The Electrostatic Cryogenic Storage Ring CSR - mechanical concept and realization

A new and technologically challenging project, the electrostatic Cryogenic Storage Ring CSR, is presently under construction at the Max-Planck-Institute for Nuclear Physics in Heidelberg. Applying liquid helium cooling, the CSR with 35 m circumference will provide a low temperature environment of only a few Kelvin and an extremely high vacuum of less than 10^-13 mbar. To realize those conditions the mechanical design has been completed and is now in the construction phase of the first quarter section. For the onion skin structure of the cryogenic system we have at the outer shell the cryostat chambers, realized by welded rectangular stainless steel frames with aluminum plates. The next two shells are fabricated as aluminum shields kept at 80 K and 40 K. The inner vacuum chambers for the experimental vacuum consist of stainless steel chambers cladded with outside copper sheets, connected to the LHe lines for optimized thermal equilibration and cryopumping. Additional large surface 2 K units are installed for cryogenic pumping of H₂. The mechanical concepts and the realization will be presented in detail.

Introduction

We report on the status of the CSR (see Fig. 1), a low energy (20 keV up to 300 keV per charge unit) electrostatic cryogenic storage ring for atomic, molecular and cluster ion experiments, which is in its construction phase. In distinction to magnetic storage rings electrostatic bending and focusing allows the storage of a large range of masses from protons up to massive clusters or bio molecules (masses of a few 10000 a.m.u.). However, it complicates the design, construction and operation by the necessity of operating all ion optical elements inside the cryogenically cooled vacuum chambers.

Figure 1: Schematic of the CSR with the connection to the cryocooler and with
in-ring experiments. Only the inner cryogenic chambers and outer cryostat
chambers are shown, all thermal shields are omitted for better visibility.
- click to enlarge -

The use of superfluid helium for the cooling is motivated by the need for achieving cryocondensation of hydrogen residual gases in order to fulfill the extreme high vacuum requirements encountered at the low ion energies. The technological challenges, connected with this cryogenic cooling and the vacuum requirements were met in the successful realization of a functional prototype. [1].

The CSR ring mainly consists of an inner cryogenic vacuum chamber (diameter from 100 mm to 320 mm), which is kept at temperatures between 2 and 10 K, surrounded by two radiation shields at 40 K and 80 K and an outer cover of 30 layers of superinsulation. An outer cryostat enclosure provides the insulation vacuum of 10^-6 mbar. It is shaped in four straight sections connected by four 90° rectangular corners. The straight sections will house an electron cooling device, the ion beam diagnostics and two experimental sections, respectively. The 90° bend of one corner consists of 9 individual cryogenically cooled chambers housing two quadrupole doublets for focusing, two 6° deflectors for injection and two 39° deflectors for high mass separation efficiency in the three detection units.

During experiments ions should not be exposed to surfaces with temperatures above 10 K to reach and maintain a low degree of internal excitation for molecular ions by radiative cooling. Long time constants of up to 1000 s for these processes and the envisaged long-time storage of multiply charged ions require residual gas densities of less than 2000 particles per cm³ (corresponding to a vacuum of better than 10^-13 mbar at room temperature). This is achievable only with 2 K pumping units to cryocondensate hydrogen as the only remaining component at extreme high vacuum. Therefore a powerful helium refrigerator system based on a Claude closed cycle process was chosen to meet the requirements. The helium lines from the helium refrigerator supply liquid helium which is fed to the inner vacuum chambers to cool the pumping units to 2 K and the inner vacuum chambers below 10 K. The returning gaseous helium is then recirculated to the radiation shields. The refrigerator system (delivering 20 W at 2 K with 600 W cooling power for shield temperatures between 40 and 80 K) is designed for stationary operation at any temperature between 2 K and room temperature. For operation at room temperature the chambers have to be baked up to 600 K in order to achieve a 10^-11 mbar vacuum. The cryogenic and vacuum concepts are described in detail in [2,3]. In order to test the cooling concepts of the CSR a 4 m long electrostatic ion beam trap with 10 K temperature of the vacuum enclosure was successfully designed and operated as a prototype, where a residual gas pressure in the upper 10^-14 mbar range was achieved [1]. The ion beam trap is presently being applied as a new physics instrument to investigate molecular and cluster dynamics at cryogenic temperatures. This report now will focus on the mechanical realization of the CSR.

