UHV Vacuum Techniques: Basic Concepts

Aims:

● What you ought to know about using vacuum

● What you might need during your PhD

● A quick starter for using vacuum equipment

● Cover entire range of vacuum - not just UHV

Lecture Contents

Lecture 1: Basics

► Levels of vacuum and subsequent applications

► Modelling gas flow

► Pressure measurement - atmosphere to UHV

Lecture 2: Achieving Vacuum

► Vacuum pumps

► Pump configurations

► Types of chambers and fittings

► Complete systems

References

● Two excellent books:

► “Basic Vacuum Technology”, A. Chambers, R.K. Fitch and B.S. Halliday.

► “A Users Guide to Vacuum Technology”, O'Hanlon.

● Vacuum equipment manufacturers:

► Catalogues, websites

► e.g. Leybold, Edwards, Varian, Pffeiffer, Alcatel etc.

● Local knowledge (valuable, but finite resource!):

► workshops

► group technical support staff

Why know about vacuum?

● Nearly everyone here needs to know a little about vacuum:

► vacuum used very widely in modern experimental physics

► mostly simple physical principles involved

► what is achievable with today's technology

► the basics - so its easier to go away and find out more

● Many people here will use vacuum systems:

► e.g. most research groups use vacuum to some extent

► need to understand the systems

► might need to analyse what is happening or design modifications

Historical Perspective

Magdeburg Hemispheres Experiment

Ancient Greeks- Vacuum

“inconcievable”

1657 1900

Manufacture ofLightbulbs

BC 1950's

Ultra HighVacuumMercury

SealedPumps

2000+

Surface treatment

research to achieve XHV

HydrocarbonSealed Pumps

What is the equipment like?

Ultra High Vacuum(~10-10 mbar)Rough Vacuum

(~10-1 mbar)

Some Definitions

Units:

►Usually use mbar (although Pa are SI, sometimes Torr)

► 1000 mbar = 1 atm = 105 Pa = 760 Torr

Levels of Vacuum:

► Low Vacuum - 1000 mbar to 1 mbar

►Medium Vacuum - 1 mbar to 10-3 mbar

►High Vacuum - 10-3 mbar to 10-8 mbar

►Ultra High Vacuum - 10-12 mbar to 10-8 mbar

► eXtreme High Vac. - below 10-12 mbar

Assume the vacuum system is a sphere...

CHAMBER(GAS)

PUMP

Kinetic Theory: reasonably valid

● Maxwell Boltzman Distribution

● Mean free path

● Impingement Rate

=1

2 d2n

v= 8kT

m

J =nv4

For nitrogen at 295K:

P (mbar) n (litre-1) λ J (cm-2s-1)

1000 2.5x1022 66 nm 2.9x1023

1 2.5x1019 66 µm 2.9x1020

1x10-3 2.5x1016 66 mm 2.9x1017

1x10-6 2.5x1013 66 m 2.9x1014

1x10-10 2.5x109 660 km 2.9x1010

Applications

Low Vacuum:

► mechanical handling

Medium vacuum:

► industrial processes

► vacuum drying/packaging

► vacuum distillation

High vacuum (HV, λ>d):

► e-beams (welding, TV)

► vacuum evaporation or coating

Ultra high vacuum (UHV):

► keeping surfaces clean for hours (surface science, epitaxial growth)

► space simulation (~10-10 mbar at 1000 km)

► achieving ultra high purities (e.g. for fusion)

Extreme high vacuum (XHV):

► storage rings

► ultra pure growth

Elementary Gas Transport

● Throughput, Q, is the volume of gas passing through an area per second at a specific temperature and pressure:

► units, e.g. mbar.l/s

● Volumetric flow rate often called the speed, S:

C=Q

P2�P1

● Conductance is a geometric property, which represents the ability of a gas to flow from a pressure gradient:

Q=pdV

dt

S=dV

dt=

Q

p

p1

p2

C

Q

S

Modes of Gas Flow

● Fluid flow is described by the Reynolds number and the Knudsen Number (both dimensionless)

● Turbulent flow if Re>2000, laminar if Re<1200.

● Viscous flow if Kn<0.01, molecular flow if Kn>1.

For vacuum systems:

● Flow is only turbulent at very high pressures and pumping speeds (e.g. during initial evacuation, if unthrottled)

● Viscous flow and the transitional regime are important at high pressures (above about 10-3 mbar)

► gas can be 'sucked out'

● Molecular flow dominant below about 10-3 mbar in 'normal' sized chambers.

