So just wanting to further understand the relationship with CFM, micron, and distillation speeds
I know the rule around here is about 4x flask size to cfm.
But I’ve also heard from a user on here that they used a 0.9cfm pump at 0.1 micron on their run on a 5L and their mantle temperature stayed around 180-190c the whole time, and about 300ml/h.
I get those same numbers with my Edwards pump but, with 16+cfm.
When started thinking about the mechanics of the situation it seems like if we’re only getting 300-500ml per hour, I doubt the vacuum pump is removing (16cfm * 60 min) 960 cubic feet of anything in an hour; gas or otherwise (I do realize the significantly higher density of an oil).
Do different pump styles have different “working” cfms? They were working with a belt drive, vs my vane.
What explains this massive discrepancy between my 16cfm and their 0.9cfm run with the same results?
So, CFM ratings are based on atmospheric pressure for most vacuum pumps. They’re not technically positive displacement but they’re pretty close, meaning they move that much volume regardless of the inlet/outlet pressure. When you look at mass transfer, you might still be moving X CFM at 10 mtorr but you’re moving a tiny fraction of the mass.
This brings up the point that with “perfect” distillation and condensation, there really shouldn’t be much need for a big pump because every molecule of vapor generated in the boiling flask should be condensed somewhere before the pump (condenser, cold trap etc.). That being said, there is always some amount of leakage, failure to condense in your cold trap, and non-condensable gases being evolved. For our application, I imagine the latter is a major culprit (decarb).
The other big factor at highish vacuum is vapor generation at the pump. Obviously if you start boiling your vacuum oil/junk in it then a lot of that “almost-fixed-volume” gets used just recompressing that oil vapor and condensing it in a circle (or not and it gets turned into mist)
Remember. It’s about cfm at vacuum rating. At desired vacuum rating what is the tottal.cfmmkt can achieve. It’s alot less at back than at ballast rating.
Like @spdking says, a lot of pump curves are not very accurate, including some from very reputable manufacturers. Maybe they would be if we were pumping N2 but there genuinely are a lot of factors with pumping what is a large differential pressure when you look at the orders of magnitude of absolute pressure being overcome.
Some of the factors that subtract from the “idealness” of the distillation can be quantified and addressed. For instance, testing the vacuum leak rate with a dry system will let you know how much “CFM” is being wasted on the leaks. Usually you can tell by smell if you need more cold trap, that is often the case. Trap vs pump capacity can be an optimization as well because sometimes it’s cheaper to just run a bigger pump.
Yea. There’s several. It’s just depends application. I’ve tested them all and I tend to keep on my inventory the heavy hitters where I don’t get complaints or customers asking where the balls are a week into ownership.
You can always do the overkill thing and run a monster Edwards diffstack backed by a Alcatel ADS 602 with a good pile of cold ass traps to protect them but I wouldn’t call it “optimized”. Works pretty good though
Think of the vacuum level as density of vapor in a given volume. At deeper vacuum there is lower density of vapor molecules in that space so even though a pump has a fixed pumping speed (motor always has same rpm) the amount of molecules is reduced and the real pumping speed is reduced as a function of that.
A rotary vane pump drop off dramatically in the range we operate in for distillation (transitioning to molecular flow) so it is generally easier to oversize a single rotary vane pump. A smaller rotary vane pump used as a roughing pump paired with a roots or oil diffusion pump will be much faster and operate a deeper vac. These style pumps operate differently and can pump much faster in the high vacuum range but require a backing pump and the added complexity is usually a deal breaker for most.
the name of the game is moving molecules out of the system.I think larger pumps became popular for two reasons:
1)due to the larger oil volume, with poor cold traps these pumps can take more junk.
2) Leaks, sometimes people just try to throw a bigger pump at a system because… leaks
sometimes a person with a sledgehammer just sees everything as a rail spike…
I call this the “brute force and ignorance” strategy, but to use it effectively requires some nuance and there’s an unfortunate amount of morons, for lack of a better term let’s call them consultants, who don’t have any concept of scale or production or maybe just lack the desire to do better or admit to the people paying their wages that they don’t know better and are unwilling to do the research.
Just going bigger does work pretty well for a lot things in this industry, we all want a falling film with more surface area, more heat and more cooling power to do more solvent recovery, no? And less vacuum depth will certainly hurt evaporation rates as anyone who’s chased down a leak on a rotovap knows all too well. You can scale vacuum depth & flow from a big pump crudely with a ball valve or finely with some of the valve offerings from Edwards, but the only way to make a small pump pump a big volume is by buying multiples and hitching them together.
What spdking says makes lots of sense. In my opinion to correctly size the pump one have to ask a question, what is the reason we run vacuum in the system. 3 immediate answers to come my head:
1. Remove oxygen from the system
2. Decrease boiling temperatures
3. Direct vapor flow
Getting lots of CFM pump hype is about the same why every Asian vendor call 3 way receiving adapter a cow shpae (spoiler: because some newbs can’t type correctly visual description of the adapter and unsuspecting asian vendors copycat errors without understanding). When beginner chemist purchases Glass made in China at 20% bargain or 50% from US China shops and every joint is leaking air the only way to overcome an issue is to purchase high CFM pump to compensate. Not bashing Asian vendors, but the quality standards they are not aware of yet praying on unsuspected inexperienced fresh chemists from 420 community.
If you run on US or Canada made glass (not China shops like AI, USLAB or BVV) your chances for bleeds are very low and therefore you do not have to concentrate on #1, but #2 and #3. At this point one must calculate its vapor pressure and amount of vapor being produced at certain temperature from certain mixture from certain volume at certain ambient temperature less glass thermoefficiency at given thickness and composition in normal atmospheric pressure. Then take into the equation what spdking is referring to as chart. The pump CFM or LPH should be able to handle it or else. Basically at lower CFM you boiling point will increase during the process, but flow will still be heading to the pump direction unless pressure will so great that it will start popping out joints. It is not very simple, but truth is you do not always require E2M28 or 30 when you run on quality glass. If you look for quality EU or US glass made in lampworked in Canada give https://labcradle.com in Toronto a try, all prices are in Canadian dollars.
I guess for starting chemist it is much easier just to spend extra $$$ for higher CFM pump than get into all this calculations and to understand the process. As long as it works!
