Materials in Vacuum FAQ
The Kurt J. Lesker Company's TechInfo Department offers a unique, free service. We try to answer any vacuum technology question posed by anyone. If you have a question contact email@example.com
Answers to previous questions on Materials in Vacuum:
- Using Brass in a High Vacuum System
- Trapping Fluorine
- Thermal Conductivity of Silver-Filled Paste
- Cycling a Viewport's Temperature
- Thread-Seal When Transferring Oxygen
- Permeation of Gases Through O-Rings
- Outgassing of Plastics
- Gamma Irradiation of Polymers
- Insulating Thermocouple Leads
- Lubrication at High Vacuum
- Gravity in a Vacuum -- Part 1
- Gravity in a Vacuum -- Part 2
- Emmisivity of Stainless Chamber
- Electrically Conducting Lubricants
- Trapping Backstreaming Oil Vapor
I wish to put brass in a high vacuum (1x10-6 torr) evaporator system. We have a sample stage that we want to use while evaporating metals. However, the stage is made of brass and I've heard that zinc has a low vapor pressure and will never stop outgassing. Is this true or a myth? If true, what can I do to keep the zinc in the brass, enamel paint?
The statement that brass should not be used in high vac systems is substantially true but, as with most such statements, shouldn't be taken as 'categorical imperative.'
The vapor pressure of zinc in brass is fairly high compared to, say, stainless steel but not unacceptable for a system with a base pressure of 1x10-6 torr. As you can see zinc's curve in Rick Honig's vapor pressure curves shown at:
The VP of zinc at ~50°C is less than 10-11 torr.
However, if you heated the brass to 200°C, the zinc's VP would rise to something in the 10-6 torr range at the surface of the brass. That means a monolayer would be put down on all the "line of sight" surfaces every few minutes. Not a pleasant thought!
I use a molecular sieve trap in a reactor for depositing thin films. Recently, the XPS results showed Fluorine content in the films, likely coming from the pump oil. Could I use one of your products, such as Zeolite, to trap fluorine?
Is the pump filled with a PFPE oil (such as Fomblin or Krytox) or a PFC oil (such as Halocarbon)? If so, then this strongly suggests the existing trap has "broken through". That is, sufficient oil vapor has reached the trap to coat all the adsorption sites. The trap is now acting like a secondary source for oil vapor into your deposition chamber.
If this explanation is correct, it is particular bad since the chamber walls and all the fixtures will be coated with at least a monolayer of oil. Changing the trap material will only prevent more oil vapor reaching the chamber. It will do nothing for the contamination already there. For that you'll need to completely clean all internal chamber surfaces --- which is far from easy.
Thanks for the message. The pump uses Fomblin oil. I will clean the chamber with a plasma process. My question now is: what is the best product to trap the oil vapor?
I am planning to clean the chamber with plasma and install a new absorbent material in the trap.
Any trapping material such as Micromaze or molecular sieve (zeolite) with provide adequate trapping for Fomblin vapor. However, three points must be made:
1. The trap must initially activated by heating to 150-180 deg C and pumping it out for many hours.
2. The trap must be 're-activated' at intervals just as in 1.
3. To prevent the vapors that come off during re-activation from reaching the chamber (a) the trap must be removed from the system or (b) a high temperature valve must be placed between trap and chamber, close to the trap. For (b) the valve is closed before heating is started and it remains closed until the trap is again at room temperature.
We are planning to use your Silver-Filled Paste, Part No. EJAGPASTE50G for Heat Sink contact to Silicon Mirrors in a Laser beam transport system and would like more detailed information on this paste, such as thermal conduction numbers and base material used.
EJAGPASTE50G is a mixture of 40% Apiezon M grease and 60% finely divided silver. We are a re-seller of the product and the manufacture does not list a combined thermal conductivity. I can tell you that Apiezon M has a listed TC of 0.192 W.m-1.K-1 and silver (in bulk) have a TC of 430 W.m-1.K-1. I suspect the combined number will be between those two, but biased to which end? Being an arch-conservative in technical matters, my guess would be >1 but <10 W.m-1.K-1.
The vapor pressure for the product is quoted as 2 x 10-9 torr at 20°C which jibes with the quoted number for Apiezon M (1.7 x 10-9 torr)
I have a 6" conflat viewport that broke. It is on a vacuum system that cycles up to 310°C to room temperature. The heat up time to 310° C is about 30 minutes. I need to know if you have a viewport that can handle that.
