Redesign, assembly and test procedure of laboratory Mars atmosphere sampling system

(1)

MASTER’S THESIS

2002:374 CIV

CAMILLA GRANSTRÖM

Redesign, Assembly and

Test Procedure of Laboratory

Mars Atmosphere Sampling System

MASTER OF SCIENCE PROGRAMME in Space Engineering

Luleå University of Technology Department of Space Science, Kiruna

(2)

Abstract

The broad interest in exploration of Mars demands instruments to be developed for the preliminary Mars Smart Lander Mission in 2009 by NASA. This report describes the development at the University of Michigan of a laboratory–scale Mars atmospheric sampling system redesigned after an earlier instrument built by Southwest Research Institute (SwRI) in Texas. This sampling system will demonstrate a method for an instrument that measures the composition of Mars’

atmosphere and the isotopic ratios of noble gases in the atmosphere, which tell us about the atmospheric loss processes. The analysis of these measurements will help determine the past and present climate on Mars and its atmospheric evolution. The technique is based on using two concentration modules with gas adsorbing material that scrubs away the gases of no interest like CO2, CO, O2

and water vapor, so that the trace amounts of noble gases can be determined by a mass spectrometer. The adsorbing materials that are used are different kinds of zeolite and one type of non-evaporable getter. Reproducibility of the tests will not only accurately measure the noble gases, it will also determine which zeolite that have the best characteristics and should be used in the future instrument as well as determine the life time of the non–evaporable getter.

A short time limit for the project resulted in redesign and an almost complete system that is ready for test. A pressure region in the low 10⁻⁸mbar was reached several times at pump down of the system before bakeout, which indicate a good vacuum. The mixing procedure for the test gas was developed and the equipment for mixing was assembled. The testing procedure is also developed, but further work needs to be done on the system and the test gas mixed before the tests can be carried out.

(3)

(4)

Preface

This report presents the results of a project work made to complete my Master of Science degree in Space Engineering at Lule¨ı¿¹₂University of Technology in Lule¨ı¿¹₂Kiruna, Sweden. The project work was performed at the University of Michigan at the Department of Atmospheric, Oceanic and Space Sciences.

I would like to thank Professor Jack Hunter Waite at the Department of Atmospheric, Oceanic and Space Sciences at the University of Michigan in USA for giving me the opportunity to do this project work and being my supervi- sor. I would also like to thank people at the department that helped me out, Bruce P Block for technical support, Ronald Rizor for drawings in AutoCAD and Sandy and Denise at the purchasing department. A special thanks goes to my advisor and examiner Bj¨ı¿¹₂n Graneli at the Department in Kiruna, Sweden for his 24–hour support.

I would also like to take the opportunity to thank my family for uncondi- tionally loving me and supporting me through good times and bad times and all my friends for always being there for me. A special thought is left for my love, the star of my heaven.

Ann Arbor, Michigan, USA Gunnarn, Sweden

2002-07-23

Camilla Granstr¨ı¿¹₂

(5)

(6)

List of Figures

3.1 Schematical drawing of the gas handling system. . . 12

3.2 Drawing of the gas handling system. . . 13

3.3 Drawing and cross section of the Zeolite module . . . 14

4.1 The general outline of the Mars atmospheric sampling system . . 15

4.2 Drawing of the whole system in the laboratory environment. . . . 16

4.3 Free standing system without table frame. . . 17

4.4 Enlargement of concentration chambers and valves. . . 17

4.5 Cross section of the Zeolite module. . . 19

4.6 Exploded view of the Zeolite module. . . 19

4.7 Digital photo of the parts for the Zeolite chamber. . . 20

4.8 Cross section of the NEG module. . . 21

4.9 Exploded view of the NEG module. . . 22

4.10 Getter assembly before mounting in NEG module. . . 22

4.11 Schematic view of the quadrupole mass spectrometer system. . . 23

4.12 A diaphragm gauge. . . 26

4.13 The inlet system . . . 27

4.14 All–metal gate valve. . . 28

4.15 Inside of a gate valve and its closing function. . . 28

4.16 Cross section of a right angle valve. . . 30

4.17 Cross section of a straight through valve. . . 30

4.18 Sealing mechanism in a CF flange . . . 33

4.19 The table frame with the table top of Marinite I. . . . 35

4.20 First assembly stage of gate valve . . . 35

4.21 Second assembly stage of gate valve . . . 36

5.1 Schematic view of the gas–mixing mode. . . 39

5.2 Illustration of the manifold with the Nupro valves. . . 39

5.3 Mixing tank with leak valve and pressure gauge. . . 40

B.1 NEG assembly exploded view. . . 58

B.2 NEG assembly. . . 59

B.3 Zeolite chamber assembly. . . 60

B.4 Zeolite heater and bushing. . . 61

B.5 Macore housing for Zeolite chamber. . . 62

B.6 Macore cap for Zeolite chamber. . . 63

B.7 Macore retaining ring for Zeolite chamber. . . 64

B.8 Frame 1 of 3 . . . 65

(9)

