Applied Science and Convergence Technology 2017; 26(5): 95-109
Published online September 30, 2017
Copyright © The Korean Vacuum Society.
Rebecca Grinham*, and Andrew Chew
Edwards Global Technology Centre, Innovation Drive, Burgess Hill, UK RH15 9TW
Correspondence to:E-mail: email@example.com
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
There are several contributions to the gas load of a system of which often the most important is outgassing. Adsorption occurs via two main processes, physisorption and chemisorption, and can be described using five (or six) classifying isotherms. Outgassing is the result of desorption of previously adsorbed molecules, bulk diffusion, permeation and vapourisation. Looking at the desorption rate, pumping speed and readsorption on surfaces, the net outgassing of the system can be calculated. There is significant variation in measured outgassing rates between different materials but also between published rates for the same materials, in part due to the number of different methods used to measure outgassing. This article aims to review the outgassing process, outgassing rates, measurement methods and techniques that can be used to reduce the outgassing of a system.
Keywords: Outgassing, Materials, Surfaces, Cleanliness, Particle-free
When pumping down a chamber below atmospheric pressure and operating under UHV conditions, one of the most important considerations is the gas load of the system. This includes contributions from multiple sources and will affect the pumping speed, time taken to evacuate the chamber and the base pressure which can be reached .
Therefore it is important to have a good understanding of how the materials used are likely to behave under vacuum and where possible take steps to reduce the outgassing rate.
Despite being an important consideration in the design of vacuum systems, outgassing is a topic which is often given only a very basic treatment. This can be attributed to several factors: measuring outgassing rates can be difficult, particularly for complex chamber designs incorporating multiple materials, rates available in literature are often widely varying for the same material and commonly used reference values are often forty years or more old. The exceptions to this are the aeronautical/space industry where outgassing rates of over 10,000 materials have been measured and recorded by NASA  and the extreme high vacuum systems used, for example, in high energy physics experiments where outgassing of materials is usually given high importance and material outgassing rates are carefully measured. Unfortunately, much of this data is either extremely specific to a particular application, for example sealing materials or adhesives, or unpublished as facilities often conduct specific outgassing trials for their vacuum system configuration which are not considered applicable to other applications.
There are five main contributions to the gas load within a system (as depicted in Fig. 1). These are:
1. Initial gas
All chambers will initially be at atmospheric pressure, requiring pumping of the enclosed gas to reach the required pressure. This will initially be the dominant contribution to gas load but is only dependent on the amount and nature of gas, temperature of the system, required pressure and pumping speed.
2. Process gas load
Sometimes additional material (usually gas) will be introduced once the required pressure has been reached. If present, this often becomes the dominant contribution to gas load but can be accounted for as the gas species and flow rate will be known.
An ideal system would have no leaks, however often both real and virtual leaks will be present and can add to the gas load. Real leaks are defects in the chamber wall, seals etc., allowing molecules from outside the system to enter whilst virtual leaks are due to trapped volumes within the chamber stopping gas being removed during initial pumpdown but allowing its release later, giving the appearance of a leak.
Back-streaming and migration are processes by which material from the pump ‘travels’ backwards into the chamber, adding to the gas load. The effect of these processes on the total gas load is usually minimal and can be minimised using cold traps, baffles, anti-suck-back valves etc.
Outgassing can often be the most important contribution to gas load and the hardest to estimate and counteract. It is the combination of several mechanisms causing gas molecules to be released from the surface of the chamber and generally also includes permeation of gas from outside the chamber into the volume.
There are four main contributions to outgassing :
Release of molecules from the surface material itself
2. Desorption (thermal or stimulated/induced e.g. by charged particles or photons)
Release of gas molecules adsorbed to the surface
Release of gas molecules from within the bulk of the material
Release of gas molecules from outside the system through the bulk
These are shown in Fig. 2. The relative importance of each will depend on the chamber surface material, treatment and condition as well as factors such as chamber temperature and humidity and previous exposure to vacuum.
The gas load of the system has a significant effect on its ability to reach the required pressure as well as the pumping speed necessary. Since:
For the average system, outgassing will be the largest contribution to gas load at around 90% of the total for a system with a base pressure in the high vacuum range or lower .
For the initial stages of pumpdown where the original bulk gas is the limiting factor pressure decays exponentially, according to the pump rate equation, as
For a system at low pressure, when the mean free path of molecules is much longer than the dimensions of the chamber, meaning that molecules are more likely to collide with the walls than each other and the physics of outgassing is governed by surface interactions.
The gases that leave the surfaces within the chamber are, in most cases, ones that became attached there earlier via a sorption process. It is key to clarify the difference between two sorption processes, absorption and adsorption.
Absorption is a physical or chemical process by which a substance enters into the bulk of a second substance, usually in a different phase (gas molecules into the solid chamber wall).
Adsorption is an accumulation of molecules from the gas phase on the surface of a second substance (the chamber wall). Adsorbed molecules can simply desorb, whilst absorbed molecules must first diffuse to the surface.
