Impact of Mn(II)-Manganese Oxide Reactions on Ni and Zn Speciation Margaret A. G. Hinkle*, Katherine G. Dye, and Jeffrey G.

Transcription

1 SUPPORTING INFORMATION FOR Impact of Mn(II)-Manganese Oxide Reactions on Ni and Zn Speciation Margaret A. G. Hinkle*, Katherine G. Dye, and Jeffrey G. Catalano Department of Earth and Planetary Sciences, Washington University, 1 Brookings Drive, Saint Louis, MO USA *Corresponding author and present address: Tel.: ; mhinkle@eps.wustl.edu; Address: Department of Mineral Sciences, Smithsonian Institution, National Museum of Natural History, Washington, DC This Supporting Information document includes 22 pages, 3 figures, & 5 tables: (page #) - SUPPORTING METHODS & MATERIALS: - Mineral and Reagent Preparation 2 - Macroscopic Ni Adsorption Experiments Details of EXAFS Fitting SUPPORTING RESULTS - Synthesized Metal-Free Materials Mn(II)-Ni Competitive Adsorption Processes Discussion of Alternative EXAFS Spectral Fits FIGURES - Figure S1: Ni EXAFS spectra and model fits 9 - Figure S2: Zn EXAFS spectra and model fits 10 - Figure S3: Effect of Mn(II) on Ni(II) adsorption and dissolved Mn 11 versus dissolved Ni for overnight Ni adsorption experiments - TABLES - Table S1: EXAFS fitting parameters for Ni data Table S2: EXAFS fitting parameters for Zn data Table S3: Final and alternate EXAFS fitting parameters for select Ni data - Table S4: Final and alternate EXAFS fitting parameters for 20 select Zn data - Table S5: Ni 24 hour adsorption Langmuir isotherm parameters 21 - SI REFERENCES 22 S1

2 SUPPORTING METHODS AND MATERIALS Mineral and Reagent Preparation The phyllomanganates prepared for this study were synthesized using previously published procedures. 1 The redox methods for synthesizing δ-mno 2 and c-disordered H + birnessite (termed HexB in this paper) described by Villalobos et al. (2003) 2 were modified as previously described. 1 The synthesis used to prepare triclinic birnessite (TriB) was based on the Na-birnessite synthesis described by Lopano et al. (2007). 3 After synthesis, excess electrolytes were washed from the minerals following procedures described elsewhere. 1 Mineral suspensions were stored in polypropylene bottles wrapped in aluminum foil in an anaerobic chamber (Coy Laboratory Products, Inc., 3% H 2 /97% N 2 atmosphere with Pd catalysts). The mineral suspensions were sparged using a secondary O 2 filtration system 4 for a minimum of 48 hours prior to use to remove dissolved O 2. X-ray diffraction (XRD) (Bruker D8 Advance X-ray diffractometers, Cu K α radiation) and surface area determinations by BET N 2 gas adsorption (Quantachrome Instruments Autosorb-1) were performed on samples dried at 70 o C. See Supporting Information for a description of these synthesized materials. NaCl, 2-(4-morpholino)ethanesulfonic acid (MES), NiCl 2 6H 2 O, ZnCl 2, and MnCl 2 4H 2 O stock solutions were made in the anaerobic chamber with deoxygenated deionized water. Experiments conducted at ph 7 used a solution consisting of NaCl and MES preadjusted to ph 7, also prepared within the anaerobic chamber. All ph adjustments were made using HCl or NaOH solutions. Macroscopic Ni(II) Adsorption Experiments A series of Ni(II) adsorption experiments were conducted for δ-mno 2, HexB, and TriB in S2

3 the absence and presence of Mn(II) at ph 4 and ph 7. All experiments were conducted within the anaerobic chamber. Each 10 ml sample consisted of 10 mm NaCl (an ionic strength buffer), 2.5 g L -1 phyllomanganate, 1 mm 2-(4-morpholino)ethanesulfonic acid (MES) buffer (if conducted at ph 7), and mm Ni(II). Samples were prepared at both ph 4 and 7 in the absence or presence of 0.75 mm Mn(II). Triplicate samples were prepared for 0.10 mm Ni(II) to estimate experimental uncertainty. Similar methods as those described in section in the main text regarding the addition of the mineral suspension, ph adjustments, and experiment conditions were followed in these experiments. After reacting overnight (25 hours of reaction for all HexB samples, 26 hours for all δ-mno 2 samples, and 32 hours for all TriB samples), the final ph values of the samples were recorded and the samples were then filtered (0.22 µm MCE filters; Santa Cruz Biotechnology), discarding the first 1 ml of filtrate. If the final sample ph substantially diverged from the target ph (i.e., more than ±0.5 ph units from the target ph), the sample results were not included in the isotherm. Dissolved Ni and Mn concentrations were determined by ICP-OES using the diluted and acidified filtrate. Details of EXAFS Fitting In the model fits, Ni-O coordination numbers (CNs) were fixed at 6, while Zn-O CNs were allowed to vary due to the potential for mixed tetrahedral and octahedral Zn ( IV Zn and VI Zn) coordination states. When split Zn-O shells were required, the CNs for these shells were constrained so that the fractions of Zn in each coordination state summed to one. The σ 2 values for Ni-Mn shells corresponding to incorporated Ni were allowed to vary for Ni-coprecipitated with δ-mno 2, HexB, and TriB. The average of these three σ 2 values was used as a fixed value when fitting the spectra all other samples (Table S1). For double-corner (DC) or triple-corner S3