^ to the top

The mechanical concept

For the prototype [1] the ion optical elements were rigidly mounted to the inner vacuum chambers and thus cooled to cryogenic temperatures. The chambers themselves are connected by CF-sealing flanges. The cryogenic chambers were suspended from the cryostat chamber by spring loaded crossed wires with low thermal conductance. This mounting technique caused difficulties with alignment and stability. The wire suspension technology was therefore changed to a more stable stem support for all chambers and shields. Within this scheme, the critical components of the CSR (mainly the electrostatic electrodes) will be mechanically decoupled from the vacuum chambers to reduce the effects of their thermal shrinkage. All beam focusing and bending elements of the cryogenic ring are supported directly by external concrete support blocks. Conversely, shrinkage of the cold vacuum chambers can be admitted; the outer cryostat and inner cryogenic chambers are almost completely decoupled mechanically from the critical beam optics elements (see Figure 2). For each bending or focusing unit only one electrode support is applied. The lower part of the 40 K shield is formed by a stiff base plate of 15 mm thickness, which will be mounted to the ground plates of the outer cryostat on low-thermal-conductance supports. This shield then supports the inner cryogenic chambers with the same support technology.

Figure 2: Cross section of the cryostat showing the quadrupole inside the
cryogenic chamber, 40 and 80 K shields as well as the supports for the
chambers and the stems for the electrodes. - click to enlarge

^ to the top

Outer cryostat chambers

The CSR will consist of 16 cryostat chambers (see Figure 1) with a cross section of roughly 1 m by 1 m, three per corner and one for each straight section. They have to fulfill several requirements: an insulation vacuum of 10^-6 mbar must be achieved to ensure thermal insulation of the cryogenic chambers, however, outside access for assembly purposes should be as easy as possible. A rectangular shape (see Figure 3) was determined as optimum for high stability and high flexibility. O-ring seals were decided to be sufficient for these chambers. A welded stainless steel frame construction with rectangular aluminum flanges was found to be the most suitable and most economic solution for these cryostat chambers.

Figure 3: View of the almost completely assembled cryostat chambers.
- click to enlarge -

The side plates have a thickness of 30 mm, the bottom and top plates of 80 mm. Each plate stabilizes the chamber by protruding a few mm into the frame. Moving the frame only without plates is therefore not possible. A limited buckling in less than 1 mm of the side plates due to vacuum forces is tolerable, but almost no movement (less than 0.5 mm) of the frame is allowed during operation. The magnetic permeability has been kept lower than 1.02 for large surfaces and lower than 1.05 for welded seams. Welding was done in such a way that O-ring sealing is still guaranteed. As one critical parameter the outgassing rate was specified to be better than 5·10^-9 mbar ls^-1 cm^-2.
In each quadrant one top flange will be used as a burst plate to relieve overpressure in case of a helium accident. The flanges are not fixed by screws but only placed on the chambers to open at overpressures below 0.5 bar.

^ to the top

Shields

The main supporting element of the thermal shielding in the 90° deflection system is a 15 mm thick aluminum base plate, which forms the lower part of the 40 K shield. This stable plate is carried by a specially designed support structure made of corrugated titanium sheet material mounted inside the outer cryostat. The stable 40 K base plate supports the innermost cold vacuum chambers via similar Ti-sheet supports. The side walls and top plate of the 40 K shield are made of 2 mm aluminum and are fixed to the ground plate by screws. A rectangular 80 K enclosure is suspended from the 40 K base plate by inconel wires. 30 layers of superinsulation around the 80 K shield are used to reduce the heat load from outside.
For baking of the inner vacuum chambers, which is necessary for room temperature operation, the 80 K shield is equipped with copper tubes for water cooling in order to prevent a destruction of the superinsulation by temperatures above 100°C.