Of these, adsorption is the most important with regards to outgassing as surface molecules are more easily removed. Adsorption is categorised into one of two processes, depending on the predominant form of interaction with the surface.
Physisorption is a process by which molecules are attracted to the surface due to the Van der Waals force between them and the surface molecules. Since this bonding is via temporary or induced dipole attraction and there is no change to the electronic structure of the molecules it is weak, with a heat of adsorption less than 0.4 eV (1–40 KJ/mol). The potential curve for physisorption is shown in Fig. 3. There is generally no barrier to this form of adsorption and it is a fast and easily reversible process.
Chemisorption is the process by which chemical bonds are formed between a gas molecule and the surface. The exact bonding mechanism will depend on the species of the adsorbate and adsorptive but bonds are much stronger than physisorption with a heat of adsorption of 0.4–10 eV (40–800 KJ/mol). An important distinction is between spontaneous and activated. The activated case is shown in Fig. 3; there is a potential barrier that must be overcome before the molecules can approach close enough to bond and fall into the chemisorption potential well. In the spontaneous case the physisorption and chemisorption curves cross below the zero in potential energy and therefore there is no barrier to overcome. A comparison of these two adsorption mechanisms is given in Table 1.
At some stage during the adsorption process, an equilibrium is reached between the number of molecules adsorbed on the surface and remaining in the gas phase. This equilibrium will depend on factors such as the relative stability of the gas species, the temperature of the system and the pressure of the gas .
The relation between the amount of material adsorbed and the equilibrium pressure (at constant temperature) is known as the adsorption isotherm . There are many commonly used isotherms such as Langmuir [7,8], Frendlich , Temkin and BET theory  a selection of which are shown in Figure 4 . These isotherms are mostly semi-empirical and differ in the way they approach adsorption; single or multiple layers, nonporous or micropores, capillary condensation, interactions between adsorbed molecules, surface coverage etc.
Desorption is the process by which material is released from a surface to the gas phase. As can be seen in Fig. 3, there will often be an energy barrier,
It is of interest to note the mean stay time (sojourn time) for a particular molecule on a particular surface. This is a measure of how likely desorption is to occur and can vary from nanoseconds to longer than the lifetime of the universe (some examples are given in Table 2). At the extremes of this range molecules will either desorb effectively instantly (weakly physisorbed rare gases, oxygen or nitrogen) or remain in place with effectively infinite surface lifetime (strongly chemisorbed oxygen, nitrogen or halogens) and therefore only add a negligible contribution to outgassing rate.
Generally, outgassing becomes a problem for gases in the intermediate range, with desorption energies of around 15–25 kcal/mol. The desorption energies for water on stainless steel and aluminium are in the range of 19–23 kcal/mol hence why water makes up the major contribution to outgassing .
Other mechanisms affecting desorption are bulk diffusion and permeation. For diffusion of molecules within the bulk of the material the important factor is the jump frequency, the rate at which atoms move from one site to another, which varies exponentially with the diffusion barrier,
Permeation requires dissolution of external gas molecules into the surface and diffusion through the bulk (usually rate defining). It is important as it can introduce molecules to the system that were not present in the original gas or materials. Permeation is strongly dependent on the type of material used. For example, diffusion through polymers is much higher than metals and almost all gases can permeate through polymers whereas at room temperature only hydrogen can permeate through metals such as aluminium or stainless steel .
An example of the energy barriers for all these processes is shown in Fig. 5 including dissociation and recombination, where a molecule must split into its component atoms before sorption or recombine upon desorption.
When calculating the net outgassing rate for a system there are three main components to consider: the gas removed from the system by the pump, the gas desorbing from the surfaces and the gas being re-adsorbed. A thorough treatment combining these contributions leads to two equations which must be satisfied, the first for the volume and the second for the surface. These lead to a single differential equation for the density of molecules in the gas phase :
A reasonable approximation of the outgassing flow rate is given by :
||clean metal surfaces|
||desorption from surfaces|
||metals, glasses and ceramics|
||highly porous surfaces|
||diffusion controlled outgassing from the bulk|
Although for most applications water vapour will be the major concern in terms of outgassing, the contributions of different species to outgassing will depend on the pressure range as shown in Table 3. In particular, for metal systems where very low pressures are required hydrogen outgassing will be the limiting factor in the base pressure of the system.
Outgassing rates for different materials vary by more than nine orders of magnitude and therefore it is critical to consider this when choosing materials to be part of a vacuum system. Generally metals and glasses have much lower outgassing rates than polymers and also larger decay constants. Outgassing rates over time for a selection of common materials are shown in Fig. 6.
Outgassing rates across various literature sources show very large variation, in some cases several orders of magnitude for the same material. A summary of average outgassing rates from the most commonly referenced and reliable literature sources is given in Table 4. Variation in measured rates can be attributed to three primary causes; the manufacturing process, surface treatments etc. used in producing the material, and the method used to measure the outgassing rate. Secondary effects include variations in local atmospheric conditions, elapsed time and pressure at the start of measurement, poor knowledge or estimates of pumping speed/conductance, errors due to gauge pumping/outgassing/accuracy and measuring total/partial pressure.
