The biogeochemical fate of nickel during microbial ISA degradation; implications for nuclear waste disposal

Intermediate level radioactive waste (ILW) generally contains a heterogeneous range of organic and inorganic materials, of which some are encapsulated in cement. Of particular concern are cellulosic waste items, which will chemically degrade under the conditions predicted during waste disposal, forming significant quantities of isosaccharinic acid (ISA), a strongly chelating ligand. ISA therefore has the potential to increase the mobility of a wide range of radionuclides via complex formation, including Ni-63 and Ni-59. Although ISA is known to be metabolized by anaerobic microorganisms, the biodegradation of metal-ISA complexes remains unexplored. This study investigates the fate of a Ni-ISA complex in Fe(III)-reducing enrichment cultures at neutral pH, representative of a microbial community in the subsurface. After initial sorption of Ni onto Fe(III)oxyhydroxides, microbial ISA biodegradation resulted in >90% removal of the remaining Ni from solution when present at 0.1 mM, whereas higher concentrations of Ni proved toxic. The microbial consortium associated with ISA degradation was dominated by close relatives to Clostridia and Geobacter species. Nickel was preferentially immobilized with trace amounts of biogenic amorphous iron sulfides. This study highlights the potential for microbial activity to help remove chelating agents and radionuclides from the groundwater in the subsurface geosphere surrounding a geodisposal facility.


Nickel in solution.
In this study the fate of non-radioactive Ni was explored in ISA-degrading, Fe(III)reducing microbial cultures. Prior to the study, abiotic factors in the system that could control the solubility of Ni were considered, e.g. the presence of Fe(III)oxyhydroxide that exhibits high adsorptive capacities for metal ions 39 . The solubility of Ni (at 0.1 mM Ni and 1 mM Ni) was tested in freshwater minimal medium (FWM) at neutral pH, containing (i) medium only, (ii) medium plus ISA, (iii) medium plus Fe(III) or (iv) medium plus ISA and Fe(III) (Fig. S1). Biosorption of Ni onto cells was not specifically tested, as it is considered to be low 40,41 . In the presence of the medium only, approximately 90-91% Ni remained in solution, whilst in the presence of ISA the solubility of Ni was increased to approximately 100%, regardless of the initial Ni concentration. In the medium that contained Fe(III)oxyhydroxides added as a solid, the solubility of Ni dropped to around 64% at 0.1 mM Ni and to 15% at 1 mM Ni, but was increased by approximately 10% in the medium containing Fe(III) and ISA at 1 mM Ni, but remained unchanged at 0.1 mM Ni.
These initial tests showed that Ni had a great solubility in the medium which was enhanced by ISA, and therefore has the potential to increase radionuclide mobility in groundwater systems. However, the presence of competing Fe(III)oxyhydroxides decreased the solubility of the Ni.

Development of an Ni-ISA biodegrading, Fe(III)-reducing enrichment experiment. Biogeochemistry.
