Design of the Helium-Cooled Lithium-Lead Breeding Blanket ...
Design and Construction of a Helium 4 Dippermsingh/assets/design-construction-helium.… · FIG. 6....
Transcript of Design and Construction of a Helium 4 Dippermsingh/assets/design-construction-helium.… · FIG. 6....
Design and Construction of a Helium 4 Dipper
Kirsten Blagg
Physics Department Colorado School of Mines, Golden CO
(Dated: December 11, 2017)
2
CONTENTS
Introduction 3
Material Properties 4
Specific Heat 4
Thermal Conductivity 7
Thermal Expansion 10
Magnetic Susceptibility 11
Material Properties for Common Cryogenic Materials 12
Design 16
Top Electronics Box 16
Main Body 21
Sample Stage 23
Wiring 28
Conclusion 29
Appendix A: Design Drawings 30
Appendix B: Purchase List 38
Acknowledgments 40
References 40
3
INTRODUCTION
Low temperature physics offers insight into a variety of interesting phenomena in materials
research, particle physics, solid state physics, and much more. There are multiple methods
to reach very low temperatures. Commercially available helium 4 cryostats can be used
to reach approximately 1 Kelvin and helium 3 - helium 4 dilution fridges can reach the
milikelvin range. These commercially available options (Oxford, Blue Forge) are simple
and effective; however, they tend to be extremely expensive and difficult to modify[1][2].
Additionally, larger dilution cryostats can take six hours or more to cycle a sample from room
temperature down then back to room temperature again. Many interesting applications
mentioned above can be accessed at the temperature of helium evaporation, 4.2K. A simple
cryostat, commonly called a dipper, can be inserted directly into a standard helium 4 dewar
allowing experiments to be run at 4.2 Kelvin. In addition to acting as a stand alone cryostat,
a helium 4 dipper can preform preliminary tests quickly and effectively compared with colder
dilution fridges. Standard helium dippers place the sample directly in liquid helium and use
evaporation to directly cool the sample. While, this offers a simple design, it does not allow
for temperature variation or for thermoelectric measurements. To this end, the sample must
be placed in vacuum. While this complicates the dipper design, it offers more flexibility and
experimental possibilities compared with a typical helium dipper.
The primary concerns in the design and use of a helium 4 dipper are cooling a sample
down to low temperature and conserving expensive helium 4. In general, the sample needs to
reach a stable thermal equilibrium with the surrounding liquid helium, measurements need
to be taken on the cold sample, and heat sources need to be minimized. For any cryostat,
the main sources of heat are: the conduction of heat along walls, wires, and tubes, thermal
radiation from room temperature to cold components, conduction by gas particles, thermal
acoustic oscillations, Joule heating, and additional electric and mechanical vibrations. Each
heat source must be offset by the cooling power of the cryoliquid. While heat transfer cannot
be completely avoided, careful attention to materials and construction is critical to ensuring
efficient use of helium.
The following paper describes basic design schematics, design challenges, material con-
siderations, and guidelines important in the design process of any fridge. It also includes a
detailed step by step guide to the construction of a vacuum 4.2K dipper.
4
MATERIAL PROPERTIES
As with any design and assembly, particular attention needs to be paid the properties
of the material used. In the construction and use of a helium fridge the most important of
these properties are specific heat, thermal conductivity, and thermal expansion. Magnetic
susceptibility also becomes important for experiments involving the production of magnetic
fields and magnetic measurements. The workability, cost, and strength of the materials also
impacts design choices. The cryogenic application of the dipper causes further challenges in
material selection because materials must operate over a wide temperature range. There-
fore, it is critical to consider the temperature dependence of material properties. In the
following sections the theoretical basis and and specific material considerations of specific
heat, thermal conductivity, thermal expansion, and magnetic susceptibility are addressed
for cryogenic applications.
Specific Heat
The specific heat, the amount of heat required to increase the temperature of the material,
is primarily controlled by the excitation of electrons and phonon vibrations. The specific
heat determines how much energy it takes for the lower part of the dipper to cool from
room temperature to 4.2K. This translates to the amount of helium boiled off each time the
dipper is inserted into the dewar.
In insulating materials lattice vibrations called phonons cause excitations. At high tem-
peratures phonons can be modeled as independent oscillators with three degrees of free-
dom. Electrons which remain bound to the atoms have negligible excitations compared
with phonons at high temperatures. These degrees of freedom (phonon modes) freeze out at
low temperatures below the Debye temperature, θD = hωD
kBwhere h is Plank’s Constant, ωD
is the Debye frequency, and kB is Boltzmann’s Constant. While these three phonon modes
freeze out at low temperatures, the electronic excitations do not. Thus, at very low temper-
atures, far below the Debye temperature, electronic contribution to specific heat becomes
important even for insulators. The specific heat of insulators at low temperatures has a
cubic dependence on temperature (C = βT 3). Thus, the specific heat is generally very low
at low temperatures.
5
In metals, the specific heat is determined by the excitation of phonons and electrons
which are free to conduct through the material. At low temperatures the electron energy
states remain close to the Fermi energy and only electrons near the Fermi energy conduct
heat. The specific heat is given by the cubic dependence of the phonon vibrations summed
with a linear temperature dependence of the electron excitations, C = βT 3 + γT where β
and γ are both constants. At very low temperatures phonon modes freeze out, just as in
the case of insulators, and electron excitations dominate (C ≈ γT )[3]. Figure 1 compares
the temperature dependence of metals and insulators.
