uhv-companion.lyx

#LyX 2.2 created this file. For more info see http://www.lyx.org/
\lyxformat 508
\begin_document
\begin_header
\save_transient_properties true
\origin unavailable
\textclass article
\begin_preamble
\usepackage[pdftex]{graphicx}
% \setlength{\parskip}{\medskipamount}
% \setlength{\parindent}{0pt}

% Uncomment the following line to allow the usage of umlauts and other non-ASCII characters
%\usepackage[utf8]{inputenc}
\end_preamble
\use_default_options false
\maintain_unincluded_children false
\language english
\language_package none
\inputencoding auto
\fontencoding default
\font_roman "default" "default"
\font_sans "default" "default"
\font_typewriter "default" "default"
\font_math "auto" "auto"
\font_default_family default
\use_non_tex_fonts false
\font_sc false
\font_osf false
\font_sf_scale 100 100
\font_tt_scale 100 100
\graphics default
\default_output_format default
\output_sync 0
\bibtex_command default
\index_command default
\paperfontsize default
\spacing single
\use_hyperref false
\papersize a4paper
\use_geometry false
\use_package amsmath 1
\use_package amssymb 0
\use_package cancel 0
\use_package esint 1
\use_package mathdots 0
\use_package mathtools 0
\use_package mhchem 0
\use_package stackrel 0
\use_package stmaryrd 0
\use_package undertilde 0
\cite_engine basic
\cite_engine_type default
\biblio_style plain
\use_bibtopic false
\use_indices false
\paperorientation portrait
\suppress_date false
\justification true
\use_refstyle 0
\index Index
\shortcut idx
\color #008000
\end_index
\secnumdepth 3
\tocdepth 3
\paragraph_separation indent
\paragraph_indentation default
\quotes_language english
\papercolumns 1
\papersides 1
\paperpagestyle default
\tracking_changes false
\output_changes false
\html_math_output 0
\html_css_as_file 0
\html_be_strict false
\end_header

\begin_body

\begin_layout Title
The CINF UHV companion
\begin_inset Newpage newpage
\end_inset


\end_layout

\begin_layout Standard
\begin_inset CommandInset toc
LatexCommand tableofcontents

\end_inset


\end_layout

\begin_layout Standard
\begin_inset Newpage newpage
\end_inset


\end_layout

\begin_layout Part
The Basics
\end_layout

\begin_layout Standard
The first part describes all the elements that will commonly be involved
 in running and maintaining a UHV chamber.
\end_layout

\begin_layout Section
Vacuum
\end_layout

\begin_layout Standard
In the high vacuum world pressures are, due to historical reasons, normally
 measured in units of mbar or Torr which differ by a factor of 1.33.
 The two units are hence used more or less interchangeablely.
 A UHV system typically has a base pressure on the order of 
\begin_inset Formula $\sim10^{-10}$
\end_inset


\begin_inset space \thinspace{}
\end_inset

mbar.
 Using rather simple and reasonable assumptions it can be shown that at
 a pressure of 
\begin_inset Formula $10^{-6}$
\end_inset


\begin_inset space \thinspace{}
\end_inset

mbar every surface atom in the vacuum chamber will on average be hit by
 a molecule once every second.
 As a rule of thumb this means that at 
\begin_inset Formula $10^{-10}$
\end_inset


\begin_inset space \thinspace{}
\end_inset

mbar you can expect a clean sample to stay clean for approximately one hour.
\end_layout

\begin_layout Section
Flanges
\end_layout

\begin_layout Standard
Every vacuum chamber is surrounded by air.
 The pressure of ambient air is 
\begin_inset Formula $10^{3}$
\end_inset


