Powder dispensing documentation, data and papers

Research notes

Latest research notes

One note a month, each one on a single measurable question. Published on the first working week of the month.

Reference data

Accuracy and precision in plate filling

Accuracy is how close the mean dose sits to target; precision is the spread across wells, reported as relative standard deviation. Most 96- and 384-well assays are written against an RSD budget, so precision is the number to qualify first.

Metric	Unit	Typical target	How it is measured
Precision across a plate (RSD)	%	Under 1% for coarse free-flowing powder; 1–5% for fine or cohesive powder	Weigh every well of a filled plate on an analytical balance; RSD of the set
Accuracy of the mean dose	% of target	Within 2–5% once bulk density is characterised	Mean well mass against nominal cavity volume × tapped bulk density
Bulk density stability	g/mL	Batch to batch spread under 3%	Tapped density per USP <616> on each incoming lot
Empty and partial wells	count per plate	Zero; any occurrence means bridging	Visual check plus per-well weighing on qualification runs
Carry-over between runs	% of dose	Below detection after standard clean	Blank plate run after cleaning, assayed for residue

Electrostatic behaviour: plastic against metal

Powder charges by friction. Insulating surfaces hold that charge, so powder clings to plastic hoppers and sprays on release; grounded metal lets it bleed away. Humidity is the other lever: charge decays far faster in moist air, which is why the working window sits at 40–60% RH rather than as dry as possible.

Parameter	Unit	Range	Effect on dispensing
Surface resistivity, insulator (most plastics)	Ω/square	10¹² and above	Charge is retained: powder clings, doses run light
Surface resistivity, dissipative	Ω/square	10⁵ to 10¹²	Charge bleeds off safely; preferred hopper and liner range
Surface resistivity, conductive metal	Ω/square	Below 10⁵	Needs grounding, then behaves best of the three
Relative humidity	% RH	40–60% working window	Below 40% static dominates; above 60% caking and clumping begin
Charge decay time	s	Seconds at high RH, far longer when dry	Long decay leaves residual charge between cycles and erratic fills

Hygroscopicity: water as a weighing error

A hygroscopic powder gains mass from the air while you weigh it. The balance reading rises, so the dose looks correct while the active content is diluted by water. The mass bias grows with exposure time, which is the strongest argument for a method that fills a whole plate in seconds.

Property	Unit	Example values	Consequence
Equilibrium moisture content, NaCl	% w/w	About 0.1–0.3% at 75–80% RH	Stable enough to weigh in open air
Equilibrium moisture content, KCl	% w/w	About 0.4% at 75–80% RH	Minor bias, correctable
Deliquescence, CaCl₂	% RH threshold	Liquefies above roughly 30–40% RH	Must be handled in a dry cabinet or glovebox
Recommended storage RH	% RH	Below 30–40%, below 5% for deliquescent solids	Prevents uptake and hopper caking
Water content verification	% w/w	Per incoming lot	Karl Fischer titration or loss on drying; correct the dose from it

Troubleshooting flow failures, in order

Symptom	Most likely cause	First fix	If it persists
Some wells empty or half full	Bridging over the cavity	Tap or vibrate the plate while levelling	Larger cavity diameter, or dry and sieve the powder
Powder clings to the hopper walls	Static on an insulating surface	Ground the frame, raise RH above 40%	Metal or dissipative hopper, ionising bar at the nozzle
Doses creep heavier through the day	Moisture uptake in the hopper	Refill in small batches, keep the lid on	Dry cabinet, or dose inside an enclosure
Hard lumps in the feed	Caking from earlier moisture exposure	Sieve before loading	Review storage conditions and container seals
Composition drifts across the plate	Segregation of a blend	Reduce hopper residence time	Blend closer to point of use, verify by assay
Residue in the next run	Carry-over in the mesh plate	Full clean and dry between powders	Dedicated mesh plate per material

Standards and further reading

Read next

The tables above say what to measure. For how a mesh plate fills a 96 or 384 well plate in one motion, read powder dispensing without the weighing.