Note: Descriptions are shown in the official language in which they were submitted.
~L22~9~
--1--
CRYOPUMP WITH RAPID COOIDOWN AND INCREASED
.
P~ESSURE STAB~LIrrY
Description
Technical Field
This invention relates to cryopumps and has
particular application to cryopumps cooled by two
stage closed cycle coolers.
Backgrou_
Cryopumps currently available, whether cooled
by open or closed cryogenic cycles, generally
follo~ the same design concept. A low temperature
second stage array, usually operating ln the range
of 4 to 25 K, is the primary pumping surface.
This surface is surrounded by a higher temperature
cylinder, usually operated in the temperature
range of 70 to 130 K, which provides radiation
shielding to the lower temperature array. The
radiation shield generally comprises a housing
which is closed except at a frontal array posi-
~ioned between the primary pumping surface and the
chamber to be evacuated. This higher temperature,
first stage, ~rontal array serves as a pumping
site for higher boiling point gases such as water
vapor.
In operation, high boiling point gases such
as water vapor are condensed on the frontal array.
Lower boiling point gases pass through that array
and into the volume within the radiation shield
and condense on the second stage array~ A surface
coated with an adsorbent such as charcoal or a
molecular sieve operating at or below the tempera-
ture of the second stage array may also be pro-
vided in this volume to remove the very low
boiling point gases. With the gases thus con-
densed and or adsorbed onto the pumping surfaces,only a vacuum remains in the work chamber.
In systems cooled by closed cycle coolers,
the cooler is typically a two stage refrigerator
having a cold finger which e~tends through the
rear of the radiation shield. The cold end of the
second, coldest stage of the cryocooler is at the
tip of the cold finger. The primary pumping
surface, or cryopanel, is connected to a heat sink
at the coldest end of the second stage of the
coldfinger. This cryopanel may be a simple metal
plate or an array of metal baffles arranged around
and connected to the second stage heat sink. This
second stage cryopanel also supports the low
temperature adsorbent.
The radiation shield is connected to a heat
sink, or heat station at the coldest end of the
first stage of the refrigerator. The shield
surrounas the first stage cryopanel in such a way
as to protect it from radiant heat. ~he frontal
array is cooled ~y the first stage heat sink
through the side shield or, as disclosed in U.S.
Patent 4,356,701, through thermal struts.
One problem that has been experienced by
certain users of cryopump systems is known as
cross over "hang up". This problem is of partic-
ular concern in systems such as sputtering systems
where the process is carried out in an argon,
oxygen or nitrogen environment. Cross over is the
processing step in which a valve between the work
chamber and cryopump is opened to expose the very
high vacuum cryopump to a lower vacuum work
chamber. The pressure of the work chamber is then
reduced by the cryopump. To bring the work
chamber pressure to a vacuum of, for example, 10 7
torr, it is necessary that, in the case of argon,
the gas be condensed on the cold, second stage
array at a temperature of 2~.6 R. Condensation of
argon at higher temperatures results in a higher
partial pressure of the argon and thus a higher
pressure in the work chamber.
During normal operation of the system in
which the first stage array is held at a tempera-
ture of, for example, 77 K, the argon does notcondense on the first stage array but passes
directly to the second stage array for proper
condensation on that array. However, under low
thermal load conditions the frontal array tempera-
ture can drop to as low as about 40 K. At that
temperature argon does condense on the frontal
array; and at that temperature the partial pres-
sure resulting from the balanced evaporation of
solid argon and condensation of argon molecules
25 results in a partial pressure of only 10 3 to 10 4
torr. So long as any argon is in this state of
sublimation on the frontal array, the pressure in
the work chamber cannot be taken down to the
desired 10 7 torr.
As the argon gas evaporates during sublima-
tion, it eventually migrates to the colder second
stage and is captured by that stage. However, the
sublimation process is a slow one and until
2~B
complete the pressure in the system "hangs ~lp" at
the higher pressure.
As a possible solution to "hang up", it has
been suggested that the first stage arrays be made
warmer by introducing an electrical heat load onto
the first stage to prevent excessive cooling of
that stage. However, a load on the stage genex-
ally increases cooldown time of the refrigerator.
Minimizing cooldown time is a significant concern
in designing cryopump systems. Further, electri-
cal elements can present a hazard where the
concentration of hydrogen is high.
Another problem associated with cryopump
systems is that a pulsed thermal load can result
in erratic pressure in the work chamber. For
example, as a low emissivity valve door is opened
to expose the frontal array to a higher emissivity
radiating surface, the thermal load is increased,
and the pressure may become unstable.