FIG. 1. Comparison of the heat capacity of silicon and copper as a function of temperature cubed.
Both materials approach the law of Dulong and Petit at high temperature. At a middle temperature
range both materials follow the Debye model and match a cubic temperature dependence. At low
temperatures the metal departs from the Debye model due to the contribution of electron specific
heat. [4]
At low temperatures, some metals become superconductors and the electrical resistivity
of the material drops to zero. The specific heat of phonons is not changed by the transition
into the superconducting state. However, the electron excitations now have an additional
degree of freedom corresponding to the possibility of forming a Cooper Pair and entering the
superconducting state. This new degree of freedom increases the specific heat at the critical
temperature of the material. Below the critical temperature the temperature relation of the
electron specific heat is given by an exponential decay (Figure 2). Due to the zero resistivity
of superconductors, they are often use in wiring, magnets, and devices in fridges.
6
FIG. 2. The heat capacity of a superconducting metal versus temperature. Above the critical
temperature, the heat capacity is dominated by a linear relation to temperature. Once the critical
temperature is reached, there is a increase in the heat capacity and the heat capacity follows an
exponential relation to temperature[5].
FIG. 3. The specific heat of common cryogenic materials is plotted with respect to temperature.
Metals are plotted with solid lines, metallic alloys with dot-dashed lines, thermal insulators with
dotted lines, and gases with dashed lines [6].
7
Thermal Conductivity
Thermal conductivity (κ) is defined as the heat flow (.q) per unit area under a temperature
gradient (5T ),.q = −κ5T . The thermal conductivity of the material is critical cooling the
sample and conserving Helium. Materials with a high thermal conductivity are required to
connect the sample to the surrounding liquid helium. Materials with low thermal conduc-
tivity are required to reduce heat flow from room temperature down to the liquid helium to
prevent wasteful boil off.
Heat is carried through the material by the conduction of electrons and phonons. This mo-
tion is defined by the limiting of conduction through scattering processes: phonon-phonon,
phonon-defect, electron-impurity, electron-electron, and electron-magnetic impurity. Con-
ductivity is thus strongly dependent on the material and temperature.
Electron scattering is negligible in insulators because of the low electronic conductivity.
Thus, thermal conductivity is defined by phonon-phonon or phonon-defect scattering. At
very low temperatures, far below the Debye temperature, the low number of thermally ex-
cited phonons decreases phonon-phonon scattering and the majority of the scattering takes
place off defects, dislocations, and grain boundaries. In metals, electronic conductivity
dominates; though electron and phonon conductivity is comparable in disordered alloys.
At high temperatures thermally excited phonons limit the conduction of electrons through
phonon-electron scattering. At low temperatures electrons primarily scatter off defects and
impurities since there are very few phonons. Figure 4 shows how the temperature depen-
dence of the thermal conductivities changes as scattering processes become more or less
dominant. Since the conduction of both electrons and phonons at low temperatures is lim-
ited by defects and impurities the thermal conductivity varies from sample to sample and
between manufacturing processes. Therefore, the purity, tempering, fabrication, and history
of a material is important. A comprehensive review of thermal conductivities for elements
has been outlined in literature[7], but variations between samples still occur (Figure 5).
8
FIG. 4. A schematic outline of the dominant scattering processes in insulators and metals versus
temperature. While the temperature and exact characteristic of these shifts in scattering are
material dependent, the schematic reveals the importance of temperature on scattering processes
and on thermal conductivity. The graph shows the thermal conductivity of AlN with respect to
temperature. The trendline changes as the different scattering processes turn off as the temperature
decreases. [8]
9
FIG. 5. The thermal conductivity versus temperature of common cryogenic materials for T > 2K.
We note that there is additional dependence on purity and defects [6].
10
Thermal Expansion
Materials at high temperatures expand and at low temperature contract. This expan-
sion and contraction caused by the aharmonic potential the atom experiences due to the
electrostatic forces of its neighbors (Figure 6). At low temperatures the atoms remain in
the parabolic harmonic region since the amplitude of the atomic vibrations is small. How-
ever, as temperature increases, the thermal vibrations of the atoms grow and experience
an asymmetric potential. This causes the atoms to spread out as they spend increasingly
more time at longer distances leading to the expansion of the lattice spacing. The thermal
expansion coefficient (α) depends on the material and is given by the ratio of the change in
length (l) by the change in temperature (T ) divided by the length at a constant pressure
(P ), α = 1l( ∂l∂T
)P . This characterizes a material’s expansion/contraction under temperature
change.
FIG. 6. A typical atomic potential as a function of distance between the atoms or ions in a lattice
is plotted above. The blue dot represents the average location of an atom at a cold temperature
and the red atom represents the average location of an atom at hot temperatures. The expansion
is seen in the difference of the two averages [3].