LIST OF FIGURES

B.9 Frame 2 of 3 . . . 66

B.10 Frame 3 of 3 . . . 67

B.11 Tubulation support. . . 68

B.12 Valve support assembly. . . 69

B.13 Mounting plate. . . 70

B.14 Bracket for dryer and afterfilter. . . 71

B.15 Tubular fix of tee. . . 72

B.16 Table top . . . 73

C.1 Bakeout zone 1 with temperature 350 ^◦C. . . 75

(10)

List of Tables

2.1 The composition of the Marsian atmosphere. . . 5

2.2 Isotopic ratios on Mars . . . 5

3.1 Composition of Molecular sieves 4A and 5A. . . 8

4.1 The components of the quadrupole mass spectrometer system. . 23

4.2 Technical data for TPU 261 and MPV 032-2 . . . 25

4.3 Data for the all–metal gate valve from [26]. . . 29

5.1 The constant K given by [32] for different gases. . . . 38

5.2 Data for mixing of gas . . . 38

5.3 Start and stop pressures in the mixing procedure. . . 41

5.4 Mixing procedure. . . 41

6.1 Test procedure in steps. . . 44

A.1 Valves. . . 49

A.2 ConFlat and QF hardware. . . 50

A.3 Measurement and pumping equipment. . . 51

A.4 Custom made parts. . . 52

A.5 Test equipment. . . 53

A.6 VCR and Swagelok. . . 54

A.7 Gaskets and bolts. . . 55

A.8 Accessories. . . 56

B.1 Drawings for the system. . . 57

(11)

LIST OF TABLES

(12)

Chapter 1 Introduction

Mars exploration is of important scientific interest in the search for life else- where in the solar system and over the next 15–20 years a planned armada of international missions are to be sent to Mars. The purposes of the missions are to understand the planets geology, search for past and present life and determine the climate history. This global interest in Mars is based on the following reasons:

• Mars is the most Earthlike planet in our solar system.

• Mars has a varied geological and climate history.

• Known planetary processes on Earth have apparently operated on Mars, under different conditions, scales and rates.

• Liquid water plays a significant role in the evolution of Mars surface.

• Speculations of life from development of indigenous life forms or coloniza- tion by terrestrial life forms from interplanetary meteorites.

• Most likely the next planet to be visited by humans.

At the present time the mission strategy is robotic and scientifically driven and is explained more completely in [1].

The goal of this report is the development of an instrument for the Mars 2009 Smart Lander Mission. The scientific objectives for the mission are divided into four parts, development and evolution of life, current and past climate, structure and dynamics of interior and preparation for human exploration and development of space. One of the aspects for determining the current and past climate of Mars is to determine the composition and isotopic ratios of the atmosphere with high accuracy. Mars’ atmosphere consists of approximately 95% CO2and only trace amounts of the noble gases that are of interest. Therefore a concentration system needs to be developed that reduces the constituents of large amounts and concentrates the constituents of trace amounts to measurable values. The normal way of doing this is based on cryogenic systems that demand a lot of energy. A new method that uses molecular sieves and getters that are not as power demanding is therefore to be tested.

A not–fully–tested, small pre–model of the instrument was build at SwRI.

(13)

The decision at the University of Michigan was therefore to build a laboratory model of the concentration system that will prove the theory for the future instrument for the Mars 2009 Smart Lander Mission. The test phase of the laboratory model will give answers to which concentration materials that should be used for best performance in the real instrument. With guidelines from the pre–model instrument, the Laboratory Mars Atmosphere Sampling System was redesigned and assembled. The procedure for mixing the test gas and the procedure for performing the final tests of the system was also developed. This report will explain the scientific background, the experimental techniques, scientific equipment, the different parts of the sampling system, the assembly, the test gas mixture and the test procedure.

(14)

Chapter 2 Scientific background

2.1 Mars atmosphere

The composition of Mars’ atmosphere has been studied since 1930 with spectroscopy methods. But it was in the middle of the 1960’s that the composition and pressure was first determined with any degree of certainty. The dynamics of Mars’ atmosphere as described in [2] is found in many ways to be less complex than Earth’s, because Mars’ atmosphere has less heat capacity and no influencing oceans on the ground that can transfer heat or develop moisture that interacts with the air. The climate system on Mars is fundamentally very complex with respect to physical and chemical processes within the interior, the crust, the surface and the different sections of the atmosphere. The interaction with the solar wind is also involved according to [3].

Compared to Earth, Mars is 150 million km farther away from the sun and does not get the same amount of solar radiation. Also Mars’ elliptical orbit gives the planet a less uniform radiation from the sun. At perihelion the planet is exposed to 40% more of the sun’s rays than at aphelion. On top of this, [3] gives the output of the sun as being lower 4 Gyr ago and makes a warmer climate on Mars in the early epoch more difficult to imagine.

A lot of information about Mars’ atmosphere comes from the Viking mission and Mars Pathfinder Mission. Table 2.1 is based on information from [4].

The main constituent in the atmosphere is carbon dioxide (CO2), followed by nitrogen (N2) and argon (Ar), and many constituents in trace amounts, like noble gases. The amount of dust present close to the surface is believed to be 5–30 particles/cm³ according to [2]. The amount of water vapor, dust, carbon monoxide and ozone is variable, especially the variation of water vapor and dust, which varies with the season and location. The average column abundance for water vapor is 15 pr-µm and this figure can vary from zero to 100 pr-µm with location and season.