4 (TC) sharing Ni complexes, σ 2 values of Ni-Mn shell were fixed to the average of the Ni-Mn σ 2 for δ-mno 2, HexB, and TriB reacted with Ni at ph 4 for 40 hours. In the Zn system, incorporated species do not exist but Zn-Mn shells for TC and DC complexes may split when both IV Zn and VI Zn are present, although this splitting may not be resolvable in all cases. When unresolved, the single σ 2 value used in fitting will account for both disorder in individual Zn-Mn interatomic distances and the disorder from the unresolved shell splitting, and will thus be larger than for individual, resolved Zn-Mn shells. Two sets of σ 2 value for split and unsplit TC and DC Zn-Mn shells were thus used, both established from initial fits of two model samples, Zncoprecipitated with δ-mno 2 and δ-mno 2 reacted with Zn at ph 4 for 40 hours (Table S2). These σ 2 values for unsplit and split Zn-Mn shells were applied when fitting the remainder of the spectra. Both Ni-Mn and Zn-Mn shells for tridentate edge (TE) sharing complexes were fixed to 0.01, using the average σ 2 value for Ni adsorbed at edge sites on phyllomanganates by Simanova et al. (2015). 5 In some cases, CNs for Ni-Mn or Zn-Mn shells refine to values larger than 6, indicating that the fixed σ 2 value is too large for that particular sample. However, the interpretation of CNs ~6 (including those greater than 6) reflecting TC Ni or Zn complexes still applies. SUPPORTING RESULTS Properties of Synthesized Phyllomanganates The chemical and physical properties of the metal-free phyllomanganates used for this research are described in detail elsewhere. 1 Briefly, the δ-mno 2 and HexB used in this study are turbostratic phyllomanganates with hexagonal sheet symmetry and high surface areas (116.4 m 2 g -1 and m 2 g -1, respectively), while TriB is a more crystalline phyllomanganate with S4

5 orthogonal sheet symmetry, rotationally ordered layer stacking, and lower surface area (24.8 m 2 g -1 ). The phyllomanganates used in this study were previously characterized by XRD and XAFS spectroscopy, 1 confirming that these materials are consistent with the expected products of the published synthesis methods. The negative layer charges in δ-mno2, HexB, and TriB arise predominantly from vacancies, a mixture of vacancies and Mn(III) substitutions, and only Mn(III) substitution, respectively. 2,3 Mn(II)-Ni Competitive Adsorption Processes To evaluate possible Mn(II) and metals competition for phyllomanganate surface sites under the conditions used in the aging studies described below, Ni adsorption was examined in the absence and presence of Mn(II) in a series of overnight uptake experiments. Macroscopic Ni adsorption (Figure S3) and corresponding Langmuir isotherm fit parameters (Table S5) indicate that 0.75 mm Mn(II) does not substantially alter macroscopic Ni adsorption. Ni adsorption onto δ-mno 2 is slightly suppressed upon addition of 0.75 mm Mn(II) at ph 4, but not at the low Ni concentrations comparable to those examined in the 25-day experiments (Figure S3A). Mn(II) has little effect on Ni adsorption onto HexB and TriB at ph 4 (Figure S3B,C). At ph 7, Mn(II) has a negligible effect on Ni adsorption onto both δ-mno 2 and TriB (Figure S3A,C). As with δ-mno 2 and TriB, Ni adsorption onto HexB at ph 7 below 0.4 mmol g -1 Ni surface loadings is substantial, with nearly all Ni adsorbing in both Mn(II)-free and 0.75 mm Mn(II) systems. With Ni loadings >0.4 mmol g -1, Ni adsorption in the ph 7 HexB system is unexpectedly greater in the presence of 0.75 mm Mn(II) (Figure S3B). The origin of this effect is unclear, and it was not further investigated because it occurs only at elevated dissolved Ni concentrations. Mn(II) uptake by all solids was substantial at ph 7, as well as at lower Ni loadings at ph 4 for δ-mno 2 and S5