When designing a helium dipper, the dipper manufacturing occurs at room temperature,
but the dipper must be heated and cooled for each experiment repeatedly expanding and
11
FIG. 7. The thermal expansion versus temperature of common materials used in cryogenic
manufacturing[6].
contracting the materials. The thermal contraction of common materials is given in Figure
7. For connections between the same materials all components will expand and contract
equally, so thermal expansion of materials is relatively unimportant. In connections between
different materials, the different rates of expansion and contraction can cause connections
to leak or break due to strain at the joints. In general, the material with the largest thermal
expansion coefficient should be on the outside and the material with the smallest coefficient
towards the center. It is important to remember that seals and contacts made at room
temperature will not necessarily hold at low temperatures. In order to ensure contact a low
expansion material such as molybdenum or tungsten can be used as a washer. In this design,
we have minimized connections between dissimilar materials to avoid leaks and additional
stress.
Magnetic Susceptibility
The magnetic susceptibility of a material (χ) defines the magnetization of that material
in response to an external magnetic field. The magnetic susceptibility is a dimensionless
constant given by χ = M/H, where M is the magnetization of the material and H is the
magnetic field strength. Often cryogenic experiments require the use of strong magnetic
12
fields, upwards one Tesla. Additionally, some cryostats employ the use of SQUIDs (Super-
conducting Quantum Interference Devices) for very sensitive magnetic field measurements.
If the materials used, particularly the material surrounding the sample or wiring have high
susceptibilities, they can produce a magnetic field influencing the sample and skewing any
magnetic measurements preformed. Again temperature considerations are important; mate-
rials which are considered non-magnetic at room temperature can have appreciable magnetic
susceptibilities at low temperatures. In addition, for sensitive experiments magnetic impu-
rities in materials can cause interference and noise. If the cryostat will be used to preform
experiments in a magnetic field the magnetic properties, including magnetic defects, of con-
struction materials must be carefully analyzed.
Material Properties for Common Cryogenic Materials
The properties of common manufacturing and cryogenic materials are given in Table 1.
While exotic materials have better properties with respect to cryostat design, the machining
ease, availability, and cost of common materials makes them the leading choice. These more
ideal materials, such as Invar in the case of thermal expansion and Suprasil in the case of
magnetic susceptibility can be used in small applications such as washers, connections, or
stages. However, readily available and easily machinable materials such as stainless steel
and copper are used for the bulk of this design.
TABLE I. The specific heat, thermal expansion and thermal conductivity of common cryogenic
materials. Magnetic susceptibility is not listed due to its extreme variability based on particular
alloy and purity.
Material Specific Heat Thermal Expansion Thermal Conductivity
(J/gK) (δL/L) (300-4K) (W/m)
Copper 7x10−5 370x10−5 200-1000(RRR 20-100)
Aluminum 3x10−4 462x10−5 6
Stainless Steel 0.4-0.5 331x10−5 0.4
Brass 2x10−4 397x10−5 7
13
The materials required for a cryostat fall broadly into two categories: materials which
conduct heat and materials which minimize heat conduction. All materials used were also
chosen due to low magnetic susceptibilities, low specific heat, and reasonably small thermal
expansion coefficients.
Stainless steel provides a durable, machinable, affordable material with a low thermal
conductivity even at low temperatures. However, different alloys, purities, and manufac-
turing can greatly affect the properties of the material. Most 300 series stainless steels are
adequate for cryogenic applications; in particular, 304 and 316 stainless are classified are
cryogenic stainless steels. While these stainless steels are idea, most important is that the
stainless steel chosen is an austenitic stainless steel. When austenitic steels cool, the iron
remains in the form of austenite (face center cubic structure) allowing the material to retain
its strength. All other stainless steels exhibit an impact ductile / brittle transition. Addi-
tionally, non-austenitic steels, due to the crystal phase of the iron, exhibit a high magnetic
susceptibility at cryogenic temperatures making them poor material choices.
304L stainless steal was chosen for the main body of this cryostat design. 304 stainless
steel is part of the 300 series of steels composed of austenic chromium-nickel alloys. 304
is the most common grade with 18% chromium and 8% nickel. As seen in table II 304
stainless steels have a low thermal conductivity over a wide temperature range compared
with other alloys. The low carbon content of L stainless steels (0.03% carbon as opposed to
0.05% carbon) allows for easy welding. Additionally, many cryogenic and vacuum parts such
as flanges are made of 304L stainless steel. Thus, by using 304L stainless steel for dipper
construction, commercially available parts can easily be welded to the dipper.
While stainless steel offers a good material for preventing thermal conduction, copper’s
high thermal conductivity and availability make it an ideal material for providing thermal
contact between the liquid helium and the sample. However, the high cost and difficult
machinability of copper requires a design which minimized copper use. Similar to stain-
less steel, copper’s many different alloys, purities, and fabrication processes can have very
different properties particularly at low temperatures (Table III/Figure 8). OFHC (Oxygen
free high conductivity copper) is widely used in cryogenic applications because of its high
conductivity at low temperatures. C1100 offers an comparable thermal conductivity and
can be used in place of OFHC. For this design, since a small amount of copper was required,
scrap copper was used due to cost and availability over OFHC and C1100 copper.