An anaerobic enrichment culture representative of Fe(III)-reducing bacteria, possibly present in the geosphere surrounding a GDF, was developed using ISA complexed with Ni as the electron donor and Fe(III)oxyhydroxide as the electron acceptor. Microbial inocula were obtained from a lime workings site, with high calcium concentrations and alkaline pH values, and were enriched to stable cultures with ISA as the sole carbon source at circumneutral pH. The microbially active incubations contained either no nickel, 0.1 mM Ni or 1 mM Ni. Medium with microbial enrichments without Ni or with 0.1 mM Ni concentration, there was an increase in pH from 7.0 to 7.3 while the E h fell to around −250 mV after approximately 14 days (Fig. 1A,B). This coincided with a drop in ISA levels, which was no longer detected after approximately 28 days (Fig. 1C), while volatile fatty acids (VFAs), comprising acetate and butyrate accumulated and peaked at 28 days (Fig. 1E,F). Thereafter, acetate remained stable in solution during an extended incubation time (62 days), whilst butyrate concentrations decreased after ISA was depleted. Fe(III) reduction, monitored using the ferrozine assay by Fe(II) ingrowth (Fig. 1D), was also initiated during the period of ISA biodegradation, and continued after ISA was depleted until day 42, after which it started to plateau at approximately 26 mmoles L −1 Fe(II). It should be noted that the medium also contained comparatively low levels of sulfate (0.13 mM), which decreased in the presence of 0.1 mM Ni and without Ni, alongside ISA removal, until it was fully depleted at day 42 (Fig. S2). In contrast, in the enrichment cultures supplemented with 1 mM Ni, data for pH, E h and concentrations of ISA, Fe(II), and sulfate all remained stable, and Fe(III) and sulfate were not reduced. Abiotic influences, two control experiments were set up, containing ISA and Fe(III) oxyhydroxide, either with an autoclaved inoculum or without the inoculum added. ISA concentration remained SCiEntifiC REPORTS | (2018) 8:8753 | DOI:10.1038/s41598-018-26963-8 stable in these controls (Fig. 1B), while the pH remained unchanged, the E h showed only a minor drop over the incubation time (Fig. 1A,D) and Fe(II) levels remained constant, all consistent with abiotic influences playing a negligible role in these experiments (Fig. 1E).
Fate of nickel. The concentrations of Ni in the supernatants of these experiments were also monitored by ICP-AES, to help identify the fate of the metal during ISA biodegradation and Fe(III) reduction (Fig. 2). At the start of incubation, soluble Ni concentrations were 64% (64 µM) of the 0.1 mM Ni added, and 12% (115 µM) of the 1 mM Ni added to the microbially-active cultures. These results were in agreement with data from the initial Ni solubility tests and were most likely a result of sorption to the surface of the Fe(III)oxyhydroxides that were added.
After partial precipitation of the Ni (about 90%) in the control incubations (with either autoclaved inoculum or no added inoculum), the remaining soluble Ni concentrations remained stable, confirming that sorption equilibrium or abiotic precipitation were not disturbed under these experimental conditions. Similarly, after initial sorption of about 90%, no decrease of Ni concentrations in solution was observed in the microbially active enrichment culture containing 1 mM Ni, where ISA degradation was absent. In the microbially active 0.1 mM Ni incubation, the Ni concentration in solution was also reduced at the start of the incubation to about 64% of Ni added (presumably due to sorption onto any residual insoluble Fe(III)). However in contrast to the treatments which showed a lack of ISA degradation, the Ni concentration started to decrease from day 14 from 64 µM to 8.5 µM (about 9%) final concentration in solution at the end of the incubation (Fig. 2), and remained at this value throughout an extended incubation time of 200 days (data not shown). Thus, following Ni sorption onto Fe(III) oxyhydroxides (36% removal), the oxidation of ISA and reduction of Fe(III) and sulfate was concomitant with the almost complete removal (90%) of the Ni added.
Molecular Ecology. A stable microbial community was enriched from lime kiln sediment samples over seven consecutive transfers in ISA/Fe(III)-containing medium, with the last transfer used in the Ni-ISA biostimulation experiments discussed above. DNA was extracted from these final cultures at selected time points and the 16S rRNA genes present were amplified and sequenced, to identify changes in microbial community composition that may have controlled the fate of the Fe and Ni. The chosen profiles were from the 0 mM Ni and 0.1 mM Ni experiments at 0 hours, at 28 days, the point at which ISA was depleted, and at 63 days, the point where Fe(III) reduction levelled off (Fig. 3). α-rarefaction curves showed a dramatic decrease in microbial diversity from the original sediment inoculum (approximately 600 distinct sequences detected) 21 to the seventh transfer of the enrichments (approximately 190 distinct sequences detected at the start of this experiment; Fig. S3). A further modest decrease in microbial diversity was noted over the duration of this experiment, whereby the 0 mM Ni and 0.1 mM Ni supplemented enrichments contained approximately 160 distinct sequences. The decrease in community microbial diversity was confirmed by calculating the Shannon H indices (values are shown above the columns in Fig. 3).