6 HexB (Figure S3D-F). TriB showed substantial Mn(II) release to solution at ph 4 in the absence of added Mn(II), even at low Ni loadings. This is also observed at ph 4 for high Ni loadings on Hex B. In systems with added Mn(II), dissolved concentrations increased at high Ni loadings for all phyllomanganates at ph 4, with final concentrations exceeding the amount of Mn(II) added for HexB and TriB. These Mn(II) concentration profiles indicate that substantial structural Mn(III) undergoes disproportionation at ph 4 and that this is further promoted by metal adsorption. Discussion of Alternative EXAFS Spectral Fits In this section, several alternative EXAFS spectral fits are assessed, with the best, most justified fits presented in the main body of the paper. Ni reacted for 40 hours at ph 7 with δ-mno 2 is best described by both Ni incorporation and TC Ni (Table S1). In the HexB system, however, the EXAFS spectrum is best fit by a model consisting of only TC Ni, with no incorporated Ni (Table S3). Attempts to fit the data with both TC and incorporated Ni increased the χ 2 ν values (Table S3), thus the inclusion of incorporated Ni into the 7HBad sample cannot be statistically justified. For Ni reacted for 25 days at ph 4 with HexB and 7.5 mm Mn(II), the addition of a Ni- Mn shell corresponding to incorporated Ni reduces the χ 2 ν relative to a fit without any incorporated Ni (Table S3). This suggests that Ni is incorporated into the HexB structure upon aging for 25 days at ph 4 with high added Mn(II). In contrast, the assessment of alternative fits demonstrates that Ni reacted at ph 4 with TriB and 0.75 mm Mn(II) does not contain detectable incorporated Ni. When a ~5.0 Å Ni-Mn incor shell indicative of incorporated Ni is included in the fit for this sample, the CN is within error of 1, the uncertainty in the fitted distance is large (5 ± S6

7 7), and the χ 2 ν is greater than for a fit that does not include a contribution for incorporated Ni (Table S3). Several alternative fits are considered for samples reacted at ph 7 for 25 days. For Ni reacted with δ-mno 2 and 0.75 mm Mn(II), the Fourier transform (Figure 2C) indicates that incorporated Ni decreases, as the feature associated with this binding mode decreases in amplitude in the presence of Mn(II). To determine if the observed dampening of the Ni-Mn incor shell is due to a decrease in the amount of incorporated Ni or an increase in structural disorder, an alternative fit was produced allowing the Ni-Mn incor σ 2 to vary (Table S3). Surprisingly, this alternative fit resulted in lower Ni-Mn incor CN and σ 2 values (Table S3) (σ 2 should increase with increasing structural disorder). It is therefore likely that the observed reduction in the Ni-Mn incor amplitude upon addition of 0.75 mm Mn(II) to δ-mno 2 is, in fact, due to a decrease in incorporated Ni. Unlike the δ-mno 2 system, Mn(II) does not reduce Ni incorporation into HexB at ph 7. The appearance of a peak at ~2.4 Å (R + R) in Fourier transforms upon aging HexB for 25 days at ph 7 (Figure 2F) suggests the presence of some incorporated Ni. Attempts to fit the ph 7 25-day Mn(II)-free sample spectrum with incorporated Ni-Mn shells were unsuccessful, resulting in higher χ 2 ν values (Table S3). However, including the Ni-Mn incor shell in the 0.75 mm Mn(II) ph 7 HexB sample fit resulted in a statistically significant improvement (Table S3). Because the Fourier transform spectra (Figure 2F) for the 0 and 0.75 mm Mn(II) Ni HexB samples are very similar in the 2.4 Å region (corresponding to the Ni-Mn incor shell), it is likely that a small amount of Ni, near the limits of detection by EXAFS spectroscopy, is incorporated into HexB at ph 7 both in the absence and presence of Mn(II). S7

8 For Zn reacted at ph 7 with δ-mno 2, substantial decreases in the TC/DC Zn-Mn shell amplitudes are apparent in the Fourier transforms relative to the 40 hour adsorbed standard (Figure 3C). However, the CNs for IV Zn- and VI Zn-Mn shells sum to , suggesting that the observed decrease in the Zn-Mn DC/TC peak may not entirely be due to a transition from TC to DC adsorbed Zn. Alternatively, an increase in structural disorder or interference from the two IV/VI Zn contributions could also result in the observed changes to the Fourier transform data. Fitting these Zn δ-mno 2 ph 7 spectra when allowing Zn-Mn TC/DC σ 2 to vary (to test if structural disorder could explain the Fourier transforms) resulted in poorer fits compared to those with a constrained Zn-Mn TC/DC σ 2 (Table S4). With 0 mm Mn(II), σ 2 and CN values increase, but with 0.75 mm Mn(II), σ 2 and CN values slightly decrease (Table S4), but neither fit is a statistically significant improvement over the fit with a constrained Zn-Mn TC/DC σ 2. Thus, these spectra likely display interference from overlapping neighboring atomic shell associated with IV Zn and VI Zn species. S8

9 Figure S1. Ni EXAFS spectra (A,D,G), Fourier transforms (B,E,H) and real part of the Fourier transforms (C,F,I), with data as points, model fits as red lines for Ni reacted with δ-mno 2 (A,B,C), HexB (D,E,F), and TriB (G,H,I) (see Tables 1 and 2 for specific conditions). Diagnostic features at 2.5, 2.7 and 3.1 Å in the Fourier transform spectra, corresponding to Ni-Mn shells for incorporated Ni at 2.88 Å, TE Ni at 3.08 Å, and DC/TC Ni at 3.48 Å, respectively, are denoted by blue lines. S9