\begin_inset space \thinspace{}
\end_inset

mbar or some 13 orders of magnitude higher than inside the chamber.
 This means that air is trying significantly hard to get inside the chamber
 and one must show a bit of carefulness to avoid this.
 Normally, every flange is packed with a copper gasket which is deformed
 from both sides by knifes cut in to the flanges.
 It is this deformation that ensures that the copper will make a perfect
 seal to both flanges.
 In order to achieve this, the bolts should be tightened reasonably hard,
 but not too much.
 Tightening too hard will only mean that it is hard to open the flanges
 and that you loose the possibility to tighten a bit extra in case of a
 leak.
 You should be careful to not tighten one bolt at a time, but distribute
 the applied force between the bolts to make sure you do not deform the
 gasket unevenly.
\end_layout

\begin_layout Section
The content of air
\end_layout

\begin_layout Standard
In daily life air might seem to be a harmless substance, but in reality
 it is terribly dirty, and you certainly do not want to have it inside your
 chamber.
 At 1
\begin_inset space \thinspace{}
\end_inset

bar every molecule that exists in concentrations of 1
\begin_inset space \thinspace{}
\end_inset

ppb will hit every surface atom once every second, basically meaning, that
 if you chamber has seen air you should regard it as polluted.
\end_layout

\begin_layout Section
Leak testing
\end_layout

\begin_layout Standard
Leak testing is the art of finding the origin of a leak in the chamber.
 The way to attack this problem depends heavily on the size of the suspected
 leak.
 If the chamber is reasonably tight so that it can be pumped to below 
\begin_inset Formula $10^{-5}$
\end_inset


\begin_inset space \thinspace{}
\end_inset

mbar, a mass spectrometer is the best form of leak testing.
 In this case leak-testing with helium is the preferred method.
 Helium is more or less the perfect molecule for this task since it a) only
 exists in absolutely minute amounts in air b) it is by far the smallest
 molecule in Nature and therefore the one that will most easily slip through
 a leak.
 The procedure is to measure mass 4 on the mass spectrometer while blowing
 helium on the chamber one flange at a time.
 If you measure a a signal on the mass spectrometer, you have found a leak.
 Remember that helium rises in air and for that reason it is wise to start
 from the top of the chamber.
 If the chamber does not have a mass spectrometer the ion gauge can be used
 to find at least rather large leaks.
 This works because the ionization energy of helium is much larger than
 that of both oxygen and nitrogen.
 If a leak is present a drop in the signal from the ion gauge will be seen
 when helium out-competes air in the vicinity of the leak.
\end_layout

\begin_layout Standard
If the leak is large enough that the chamber cannot be pumped low enough
 to turn on the mass spectrometer life is a bit more difficult.
 In that case one is forced to randomly tighten flanges under the suspicion
 of being leak.
 If this also fails, another option is to back-fill the chamber with helium
 and use an outside sniffer connected to the leak tester to find the leak.
\end_layout

\begin_layout Standard
Once the leak is found often the problem can be solved simply by tightening
 the flange.
 If this does not solve the problem, chances are, that the gasket is not
 mounted properly and you will have to open the flange.
\end_layout

\begin_layout Section
UHV pumps
\end_layout

\begin_layout Subsection
Turbo pumps
\end_layout

\begin_layout Standard
The turbo pump is the true workhorse in a UHV-system.
 It works by spinning plates tilted at an angle very fast and thus it mechanical
ly kicks the molecules out of the chamber.
 Once the molecule is out of the pump it reaches a roughing pump of the
 backside of the turbo that takes it the rest of the way to atmospheric
 pressure.
 For this kind of pump to work the pressure needs to be low enough that
 the molecule-molecule interactions are very small – meaning somewhere below
 1
\begin_inset space \thinspace{}
\end_inset

mbar.
 A turbo pump is a more or less perfect pump in the sense that practically
 no back-flow occurs – no gas from the pre-vacuum side will enter the chamber.
 This gives the very nice property that even though it will of course not
 provide a perfect vacuum the turbo pump will provide a very clean vacuum
 consisting only of the gasses that are already in chamber.
 Thus it is possible to have a residual gas mixture containing only a few
 well known species.
\end_layout

