Hoffman D.M., Singh B., Thomas J.H. (Eds). Handbook of Vacuum Science and Technology
Подождите немного. Документ загружается.
502
Chapter 4.7: Outgassing of Materials
Table 7
Outgassing Rate Constants for Metals and Ceramics Calculated from Experimental Data
Material
Alundum (901 Norton)
Aluminum (spray-coated mild steel)
Copper
Degussite (AI2O3)
Graphite wool (Morganite)
Iron
Mild steel (slightly rusty)
Molybdenum
Mullite (800, United Fireclay)
Mullite (Morganite)
Nickel
Porcelain, glazed
Pyrex
Pyrophyllite
Pyrophyllite (fired)
Sillimanite (Zirconal)
Silver
Stainless steel (ICN sanded)
Stainless steel (ICN fresh)
Stainless steel (NS22S)
Stainless steel (NS22S electropolish)
Steatite (AI2O3)
Steel (chromium plated, fresh)
Steel (nickel plate)
Steel (niclel plated, fresh)
Steel (chromium plated, polished)
Tantalum
Tungsten
Zirconia (Zirconal Ltd.)
Zirconium
A,n
39
65
12
12
39
12
—
12
39
39
12
30
103
12
39
39
12
10^
103
103
103
30
103
103
103
103
12
12
39
12
s.
1
0.4
0.08
0.1
1
0.07
1
0.12
1
1
0.08
0.7
1
0.1
1
1
0.07
1
1
1
1
0.7
1
1
1
1
0.07
0.07
I
0.15
10^/r,,.,
400
60
20
75
1000
5
600
14
800
620
10
6.5
7.2
20
1500
500
50
8.3
13.5
14.4
4.3
90
7.1
2.8
4.2
9.1
9
7
1200
8
Oi\
1
0.75
1.1
3.1
0.5
1.1
1.2
0.8
0.6
1.2
0.9
1.3
1.0
1.1
0.9
Ref.
No.
[14]
[5]
[4]
[13]
[14]
[4]
[3]
[4]
[14]
[14]
[4]
[1]
[11]
[13]
[14]
[14]
[4]
[11]
[11]
[11]
[11]
[1]
[11]
[11]
[11]
[11]
[4]
[4]
[14]
[4]
Note:
A„ = exposed geometric area of outgassing material (cm^)
5a = pumping speed for air at 25°C (L/s)
K^i
= measured outgassing rate after
1
h of pumping (torr L/s
"'
cm ~^)
a
I
= absolute value of slope of log-log graph of measured outgassing rate vs. time after 1 h of
pumping
References
1.
R.Geller.L^VzV/^, 13(74) (1958) 71.
2.
J. C. Boulassier, Le Vide, 14(80) (1959) 39.
3.
B. D. Power and D. J. Crawley,
Trans.
1st
Internat.
Vac.
Congress, Namur, Vol. 1 (Pergamon
Press,
New York, 1960), pp. 206-211.
(continues)
4.7.5 The Outgassing Rate of Metals and Ceramics 503
Table 7—Continued
4.
R. Jaeckel and F. J. Schittko,
Forschimgsberichte
des
Wirtscliafts-
iind
Verkehrsministenum
Nordrhein-Westfalen, Nr. 369. Westdeutscher Verlag, Cologne, 1957.
5.
B. B. Dayton, Trans. 6th Nat. Vac. Symp. (Pergamon Press, New York, 1960), pp. 101-119.
6. G. Thieme, Vacuum, 13 (1963) 137.
7.
J. Blears, E. Greer, and J. Nightingale, Advances in Vacuum Science and Technology, vol. 2,
edited by E. Thomas (Pergamon Press, New York, 1960), pp. 473-479.
8. W. Beckmann, Vacuum, 13 (1963) 349.
9. B. B. Dayton, J.
Vac.
Sci.
TechnoL,
A13 (1995) 451.
10.
H. von Munchausen and F. Schittko, Vacuum, 13 (1963) 549.
11.
A.Schram,L€V/Je, 18(103)(1963)55.
