Sciencemadness Discussion Board

Vacuum From Peristaltic Pumps

jpsmith123 - 13-1-2008 at 13:48

I'm wondering, has anyone ever used a peristaltic pump as a vacuum pump?

It seems to me that a peristaltic pump should be capable of achieving a fairly low ultimate pressure...albeit at a very low throughput.

chromium - 13-1-2008 at 14:20

I once tried to make one but i ran into problems which i was unable to solve.

Tubing did not stay in the place. It got streched and jammed in every possible way and this often led to losing vacuum. I still saw less than 10 torr at my vacuum meter once or twice.

My pump was with rather good ball bearings as rolls but this did not help. It seems that tube must be very well fixed (there is great force applyed to the tube in the axial as well as in the radial direction).

To be honest i have no idea what type of tubing should be used for good results. I did my tests with rubber tubing but, as i said, results were disapointing.

If someone has seen such vacuum pump working, i would like very much to hear more about it.

[Edited on 14-1-2008 by chromium]

Magpie - 13-1-2008 at 14:31

We used a peristaltic pump to feed a brine where I once worked. Polyurethane hose gave the best life in our experience.

I have never heard of a peristaltic pump being used to pump gases. I'm surprised that you could develop a good vacuum with one.

jpsmith123 - 13-1-2008 at 14:38

You tried to build the pump from scratch? I was thinking more along the lines of buying a used commercial pump on Ebay, or at least buying the pump head.

It seems to me it should go well below 1 torr...if you're pumping a small volume.

Also, I'm thinking that maybe a multichannel pump could be connected to a manifold (i.e., in parallel) so that the throughput could be increased?

microcosmicus - 13-1-2008 at 14:44

As for ultimate pressure, the situation is going to be the same as with
any other mechanical pump --- you cannot obtain a vacuum less than a
millitorr because, at that point, the gas laws on which mechanical
pumps are based no longer hold. As for attaining that theoretical limit,
maybe try some sort of multistage arrangement like having the different
stages as different portions of the same hose.. If each stage reduces
the pressure by a factor of x, then n stages will reduce it by x^n so
even an inefficient pump can attain a low vacuum with enough stages.

jpsmith123 - 13-1-2008 at 14:53

I'm not sure I see how 1 millitorr should be a theoretical limit for a mechanical pump. I know a mechanical pump with a molecular sieve trap can go below 1 millitorr.

microcosmicus - 13-1-2008 at 15:33

A molecular sieve is not a mechanical pump --- it relies on
adsorption of individual molecules as opposed to compressing
gas in bulk.

Basically, the reason for the limit is that, once you get to a millitorr,
the gas is too dilute for there to be significant collisions between
molecules. Analyze how a mechanical pump works and you will
realize that the molecules in the molecules in the space you are
trying to evacuate bumping into each other and thereby being pushed
into the pump is important. Once the pressure becomes so low
that a molecule can travel across the length of the apparatus with
a good chance of not encountering another molecule, all that will
happen is that molecules will randomly move between the pump
and the bell jar with no net effect.

For this reason, pumps which make high vacuum, such as sorption
pumps, diffusion pumps, turbomolecular pumps, getters, etc. deal
with molecules on an individual basis rather than by compressing
gas in bulk.

jpsmith123 - 13-1-2008 at 16:04

I didn't mean to imply that a mol. sieve trap "is a mechanical pump", I was merely relating my personal experience that a so-trapped mechanical pump (with good oil) will go well below 1 millitorr; in part because of its ability to adsorb backstreaming oil which will occur when the system goes into the molecular flow regime (or perhaps out of the "viscous flow regime"). (And yes, I know how a mol. sieve trap works).

You said that a mechanical pump will not go below *1 millitorr*; I was merely questioning that statement.

[Edited on by jpsmith123]

microcosmicus - 13-1-2008 at 16:34

Sorry about that -- I was merely pointing out some basic facts which
you obviously already know.

