Eleanor C. Fadely¹, Gaitan Gehin¹, Sharon E. Bone², Samuel M. Webb³, Verónica L. Morales¹, Jasquelin Peña^1,4

1. Department of Civil and Environmental Engineering, University of California Davis, Davis, CA 95616, USA.
2. Institute of Bio- and Geosciences: Agrosphere, Forschungszentrum Jülich, Jülich 52428, Germany.
3. Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA.
4. Energy Geosciences Division, Earth and Environmental Sciences Area, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA.

Data acquisition

Microfluidic experiments were set up and run following the methodology outlined in the manuscript main text and supporting information. Color brightfield and mCherry fluorescence images of each microfluidic replicate were collected every 1 h during the experimental growth phase and every 30 min during the experimental mineralization phase. Images were acquired as .nd2 files and exported as .tiff files. All time-lapse image data were collected using the same microscope and same imaging settings. Measurements of aqueous Mn(II) and total Mn mass in the microfluidic system were made using inductively coupled plasma mass spectrometry (ICP-MS). Pixel-scale measurements of Mn mass for image calibration were made using micro-X-ray fluorescence (µ-XRF) mapping at the Stanford Synchrotron Radiation Lightsource (SSRL). All data were processed using MATLAB (2024b).

Data organization

Experimental data for the manuscript main text and supporting information are organized in the compressed file "Fadely_Pena_EST_2025_DataRepository.zip" as follows.

1. Data for each experimental replicate (Replicate_1, Replicate_2, Replicate_3) are stored in the Microfluidic_Experiments directory.
   1. ICP-MS data are stored in ICP_MS. Data are stored as .xlsx and/or .csv files.
   2. Processed microscopy data are stored in Microscopy. Registered and cropped color brightfield and mCherry fluorescence images, as well as all subsequent variables, are provided as .mat files in Variables_Growth or Variables_Oxidation. Requests for raw .tiff images can be addressed to the corresponding author.
2. MATLAB scripts used for data processing are stored in Data_Processing. Specific scripts for each figure are stored in their corresponding subfolders in Figures. Scripts are stored as .m files.
3. Data and scripts for main text and supporting information figures are stored in Figures as described below.

Naming conventions and units

In general, MATLAB scripts use the following naming conventions to denote experimental phase and image modality.

1. Color brightfield data. BF, Color.
2. mCherry fluorescence data. FL, Fluor.
3. Growth phase images. Growth, GrPhase, Gr.
4. Mineralization phase images. Oxidation, OxPhase, Ox.

ICP-MS results have units of ppb or µM. Uncalibrated color brightfield and mCherry fluorescence images have arbitrary units (a.u.). The color calibration converts pixel intensity in a.u. to Mn mass in ng pixel^-1. For ease of comparison with the ICP-MS results, all calibrated image matrices have units of µg pixel^-1, and the resulting total Mn mass is reported in µg. Distances are reported in µm or mm and areas are reported in mm². Any missing values are indicated by NaN.

Data processing

Aqueous and total Mn quantification. Aqueous Mn(II) in the reactor influent and effluent was measured with ICP-MS. Total Mn retained in the reactor and outlet tubing was dissolved and measured with ICP-MS. Results are provided in spreadsheets for each experimental replicate (.xlxs and .csv) containing the following: Sample name and number, measured sample mass, dilution factor, instrument readout of Mn concentration (ppb), and corrected Mn concentration (µM or ppb).

Mass balance quantification. Using the ICP-MS results for aqueous and total Mn, a mass balance was calculated for each experimental replicate. Input, output, and stored Mn, as well as mass balance percent error, were calculated from input .csv files, and the result was written to a new .xlsx file for each replicate. Data were processed using the script “BTC_Processing_Final.m.”

Image post-processing and registration. All growth and mineralization phase images (color brightfield and mCherry fluorescence) were straightened as needed, registered to fit a standardized grain mask, and cropped to the dimensions of the pore space. These processed images are saved as .mat files in the following Variables_Growth or Variables_Oxidation directories: Processed_Images. Registration parameters are saved as .mat files in the following Variables_Growth directory: Image_Registration. The grain mask for each replicate is saved as a .mat file in the following Variables_Growth directory: Grain_Mask. The original AutoCAD geometry of the reactor pore space is saved in Figures > Figure_S2. Data were processed using the scripts “Step1_GrowthPhase_Process_Final.m” and “Step2_MineralizationPhase_Final.m.”

Biofilm segmentation. All processed mCherry fluorescence growth and mineralization phase images were contrast-adjusted, binarized, eroded, and multiplied by each replicate’s corresponding grain mask to isolate regions of biofilm within the reactor pore space. The binary mask of each timepoint was then summed to calculate the total number of biofilm pixels. Contrast-adjusted images, binary images, and total biofilm pixel arrays are saved as .mat files in the following Variables_Growth or Variables_Oxidation directories: Biomass_Variables. Data were processed using the scripts “Step1_GrowthPhase_Process_Final.m” and “Step2_MineralizationPhase_Final.m.”

Mn oxide segmentation. All processed color brightfield mineralization phase images were binarized, then all images were multiplied by the mask of the final timepoint to isolate the maximum biofilm-biomineral footprint. All masked image timepoints were then subtracted from the masked reference timepoint to isolate the contribution of Mn oxides to total pixel intensity, and the resulting matrices were converted to grayscale. The reference and final image indices, binary masks, and reference-subtracted image matrices are saved as .mat files in the following Variables_Oxidation directory: Reactor_Calibration. Data were processed using the script “Step3_OxideCalibration_Final.m.”