10 Figure S2. Zn EXAFS spectra (A,D,G), Fourier transforms (B,E,H) and real part of the Fourier transforms (C,F,I), with data as points, model fits as red lines, of Zn reacted with δ-mno 2 (A,B,C), HexB (D,E,F), and TriB (G,H,I) (see Tables 1 and 2 for specific conditions). Diagnostic features at 6.1 Å -1 in k space (sensitive to Zn coordination) and at 2.7 Å and 3.1 Å in the Fourier transform spectra (Zn-Mn shell for TE Zn at 3.1 Å, and Zn-Mn shell for DC/TC Zn at 3.48 Å, respectively) are denoted by blue lines. S10

11 Figure S3. Effect of Mn(II) on Ni(II) adsorption overnight onto δ-mno 2 (A), HexB (B), and TriB (C) at ph 4 or 7, and amount of Mn in solution as a function of dissolved Ni on δ-mno 2 (D), HexB (E), and TriB (F). Lines in A, B, and C represent Langmuir isotherm fits to the data (Table S5). Horizontal dashed lines in E and F correspond to the initial amount of added Mn(II) in the experiments conducted with 0.75 mm Mn(II). S11

12 Table S1. Ni EXAFS spectra fitting parameters. Sample Shell N a R (Å) b σ 2 (Å 2 ) c E 0 (ev) d 2 e χ ν Ni Coprecipitated δ-mno 2 Ni-O (5) f (4) 2.5(6) 6.38 co- δ Ni-Mn incor1 3.1(5) 2.895(6) 0.008(1) Ni-Mn DC/TC1 3.8(4) 3.492(9) Ni-Mn incor (2) Ni-Mn DC/TC (3) Ni-Mn incor (3) HexBirn Ni-O (5) (5) 2.1(7) 5.66 co-hb Ni-Mn incor1 3.2(5) 2.858(5) 0.005(1) Ni-Mn DC/TC1 5.5(5) 3.480(8) Ni-Mn incor (2) Ni-Mn DC/TC (2) Ni-Mn incor (3) TriBirn Ni-O (7) (5) 1(1) co-tb Ni-Mn incor1 3.4(8) 2.881(8) 0.007(2) Ni-Mn DC/TC1 2.6(6) 3.49(2) Ni-Mn incor (3) Ni-Mn DC/TC (7) Ni-Mn incor (5) Ni 40 hr Rxn ph 4 δ-mno 2 Ni-O (5) (4) 3.0(7) δad Ni-Mn DC/TC1 8(1) 3.496(6) 0.009(1) Ni-Mn DC/TC (1) HexBirn Ni-O (4) (3) 2.3(6) HBad Ni-Mn DC/TC1 7.3(9) 3.488(6) 0.010(1) Ni-Mn DC/TC (1) TriBirn Ni-O (5) (4) 2.4(7) TBad Ni-Mn incor1 0.8(2) 2.88(1) Ni-Mn DC/TC 5.0(8) 3.489(6) 0.007(1) Ni-Mn incor (6) Ni-Mn DC/TC (2) Ni-Mn incor (9) Ni 40 hr Rxn ph 7 δ-mno 2 Ni-O (7) (4) 2.4(7) δad Ni-Mn incor1 1.9(2) 2.861(7) Ni-Mn DC/TC1 6.8(4) 3.483(6) Ni-Mn incor (3) Ni-Mn DC/TC (2) Ni-Mn incor (5) HexBirn Ni-O (7) (5) 2(1) HBad Ni-Mn DC/TC1 6.9(6) 3.48(1) Ni-Mn DC/TC (3) TriBirn Ni-O (5) (3) 2.3(8) TBad Ni-Mn TE 1.6(3) 3.06(1) 0.01 Ni-Mn DC/TC 1.4(4) 3.52(2) Ni 25 day Rxn ph 4 δ-mno 2 Ni-O (5) (4) 2.7(7)