I looked in the catalog and saw a 6" viewport (VPZL-600) will this be able to do the job?
The maximum temperature for viewports depends somewhat on the material of the "glass" part but mostly on the method by which the "glass" and the metal are bonded. For example, two companies well- known in the vacuum industry as suppliers of viewports both offer sapphire ports: one gives the temperature limit of 420°C for bakeout; the other states the limit is 250°C. Optically, both products are the same.
Some manufacturers differentiate between the bakeout temperature and the operating temperature, often a difference of 50°C. So, your first concern should be to make sure the viewport you selected is suitable continuous operation at 310°C.
However, from the description you give I am fairly confident your problem is related to rate of temperature rise. It is "common wisdom" in this company that a viewport must not be heated at a rate greater than 2.5 to 5°C per minute (different people here hold different common wisdom numbers!). To my surprise, a reasonable extensive search of competitors' literature and the web shows only one reference to the subject:
which is even more conservative than we are (2 - 3°C per minute). Your value of ~10°C per minute is much too high.
But we all agree that the viewport must be covered with many layers of aluminum foil during heating and cooling. It is important to avoid temperature differentials (caused by radiation, conduction, and convection) between glass at the viewport's center and rim and aluminum foil helps.
I am trying to locate a suitable thread sealant that is compatible with oxygen and suitable for our system pressures.
It is for use in our gas control system on our glass coater and the expected pressures are: During sputtering 2x10-3 Torr System base pressure 10-7 Torr and better. Another important factor is that it can not contaminate our process gasses.
I have had conflicting advice that Apiezon L could be used. Can you confirm or deny its suitability? And can you recommend any other seal.
One of the first things I learned about handling pressurize oxygen gas cylinders was "Never lubricate the regulator's threads!" The danger of fire when a 'fuel' is surrounded by 100% O2 is too great.
And here, fuel can mean something as innocuous as a cotton wool ball. One character I worked with many years ago would dip a cotton wool ball in liquid O2, pull it out, throw it on the floor and approach it with a gas lighter. The resulting explosion gave him great pleasure. It made the rest of us question his sanity.
Even if you are discussing only a very low total pressure of O2 (that is, the threads are at chamber pressure), it is still 100% partial pressure and Apiezon L is a wonderful 'fuel'. Why risk it?
My thread seal choices would be:
1. Get rid of the thread that needs sealing
if this is not possible
2. At high pressure (>1 torr) - PTFE (plumber's) tape
3. At chamber pressure (<1 millitorr) - a silicone sealant like our KL-5 (correctly cured before exposed to the oxygen) sprayed on the joint after it has been made.
For solution 3 you should seal the atmosphere surface. You don't want a large area of KL-5 inside the chamber even though its outgassing characteristics (after curing) are good. However, this leaves the thread as a wonderful virtual leak that will probably cause you pumpdown time grief.
For the various types of elastomer seals ( say for a KF flange) which has the lowest permeation rate? Is this the same as the outgassing rates? Do you have quantitative values or references to the original tests. We are generally containing helium at room temperature, though it could be as high as 65°C during a bakeout. Are there any new materials on the horizon or special order materials that have better performance than the traditional ones (e.g. Fluorocarbon, Buna-N, Neoprene, etc.)
Permeation rates and outgassing rates are different. Permeation requires a pressure differential across the o-ring. The rate is initially small (since the gas permeating from the high pressure side takes time to reach the vacuum side) but once the gradient is fully established, the permeation rate is constant.
Outgassing describes desorption of gases and vapors adsorbed on or near the o-ring's surface. The desorption rate is large to start and exponentially decays. Theoretically after an infinite time the outgassing would become zero.
Permeation rates are temperature dependent and different elastomers have different dependences. Less strongly established (but our experience suggests it is correct) is the fact that different formulations of, say, a Fluorocarbon o-rings seem to have different permeation values. There is an example of "black" and "brown" Fluorocarbon o- ring permeation given at:
A pure gas/vapor permeating through different elastomers gives different permeation rates. Different pure gases/vapors permeating through the same elastomer give different permeation rates. For example, O'Hanlon's "Users Guide to Vacuum Technology" quotes the permeation (in units of 10-12 m2/s) helium through
PTFE 570 Fluorocarbon A 8.9 Nylon 31 0.3
PTFE 20 Fluorocarbon A 2.2 Nylon 31 0.13
When trying to find a "lowest" permeation rate, you will surely run into questions of outgassing and temperature acceptability since Nylon is not a material I would chose to have in my system.