14
TABLE II. The thermal conductivities of stainless steel alloys at room temperature and at 4.2K
Stainless Steel Alloy Thermal Conductivity (W/m ∗K)
Room Temperature 4K
304 0.3 14.5
316 0.3 14.5
16 Cr 20 Ni 0.4 13.6
15 Cr 26 Ni 1.0 11.2
Armco Iron 13 76
TABLE III. The thermal conductivity, magnetic susceptibility, and strength of different copper
alloys
UNS Number Material Thermal Conductivity (4K) Magnetic Susceptibility (4K)
(W/mK)
C10100, C10200 Oxygen Free, Annealed 183 −2.98x10−6
C11000 Electroytic Tough Pitch, Annealed 325 2.53x10−5
C17000-C17300 Beryllium Copper, Annealed 20 - 0.02 1.82x10−3
C22000 Commercial Bronze Annealed 6.1 7.63x10−6
C36000 Free Cutting Brass 2 −1.41x10−2
Copper and 304L stainless steel made up the majority of the materials used in this design.
Other materials used for wiring, connections, and electronics will be covered in the following
sections.
15
FIG. 8. The temperature dependence of the thermal conductivity of different copper alloys.
16
DESIGN
The helium 4 dipper is a multi-part assembly consisting of a top box at room temperature
for electronics, a long tube through which electronic wiring is run to reach the sample, and a
bottom sample container and stage. Figure 9 shows a diagram and pictures of the completed
helium dipper. The following subsections describe the design of each part as well as an
explanation of the intent and restrictions which influenced the design choice.
FIG. 9. Schematic and picture of the helium 4 dipper outlined in this manual.
Top Electronics Box
In order to run experiments at low temperatures a variety of electronics need to be con-
nected to the sample at 4.2K. This involves wiring (covered in a following section) and
electronic hook ups. A large, 8.7in x 5.7in x 2.2in, aluminum enclosure box (Mouser Elec-
tronics) houses the connections and wiring. Since changing wiring and connections after
construction of the fridge can be difficult, it is important to have a large number of versa-
17
tile connections such that the same wiring can be used in a variety of experiments. There
are a variety of hook up options available through distributors such as Mouser electronics,
Newark, and Digikey. These hook ups can be tailored to any specific project. Our enclosure
box houses 24 coaxial cable feedthroughs purchased from Newark electronics (Figure 10).
Coaxial hook ups allow for a diverse set of connections, a large bandwidth, and a variety
of input/output adapters. In order to prevent electrostatic build up on the enclosure, it
is important to have coaxial feedthroughs with floating insulation rather than connections
without a grounding loop. Each coaxial feedthrough is attached to a switch (Newark) which
connects the central pin to either a wire going down to the sample stage (on) or to a ground-
ing pin (off). An electronic schematic is shown in Figure 11. This allows each hook up to
be turned on only when in use ensuring unused hook ups and wires do not cause additional
noise and thermal conduction to the sample. Each coaxial grounding pin is connected to
a single coaxial connection which can be connected to an external grounding bar (Figure
12). It is important when wiring to avoid grounding loops which can short the circuit, cause
inaccurate measurements, and prevent current.
The enclosure box is connected to the vacuum can tube with a 25KF four way cross
and hermetically sealed wire feedthroughs. Originally, a high vacuum epoxy (Torr Seal)
was used to create a vacuum seal between the top enclosure box and the tube of the main
body. However, this proved to only hold a vacuum to 10−3Torr and thus was replaced
with hermetically sealed feedthroughs. Since a single feedthrough with both dc and coax
lines is expensive, the wiring to the top box was broken into two feedthroughs, one 19 pin
circular connector and a BNC connector feedthrough (Figure 13). The larger top enclosure
is connected via clamps to the 19 pin hermetically sealed flange. The BNC connector leads
4 connections to one side of the four way cross. This is connected to 4 BNC hookup via a
small breakout box. In order to connect the 25 KF cross to the 40 KF hermetically sealed
DC connection and the 50 KF BNC connector an adapter was used. The remaining side
of the cross is used to pump down the vacuum with a standard roughing pump and turbo
pump.
18
FIG. 10. Schematics and pictures of the enclosure box with 24 coax connections and 24 switches
is shown above. One additional coax connector is added on the side to connect the ground to an
external grounding bar.
19
FIG. 11. A wiring schematic of the coax connections and an image of the wiring in the box. Each
coax plug is connected to a switch which toggles between a wire going to the sample stage and a
grounding bar.
FIG. 12. The enclosure box with all coax connections wired to switches connecting the plug to
either ground or the sample. Wiring choices are described in a following section. A strip of copper
was used as a grounding bar to ease soldering.
20
FIG. 13. Assembly of the hermetically sealed connection between the main body of dipper and the
top electronics box. This allows for a vacuum sealed lower sample stage and main body. The top
electronic box connects via wiring to the 19 pin connector, while the main body is connected on
the opposite side.
21
Main Body
Connecting the room temperature enclosure to the sample at 4.2K is a vacuum can
made of a 50 inch long 304L stainless steel tube with a one inch outer diameter and a wall
thickness of 0.060 inch (McMaster Carr). The tube must fit in the opening of the helium
dewar and allow the sample to reach the bottom of the dewar. The length and diameters
choosen for this design was optimized for insertion into a Cryofab CMSH 100 LHe dewar
(Matheson Trigas). Since the tube goes from room temperature to liquid helium at 4.2K,
heat flow along the can walls must be minimized. Since significant thermal conductivity
requires more cooling power from the helium, the vacuum tube is designed to minimize heat
conduction. The 304L Stainless steel used has a low thermal conductivity at both high and
low temperatures. Additionally, a thin tube further minimizes the conduction by decreasing
the area in which heat can flow.
