File Descriptions:


To read HDF5 (.h5) files in Matlab, you must follow the instructions found here: http://hasyweb.desy.de/services/computing/nexus/hdf5-external-filters/install_on_windows.html
Briefly, this instructs you to install the needed C++ runtime environent on your PC (with the file vc_redist.x64.exe which is included here). Then you must install the plugin which enables Matlab to read Lz4 compressed data (using the included file: h5lz4-1.0-win64.msi). Additionally, the detection limit data is bitshuffled and to read it you must copy the file H5bshuf.dll into the same folder were the LZ4 filter is installed.

The Detection_Limit.h5 files contain DMC-XRD patterns from the indicated nanoparticle and concentration within droplets

The DMC_XRD.h5 files contain representative DMC-XRD data from an experiment with Porous 58S Bioactive Glass at the indicated channel position (pos).

DMC_XRD_processing.m is a sample Matlab script used in processing DMC-XRD files

The jpg. files contain optical micrographs of from "bulk" 2 muL droplets utilised in preliminary mineral nucleant experiments. 

The optical.tif files contain polarised light optical micrographs used in the PDMS device based droplet experiments utilising the indicated nucleant.The "1day_old" and "2month_old" files were obtained from experiements using Porous bioactive glass that was made into a solution 1 day before experiments and 2 months before experiments, respectively.

The collected_SEM.tif files contain scanning electron micrographs of crystals collected from droplets after microfluidic experiments with the indicated nucleant after 2 days.

The bulk_SEM.tif files contain scanning electron micrographs of crystals precipitated in bulk on silicon wafers with the indicated nucleant. The Non_Porous_BG_1.77_wt_percent_ files were conducted with the increased concentration of 1.77 wt% Non-porous 45S5 bioglass to compare with the same surface area of the porous 58S bioactive glass.

The TEM.tif files contain transmission electron micrographs to confirm the size of the indicated nanoparticles used in detection limit or seeding experiments.

The .csv files contain time-resolved turbidity measurements of crystallization in the presence of porous 58S bioactive glass solutions made immediately before experiments ("fresh") or 1 day before experiments ("1day"). Each cased was repeated 3 times.

BET_BJH.xlsx - Sheet 1 contains BET surface area values and average BJH pore sizing values. Sheet 2 contains the entire BJH results for each porous nucleant.





Technique Descriptions:


Droplet Microfluidics Coupled X-Ray Diffraction (DMC-XRD)
DMC-XRD analysis at ESRF beamline ID13 (Microfocus) was performed with an X-ray beam of 13 keV and 12 x 15 µm^2 spot size using an EigerX 4M detector at 116 mm sample-to-detector distance. Microfluidic devices were mounted on a computer-controlled XYZ stage, where alignment and positioning were facilitated with an inline optical microscope. After the coordinates of each analysis position were determined, the source flows were switched on and allowed to equilibrate. Then 10-20 second exposures were collected at each position at 50 frames-per-second (fps). A similar procedure was also utilised at Diamond beamline I22 (Small Angle X-Ray Scattering).

DMC-XRD Data Processing
A Matlab algorithm was developed to cycle through the frames of a particular time-resolved exposure, where frames containing oil scattering are discarded and frames containing water scattering are background subtracted. The background subtraction routine consists of subtracting a frame from the same exposure, but one not containing any crystals, from the target frames. It is not possible to use a single background reference for all channels for all experiments as small differences in sample-to-detector distance, texturing/imperfections in the windows, and possible beam clipping of channel walls, make each exposure too unique for application of a universal background reference. Any remaining background noise is removed with a threshold identified for each experiment. These frames are summed together to form a composite 2D pattern incorporating all the diffraction observed during that exposure. This pattern is then integrated, and the detector parameters (pixel size, aspect ratio) and the sample-to-detector distance are taken into account to produce a 1D pattern displaying intensity as a function of 2theta. Reference data for particular crystal polymorphs are then plotted against these 1D patterns to identify particular peaks, where errors in peak position are typically < 0.05 deg.

Scanning electron microscopy (SEM)
SEM was conducted on FEI NanoSEM Nova 450 from samples dispersed in ethanol and dried on a silicon wafer. Samples were mounted on aluminium stubs with double sided Cu tape. All samples were coated with 2 nm Ir conductive layer prior to analysis. 

Turbidity measurements
The turbidity measurements were conducted with a Perkin Elmer Lambda 35 UV-Vis double-beam spectrometer in time-drive mode at 500 nm wavelength, 2 nm slit size and 1 sec exposure time for 600 sec. Samples were mixed in PMMA cuvettes and placed into the spectrometer for analysis. 100% transmission was calibrated using a cuvette filled with deionised water.

Transmission electron microscopy (TEM)
TEM was conducted using an FEI Tecnai TF20 FEG-TEM after dispersing a powder sample in ethanol and drying it onto a carbon-coated Cu grid.

Optical and polarised microscopy
Optical microscopy was conducted with a Leica M165 FC stereo microscope in bright field transmission mode. Images and videos were recorded using a USB 3.0 Leica DMC2900 color camera with a 3.1 Megapixel CMOS sensor using the Leica Application Suite (LAS) software installed with Single and MultiTime modules. Polarized images and videos were obtained by orienting the analyzer above the sample at close to 90 deg to the polarizer below the sample.

Brunauer–Emmett–Teller (BET) and Barrett-Joyner-Halenda (BJH) analysis
BET was performed with an ASAP 2020 Plus system (Micrometrics), where the pore size distributions were determined from the BJH model of N2 desorption