13 Mn(II)-free Ni-Mn DC/TC1 8.4(4) 3.492(6) δno Ni-Mn DC/TC (1) δ-mno 2 Ni-O (5) (4) 2.6(7) mm Mn(II) Ni-Mn DC/TC1 8.3(4) 3.500(6) δlo Ni-Mn DC/TC (1) δ-mno mm Mn(II) Ni-O (6) (4) 2(1) δhi Ni-Mn DC/TC 2.8(4) 3.48(1) Ni-Mn DC/TC (4) HexBirn Ni-O (5) (3) 2.2(8) 4.41 Mn(II)-free Ni-Mn DC/TC1 6.4(4) 3.486(7) HBno Ni-Mn DC/TC (2) HexBirn Ni-O (5) (5) 2(1) mm Mn(II) Ni-Mn DC/TC1 4.4(4) 3.47(1) HBlo Ni-Mn DC/TC (3) HexBirn Ni-O (4) (3) 2.6(7) mm Mn(II) Ni-Mn incor1 0.4(1) 2.87(3) HBhi Ni-Mn DC/TC1 2.9(3) 3.473(9) Ni-Mn incor (2) Ni-Mn DC/TC (4) Ni-Mn incor (2) TriBirn Ni-O (7) (5) 2(1) Mn(II)-free Ni-Mn incor1 0.7(2) 2.88(2) TBno Ni-Mn DC/TC1 6.2(5) 3.481(9) Ni-Mn incor (1) Ni-Mn DC/TC (2) Ni-Mn incor (2) TriBirn Ni-O (6) (4) 2.4(8) mm Mn(II) Ni-Mn DC/TC1 6.9(4) 3.494(7) TBlo Ni-Mn DC/TC (2) TriBirn Ni-O (1) (8) 2(1) mm Mn(II) Ni-Mn incor1 0.8(3) 2.88(3) TBhi Ni-Mn DC/TC1 6.7(7) 3.47(1) Ni-Mn incor (1) Ni-Mn DC/TC (3) Ni-Mn incor (2) Ni 25 day Rxn ph 7 δ-mno 2 Ni-O (6) (5) 2.5(8) 9.23 Mn(II)-free Ni-Mn incor1 2.0(2) 2.871(9) δno Ni-Mn DC/TC1 6.4(5) 3.489(8) Ni-Mn incor (4) Ni-Mn DC/TC (2) Ni-Mn incor (6) δ-mno 2 Ni-O (4) (4) 2.7(6) mm Mn(II) Ni-Mn incor1 1.3(2) 2.863(8) δlo Ni-Mn DC/TC1 7.1(3) 3.484(5) Ni-Mn incor (4)

14 Ni-Mn DC/TC (1) Ni-Mn incor (6) HexBirn Ni-O (5) (4) 2.2(8) 4.41 Mn(II)-free Ni-Mn DC/TC1 6.4(4) 3.486(7) HBno Ni-Mn DC/TC (2) HexBirn Ni-O (4) (3) 2.6(7) mm Mn(II) Ni-Mn incor1 0.7(1) 2.88(2) HBlo Ni-Mn DC/TC1 5.1(3) 3.482(6) Ni-Mn incor (7) Ni-Mn DC/TC (2) Ni-Mn incor (1) TriBirn Ni-O (6) (4) 3.1(9) 7.79 Mn(II)-free Ni-Mn TE 2.7(3) 3.06(1) TBno Ni-Mn DC/TC 1.0(5) 3.59(3) TriBirn Ni-O (6) (4) 2.9(9) mm Mn(II) Ni-Mn TE 1.9(3) 3.06(1) TBlo a Coordination Number. b Interatomic distance. c Debye-Waller factor. d Difference in the threshold Fermi level between data and theory. e Goodness of fit parameters. 6 f Statistical uncertainties at the 68% confidence level are reported in parentheses. Any parameters without reported uncertainties were not allowed to vary during fitting. h Ni-Mn shell names correspond to the type of site, with incor referring to Ni incorporated into the phyllomanganate sheet, DC/TC referring to either double-corner sharing Ni bound at edge sites ( DC ) or triple-corner sharing Ni bound above vacancies ( TC ), which have similar Ni-Mn interatomic distances, and TE referring to tridentate-edge sharing Ni complexes. 14

15 Table S2. Zn EXAFS spectra fitting parameters. Sample Shell N a R (Å) b σ 2 (Å 2 ) c E 0 (ev) d 2 e χ ν Zn-Coprecipitated δ-mno IV 2 Zn-O 2.6(3) f 1.97(2) 0.005(1) 3(1) 8.72 co-δ Zn-O (2) g h Zn-Mn DC/TC1 2.2(5) 3.35(2) 0.007(3) VI Zn-Mn DC/TC1 3(1) 3.51(2) IV Zn-Mn DC/TC (3) VI Zn-Mn DC/TC (3) HexB co-hb IV Zn-O 3.0(3) 1.97(2) 0.005(1) 5(2) 9.78 Zn-O (4) (1) Zn-Mn DC/TC1 1.7(4) 3.32(3) VI Zn-Mn DC/TC1 1.8(4) 3.47(3) IV Zn-Mn DC/TC (4) VI Zn-Mn DC/TC (4) Zn 40 hr Rxn ph 4 δ-mno 2 Zn-O 6.4(6) 2.075(7) 0.008(1) 4.1(9) δad Zn-Mn DC/TC1 8(1) 3.521(8) 0.010(1) Zn-Mn DC/TC (2) HexBirn Zn-O 6.8(4) 2.079(5) (7) 5.6(7) HBad Zn-Mn DC/TC1 2.7(3) 3.512(9) 0.01 Zn-Mn DC/TC (3) 0.01 TriBirn Zn-O 6.4(8) 2.07(1) 0.011(2) 6(1) TBad Zn-Mn DC/TC1 6.0(5) 3.50(1) 0.01 Zn-Mn DC/TC (3) 0.01 Zn 40 hr Rxn ph 7 δ-mno 2 Zn-O 6.1(7) 2.051(8) 0.011(1) 3(1) δad IV Zn-Mn DC/TC1 2.0(5) 3.35(3) VI Zn-Mn DC/TC1 4.9(6) 3.50(1) IV Zn-Mn DC/TC (4) VI Zn-Mn DC/TC (2) HexBirn 7HBad IV Zn-O 2.4(2) 1.96(2) 0.003(1) 4(1) 3.95 VI Zn-O (2) Zn-Mn DC/TC1 1.8(3) 3.33(2) VI Zn-Mn DC/TC1 2.3(4) 3.49(2) IV Zn-Mn DC/TC (3) VI Zn-Mn DC/TC (3) TriBirn Zn-O 6.9(5) 2.057(6) (9) 3.0(9) TBad Zn-Mn TE 1.6(3) 3.08(1) 0.01 Zn-Mn DC/TC1 0.8(4) 3.32(3) 0.01 Zn-Mn DC/TC (8) 0.01 Zn 25 day Rxn ph 4 δ-mno 2 Zn-O 6.5(6) 2.080(6) 0.008(1) 4.5(8) mm Mn(II) Zn-Mn DC/TC1 7.5(5) 3.524(8) δlo Zn-Mn DC/TC (2) 0.01 δ-mno 2 IV Zn-O 2.4(2) 1.96(1) (9) 3(1)