► gas flows through random collisions with walls

R e=u D

Kn=

D

for round pipes

ρ = density u = stream velocity D = diameter λ = m.f.p.

Molecular Flow

● dominant for HV and UHV and well understood

● particles flow in all directions to reach dynamic equilibrium

● pumps 'wait and catch' gas particles - high vacuum pumps do not 'suck'

PUMP

Particles scattered from surface with cosine distribution to surface normal(Knudsen's cosine law)

Molecular Conductance of an Aperture

P1, n

1P

2, n

2

A

Q=k T J 1�J 2 A

= k T

2 m p1� p2 A

C0= k T

2 mA

Net flow through an aperture corresponds to the rate of impingement from both sides:

► conductance depends on (T/m)1/2

► C0 = 11.8A l/s (A in cm2) for nitrogen

(air) at room temp.

► need to account for gas and temp.

J1

J2

Conductance of Pipes

● Various methods for calculating conductance of pipes give same results (see O'Hanlon for details)

● For long pipes (L>>D):

● l/s

(D, L in cm) for nitrogen (air) at room temperature.

● For short pipes is convenient to reciprocally add the conductances for a long pipe and equivalent aperture:

● l/s

(D, L in cm) for nitrogen (air) at room temperature.

● Accurate to about 10%

C L=D

3

6 L 2k T

m

CS=C 0×CL

C 0CL

CS=12.4 D

3 / L

14D /3L

C L=12.4 D3/ L

Combining Conductances

● For conductances in series, add reciprocally:

► 1/CT = 1/C

1 + 1/C

2 + 1/C

3

● For conductances in parallel, add normally

► CT = C

1 + C

2 + C

3

● this assumes the volumes are independent - need to be careful

C1

C2

C3

C1

C2

C3

Effect of Conductance on Speed of a Pump

CHAMBER PUMP

Conductance, C

Effective pumping speed atchamber is reduced by conductanceof any connecting pipe

PUMP

The molecular conductanceof the entrance aperturedetermines maximum speed ofany pump

Viscous and Transitional Flow

● higher pressures in pipes to mechanical pumps mean transitional and viscous flow may be important

● conductance increases with pressure

● important in connections to mechanical pumps - can use smaller connections

● accurate calculations complex

● results tabulated - often in mfr's catalogues

From BOC Edwards Catalogue

Molecular Transitional Viscous

Conductances of Complex Shapes

● Monte Carlo simulations required in general

● Refer to O'Hanlon for simple 'standard' shapes

● Quite often, just need approximate value

► can get quite a long way using simple approximations

Approximate conductance isthe conductance of the two apertures,added reciprocally...

A closer look at a generic system...

CHAMBER

PUMP

Evaporation

Process

Outgassing

Leak

Virtual Leak

Outgassing: gradual loss of particles adsorbed on walls of the chamber - water is the major problem.

Real Leak: a finepassageway to theair outside.

Virtual Leak: usually a small trapped volume, which acts like a real leak, but will deplete with time.

Backstreaming

Backstreaming: many pumps can lose fluids into the vacuum system, causing contamination, which can be difficult to remove.

QT = Q

P + Q

O + Q

L + Q

VL + Q

E + Q

B

Evaporation: liquids (and greases) will limit the vacuum until evaporated. (Clean components and WEAR GLOVES!)

Initial Air

Quantitative Description of Pumping

● Constant volume system governed by

● i.e. change in gas in chamber, d(pV)/dt is load minus gas removed by pump

● Solving fully requires detailed knowledge of Q

T - not

generally available

► evaporation and outgassing of range of gases gives complex behavior...

● Ultimate pressure (dp/dt=0) pressure is simply

p = QT / S

► in the case of a leak, Q

T~Q

L

● For initial pumpdown, QL is

not important, so we obtain

p = p0 exp { -t / (v/S) }

► easy to calculate initial pumpdown times using the pumping speed at the chamber

Vdp

dt=QT�Sp

Limit of Pressure - Outgassing

● outgassing limits vacuum in a clean, leaktight HV/UHV chamber

● made up of general 'grot', greases, water vapour etc.

● for UHV, must accelerate degassing of water by baking to ~200°c for ~24 hours

► desorption is activated with Boltzmann factor

► rule of thumb: rate doubles for every extra 10°c

► can get to UHV in days instead of years

● what is the ultimate limit?

► diffusion of H2 through

chamber wallslog(time)

log (

pre

ssure

)

Measurement of Pressure

● Require a measurable property which changes (linearly) with pressure, preferably independent of gas type

● No single principle is good between atm. and UHV.