Analysis of 16S rRNA gene sequences obtained from the 0.1 mM Ni experiment after ISA depletion (28 days) showed enrichment in organisms most closely affiliated with the Gram-positive Class Clostridia, representing approximately 56% of the total abundance of bacteria detected. This Class comprised mainly of members of the family Ruminococcaceae (53% of genes detected), known to possess a multi-enzyme cellulosome complex that could play a central role in carbohydrate metabolism 42 . Another important Class detected was the Gram-negative Deltaproteobacteria comprising approximately 18% of sequences detected. This Class was dominated by sequences affiliated most closely (97% to 99% match) with the Fe(III)-reducing family Geobacteraceae (16% of sequences detected). Despite representing less than 2% of the sequences, detection of members of the Desulfovibrionaceae was considered potentially important, as these organisms were affiliated with the sulfate-reducing bacteria Desulfovibrio putealis B7-43 (1% of sequences, 98% match) and Desulfovibrio L7 sp. (0.65% of sequences, 100% match). A second 16S rRNA gene profile was obtained from the 0.1 mM Ni experiment when Fe(III) reduction had finished (day 63). This had a higher percentage of sequences affiliated with Betaproteobacteria (26% of sequences), of which most of the sequences (25%) were related to Dechlorosoma suillum PS (100% match) from the genus Azospira, an anaerobic, perchlorate-reducing organism 43 . The second most abundant Class detected was again most closely affiliated with Clostridia (20% of sequences) followed by the Bacteroidia (17% of sequences), of which most sequences were affiliated with an uncultured bacterium from the vadinBC27 wastewater-sludge group (11% of sequences), from the genus Rikenellaceae, which are described as anaerobic, mesophilic carbohydrate-fermenting organisms 44 .  Biomineralogy. To help define the fate of Ni in the ISA-degrading, Fe(III)-reducing cultures, the mineralogical end-products in these experiments were assessed using X-ray Diffraction (XRD), Transmission Electron Microscopy (TEM) and Environmental Scanning Electron Microscopy (ESEM).
At the start of the Ni-ISA biostimulation experiment, a brick-red precipitate was observed, which is likely to be ferrihydrite as detected in XRD analyses (Fig. S4). This starting material remained unchanged in the sterile and no electron donor controls, and also the incubations with 1 mM Ni added. This mineral phase was converted to a dark grey precipitate where ISA metabolism was observed (microbially active experiment without Ni and with 0.1 mM Ni). TEM analysis of the red precipitate in the sterile control showed the presence of fine grained material, while selected area electron diffraction (SAED) revealed two broad rings at about 1.49 Å and 2.5 Å, all consistent with 2-line ferrihydrite that is of irregular, poorly ordered nanoparticulate dimensions (2-4 nm) [45][46][47][48] . The XRD patterns (Fig. S4) obtained for the bio-reduced samples in the absence of Ni and with 0.1 mM Ni added, revealed two sharp peaks at 2Ɵ = 13.2° and at 2Ɵ = 31.9°, corresponding to siderite (FeCO 3 ) and vivianite Fe 3 (PO 4 ) 2 ·8H 2 O (Fig. S4). The relative abundance of vivianite:siderite was very similar at both ISA-degrading treatments of about 40%:60% in the absence of Ni and 35%:65% at 0.1 mM Ni, respectively. Fe(II) in the siderite structure was partially replaced by Ca(II), as previously noted (Kuippers et al., Submitted 2017).