16 7.5 mm Mn(II) 4δhi VI Zn-O (1) Zn-Mn DC/TC1 1.6(2) 3.27(2) VI Zn-Mn DC/TC1 2.3(3) 3.43(1) IV Zn-Mn DC/TC (3) VI Zn-Mn DC/TC (2) HexBirn Zn-O 7.0(5) 2.073(6) (9) 4.8(8) mm Mn(II) Zn-Mn DC/TC1 1.3(3) 3.47(2) HBlo Zn-Mn DC/TC (6) 0.01 HexBirn Zn-O 6.9(6) 2.033(8) 0.012(1) 2(1) mm Mn(II) IV Zn-Mn DC/TC1 1.6(4) 3.27(2) HBhi VI Zn-Mn DC/TC1 1.8(4) 3.43(2) IV Zn-Mn DC/TC (4) VI Zn-Mn DC/TC (4) TriBirn 0.75 mm Mn(II) 4TBlo TriBirn 7.5 mm Mn(II) 4TBhi IV Zn-O 2.3(2) 1.96(2) 0.002(1) 2(1) VI Zn-O (2) Zn-Mn DC/TC1 2.2(5) 3.32(2) VI Zn-Mn DC/TC1 4.9(5) 3.48(1) IV Zn-Mn DC/TC (4) VI Zn-Mn DC/TC (2) IV Zn-O 2.4(2) 1.96(1) 0.004(1) 5(1) VI Zn-O (2) Zn-Mn DC/TC1 2.5(4) 3.33(2) VI Zn-Mn DC/TC1 2.7(4) 3.48(2) IV Zn-Mn DC/TC (3) VI Zn-Mn DC/TC (2) Zn 25 day Rxn ph 7 δ-mno 2 Mn(II)-free 7δno δ-mno mm Mn(II) 7δlo HexBirn Mn(II)-free 7HBno HexBirn 0.75 mm Mn(II) 7HBlo IV Zn-O 2.4(1) 1.95(1) (9) 3(1) 4.84 Zn-O (1) Zn-Mn DC/TC1 1.7(4) 3.30(2) VI Zn-Mn DC/TC1 3.6(4) 3.48(1) IV Zn-Mn DC/TC (5) VI Zn-Mn DC/TC (2) IV Zn-O 2.2(2) 1.95(2) 0.004(1) 3(1) 6.73 VI Zn-O (1) Zn-Mn DC/TC1 2.2(3) 3.31(2) VI Zn-Mn DC/TC1 3.4(3) 3.48(1) IV Zn-Mn DC/TC (3) VI Zn-Mn DC/TC (2) IV Zn-O 2.61(8) 1.98(1) (7) 6(1) 2.16 VI Zn-O (1) Zn-Mn DC/TC1 1.8(3) 3.32(2) VI Zn-Mn DC/TC1 2.5(4) 3.49(2) IV Zn-Mn DC/TC (4) VI Zn-Mn DC/TC (3) IV Zn-O 2.7(2) 1.98(1) (9) 5(1) 9.48 VI Zn-O (2) IV Zn-Mn DC/TC1 1.5(3) 3.29(2) VI Zn-Mn DC/TC1 2.1(4) 3.45(2)

17 IV Zn-Mn DC/TC (4) VI Zn-Mn DC/TC (3) TriBirn Zn-O 6.3(7) 2.05(1) 0.012(2) 3(1) Mn(II)-free Zn-Mn TE 0.9(5) 3.17(3) TBno Zn-Mn DC/TC2 2.1(5) 3.43(2) 0.01 Zn-Mn DC/TC (5) 0.01 TriBirn Zn-O 6.5(6) 2.067(7) 0.010(1) 4(1) mm Mn(II) Zn-Mn TE 2.5(4) 3.08(1) TBlo Zn-Mn DC/TC1 2.0(5) 3.31(2) 0.01 Zn-Mn DC/TC (5) 0.01 a Coordination Number. Interatomic distance. Debye-Waller factor. d Difference in the threshold Fermi level between data and theory. e Goodness of fit parameters. 6 f Statistical uncertainties at the 68% confidence level are reported in parentheses. Any parameters without reported uncertainties were not allowed to vary during fitting. g σ 2 fixed for all linked Zn-O and Zn-Mn shells, denoted by those with no uncertainties reported. h Zn-Mn shell names correspond to the type of site, with TE referring to tridentate-edge sharing Zn complexes, and DC/TC referring to either double-corner sharing Zn bound at edge sites ( DC ) or triple-corner sharing Zn bound above vacancies ( TC ), which have similar Zn-Mn interatomic distances. 17