\begin_layout Standard
A turbo pump is not made to move large amounts of gas but it will work reliably
 up to at least 
\begin_inset Formula $10^{-2}$
\end_inset


\begin_inset space \thinspace{}
\end_inset

mbar, even for extended amounts of time, provided sufficient cooling and
 a good roughing pump capable of removing the gas from the backside of the
 pump fast enough.
 The workload of the pump can to some degree be measured from the current
 drawn by the pump, which can normally be seen on the control unit.
 A turbo pump should typically not draw more than 1–1,2
\begin_inset space \thinspace{}
\end_inset

A for extended amounts of time.
\end_layout

\begin_layout Standard
On a practical note; a turbo pump is spinning very fast (typically between
 800 and 1500
\begin_inset space \thinspace{}
\end_inset

Hz depending on the size of the pump) which means that you should 
\series bold
NOT
\series default
 try to move a spinning turbo unless you want to perform a very expensive
 demonstration of applied mechanics.
 Most pumps are by nature stationary but for instance the leak-tester has
 a turbo-pump that can in principle be moved.
\end_layout

\begin_layout Subsection
Ion pumps
\end_layout

\begin_layout Standard
An ion pump is closed pump meaning it will store the pumped molecules inside
 the pump itself.
 This obviously means that ion pumps should only be used at low pressures
 (typically below 
\begin_inset Formula $10^{-7}$
\end_inset


\begin_inset space \thinspace{}
\end_inset

mbar).
 At such low pressures it is possible to accumulate the pumped gasses even
 for extended amounts of time.
 Once in a while (typically at unpredictable times) the ion pump will release
 the some of enclosed gasses, typically noble gases which do not react very
 well with the wall of the pump.
 These gases must then be pumped away with a turbo pump.
 Ion pumps provides excellent base pressures and very conveniently have
 no moving parts.
\end_layout

\begin_layout Subsection
Titanium supplimation pumps (TSP)
\end_layout

\begin_layout Standard
A TSP is also a closed type of pump.
 It works by evaporating titanium on the surface of the pump itself.
 This titanium will then very efficiently bind some of the residual gas
 in the chamber.
 Next time the pump is run, the gas adsorbed close to the filamant will
 be released but in much more concentrated form, meaning that it will be
 pumped more efficiently by the other pumps in the system.
 A TSP is very good at 'cleaning up' after an experiment that left the camber
 more dirty than what is normally wanted.
\end_layout

\begin_layout Section
Pressure gauges
\end_layout

\begin_layout Standard
FIXME WRITE
\end_layout

\begin_layout Section
Roughing pumps
\end_layout

\begin_layout Standard
A roughing pumps purpose in life when placed in a UHV lab is normally to
 act as backing for the turbo pump.
 At CINF you will find typically two main kinds; the oil-lubricated pumps
 and non-lubricated pumps (so-called Scroll pumps).
 The oil-lubricated kind is by far the most common one.
 It will normally provide a vacuum between 
\begin_inset Formula $10^{-2}$
\end_inset

 and 
\begin_inset Formula $10^{-3}$
\end_inset


\begin_inset space \thinspace{}
\end_inset

mbar, preferably closer to the latter value.
 If the pump is not able to reach 
\begin_inset Formula $5\times10^{-2}$
\end_inset

 it is most likely in need of service.
 A scroll pump will not reach nearly the same ultimate pressure as a oil-lubrica
ted pump.
 Normally one should keep an eye on the pressure of the roughing pump, it
 should stay quite constant, and any sign of an increased pressure is properly
 a sign of a pump that should be serviced.
\end_layout