In order to control temperature and run thermoelectric measurements, the tube is pumped
down to a few militorr with a roughing pump. The vacuum has the additional benefit of
minimizing the conduction of heat through gas particles. The pump is attached to the
stainless steal tube with a 25KF flange vacuum hose adapter (Kurt Lesker) to the four way
cross at the top of the can. While there are other insulation options available such as MLI
sheets, these are difficult to effectively install in the narrow vacuum can.
Finally, in order to prevent thermal radiation from room temperature, radiation shields
made of high conductivity copper are equally spread throughout the tube. The shields,
copper baffles, consist of small thin disks which just fit in the inner diameter of the stainless
steal tube, seen in Figure 13. To hold the baffles in place the copper disks are brazed to
a thin (0.060 inch) stainless steel tube using Stay-Silv 45. Each disk has two large holes
through which wires can be threaded. To increase radiation shielding the copper is buffed
until shinny. The rod of baffles is threaded with wire and slid into the long main body
tube (Figure 14). These copper baffles act as a radiation shield ensuring that the 4.2K part
of the fridge is not in direct radiation contact with the room temperature portion. Since
thermal radiation goes as the temperature difference to the fourth, even small decreases
in the temperature difference can drastically decrease parasitic heating. By minimizing
conduction and radiation heat sources can be reduced and helium usage optimized.
A 50KF flange with a one inch vacuum tube adapter (Kurt Lesker) was used to connect
22
the vacuum tube to the top of the dewar. While the top of the dewar is fitted with a triclover
flange, a 50KF offers a reasonably good seal and more fitting options (Figure 15). However,
an adapter can be custom made with a triclover flange offering a better fit. The flange was
attached along the main body tube between the top KF flange and the bottom sample stage
such that it is free to slid along the length and can be tighten with an o-ring seal. This
allows the top of the dewar to be sealed when the dipper is in the dewar minimizing leakage.
FIG. 14. Above are pictures and schematics of the copper baffles used as radiation shields.
FIG. 15. 50KF flange with vacuum tube adapter used to seal the dipper to a liquid helium dewar.
23
Sample Stage
The sample stage should be in good thermal contact with the surrounding liquid helium
and easily accessible for sample placement and wiring. In order to exchange samples a
vacuum grade demountable seal must connect the sample can to the main body. There are
many options available for demountable cryogenic seals: Conflat flanges, Indium corner and
face joints, Helicoflex gaskets, Kapton and Mylar gaskets, and conical seals. Flange seals
and gaskets are simple and effective, but too large to fit in the mouth of the helium dewar.
Indium seals are often used in cryogenic applications, but these seals need to be replaced
frequently, are expensive, and are environmentally toxic. This design uses a simple conical
seal consisting of a short 304L stainless steel cone with a 5 to 15 degree incline and holes
that allow wiring to reach the sample stage and a 304L stainless steel can which connects to
the can with vacuum grease (Kurt Lesker) (Figure 16 and Figure 17). The cone is welded to
the bottom of the long stainless steel tube with a vacuum tight weld. The cone and can seal
when the dipper pumped down to vacuum using the pressure difference to hold a vacuum
tight seal.
The sample stage is made of a copper rod and a copper paddle (Figure 18). The stage
is connected to the vacuum can by treading the copper rod and allowing the sample stage
to screw into the cone. This allows sample paddles of different configurations to be inter-
changed. The similar thermal contraction of stainless steel and copper ensures that the
sample paddle remains connected even at low temperature. Connecting the stage to the
upper part of the dipper simplifies wiring and minimizes the potential for breakage.
A sample chip holder was designed in order to connect the sample to the paddle and to
the electronics. While, there are sample chips available for purchase (Montana Instruments)
these are often expensive and cannot be customized. To fabricate the chip holder EAGLE
(a electronic circuit board design program) was used to fabricate a PCB board to which
a microD connector could be soldered, a sample could be placed, and connections made
between the sample and small connector pads (Figure). The PCB is attached to the paddle
using 0-80 screws. To create thermal contact between the sample chip and the sample paddle
the center of the PCB board was removed so that the sample was placed directly in contact
with the copper paddle.
Finally, the sample stage must be in thermal contact with the liquid Helium outside the
24
bottom stainless steel can. The stainless steel cone and can have a low thermal conductivity
and thus do not provide good thermal contact. Furthermore, the vacuum surrounding the
sample stage prevents conduction through gas particles. To ensure that the sample stage
and thus the sample reaches 4.2K a copper wire is threaded from outside through the bottom
of the vacuum can then wrapped around the rod of the sample stage. The hole is then sealed
by brazing. This design allows the sample to be in contact with the surrounding helium and
in vacuum.
Multiple designs for the bottom can and sample stage were considered. Attaching the
sample stage to the bottom can with the use of a copper cold finger allowed for good thermal
contact with the liquid helium at the bottom of the dewar, but the wiring and the sample
would be separated until sealed with the vacuum making electrical connections to the sample
difficult. Attempting to have an upper stage connecting to a lower cold finger would require
extremely precise machining and careful calculation of thermal expansion. Copper is too
malleable and difficult to machine to be used to construct a thermally conductive cone and
can. The final design provides a simple way to connect the sample stage to the liquid helium
and remained anchored to the upper can and electronics.
25
FIG. 16. The figure shows the cone portion of the conical seal. The cone has a threaded hole to
hold the sample stage and four holes to thread the wiring.