18 Table S3. Alternate fitting results for select Ni EXAFS spectra. Sample Shell N a R (Å) b σ 2 (Å 2 ) c E 0 (ev) d 2 e χ ν Ni 40 hr Rxn ph 7 HexBirn Ni-O (7) f (5) 2(1) 9.38 No Ni-Mn incor Ni-Mn DC/TC1 6.9(6) 3.48(1) g Ni-Mn DC/TC (3) HexBirn Ni-O (8) (6) 2(1) Ni-Mn incor included Ni-Mn incor1 0.4(3) 2.86(5) Ni-Mn DC/TC1 6.9(7) 3.48(1) Ni-Mn incor (2) Ni-Mn DC/TC (3) Ni-Mn incor (4) Ni 25 day Rxn ph 4 HexBirn Ni-O (4) (3) 2.6(7) mm Mn(II) Ni-Mn incor1 0.4(1) 2.87(3) Ni-Mn incor included Ni-Mn DC/TC1 2.9(3) 3.473(9) Ni-Mn incor (2) Ni-Mn DC/TC (4) Ni-Mn incor (2) HexBirn Ni-O (5) (3) 2.7(7) mm Mn(II) Ni-Mn DC/TC1 2.8(3) 3.47(1) No Ni-Mn incor Ni-Mn DC/TC (3) TriBirn Ni-O (6) (5) 2.3(9) mm Mn(II) Ni-Mn incor1 0.2(2) 2.88(7) Ni-Mn incor included Ni-Mn DC/TC1 6.9(5) 3.494(8) Ni-Mn incor (7) Ni-Mn DC/TC (3) Ni-Mn incor (1) TriBirn Ni-O (6) (4) 2.4(8) mm Mn(II) Ni-Mn DC/TC1 6.9(4) 3.494(7) No Ni-Mn incor Ni-Mn DC/TC (2) Ni 25 day Rxn ph 7 δ-mno 2 Ni-O (4) (4) 2.7(6) mm Mn(II) Ni-Mn incor1 1.3(2) 2.863(8) Ni-Mn incor sig2 fixed Ni-Mn DC/TC1 7.1(3) 3.484(5) Ni-Mn incor (4) Ni-Mn DC/TC (1) Ni-Mn incor (6) δ-mno 2 Ni-O (4) (4) 2.7(6) mm Mn(II) Ni-Mn incor1 1.1(4) 2.860(8) 0.005(2) Ni-Mn incor sig2 Ni-Mn DC/TC1 7.1(3) 3.483(5) unconstrained Ni-Mn incor (4) Ni-Mn DC/TC (2) Ni-Mn incor (6)

19 HexBirn Ni-O (5) (4) 2.0(7) 4.48 Mn(II)-free Ni-Mn incor1 0.3(2) 2.86(3) Ni-Mn incor included Ni-Mn DC/TC1 6.4(4) 3.485(7) Ni-Mn incor (2) Ni-Mn DC/TC (2) Ni-Mn incor (3) HexBirn Ni-O (5) (4) 2.2(8) 4.41 Mn(II)-free Ni-Mn DC/TC1 6.4(4) 3.486(7) No Ni-Mn incor Ni-Mn DC/TC (2) HexBirn Ni-O (4) (3) 2.6(7) mm Mn(II) Ni-Mn incor1 0.7(1) 2.88(2) Ni-Mn incor included Ni-Mn DC/TC1 5.1(3) 3.482(6) Ni-Mn incor (7) Ni-Mn DC/TC (2) Ni-Mn incor (1) HexBirn Ni-O (6) (5) 3(1) mm Mn(II) Ni-Mn DC/TC1 5.1(4) 3.484(9) No Ni-Mn incor Ni-Mn DC/TC (2) a Coordination Number. b Interatomic distance. c Debye-Waller factor. d Difference in the threshold Fermi level between data and theory. e Goodness of fit parameters. 6 f Statistical uncertainties at the 68% confidence level are reported in parentheses. Any parameters without reported uncertainties were not allowed to vary during fitting. g σ 2 fixed for all linked Zn-O and Zn-Mn shells, denoted by those with no uncertainties reported. h Ni-Mn shell names correspond to the type of site, with incor referring to Ni incorporated into the phyllomanganate sheet and DC/TC referring to either double-corner sharing Ni bound at edge sites ( DC ) or triple-corner sharing Ni bound above vacancies ( TC ), which have similar Ni-Mn interatomic distances. 19