An example of the variations in measured rates of the same material from multiple sources is shown in Fig. 7. Here measurements by various authors are compared for the same material (stainless steel) with several surface treatments. Initial outgassing rates vary by up to five decades. The inconsistency displayed in published outgassing rates often leads to system design which is conservative rather than optimal.
As has been demonstrated, the method used to measure outgassing rate can have a significant effect on the results. It is often difficult to get accurate data on outgassing rates for materials due to variation in exact surface structure, even when the same material composition, preparation and surface treatment are used. For this reason is it important that when measuring outgassing rates, a robust method is used so that values for rates are as accurate and reproducible as possible.
There have been many methods used for measuring outgassing rates and a selection of these have been defined as recommended practices by the American Vacuum Society . As summary of key attributes for each method are presented in Table 5.
The choice of measurement method will depend on the specifics of the systems and the material being measured. There are several popular methods:
Measured effective pumping speed throughput
Conductance modulation throughput
Rate of rise
Rate of rise and throughput
Of these the throughput and rate of rise methods are the most often used.
For all methods the outgassing rate is measured once the pressure in the chamber is below 7.5 × 10−3 Torr, ensuring it is within the range accurately measurable with a single vacuum gauge. For all methods, if the outgassing of a sample is being measured, a second chamber only measurement is needed to correct for background outgassing. Outgassing is usually expressed on a log-log plot of rate against time or as a rate after a particular time interval (1, 10, 100... hours). The exponential decay constant,
To ensure measurements are as accurate as possible the vacuum system should be all metal and in particular, elastomers should only be used for valve seatings which are not exposed to air. The pumps used must be reliable and not change pumping speed over the measurement period; turbomolecular, diffusion or cryogenic pumps are recommended. All gauges used should be as accurate as possible and recently calibrated . Recently, outgassing reference samples have become available commercially which can be used to calibrate the measurement system.
This method uses a simple set-up and is most often used for measuring the outgassing rate of a sample of a material. The sample is placed in a vacuum chamber separated from the pump by a restriction with a known conductance as shown in Fig. 8. Gauges, usually ionisation type or residual gas analysers (RGAs) for greater accuracy, are placed either side of the restriction to measure pressure.
If the outgassing of the chamber itself is being measured, then pressure is measured as the system is pumped down to its base pressure. If the outgassing of a sample is being measured then this must be done twice; once with the sample and once without to measure the background outgassing rate.
Since the pressure on the outlet side of the restriction is generally much less than on the inlet side and conductance is much lower than the speed of the pump −
There are several variations on this method such as the measured effective pumping speed throughput method which measures the speed of the pump and uses this to calculate the outgassing rate rather than the conductance. The method used is the same except for the introduction of a known leak rate of nitrogen,
Another method similar to the throughput method is the two-path method. In this, there are two paths from the sample chamber one with its outlet before the restriction and one after. By switching between these the outgassing rate can be found from the difference in pressures between the two,
This method is most commonly used for measuring very low outgassing rates as it removes the errors due to outgassing from other surfaces by using the difference between two measurements.
The conductance modulation throughput method is a variation which aims to remove error due to pumping speed and gauges. A variable orifice is used and pressure measured as for the throughput method with two different conductances (usually one much larger than the other). Outgassing rate is calculated by :
This method can also be used to calculate the pumping speed (extrapolating for 1/
This method is also known as the pressure rise or gas accumulation method. It is a simple system, as shown in Fig. 9, in which the sample is held within a chamber which is pumped down and isolated. As the sample outgasses, the pressure within the chamber increases and outgassing rate can be calculated by:
This method is most often used for samples expected to have high outgassing rates. For reliable results, the process should be repeated several times following a reproducible procedure in which the pressure shouldn’t be allowed to exceed three times its initial value before the isolation valve is opened. An example trace of this is shown in Fig. 10 where the green dashed line shows an extrapolation of the chamber pumpdown if the isolation valve were not closed, the red dashed line shows an extrapolation of the pressure rise after isolation due to outgassing, and the blue dashed line shows an extrapolation of the initial linear increase in pressure with time after isolation. The black line shows an example of the pressure trace achieved with multiple measurement cycles.
There are many options available to reduce the outgassing in a vacuum system relating to four aspects; choice of materials, cleaning and handling, surface treatment and baking. All the techniques outlined here amount to two general techniques: temporarily stimulate desorption to remove as many adsorbed molecules as possible prior to use or create a barrier on the surface to prevent adsorbed molecules desorbing during operation. A summary of all the outgassing reduction methods is given in Table 6.