12.
E J. Schittko, Vacuum, 13 (1963) 525.
13.
R. Jaeckel, Trans. 8th Nat. Vac. Symp. and 2nd
Internat.
Vac. Congress, vol. 1 (Pergamon
Press,
New York, 1962), pp. 17-26.
14.
B. H. Colwell, Vacuum, 20 (1970) 481.
cal equations
^n =
K,,,,
[1
+ (A,/5,)/w^]/(2^op/e) (47)
where/w =14.8 liters
s"^cm"'^
for water vapor at 25°C, the sticking probability
cr
is arbitrarily set equal to 6 X
10
"^ as a constant average value [13], and at 25°C
^op = (5.2lXlO-Ve)-^^ (48)
is the fraction of the available sites for physisorption of H2O in a precursor state
that would be actually occupied by H2O in equilibrium, according to a Freundlich
isotherm, with the partial pressure
Pe
at 25°C (^^p = 0.5 at the saturation pressure
23 torr), prior to dissociation and chemisorption, and at 25°C
/e = [l-exp(-re/4.46)]-^ (49)
gives the fraction of the saturation coverage ^op attained after an exposure to the
partial pressure
Pe
for
/g
hours according to a semiempirical equation based on the
data of Gakon [14]. The numerical constants in Equations (48) and (49) will be
different at temperatures other than 25°C.
When (A,„/5w)/w^ ^^
1»
the rate of readsorption is so high that the outgas-
sing process is near the state of equilibrium corresponding to an adsorption-
desorption isotherm. Since it can be shown that the activation energy for desorp-
tion of
H2O
from the oxide layer varies with the fractional coverage ^ in a manner
similar to that for a Temkin isotherm, several authors [15,16,17] have modeled
the outgassing process in terms of a quasi-equilibrium approach to a Temidn iso-
therm, either the simple isotherm for a uniform surface or the more complicated
isotherm equation for nonuniform surfaces [18]. These models predict different
dependence of outgassing rate on temperature, but only limited experimental data
[19] exist with which to determine the correct model.
504 Chapter 4.7: Outgassing of Materials
The measured value K„ji for metals exposed to moist atmosphere prior to
pumpdown should preferably be reported using the net pumping speed, S^,, for
water vapor, but in most cases the reported value appears to be for the air pump-
ing speed, S^. If the pressure is measured with an ion gauge using the air cali-
bration factor, the indicated pressure,
p^^^.,
will be lower than the true pressure,
/7^,
because the gauge sensitivity for H2O is about 0.75 times the sensitivity for
air (or
A^2)
^^^ H2O usually constitutes about 90% of the total pressure due to
outgassing from an unbaked sample. Since SJS^, = 0.79, the product p^^S^ =
0.6 p^^S^ approximately. The measured outgassing rate Ar,„h =
p„S„/A„j,
and the
reported rate using air calibration will thus be somewhat less than the true out-
gassing rate. However, the behavior of a vacuum system in which pressure is
measured with ion gauges using the air calibration factor can still be predicted us-
ing the outgassing rate tables.
From Table
1
it is evident that the free outgassing rate for aluminum alloy is an
order of magnitude greater than that for stainless steel, which is presumably due
to the greater thickness of the oxide layer on the former
[20].
Although in general
the outgassing rate for water vapor is likely to increase with the thickness of the
oxide layer, various attempts to reduce the outgassing rate, in systems that cannot
be baked, by reducing the surface roughness and oxide layer thickness by polish-
ing and detergent cleaning have not been entirely successful, but in situ glow dis-
charge cleaning with He gas reduces the outgassing rate by an order of magnitude
[21,22].
