Sciencemadness Discussion Board

No experience with fractional distillation. help me decide what column to buy?

Cou - 1-1-2020 at 21:50

I don't know if I was using wrong technique, but I recently tried out a 200 mm vigreux column at my local makerspace, and didn't like the results. i was distilling the mixture resulting from fischer esterification of methanol, and a large excess of acetic acid. these azeotropes should have come out:

95% methyl acetate / 5% water: 56.1 C

methanol: 64.7 C

water: 100 C

acetic acid: 118 C

however i found that the boiling point stayed around 54 C for like 20 minutes, then gradually rose up until I decided to stop at 94 C. again i think maybe i went too fast. I looked at Vogel's fractional distillation section after my attempt, and i learned i had a drop every second, instead of every 2-3 seconds, and no insulatoin.

so i don't think vigreux columns are very good. i'm considering a 400 mm hempel column that i will pack with raschig rings, but that costs $90. Might be overkill for most applications, but if it can seperate a few degrees then it would be very useful for recycling solvents, worth investing in. maybe even allow u do to fischer esterifications without workup. how about this snyder column? seems kinda short https://www.amazon.com/gp/product/B01ASZ721G?pf_rd_p=ab873d2...


[Edited on 2-1-2020 by Cou]

[Edited on 2-1-2020 by Cou]

j_sum1 - 1-1-2020 at 23:04

Not that I have ever had a need to do it, but you can improvise your own packed column fairly easily. Glass beads or anything similar can be used for packing and you can use a liebig vertically. You just need some kind of porous plug to make sure your packing is contained.

A 200mm vigreux will give you maybe 2-3 theoretical plates if used correctly. That snyder is designed to be 3 plates so it is not likely to be a whole lot better even though it would look and sound very cool.
The hempel + rasching rings will be your best performer. But as a budget option, just break up some glass tubing into short lengths and go with that.

The key is to go slow. Dont boil too quickly or your column will flood. Insulate so that you get a small
temperature gradient: which is what you actually need for separation. If it is going well you will have most of your liquid reflux back into your flask and only a small fraction coming over. I would think that anything less than a 3:1 ratio and you are wasting your time. You may as well do simple distillation.

With anything like this, getting a good understanding of phase diagrams and working out what is actually happening throughout the process is essential for full understanding. (What is evaporating? What is condensing and where? What is the concentration at each point etc. It is not that you need quantitative answers but you do need to understand the separation process.)


For me I find that I am noramlly a few degrees off in my temperature measurements when I do distillations. A combination of altitude, thermometer calibration and measurement lag. Still, 54° seems a bit low to be methanol so I am confused as to what it might be. Always worth looking up all possible azeotropes to see what is happening and remember that ternaries exist also. And don't forget impurities from side reactions that may also be present.

Chemetix - 1-1-2020 at 23:21

You have a few substances that will be difficult to separate cleanly. For methanol/ water a practical column will be about 800mm to 1000mm with a 30:1 height to diameter ratio. A packed column with Raschig rings or my personal favorite, expanded clay beads for hydroponics. I give them a sieve with some mesh to select the 8-10mm diameter ones. Ethanol water will give the azeotrope with 2m column. I'd say contact your local glassblower and get a custom column made for you.

DavidJR - 1-1-2020 at 23:28

Quote: Originally posted by Cou  

i'm considering a 400 mm hempel column that i will pack with raschig rings, but that costs $90. Might be overkill for most applications, but if it can seperate a few degrees then it would be very useful for recycling solvents, worth investing in. maybe even allow u do to fischer esterifications without workup. how about this snyder column? seems kinda short https://www.amazon.com/gp/product/B01ASZ721G?pf_rd_p=ab873d2...


I have a a 400mm hempel column packed with ceramic raschig rings and it works well (and cost me a lot less than 90 USD) but you do still need to go slowly to get good separation. I really only use it for solvent recycling because it has a large holdup volume and is a bit of a pain to clean if you get nonvolatile crud in the packing. Packed columns are also susceptible to flooding if you heat too strongly, which is not an issue with Vigreux columns.

Insulating the column will allow better separation.

However, it is a tradeoff because it also decreases the effective capacity meaning you have to go slower. This is because decreased cooling of column also decreases the reflux ratio, which in turn increases the HETP (height equivalent of a theoretical plate). This can be avoided however by using a partial takeoff head (because the cooling required for reflux is external to the column itself, and also you can easily dial in the reflux ratio).
Quote: Originally posted by Cou  

maybe even allow u do to fischer esterifications without workup.

Distillation is workup. Also, in my opinion, fractional distillation is not a particularly convenient workup method anyway because it's time consuming and creates lots of dirty glassware to clean.

P.S. Please use actual words ("you" instead of "u") and capital letters at the beginning of sentences...


[Edited on 2-1-2020 by DavidJR]

XeonTheMGPony - 2-1-2020 at 03:37

Quote: Originally posted by DavidJR  

P.S. Please use actual words ("you" instead of "u") and capital letters at the beginning of sentences...


This can not be overstated!

With that said, for a separating column I don't even waste my time looking at any thing under 300mm min (Same for condensers really).