As for the millitorr, that was only meant as a round figure for when mean
free path becomes macroscopic (it is 5 cm long at that pressure and room
temperature) so one has molecular flow. I definitely wasn't trying to say that
the lower limit is precisely 1 militorr.

[Edited on 13-1-2008 by microcosmicus]

vulture - 14-1-2008 at 15:02

Peristaltic pumps are designed for pumping liquids, they suck at vacuum production. Well, they don't suck well enough.

jpsmith123 - 14-1-2008 at 16:50

I don't think there's any theoretical reason why they can't make a half-decent vacuum. Assuming that the rollers are completely sealing the tubing, I think that for small volumes, the ultimate pressure attainable depends on the vacuum rating of the tubing used...IOW if the walls are too thin the tubing will collapse.

Here's a really crude and cheap linear peristaltic pump that someone threw together and it apparently goes lower than 1 torr.

[Edited on by jpsmith123]

chromium - 15-1-2008 at 01:41

Theoretically it can indeed go to very low pressures but there are tehnical difficulties that must be overcome and it does not seem easy.

If you build this very simple apparatus from your link you will see that to close tube relialby you must apply strong force. This not only presses walls of tubing together but also deforms it in great extent. Tubing starts to twist, to move chaotically in sideways and most important - along with rollers. This axial force is so strong that it can stretch your tubing to double of its length if it is anything that can be stretched at all.

I know, there are some commercial vacuum pumps that use this technique. Unfortunately i have not seen one myself but i am sure this needs special sort of tubing and there must be some nontrivial tricks to hold it in place and to release unnecessary tensions that otherwise will destroy the tubing or tear it apart from connectors.

[Edited on 15-1-2008 by chromium]

[Edited on 16-1-2008 by chromium]

vulture - 15-1-2008 at 15:18


I don't think there's any theoretical reason why they can't make a half-decent vacuum. Assuming that the rollers are completely sealing the tubing, I think that for small volumes, the ultimate pressure attainable depends on the vacuum rating of the tubing used...IOW if the walls are too thin the tubing will collapse.

And that is exactly why it doesn't work very well. If the vacuum rating of the tubing is high, it is rather unflexible and won't seal completely when pumping. Also it'll require considerable force.

A fridge compressor is much better as long as you keep nasty gasses out of it. Or build your own membrane pump.

jpsmith123 - 15-1-2008 at 17:13

So you agree with me that there are no theoretical obstacles, only engineering issues, right?

And those are the issues I'm exploring. They may turn out to be show-stoppers, and they may not.

This tubing caught my eye. I think it's intended for peristaltic pumps (although it doesn't say that I'm going to ask them).

I'm interested in the 3/8" ID 5/8" OD tubing. Looks Like it's rated for full vacuum, yet the wall thickness is only 1/8". I know that "Masterflex" makes pump heads that will take 3/8" ID tubing, but I don't know what the OD of the Masterflex tubing is.

chromium - 16-1-2008 at 12:42

I may be wrong but i am sceptical on this type of tubing. Mainly because it is too strechable and smoothness of inner surface may be downside as stronger force might be needed to close it completely.

I think that one needs something trully special like soft silicon tubing which is armored with polyamide threads to make it nonstretchable and with special soft inner surface so that it closes vacuum-tight even if walls are pressed together with only slight force.

I agree that problems are technical and if you can find type of tubing that works and lasts, you can go further than 0.1 torr. With ideal tubing (which, of course, does not exist) and infinite amount of time you could eliminate even last gas molecules. In reality it is not easy way. Good luck!

[Edited on 16-1-2008 by chromium]

microcosmicus - 16-1-2008 at 23:20

This topic has gotten me interested enough that I am going to try to
build a peristaltic vacuum pump. Even though such a contraption may
be too inefficient for consideration as a practical vacuum pump, it at
least promises to be a fun project.