Pressure (mbar) 103 100 10-3 10-6 10-9 10-12

Capsule gauge

Diaphragm

Pirani (thermal cond.)

Capacitance manometer

Spinning rotor

Penning

Bayard-Alpert (Ionis.)

Inverted Magnetron

Extractor Ion

RGA Mass Spec.

Vacuum Gauges

● In practise, integrated systems now widely available

► almost 'plug and play'

► connect to PC for logging

► can link to pump controllers

● Composite gauge heads can measure over wide ranges

► combined pirani & ion gauge

► measure from atm. to UHV

► Still need care in operation - e.g. responses different for different gases

● e.g. integrated ion gauge and controller

Mechanical Total Pressure Gauges

● Sense pressure by mechanical deformation

● Measure total pressure - independent of gas type

● Capsule gauge:

► simple mechanical lever from expanding capsule to dial

► good for 1-1000 mbar

● Diaphragm gauges

► sense by mechanical deflection

► sense by change in capacitance, very accurate and good to 10-5 mbar, but expensive

Capsule Gauge

Capacitance Manometer

Diaphragm

ChemicalGetter

Annularelectrode

Discelectrode

pressure undermeasurement

Pirani Gauge - Low to Medium Vacuum

● measures thermal conductivity from hot wire to surroundings through vacuum

● typically set up in bridge arrangement with a compensating filament, and calibrated at HV

● 'standard' backing line gauge, cost few £100

● needs calibrating to atm. and high vacuum

Pirani gauge and controller

RF

RF

RC

10-4 1 P (mbar)

He

at

Loss

V

Bayard Alpert Ion Gauge - HV and UHV

● hot filament emits electrons, which are attracted to grid and spiral around

● electron impact ionised residual gas inside grid

● positive gas ions reach collector and are detected by electrometer

● usable between about 10-3 and 10-11 mbar, cost ~£1000

● limited by x-rays from grid

Response of gauge depends on:

► geometry of gauge

► emission current

► ionisation probability of gas

Gas Typical Correction Factor He 0.16 N

21.0

CO2

1.4

Xe 2.4 C

6H

65.8

30 V 180 V

Penning Gauge

● cold cathode ionisation gauge - no filaments to blow

● single ion initially created spontaneously (e.g. cosmic ray)

● electrons attracted to anode, but crossed B and E fields cause long spiraling paths

● electrons cause further ionisation to maintain a stable discharge

● ion current given by i = kpn where k and n are constants, and 1.1<n<1.2

+2kV

B

● response non-linear, so not regarded as so accurate

● some pumping effect

● robust, widely used gauge (industrially) for range 10-3 to 10-7 mbar

Inverted Magnetron Gauge

● similar principle, but works down to ~10-10 mbar

● single initial ionisation event gives electron

● electrons perform cycloid motion in crossed E and B fields, ionising gas within

● produces stable discharge similar to Penning gauge

● some pumping effect

● considerable sputtering at high pressures (avoid >10-5 mbar for long periods - deposits can cause leakage currents)

● provides alternative to ion gauge

► no filaments to blow

► no light, no heating

► but, slow to start at low pressures and requires big magnet

fewkV

B

Precision Vacuum Gauges

● most gauges are only accurate to ~20% at best (often worse)

● a few 'precision' gauges are available

► capacitance manometer (limited to ~10-5 mbar)

► spinning rotor gauge (limited to ~10-7 mbar)

● precision gauges are expensive, and limited range but can be used to calibrate ion gauges etc. when pressure critical Spinning Rotor Gauge

Vacuum Pumps

● wide variety of pumps used at all pressure levels

● concentrate on main types of pumps used in research

PositiveDisplacement

Kinetic Entrapment

Moving components displace and eject a volume of gas.

e.g. reciprocating piston or rotating vanes

Gas Transfer

Momentum imparted to individual gas particles, driving them to exhaust

e.g. rotor in a turbo-molecular pump

Gas particles react chemically and trapped, or ionised, accelerated and embedded in pump walls.

All Vacuum Pumps

Vacuum Pumps

103 100 10-3 10-6 10-9 10-12

Pressure (mbar)

Viscous Flow Molecular Flow

Mechanical (Rotary) Pump

Diffusion Pump

Turbomolecular Pump

Rotary Vane Pump

● 'standard' mechanical pump, used to achieve rough vacuum

● sliding vanes rotate, compressing and ejecting gas to atmosphere

● sealed with oil

► needs replacing periodically

► need foreline trap to keep oil from backstreaming into inlet lines

► special oils available for pumping oxygen/aggresives - else BANG!