^ to the top

Helium tubing

The supply lines for liquid helium to the cryogenic chambers and to the cryocondensation surfaces are routed inside the 40 K shielding enclosure. According to the experiences from the prototype, the vacuum chambers and the pumping heads are supplied separately. While the walls of the vacuum chambers may be kept at up to 10 K to fulfill the experimental requirements, a surface temperature of 2 K is necessary for efficient cryocondensation of hydrogen, which is mandatory to achieve the target pressure of less than 10^-13 mbar.
Therefore special cryopumping units were developed, consisting of copper blocks cooled by superfluid helium which protrude into the experimental vacuum with a large surface. The copper structures (see figure 4) will be manufactured by brazing copper to stainless steel. Five of these cold pump units will be mounted on flanges in the stainless steel vacuum chambers in each 90° corner.

Figure 4: Pumping unit with 200 cm² active 2 K pumping surface inside the cryogenic vacuum chamber and the connection to the helium distribution line. - click to enlarge

Outside the cryogenic vacuum chamber, parts of the copper blocks cooled by superfluid helium are used as thermal anchors to fasten copper strips of high thermal conductivity for cooling of the inner vacuum chambers. Roughly 0.5 W may be dissipated per block without influencing the pumping efficiency.

^ to the top

Inner Cryogenic Vacuum Chambers

One quadrant section consists of 9 cryogenic vacuum chambers (see Figure 5), coupled either directly or via bellows.

Figure 5: All chambers of one quadrant. Both sides from the middle detector
chamber are nearly symmetric - click to enlarge

Using Helicoflex seals enables us to use a softer but also less expansive stainless steel (DIN 1.4435 ESU/ANSI 316L) for all chambers instead of the harder DIN 1.4429, which would be needed for CF flanges. Lower compression of the flanges results in fewer and less demanding quality of screws. A low magnetic permeability of 1.01 for all sheet material and 1.05 for all welding seams was required. For welding, only completely filled seams are allowed and specified filler material has to be used for the cryogenic requirements. Using filler material also helps to fulfill the requirements for low magnetic permeability. From the contractor it was required during the chamber manufacturing, to produce conform to the pressure-vessel directives according to German standard AD2000. Though the chambers will not be used as pressure-vessels the directive enforces high welding quality and a reasonable traceability by restamping of materials before cutting. To achieve low mechanical tolerances the milling and drilling of each chamber was carried out on 5-axes CNC machines in one step without re-orienting the part. The flanges are therefore first welded with excess dimensions and then machined to final tolerances as the last production step.
For the focusing and deflector units, electric voltages of up to 10 or 30 kV respectively will be needed. Special high voltage feedthroughs from nonmagnetic materials are being developed. The electrical connection from the outer feedthrough to the cryogenic chamber will be made from stainless steel wire to minimize the heat load from outside. Inside the cryogenic vacuum chamber the high-voltage wires will be made from 1.5 mm diameter copper in order to use them additionally for heat transfer, because the ceramic supports of the electrodes are unsuitable in that respect. Specially designed sapphire blocks are used as electrically insulating thermal anchor points on the 40 K shield and again inside the cryogenic vacuum chamber to improve the cooling of the electrodes.
Though a temperature of 10 K is sufficient for the inner vacuum chamber, the thermal conductivity of the stainless steel walls is not sufficient to provide a uniform temperature distribution on the surface. As galvanic copper plating was too complex, the outside of the chambers is wrapped (see Figure 6) with copper foil (0.25 mm with 99.95% purity). High conductivity copper strips to the 2 K pump units with 99.995% purity are used to enforce a uniform cooldown with high efficiency.

Figure 6: The middle detector chamber wrapped in copper foil.
- click to enlarge -

^ to the top

Summary

The main effort after the almost completed assembly of the outer cryostat is directed to the completion of one quadrant in order to test the new suspension system under cryogenic conditions and in the final working environment. All cryogenic vacuum chambers for this first quadrant have been delivered and are being wrapped with copper foil. The base plates of the shields are also delivered and the copper lines for water cooling are fixed on the 80 K base plate. Following the mounting of these shield parts, which is then the installation and adjustment of the critical electrodes will be preformed. Then helium lines will be mounted in parallel to the progressive closing of the radiation shields. A test of the cooling speed and the temperature distribution of the vacuum chambers will be the next step after the measurements of the accuracy of the adjustment.

^ to the top

Up-to-date pictures

Some up-to-date pictures to give an impression of the current status (please click to enlarge):

Acknowledgement

It is a pleasure to thank our technicians for their outstanding support as well as the Max-Planck-Society for their continuous financing support.

References