The first order of business for me was to determine the force necessary
to seal the tubing. To test this, I attached a metal beam to my workbench,
on which I placed the tubing. To simulate the roller in a pump, I placed a
piece of pipe on top of the tubing (9.5 mm diameter). Through this pipe
ran a string attached to a bucket underneath the beam. By pouring water
into the bucket, I gradually increased the force of the pipe on the tubing.
I blew into one end of the tubing and closed the other end with a bit of
soapy water; when the weight in the bucket had gotten to the point that
bubbles no longer formed, I weighed the water.

I had on hand both some regular aquarium tubing and silicone tubing.
The regular tubing required 10 kg (100 N force) to pinch it off, whilst the
silicone tubing required 3 kg (30 N force) to pinch off.

I plan to go with the silicone tubing in building the pump. For the record,
this tubing has an inner diameter of 4 mm and 1 mm thick walls. A quick
test with a roller (a crude version of the linear pump in the link given
above) shows that this tubing should be suitable for making a
peristaltic vacuum pump.

To be safe, I will throw in a safety factor and apply a force more like
50 N in the actual pump. Also, 1 cm might be too small a roller size,
so I will repeat my measurement with a bigger pipe to see what force
would be needed with a larger roller. One of these days, it might make
for an entertaining exercise in continuum mechanics to compute the force
for pinching off tubing in terms of elastic constants but, for the purpose
of this project, the measured values are all that is needed.

As for the problem of stretching and tubing moving around, I am quite
sure that can be overcome by intelligent friction management with no
need for threads or similar recourses. As for sealing the inside well, I
suspect oiling the inside of tubing should help. Even if some of the
gas does leak past the seal, as long as the rate of pumping exceeds
the rate of leakage by a reasonable factor, that difficulty should be
surmountable by using multiple stages.

Next, I will repeat my measurement with a different size pipe as
mentioned above, measure friction coefficients and try out my plan
for dealing with stretching. Hopefully, I'll have some more results
to share tommorow or so.

[Edited on 17-1-2008 by microcosmicus]

jpsmith123 - 18-1-2008 at 05:35

I sent some emails to a few companies asking a few questions about the ratings of their tubing and suitability for use in peristaltic pumps.

I see a peristaltic pump as a potentially very handy, general purpose pump that could be used not only to pump liquids, but to pull vacuum for filtration, distillation, CVD experiments using methanol-water vapor, or anything else where you don't want to worry about condensable vapors contaminating the pump, or pump oil contaminating anything, etc.

If it seems possible to use tubing that's 3/8" ID or larger, and rated for "full" vacuum, I may start looking for a pump on Ebay.

microcosmicus - 18-1-2008 at 10:08

Filtration and distillation sound like a poor match for a peristaltic vacuum
pump to me. Those applications don't need much of a vacuum but do
require high throughput rates to keep pulling out vapor being boiled off
and the like. By contrast, the discussion above leads me to expect that a
peristaltic pump would have the opposite characteristics --- high vacuum
at low throughput. For filtering and distilling, I would rather use the peristaltic
pump to feed liquid through an aspirator than directly pull vacuum.

When you hear back from the companies, please let me know what they
say about the ratings of tubing. For the time being, I am fine with the
silicone tubing I already have and don't want to spend money on this
project yet. However, if I get a prototype working reasonably well, then I
would certainly be interested in seeing whether it could be improved by
choice of tubing.

I've thought about and puttered around a little more with this project with the
last post. As chromium pointed out, for a peristaltic pump to pull good vacuum,
one is going to need to apply enough force to seal the tubing well. Based on
my measurements, the force required is not inordinately large, but it would
be good to minimize it. Therefore, I am thinking of small rollers rather than
the big brayer in the link above or the usual pump rollers. Last night, I made
a quick measurement and noticed that, by going to a roller with a diameter of
3 mm, I could cut down the force to 20 N.