Disclosure of the Invention
In accordance with the principles of this
invention, cross over hang up in a cryopump is
avoided by providing a passive heat load to the
first stage to assure that the first stage is held
at a temperature above about 50 K. During initial
stages of cooldown, the passive heat load is
substantially less than that at the final cooldown
temperature condition, so that cooldown time is
not substantially affected.
Preferably the heat load is due to radiant
heating of a radiation shield. To increase the
radiation heat load to the first stage, the
!
~ 2 ~ ~ ~8
effective emissivity between at least a po.rtion of
the radiation shield and the vacuum vessel is
increased. At low temperatures of the first
stage, the radiation heat load on the first stage
5 is great due to the fact that the heat flux is a
function of the difference in temperatures to the
fourth power. As a result, when the first stage
drops to a temperature near 50 K the heat load is
substantial and prevents the first stage from
dropping to a temperature below 50 K. It has been
found that, so long as the temperature is held
above 50 K, cross over hang up is avoided. At
higher temperatures, the temperature differential
between the radiation shield and the vacuum vessel
is less and, due to the fact that the radiation
heat flux is a function of the difference in
- temperatures, the l.oad is substantially less.
When the system is initially at ambient tempera-
ture, the heat load is negligible. Thus, by
providing a radiation heat load to the first
stage, that heat load is minimized at cooldown
temperatures but is significant enough at very low
temperatures to prevent the first stage from
dropping to a temperature below 50 K. Cooldown
time is not significantly hampered and cross over
hang up is avoided.
Preferably, the effective emissivity between
the radiation shield and vacuum vessel is obtained
by painting the outer surface of the radiation
shield black. Painting of the inner surface of
the vacuum vessel would also increase the effec-
tive emissivity, but might result in outgasing
~z~
--6--
from the paint at the higher temperatures of the
vacuum vessel.
A problem related to cross over "hang up" can
occur as a result of condensation of gases on the
side of the second stage refrigerator cylinder.
This problem is particularly apparent where an
open second stage array is used to provide for
maximum flow to an adsorbent material on the back
side oE the array. At normal operating tempera-
tures, there is a temperature gradient along thelength of the refrigerator cylinder from the
approximately 77 K first stage heat sink to the 15
K second stage heat sink. Argon and other gases
can condense along a zone of the refrigerator
cylinder which is at a temperature of less than
50K. The temperature of that zone is determined
by the system pressure. When a thermal load is
applied to the first stage, as by opening a valve
in the system, the first stage temperature in-
creases and shifts the 50 K zone along the lengthof the refrigerator cylinder. As that zone
shifts, gas which had been frozen out on the
cylinder is rapidly liberated. Tha' rapid evapo-
ration results in a sharp increase in the work
chamber pressure. Further, even when the thermal
load on the first stage is constant, a displacer
within the refrigerator cylinder reciprocates and
causes continuous movement of the critical zone.
That movement of the critical zone results in a
high frequency fluctuation of the pressure in the
work cha~ber.
To avoid the problems caused by condensation
of argon and other gases on the second stage
!
--7--
refrlgerator, a close fit-ting sleeve surrounds the
refrigerator cylinder. That sleeve is in thermal
contact with the second stage heat sink but is not
in contact with the refrigerator cylinder. Most
gas which passes the second stage array is con-
densed on the shield before it reaches the cylin-
der. The narrow gap of about .1 inch or less
between the shield and the cylinder assures that
even gas which passes beneath the cylinder is
quickly condensed on and thus captured by the cold
shield. With the shield held at the low tempera-
ture of the second stage heat sink, gas which
condenses on the shield is held there and does not
subsequently evaporate with displacer motion or
high heat load to the first stage.
Brief Description of the Drawings
The foregoing and other objects, features and
advantages of the invention will be apparent from
the following more particular description of
preferred embodiments of the invention, as illus-
trated in the accompanying drawings in which like
reference characters refer to the same parts
throughout the different views. The drawings are
not necessarily to scale, emphasis instead being
placed upon illustrating the principles of the
invention.
Fig. 1 is a perspective view of a cryopump
embodying this invention;
Fig. 2 is an elevational cross sectional view
of the cryopump of Fig. l;
Fig. 3 is an illustration of an alternative
thermal switch embodiment.
"~
~2~1~3~L~
--8--
Preferred Embodiments of the Invention
..
The cryopump of Figs. 1 and 2 comprises a
vacuum vessel 12 which is mounted to the wall of a
work chamber along a flange 14. A front opening
16 in the vessel 12 communicates with a circular
opening in a work chamber. For shipment, a
removable cover 17 is provided over the opening as
shown in Fig. 2. Alternatively, the cryopump
assembly may protrude into the chamber and a
vacuum seal be made at a rear flange. A two stage
cold finger 18 of a refrigerator protrudes into
the vessel 12 through an opening 20. In this
case, the refrigerator is a Gifford-MacMahon
refrigerator such as disclosed in U.S. Patent No.
