Vacuum array for distillation

I have been seeing all types of vacuum array for short path distillation and am curious

@spdking suggests a Sogevac, twin Edwards e2m28 and a diffusion pump

I also see people such as myself using single Alcatel 2021i, single Welch 1400 etc.

What do you vacuum do you run and what for what size system are you using on. What vacuum do you achieve with this setup?

I use a single Alcatel 2021i for 500ml and 1000ml systems. Running full bore and oddball China glass I have achieved 6 micron and this works great for what I do.
That being said I am wanting to upgrade vacuum system to a more serious setup and want ideas

1 Like

Used leybold trivac 2.5
2L spd (elcheapo china glass)
Best was 130micron during mains


Question how good do you clean your crude before spd?
Real good clean a good Rotary. Vane Will do the trick
No pre cleaning get a volatile pump and a mains pump
Diffusion Nice but notneccesary until flask is 10 L or more
Size the larger the better but 8 cfm per liter of flask is more than enough

Ok so you are saying for rule of thumb 8cfm per liter is sufficient. This is a good starting point I run 14.5 cfm in a 1000ml . My crude is very clean from the start and my first pass is always golden. I just bought a Welch 1400, a Sogevac sv25, and a Leybold Trivac vde 0530 to try out . They should start arriving this week will see what we can do.

I have pulled 6um on my 1000ml with mostly cheap China glass and single Alcatel 2021i . I believe @Killa12345 is seeing similar results


This making me want to slash my face into a brick wall. There is so much miss information here it’s making we wonder if people purposely mislead each other for fun or what. How would you like me to clear all this up for you. Free of charge. Next time I’ll bill you by the hour


Misleading info ?
There is a boiling surface there is vapor that needs to be condensated
And a time frame in wich to do iT with as little Harm to compounds dooing iT
What mis info do we have in. This tread that makes you. Want to mutilate yourself

1 Like

Do you want me to explain how to guage pumps to your process and understand why or you want to go back and forth some more.

Please do
Also Please remember that @Poizzin
Is asking for help on a 1000ml flask
We could recomend a sogevac as roughning pump a 2 stage Rotary vane a diffusion and pirrani sensors before and after the dryice coldtrap backed by the savant but. I think he yust wants advice
For a. Nice and steady distillate

1 Like

What determines a proper pump in application is based on some math. I will not divulge it but it’s pretty simple to figure out.

Wat is taken into consideration is the CFM.

E2m30 may have around 20cfm are free air ballast.
It will have approx 2.5-3.8 cfm at 20 microns.

Adixen 2021 may have 14cfm at free air ballast.
It will have about 1/4-1/3 cfm at 20 micron.

These numbers are general.

So. At low vacuum you lose CFM. Meaning the pump can no longer carry and drag molecules to the pump as efficient. So a solution is to add CFM via sweep(that’s a different topic). Anyways. So the faster you want your short path to run the faster the CFM must be at vacuum and lower the boiling point/vacuum pressure cusp will be at.

So when you have a 1l short path basically any small pump will work because the pressure at the pot isn’t very high and boil off is low.

See at vacuum you boil compounds. Meaning the pump uses vacuum conductivity(like electricity when a wire conducts electrons in a path) vacuum conducts melecules side by side in a open pathway. While the boil off occurs you are backing molecules side by side. And relying on the mechanical function of the pump to grasp each molecules and drag in down the pump throat. With pumps at low CFM at vacuum we see lower speeds and higher temps. Because these pumps cannot overcome the boil off.

Don’t just use a pump down curve from a supplier. Use advice from people with analytical pump controllers. But the pump down curves by manufacturer do help. The smallest drop off and largest curve is what you want. Mellow curves and hard drop offs won’t make it easy at all to perform. Infact it’s counter productive.

Search for pumps with highest CFM at vacuum.
And bank them for larger systems.

Take for example the xtractor depot pump they stole from cacejen - the price is about three times what they cost. However it’s pump down curve is not the same and is entirely lower cfm at vacuum. So when making comparison to that woosung pump that cost 900 white labeled from Korea to a Edwards it’s night and day.


What about using a dry scroll and a cold trap? Jw

Nice write up and completly. Right reason i always recomend EU laboratory brands instead of hvac pumps
Thx for this :+1:

1 Like

Rotary plate scroll pumps aren’t efficient for short paths. I’d rather use a single stage brick rotor pump style to pull volatiles because they can more efficiently eject the contaminants out of the hot oil. Where scrolls it gets stuck inside the pulps.


great info here

Thank you for this info. I do find running full bore at 1000ml is not the same as when I was running a cow. The Alcatel 2021i is working but I am finding as I upgrade glass, I find different boiling points and am actually learning what is happening.

So how does the Welch 1374 rate compared to the Edwards 30 with the curve and lower micron rate?

When they were actually make in Europe those were decent. Almost every part is not made in China. The Welch pump down curve is sorta honest but I think it’s on a 1-2cuft chamber. So in reality it’s performance curve is decent but it’s very slow.