Sciencemadness Discussion Board

Calculate your fractionating column

Sulaiman - 11-9-2016 at 14:22

I came across this document which compares glass, lead & ceramic spheres of 2 to 4 mm diameter in a 28.8 mm i.d. vacuum insulated fractionating column.

e.g. based on the document I estimate that my 1" i.d., 10" effective height quickfit FC7/23
packed with 3mm dia. glass spheres, should give about
6 theoretical plates (with vacuum insulation)
throughput 15 ml/min. or 0.7 ml/min. at reflux = 19:1
holdup c11.5 ml.

it cleared a few questions lurking in my head, nice simple document.

http://nvlpubs.nist.gov/nistpubs/jres/19/jresv19n5p593_A1b.p...

[Edited on 11-9-2016 by Sulaiman]

Magpie - 11-9-2016 at 17:32

My columns are smaller ID, about 1/2". I use some ss scrub pad when I can but go to broken glass when I have to for chemical resistance. I use these open packings to prevent flooding. The broken glass does not distribute especially well so there's usually some channeling.

I would like to find better packing, split helices and the engineered honey comb kind (structured packing) but have not been able to find them in small enough quantity to be reasonably priced. This might be a good candidate for a group buy or for one of our member sellers.

[Edited on 12-9-2016 by Magpie]

[Edited on 12-9-2016 by Magpie]

wg48 - 12-9-2016 at 07:39

Quote: Originally posted by Sulaiman  
I came across this document which compares glass, lead & ceramic spheres of 2 to 4 mm diameter in a 28.8 mm i.d. vacuum insulated fractionating column.

e.g. based on the document I estimate that my 1" i.d., 10" effective height quickfit FC7/23
packed with 3mm dia. glass spheres, should give about
6 theoretical plates (with vacuum insulation)
throughput 15 ml/min. or 0.7 ml/min. at reflux = 19:1
holdup c11.5 ml.

it cleared a few questions lurking in my head, nice simple document.

http://nvlpubs.nist.gov/nistpubs/jres/19/jresv19n5p593_A1b.p...

[Edited on 11-9-2016 by Sulaiman]


Yes a nice find.

From that paper perhaps the most useful equation is the equivalent number of theoretical plates for a column. With my own adjustments, the number of theoretical plates is 30 times the total area of column (packing + inner wall area m^2) plus one for the boiling flask. Review the paper for the assumptions. A major one is total reflux which could be interpreted as say a reflux to take off ratio of greater than 10:1.

Irritatingly that makes my Vigreux column and flask just 3 theoretical plates.

For completeness (from Hive, no assumptions given)

BP difference- No. of theoretical plates
108- 1
72 - 2
54- 3
43- 4
36- 5
20- 10
10- 20
7 - 30
4- 50
2- 100

I guess the figures are for 95% purity and ideal mixtures.

and if you don't like tables from inspection to 5% accuracy :
BPdiff is 216/(number of theoretical plates)



[Edited on 12-9-2016 by wg48]

Sulaiman - 12-9-2016 at 08:32

WG48, thanks for the other half of my quest,
so on a good day I may be able to separate fractions with b.p. difference of 36 C with one column, and 20 C using both,
at least a target to aim for.... close enough for a first approximation.

but quite far from my initial/naive imagined simplicity of distillation :o

"Irritatingly that makes my Vigreux column and flask just 3 theoretical plates."
it is probably worse ... unless your vigreaux is vacuum insulated :P

I now realise that my earlier experiments with my 10/19 kit failed due to my impatience / over-estimation of throughput.
Edit: forgot to add; and no understanding of plates, heating control, insulation, reflux .....
I'd read it but did not quite 'get it' until recently.
I'm not even up to understanding pressure, zeotrope/azeotrope etc. yet :(

[Edited on 12-9-2016 by Sulaiman]