3,218,815 to Chellis et al., but others may be
used. A two stage displacer in the cold finger 18
is driven by a motor 22. With each cycle, helium
gas introduced into the cold finger under pressure
through line 24 is expanded and thus cooled and
then exhausted through line 26. A first stage
heat sink, or heat station, 28 is mounted at the
cold end of the first stage 29 of the refrigera-
tor. Similarly, a heat sink 30 is mounted to the
cold ~nd of the second stage 32. A suitable
temperature sensor element 34 is mounted to the
rear of the heat sink 30.
The primary pumping surface is an array
mounted to the heat sink 30. This array comprises
a disc 38 and a set of circular chevrons 40
arranged in a vertical array and mounted to disc
38 by thermal struts 41. The struts 41 extend
through the chevrons 40 and cylindrical spacers 43
bet~een the chevrons, and nuts at the ends of the
``~
~z~
_9_
struts compress the chevrons and spacers into a
tight stack. A low temperature adsorbent such as
charcoal particles is adhered to the lower,
backside surface area of the chevrons. Access to
this adsorbent by low boiling point gases is
through the open chevrons 40. This open arrange-
ment with the chevrons supported by struts, allows
for simple assembly and also ready flow of gases
past the front side of the chevrons 40 to the
adsorbent. As an alternative, the chevrons could
be supported on an inner cylinder to which ad-
sorbent could adhere.
For reasons to be discussed below, a sleeve
52 is positioned over the second stage refrigera
tor cylinder 32. The sleeve 52 is formed of two
hemicylindrical elements 54 and 56 which are
mounted to and extend downward from the second
stage heat sink 30. A small gap 55 is provided
between the sleeve and the cylinder 32.
A cup shaped radiation shield 44 is mounted
to the first stage, high temperature heat sink 28.
The second stage of the cold finger extends
through an opening 45 in that radiation shield.
This radiation shield 44 surrounds the second
stage array to the rear and sides to minimize
heating of the array by radiation. Preferably the
temperature of this radiation shield is less than
about 120 K.
A frontal cryopanel array 46 serves as both a
radiation shield for the primary cryopanel and as
a cryopumping surface for higher boiling tempera-
ture gases such as water vapor. This array
comprises louvers 4~ joined by rim 50. The
`~
9~3
-ln-
frontal array 46 is mounted to the radiation
shield 44, and the shield both supports the
frontal array and serves as the thermal path from
the heat sink 28 to that array. The configuration
of this array need not be confined to the arrange-
ment shown but it should be an array of baffles so
arranged as to act as a radiant heat shield and a
higher temperature cryopumping panel while provid-
ing a path for lower boiling temperature gases to
the second stage array.
As noted above, the problem of cross over
hang up results from argon and other gases freez-
ing on the first stage frontal array rather than
passing directly through to the second stage
array. Experiments have shown that hang up due to
argon can be avoided by holding the temperature of
the frontal array above 50 degrees. This in turn
can be accomplished by providing a heat load to
the first stage at low temperatures. On the other
hand, it is preferred that the heat load of the
first stage be minimi~ed at higher temperatures in
order to maintain high cooldown speeds. To that
end, a radiation heat load is applied to the first
stage by painting the outside of the radiation
shield 44 with flat black paint. This increases
the emissivity of the shield and increases the
radiant heat flow from the vacuum vessel to the
shield. That radiant heat flow is a thermal load
on the first stage refrigerator.
The thermal load on the first stage is due to
the radiant heat flow Q to the radiation shield
~4:
Q = A ~eeff (TH ~ TL ) (1
9~3
where A is the surface area, u is a constant, eefE
is the eEfective emissivity, TH is the temperature
of the vacuum vessel and Tl is the temperature of
the radiation shield.
The effective emissivity is a function of the
emissivity eO of the outer surface of the radi-
ation shield and the emissivity ei of the inner
surface of the vacuum vessel:
e 1 (2)
eff 1/eO + 1/ei-1
In the past, these surfaces have been polished to
obtain very low emissivities of less than about .1
for an effective emissivity of less than about
.05. That low effective emissivity minimizes
radiant heat flow and the resultant load on the
first stage. To provide a proper heat load to the
first stage in accordance with this invention the
effective emissivity should be at least about .10.
This effective emissivity is obtained by an
emissivity of the outer surface of the radiation
2~ shield 44 approaching one and the emissivity of
the inner surface of the vacuum vessel 12 of about
. 1 . .
It is significant that the high emissivity is
provided on the radiation shield 44 and not on the
frontal array 46. With a high emissivity on the
array 46, the effective emissivity could vary
greatly. As a valve door to the work chamber
opens, the emissivity seen by the array would
change from .l to near one. With an emissivity on
the array of near one, the effective emissivity
would change from about .1 to about one. This
~ `~
34L8
~12-
would result in a change in thermal load of
several watts.