So what I am now thinking of is an odd looking roller in the form of a shaft which
goes through the inner holes on two ball bearings at either end. As time permits,
I will make more careful measurements, look into keeping the hose from
slipping and stretching, and try repeating the demo in the link above with
this new sort of roller.

Some Correspondence

jpsmith123 - 22-1-2008 at 19:46

Here's an email exchange I had with Tech Support at Masterflex/Cole-Parmer:

I'm wondering whether any of your peristaltic pumps can be equipped with tubing rated for full vacuum, i.e., 29.9" Hg or better? I'm particularly interested in the larger size tubing, e.g., 3/8" ID.

Thank you for your recent inquiry and interest in Cole-Parmer.

Unfortunately all the tubing that is soft enough to be occluded in the peristaltic pumps is not strong enough to withstand full vacuum. The tubing will collapse.

Some of the more rigid tubing, like Masterflex Norprene 06404-35 will handle some vacuum (hard to determine the actual amount), but definitely not full vacuum.

Thanks for your reply.

I wonder if I may ask you one more question: Most 3/8" ID plastic vacuum tubing I've seen has very thick walls; however, Saint-Gobain part #ABW00029 (as per the below link) seems to be rated at "full" vacuum (i.e., 29.9" Hg in this case). Is it possible that this tubing would work in any of your pumps?

Thank you for your continued interest.

Masterflex has BioPharm tubing very similar to the Tygon 3350.

96420-35 is BioPharm tubing that will work with our Masterflex L/S line of peristaltic pumps and according to St. Gobain, may be able to handle full vacuum.

I do not have any data on this, however, we do trust the specifications from St. Gobain for their tubing.


So I suppose that sounds somewhat promising.

microcosmicus - 22-1-2008 at 23:46


So I suppose that sounds somewhat promising.

Another promising fact which showed up in my measurements was that the
pressure needed to seal off even the soft silicone tubing was several times
atmospheric pressure. Using the value of 20 N for a 3mm diameter roller
given above and the fact that the tubing is 8mm wide gives a quick estimate
of 20/(3x8) = 0.8 N/mm^2 = 8 atm.

To be sure, my silicone tube does not do well under vacuum --- it collapses.
However, looking more closely at the matter, one notices that collapsed
does not mean completely shut --- rather, while the tubing is flattened, near
the ends, where it is most bent, there remain small channels:


Of course, the cross-sectional area in this case is ridiculously small so,
while this would work in a pump, it would also be preposterously
inefficient. Rather, one needs tubing with a much thicker wall relative
to the inner diameter..

When I get a chance. I'll have to sit down and do some elasticity theory
to work out how the cross-sectional area depends on applied pressure,
in particular look at the critical pressure where the tubing shuts off.
Basically, there are two pressures involved here, the pressure of the
rollers and the atmospheric pressure. The former pressure, of course,
should be set equal to or slightly larger than the critical pressure. Then
the question becomes what is a good wall thickness for a given tubing
material so that the area of the tubing under atmospheric pressure is
reasonably big while the critical pressure is not too large.

[Edited on 23-1-2008 by microcosmicus]

jpsmith123 - 31-1-2008 at 17:53

Here's an interesting pump design:

unionised - 1-2-2008 at 05:22
gives a design for a peristaltic vac pump.
Incidentally, the last time I looked, turbomolecular pumps were mechanical and could get well below 1mTorr

indigofuzzy - 16-5-2008 at 01:41

Today, I successfully built and used my first peristaltic vacuum pump. I used aquarium air hoses and an aquarium check valve, routing the output gas into a tub of mineral oil. I also put some mineral oil in the main pump tube to aid in sealing. I was able to pull enough vacuum on small glass tubes (about the size of miniature holiday tree lights, actually, that was the first tube I pulled a vacuum on) to see purple plasma arcing around with a few kV thrown in. I'll post some pictures when I can convince my roommate to play cameraman. I've been using a tiny solid state tesla coil as a power source, and need my fingers to be on the glass to pull the arcs, so I really don't feel like holding my digital camera in one hand and a few kV in the other :o

Trouble is, the tubing collapses under the weight of the atmosphere before I can get down to a low enough pressure to see a discharge without touching the glass.