The evolution of Mars’ atmosphere can be divided into different time epochs over the last 4.6 Gyr. The earliest epoch, Noachian was present 4.6 to 3.7 Gyr ago. Four Gyr is as far back as the history of the atmosphere and volatiles can be tracked due to observations, in–situ measurements and measurements from meteorites. It is a general theory given in [5] that Mars’ atmosphere and volatiles are primarily derived from a major reservoir such as the solar nebula or

(15)

2.2. SCIENTIFIC OBJECTIVES

gases modified during the early planetary formation. Loss processes of different kinds have had a modifying effect on the atmosphere during its evolution. Al- though according to [3], atmospheric gases were present long before 4 Gyr ago.

A possible evolution of the atmosphere is given here, based on [3]. The heavy impacts of the large asteroids in the early Noachian epoch may have removed 50–90% of the atmosphere. But impacts of volatile rich objects may have also enriched the atmosphere as well as produced outgassing from the interior of the planet and released H2O and CO2 from Tharsis volcanism later in this epoch.

The analysis of isotopic ratios of heavy noble gases from SNC meteorites that are assumed to represent the atmosphere of Mars suggests that comets could have supplied much of the planets volatiles. The atmosphere was then lost again. The atmospheric escape on Mars is explained in [6]. There is evidence for many of the processes, but it is not strong enough to determine the absolute consequences for the atmosphere. Water and CO2 were also lost into the polar caps and into the planet’s crust. Other events that likely have affected the climate and atmosphere include the shut down of the intrinsic magnetic field, changes in the solar wind over time and the decline in volcanism.

The atmospheric density measured by Mars Pathfinder was measured from 160 km down to the surface. It varies from 5 × 10⁻¹¹ kg/m³ to 0.02 kg/m³. This suggests that the Martian air is very thin and has a low heat capacity.

It will therefore heat up and cool down more rapidly than the atmosphere on Earth. The dust amounts therefore play a big role in temperature differences in the atmosphere. Dust absorbs the solar radiation directly and heats up the atmosphere. When there is little or no dust in the air the atmosphere absorbs very little solar energy directly and conduction or convection from the surface controls the temperature profile. This gives diurnal fluctuations on the surface up to 50 degrees from the normal average surface temperature 220 K given by [3].

According to [4] the noble gases exhibit an abundance pattern fairly similar to the pattern found in the terrestrial atmosphere and the primary components of meteoric gases found in meteors on the moon and the earth. In Table 2.2, a comparison of the isotopic ratios in Mars’ atmosphere versus the terrestrial value is presented based on the information from the two Viking landers. The Viking landers used a scrubbing system containing two cavities that scrubbed away water, CO and CO2 to concentrate the amount of noble gas before analysis. The uncertainties in this table are ±10%, except for Ar and Xe. If the errors are ap- plied on the other gases their values will be in the range of the terrestrial values.

2.2 Scientific objectives

Studies of atmospheric isotopes can lead to an understanding of the evolution of the Martian atmosphere and climate history. Present day isotopic measurements of climate related species, H, C, O and N, are important as well as the isotopic ratio of the noble gases. The noble gases are in particular useful indi- cators since being inert they are not removed from the atmosphere or surface by chemical reactions. Therefore they give direct information about the atmospheric loss processes on Mars.

Determination of the loss processes and their effects on the atmosphere will

(16)

CHAPTER 2. SCIENTIFIC BACKGROUND

Table 2.1: The composition of the Marsian atmosphere.

Gas Proportion

Carbon dioxide (CO2) 95.32%

Nitrogen (N2) 2.7%

Argon (Ar) 1.6%

Oxygen (O₂) 0.13%

Carbon monoxide (CO) 0.07%

Water vapor (H2O) 0.03%

Neon (Ne) 2.5 ppm

Krypton (Kr) 0.3 ppm

Xenon (Xe) 0.08 ppm

Ozone (O3) 0.03 ppm

Table 2.2: Isotopic ratios on Mars

Ratio Earth Mars

12C/¹³C 89 90

18O/¹⁹O 499 500

14N/¹⁵N 277 165

40Ar/³⁶Ar 292 3000

129Xe/¹³²Xe 0.97 2.5

help construct a global inventory of the species on Mars including the quantity of water. The measurements by Viking have been helpful in many ways, but new measurements with improved methods need to be performed to fulfil the scientific goals. Accurate measurements of the isotopes have been carried out on meteorites like SNC and ALH84001 but there is nothing that proves that this is the composition of the present day Martian atmosphere. In a proposal to NASA by Professor J Hunter Waite, [7],one can read that the measurements of Xe, Kr,³⁸Ar/³⁶Ar, ²²Ne/²⁰Ne, ¹⁸O/¹⁶O and ¹³C/¹²C needs to be improved. The measured C and O isotopic ratios contains many uncertainties and conclusions about mass losses that cannot be determined from the present information.By the new techniques it is also desired to first detect He and accurately determine the²²Ne/²⁰Ne ratio.