With the present arrangement the frontal
array has an emissivity of about .1 so that as the
valve opens the frontal effective emissivity only
changes from about .05 to about .1. The effective
emissivity between the radiation shield and vacuum
vessel remains at about .1 regardless of the valve
position. Thus, the first stage load remains much
more constant at about one or two watts.
It can be noted that the radiation heat flow
is a function of the difference in temperatures
raised to the fourth power. Thus, as the tempera-
ture differential increases, the heat flow in-
creases. It has been found that by painting theradiation shield 44 black, which provides a shield
emissivity of about .9, a significant heat load on
the first stage due to radiant heat flow is
obtained at low temperatures of the first stage.
That heat load is sufficient to keep the tempera-
ture of the first stage, including the frontal
array 46, above 50 K. However, at higher tempera-
tures the radiant heat load is much less signifi-
cant and thus does not appreciably hamper cooldown
of ~he system.
Another means for obtaining the desired load
at only lower temperatures is illustrated in Fig.
3. In this arrangement, a thermal switch provides
a conductive heat flow path between the vacuum
vessel 12 and the radiation shield 44 at low
temperatures. The switch is formed of bimetallic
e]ements 56 and 58. At low temperatures approach-
ing 50 K, these bimetallic elements come into
t. .,~j \
94~3
-13-
contact and provide a heat flow path to the
radiation shield ~4 to prevent the temperature oE
the Erontal array from dropping below 50 K. At
higher temperatures, however, the elements are
separated and the vacuum between the elements 56
and 58 provides good insulation.
A radiation heat load is preferred over the
conductive heat load because it provides more
uniform loading of the second stage and because it
does not result in any structural changes to the
system. Both radiation and conductive heat loads
avoid the need for an electrical heating element
in the system, and both provide greater thermal
loading as the first stage temperature decreases.
The heat load provided by the increased
radiation to the radiation shield 44 prevents the
condensation of argon and other low condensing
temperature gases on the frontal array, but it was
found that a problem still existed with the
condensation of argon on the second stage refrig-
erator cylinder 32. Even at normal operating
temperatures with the first stage heat sink 28 at
77 K and the second stage heat 30 at 15 K, a
temperature gradient exists between those heat
sinks along the length of the cylinder 32. The
pressure of the chamber, for example 10 4 torr,
determines a limited temperature range less than
50 K at which argon gas condenses and evaporates
in equilibrium. Thus at all times, at some point
along the length of the cylinder, there exists a
critical zone on the cylinder 32 at a temperature
at which argon gas condenses and evaporates in
equilibrium. As the displacer within the cylinder
~L2~
32 reciprocates up and down, that critical zone
moves up and down along the cylinder. As the zone
moves up, the region which had supported condensed
argon warms to a higher temperature at which the
argon evaporates. The fairly rapid evaporation of
the argon results in a rise in the pressure of the
system. As the displacer reciprocates, this
oscillating movement of the critical region can be
seen as an oscillation in the chamber pressure.
Another result of the argon condensation on
the cylinder 32 is pressure instability with
changes in the thermal load on the first stage.
For example, when a valve is opened to the work
chamber, the first stage is subjected to a large
thermal load which increases the tempexature of
the first stage heat sink 28. This in turn causes
a rapid shift in the critical zone and unstable
pressure in the chamber.
It has been found that the condensation of
argon on the cylinder can be virtually eliminated
by positioning a close fitting shield 52 over the
cylinder and maintaining that shield at a stable,
low temperature. Most of the gas which passes
through the second stage array and which would
otherwise come into contact with the second stage
cylinder 32 is intercepted by the shield. Fur-
ther, any gases which are able to pass from below
the shield into the region between the shield and
the cylinder are soon captured on the inner
surface of the shield. Once argon is condensed on
the 15 K shield, evaporation is very limited. On
the other hand, any gas which should condense on
the cylinder does evaporate at a relatively faster
-15-
rate. On balance, -then, gas which enters the gap
between the shield and the cylinder is quickly
captured by the cylinder and condensation on the
cylinder is virtually eliminated. A gap of .085
inch has been found suitable for this purpose.
While the invention has been particularly
shown and described with reference to a preferred
embodiment thereof, it will be understood by those
skilled in the art that various changes in form
and details may be made therein without departing
from the spirit and scope of the appended claims.
For example, a closed cycle, two stage refrigera-
tor is shown. A cryopump cooled by an open cycle
refrigerator such a liquid nitrogen, hydrogen or
lS helium may also be used. Also combinations of
single and two stage closed cycle refrigerators
may be used to provide the cooling.