The simplest way to reduce outgassing is by making an appropriate selection of materials when constructing the system. As has been shown, there is significant variation of outgassing rate with material type however there are other factors which must be taken into account such as mechanical properties (ability to be machined/fabricated, strength, retention of properties such as elasticity/fluidity at extremes of temperature), thermal properties (vapour pressure, expansion), gas loading (non-porous/no surface features to trap material and cause virtual leaks) as well as the cost, magnetic permeability, reactivity/volatility, radiation and corrosion resistance and many others . Often the choice of material must be a compromise between these factors depending on the needs of the application.
Generally, metals have lower outgassing rates than other materials and should be used preferentially with elastomers, oils and greases avoided wherever possible. Of metals, stainless steel is most commonly used in vacuum systems as it has many desirable properties including corrosion resistance, reasonable cost, simple manufacture and is bakeable to high temperature (>1000 k) alongside a low outgassing rate. For similar reasons, aluminium and copper are also popular choices for use in vacuum. If polymers, elastomers etc. must be used in the system (for example in O-rings) then a fluoroelastomer should be used as they have lower outgassing rates than nitriles.
It has also been shown that for metals (in particular, stainless steel) a significantly lower outgassing rate can be achieved by using thinner material as this gives a smaller bulk volume for diffusion and permeation and additionally allows reduced baking times. However, if especially thin walls are used this will be detrimental to the mechanical strength of the system .
Proper material preparation is absolutely vital to achieve low outgassing rates and reach UHV. Generally, the cleaning process will follow a series of steps :
Remove gross contaminants such as grease or paint
Remove fine contaminants such as oils, cutting lubricant and adsorbed species such as water vapour
(For most metals) Remove hydrogen dissolved in the bulk
The most basic cleaning procedure is washing. This is often uses a series stages such as hot water, deionised water, ultrasonic, detergent and combinations of these [20, 21]. Another simple method is vapour degreasing which is most often used for small components. The part is placed in a chamber where solvent is heated till it vapourises and condenses on the cool surface of the part. This method is effective at removing greases but does not clean inaccessible areas well.
Another basic cleaning technique is blowing. This technique uses a jet of compressed air, nitrogen or other inert gas to knock particulate from the surface and is especially useful for rough surfaces or recessed/confined areas. Similar techniques which use small particulate pellets rather than gas are bead blasting (aka shot blasting) and snow cleaning.
In bead blasting, the jet is comprised of small beads of alumina or silica (contained within air or water) which physically remove material from the surface. This method is rough and can remove unwanted surface coatings. Bead blasting has been shown to be an effective method of reducing outgassing with rates lower than H2 firing, electropolishing and in some cases chemical cleaning .
Snow cleaning uses a jet of small dry carbon dioxide pellets (crystals) within a gas or liquid carrier. In this method, the snow causes surface adhesion to be overcome, allowing the particles to be removed by the high velocity gas. Liquid CO2 also has good solvent properties, meaning it can effectively remove hydrocarbons from the surface. This method has the advantages that it doesn’t cause any damage to the surface and is residue free [23,24].
Chemical cleaning is generally used in addition to other techniques and generally is comprised of an alkaline clean, a water rinse and an acid clean. The alkaline clean takes place at a temperature of between 60–90°C where the pH is around 10–11. The material remains in the alkali for around 15 minutes. The acid dip is usually short, lasting from 20 second to a couple of minutes, typically using a mix of chemical including H2O, HCl, HNO3 and HF. This technique is most effective when used on rough surfaces where many contaminants such as oxygen, carbon, phosphorus and chlorine can be held.
In this procedure, gases that are either oxidising (such as O2 and NO) or reducing (such as H2 and NH3) are flowed over the surface where they react with contaminants. It is most effective for carbon and hydrocarbons but also often used on CO and H2. This method can be used on stainless steel as well as a range of other materials including Fe, Ni, Pd, Cu and Ag . This process usually takes place at between 10–100 times the base pressure of the system with pressure increasing as it proceeds . Reactive gas treatment can be used in situ or ex situ and it is common to combine this process with the early stages of bakeout as the higher temperature increases the reaction rate.
The glow discharge cleaning method is often a final stage in the cleaning process and is most effective for removal of carbon and oxygen based molecules. It takes place in situ once pressure has been reduced below 10−2 to 10−5 Torr. The chamber is filled with gas and a voltage (100–1000 eV) applied. Gas molecules are ionised and accelerated to the walls of the chamber causing ion induced desorption. Generally the gas will be flowed as the process continues to encourage desorbed molecules to leave the chamber .
Argon with 5–10% oxygen is the most common gas used as it is very efficient, but mixtures of N2, H2, He and CH4 are also used. Glow discharge cleaning creates an atomically smooth surface.
Surface treatments are often used to reduce surface area by reducing roughness or develop a surface layer that is impermeable to gas or unreactive to reduce degassing and readsorption.
