Refer to Supplemental File, "57Fe Wall Loss Experiment diagram" for the steps indicated in bold.

All materials were soaked overnight with heating in 1.5% Citrad Citric Acid liquid cleaner in deionized water, rinsed in RO water, and soaked in 10% HCl in Milli-Q ultrapure water for one week (due to time constraints), then rinsed with MilliQ ultrapure water, let dry in an AirClean 400 work station overnight, and double-bagged in polyethylene bags (Mellett et al., 2017). M9 Minimal Media for bacterial cultures was made using Milli-Q ultrapure water, containing final concentrations of 33.7 mM Na₂HPO₄·2H₂O, 22 mM KH₂PO₄, 8.56 mM NaCl, 18.7 mM NH₄Cl, 0.1 M magnesium chloride, 0.1 M calcium chloride, 2 mg/ml Thiamine HCl in 70% EtOH, and 20% Glucose (Kutter & Sulakvelidze, 2004). Half of the media was chelexed using Chelex-100 resin (Pai et al., 1988) that was not acid-cleaned (C), and half remained un-chelexed (B). Half of each treatment was spiked with 10 µM labeled ⁵⁷FeSO₄(2) while half remained un-spiked (1). A volume of 25 ml was then filtered through either a 0.2 µm Sterivex PVDF syringe filter for the dissolved fraction (D), or a 0.02 µm Whatman Anatop syringe filter for the soluble fraction (S) (Gledhill & Buck, 2012). Samples were directly filtered into trace metal clean polycarbonate Erlenmeyer flasks, placed in a clean bag with a small opening to vent, and left shaking at 37 °C for one week. Samples of each treatment were taken initially, then after the 1st, 3rd, and 7th days (t= 00, 01, 03, 07). The samples were diluted, acidified, and analyzed using an ELEMENT XR High Resolution Inductively Coupled Plasma Mass Spectrometer (HR-ICP-MS). The limit of detection (LOD) for ⁵⁶Fe and ⁵⁷Fe were 0.3 nM and 0.012 nM, respectively (Shrivastava & Gupta, 2011).

Problem Report:
The original measurements of the M9 minimal media ⁵⁶Fe contamination was high, so in an attempt to lower background contamination the media was chelexed. However, the Chelex-100 resin was not rinsed with acid, so was less active for removal of metals. This resulted in higher than expected background concentrations of metals.

Three samples were contaminated with bacterial growth (as indicated by visual turbidity and confirmed using SYBR nucleic acid stain and epifluorescence microscopy) over the course of the week. On t=3, sample B2 was contaminated. By t=7, sample B1D and B1S were contaminated.