By refinement of the isotopic ratios of the noble gases, the predicted loss processes of the early atmosphere; hydrodynamic escape, drag and fractionation can be studied. Existing data suggest that xenon fractionated in the primal atmosphere and that all other lighter species were lost to space through hydrodynamic escape. But measurements from the meteorite ALH48001 may represent an ancient atmosphere of Mars where non-fractionated N and Ar trapped in the meteorite indicate that solar wind stripping of the atmosphere did not occur 3.9 Gyr ago according to [3]. This hydrodynamic loss process may have appeared later in the evolution. If all other species were lost by indication of fractionated Xe the current measurements of Kr indicate that species must have been added to the atmosphere by impacts of comets or outgassing from the planets interior. However, keep in mind that impacts of meteorites can also remove large amounts of the atmosphere, a removal that will not change the isotopic ratio.

(17)

2.2. SCIENTIFIC OBJECTIVES

Detection of radioactive decay products such as¹²⁹Xe from¹²⁹I indicates loss of an early atmosphere. And by identifying isotope products from spontaneous fission of ²²⁴Pu and ²³⁸U, a time determination of the atmospheric degassing and the early losses can be performed. Also significant is the determination of loss of key species throughout geological time. O, C and N can be lost by pho- tochemical escape and also by sputtering together with Ne and Ar. These loss processes are mass dependent and non–thermal; this can be seen by fractionation among the species in the atmosphere. The upper atmosphere is enriched with the lighter isotopes due to diffusive separation where the lighter isotopes are removed.

New measurements are expected to confirm the ¹⁵N/¹⁴N ratio and give accurate isotopic ratios for ³⁸Ar/³⁶Ar. Both indicate loss of the atmosphere to space. The ³⁸Ar/³⁶Ar ratio is a very good indicator since it does not fraction- ate by interaction with reservoirs in the crust. There are suggestions that water played a significant role in the creation of many surface features on Mars and that its cycle involves exchanges between the surface and the atmosphere. The primary reservoir sources for water are believed to be the atmosphere, the soil, the permafrost or subsurface water and the polar caps. Among these the atmosphere is the easiest reservoir to perform measurements on.

The D/H ratio in Martian water is in this case interesting and [3] states that there is a five–fold enrichment of D compared to H, which can escape thermally into space. The hydrogen and oxygen lost to space probably comes from water;

therefore the loss of water to space depends on the relative supply rates to the upper atmosphere of H compared to D, their relative escape rates and the initial D/H ratio on Mars. The escape rates of H and D can modify the initial D/H value on Mars and the initial D/H ratio is suggested to be twice as much as the terrestrial value due to meteoritic measurements. It is also determined that the fractionation of D/H and ¹⁸O/¹⁶O are not related even though the losses of O and H from water are. This means that a different reservoir source for O must exist.

(18)

Chapter 3 Experimental technique and scientific equipment

3.1 Concentration and separation methods

Concentration and separation of noble gases in trace amounts has been performed in laboratories around the world for many years. Accurate measurements in laboratory environments can be made with different getters for concentration and cryogenic traps for separation.

One method for future space application is presented in [8]. A small oven with a titanium getter is used for purification and concentration of the noble gases at a temperature of 800^◦ C. After concentration the sample is introduced to the cryogenic trap for separation. The trap will sorb the gases at a low temperature and release them as the temperature rises. In this method a time of flight mass spectrometer with an electrostatic mirror is used for analysis; it will provide a long flight path and correct for energetic aberrations. A pump is present for creating vacuum and pumping residues. In this method, the noble gases under investigation are Ar, Kr and Xe since He and Ne are too difficult to separate in the cryogenic trap, although a new cryogenic trap has been developed that uses stainless steel as an active trapping surface. The advantage with this cryogenic trap over the commonly used charcoal trap is its ability to trap the gases at a low temperature and release them at much lower temperatures.

The release temperature of the gases is mass dependent; it increases with the increasing mass. Using this trap all the noble gases can be separated. A more detailed description of the trap can be found in [9]. The cryogenic method of separation demands a lot of power and is therefore not a suitable separation source for the noble gases in a space application. Therefore another concentration and separation method is preferred that is more power efficient.

Greg Miller, at the Southwest Research Institute (SwRI), developed the method used in our system. The concentration is achieved by using two concentration modules that can scrub away gases of no interest. The first module consists of zeolite in powder form that works like a bulk scrubber of CO2and a water concentrator. The zeolite efficiently traps CO2, CO, O2 and H2O. When the sample gas is scrubbed in the first module it is introduced into the second module. The second module consists of non–evaporable getter of sintered zir-

(19)

3.2. ZEOLITE

conium powder. This getter can absorb all other gases like; CO2, CO, O2, N2, H2O and H2, that the first module did not fully trap, but the getter will not trap the noble gases that are of interest for analysis. The separation of the noble gases for analysis is done by the quadrupole mass spectrometer itself. A pump is also part of the system to create vacuum and pump out residual gases.

3.2 Zeolite

Zeolite is a naturally occurring mineral formed under low metamorphic conditions. Because of its importance in industry, synthetic zeolite also has a broad market. Zeolites can perform ion exchange, filtering, odor removal, chemical sieve and gas absorption tasks. The molecular sieves used for concentration of the noble gases consist of synthetic zeolite. Zeolite is a framework of silicates consisting of interlocking tetrahedrons of SiO4 and AlO4 according to [10]. In order to be a zeolite the ratio in composition

Si + Al

O = 0.5 (3.1)

must be fulfilled. The aluminium–silicate structure is negatively charged and attracts positively charged particles and retains them in the structure. The zeolite structure contains many vacant spaces in the structure where large particles like water can be captured. Zeolites can adsorb and desorb water without dam- aging the crystal structure. Desorption of water takes place when the zeolite is heated. Heating the zeolite can also drive off other gases that are sorbed in its structure. As stated earlier the zeolite efficiently adsorbs CO2, CO, O2 and H2O.