\begin_layout Standard
An extremely important thing to be aware of is the composition of the residual
 gas in a roughing pump.
 For an oil-lubricated pump, the residual gas consist mostly of various
 organic compounds and for a scroll pump it consists mostly of air.
 In both cases the residual gas must be considered to be extremely dirty,
 why it is important that you never apply the pump in a way that will allow
 for backflow from the pump to reach your clean vacuum chamber.
 In normal operation this is ensured by the turbo pump but in situations
 where the roughing pump is pumping directly on the chamber one needs to
 think.
 If you are pumping down a vented chamber, everything is OK, you are going
 to bake the chamber anyway.
 If you are pumping down a system considered clean (high-pressure cells,
 load-locks and gas-lines come to mind) you must 
\series bold
NOT
\series default
 pump too low in pressure with the roughing pump.
 Stay in a pressure range no lower than 2 to 10
\begin_inset space \thinspace{}
\end_inset

mbar, which will ensure a high enough flow into the pump that essentially
 no residual gas will back-flow.
 From this pressure you can leak the gas into a turbo pump using a good
 quality valve.
\end_layout

\begin_layout Section
Valves
\end_layout

\begin_layout Standard
At CINF you will typically find a certain variety of valves.
 Three main kinds exists as well as a number of other less common types.
 The three main types are the VAT valve, the Swagelok valve and various
 kinds of all-metal valves.
 The Swagelock and to a certain extent the VAT valve has the rather nice
 property, that the can be opened in a controlled way, allowing for a slow
 opening between two compartments.
 This is very practical when you want to leak a semi-high pressure into
 a turbo pump, which is for instance the case after pumping down a clean
 system with a roughing pump.
\end_layout

\begin_layout Standard
Every time one operates a valve, one needs to be aware of the current conditions
 on the two sides of the valve.
 Often one side of the valve is exposed to the vacuum and the other side
 may or may not also be pumped down.
 If it is not, and you open the valve, the gas from the high pressure side
 will equilibrate with the vacuum side resulting in contamination of the
 vacuum chamber.
\end_layout

\begin_layout Section
Baking
\end_layout

\begin_layout Standard
After a chamber has been vented it has, as mentioned, been exposed to basically
 any unpleasant gas one can imagine.
 All this pollution needs to be removed.
 Some substances is easily removed just by pumping, but others forms layers
 on the inside of chamber which is not easily pumped away.
 Water and pumping oil are two notable substances showing this behavior.
 Even though the gasses will of course eventually be pumped away, the time
 constant for this is basically infinite and something must be done to increase
 the speed.
 The solution is to increase the temperature of the entire chamber, so-called
 baking.
 Normally baking is performed at temperatures around 150
\begin_inset space \thinspace{}
\end_inset


\begin_inset Formula $^{\circ}C$
\end_inset

.
 At these temperatures the vapor pressure of the contaminants rises significantl
y and will desorb from the walls into the gas phase where it can be pumped
 efficiently.
 After a few days of baking, your chamber should be reasonable clean.
\end_layout

\begin_layout Standard
An important thing to remember when baking is that you should always bake
 all parts of a given volume at the same time.
 If any part of the chamber is not heated, this part will act as a cold
 finger where all the desorbed contaminants will tend to re-adsorb.
 Once the chamber is cooled the cold spot will soon equilibrate with the
 rest of the chamber and your baking is ruined.
\end_layout

\begin_layout Part
Additional Equipment
\end_layout

\begin_layout Standard
This part describes general operating principle and operational range for
 different additional equipment commonly used in conjunction with a UHV
 chamber.
\end_layout

\begin_layout Section
Leak valves
\end_layout

\begin_layout Standard
FIXME WRITE OUT
\end_layout

\begin_layout Section
Mass flow controllers (MFC)
\end_layout

\begin_layout Standard
FIXME WRITE OUT
\end_layout

\begin_layout Section
Pressure controllers (PC)
\end_layout

\begin_layout Standard
FIXME WRITE OUT
\end_layout

\begin_layout Section
Electron detectors
\end_layout

\begin_layout Standard
FIXME WRITE OUT
\end_layout

\end_body
\end_document