There are two main polishing methods; mechanical polishing and electropolishing. Electropolishing takes place ex situ and works by smoothing the surface and therefore reducing the surface area for adsorption and desorption. Electropolishing can be applied to several metals including aluminium, copper and brass but is most often used for stainless steel . Although this alone does not have a significant effect on the outgassing rate, electropolishing also replaces the amorphous surface layer with an ordered oxide layer which can act as a barrier to diffusion. The downside of electropolishing is that it can actually increase outgassing of certain molecules such as hydrogen by ingraining them in the surface layer. Therefore if electropolishing is used the system should also undergo bakeout to degas the surface.
Mechanical polishing is one of the first steps in the cleaning process and uses a combination of techniques including grinding, milling, filing, brushing and buffing for the removal of gross contaminants. These processes can be rough and care should be taken not to leave additional contaminants on the surface or damage the material, particularly sealing surfaces. Mechanical polishing only gives a small decrease in outgassing rate and should be used in conjunction with other, more refined, techniques.
The technique known as compound electropolishing is the simultaneous application of both electropolishing and mechanical polishing. This results in extremely low surface roughness and reduced outgassing rates compared with electropolishing alone .
Passivation is the process by which a barrier layer is formed to prevent contaminants adsorbing on the surface of the chamber and gas permeating through to the vacuum from the bulk. This can be in the form of coating (both passive and active) or surface oxidation achieved through baking.
The aim of using a coating is to create a barrier to diffusion and desorption on the surface of the material. These can either be passive coating that are a simple barrier or active coatings which effectively pump gases from the chamber to trap them. Passive coatings are primarily used as a barrier to hydrogen diffusion from the bulk whilst active coatings can be used to pump CO and H2.
Coatings are usually applied to the material via chemical vapour deposition (CVD), cathodic arc physical vapour deposition or sputter coating at raised temperature (200–500°C). The thickness of the layer can vary from 100 nm to 10 μm with films of around 1 μm being most common in general applications. A range of materials are used for passive coating, including Si (usually amorphous), TiN, BN, Al2O3 and ZrO2. These are effective at blocking CO, CO2 and H2O and particularly H2.
Active barriers act to absorb gases such as H2, H2O, CO, O2 and N2 exothermically but cannot pump the noble gases. Generally materials used for this type of coating are first group transition metals, rare earth elements or fourth group and their alloys after surface activation. These are known as Non Evaporable Getters (NEG) . The particular materials used are chosen as they will readily bond with typical gases found in vacuum systems to form stable compounds. When the NEG surface becomes saturated, it can be activated by heating which causes the dissociated hydrogen to recombine and be released into the system where it can be pumped away and other gases to move deeper into the bulk, leaving the surface sites free for further pumping.
Fig. 11 demonstrates the effect of activation temperature on pumping for some commonly used NEG materials in various combinations. Generally, higher temperature activation results in lower outgassing, however NEG materials with low temperature activation are desired for systems where high temperature would damage sensitive components. Activation of coatings for stainless steel should take place at temperatures below 350°C and for copper or aluminium at less than 200°C [29,30].
Bakeout is one of the most commonly used procedures since it gives reliable reduction in outgassing rate; it can result in a decrease of over two decades from a bake of a few hours at a moderate temperature although it is common to bake systems for 10–20 hours to get lower outgassing rate. Baking is either carried out under vacuum or in air and generally, the higher the temperature during bakeout, the lower the outgassing rate afterwards since higher temperatures lead to faster outgassing as long as any gas removed from the surfaces during bakeout are pumped from the chamber. Empirically, this relation was found for stainless steel by Odaka, Ishikawa and Furuse  to be given by:
From 100–500°C baking is used to remove water vapour; generally temperatures up to 150°C are sufficient for most applications, temperatures at the upper end of this range can be used for removing chemisorbed water vapour from stainless steel but also gases such as CH4, CO and CO2. At higher temperatures (up to 1000°C for stainless steel) baking is mainly used to remove dissolved hydrogen from the bulk. Care should be taken when baking at high temperature to avoid changes to the mechanical properties of the material (magnetism, structure etc.). During a moderate temperature bakeout of stainless steel around 90% of the gas desorbed from the surface is hydrogen with the remainder being composed of CO, CO2 and CH4 .
Vacuum stoving is a similar process to vacuum bakeout and is used to precondition individual components prior to construction of the system. Vacuum stoving should be carried out for as long as is practical although, as with bakeout, heating for a few hours is usually sufficient for most applications. It has been shown that stoving individual parts can reduce outgassing rates of the final system by an order of magnitude .
Oxidation is another process used to create a surface barrier layer against diffusion and permeation from the bulk. Oxides also encourage the formation of stable compounds with low outgassing rates for both stainless steel and aluminium.
There are several methods which can be used for oxidation, however the most common is to expose the surface to air at a raised temperature (100–250°C), known as air bakeout. The duration of air bakeout is usually less than for a vacuum bakeout; typically between 2 and 5 hours. This process results in an oxide layer over ten times thicker than for a cleaned sample. This is particularly effective at reducing hydrogen outgassing however does not give outgassing rate as low as vacuum bakeout.
