1	liquid dispensing validation LESSONS LEARNED IN THE TRENCHES
2	lessons learned… Robbins Scientific Hydra 96 ScreenStation dispenser integrated with multi-mode detection uL, non-contact in 96, 384, 1536 AquaMax Line uL, non-contact in 96, 384, 1536 plate washer IonWorks HT dispenser integrated with automated patch clamp (e-phys.) uL, contact in pseudo-1536 Hand pipetters
3	topics Selecting a detection method Meniscus effects on precision Pitfalls in air displacement dispensers
4	selecting a detection method Method is very robust Sensitivity to dust and other contaminants is rare Absorbance is uM to mM Fluorescence is pM to nM Sensitivity to meniscus effects is small Dyes are stable and cheap SpectraMax Plus³⁸⁴ good to ~0.001ABS in 96 and 384 Does not read 1536 Analyst GT good to ~0.005ABS in 96 and 384 Can work for 1536, but cross-talk is a consideration Black wall, clear bottom plates help cross-talk, but are expensive
5	selecting a detection method Method is less robust than absorbance Sensitive to dust and other contaminants 10X, 100X readings not uncommon Appears to be more sensitive to meniscus effects Dyes are cheap, but not as stable (photobleaching) Analyst GT precision is ~0.5% on stable, dry targets CVs of 2% possible in 96, 384 CVs of 5-7% possible in 1536 Cross-talk is negligible And the winner is… 96, 384 – Absorbance on SpectraMax 1536 – Fluorescence on Analyst
6	why are fluorescent 1536 CVs high? Imprecise dispensing? Same volume dispensed to 384 plates gives lower CVs Readout artifacts increase CV? Does meniscus non-uniformity lead to: Readout artifacts? Higher CVs? Indirect evidence CVs improve slowly with time (aging) Centrifuging improves CVs Could help make meniscuses more uniform Could be due to bubble removal Surfactants improve CVs Meniscuses look more uniform to the eye
7	absorbance and fluorescence differ
8	meniscus shapes
9	what factors affect meniscus shape Well size Well shape Round vs square Slope of well side walls Properties of liquid (relative to air) Difference in density (gDr, ~9.81 g/mm²/s²) Surface tension (s_LV, 20-70 g/s²) Capillary constant a = (2g/gDr)^1/2 = 2-4 mm Relative wet-ability of microplate for liquid, air Contact angle (0-180°) cos(q) = (s_SV – s_SL)/s_LV When the liquid wets the solid, q=0 Depends on surface material, roughness, cleanliness, etc.
10	surface tension of aqueous solutions Inorganic salts, acids, bases (>10 wt%) Organics (alcohols, acetone, etc., <0.5 wt%) Detergents and surfactants (<0.1 vol% ) Contact angle???
11	properties of selected liquids Surface tension Tested in capillaries of unknown diameter --- calibrated to water Contact angle Observed on bottom side of PS 384 microplate Angles estimated by eye
12	theoretical model of meniscus shape Round wells only Assumes cylindrically symmetric shape Assumes physical properties known and constant
13	model results
14	is a curved meniscus better?
15	theory: meniscus artifacts Meniscus shapes are non-uniform “Flat” meniscuses are not necessarily uniform Well-to-well meniscus variability leads to detection artifacts Absorbance Beam steered away from detector Path length changed Fluorescence Meniscus reflections affect background Size, shape and position of sensed volume change Liquids that wet the plate can reduce detection artifacts Wetting helps make meniscus shapes more uniform Recommend surfactants for testing liquid dispensing precision Speculative theory
16	air displacement dispensers
17	fountaining and coring
18	system dynamics of air displacement
19	acknowledgements