The compositions of the washed zeolites used in the test cycles for observa- tion of its sorption abilities are listed in table 3.1 given by [11]. The difference between the two molecular sieves is the pore size and the ability to capture water. Given in [12] type 4A has a pore size of 4¨ı¿¹₂and a water capacity of 22% and type 5A has a pore size of 5¨ı¿¹₂ and a water capacity of 21.6%. The different mesh sizes on the zeolite for the test is 1/16 inch pellets and mesh size 60/80 and 100/120.

Table 3.1: Composition of Molecular sieves 4A and 5A.

Ingredient 4A 5A

Silicon oxide <50% weight <65% weight Aluminium oxide <30% weight <40% weight Sodium oxide <30% weight <20% weight Magnesium oxide <5% weight

Calcium oxide <20% weight

Quartz <1% weight

3.3 Non–evaporable getter

Non–evaporable getters, are according to [13], very reliable, porous up to 70%

and can operate over a wide range of temperatures. Getters are known to be able

(20)

CHAPTER 3. TECHNIQUE AND EQUIPMENT

to reversibly absorb hydrogen at low equilibrium temperature and irreversibly absorb other gases, water vapor, nitrogen, carbon monoxide, carbon dioxide and oxygen. The irreversible absorption is normally a formation of thermo–stable chemical compounds. The noble gases are scarcely absorbed by the getters.

The material used for getters have relatively low atomic packing density and therefore an increasing rate of diffusional transfer of absorbed gases in to the crystal lattice, all the way into the bulk is possible. It is also stated by [13] that the efficiency of the getters are based upon their ability to absorb hydrogen.

This absorption ability increases along the Hf-Pd-Ta-V-Nb-Ce-La-Ti-Zr series.

This makes zirconium and titanium two perfect base material in getters.

The getter for this system is a sintered porous getter St171 from SAES Getters. This getter contains 83% of a fine zirconium powder as a base, which is mixed with graphite powder. Around an internal heater made of refractor metal coated by an insulating alumina surface layer the getter mass is formed during a sintering process at high temperature and under high vacuum conditions.

Running current through the heater activates it and its construction can also support the mounting of the getter. According to the St 171’s data sheet, the porosity of the getter is 50% and the recommended activation temperature is 900^◦C. The data sheet also provides the following absorption processes. The St 171 can sorb CO, CO2, O2and N2by surface absorption, but for more efficiency a work temperature above 300^◦C will force the gases to diffuse into the bulk.

Water vapor is absorbed as H2and O2after a dissociation process on the getter surface has taken place. The absorption of H2 depends on the temperature and the equilibrium pressure of H2according to Sievert’s law given in the more convenient form by [13] as

log Peq = 2 log G0+ A −B

T (3.2)

Where Peq expressed in Pascals is the equilibrium hydrogen pressure over the getter surface that limits the ultimate pressure that can be attained. G0 is the quantity of hydrogen sorbed by a unit mass of the getter given in m³Pa/kg. T is the temperature given in Kelvin. The absorption constants A and B are 3.6 and 5200 according to St171’s data sheet.

The getter has a passivating layer on its surface that has to be eliminated.

This can be done by an activation process where the getter is heated up to 900

◦C for 10 minutes in a vacuum environment better than 10⁻³mbar. If the getter ability to absorb decreases so much that it does not give the desired results the getter can be reactivated at 850^◦C for 5 minutes, and the absorption ability shall increase again due to the new clean surface created. The getter is also good in the bake–out temperature range of 85–500^◦C. The lifetime of the getter can be calculated in thermal cycles given by [13] based on empirical data provided by SAES Getters. The number N of cycles the getter layer can withstand without cracking and peeling are

N = 6.6 · 10⁷t^−0.4e^−1.2·10⁻²^T (3.3) where t is the operating time in minutes of the getter at a temperature T given in Kelvin. The limiting number N is given with the number of reactivation included.

(21)

3.4. INTRODUCTION TO MASS SPECTROMETRY

3.4 Introduction to mass spectrometry

Mass spectrometry is an analysis method where atoms and molecules are ionized, separated by their mass to charge ratio, m/Z, and recorded usually electrically.

The resulting ion–current, responding to the different masses, is amplified and displayed on oscilloscopes or stored in a computer where software transforms it into displayable data. Compared to other types of spectroscopy methods based on absorption of electromagnetic radiation where the sample eventually returns to its ground state, mass spectrometry is a destructive method of analysis due to the fact that the ions cannot return to molecules or atoms again after ionization.

