochim. Acta 71 (2007) 95–128. 9) D. Dong, Y. M. Nelson, L. W. Lion, M. L. Shuler and W. C. Ghiorse: Wat. Res. 34 (2000) 427–436. 10) R. G. Burns: Geochim. Cosmochim. Acta 40 (1976) 95–102. 11) S. B. Kanungo and R. P. Das: Hydrometallurgy 20 (1988) 135–146. 12) G. P. Tindall: PhD Thesis, (Murdoch University, Perth, Western Australia, 1998). 13) J. B. Maynard: Manganiferous Sediments, Rocks, and Ores, in Treatise on Geochemistry, ed. by Heinrich D. Holland and Karl K. Turekian, (Elsevier, Amsterdam, 2007) pp. 289–308. 14) K. Takano, Y. Itoh, T. Ogino, K. Kurosawa and K. Sasaki: Limnology 47 (2006) 2457–2461. 15) J. W. Murray, L. S. Balistrieri and B. Paul: Geochim. Cosmochim. Acta 48 (1984) 1237–1247. 16) D. A. Shirley: Phys. Rev. B 5 (1972) 4709–4714. 17) J. Hem, C. Roberson and C. J. Lind: Geochim. Cosmochim. Acta 49 (1985) 801–810. 18) Y. Tani, N. Miyata, K. Iwahori, M. Soma, S. Tokuda, H. Seyama and B. K. G. Theng: Appl. Geochem. 18 (2003) 1541–1554. 19) G. Limousin, J. P. Gaudet, L. Charlet, S. Szenlnect, V. Barthes and M. Krimissa: Appl. Geochem. 22 (2007) 249–275. 20) M. Ocana and A. R. Gonzalez-Elipe: Colloids and Surfaces A: Physicochem. Eng. Aspects 157 (1999) 315–324. 21) M. Villalobos, B. Lanson, A. Manceau, B. Toner and G. Sposito: Am. Mineral. 91 (2006) 489–502.