Single stage rotary vane pump

Rotary Vane Pump

Inlet Connection(to chamber)

Flexible pipelineto chamber

Motor

Pump housing(oil casing)

Oil sightglass

Exhaust conn.(always exhaustto roof/outside -SAFETY ISSUE)

Foreline trap(catches oil inadsorbant beads -regen. periodically)

Rotary Vane Pump

● 2 stage pump gives better ultimate pressure

► 2 stages in series

● difficult to expel condensables

► e.g. water vapour

► gas ballast helps (leak air in between stages)

● typically get down to 2x10-3 mbar with a good rotary pump

● pumping speeds between about 0.5 and 80 m3/hr

● cost ~£1000 for few m3/hr

Two stage rotary vane pump

'Dry' Pumps

● can avoid contamination altogether with dry pumps, if necessary

► e.g. silicon wafer processing

● several mechanisms available, e.g. diaphragm, PTFE sealed pistons

● Dry pumps generally:

► lower pumping speed for same size or cost

► poorer ultimate pressures

► more expensive

► noisier

Other mechanical pumps

● Other mechanical pumps are available, e.g.:

► piston pumps

► roots pump

● Usually designed for high pumping speeds needed in industry

Large roots and piston pumpsused on a helium beam source

Rootspump

15cm dia.inlet flange

Pistonpump

Principle ofa roots pump

Getting to High Vacuum

● Need different type of pump to get below about 10-3 mbar

● Usually means using either:

► diffusion pump

► turbo pump

● High vacuum pumps can't discharge to atmospheric pressure

► permanently need a rotary pump as a 'support' or 'backing' pump

Chamber

HVPump

BackingPump

Vent toroof

1000 mbar

~10-2 mbar

<10-6 mbar

Diffusion Pump

● boils highly refined, high molecular weight fluid

● vapor jets impart downward momentum to gas entering pump

● oil condenses on water cooled body & recycled

● Discharges to rough vacuum

► 'critical backing pressure' ~0.5 mbar (depends on model)

● reliable 'standard', gets down to between ~10-7 and ~10-11 mbar depending on setup

Diffusion Pump

● Ultimate pressure depends on quality of fluid used

► cheap fluid (e.g. Corning DC704) ~10-6 mbar

► good fluid (e.g. Edwards L9) ~ 10-9 mbar

► best fluid (Santovac 5) ~10-10 mbar (with baffle)

● Different fluids have different safety issues

● Need chevron baffle to catch backstreaming pump fluid and achieve best pressures Heater

WatercoolingWatercooling

ThermalSensor

Inletbaffle

Diffusion Pump: Chevron Baffles

● all diff pumps backstream pump fluid when hot

● cold, optically dense baffle catches oil, but reduces the pumping speed

● need liquid nitrogen temp. to reduce vapour pressures to UHV level

► once nitrogen reservoirs filled, need to be kept filled

► internal water condensation can re-freeze in cracks and cause leaks

To diffusion pumpTo diffusion pump

To rest ofsystem

Liquid N2

reservoir

Optically densechevron baffle,connected tocold reservoir

Diffusion Pump

Advantages:

● fairly cheap (start ~£1000)

● reliable - little to go wrong

► heater is replaceable

► cooling coils can be descaled

► can take to bits and scrub out inside

● often found 'lying around' lab

Disadvantages:

● slow to start and stop

► ~½ hr to warm up

► ~1 hr to cool down before venting

● expensive to run large pumps (£1000s per year)

● require liquid nitrogen baffles for true UHV (require daily filling!)

● Dependant on cooling water – MUST BE INTERLOCKED

Turbomolecular Pump

● fast moving rotors and stators impart momentum to gas molecules

● frequencies of up to ~1000 Hz

● high precision greased/oiled or magnetic bearings

Turbomolecular Pump

● pumping speed varies with gas and pressure

● compression varies with molecular mass

► typically 109 for N2

► typically ~104 for He

► typically ~103 for H2

● difficult to remove lighter gases

► can add extra pump in series to improve compression

► add chemical pump in parallel to pump reactive species (e.g. H

2)

Chamber

TurboPump

BackingPump

Vent toroof

TurboPump

GetterPump

Turbomolecular Pump

Advantages:

● quick to start and stop

● low electrical costs

● clean – completely so for a mag-lev - no baffles required

● can mount in more orientations

● water cooling much less critical

► pump should turn off if too hot, usually after >1/2 hr

► air cooling options available

● mag-lev turbos good for aggressive gases

Disadvantages

● much more expensive to buy and service (typically around £10 000)

● occasional catastrophic failure

► something dropped in top

► bearing seizes

► magnetic controller fails

● need to be well secured (SAFETY RISK)

► e.g. Leybold T1600 has to be secured to withstand an impact torque of 20 000 Nm

Ion Pump

● type of entrapment pump, operating like a series of Penning gauges

● electrons spiral in strong magnetic field, ionising gas species on impact

● ions are accelerated to and get buried in surface of anode

● clean pumping to UHV, but requires large magnet

● needs additional pump to get down to about 10-4 mbar before ion pump will start

B

Ion pump cell

+kV

-kV

Getter Pump

● chemical 'getter' pump, reacts with and contains reactive elements

● various types of getter

► sublimated coatings (TSP uses Ti/Mo alloy)

► blocks of reactive, sintered material (NEG)

● particularly good at pumping H

2, unlike many other pumps

● limited capacity - has to be used at very low pressures, e.g. in a load-locked chamber

● cannot pump inert/rare gases

● use strip getters along entire length of particle accelerator beamlines to achieve XHV

● once activated, pumps without power - portable!

Choice of Pumps

● Choice of pump depends on

► level of vacuum

► size of chamber (outgassing)

► gas throughput of process

► pumpdown time

► type of gas (nasty?)

► level of cleanliness required

► size and positioning

► finances available (!)

► ...

● careful consideration and risk assessment is vital

● number of big mfrs.:

► never pay list price for new vacuum equipment

► discounts of up to ~50% are routine for academia

Low to High Vacuum Fittings

High vacuum fittings:

● elastomer seals

● usually use KF (Klein Flange) fittings with o-rings

Ultra High Vacuum Fittings

To achieve UHV:

● not difficult, just need to do the job properly and cleanly

● require materials which are not gassy

► all metal fittings, plus ceramics, glass(fibre), PTFE

► no plastics, adhesives, solder, brass

► http://outgassing.nasa.gov

● usually use CF (Conflat) knife edge flanges and copper gaskets

► can seal with elastomer gaskets temporarily (indicate by using few bolts) Knife Edge

Vacuum Feedthroughs

● Need electrical, fluid and mechanical connections to experiment

► commercial feedthroughs

● Electrical connections

► ceramic to metal seals

► delicate - avoid bending pins!

● Mechanical connections

► o-ring sealed for HV

► differentially pumped or metal bellows sealed for UHV

Push-pull linear feedthrough

metal bellows to allowmotion in vacuum

Electrical feedthrough

metal flange

vacuum sideceramicinsulator

Vacuum Valves

Gate valve(high conductance + UHV)

Butterfly valve(high conductance)

HV 90° valve UHV 90° valveDiaphragm valve(typically on backing lines)

Putting it all together

Standard Vacuum Symbols

Vacuum Pump,General

Rotary VanePump

DiffusionPump

TurbomolecularPump

AdsorptionPump

Cryopump

Sputter-IonPump

GetterPump

ManualValve

VariableLeak Valve

Electromag.Valve

Cold Trap

SorptionTrap

VacuumGauge

Another Simple System

GetterPump

UHVGateValve

Backing Line

Exhaustto roof

Exhaustto roof

Leak Test

Roughing Line

TurboPump

All MetalValve

UHVChamber

Reaching the Ultimate Pressure

Normal procedure:

● Design system carefully

● Avoid contamination

► degrease and clean all components

► assemble with clean gloves

● Locate and address leaks

● Bake the system to achieve UHV

Then, if you have problems...

● Try to identify source of contamination

● Check system processes

● Check system design and construction

► a single unsuitable component will limit the pressure achieved

► avoid gassy materials

Cleaning How-To

● Not very exciting, but very important to achieving good high vacuum performance

● Specialist cleaning possible for certain materials

► ceramics, glass

► see textbooks

● Some components difficult to clean:

► clean as well as possible by hand

► then use ultrasonic cleaner

1. Start with 'mechanical cleaning' (scrub with detergent)

2. 'Buzz' in detergent in ultrasonic cleaner for 5-15 minutes.

3. Rinse in clean tap water

4. Rinse in deionised water

5. Dry carefully

6. Rinse in solvent

A generic procedure:

Leaks

● Often present - is it important?