The mass resolution of a mass spectrometer is often referred to as the re- solving power, R, given by [14] as:

R = M

∆M (3.4)

where M is the mass of detection and ∆M is difference in the masses identi- fied. This ratio will determine the ability of the mass spectrometer to separate the mass M from its neighbouring mass M + ∆M . The mass to charge ratio is usually displayed as a mass spectrum, a graph of the ion intensity as a function of mass to charge ratio. The mass spectrum can be explained as an ordinary histogram, normalized after the highest peak often referred to the value 100. A mass spectrometry system is normally operated under vacuum conditions and consists of the following parts:

• An ion source

• A mass analyzer

• A detector

• A recorder

The names of the mass spectrometer techniques are based on the type of mass analyzer used. A brief description of different analyzers is given here from more detailed explanations in [14] and [15].

The magnetic analyzer is based on focusing or bending the trajectories of the ions in to circular paths by using magnetic fields. Only the ions with the right mass to charge ratio will be able to pass through the curved analyzer and be detected. By changing the strength of the magnetic field different ion trajectories will pass through and therefore different mass to charge ratios can be detected.

The double focusing mass analyzer uses both magnetic and electric fields to focus the ions on the detector. The electrostatic analyzer will focus ions with the same kinetic energy on the magnetic analyzer that can separate the ions by their mass differences. This analyzer offers a very high mass resolution.

Time of flight analyzers use a flight tube where the ions are introduced by accelerating electrical fields. Ions with equal charge will have the same kinetic energy and therefore travel with a velocity in the flight tube dependent on the mass per charge of the ion. The smaller the mass, the higher the speed. The

(22)

travel time in the accurately length defined tube is measured and used for cal- culation of the mass to charge ratio.

The quadrupole mass analyzer uses four parallel rods with hyperbolic, elliptical or circular cross–sections. On the rods a constant electrical field and a radio frequency electric field are applied. The applied fields will influence the ions motion through the rods and ions not transmitted collide with the rods and lose their charge. To scan a mass spectrum different strengths of the fields are used to sort among the ions.

The Fourier–transform cyclotron resonance analyzer traps the ions in a cubic cell with a constant magnetic field and an orbiting motion is generated among the ions by an applied radio frequency field. The ions will generate a faint signal that is detected and the frequency of the signal is related to the mass to charge ratio. This method offers a high mass resolution.

The omegatron resonance analyzer also uses a constant magnetic field to trap the ions in a cubical cell. Perpendicular to the magnetic field a radio frequency electric field is applied. When the radio frequency field is in resonance with the cyclotron oscillating frequency of the ions in the magnetic field the ions motion is forced to follow an Archimedes spiral to the detector. To detect ions with different masses the frequency is scanned.

3.5 Method of ionization

There are many ways to ionize neutral gas, the oldest and most used method is electron impact ionization that is explained in [16]. The sample gas is introduced to an ionization chamber where an orthogonal beam of electrons interacts with the neutral gas. The energy of the ions may be varied to create different efficiency of the ionization. Energy in the range of 50–100 eV is normally used and the higher values lead to a stronger ionization as well as a larger amount of fractionation. Ionization occurs when the electron passes close to or through the atoms or molecules field. In this interaction the molecule or atom turned into an ion can be left in highly exited vibrational and rotational states. When the ion stabilizes fractionation can occur. Therefore the use of to high ionization energy among the electron is not preferable, because they create complex mass spectrum. The ionization process created can be explained by the following relationships from [16].

For positive ions from atoms:

A + e⁻→ Aⁿ⁺+ (n + 1)e⁻ (3.5) For positive ions from molecules (2 alternatives):

AB + e⁻→ AB⁺+ 2e⁻ (3.6) AB + e⁻→ A⁺+ B⁰+ 2e⁻ (3.7) For negative ions:

Y Z + e⁻→ Y Z⁻ (3.8)

3.6 Description of the old prototype system

Greg Miller, at the Southwest Research Institute, built the old system for a proof of principle of an enrichment system for noble gases and water for a proposal

(23)

3.6. DESCRIPTION OF THE OLD PROTOTYPE SYSTEM

to a NASA Mars mission. The enrichment system for the mission was to be combined with a mass spectrometer on a chip developed by SwRI and Northrop Grumman, for detection of elemental and isotopic composition of Martian atmosphere.

This laboratory scale prototype consists of two vacuum or concentration modules, valves and heating system associated with the mass spectrometer for gas analysis. The mass spectrometer used during test was a sophisticated laboratory quadrupole mass spectrometer. A schematic drawing of the gas handling system is shown in Figure 3.1. The circles with a cross represent valves and the lines represent stainless steel tubulation.

The first concentration module works like a bulk scrubber of CO2 and a water concentrator. Inside this module a cylindrical cage filled with zeolite and carbon material works as the scrubber and effectively traps CO2, CO, O2 and water. This module could also be cooled down to the expected surface temperature of Mars. When the sample gas is scrubbed in the first module a valve introduces the gas into the second module, the NEG module. The NEG module consists of a non–evaporable getter of sintered zirconium aluminium powder.

This getter absorbs all other gases like; CO2, CO, O2, N2, H2O and H2, that the first module did not fully trap, but the getter will not trap the noble gases that are of interest for analysis. To be able to evacuate the whole system a turbo pump was connected to the common pump out valve, the port out. When all valves were opened the system could easily be pumped out. The system was also equipped with a Pirani gauge for monitoring the pressure in the prototype system. A drawing illustrating the pre–model is shown in Figure 3.2.