ShadowWarrior4444 - 16-5-2008 at 02:05

Originally posted by indigofuzzy
Trouble is, the tubing collapses under the weight of the atmosphere before I can get down to a low enough pressure to see a discharge without touching the glass.

Thick, braided low pressure tubing--should be available from your nearest large hardware store (Lowes, Home Depot, for those in the US.) I seem to recall the clear vinyl ones being quite useful, though you may want something more resistant if you’re using them near plasma. (The clear ones are perhaps... vinyl acetate? I didn't dig too deeply.)

indigofuzzy - 17-5-2008 at 14:50

Rarefied air makes such lovely discharges!

And a closeup of the discharge:

[edit] and now, I'm officially a hazard to others! yippee!

[Edited on 5.17.2008 by indigofuzzy]

vulture - 18-5-2008 at 00:58

Can we keep this on topic?

ShadowWarrior4444 - 18-5-2008 at 01:11

Originally posted by vulture
Can we keep this on topic?


That wasn't vacuum tubing in those pictures. You're grievous acts of heresy have not gone unnoticed. It is quite important to use vacuum-tolerant equipment while employing a vacuum, or you may have a nasty surprise when your discharge tube repressurizes or your chemical vapor deposition chamber happily explodes throwing glass and diamond dust all about the lab.

That said, have you tested how high a vacuum you can achieve with the pump? And, does the collapse of the tube prevent further depressurization?

[Edited on 5-18-2008 by ShadowWarrior4444]

bio2 - 18-5-2008 at 11:05

Nalgene tubing recommended for peristaltic pumps. Couldn't find a vacuum rating for this.

NALGENE 280 PUR Tubing

Materials: ester-grade

NALGENE 280 Ester-based Tubing is an extremely tough product featuring excellent abrasion resistance, low-temperature resilience and flex-fatigue resistance. Pure polyurethane-contains no plasticizers and low levels of extracables, which make it ideal for high-purity applications. Resistant to atmospheric ozone, aliphatic hydrocarbons and petroleum products. Higher physical properties than PVC products (tensile strength, tear resistance, elongation, etc.) makes it

ideal for peristaltic pump applications.

Not recommended for aqueous solutions. Note: NALGENE PUR tubing is not autoclavable, but can be gas-sterilized.

indigofuzzy - 18-5-2008 at 19:27

@vulture: are you referring to my pictures? I apologize if that was off topic, my intent was to share the results of constructing a peristaltic vacuum pump, by showing that I had achieved a low enough pressure to support a continuous discharge.

IrC - 19-5-2008 at 18:04

"To be honest i have no idea what type of tubing should be used for good results."

Tygon would be the best of the choices I saw on the wiki page mentioned.

"It seems to me that a peristaltic pump should be capable of achieving a fairly low ultimate pressure...albeit at a very low throughput"

I do not think so, I imagine this design of pump would be very poor for high vacuums. Two stage pumps using high quality vacuum oil is a much better idea for basic roughing. One more point, having put considerable time and money into a good vacuum station myself, you must use copper tubing or steel brake line with flare fittings or you will never hold even a remotely decent vacuum, trust me on this one. A flaring tool kit and high quality tubing cutters is a must have! You will find it is very quick and easy to plumb for whatever experiment you are doing with say 1/8 inch Cu tubing and when you get good at flares leaks will not even be a problem.