The key aspects of the main mechanisms behind outgassing have been presented. The gas load of the system has multiple contributions of which molecules desorbing from the surface (outgassing) are often the limiting factor in the ultimate pressure achievable and for pump down time. Surface interactions are extremely important in understanding outgassing. Adsorption can proceed via either weak bonding physisorption or strong bonding chemisorption. Various isotherms can be used to model and understand adsorption and are usually classified into five of six basic isotherm types.
Molecules with very short or very long surface residence times can be ignored with regards to outgassing as they do not have a significant effect on the time to reach ultimate pressure. Molecules physisorbed to the surface will be the first to desorb, followed by the more strongly bonded chemisorbed molecules. In High Vacuum the largest contribution to outgassing (at around 80%) will be water vapour and then at lower pressures H2 and CO.
Overall, there are many theoretical aspects which must be considered when looking at the outgassing of a system and selecting materials. In general all-metal systems are preferred due to lower saturated vapour pressure and permeability meaning outgassing is lower by several orders of magnitude than for polymers and elastomers. Stainless steel is the most commonly used materials for vacuum systems.
It is important that the outgassing rate of materials, components and systems can be measured accurately. There are several methods and variations on these which can be used depending on the application. Although rates for many materials are available in literature there is significant variation between these due to factors such as material composition, history and treatments and measurement method and conditions.
There are many techniques which can be used to reduce outgassing and the choice of which to use will depend on the application; the materials used, process, gas load, time and budget and other limitations. Simple processes such as polishing and washing or more rigorous ones such as blowing, bead blasting, snow treating or chemical cleaning should be used during the fabrication of the systems to give smooth surfaces and remove gross and fine contamination from the surface. Surface treatments such as reactive gas, glow discharge or coating can be used to further reduce outgassing. All of these surface techniques can be combined with the early stages of bakeout or used alone as alternatives when baking is not possible.
Bakeout, both under vacuum and in air, is the most effective technique to reduce outgassing. Water vapour is a major concern in vacuum systems and this is significantly reduced by bakeout at temperatures between 100 and 500°C. At lower pressure, after water vapour is removed, hydrogen permeation and diffusion from the bulk are the principle causes of outgassing. These are effectively reduced by a high temperature bakeout in air with the formation of an oxide layer.
With the appropriate combination of material selection, cleaning and preparation, outgassing rates as low as 10−14 Torr · l ·s−1·cm−2 can be achieved. Considering the many methods for measuring outgassing rate and the large effect small differences in material properties, surface preparation and treatment can have on outgassing rate, for cases where it is of extreme importance to accurately know the outgassing rate, it is recommended to make independent measurements and use these to guide which outgassing reduction strategies should be employed. For applications where outgassing requirements are less stringent, there are a wealth of techniques available to reliably reduce outgassing by an order of magnitude or more.
Of the many possible contributions to the gas load of a vacuum system, outgassing if often the most important (especially beyond medium vacuum), in particular with regards to pumpdown time and ultimate pressure achievable. The mechanisms behind bulk transport mechanisms, sorption and desorption are of particular importance as they govern the rate of outgassing and the prevalent gas species for each pressure regime. A main contribution to outgassing is from water vapour, with hydrogen outgassing also critical for achieving UHV with metal systems.
There are many sources of outgassing rate data available, however these can often give widely varying values or be too specific in terms of material and treatment to be of general use. Although outgassing is a fundamental parameter and indeed limitation for vacuum systems there is a need for an updated/additional set of experimental rate measurements. It is often important to accurately measure the outgassing rate, either of a chamber or a sample and there are several commonly used methods to accomplish this. The selection of a measurement method can be made based on factors such as the accuracy required, whether a sample or chamber is being measured and if identification of gas species is needed. The development of a standard for rate measurement techniques will be highly valuable to be able to enable ‘direct’ comparisons between measured rates.
There have been many and varied techniques developed to reduce outgassing. These vary from simple techniques such as cleaning or polishing which give small reductions in outgassing rates to complicated coatings or vacuum/air bakeout which allow very low outgassing rates to be attained. These techniques are clearly described and proven, with the choice of outgassing reduction technique depending on factors such as length of treatment time, material type and desired ultimate outgassing rate. Generally, multiple outgassing reduction methods will be combined, along with appropriate selection of materials.