ESEM imaging of the post-reduction precipitates (with 0.1 mM Ni; Fig. 4A,B), confirmed the identity of (1) siderite and (2) [51][52][53][54][55] . Another area was amorphous, with no rings in the diffraction pattern, but consisted of mainly Fe, S and Ni (area 6 in Fig. 4C). Finally, TEM mapping suggested close association of Ni and S areas, whilst P, Ca and O were accumulated in separate locations, and Fe was detected throughout the mapped area (Fig. S5). Additional PHREEQC modelling was carried out to estimate the Ni solubility. Results from the modelling suggested that Ni was undersaturated in solution under anoxic conditions and before reduction (Fig. S6) while after bio-reduction precipitation of millerite and mackinawite was predicted (Fig. S7).

Discussion
Ni is an important constituent in steel alloys used for nuclear reactor components. Irradiation of stable Ni leads to the formation of Ni-63 and Ni-59 activation products, hence there is great interest in understanding the fate of Ni during the (bio)geochemical evolution of a GDF and the surrounding geological environment.
Here we studied the fate of a Ni-ISA complex in circumneutral enrichment cultures containing Fe(III) as the main terminal electron acceptor. It should be noted, however, that to facilitate mineralogical investigations that underpin this study, significantly higher concentrations of Ni were added, than those expected in a GDF environment for ILW, which will range around 3.8 * 10 −7 M within 10 6 years of GDF evolution 9 . In these incubations, ISA biodegradation coincided with Fe(III) reduction, and led to a significant decrease in the concentration of Ni when supplied at 0.1 mM. At 1 mM Ni concentration, no ISA removal occurred, highlighting the strongly inhibitory impact of higher Ni concentrations on microorganisms. In the microbially active experiment at 0.1 mM Ni, fermentation played a significant role as approximately 60% of the carbon from ISA degradation was recovered as acetate and butyrate. Other end-products identified included the crystalline Fe(II) biominerals, siderite (Fe(II)-carbonate) and vivianite (Fe(II)-phosphate). Although ferrihydrite is considered highly bioavailable to Fe(III)-reducing bacteria 56 , the production of fully reduced Fe(II) minerals from this substrate, rather than a mixed Fe(II)/Fe(III) valence iron oxide, such as magnetite, is not commonly observed in laboratory studies 57,58 , but has been previously noted when ISA was supplied as an electron donor 21 . Chelating ligands are known to enhance microbial Fe(III) reduction rates [59][60][61][62] , and since ISA has been shown to complex with Fe(III) minerals 8 , ISA may have increased the solubility and hence bioavailability of ferrihydrite to microbial Fe(III) reduction in our experiments. We conclude therefore, that chelation of ISA with Fe(III) will result in the production of Fe(II) biominerals not normally observed in natural environments.
The negative standard reduction potential of the Ni(II)/Ni(0) couple (E0 = −0.26 V) 63 makes Ni highly resistant to microbially-mediated reduction under pH-neutral conditions (Tsezos et al. 41 ). Therefore, capture of Ni via interaction with bio-mineralogical products is a far more feasible endpoint under anoxic ISA-degrading conditions. In these Fe(III)-reducing experiments, the endpoint minerals included an Fe(II)-carbonate phase (siderite) and Fe(II)-phosphate phase (vivianite). Siderite is known to play a role in trace metal immobilization because cations of similar size to Fe(II), such as Co(II), Cr(III), Ni(II), can be readily incorporated into the crystal structure in place of ferrous iron 64,65 . In these experiments only Ca(II) was incorporated into the siderite structure upon metabolism of Ca(ISA) 2 , whereas Ni(II) was not found to be associated with the biomineral, nor was Ni associated with vivianite. However it should be noted that Fe(II)-bearing minerals, including siderite and vivianite, are known to mediate the reduction of soluble Tc(VII) 62,66,67 or Np(V) 68 to insoluble tetravalent forms. Additionally, vivianite has also been shown to reduce soluble U(VI) to immobile U(IV) 69 . Thus, siderite and vivianite may decrease the mobility of this and other priority radionuclides in the wider geosphere.