During a web search, in addition to O'Hanlon's book, I found two reference books (this is not a recommendation since I've no idea of their content):
On our website I referenced J Vac Sci Tech Vol. 17, No. 1, Jan./Feb. 1980, p.334 which has some information about permeation rates.
But in case you've missed my (unstated) overall message.... I think the applicability of any published permeation data to the o-rings you can purchase is, at best, suspect.
I need to manufacture a special high Voltage vacuum feedthru and would like to know if Delrin or polycarbonate can be used for 10-6 torr and better. Is there a vacuum grade plastic? PTFE is not structurally strong enough as it flows too easily. Macor ceramic is not made large enough and I need something quick, (not special).
All plastics outgas at much higher rates than 'vacuum compatible' metals. So, in one sense, there is no vacuum grade plastic. But if you are forced the use plastic and Delrin, Lexan, or PEEK are the one that might work mechanically, electrically, and on the time scale you need, then you have to consider what you must do to make one of them work.
Outgassing depends of many things but there are two you can control: (1) the area exposed to vacuum and (2) the component's temperature. Unfortunately, in high voltage situations, you can't just hack away at the area. There is some minimum distance to prevent flash over. But you should design to that minimum as close as you can. With plastics, the temperature limit is pretty low too, so you can't just bake then under vacuum to reduce outgassing rate.
Does this mean there's nothing you can do? Not quite. Remember, that old vacuum lore: At constant pressure
Gas in = Gas Out
That is, at constant pressure the Gas Load on the system (caused by all the internal surfaces outgassing) equals the Pump Throughput (volume of gas removed by the pump multiplied by the pressure at the pump's inlet).
Since seriously reducing the outgassing isn't possible then you must tackle the throughput. And the solution is: put huge pump (or pumps) on the system in a way that will give a huge pumping throughput. Yes, this will be expensive but I don't see any other options.
You can "calculate" roughly what size pumps you need. If you have a system that reaches 1 x 10-6 torr, look at the pumping speed of the pump (assuming it's connected to the system in a "reasonable" way following vacuum principles). This give you a baseline, you'll need that pumping speed just to handle the 'normal' gas load.
Calculate the surface area of plastic that will be exposed to the vacuum and make a stab at finding an appropriate outgassing rate. (I searched for "outgassing" and "delrin" using www.alltheweb.com and got 60 hits). Ideally, you want a rate quoted in sensible terms of "pressure.volume/unit time/unit area" but chances are you'll get something in "percent weight loss". Anyway, when you've calculated a rough gas load from the plastic surface it should be in units of "pressure.volume/unit time". Divided that by the pressure (1 x 10-6 torr) and the result is the pumping speed you'll need to achieve that pressure.
I was wondering if you had some information about the outgassing of polymers in vacuum and irradiated with gamma-irradiation..
No more specific information than is available in the literature and in the appendix of O'Hanlon's book "A User's Guide to Vacuum Technology".
There are three aspect to gamma irradiation of polymers:
a. Photo desorption will increase outgassing rates. There is an enormous amount of photo-desorption information in the literature generated by people at various accelerator and beam line facilities.
b. Some polymers degrade under ionizing radiation. That is, their average molecular weight is reduced. This will increase volatility and, hence, increase the outgassing rate with time.
c. Other polymers cross link under ionizing radiation, for example polyethylene. From such a polymer one might expect the average molecular weight to increase with time and the outgassing rate to decline slightly as a result.
I'm looking to insulate some thermocouples (K) and some power lines (110VAC, 15Amp) inside a vacuum chamber. My questions are:
1) What is the standard flexible insulation for a thermocouple? Both thermocouples in the chamber need to flex/move every time the chamber is opened.
2) What is the standard method/material to insulate power lines inside a vacuum chamber? Two of my lines are stationary. Two of my lines flex each time the chamber is opened.
1. the temperature the T/C in measuring
2. how close to the measurement zone the T/C must be insulated
3. the vacuum level required
there are probably three different types of "flexible" insulation that are useful in vacuum. In order of temperature acceptance, they are: PTFE or polyimide; woven glass (or mineral wool) fiber; fish-spine beads. For rough to high vacuum, the order changes to: glass fiber; PTFE; fish-spine beads.