||clean metal surfaces|
||desorption from surfaces|
||metals, glasses and ceramics|
||highly porous surfaces|
||diffusion controlled outgassing from the bulk|
|Type of interaction||Long range Van der Waals attraction (dipole)||Chemical bonding via hybridisation of electron orbitals|
|Energy of bond (adsorpton enthalpy)||10–400 meV (0.96–38.5 kJ/mol) related to molecular mass, polarity and polarisability||0.4–10 eV (38.6–965 kJ/mol) related to the type/strength of the chemical bond|
|Distance from surface||1–3 Å||3–10 Å|
|Saturation uptake||Can form multilayers||Can only form a single monolayer|
|Material specificity||There is only a slight dependence on the composition of the surface material||There will be substantial variation depending on the material|
|Crystallographic specificity (surface symmetry)||Effectively independent||There will be noticeable variation depending on the crystal plane exposed|
|Adsorption temperature range||Around (or below) the condensation point of the gas being adsorbed||Small range dependent on individual molecules, effectively unlimited range overall|
|Other properties||Always atomic/molecular (non-dissociative), reversible, usually fast||Can proceed by causing molecules to dissociate, is often reversible, speed varies significantly depending on activation energy|
Comparison of mean stay time (time for desorption) for a range of desorption energies and various cases 
|Desorption energy (kcal/mole)||Cases||τd (s)|
|0.1||Helium||1.2 × 10−13|
|1.5||H2 physisorption||1.3 × 10 × −13|
|3–4||Ar, CO, N2, CO physisorption||1 × 10 × 11|
|10–15||Weak chemisorption||3 × 10 × 6|
|25||6 × 105-one week|
|30||CO/Ni chemisorption||4 × 109-100 years|
|40||1 × 1017-age of the Earth|
|150||O/W chemisorption||> age of the universe|
Variation in contribution to gas load for various gas species over a range of pressures 
|Pressure (mbar)||Major Gas Load|
|Atmosphere||Air (N2, O2, H2O, Ar, CO2)|
|10−3||Water vapour (75%–95%)|
Averages and standard deviations of collected literature values for outgassing rates of various common vacuum materials after different treatments (unbaked unless otherwise stated) [37–46], N/A for standard deviation indicates that only a single literature value was found for a particular material
|Material||Treatment||Average of literature values for outgassing rate Torr · l · s−1 · cm−2||Standard deviation of literature values for outgassing rate Torr · l · s−1 · cm−2|
|Aluminium||Fresh||2.2 × 10−7||5.3 × 10−7|
|Degassed (24 hr)||8.7 × 10−8||N/A|
|Baked (15 hr @ 250°C)||4.0 × 10−13||1.7 × 10−15|
|Baked (20 hr @ 100°C)||3.9 × 10−14||1.6 × 1015|
|Duraluminium||1.5 × 10−7||3.0 × 10−8|
|Iron||2.0 × 10−7||2.8 × 10−7|
|Brass||Cast||1.1 × 10−6||1.4 × 10−7|
|Waveguide||3.5 × 10−7||7.1 × 10−8|
|Copper||Fresh||1.8 × 10−8||1.2 × 10−6|
|Mechanically polished||2.2 × 10−8||9.0 × 10−10|
|OHFC copper||Fresh||1.8 × 10−8||3.1 × 10−9|
|Mechanically polished||2.2 × 10−9||9.0 × 10−10|
|Gold||Wire, fresh||8.5 × 10−8||1.0 × 10−7|
|Mild steel||Fresh||4.67 × 10−7||6.5 × 10−8|
|Slightly rusty||1.5 × 10−6||2.0 × 10−6|
|Chromium plated, polished||1.0 × 10−8||N/A|
|Aluminium spray coated||6.0 × 10−8||1.1 × 10−10|
|Steel||Chromium plated, fresh||5.1 × 10−9||3.0 × 10−9|
|Chromium plated, polished||9.1 × 10−9||1.2 × 10−9|
|Nickel plated, fresh||3.9 × 10−9||6.2 × 10−10|
|Nickel plated||1.3 × 10−7||2.5 × 10−7|
|Descaled||2.7 × 10−7||5.4 × 10−8|
|Molybdenum||3.1 × 10−7||3.6 × 10−7|
|Stainless steel||2.2 × 10−7||2.0 × 10−7|
|Fresh||1.5 × 10−8||4.1 × 10−9|
|Sanded||8.3 × 10−9||3.4 × 10−11|
|Electropolished||7.0 × 10−9||9.3 × 10−9|
|Mechanically polished||4.2 × 10−9||6.0 × 10−9|
|Baked (30 hr @ 250 °C)||3.0 × 10−12||4.3 × 10−9|
|Zinc||1.9 × 10−7||4.0 × 10−8|
|Titanium||7.6 × 10−9||5.1 × 10−9|
|Pyrex||Fresh||7.4 × 10−9||6.3 × 1011|
|Neoprene||3.0 × 10−5||4.3 × 10−9|
|Polystyrene||2.0 × 10−5||N/A|
|Plexiglas||2.0 × 10−6||1.2 × 10−6|
|Viton A||Fresh||8.5 × 10−7||4.7 × 10−7|
|Baked (4–16 hr @100–200°C)||1.7 × 10−9||3.5 × 10−9|
|PVC||24 hr @ 95% RH||2.4 × 10−6||2.5 × 10−6|
|Teflon/PTFE||1.1 × 10−6||1.9 × 10−6|
Comparison of suitability of outgassing rate measurement methods against various criteria (based on table in ).