Both of the modules were built of stainless steel and connected to each other by stainless steel tubing and inline metal to metal sealed valves. The tubulation was used for pumping, introduction and transfer of the noble gas to the mass spectrometer. The modules were electro–polished and built as com- pactly as possible to be able to fit in a bakeout envelope. The valves were also protected from contamination from the zeolite by a fine screen welded in front of the valve inside the tubing on the zeolite side. The zeolite module was also constructed so that the sorbent material, zeolite, could easily be exchanged by other sorbent materials through a removable flange. As can be seen in Figure 3.3 below, the approximately 5.5 inch high zeolite module consists of an inner cage structure with elliptical windows covered by a fine mesh. The mesh holds the

Figure 3.1: Schematical drawing of the gas handling system.

(24)

Figure 3.2: Drawing of the gas handling system.

fine–grained zeolite in and keeps it packed around a centred cylindrical heater made of corrosion resistive incoloy. Keeping the zeolite in a cage like this gives the gas an easy way to diffuse in and out of the zeolite material. It will also isolate the hot zeolite from the chamber walls. Figure 3.3 also shows the upper removable flange for exchange of zeolite.

Different types of zeolite, 3A, 4A, 5A and 13X were tested by the prototype where the major difference among the zeolites is the pore size. The ability for all the different types to scrub CO₂and trap water was good, but 5A seemed to show the best desorption and sorption characteristics. The gas mixtures used in the test cycles were a 5, 10, 25 and 50% CO2 mixture with 0.5% water and 0.2 ppm Xe together with air as the balancing gas. When the test was running the measurements from the pressure gauge and the spectrum from the mass spectrometer showed that the prototype system was working, but not fully sat- isfactorily. In the prototype system, it was desired to test solenoid actuated valves, but valves could not be found of the right size. And the decision to use bigger valves caused leaks in the valve seal and tripping of the gauge due

(25)

3.6. DESCRIPTION OF THE OLD PROTOTYPE SYSTEM

Figure 3.3: Drawing and cross–section of the Zeolite module showing its inner construction.

to sticking in the seat material. A modification into manually operated Nupro valves was carried out when leak free solenoid valves could not be found. Dur- ing test a leak of bigger concern was also found in a weld such that only a few test cycles could be performed. This gave a vague proof of the principle of the system.

A comparison of the enriched spectra with a background spectrum was also desired, but due to the mass spectrometer’s poor signal to noise ratio, this could not be done. Therefore the theory needs to be proven with better scientific equipment.

(26)

Chapter 4 Assembly of the new system

4.1 Description and general outline

The new system is based on using the same type of modules for concentration as the old system. But the new system is enlarged to a laboratory model where precision measurements are more important than the size of the equipment.

None of the parts in the laboratory model will be flight hardware but they will the foundation for the developing the real instrument. An advantage with the basic construction and the parts of the system is that it may be used in other experiments in the future. The schematic structure of the system is shown in Figure 4.1. The circles with a cross represent valves and the lines that represent stainless steel tubulation connect the different parts. The gas tank is the

Figure 4.1: The general outline of the Mars atmospheric sampling system. V1- V8 represent valves.

(27)

4.1. DESCRIPTION AND GENERAL OUTLINE

container for the mixed gas and is filled up according to the mixing procedure of the test gas. The whole assembly with the gas tank, the pumping port, the leak valve and the full range CC gauge is connected to the inlet of the Mars atmospheric sampling system at the connection point, between the leak valve and the pressure–regulating valve. When mixing the gas the assembly is instead connected to the gas–mixing manifold. The valve V1 represents the inlet valve to the Zeolite module and during the test the mixed gas in the Mars atmospheric sampling system will flow in the right directions because of the pressure gradient between the chambers. The pressure in the NEG module will always be lower than the pressure in the Zeolite chamber thanks to the non–evaporable getters.

The mass spectrometer will have a lower pressure than both of the modules and therefore lead the gas into the ionization source regardless of which module the gas comes from.

For a better understanding of how the system is constructed a series of figures will follow, where Figure 4.2 shows a drawing of the whole sampling system, mounted on the special made table frame, and the gas mixing equipment in the laboratory environment. Figure 4.3 is a view of the system without the table and the mixing equipment. It only shows the mass spectrometer chamber, the concentration modules, the different pressure gauges, the different valves such as the leak valve and gate valve and the pumping cube with the pump and pump electronics inside of it. Figure 4.4 is a close up of the concentration modules showing also the valves and the support–stands for the tubulation.

Figure 4.2: Drawing of the whole system in the laboratory environment.

The design of the system is based on user friendliness; it should be easy to move around and the work interface of the system is suited for a nice work environment. As can be seen in Figure 4.4 all the knobs on the manual valves between the concentration modules are turned in one direction except for the straight through valve, V5 in Figure 4.1, that has a small angle. Another part of the user friendliness is to use standard components as the tubulation. In case of redesign or failure of parts it should be easy to get new parts in a short time and at a reasonable price. Therefore the system consists of stainless steel

(28)

CHAPTER 4. ASSEMBLY OF THE NEW SYSTEM

Figure 4.3: Free standing system without table frame.

Figure 4.4: Enlargement of concentration chambers and valves.