Regarding the fate of Ni, EDS profiles from TEM and ESEM images showed that Ni accumulated together with Fe and S in amorphous precipitates, consistent with mackinawite (FeS), although this mineral could not be detected with XRD. Mackinawite is a common metastable iron-sulfide mineral in anoxic sediments and a possible precursor to greigite and pyrite formation 52,70-72 . Furthermore it is known to substitute Ni for Fe, into the sulfide mineral structure 51,73-75 . The identification of Ni-containing iron-sulfide phases in the ISA bio-stimulated experiment, accompanied by an enrichment of Desulfovibrio sp., is consistent with the production of sulfide by sulfate-reducing bacteria that can act as a precursor to precipitate Ni. This is in agreement with PHREEQC modelling that suggested formation of mackinawite (FeS) and millerite (NiS) in these experiments. The PHREEQC database does not include data for mixed phases, such as Fe (1-x) ,Ni x S, and so it predicts the formation of these separate phases. However, since pure Ni-sulfides, such as heazelwoodite (Ni 3 S 2 ) or millerite (NiS), are generally only observed in systems that are low in Fe 75 , and (Fe,Ni)S has a greater stability over pure FeS 70,73,74 , the modelling results can be interpreted as formation of a mixed Fe (1-x) ,Ni x S phase, which is in agreement with TEM and ESEM imaging, following the equation: The immobilization of Ni precipitated as a mixed (Fe,Ni)S phase is thus a result of combined microbial activity by Fe(III)-and sulfate-reducing bacteria. 16S rRNA gene profiling of the enrichment identified key soil microorganisms that may have been involved in the degradation of ISA and thus supported Ni immobilization. The 16S rRNA gene profiles of the 0.1 mM Ni and without Ni experiments were very similar, being dominated by fermenting organisms associated with members from the family Clostridia during ISA degradation (day 28), whilst Betaproteobacteria increased in later stages of the incubation, when Fe(III) reduction became more important (day 63). Although close relatives to well-known Fe(III)-reducing bacteria such as Geobacter species were detected in these cultures, we cannot discount the involvement of other more abundant organisms in mediating this process. Finally, although Ni is an essential trace element for several enzymes [76][77][78][79] , at higher concentrations it poses a toxic threat to microorganisms 80,81 . Such an inhibitory effect of Ni was observed at 1 mM concentration, where no ISA biodegradation (or Fe(III) reduction) was recorded.
In summary, it was shown that the solubility of Ni in ISA-degrading Fe(III)-reducing enrichment cultures was influenced by two dominant factors; the mobilization of Ni by complexation with ISA, and the competing immobilization of Ni by sorption to Fe(III)oxyhydroxides. The overall fate of Ni in the system was defined by microbial metabolism, which was driven by biodegradation of ISA. The combined activity of Fe(III)-and sulfate-reducing microorganisms resulted in >90% Ni incorporation into iron-sulfides, presenting a stable sink for the Ni. Given the high abundance of sulfate as electron acceptor in deep subsurface groundwaters in the UK 82 and elsewhere, the fate of Ni and other radionuclides in the deep geosphere surrounding a GDF, is likely to be governed by sulfidation reactions. These processes will be potentially mediated by Fe(III)-and sulfate-reducing bacteria in the deep subsurface that have been stimulated by the release of organics such as ISA, that may migrate in the alkaline groundwater plume from a cementitious GDF. Since ISA oxidation coupled to Fe(III) and sulfate reduction is energetically unfavorable in the alkaline GDF "near field" 83 , precipitation of Ni and other radionuclides is expected to occur in the geological "far field" of a GDF and in low pH niches of the "near field". It should be noted that also direct abiotic interactions within the geosphere are possible once the strong complexant ISA has been biodegraded, e.g. with clay minerals, known to adsorb cationic Ni due to their negatively charged surfaces, and thus are able to retard the movement of Ni 77,84,85 . Additionally to the trace levels of