26
FIG. 17. The figure shows the can portion of the conical seal consisting of a stainless steal can
with a bevel at the top matching the angle of the cone seal.
27
FIG. 18. The figure shows the schematic and pictures of the copper sample stage.
28
Wiring
To conduct experiments wiring must run from the room temperature enclosure to the
sample at 4.2K. The material, diameter, and number of wires should be chosen to reduce
the heat conduction along the wires. A material with a large thermal resistivity but a small
electrical resistivity at a temperature range from 300K to 4.2K is ideal. Typical copper
wiring, while having a low electrical resistivity, has a high thermal conductivity so is a poor
option. Thin Manganin and Phosphor Bronze wires have low electrical resistivity and high
thermal conductivity (Table IV).
TABLE IV. The table shows the thermal conductivity and electrical resistivity of common wire
material and 300K and 4K.
Material AWG Thermal Conductivity (W/m ∗K) Resistivity (Ω/m)
300K 4K 300K 4K
Copper (C110) 30 400 300 0.32 0.003
Copper (C110) 34 400 300 0.81 0.0076
Phosphor Bronze 32 48 1.6 4 3.3
Phosphor Bronze 36 48 1.6 10 8.6
Manganin 30 22 0.5 9.7 8.6
Manganin 36 22 6 39 35
Quad-twist 36 Phosphor Bronze wire was purchased from Lakeshore electronics and sol-
dered to 14 of the coax connections. 25 C Coaxial Cable from Lakeshore electronics were
soldered to 4 coax hookups, Thermal couple wire for Lakeshore to two, and plain copper
wire to 4 for a magnet and heater. Due to the thermal contraction of the wires at low tem-
peratures, all wiring needs to have extra length, slack, and large loops to prevent the wires
for breaking while cooling. The wiring should be long to the radiation shield and coiled up
at low temperature region to take up less space. Additionally, using twisted pairs or wire
loops can simplify wiring and reduce electronic noise. To further reduce thermal conduction,
all wires should be heat sunk by wrapping them around copper shielding and the sample
stage (Table V). To ensure accurate measurements wires should be heat sunk adjacent to
29
any sensors.
TABLE V. The table show recommended heat sinking for common cryogenic wiring.
Heat sinking length (mm)
Material 0.21mm2 0.032mm2 0.031mm2 0.005mm2
(24 AWG) (32 AWG) (36 AWG) (40 AWG)
Copper 688 233 138 80
Phosphor Bronze 38 13 7 4
Maganin 20 7 4 2
Thermal conduction through the electrons traveling down the wires is unavoidable. How-
ever, allowing each connection to be turned off reduces the heat conduction through electrons
and minimizes another source of heat.
CONCLUSION
As an alternative to larger commercially available cryostats, we have designed an effective
helium 4 dipper which can cool samples to 4.2K by inserting the cryostat directly into the
dipper. In contrast to many dipper designs this dipper keeps the sample in vacuum allowing
the sample to be heated and thermoelectric measurements to be made. The design concerns
described in this paper can be applied to any cryostat design and modified for a variety of
experiments.
30
APPENDIX A: DESIGN DRAWINGS
Appendix A provides the Solidworks Drawings for each of the machined components of
the helium 4 dipper. All units are given in inches.
31
50.
00 7
0.0
7.0
0 4.7
6 1.
25
1.
00
Cop
per B
affe
ls
Con
e/C
onic
al
Seal
Can
Sam
ple
Stag
e
Top
Elec
troni
cs
Box
4 W
ay 2
5 KF
Fl
ange
Dip
per B
ody
AA
BB
22
11
Heliu
m 4
Dip
per
DO
NO
T SC
ALE
DRA
WIN
G
Cry
oSH
EET 1
OF
1
Kirst
en B
lagg
UNLE
SS O
THER
WIS
E SP
ECIF
IED
:
SCA
LE: 1
:12
WEI
GHT
:
REV
DW
G.
NO
.
ASIZE
TITLE
:
NA
ME
DRA
WN
FIN
ISH
MA
TERI
AL
304L
Sta
inle
ss S
teel
, Alu
min
um, C
oope
r
INTE
RPRE
T G
EOM
ETRI
CTO
LERA
NC
ING
PER
:
DIM
ENSI
ON
S A
RE IN
INC
HES
32
8.5
0
5.7
0
.5
5 .2
7
1.1
0
1.1
0
.60
5.7
0
2.2
0
1.
01
Epox
ed
to 2
5 KF
fla
nge
Coa
x ca
ble
hook
upSwitc
h
Gro
und
ing
coax
hoo
kup
AA
BB
22
11 Top
Elec
troni
cs
Box
DO
NO
T SC
ALE
DRA
WIN
G
Brea
kout
Box SH
EET 1
OF
1
Kirst
en B
lagg
UNLE
SS O
THER
WIS
E SP
ECIF
IED
:
SCA
LE: 1
:4W
EIG
HT:
REV
DW
G.
NO
.
