Vacuum II
G. Franchetti
CAS - Bilbao
30/5/2011
G. Franchetti
1
Index
Creating Vacuum (continuation)
Measuring Vacuum
Partial Pressure Measurements
30/5/2011
G. Franchetti
2
Diffusion Ejector pump
operating pressure:
10-3 – 10-8 mbar
Laminar flow
Schematic of the
pump
Cold surface
Inlet
Outlet
Boiling oil
30/5/2011
G. Franchetti
3
Inlet
Pi
Laminar flow
Cold surface
Pump principle:
The vacuum gas diffuses into the jet
and gets kicked by the oil molecules
imprinting a downward momentum
Po
Outlet
The oil jets produces a skirt which
separate the inlet from the outlet
Boiling oil
30/5/2011
G. Franchetti
4
Inlet
Diffusion
Outlet
30/5/2011
G. Franchetti
5
Problems
Inlet
Pi
Back streaming
Cold surface
Back-Migration
Po
30/5/2011
G. Franchetti
6
Cures
Inlet
Baffle
Cold Cap
Cold surface
Pi
Cold surface
(reduces
the pumping
speed of 0.3)
Po
30/5/2011
G. Franchetti
7
with
The pumping speed Sm is proportional to the area
of the inlet port
100 mm diameter  Sm = ~ 250 l/s for N2
LEYBOLD (LEYBODIFF)
30/5/2011
G. Franchetti
8
Capture Vacuum Pumps
Getter Pumps
Principle
(evaporable, non-evaporable)
Capture vacuum pumps are based on the
process of capture of vacuum molecules
by surfaces
Sputter ion Pumps
Cryo Pumps
30/5/2011
G. Franchetti
9
Getter Pumps
Gas
Surface
Solid
Bulk
30/5/2011
G. Franchetti
10
Definition of NEG
Getters are materials capable of chemically adsorbing gas molecules. To do so their surface
must be clean. For Non-Evaporable Getters a clean surface is obtained by heating to a
temperature high enough to dissolve the native oxide layer into the bulk.
T = Ta
T = RT
T = RT
Native oxide layer
-> no pumping
Heating in vacuum
Oxide dissolution -> activation
Pumping
P. Chiggiato
NEGs pump most of the gas except rare gases and methane at room temperature
11
Sorption Speed and Sorption Capacity
C.Benvenuti, CAS 2007
30/5/2011
G. Franchetti
12
Choice of the coating technique for thin film: sputtering
substrate to coat: vacuum chamber
target material: NEG (cathode)
driving force: electrostatic
energy carrier: noble gas ions
NEG composition
Ti
V
Zr
 The trend in vacuum technology consists in moving the pump progressively closer to the vacuum chamber
wall.
 The ultimate step of this process consists of transforming the vacuum chamber from a gas source into a
pump.
 One way to do this is by “ex-situ” coating the vacuum chamber with a NEG thin film that will be activated
13
during the “in situ” bakeout of the vacuum system.
Dipole Coating Facility
Quadrupole Coating Facility
M.C. Bellachioma (GSI)
30/5/2011
G. Franchetti
14
Sputter ion pumps
Cathode
Anode
-
+
E
-
E
P2
B
Titanum
P1
30/5/2011
G. Franchetti
15
Sputtering process
+
trapped electrons
30/5/2011
The adsorbing material is sputtered around the pump
G. Franchetti
16
A glimpse to the complexity
ion bombardment
with sputtering
Adsorbed ions
ionization
process
Trapped electron
30/5/2011
G. Franchetti
17
Example of Pumping speed
J.M. Laffterty, Vacuum Science, J. Wiley & Son, 1998
30/5/2011
G. Franchetti
18
Cryo Pumps
m
Stick to the
Wall !!