Since no T/Cs I know are multi-stranded wires, I don't have to warn you about the bad outgassing characteristics of a multi-strand coated with a close fitting plastic insulator (like PTFE). But for low power leads you do need to know that. So, given bare, single strand power wires, I'd look at using glass fiber or fish-spine beads.
But remember, a single strand wire doesn't have nearly the repeated flexibility of a multi-strand capable of carrying the same current. Think about having a routine schedule of replacing any single strand power lines and any thermocouple leads that have to flex.
I have a translation stage in a large SS vacuum chamber which does not work well without some lubrication. It can be re-engineered to work without lubrication, but, I would prefer not to do this. I've heard, and read on the web, that Apiezon greases are very good for vacuum service. I plan to reach a vacuum of 10-8 Torr (The vacuum chamber has all conflat flanges except one port which attaches to a beam line sealed with Fluorocarbon o-rings.) which may necessitate a soft bake. I have two questions:
First, despite the low vapor pressure of the Apiezon grease, will it migrate on surfaces and slowly contaminate the vacuum chamber and ultimately the vacuum pumps. I have a new turbo/scroll combination which I would like to keep clean for as long as possible. We may eventually use these pumps in applications which would be sensitive to hydrocarbon contamination.
Second, what happens to the grease as the temperature is raised? H grease has an operating range to 240°C, but, will the vapor pressure increase substantially at that temperature and then migrate to other surfaces? I don't expect baking the chamber to greater than 150°C.
While I don't have VP curves for the Apiezon greases, the VP of any material increases with temperature. If you look at the Honig's VP's of the elements:
You will see (very roughly) they all have the same slope. From time to time, using other sources, I've plotted hydrocarbons VP curves and found they too have the same slope. From
using Cd as my 'model', the VP changes from ~1 x 10-11 at room temp to ~1 x 10-5 torr at 150°C. That is, 6 orders of magnitude for ~130degC rise. I'd expect hydrocarbon greases to start higher than 1 x 10-11 torr at room temp and be correspondingly higher at 150°C. Bottom line, if you don't want hydrocarbon grease contaminating every surface then don't lubricate the manipulator with it.
So, what can you do? Typically, we recommend a dry lubricant like moly disulfide, tungsten disulphide, or graphite. There are a variety of ways of applying such a coating (which I can't help you with) ranging from application of dry powder to sputtering or plasma spray coating. You should be able to find something on the web if you want to go that route.
Does a vacuum produce a microgravity environment?
I'm not sure I understand the question. But if you are asking does gravity differs inside and outside a vacuum chamber the answer is definitely NO!
As proof, go outside and look at the sun (or the moon). There's not much of anything between those bodies and us except a pretty good vacuum. Yet we are still circling (and being circled by) after all those billions of years, thanks to gravity.
How would I calculate the speed of a particle in vacuum at, say 10-6 torr at room temperature?
A particle (gas atom or molecule) of given mass (at a given temperature) will be somewhere on its Maxwell-Boltzmann velocity distribution curve for that temperature.
There are a variety of velocity "averages" one can calculate:
1. most probably velocity given by 12,895 x (T/M)0.5 cm.sec-1
2. arithmetical average velocity given by 14,551 x (T/M)0.5 cm.sec-1
3. root-mean-square velocity given by 15,794 x (T/M)0.5 cm.sec-1
Note that the pressure is not a consideration in the distribution function. At a fixed temperature, it doesn't matter if you are discussing 10-6 torr or atmospheric pressure, the velocity distribution is the same.
I originally asked the "speed of molecules in vacuum" question because someone here was suggesting that gravity had a major effect in vacuum (ie pumps on top of the system would be less efficient than on the bottom). I was hoping to use the speed of the molecules to show that gravity is not an issue. Can you give a simple explanation as to why gravitational effects are not a consideration in vacuum (all I can think of now is entropy... the fact that if all the gas atoms "fell" to the bottom of the chamber etc...). As you can tell this discussion is getting out of hand in house so I would like to close it!
The guy who suggested gravity has a big effect of vacuum is, in scientific parlance, talking out the back of his pith helmet.
There may be a number of ways to approach this the subject. Let me start with the premise --- gravity causes a homogenous gas mixture to spontaneously stratify and separate into it's components. Then some relevant questions are:
1. Why do we bother with ionization and electric/magnetic fields to make something called a mass spectrometer?
2. Why aren't we all dead from carbon dioxide poisoning from atmospheric stratification (particularly since we and flora generate most of the CO2 at sea level)?