|Measurement method||Vapour outgassing||Time dependent outgassing||Identifies gas species||Outgassing >10−9 Pa · m3 · s−1||Outgassing ≤10−9 Pa · m3 · s−1||Accuracy||Effort/experience required||Chamber as sample|
|Measured effective pumping speed throughput||+||++||−||+||−||− −||Low||+|
|Conductance modulated throughput||−||−||++||+||++||+||High||+|
|Rate of rise||− −||−||− −||+||++||+||Low||++|
|Rate of rise and throughput||− −||−||+||+||++||+||High||+|
|Mass loss||+||++||+||+||− −||+||High||− −|
|Continuous expansion||+||++||++||+||−||++||Very high||+|
Comparison of methods to reduce outgassing
|Method||Materials||Effective for||Not effective for||Recommended combination with other methods||Vacuum or outgassing level||In/ex situ||Time||Difficulty|
|Wash||Hot water, detergent, hot solvent||Gross/fine contaminants||Grease||Followed by surface treatment and/or bake||Pressures >10−3 mbar||Ex situ||< 30 min||Easy|
|Vapour degrease||Heated solvent||Gross/fine contaminants, loosely bound heavy molecules||Recessed/confined/inaccessible areas, large parts||Followed by surface treatment and/or bake||Order of magnitude reduction in outgassing rate||Ex situ||< 1 hour||Easy|
|Blowing||Air, nitrogen or other dry inert gas||Gross/fine contaminants, recessed/confined/inaccessible areas||Grease||Preceded by washing, followed by non-particulate generating process||Ex situ||< 30 min||Easy|
|Bead blasting||Alumina or silica beads in air or water||Gross contaminants, films or other surface coatings||Smooth surfaces||Followed by method to reduce surface roughness||Ex situ||< 1 hour||Moderate|
|Snow cleaning||CO2ingasorliquid||Gross/fine contaminants, hydrocarbons||Followed by bake (can be used alone if required)||Ex situ||< 1 hour||Moderate|
|Chemical treatment||H2O,HCl,HNO3 or HF||Rough surfaces (O2, C, P or Cladsorption)||Preceded and followed by wash, followed by bake (can be used alone if required)||Order of magnitude reduction in outgassing rate||Ex situ||< 30 min||hard|
|Reactive gas||O2 or NO for oxidations, H2 or NH3 forreduction||C, hydrocarbons, CO/H2 on stainlesssteel, Fe/Ni/PdonCu/Ag, surface reconditioning after atmospheric exposure||Can form unwanted oxides||Preceded by cleaning, can be combined with early stage bake||Reduction in outgassing rate by 50% up to 5 orders of magnitude||In situ or ex situ||30 min – 2 hours||Hard|
|Glow discharge||Ar/5–10%O2, Ar,O2, N2, H2, He, He/10%O2, He/1–3%H2, He/1–3%CH4||C or O based molecules, large surface areas, reconditioning after atmospheric exposure||Large components||Preceded by wash/clean, can be combined with oxidation, can replace bake||13 × reduction in outgassing rate||In situ||~2 hours||Hard|
|Method||Materials||Effective for||Not effective for||Recommended combination with other methods||Vacuum or outgassing level||In/ex situ||Time||Difficulty|
|Mechanical polishing||Reducing surface roughness, metals||Reducing gross/fine contaminants||Followed by wash and other surface treatment and/or bake||Reduction proportional to decrease in roughness (2 × 10−10 per unit roughness factor)||Ex situ||Depends on size/complexity of parts and finish||Moderate|
|Electro-polishing||Electrolyte solution (phosphoric acid)||Metals (stainless steel, aluminium, copper, brass), forming oxide layer||H2 (can ingrain in surface)||Followed by bake||Ex situ||Moderate|
|Passive coating||Si, TiN, BN, Al2O3,ZrO3||H2 from metals, CO, CO2, H2O||Preceded by clean/polish||0.1–100 × reduction in H2outgassing||Ex situ||Depends on size/complexity of parts||Moderate|
|Active coating (NG)||Hf, Zr, Ti, Pd, V and combinations||H2, H2O, CO, O2, N2, smallspaces, where bake is no tpossible||Hydrocarbons, inert gas, continual operation (needs re-activation and replacement)||Preceded by clean/polish, followed by heat activation||Outgassing reduced to <10−13 with no bake||Ex situ coating, in situ activation||Depends on size/complexity of parts||hard|
|Vacuum bake||Metals, UHV/XHV, H2O||Delicate/temperature sensitive components||Follows any other method||Reduces outgassing of H2to<10−14||In situ (ex situ for vacuum furnacing components)||2–400 hours||Moderate|
|Air bake (oxidation)||Air||Stainless steel, aluminium, H2,CO,CO2CH4||Plastic, brass, cadmium plating or other temperature sensitive materials||Follows any other method, can be followed by vacuum bake||Reduces outgassing of H2to<10−14||In situ||2–400 hours||Moderate|