4.7.6
OUTGASSING RATE OF PRECONDITIONED VACUUM
SYSTEM AFTER SHORT EXPOSURE TO THE ATMOSPHERE
4.7.6.1 Formula for Outgassing Rate
During Second Pumpdown
If the outgassing rate varies with time ti (in h) as
Ki = K^/ti" (50)
during the first pumpdown where a is nearly constant up to a time r^ >
1
h. then
we can designate the outgassing rate during the second pumpdown of a pre-
viously evacuated system by Ku and successive pumpdowns by appropriate
Roman numerals, while the subscript
1
always identifies the state of
a
variable af-
ter one hour of pumping on the first pumpdown. The dimensions of
K^
may be
taken as torr liter
s"^cm~^
provided it is realized that the preceding equation is
4.7.6 Outgassing Rate of Preconditioned Vacuum System 505
a simplification of the formula
A:,
=
A:,
t^vtf
(51)
where t^ = 1 h, so that in Equation (50) K^ is actually multiplied by a factor 1
having dimensions (h)".
For many materials that have not been previously degassed, the value of a is
between 0.3 and 1.0 when fi < r^. For most plastics and elastomers (except sili-
cone rubber) the value of a is approximately 0.5 corresponding to the ordinary
law of desorption limited by the rate of diffusion of gas (usually mosdy water va-
por) through the solid. For silicone rubber, unpolished metals, glass, and ceram-
ics,
the initial outgassing (due primarily to water vapor) corresponds to an expo-
nent a that is usually in the neighborhood of 1.0.
When a solid sorbent is not altered by excessive heating or sintering and the
gas content is not too high, the process of physical adsorption and absorption is
often simply the reverse of desorption. In the absence of externally generated
force fields, the diffusion coefficient for a gas thiough a solid is not changed when
the direction of the concentration gradient is reversed. Therefore, it is to be ex-
pected that the sorption rate of a partially degassed elastomer for water vapor, ni-
trogen, and oxygen when exposed to the atmosphere at a fixed relative humidity
will vary with time in a manner similar to the variation of the outgassing rate of
these gases with time.
When the sorption and desorption rates depend primarily on the diffusion of
water vapor and are reversible in the above sense, then it can be shown by an ap-
plication of Duhamel's theorem [23] that the outgassing rate, Kn, during the sec-
ond pumpdown, for materials that obey Equation (50) on the first pumpdown, is
given by the equation
K ^^ I ^^' ^^" (52)
where tu is the time (in h) from the beginning of the second pumpdown, t^ is the
conditioning time (total time in hours of the first pumpdown), t^ is the time in
hours during which the surface is exposed to atmosphere having a constant partial
pressure,
P^,
of water vapor after the first pumpdown, and as a first approximation
it is assumed that
Kn = (P^/PoVK, (53)
where n is a constant that depends on the adsorption isotherm,
PQ
is the partial
pressure of water vapor during exposure before the first pumpdown, and the sorp-
tion and desorption of nitrogen and oxygen is neglected. Equations (52) and (53)
seem to hold for outgassing from plastics and elastomers where a = j and n = I
(Henry's law), but experimental data on the outgassing of water molecules from
506 Chapter 4.7: Outgassing of Materials
the oxide layer on metals (for which
a
is in the neighborhood of 1 and n is less
than 1 corresponding to a Freundlich isotherm) indicate that Equation (52) is not
obeyed. The expected outgassing rate and system pressure on
a
second pump-
down for metal chambers that have been preconditioned by long pumping (with
or without baking) and then exposed to the atmosphere for a time t^ can be esti-
mated from
K„a
=
Kn(2d,^f,)/n
+
iAJS^)AcT]
(54)
using Equations (48) and (49), together with appropriate values ofKfi and ai from
Table 1.
Equation (52) is only valid when the venting and roughing times of the vacuum
chamber are small compared to t^ and t^, the material is maintained at constant
temperature, and the time tu is within die limits indicated by 0.01
<
tn
<
^i
<
^m»
where
ti = tu
+
fe
+
^c
(55)
When the conditioning time, t^, is long compared to
^n
and the exposure time is
brief so that 100 t^ < tu
<
10"^ t^. Equation (52) reduces to
Ku =
KnatJ{tur^' (56)
which can only be applied to elastomers and plastics with
a
about 0.5. For elas-
tomers with
a = J,
Equation (55) gives a slope of — y when plotted on log-log
graph paper as shown by Rogers [24]. When tu
+
fc
^
^m*
the slope can increase
to values less than —2. Log-log plots
of
an equation similar to Equation (52)
with
a = J
are given in Roger's article.