20 Table S4. Alternate fitting results for select Zn EXAFS spectra. Sample Shell N a R (Å) b σ 2 (Å 2 ) c E 0 (ev) d 2 e χ ν Zn 25 day Rxn ph 7 δ-mno 2 Zn-O 2.4(1) f 1.95(1) (9) 3(1) mm Mn(II) Zn-O (1) g Zn-Mn DC/TCr sig2 fixed h Zn-Mn DC/TC1 1.7(4) 3.30(2) VI Zn-Mn DC/TC1 3.6(4) 3.48(1) IV Zn-Mn DC/TC (5) VI Zn-Mn DC/TC (2) δ-mno mm Mn(II) Zn-Mn DC/TCr sig2 unconstrained IV Zn-O 2.4(1) 1.95(1) (9) 3(1) 4.65 VI Zn-O (1) Zn-Mn DC/TC1 2.1(7) 3.28(3) 0.010(3) VI Zn-Mn DC/TC1 5(2) 3.47(2) IV Zn-Mn DC/TC (3) VI Zn-Mn DC/TC (1) δ-mno mm Mn(II) Zn-Mn DC/TCr sig2 fixed IV Zn-O 2.2(2) 1.95(2) 0.004(1) 3(1) 6.73 VI Zn-O (1) IV Zn-Mn DC/TC1 2.2(3) 3.31(2) VI Zn-Mn DC/TC1 3.4(3) 3.48(1) IV Zn-Mn DC/TC (3) VI Zn-Mn DC/TC (2) δ-mno mm Mn(II) Zn-Mn DC/TCr sig2 unconstrained IV Zn-O 2.2(2) 1.95(2) 0.004(1) 3(1) 7.12 VI Zn-O (1) IV Zn-Mn DC/TC1 2.1(4) 3.32(2) 0.006(2) VI Zn-Mn DC/TC1 3.1(8) 3.48(1) IV Zn-Mn DC/TC (3) VI Zn-Mn DC/TC (2) a Coordination Number. b Interatomic distance. c Debye-Waller factor. d Difference in the threshold Fermi level between data and theory. e Goodness of fit parameters. 6 f Statistical uncertainties at the 68% confidence level are reported in parentheses. Any parameters without reported uncertainties were not allowed to vary during fitting. g σ 2 fixed for all linked Zn-O and Zn-Mn shells, denoted by those with no uncertainties reported. h Zn-Mn shell names correspond to the type of site, with DC/TC referring to either doublecorner sharing Zn bound at edge sites ( DC ) or triple-corner sharing Zn bound above vacancies ( TC ), which have similar Zn-Mn interatomic distances. 20

21 Table S5. Langmuir isotherm parameters for 24-hour Ni(II) adsorption onto Mn oxides. Sample Γ max (mmol g -1 ) K (ml mmol -1 ) ph 4 δ-mno 2, Mn(II)-free 1.02 ± ± 800 δ-mno 2, 0.75 mm Mn(II) 0.84 ± ± 260 HexB, Mn(II)-free 0.57 ± ± 9 HexB, 0.75 mm Mn(II) 0.52 ± ± 2 TriB, Mn(II)-free 1.9 ± ± 0.09 TriB, 0.75 mm Mn(II) 1.3 ± ± 0.2 ph 7 δ-mno 2, Mn(II)-free 1.6 ± ± 800 δ-mno 2, 0.75 mm Mn(II) 1.6 ± ± 840 HexB, Mn(II)-free 1.3 ± ± 9 HexB, 0.75 mm Mn(II) 1.5 ± ± 80 TriB, Mn(II)-free 2.7 ± ± 30 TriB, 0.75 mm Mn(II) 2.8 ± ± 40 21

22 REFERENCES (1) Hinkle, M. A. G.; Flynn, E.; and Catalano, J. G. Effect of Mn(II) on manganese oxide sheet structures. Geochim. Cosmochim. Acta. 2016, 192, (2) Villalobos, M.; Toner, B.; Bargar, J. R.; and Sposito, G. Characterization of the manganese oxide produced by pseudomonas putida strain MnB1. Geochim. Cosmochim. Acta. 2003, 67, (3) Lopano, C. L.; Heaney, P. J.; Post, J. E.; Hanson, J.; and Komarneni, S. Time-resolved structural analysis of K- and Ba-exchange reactions with synthetic Na-birnessite using synchrotron X-ray diffraction. Am. Mineral. 2007, 92, (4) Hinkle, M. A. G.; Wang, Z.; Giammar, D. E.; and Catalano, J. G. Interaction of Fe(II) with phosphate and sulfate on iron oxide surfaces. Geochim. Cosmochim. Acta. 2015, 158, (5) Simanova, A. A.; Kwon, K. D.; Bone, S. E.; Bargar, J. R.; Refson, K.; Sposito, G.; and Peña, J. Probing the sorption reactivity of the edge surfaces in birnessite nanoparticles using nickel(ii). Geochim. Cosmochim. Acta. 2015, 164, (6) Kelly, S. D.; Hesterberg, D.; and Ravel, B. In Methods of soil analysis. Part 5 - Mineralogical Methods; Drees, L. R. and Ulery, A. L., Editor.; Soil Science Society of America: Madison, WI, 2008; p