v1
Cold Wall
Dispersion forces between molecules and surface
are stronger then forces between molecules
30/5/2011
G. Franchetti
19
Schematic of a cryo pump
Vessel
nw
Pump
pw
= pressure warm
= pressure in the cold
= flux Vessel  Pump
= flux Pump  Vessel
n c pc
cold
30/5/2011
Pw
Pc
Iw
Ic
molecules stick to the cold wall
G. Franchetti
20
By using the
state equation
Now
In the same way
If
no pumping although
Relation between pressures
Thermal Transpiration
30/5/2011
G. Franchetti
21
When the two pressures breaks the thermal transpiration condition a particle flow starts
We define
and
We find
Pc depends on the
capture process
30/5/2011
Cryocondensation: Pw is the vapor pressure of the gas at Tc
G. Franchetti
22
Summary on Pumps
30/5/2011
G. Franchetti
N. Marquardt
23
Gauges
Liquid Manometers
MacLeod Gauges
Viscosity Gauges
Thermal conductivity Gauges
Hot Cathode Gauges
Alpert-Bayard Gauges
Penning Gauges
30/5/2011
G. Franchetti
24
Liquid Manometers
P2
P1
Relation between the two pressure
h
issue: measure precisely “h” by eye +/- 0.1 mm
but with mercury surface tension depress liquid surface
High accuracy is reached by knowing the liquid density
Mercury should be handled with care: serious health hazard
30/5/2011
G. Franchetti
25
McLeod Gauge
First mode of use:
P2
The reservoir is raised till the mercury reaches the level
of the second branch (which is closed)
Closed
Δh
h0
Quadratic response to P2
Second mode of use: keep the distance Δh constant and measure the distance of the
two capillaries  linear response in Δh
30/5/2011
G. Franchetti
26
Viscosity Gauges
ω
φ
Gas
molecule
It is based on the principle that gas molecules
hitting the sphere surface take away rotational
momentum
R
The angular velocity of the sphere decreases
ρ = sphere density
α = coefficient of thermal dilatation
T = temperature
va = thermal velocity
30/5/2011
G. Franchetti
27
(K. Jousten, CAS 2007)
F.J. Redgrave, S.P. Downes, “Some comments on the stability
of Spinning Rotor Gauges”, Vacuum, Vol. 38, 839-842
30/5/2011
G. Franchetti
28
Thermal conductivity Gauges
A hot wire is cooled by the
energy transport operated by
the vacuum gas
hot wire
Accommodation factor
By measuring dE/dt we measure P
30/5/2011
G. Franchetti
29
Energy Balance
 Energy loss by gas molecules
Wc/2
WR
 Energy loss by Radiation
Wc/2
V
ε = emissivity
 Energy loss by heat conduction
When the energy loss by gas molecule is dominant
P can be predicted with contained systematic error
30/5/2011
G. Franchetti
30
Pirani Gauge
Example of power dissipated by
a Pirani Gauge vs Vacuum pressure
The Gauge tube is kept at
constant temperature and
the current is measured
So that
.
Through this value  E
30/5/2011
G. Franchetti
31
Ionization Gauges Principle
electrons
+
V
L
accelerating
gap
30/5/2011
G. Franchetti
32
E
new
electrons
i+ = current proportional
to the ionization rate
+
i-
V
positive
ions
30/5/2011
G. Franchetti
33
Electrons ionization rate
Sensitivity
30/5/2011
G. Franchetti
34
Hot Cathode Gauges
Electric field
Anode
Grid
30 V
180V
+
+
30/5/2011
Hot
Cathode
G. Franchetti
35
Hot Cathode Gauges
In this region
the electrons
can ionize vacuum
particles
Anode
Grid
i+
30 V
180V
Hot
Cathode
+
A current between grid
and anode is proportional
to the vacuum pressure
+
i+
30/5/2011
G. Franchetti
36
Limit of use
Upper limit:
it is roughly the limit of the linear response
Lower limit:
X-ray limit
Anode
X-ray