A general method of computing outgassing rate for elastomers and plastics af-
ter several pumpdown and exposure cycles is given in Reference [23].
4.7.7
METHODS OF DECREASING THE OUTGASSING RATE
4.7.7.1 Selection of Gasket Materials
If an elastomer gasket must be used, the preferred materials are
fluorinated
hy-
drocarbons such as Kakez or Viton-A, which if necessary can be baked at 150°C.
The exposed surface of gasket material can be minimized by proper design of the
gasket grooves
in
the flanges, but the effective exposed area depends on how
the gasket fits in the groove and usually is greater than the inside gap between
the flange surfaces. Metal gaskets and suitably designed flanges must be used on
4.7.8 Measurement of the Outgassing Rate of Materials 507
ultra-high-vacuum systems both because of the lower outgassing rate of metals
and the ability to bake the system without damaging the gaskets.
4.7.7.2 Recommended Bakeout Temperatures and Times
Systems with metal walls and gaskets and not containing substances such as plas-
tics,
brass, cadmium-plated metal, etc., which are damaged by temperatures over
200°C can be baked under vacuum for 24 hours at 200-250°C to remove most of
the water vapor adsorbed in the oxide layer on the metal walls [25]. The residual
outgassing will then usually be due to hydrogen dissolved in the metal. To lower
the outgassing rate for hydrogen appreciably it is necessary to bake at 450-
500°C for 12 to 24 hours [5]. The fabrication of the metal chamber from vacuum
degassed steel only reduces the outgassing rate by a factor of about 2. Pyrex glass
walls can be baked up to 450°C.
4.7.7.3 Cleaning Methods and Surface Treatment
for the Interior Walls of the Vacuum Chamber
The oxide layer on the metal walls can be reduced in thickness by glass bead
bombardment [26] (which may introduce undesired contaminants on the walls) or
mechanical polishing. Organic contaminants and water vapor can be partially re-
moved by vapor degreasing or by the use of glow discharges or electron bom-
bardment under vacuum and by flushing with dry nitrogen [27]. Surface rough-
ness can be reduced somewhat by electropolishing and chemical etching. Nitrile
or polyethylene gloves should be used when necessary to handle parts to be pro-
cessed in the vacuum chamber, and air conditioning to lower the atmospheric hu-
midity is desirable. Venting the vacuum system with dry nitrogen, and using a
positive pressure of dry nitrogen inside a covering of plastic sheeting with entry
flaps while removing and inserting parts to be processed, will reduce the adsorp-
tion of water vapor on the chamber walls. Parts to be processed should be stored
in sealed containers with dessicants.
4.7.8
MEASUREMENT OF THE OUTGASSING RATE
OF MATERIALS
Various methods for measuring the outgassing rate have been summarized by
Elsey [28]. Recommended practices for measuring and reporting outgassing rates
508 Chapter 47: Outgassing of Materials
have been published by the American Vacuum Society [29]. Frequently, reported
outgassing rates have been much lower than the "free" outgassing rate because of
the use of a small ratio of pumping speed to sample area, resulting in appreciable
readsorption of gas. The reported rates are thus of little value unless SIA and
the sticking coefficient a are known. The direct molecular beam method of mea-
surement [30] avoids this problem but may give incorrectly high values due to
noncosine-law distribution of molecules in the intercepted beam.
REFERENCES
1.
B. B. Dayton, Vacuum, 15 (1965) 53.
2.
B. B. Dayton, Trans. 6th Nat.
Vac.
Symp. (Pergamon Press, New York, 1960), pp. 101-119.
3.
B. B. Dayton, Trans. 8th Nat.
Vac.
Symp. (Pergamon Press, New York, 1962), pp. 42-57.
4.
T. Kraus, Trans. 5th Nat.
Vac.
