Powder sampling / 

subsampling

* Store samples in dry conditions.

* Subject sample to drying / light pestle & mortar action before measurement. CAUTION –possible primary particle breakdown.

* Has separation of fines and coarse fractions occurred during transport?

* Consider re-mixing sub-sample by tumbling in its sample tube immediate prior to measurement.

* Securing a representative sample: has separation of fine and coarse fractions occurred (e.g. during transport or whilst stored in a silo)?

* If dry powder flows as part of the production /delivery process, take samples at different times, positions and combine.

* Avoid sampling from a heap as fines and coarse particles tend to separate.

i) Riffling: Not so good for powders with poor flow.

ii) Cone and quartering: Good for powders with poor flow.

iii) Tumble in container / spatula.

iv) Spatula.

(ii) Creating a fluid suspension in a beaker and using a pipette to transfer to cell can result in large particles quickly sedimenting in the beaker and not being picked up in the pipette. A PSD with smaller particles dominating results.

(iii) Simply throwing powder into the liquid phase in cell could result in the powder taking a considerable time to de-agglomerate or never fully breaking down to primary particles.

(ii) Pestle and mortar – Employ to gently break down agglomerates. CAUTION: Possible primary particle break down.

(iii) Doing nothing means there is a risk of agglomerates not breaking down and PSD appearing coarser.

* Using denser liquids such as glycerol (along or mixed with water) can reduce sedimentation when measuring higher density powders like Ni, WC.

* Check sample is not soluble in water and seek less polar solvents if this is the case.

(ii) Tap water. The soluble ions in tap water (electrolyte) can influence zeta potential and hence how well dispersed the powder in the suspension is. Generally, higher electrolyte concentration and the presence of 2+ or 3+ ions (like Ca2+, Al3+) will encourage unwanted flocculation.

Measurement (I): Fixing software settings.

* Refractive Index

* Imaginary RI

* Pump speed / rate

* Does my instrument use MIE or Frauenhofer calculations?

* Does my instrument have other methods for coping with smaller particles?

* Can the equipment offer corrections for particle shape?

* Changes in this value can alter whether the PSD is broad or narrow

* Energy put into mixing can affect quality of de-agglomeration and (with excess energy) even agglomeration. Watch for temperature increases that alter level of solubility of dispersed powder.

* MIE and Frauenhofer both have a particle size range that they can cope with

* Some instruments can bring an alternative detection mode into play to better measure fines (e.g. use of polarised light)

* Shape will influence laser diffraction patterns with software simply calculating a sphere equivalent size

* Measure RI of the sample and use value.

* Agree a common literature RI value to use.

* Use SEM analysis to get an idea of whether PSD is actually broad or narrow. Apply a fixed appropriate Imaginary RI value (as this value rises from 0.01 to 1.0, the PSD narrows 

* Agree / fix any imaginary RI value to use

* Compare flow rates, pump power and time of mixing for different models/manufacturers of equipment. Carry out calibration checks (how does PSD change versus mixing time?) Identify, then nominate, a protocol for different laboratories / machine that results in a comparable energy input.

* For powders with a significant level of fines, Frauenhofer theory (which can deal with sizes down to 2 μm) may not be suitable. Mie theory can cope with 0.1 μm upwards. Select Mie or Frauenhofer theory option accordingly if available.

* Decide whether additional modes will / will not be used and introduce decision into protocol.

* Again, decide on whether to select or avoid correction factors based perhaps on SEM analysis.

Measurement (II): Fixing other variables

* Checking equipment performance prior to measurement

* Level of aliquot added during a measurement

* Use of ultrasound

* Use of surfactants

* Use of electrolytes and / or pH

* Number of measurements per aliquot

* Number of aliquots to run tests on.

* Background runs (made with solvent only in the cell) allow determination of very low levels of laser light reaching detectors in absence of diffracting powder. How often do I check this?

* Background reading taken is not satisfactory

* Calibration run : What material do I select for calibrating?

* Optical alignment check

* Don not add to little or too much. Too much risks multiple scattering of light and erroneous results. Too little means not enough scatterings to give statistically meaningful data.

* I only have a limited amount of powder to disperse.

* Can assist in reducing agglomeration.

* Can assist in reducing agglomeration.

* Can assist in reducing agglomeration in aqueous systems

* By running repeat measurements, you can check for trends in PSD

associated with the sample gaining greater homogeneity.

* Running greater than one aliquot gives more data to ensure a reliable

result.

* Run before each sample is run.

* Run according to manufacturers recommendations (with built in timing to force user to measure at specified periods).

* Run every hour.

* Run at the start of each day.

* Remove and clean windows, lenses according to instructions to avoid damage to coatings.

* Consider pumping cleaning agent in water in between measurements.

* Air bubbles may be present – circulate water for longer period / avoid vigorous circulation and re-check background.

* Mono-sized glass spheres (avoids variations in data due to shape) with narrow PSD. Run once a month.

* Secure other non-spherical “reference” and run monthly

(i) If % obscuration figures are given aim for adding enough sample to deliver the same % figure on every run.

(ii) Often equipment offers a “window” or range of acceptable obscuration. Make sure you are in this range.

(i) Run measurement at highest % obscuration value below acceptable level. Retain the % as the value always used.

* Watch for excess energy input cause agglomeration (manifesting as temperature rise). Worth doing trials to track PSD vs. U/S power setting, U/S time.

* For aqueous systems, need to check surfactant is suitable for powder being tested. Zeta potential vs. pH plots for powder and powder + surfactant will assist. A zeta value of > ± 30 mV is recommended.

* Monitor zeta potential as function of electrolyte and pH checking for < + 30 mV or > - 30 mV.

* Take three measurements if results are almost identical; take five measurements if not. For five measurements ignore the first one or two measurements if subsequent ones are lower and consistent.

* Generally not necessary if good consistency seen with first aliquot.

* How can I be sure my measured PSD will mean good or bad performance down the processing line?

* How should I display the data?

* I use a LD unit that represents a different model / manufacturer from the one my supplier / client uses.

* Need to do correlation studies.

* Consider what y-axis should be.

* Should I display fractional or cumulative PSD?

* Need to check that a good correlation between different units exists.

* Apply Factorial Experimental Design (FED) or alternative to show the level of correlation between PSD values (D10, D50, D90 etc.) and intermediate or final products made with different powder batches.

(i) Vol.%. The format usually adopted.

(ii) Wt.%

(i) Number % (good for seeing the presence of fines) – can this be done??

(i) Super-impose both.

(ii) Fractional only (good for assessing mono, bi-modal etc. but not easy to see d10, d50, d90, etc.)

(iii) Cumulative only (good for quickly identifying d10, d50, d90, etc.)

* Carry out PSD runs on a series of standard materials with different PSD, Dispersity. Use the same protocol as you have defined in column 4. Plot d10, d50, d90 and d4,3 data for the two instruments against each other.