A good thing to do is get a 3 channel thermal couple and attache it to your column 4 is fine too, it will help you get a feel of how things are going in it

All so as stated heat control is every thing! Start slow and bring heat up very slowly with lots of time between adjustments, keep a detailed log and eventually you'll find a sweet spot for your heat source in terms of settings for the mix you're distilling

[Edited on 2-1-2020 by XeonTheMGPony]

Amos - 2-1-2020 at 06:38

Get yourself some fine bits of inert porous material as boiling chips. Rough, gritty bits of porcelain or quartz make great chips if you don't want to buy them. By insuring you have lots of nucleation sites for bubbles to form you'll achieve a very even boil and less burping. I've never had much complaint with a 300mm vigreux wrapped in a thin foil jacket.

SWIM - 2-1-2020 at 14:31

Distillation is always harder than people think.

But you ought to be able to find a good Hempel and packing

https://www.ebay.com/itm/New-In-Box-Kimble-Kimax-Condenser-C...

This is a Hempel that also has a jacket so it can be used as a condenser too.
Quality Amrtican glass, new in box!
(Notice the indentations near the bottom. They're to keep the packing from falling out.)

Plug the hose fitting when using it as a distillation column, or plug one and apply vacuum to the other for a half-asssed vacuum jacket.

Note: beware of condensers on Ebay being sold as columns. many are too thin inside to take most packing materials, and the don't have those dents or spikes at the bottom to retain the packing.



[Edited on 2-1-2020 by SWIM]

There's stuff like this too, if you don't mind Chinese mystery glass.

https://www.ebay.com/itm/400mm-24-29-Distilling-Column-Lab-D...

[Edited on 2-1-2020 by SWIM]

Cou - 2-1-2020 at 15:23

Quote: Originally posted by SWIM  
Distillation is always harder than people think.

But you ought to be able to find a good Hempel and packing

https://www.ebay.com/itm/New-In-Box-Kimble-Kimax-Condenser-C...

This is a Hempel that also has a jacket so it can be used as a condenser too.
Quality Amrtican glass, new in box!
(Notice the indentations near the bottom. They're to keep the packing from falling out.)

Plug the hose fitting when using it as a distillation column, or plug one and apply vacuum to the other for a half-asssed vacuum jacket.

Note: beware of condensers on Ebay being sold as columns. many are too thin inside to take most packing materials, and the don't have those dents or spikes at the bottom to retain the packing.



[Edited on 2-1-2020 by SWIM]

There's stuff like this too, if you don't mind Chinese mystery glass.

https://www.ebay.com/itm/400mm-24-29-Distilling-Column-Lab-D...

[Edited on 2-1-2020 by SWIM]


I ordered from Deschem and they sent me the wrong frikin glassware! An erlenmeyer flask, when the customs label said "glass tube".

Sulaiman - 2-1-2020 at 20:28

When it comes to fractional distillation I consider myself quite expert
- at discovering things not to do ;

. use a cheap on/off heating regulator
- the changes in evaporation rate cause a change in gas velocity which upsets the equilibrium of the column
+ use a variac or phase-controller (dimmer) for constant heating rate

. boil the pot contents vigorously
- also causes changes in gas velocity, disturbing column equilibrium
+ use a constant simmering boil

. work outdoors
- rapid changes in ambient temperature (direct Sunlight etc.) or an intermittent breeze will disturb column equilibrium
+ work in the shade, preferably in a climate controlled room

. be optimistic
- in theory, theory and reality are the same, in reality .....
+ be realistic;
distilling diy ethanol using 2x Quickfit FC7/23 (25.4mm i.d. 250mm packing height) columns, packed with 6mm glass spheres ,with reflux, gives about 5 'plates' each,
so I expected 11 (1+5+5) plates performance.
The boiling pot + lower column half produce 90+ %ABV (from 16%ABV to 2%ABV)
so the upper half of the column should have a temperature difference from top to bottom,
of significantly less than 1oC spread over five theoretical plates,
... extremely difficult to maintain equilibrium.

I think that if you want to reliably and predictably perform fine fractional distillations
you will want a reflux condenser and variable takeoff head, on top of the column.
You could make your own column, or buy a nice silvered vacuum jacketed one
or go for something exotic like Heptylene has https://www.sciencemadness.org/whisper/viewthread.php?tid=71...
(you need to decide on a scale of operations first ... ml/hour)
and have a semi-permanent setup
(with entertainment, food, drinks and toilet not too far away)
OR
have a modest fractionating column that is convenient to set up, clean and store.
e.g. be sattisfied with less plates (2,3,4,5 or 6 is better than 1)

Stainless steel wool (the shiny type of kitchen scrubber) is compatible with many distillates
it is self-supporting, fairly good at stopping channeling, and does not drop little pieces of itself into the boiling pot
and if nicely packed, gives better performance than 6mm dia glass spheres,
with moderate holdup.

My dream of a simple fine fractionating still has gone - I've given up on the 'simple' part.
I will probably be unable to start on my next version for another year (I need to have a house built first).

P.S. if you do go for a complex column, arranging both columns and condensers vertically saves a lot of space.

[Edited on 3-1-2020 by Sulaiman]

Cou - 2-1-2020 at 20:46

Quote: Originally posted by j_sum1  
Still, 54° seems a bit low to be methanol so I am confused as to what it might be. Always worth looking up all possible azeotropes to see what is happening and remember that ternaries exist also. And don't forget impurities from side reactions that may also be present.


Methyl acetate and water form an azeotrope at 56 C. Methyl acetate itself boils at 57 C

j_sum1 - 2-1-2020 at 22:49

Thanks Cou. I somehow missed methyl acetate on the list.
Still, I was thinking general principles rather than the specifics of this problem. Thete is a lot of good advice in this thread.