ASIZE
TITLE
:
NA
ME
DRA
WN
FIN
ISH
MA
TERI
AL
Alu
min
um
INTE
RPRE
T G
EOM
ETRI
CTO
LERA
NC
ING
PER
:
DIM
ENSI
ON
S A
RE IN
INC
HES
33
.1
5
.80
.0
6
.50
Braz
ed u
sing
Stay
-Silv
45
and
Sta
y-sil
v W
hite
Bra
zing
Flux
to 0
.06
dia
met
er
stai
nles
s ste
el tu
be
Hole
s to
allo
w
wiri
ng to
reac
h th
e bo
ttom
sa
mpl
e
.05
.1
5
.8
0
.0
6
AA
BB
22
11 Cop
per B
affe
l/ Ra
diat
ion
Shie
ld
DO
NO
T SC
ALE
DRA
WIN
G
CuB
affle
SHEE
T 1 O
F 1
Kirst
en B
lagg
UNLE
SS O
THER
WIS
E SP
ECIF
IED
:
SCA
LE: 2
:1W
EIG
HT:
REV
DW
G.
NO
.
ASIZE
TITLE
:
NA
ME
DRA
WN
FIN
ISH
MA
TERI
AL
Cop
per
INTE
RPRE
T G
EOM
ETRI
CTO
LERA
NC
ING
PER
:
DIM
ENSI
ON
S A
RE IN
INC
HES
34
.1
5
.2
5
.8
615
1.
00
1.
11
.30
1.
00
.8
615
.75
.875
.50
.8
750
.7
5
Wel
ded
to lo
ng
stai
nles
s ste
el tu
be
bod
y
Thre
aded
hol
e fo
r cop
per
sam
ple
stag
e
Hole
s for
w
iring
to
feed
thro
ugh
8 d
egre
e an
gle
man
ufat
ured
w
ith th
e us
e on
a
sin b
ar a
nd
forc
e ga
uge
Inne
r and
out
er
dia
met
er
mat
ch th
e st
ainl
ess s
teel
tu
be b
ody
for
ease
of
wel
din
g
AA
BB
22
11 Stai
nles
s Ste
el
Con
e
DO
NO
T SC
ALE
DRA
WIN
G
SSC
one SH
EET 1
OF
1
Kirst
en B
lagg
UNLE
SS O
THER
WIS
E SP
ECIF
IED
:
SCA
LE: 1
:1W
EIG
HT:
REV
DW
G.
NO
.
ASIZE
TITLE
:
NA
ME
DRA
WN
FIN
ISH
MA
TERI
AL
304L
Sta
inle
ss S
teel
INTE
RPRE
T G
EOM
ETRI
CTO
LERA
NC
ING
PER
:
DIM
ENSI
ON
S A
RE IN
INC
HES
35
1.
25
1.
13
1.
06
Wel
ded
cap
to
the
botto
m
of th
e tu
be
1.
06
1.
25
1.
13
.70 4.0
0
.75
.70
1.
25
1.
13
Tube
Cap
Ass
embl
y
AA
BB
22
11 Stai
nles
s Ste
el
Can
DO
NO
T SC
ALE
DRA
WIN
G
SSC
anSH
EET 1
OF
1
Kirst
en B
lagg
UNLE
SS O
THER
WIS
E SP
ECIF
IED
:
SCA
LE: 1
:2W
EIG
HT:
REV
DW
G.
NO
.
ASIZE
TITLE
:
NA
ME
DRA
WN
FIN
ISH
MA
TERI
AL
304L
Sta
inle
ss S
teel
INTE
RPRE
T G
EOM
ETRI
CTO
LERA
NC
ING
PER
:
DIM
ENSI
ON
S A
RE IN
INC
HES
36
1.
00
.2
5
3.0
0
.10
Top
tread
ed to
be
scre
wed
into
stai
nles
s st
eel c
one
Plac
men
t of
sam
ple/
sam
ple
chip
car
rier c
an b
e on
top
or o
n bo
ttom
AA
BB
22
11 Sam
ple
Stag
e
DO
NO
T SC
ALE
DRA
WIN
G
CuS
tage
SHEE
T 1 O
F 1
Kirst
en B
lagg
UNLE
SS O
THER
WIS
E SP
ECIF
IED
:
SCA
LE: 1
:1W
EIG
HT:
REV
DW
G.
NO
.
ASIZE
TITLE
:
NA
ME
DRA
WN
FIN
ISH
MA
TERI
AL
Cop
per
INTE
RPRE
T G
EOM
ETRI
CTO
LERA
NC
ING
PER
:
DIM
ENSI
ON
S A
RE IN
INC
HES
(1) K
F-25
4-w
ay c
ross
(2) 1
9-pi
n he
rmet
ical
ly
seal
ed K
F-25
circ
ular
con
nect
or (4
) Ele
ctro
nics
Box
(5
) KF-
25 h
inge
cla
mp
(6) K
F-40
hin
ge c
lam
p (7
) K
F-50
hin
ge c
lam
p (8
) KF-
40 to
KF-
25 c
onic
al
redu
cer
(9) K
F-50
to K
F-25
con
ical
redu
cer (
10)
KF-
50 c
ap (1
1) K
F-25
cap
Wire
Sea
l
15
11
5
5
9
7
10
82
6
4
15
11
5
5
9
7
10
82
6
4
1
5
115
5
9
7
108
2
6
4
1
5
115
5
9
7
108
2
6
4
15
11
5
5
9
7
10
82
6
4
15
11
5
5
9
7
10
82
6
4
15
11
5
5
9
7
10
82
6
4
15
11
5
5
9
7
10
82
6
4
37
.062
.920
.440
.394
.046
1.130
1.500
.118 .394
TRUE R.050
.070
.150
.920
.240
.440
.051
A A
B B
2
2
1
1
WEIGHT:
SW988C
PROPRIETARY AND CONFIDENTIALTHE INFORMATION CONTAINED IN THIS DRAWING IS THE SOLE PROPERTY OF <COMPANY NAME >. ANY REPRODUCTION IN PART OR AS A WHOLE WITHOUT THE WRITTEN PERMISSION OF <COMPANY NAME> IS PROHIBITED.