One station I built uses a gast diaphram pump for the quick pump to 18 to 20 inches of Hg, then a oil rotary surplus medical pump gets down to 28 inches. It likely used to get down to 100 microns or so but being surplus the seals are in need of repair. After this I shut off the valve to the above mentioned pair of pumps and turn on one I bought new from frostee freeze or whatever the ebay name was for an AC vacuum pump used in AC vacuum work. This goes to 70 microns or less, and when it has been on a short while I open the valve to this one. All 3 pumps have a common high vacuum tank about a foot in diameter with valving and guages to all 3 pumps. The next pump is a 4500 dollar new scientific pump which is oil 2 stage, and this gets me down to a micron or lower. Below this I still cannot get as money has been too low for a while to have enough for a super exotic molecular type pump. Maybe someday.

A test I did, I tried various tubings and fittings with my reservoir tank held at 100 microns, and in all cases the vacuum did not last real long, whereas with my flare fittings and metal tubing I pumped the tank down to 100 microns just before I moved from Arkansas to Arizona. A year went by before I even had any experiments to do using the pumps, and the guage was still sitting on 100 microns. After a year I could see no difference in vacuum in the tank. Yes I let the air in a bit to make sure the guage was not just stuck or something, pumped it back down to 100 and shut the valve. A year after this I moved back east again, having never needed the pump for any work I was doing, and you guessed it after reading this thread I went and looked and sure enough it still sits on 100 microns. Metal tubing with flare fittings rule for doing high vacuum work in my opinion.

indigofuzzy - 20-5-2008 at 08:27


That said, have you tested how high a vacuum you can achieve with the pump? And, does the collapse of the tube prevent further depressurization?

Sadly, I don't have a vacuum gauge to test that with. All I can state is the length of discharge I can get from my mini Tesla coil. (And my volt meter doesn't read kilovolts, nor do I have hefty enough resistors for a proper voltage divider.) What I know is this: the best vacuum I've gotten, evacuating a 5cm diameter spherical lightbulb, was sufficient for diffuse violet plasma streamers to extend to the glass in response to touch. Without touching the glass, there was a small (millimeter or so) ball of violet plasma around one electrode.

indigofuzzy - 28-7-2008 at 23:12

Ok, a few weeks ago, i acquired a hand operated vacuum pump with a gauge. Sadly, the gauge is a bit miscalibrated, and reads atmospheric pressure as -60mmHg. My peristaltic pump will drag the needle to just below -760mmHg, so I'm guestimating the final pressure at around 50mmHg.

Now, IIRC, an aspirator using room temperature H20 will pull more vacuum than that. I think Evil_Lurker had some positive experiences with a waterbed draining pump.

Tacho - 21-1-2009 at 13:37

The attached pdf describes the development of a peristaltic vacuum pump that can reach 5 x 10-2 Torr. That is a pretty impressive number for this kind of pump.

At some point they suggest a possible improvement by using tubes with thinner walls imersed in a fluid under negative pressure to prevent colapsing under vacuum. I think that negative pressure is not even necessary. Merely imersing the tube in an hermetically sealed fluid would prevent the walls from colapsing.

Attachment: Peristaltic001.pdf (662kB)
This file has been downloaded 1686 times

Swede - 25-1-2009 at 05:29

I picked up a pair of pump drives off eBay for nearly nothing, and a pair of good Cole-Palmer pump heads. I've tried a number of quality St. Gobain Tygon plastics, and the problem has already been stated.... if you set the pump head occlusion up such that you obtain a full seal between rollers, the forces on the pump rollers are quite high, assuming you are using tubing stiff enough to handle vacuum. I think the best option is to have the tubing INSIDE THE PUMP HEAD be as short as possible, and couple the stub leading out to a proper vacuum line, like braided vinyl, perhaps.

I think it can be done, but how low it can go, I have no idea. The best chance for a harder vacuum is a smaller pump with its associated smaller tubing. Much slower, of course, but more able to handle the necessary forces.

A peristaltic pump is a handy beast. Even if it cannot create a vacuum hard enough for one's purposes, it'll always find use in some other project. And peristaltic pumps that retail for $600+ can be had surplus for $20 to $40. Good deals all around.

Daisy-chaining pumps? Might be the key and is an interesting concept.