new
electron
The new electron change the
ionization current
Pm
X-ray
limit
X-ray
Grid
P
30/5/2011
G. Franchetti
37
Alpert-Bayard Gauge
The X-ray limit is easily
suppressed by a factor
~100-1000
X-ray
Hot
Cathode
Anode
X-ray
the probability
that a new electron
hits the anode is now
very small
Grid
30/5/2011
G. Franchetti
38
Schematic of the original design (K. Jousten, CAS 2007)
30/5/2011
G. Franchetti
39
Penning Gauge (cold cathode gauges)
anode
cathode
E = electric field
B = magnetic
field
Electrons motion
30/5/2011
G. Franchetti
40
Under proper condition of (E,B) electrons get trapped in the Penning Gauge
-
+
+
One electron is trapped
+
+
+
+
+
+
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
30/5/2011
+
+
+
+
+
+
+
+
+
-
G. Franchetti
41
-
+
+
One electron is trapped
+
+
+
+
+
+
-
-
-
-
-
-
one vacuum
neutral gas
enters into the
trap
-
-
-
-
-
-
30/5/2011
+
+
+
+
+
+
+
+
+
-
G. Franchetti
42
Ionization process
Collision  Ionization
Before Collision
Neutral vacuum atom
-
Charged vacuum atom
-
Electron at ionization
speed
30/5/2011
Electron at
ionization
speed
G. Franchetti
+
-
New
electron
43
-
+
+
+
-
+
+
Ionization
-
Charged vacuum atom
-
-
-
30/5/2011
-
-
Electron at
ionization
speed
-
-
-
+
-
-
+
+
+
+
+
+
this ion has a too
large mass and relatively
slow
therefore+ velocity,
+
its motion is dominated
by the electric field and
not by the magnetic field
New
electron
+
+
+
+
-
G. Franchetti
44
-
+
+
+
-
+
+
+
Charged
vacuum atom
Ionization
-
Motion of stripped ion
+
+
-
+
-
-
-
-
New
electron
-
30/5/2011
-
-
-
-
+
+
+
+
+
+
+
+
+
-
G. Franchetti
45
More electrons are formed through the ionization of the vacuum gas and
remains inside the trap
-
+
+
- +
- +
+
+
- +
+
+
- +
-
+
+
+
-
-
+
+
+
+
+
+
+
+
+
+
+
-
-
+
+
+
+
-
+
+
+
-
+
-
-
Negative space charge formation
30/5/2011
G. Franchetti
46
When the discharge gets saturated each new ionization produce a current
- +
- +
+
+
- +
+
+
- +
-
+
+
+
-
+
-
-
-
-
-
+
+
+
+
-
+
+
+
+
+
+
+
+
+
+
--
-
30/5/2011
+
-
+
+
+
+
+
I
-
-
G. Franchetti
47
The time necessary for the discharge formation depends upon the
level of the vacuum
lower pressure  longer time of formation
Sensitivity of a SIP (the same as
for the Penning Gauge).
J.M. Lafferty, Vacuum Science,p. 322
30/5/2011
G. Franchetti
48
Summary on Gauges
N. Marquardt
30/5/2011
G. Franchetti
49
Partial Pressure Measurements
These gauges allow the determination of the gas components
Partial pressure gauges are composed
Ion
Source
30/5/2011
Ion current
detection
System
Mass
Analyzer
G. Franchetti
data
output
50
Ion Sources
Vacuum gas is ionized via
electron-impact

the rate of production of ions is
proportional to each ion species
Electron-impact ionization process
Before Impact
e
After impact
M
v
M+
inelastic scattering:
kinetic energy transfer
from the electron to
the molecule
e m
e
me
e
v
me
v
30/5/2011
G. Franchetti
51
At the appearance potential the production
rate is low
The minimum energy to ionize M
is called “appearance potential”
Electron
energy
(eV)
15
IP = 15 eV
M+ + e-
appearance
potential
30/5/2011
G. Franchetti
A. von Engel, Ionized Gases, AVS Classics Ser., p. 63. AIP Press, 1994
52
Ion source
acceleration gap
Open source
Bayard-Alpert gauge
electrons
source
ion collector
ionization
region
ions
electrons
E
+
Vf
E
+
+
Vg
30/5/2011
G. Franchetti
Vg < Vc
Vc
53
A schematic
+
+
+
+
+
+
+
+
Mass analyzer