3. How could Boyle's law (P1V1 = P2V2) work?
4. But the real kicker for me is: if you "eyeball" the Maxwell Boltzmann distribution curves for xenon, nitrogen, and helium and determined the average velocity and then calculate the average kinetic energy (1/2.m.v^2), you'll get something like this:
Gas Mass (g/mole) Velocity (m/sec) Kinetic Energy Xenon 131 250 8.2x106 Carbon Dioxide 44 325 4.6x106 Nitrogen 28 400 4.5x106 Helium 4 1175 5.5x106
I'm pretty certain, if I'd used the 'correct' average and been a little more careful, I would have found the average KE of ALL gases is identical (I think that's either a starting premise or a conclusion of Kinetic Theory). So, how do heavier molecules get to sink to the container's bottom when all are being "stirred up" by equal violent collisions no matter what their neighbors' chemistry?
I suggest you declare this subject "off limits" until your chief antagonist has read, learned, and can parrot back without peeking the Kinetic Theory of Gases (Chapter 1) of Lafferty's Foundations of Vacuum Science and Technology.
We have a custom vacuum chamber from you, a small cube constructed from your standard stainless steel. For our experiment, it would be useful to know the thermal emissivity of the interior surface of the chamber. Would you have this data?
Thermal emissivity of stainless steel (and, come to think of it, every other metal) varies depending on the surface's mechanical finish, its chemistry, and its temperature.
I've seen numbers for 18-8 stainless (type 304L we use for chambers would fall into this category) quoted as:
Condition Emissivity Buffed 0.16 Sandblasted 0.44 Oxidized (800°C) 0.85
and for type 316 (also an 18-8 stainless)
Condition Emissivity Polished (24°C) 0.28 Polished (232°C) 0.57 Polished (1740°C) 0.66
I have no idea of the "authority" of these numbers, who did the work, or how valid their techniques.
Typically, if I have thermal transfer questions and need emissivities, I look on the web for the largest and smallest numbers, do the calculation at both extremities and take the result that's worst case for whatever I'm doing.
Sorry to be so uninformative.
We are designing an electrical contact system for high altitude aerospace application, desire is to find a contact lubricant that has low/no outgassing, and has low electrical resistance for sliding metal to metal contact. Any suggestions would be sincerely appreciated.
I don't think I have much information that will help. For example, what little I know about non-lubricated sliding surfaces suggests the wear can only be worse in a vacuum environment.
Organic based lubricants (and I include Telfon with this group) have a vapor pressure above, say, 10-9 torr at normal temperatures and/or high resistivities.
Sulfide-based lubricants (such as MoS2 and WS2) have low vapor pressures but still high resistivities.
Graphite might fit your requirements but I suspect the prospect of graphite particles near a space device will be frowned on.
Diamond-like carbon films have excellent friction, wear and thermal conductivity properties but high resistance.
The only materials I could suggest (and I have no idea about their commercial availability) are carbon nanotubes (the tubular analog to the C60 fullerene spheres and C70 'footballs'). If relative short nanotubes could be clamped or grown on the end of a stationary electrode and it is placed close enough to the moving electrode for the nanotubes to make contact, I think vapor pressure, wear, and electrical conductance would not be an issue. What would matter, of course, are factors such as power, voltage, and clamping the nanotubes in the first place..
I have a Stokes 013-2 Rotary vane pump with a Molecular Sieve trap mounted directly above the inlet port. What absorbant material would you recommend? The mechanical pump is on only for rough down of the chamber 5 times/day about 6 minutes/time and for regen of the 10" cryo pump once / month. The system is backfilled with argon during processing and pumped into the cryo; nitrogen is used for purge gas during venting.
The worst oil vapor backstreaming from a mechanical pump occurs when the foreline is at a low pressure and the flow of gas from the chamber is insignificant. So, your operating conditions are quite favorable for long life in the trap. Probably the worst backstreaming will occur during cryo regen if you are not using a nitrogen purge (which is a good idea for other reasons).
While there are thousands of different mol sieves available with different pore sizes and structure, the one most often used to trap oil vapor is called by most vacuum companies "13X".
Remember, just filling the trap with fresh sieve doesn't do much for you since it is usually loaded with water. To "regenerate" the sieve you should bake it up to say 200°C in a vacuum oven for some hours. If the trap has a heater, you can do an in situ bake but the trap **must** be isolated from the chamber or the oil vapor will distill in that direction too.