Mn oxide mass calibration. The binary mCherry fluorescence masks and binary color brightfield masks were used to index mineral precipitation inside and outside the biofilm footprint. Pixel intensity in the reference-subtracted image matrices was then calibrated to Mn mass based on each pixel’s index (calibration curve data are stored in Figures > Figure_S4_S5). Mineral mass at each timepoint was calculated for the whole pore space and each reactor quintile. Rates of mineral accumulation between sequential timepoints were calculated for the whole pore space and each reactor quintile. Indexed image matrices and calibrated image matrices are saved as .mat files in the following Variables_Oxidation directory: Reactor_Calibration. Mineral mass and mineral accumulation rates are saved as .mat files in the following Variables_Oxidation directory: Segmented_Mass_Rates. Data were processed using the script “Step3_OxideCalibration_Final.m.”

Figure data and scripts

The figure data and scripts described below are stored in subfolders in the Figures directory. Subfolders for specific figures contain the script used to generate that figure. Unless otherwise noted, variables used to produce figures are stored in subfolders in the ICP-MS, Variables_Growth, or Variables_Oxidation directories of each triplicate in Microfluidic_Experiments and are loaded into the MATLAB workspace by each figure script.

Figure_1_S7_S8_S9. This folder contains color brightfield and contrast-adjusted mCherry fluorescence images of a representative biofilm at select timepoints during the growth and mineralization phases, false-color images showing the biofilm boundary, and images of the whole pore space at the reference and final timepoints. It also contains the script “Figure1_Images_Final.m” used to crop, process, and save images for Figure 1, Figure S7, Figure S8, and Figure S9.

Figure_2. This folder contains the script “Figure2_Plots_Final.m” used to plot aqueous Mn(II) in the reactor effluent, biofilm growth in the pore space, and Mn oxide accumulation in the pore space over time for Figure 2.

Figure_3_S10_S11. This folder contains the script “Figure3_SegmentedImages_Plots_Final.m” used to produce the false-color pore space map of Mn oxide mass, Mn mass profiles in the longitudinal and transverse dimensions, and biomass distribution in the longitudinal dimension for Figure 3, Figure S10, and Figure S11.

Figure_4_S13. This folder contains the script “Figure4_CroppedImages_Plots_Final.m” used to select representative inlet and outlet biofilms, extract transect coordinates, plot the corresponding Mn oxide mass profile as a function of time, and produce the false-color Mn oxide mass map of each biofilm for Figure S13. Resulting biofilm transect parameters are saved in the subfolders for each replicate (Replicate1_Biofilms, Replicate2_Biofilms, Replicate3_Biofilms). The script also normalizes and averages all individual inlet and outlet biofilms for Figure 4.

Figure_5_S14_S15_S16. This folder contains the script “Figure5_Plots_Final.m” used to calculate total Mn oxide mass inside and outside the biofilm footprint, total and area-normalized rates of Mn oxide accumulation inside and outside the biofilm footprint, Mn oxide mass inside and outside the biofilm footprint in vertical quintiles of the pore space, and the increase in biofilm-mineral assemblage area for Figure 5, Figure S14, Figure S15, and Figure S16.

Figure_S1. This folder contains .csv files batch growth data containing optical density measurements at 600 nm (OD₆₀₀) for the P. putida GB-1 wild-type and fluorescent-tagged strains. It also contains the script “FigureS1_BatchGrowthRates_Final.m” used to average and plot experimental replicates and calculate growth rates for Figure S1.

Figure_S2. This folder contains a high-resolution TIFF of the AutoCAD grain geometry and a PNG of the watershed-segmented pore space. It also contains the script “FigureS2_GrainPoreDistribution_Final.m” used to measure the grain size and pore size distribution of the microfluidic pore space for Figure S2.

Figure_S4_S5. This folder contains the image used to produce a calibration curve to convert color brightfield pixel data from a.u. to µg. This includes color brightfield images of Mn oxide droplets collected at University of California Davis (Color_Brightfield), optical images from SSRL showing the scan region for image registration (Optical_Regions), and calibrated µ-XRF map transects of the droplet samples collected at SSRL (XRF_Maps). Map data in the XRF_Maps folder were collected at Beamline 2-3, converted from counts to Mn mass in the MicroAnalysis Toolkit software, and exported as .npy arrays for import into MATLAB. This folder also contains associated image processing variables and the resulting calibration curve parameters (Variables). Finally, it contains the script “Calibration_Color_to_Mass_Final.m” used to process the optical and µ-XRF data for Figure S4 and produce calibration curves for Figure S5.

Figure_S6. This folder contains representative color brightfield images, contrast-adjusted mCherry fluorescence images, binary masks, masked images, and segmented images. It also contains the script “FigureS6_ProcessedImages_Final.m” used to crop and save images for Figure S6.

Figure_S12. This folder contains 40x color brightfield images, raw mCherry fluorescence images, and contrast-adjusted mCherry fluorescence images of representative biofilms. It also contains the script “FigureS12_40xImages_Final.m” to adjust the image contrast for Figure S12.