COMMENTS:
SHEET 1 OF 1
Q.A.
MFG APPR.
ENG APPR.
CHECKED
DRAWN
DATENAMEDIMENSIONS ARE IN INCHESTOLERANCES:FRACTIONALANGULAR: MACH BEND TWO PLACE DECIMAL THREE PLACE DECIMAL
NEXT ASSY USED ON
APPLICATION DO NOT SCALE DRAWING
FINISH
MATERIAL
REV.
ADWG. NO.SIZE
SCALE:2:1
1.00
.25
3.00
.10
Top treaded to be screwed into stainless steel cone
Placment of sample/sample chip carrier can be on top or on bottom
A A
B B
2
2
1
1
Sample Stage
DO NOT SCALE DRAWING
CuStageSHEET 1 OF 1
Kirsten Blagg
UNLESS OTHERWISE SPECIFIED:
SCALE: 1:1 WEIGHT:
REVDWG. NO.
ASIZE
TITLE:
NAME
DRAWN
FINISH
MATERIALCopper
INTERPRET GEOMETRICTOLERANCING PER:
DIMENSIONS ARE IN INCHES
PCB board designed in
EAGLE
Cut out for sample
placement
38
APPENDIX B: PURCHASE LIST
Below lists the parts and materials purchased for the manufacturing of the helium 4
dipper described in this manual. Most parts can be purchased from various distributors.
Extra raw material was purchased in order the ease machining and allow for practice parts
and welding.
39
TABLE VI.
Part Quantity Supplier Cost
Stainless Steel Tube, 60” length, 1” OD, 0.88” ID 1 Kurt Lesker 124.20
Triclover Flange, 3” 1 Brewer’s Hardware/Amazon 36.25
Stainless Steel Cylinder, 1-3/8” Dia, 0.5’ Length 1 McMaster Carr 21.13
Stainless Steel Tube, 1-1/4” OD, 10.084 ID, 0.5’ length 1 McMaster Carr 19.34
Copper Cylinder, 1” Dia, 1’ Length 1 Scrap Material NA
Thin Stainless Steel Tube, 60” length, 0.06” OD 1 Kurt Lesker 68.00
Enclosure, 8.7” x 5.7” x 2.2” 1 Mouser Electronics 74.84
Four Way Cross, 25 KF Flange 1 Ebay 50.00
25 KF Flange 2 Kurt Lesker 20.52
BNC Hookups 30 Newark 73.50
Switches 30 Newark 69.30
Thermocouple Wire 3 m Lakeshore Electronics 81.00
Coax Cable 25 ft Lakeshore Electronics 173.00
Phosphor Bronze Wire 50 ft Lakeshore Electronics 300.00
Vacuum Grease 1 Amazon 6.55
Torr Seal 1 Kurt Lesker 63.75
Brazing Supplies 1 Amazon 29.16
PCB 1 Ebay 2.99
62GB-12E14-19SN Circular Connector 1 Newark 33.22
Flat Ribbon Cable Black 20 Conductors 1 Digikey 37.78
Reducer conical KF-25 to KF-50 1 Ideal Vacuum 48.00
Hinge Clamp KF-50 1 Ideal Vacuum 10.95
Centering Ring KF-50 1 Ideal Vacuum 10.85
KF-25 Cap 1 McMaster Carr 10.53
KF-25 Ring 4 McMaster Carr 33.60
KF-25 Hinge Clamp 1 McMaster Carr 33.75
40
Thank you to Mike Manz and Randy Bachman in the Colorado School of Mines Physics
Machine Shop for their invaluable expertise on machining and welding all of the components
of this assembly.
Thank you to Jonathan Watson and Devon Gonzales in the Colorado School of Mines
Brazing and Welding Lab for help with the copper to stainless steel brazing.
[1] E. T. Swartz, Review of Scientific Instruments 57, 2848 (1986).
[2] A. M. Putnam, D. A. Geller, and V. Alexis, Physica B: Condensed Matter 194-96, 57 (1994).
[3] F. Pobell, Matter and Methods at Low Temperatures, 3rd ed. (Springer, 2007).
[4] J. W. Rohlf, Modern Physics from a to Z0 (Wiley, http://hyperphysics.phy-
astr.gsu.edu/hbase/thermo/heatrf.html, 1994) an optional note.
[5] “The great soviet encylopedia,” (The name of the publisher) 3rd ed.
[6] P. Duthil, arXiv preprint arXiv:1501.07100 (2015).
[7] C. Y. Ho, R. W. Powell, and P. E. Liley, Journal of Physical and Chemical Reference Data 1,
279 (1972).
[8] G. A. Slack, R. A. Tanzilli, R. Pohl, and J. Vandersande, Journal of Physics and Chemistry
of Solids 48, 641 (1987).