I am using a Beckman ultracentrifuge to settle out colloids (less than 1um in diameter, with filtration occurring beforehand) from soil column leachate and trying to figure out the lower limit of particle size I am trying to pellet. I used the ka factor corresponding to my planned rpm, derived from the planned g-force of 300,000 at maximum rotor radius, to find how much time it would take to settle particles of certain Svedberg values. For instance, a particle of 80S (eukaryotic ribosome for my own reference for organic material, though I am still searching for values for inorganic) would take a little less than an hour. However, even though the centrifuge tubes are made for the rotor, they do not hold the rotor's maximum capacity (26.6mL instead of 38.5). Is there a convention for correcting for this? For reference, this is the rotor manual: http://www.laborgeraete-beranek.de/info/Type%2060%20Ti.pdf
I also tried using the algebraic Stokes settling equation (Vt=gd^2(p_p-p_m)/(18*u)) for pelleting spheres to find out the maximum particle size that would not be settled out (assuming same speed throughout settling from the top of the tube for a very conservative estimate and assuming particles are of a relatively low density, as material in the leachate is both organic and mineral matter). But I am cautious to use this; does Brownian motion of such small particles make this calculation unuseful? If so, is it useful to try to account for different particle densities, or, when applied to mixtures, is it convention to only stick to Svedberg units? If accounting for different densities, is there a convention for possibly very high-surface area minerals (I have seen that man-made nanoparticles are pretty straightforward but the case for soil is more complicated)?
- Badges
- Science topic