FeS formed, ISA degradation in our experiments supported dissimilatory Fe(III) reduction that produced the Fe(II) minerals siderite and vivianite. These Fe(II) minerals of are known to form at moderately high pH 86 , and to abiotically mediate the reduction of priority radionuclides to insoluble forms. Thus, it is likely that microbial metabolism in the deep subsurface biosphere surrounding a GDF will play a significant role in minimizing the transport of mobile radionuclide-ISA complexes, through the fermentation of ISA at high pH 31 , and further coupling of biodegradation reactions to the reduction of electron acceptors such as Fe(III) and sulfate in and around the GDF, extending to the circumneutral deep subsurface biosphere. Consequently it is likely that a "bio-barrier" will develop, which could support the containment of priority radionuclides, including Ni-63 and Ni-59, in addition to other physical and chemical barriers. Clearly further research is warranted in this area, including studies to define the impact of microbial metabolism on ISA complexed to other radionuclides, including Am(III), Th(IV) and U(VI). There exists also the need to focus on more realistic subsurface in situ conditions to better quantify the efficiency of metal and radionuclide immobilization upon ISA degradation. Although it is clearly a significant challenge to quantify the impact of microbial metabolism in GDF scenarios, over long time periods within highly complex biogeochemical gradients, it is reasonable to expect that such studies would support a reduction in conservatisms in future GDF safety cases. Ni-ISA biodegrading, Fe(III)-reducing enrichment experiment. Anaerobic Fe(III)-reducing microbial communities were enriched from shallow subsurface sediments that were collected from a site contaminated by legacy lime workings at Harpur Hill, in Buxton, Derbyshire, UK 34 . Cultures were prepared with a 1% (v/v) inoculum added to 100 mL serum bottles containing 30 mL FWM. ISA (as Ca(ISA) 2 ) was added to 4 mM as sole carbon source and 25 mmoles L −2 of poorly soluble Fe(III)oxyhydroxide was added as the terminal electron acceptor (TEA) to support anaerobic growth. The bottles were sealed with butyl stoppers and flushed with a N 2 /CO 2 (80:20) gas mixture for 5 min to provide anaerobic conditions and to adjust a pH of 7 and incubated in the dark at 20 °C. After Fe(III) reduction had reached a plateau, a 1% vol/vol inoculum was transferred to fresh medium. For this study, the seventh consecutive transfer was used to prepare five tests in the medium as above in triplicate: to three incubations containing a microbially active inoculum, ISA and Fe(III)oxyhydroxide, nickel was added at the following concentrations: (i) 0 mM Ni, (ii) 0.1 mM Ni, or (iii) 1 mM Ni, alongside (iv) a sterile microbial inoculum (autoclaved), Geochemical analyses. The pH was monitored throughout the experiment with a calibrated Mettler Toledo FEP20 digital meter equipped with a Fisherbrand FB68801 electrode and the redox potential with a Denver Instrument Accumet digital meter equipped with a Mettler Toledo Inlab Redox Micro ORP. Microbial Fe(III) reduction was monitored by the ferrozine spectrophotometric assay quantifying Fe(II) following the protocol from Stookey 89 . Total bioavailable Fe was determined at the beginning and end of the experiment by digesting 20 µL of sample in 980 µL of 0.5 N HCl and 0.25 N hydroxylamine-HCl, and biogenic Fe(II) was determined by digestion of 20 µL of sample in 980 µL of 0.5 N HCl, followed by the ferrozine assay 90 .

Methods
Ion-Exchange Chromatography. ISA and organic acids were analyzed using ion exchange high performance liquid chromatography (IE-HPLC), using a Dionex ICS5000 Dual Channel on Chromatograph, fitted with a Dionex AS-AP auto sampler that was connected to a CD20 conductivity detector. The columns were a Dionex Capillary (50 × 0.4 mm) AG11-HC 4 µm guard column and Dionex Capillary (250 × 0.4 mm) AS11-HC 4 µm analytical column. A Dionex ACES300 Chemical Suppressor was used for background reduction. The mobile phase was concentrated KOH mixed with high purity water at a flow rate of 0.015 mL/min. Background suppression was hold at 13 mA.