Symp. (Pergamon Press, New York, 1959), pp. 38-40; Trans. 6th
Nat.
Vac.
Symp. (Pergamon Press, New York, 1960), pp. 204-205.
5.
R. Calder and G. Lewin, Brit. J. Appl. Phys., 18 (1967) 1459-1472.
6. R. Barton and R. Govier, Vacuum, 20 (1970) 1-6.
7.
F. J. Norton, J. Appl. Phys., 11 (1940) 262; J. Appl. Phys., 28 (1957) 34.
8. R. M. Barrer, Diffusion in and Through Solids (Cambridge University Press, London, 1941).
9. M. Abramowitz and I. A. Stegun (eds.). Handbook of Mathematical Functions (Dover, New
York, 1972).
10.
R. Jaeckel and F. J. Schittko, Forschungsberichte des Wirtschafts-und
Verkehrsministerium
Nordrhein-Westfalen Nr. 369. Westdeutscher Verlag, Cologne, 1957.
11.
G. Thieme, Vacuum, 13 (1963) 137.
12.
J. R. Young, J. Appl. Phys., 30 (1959) 1671.
13.
B. B. Dayton, unpublished paper.
14.
H. Galron, Vacuum, 23 (1973) 177.
15.
R A. Redhead, J.
Vac.
Sci. Technol., A13 (1995) 467.
16.
G. Horikoshi, J.
Vac.
Sci. Technol., A5 (1987) 2501.
17.
K. Kanazawa, J.
Vac.
Sci. Technol., A7 (1989) 3361.
18.
D. Hay ward and B. Trapnell, Chemisorption, 2nd ed. (Butterworths, Washington, 1964).
19.
A. Glassford, R. Osiecki, and C. Liu, J.
Vac.
Sci. Technol., A2 (1984) 1370.
20.
N. Yoshimura, H. Hirano, T. Sato, I. Ando, and S. Adachi, J.
Vac.
Sci. Technol., A9 (1991) 2326.
21.
H. Dylla, D. Manos, and P LaMarche, J.
Vac.
Sci. Technol., All (1993)
2623.
22.
M. Li and H.
F.
Dylla, J.
Vac.
Sci. Technol., A13 (1995) 71.
23.
B. Dayton, J.
Vac.
Sci. Technol., A13 (1995) 451.
24.
K. W. Rogers, Trans. 10th Nat.
Vac.
Symp. (Macmillan, New York, 1963), pp. 84-87.
25.
Y. E. Strausser, Proc. 4th Intematl.
Vac.
Congress. (1968) (Inst. Physics and Physical Soc., Lon-
don, 1969), p. 469.
26.
J. Amoignon and J. R Couillaud, Le Vide, No. 141 (May-June 1969), pp. 181-189.
27.
A. Berman, Vacuum, 47 (1996) 327.
28.
R. J. Elsey,
Vacuum,
25 (1975) 347.
29.
AVS Standards Committee, J.
Vac.
Sci. Technol., 2 (1964) 314.
30.
S. Komiya and Y. Sugiyama, /
Vac.
Sci. Technol., 16 (1979) 689.
CHAPTER
4.8
Aluminum-Based
Vacuum Systems
James L Garner
SMC Corporation
The many advantages of aluminum for vacuum applications have driven industry
to overcome most of its technical disadvantages, making it a practical alternative
to stainless steel in vacuum system design. Since the properties and techniques for
stainless steel are well understood, these are used as a departure point for com-
paring and contrasting with those for aluminum.
The physical properties of aluminum differ from those of stainless steel in
many ways. Usually, one is of such overriding importance that it drives the selec-
tion of aluminum. Sufficient information is provided here to make such engineer-
ing judgments possible.
Except where noted otherwise, the equation variables in this section can be in
any consistent unit system. Sample values are given in convenient engineering
units but, in some cases, must be converted to a consistent unit system when used
with the other variables in the sample calculations.
4.8.1
OUTGASSING
Properly cleaned aluminum surfaces outgas at rates comparable to, or better than,
those of stainless steel. Outgassing in unbaked and baked regimes must be dis-
cussed separately, since the mechanisms differ.