F = ion transmission factor
Ion Detection
30/5/2011
G. Franchetti
54
Ion Detection
 Faraday Cup
 Secondary electron Multiplier (SEM)
Gain
G=
# electrons output
# ion input
Idea: an ion enters into the tube, and due to the potential
is accelerated to the walls. At each collision new electrons are
produced in an avalanche process
V0 = applied voltage
d
V = initial energy of the electron
L
K = δ Vc
where
J.Adams, B.W. Manley IEEE Transactions on Nuclear Science, vol. 13, issue 3, 1966. p. 88
30/5/2011
G. Franchetti
δ = secondary emission coefficient
Vc = collision energy
55
Mass Analyzers
Quadrupolar mass spectrometer
r0
y
s
Ion equation of motion
U, V constant
x
define:
30/5/2011
G. Franchetti
56
By rescaling of the coordinates the equation of motion becomes
Stability of motion
Mathieu
Equation
The ion motion can be stable or unstable
horizontal
vertical
q=0, a>0
unstable
stable
q=0, a<0
stable
unstable
The presence of the term q, changes the stability condition
Development of stable motion
Stable or unstable motion is referred to a channel which is infinitely long, but typically
a length correspondent to 100 linear oscillation is considered enough
Example:
 vs = 8301 m/s
For N2 at Ek = 10 eV
For a length of L = 100 mm
100 rf oscillation
30/5/2011
G. Franchetti
f = 8.3 MHz
typically
f ~ 2 MHz
57
a
Stability Chart of Mathieu equation
10
10
a
5
5
Stable in y
0
5
10
5
10
15
15
5
5
Stable in x
10
10
30/5/2011
G. Franchetti
58
q
0
q
a
10
a
a
5
q0=0.706
a0=0.237
q
0
5
10
15
q
5
Stability region
in both planes
10
30/5/2011
G. Franchetti
59
a
Given a certain species
of mass M1 there are two
values U1, V1 so that
q=q0 and a=a0
(0.706; 0.237)
By varying V and keeping
the ratio V/U constant, the
tip of the stability is crossed
and a current is measured at V=V1
q
Stability region
in both planes
30/5/2011
i+
V
G. Franchetti
60
Magnetic Sector Analyzer
Example: A 900 magnetic sector mass spectrometer
B is varied and when a
current is detected then
Ez is the ions
kinetic energy
Resolving power
Wsource = source slit width
Wcollector = collector slit width
30/5/2011
J.M. Lafferty, Vacuum Science, p. 462
G. Franchetti
61
Omegatron
30/5/2011
G. Franchetti
62
The revolution time is independent
from ion energy
collector
B
If the frequency of the RF is 1/τ
a resonant process takes place
and particle spiral out
E
Resolving power
30/5/2011
G. Franchetti
63
Conclusion
Creating and controlling vacuum will always be a relevant
part of any accelerator new development.
THANK YOU FOR YOUR ATTENTION
These two lectures provides an introduction to the
topic, which is very extensive: further reading material is
reported in the following bibliography
30/5/2011
G. Franchetti
64
Kinetic theory and entropy ,C.H. Collie, Longam Group, 1982
Thermal Physics, Charles Kittel, (John Wiley and Sons, 1969).
Basic Vacuum Technology (2nd Edn), A Chambers, R K Fitch, B S Halliday, IoP
Publishing, 1998, ISBN 0-7503-0495-2
Modern Vacuum Physics, A Chambers, Chapman & Hall/CRC, 2004, ISBN 0-84932438-6
The Physical Basis of Ultrahigh Vacuum, P. A. Redhead, J. P. Hobson, E. V.
Kornelsen, AIP, 1993, ISBN 1-56396-122-9
Foundation of Vacuum Science, J.M. Lafferty, Wiley & Sons, 1998
Vacuum in accelerators, CERN Accelerator School, CERN-2007-003 11 June 2007
Vacuum technology, CERN Accelerator School, CERN 99-05 1999
Acknowledgements: Maria Cristina Bellachioma
30/5/2011
G. Franchetti
65