● Need to work out the source of the gas to fix the leak

► seal off system, monitor pressure

► leak test

► analyse gases in chamber

● Fix real leaks:

► re-seal leaking flanges

► re-weld crack in welding (avoid leak-sealant!)

► leak-seal cracks in feedthrough connections

pressure

tim

e

pressure

tim

e

pressure

tim

e

real leak

virtual leak

both

Virtual Leaks

● Trapped volumes of air (or other gas / vapour) constitute a virtual leak

● Difficult to 'degas' - can take a long time

● Avoid creating virtual leaks by careful design in the first place:

► use internal welding

► vent trapped volumes directly

► drill screw holes right through

Helium Leak Testing

● Almost VITAL to have some leak testing facilities - especially at UHV level

● Two common approaches

► apply acetone + look for ANY pressure change

► use mass-spec helium leaktester

● Helium leak-tester is a simple mass-spec designed to detect helium

► connect up and spray Helium around potential leaks

► versatile technique, used widely outside 'vacuum field'

Chamber

HVPump

BackingPump

HeliumLeak Tester

Heliu

m

Helium Leaktest How-To

1. First try to detect the leak with acetone

2. Connect the leaktester to the system (~4 helium leak testers around Cavendish)

1. use a leak test port, or

2. replace one of the backing pumps

3. Apply a fine jet of helium from a leak test cylinder (borrow from liquifier facility)

4. Follow response of leaktester over period of 10's of seconds (depending on pumping path)

5. Use as little helium as possible

6. Ventilate the area

7. Work from the top down

8. Isolate sections of vacuum system in plastic bags

9. If gas load is small, divert entire backing load through leaktester, to maximise response.

10.Once leak identified, locate its position precisely to help fix it

Bakeout

● Generally necessary to achieve p < 10-8 mbar without waiting too long

● Bake for as hot and as long as possible!

● Max. temperature determined by components

► Viton seals limit ~150°c

► PTFE limit ~200°c

► All metal systems are OK to higher temperatures

Baking allow UHV to be reached in days, rather than months or years

Bake OffBake On

Bakeout How-To

1. Check that the components in the system are OK for baking & decide on max. temp.

2. Disconnect all non bakeable connections/fittings and store

3. Wrap sensitive components (feedthroughs, windows) in foil.

4. Position thermocouples at various characteristic points of the system (test them now!).

5. Either:

1. Wrap chamber in heating tape and several loose layer of foil

2. Enclose chamber in custom made oven

6. Do final check on bake zone

7. Slowly increase temperature to e.g. 200°c for 24 hours (heat up over a period of a few hours)

8. Monitor temperatures and ensure an even heat - avoid cold spots

9. Hold at temperature until pressure sufficiently low

10.Cool down over a few hours

11.Degas filaments while chamber still warm

RGA Mass Spectrometer

● Provides lowest pressure measurements - down to pp. of ~10-15 mbar

● provides pressure breakdown by mass number

● allows analytical measurements and diagnosis

● modern systems are run from a PC

● start at ~£5000

ioniserion

countermassfilter

ResidualGases

controlelectronics

RGA of a High Vacuum System

18 - water, chamberneeds baking

28, 32 - N2 (or CO)

and O2, suggest a

leak to atmosphere

4 (He), from process, or diffusing through elastomer gaskets

Forest of peaks around 40-70 suggests pump oil contamination (fit baffle or trap to pump)

only 2 (H2), 28 (CO)

and 44 (CO2) should

be present inclean, leaktight UHV

Cracking Patterns

● different species give different mass patterns

► cracking, isotopic abundances etc.

● allows individual components to be identified

► understanding a process

► diagnosing contaminants

● cracking patterns tabulated by RGA mfrs. and in O'Hanlon

Species Mass Abund. Mass Abund. Mass Abund.Hydrogen 2 100 1 2.7 3 0.31Helium 4 100 2 0.12Methane 16 100 15 83 14 15Water 18 100 17 27 16 3.1Argon 40 100 20 5.0 36 0.36Acetone 43 100 58 27 27 8.0

Good Working Practice for HV / UHV

Do:

● keep a log of the behavior of system - it can help identify problems

● record reference RGA spectra of the chamber, if possible

● vent system to dry nitrogen

● open smallest flanges possible (keep under slight overpressure of nitrogen?)

● ensure regular maintenance on pumps (if possible!) - no one else will!

Don't:

● vent HV / UHV systems with air

● vent system while components are cold (e.g. nitrogen traps, cryopumps)

● handle any components without gloves

● get any fluid/oil on your skin

● burn yourself on diff. pumps(!)