ISBN 0-12-325065-7 Copyright © 1998 by Academic Press
$25.00 All rights of reproduction in any form reserved.
509
510
Chapter 4.8: Aluminum-Based Vacuum Systems
4.8.1.1 Unbaked System Outgassing
In unbaked systems, desorbing moisture dominates the gas load and comes off
clean aluminum at about the same rate as from stainless steel [1]. A comparison
of this performance is shown in Table 1. Outgassing rates decline in approximate
inverse proportion to pumping time and, after 100 minutes, remain within about a
3:1 range for both aluminum and stainless steel on all the various surface finishes
tested. Indeed, comparisons of published data suggest that moisture sorbs and de-
sorbs from virtually all nonporous surfaces at similar rates [2]. While outgassing
is proportional to effective surface area, this area is relatively independent of sur-
face roughness, as defined in conventional engineering terms such as /?a [3]. The
thick, porous scale on aluminum received from a mill can create large effective ar-
eas and produce high outgassing rates. However, conditioning of such surfaces is
relatively easy as described in Section
4.8.3.
Table
I
Outgassing Rates of Stainless Steel and Aluminum with Various Finishes
After Exposure to Air for One Hour
Material/Treatment
Q{t = 100)
torr-L/cm-sec
Stainless Steel:
ElectropoUshed
(Summa Method^ with phosphoric
acid commonly used in vacuum components) 7.82X10"^ 1.11 4.72X10"^
Detergent-Cleaned,
vacuum
remelted
SS304
(3X lower
C
than 304L)
Mill
Finish
(3)um roughness as received) with
Alconox'*^ detergent cleaning
Aluminum:
•^iXf>d^«<Mto ..;
:^'
Mirror Processed 6063 (0.02/xm roughness using
ethanol-cooled, diamond-tip cutting tool)
5.49 X 10"
1.17 2.55X10"
5.27 X 10-'' 1.14 2.75 X 10"^
6.38 X 10"'' 1.12 3.62 X 10"^
Source: Adapted from H. F. Dylla et al., Correlation of outgassing of stainless steel and aluminum
with various surface treatments, J.
Vac.
Sci.
TcchnoL,
All(5) (September-October 1993) 2629.
Note: Outgassing
rate:
Q =
QQI'^
torr-L/cm-sec, where / = time in minutes.
4.8.1 Outgassing
511
4.8.1.2 Baked System Outgassing
Moisture can be rapidly removed
by
baking a chamber. The thermal degassing spec-
trum in Figure I shows that substantial degassing starts at about 80°C and is very
high at 150°C [4]. It is believed that the moisture desorption rate of aluminum in
this temperature range is higher than that from stainless steel at even higher tem-
peratures and adequate for practical systems. Baking temperatures above 150°C
should be avoided, because aluminum alloy strength falls and its tempered prop-
erties begin to diminish above
170°C.
Of course, lower strength and annealed
prop-
erties can be accounted for in the design if high temperature operation is necessary.
In fully baked-out systems, adsorbed moisture is virtually eliminated, leaving
hydrogen as the dominant gas load. The principal mechanism for the release of
hydrogen is believed to be permeation from pockets inside the bulk metal and
from the exterior surface. At room temperature, hydrogen permeation rate through
aluminum is approximately seven orders of magnitude lower than through stain-
less steel [5]. Therefore, aluminum normally has a much lower UHV outgassing
rate than stainless steel, typically on the order of
10 "^"^
to
10 "^"^
torr-L/sec-cm^
[6-12].
These rates have been reported approachable with stainless steel only
using elaborate vacuum melting and high-temperature baking processes. The sim-
Fig.l.
Thermal desorption spectra of moisture on EX processed aluminum 6063.
500 400 500 600 700 800 CK)
20
100 200 300 400
Temperature (°C)
500
(Courtesy of M. Mohri et al., "Surface study of Type 6063 aluminum alloys," Vacuum, 34(6)
(1984)644.