Inductively Coupled Plasma Atomic Emission Spectrometer (ICP-AES). ICP-AES was used to quantify total Ni in solution. Samples were prepared by centrifugation to remove any suspended particles and the supernatants were diluted to below <10 ppm Ni in 2% nitric acid. Samples were analyzed together with standards (from 10 ppm stocks obtained from VWR) on a Perkin-Elmer Optima 5300 DV.  91,92 . PCR was performed in 50 µL reactions, starting with the denaturation at 95 °C for 2 min, followed by 36 cycles of 95 °C for 30 s, 55 °C for 30 s, 72 °C for 1 min, and a final extension step of 5 min at 72 °C. PCR products were cleaned up and normalized to ~20 ng each using the SequalPrep Normalisation Kit (Fisher Scientific, Loughborough, UK). PCR amplicons were pooled in equimolar ratios, using a 4 pM sample library spiked with 4 pM PhiX, to a final concentration of 10% 93 . Sequence analysis was performed using an Agilent BioAnalyzer at the Centre for Musculoskeletal Research, Manchester 93 . Raw sequences were divided into samples by barcodes (up to one mismatch was permitted) using a sequencing pipeline. Quality control and trimming was performed using Cutadapt 94 , FastQC 95 , and Sickle 96 . MiSeq error correction was performed using SPADes 97 . Forward and reverse reads were assembled to full-length sequences with PANDAseq 98 . Using ChimeraSlayer the sequences were screened for chimeras which were removed subsequently 99 . OTU's were generated with UPARSE 100 and classified by USEARCH 101 at the 97% similarity level, and singletons were removed. The taxonomic assignment was performed by the RDP classifier 102 and α-rarefaction curves were generated in Qiime 103 . The raw sequencing data have been deposited at the NCBI Sequence Read Archive (http://www.ncbi. nlm.nih.gov/sra/) with the accession numbers SRR5786557 and SRR5786559 to SRR5786562.
Mineralogical analyses. X-Ray Diffraction (XRD) crystallography. XRD was used for mineral phase identification by centrifugation of sample aliquots and coating wet pellets onto a XRD glass slide which was left to dry anaerobically. Samples were analyzed in an anaerobic holder fitted with an internal knife-edge on a Bruker D8 Advance, fitted with a Gobel Mirror and a Lynxeye detector. Cu Kalpha1 X-rays were used from 5-70 degrees with step size 0.02 degrees and count time of 0.5 sec per step. Eva v14 was used to match crystal patterns to standards from the ICDD (International Centre for Diffraction Data) database and a semi-quantitative analysis of relative crystallite proportions was achieved with Topas v4.2 software.

Transmission Electron Microscopy (TEM) and Environmental Scanning Electron Microscopy (ESEM)
. TEM and ESEM were used to characterize biominerals formed during ISA degradation and to trace the fate of Ni in the solid phase. For TEM analysis, samples were centrifuged and the pellet was added to a TEM copper grid with a holey carbon film coating and let to dry anaerobically. Samples were viewed over a range of magnifications (5,000-200,000x) on a Philips FEI Technai T20 (200 kV LaB6) instrument fitted with an Oxford Instruments X-Max 80 mm 2 SDD EDS system operating Aztec software, and a Gatan Orius SC200 CCD camera operating GMS 2 software. For ESEM imaging sample aliquots were washed in a bicarbonate buffer and dried anaerobically on an aluminum pin stub (Zeiss, Ø12.7 diameter top) before coating with carbon to enhance conductivity. The instrument used was a FEI XL30 ESEM-Field Emission Gun (ESEM-FEG) operating at 15 kV in high vacuum mode (10 −5 to 10 −6 mbar) secondary electron (SE) mode fitted with an EDAX Gemini EDS system for elemental analysis.