(29)

4.2. PARTS DESCRIPTION

components with standard connections as ConFlat, Quick–Flange and VCR.

All parts oriented above the tabletop are supported by support stands to prevent leaks and deformation caused by the weight of the parts. The support stands will also prevent twisting of the surrounding tubulation when the valve knobs are turned and a torque is created. Also the vibrations in the tubulation created by the pumping equipment will be reduced. For more detailed information about the systems a part list with vendors can be found in Appendix A, and the drawings for the different parts can be reviewed in Appendix B. The drawings are sorted by drawing number and show for example basic construction and dimensions.

4.2 Parts description

4.2.1 Zeolite module

This module is the first concentration module introduced to sample gas and it will mainly scrub CO2and water, but also CO and O2. The design of the module is fairly similar to the basic design of the module in the old prototype system.

There are changes in the choice of the surrounding components, dimensions, materials, window constructions and the type of heater.

In Figure 4.5 and Figure 4.6 a cross section and an exploded view of the module are shown. The outer surrounding component is a reducing cross with ConFlat flanges of tubing size 2.75 to 1.33 inches. In the 0.75 inch diameter tubing a little snap ring with a mesh is placed on both sides of the main chamber to prevent dust and particles of zeolite from travelling out in the system and contaminating valves and other parts. The cylindrical cage holding the zeolite is custom–made in macor, a hard and heat–resistive ceramic material. The cage consists of a cylindrical housing with windows formed as long slots, a retaining ring and a housing cap. The inside of the cylindrical housing is covered with a fine stainless steel mesh of size 120 × 120 lines per inch. The mesh is rolled in two laps and springs out by its own force to cover the window slots. Along the edge of the mesh a seal will be created by self–weld of the material in the high temperature during heating of the module. The large size of the window slots compared to the size of the cylindrical housing is to create a good flow of gas into the zeolite during process. The mesh is also rolled over both edges on the cylindrical housing to create a seal against the flange and the retaining ring.

The fitting between the retaining ring and the cylindrical housing as well as the fitting for the cylindrical housing into the grove of the flange needs to be so tight that the assembly stays in position when assembled. Two threaded titanium rods screwed into the flange clamp the housing cap to the retaining ring by nuts tightened from the outside of the cap. By this construction the zeolite can easily be replaced by removing the flange with the cage construction from the module and remove the housing cap. The mesh was tested before mounting to investigate if it could hold the smallest grain size of zeolite, 100/120 and the result was satisfying.

The heater used for heating up the zeolite during the test procedure is a 6 inch long cartridge heater of diameter 3/8 inch and the sheet material for the heater is inconel that is more resistive against corrosion than stainless steel. The heater is rated for 240 V and 575 W and has an built–in thermocouples of J

(30)

Figure 4.5: Cross section of the Zeolite module.

Figure 4.6: Exploded view of the Zeolite module.

(31)

type. With the thermocouple the temperature from the inside of the zeolite cage can be monitored and adjusted to find the right temperature in the different phases of the test procedure. When attaching the heater to the flange to create a welded vacuum seal the two first attempts failed caused by a leak in the weld.

A third solution consisted of a sleeve covering the part of the heater penetrating the flange. The sleeve was silver soldered on the inside of the module between the heater and the sleeve and welded on the outside of the module between the sleeve and the flange. This solution was leak tested without satisfying results and the discovery that the heater itself was leaking was done. The leak was from the inside of the heater through a seam into the system. Under a microscope small cracks could be seen on the surface material of the heater. The last solution was to design and build a flange with a stainless steel tub welded to the flange with the same inside diameter as the outside diameter on the heater.

The tube is bored out from a solid piece so that one end is capped off without a weld and the other end of the tube is welded to the flange. The welded is on both sides of the flange to prevent virtually leaks. By sliding the heater into the tub a very tight fitting is created. This will prevent the heater from having contact with the vacuum side of the system and the leaks should be minimized.

The problems that might arise from this solution are bad conduction between the heater and the stainless steel tubing and corrosion on the stainless steel in contact with the zeolite during process.

A more detailed drawing of the Zeolite module’s different parts are shown in Appendix B in drawing No 090-0024 and No 090-0021. The macore pieces with the stainless steel mesh and the titanium rods before assembly of the module are shown in Figure 4.1.

Figure 4.7: Digital photo of the parts for the Zeolite chamber.

4.2.2 NEG module

This module is the second concentration module that will remove undesirable gases from the sample gas in the zeolite module and improve the concentration before analysis. It consists of two St 171 getters mounted on an electrical feed- through rated to 2000 V and 30 A.

The mounting assembly to hold the getters in position is made of titanium and can be seen in Figure 4.8 and Figure 4.9 showing the module in a cross

(32)

section and in a 3D exploded view for better understanding of the assembly.

The reason for using titanium as a material for the mounting of the getters is its high melting temperature of 1948 K given in [17] as well as its low thermal conductivity compared to other materials suitable for the task. The titanium pieces where designed and custom made to hold the getters in the middle of the surrounding reducing tee of stainless steel and to keep the assembly from touching the inner walls of the chamber. The reducing tee has a flange diameter of 2.75 inches and an outer tubing diameter of 1.5 inches, where the getters are assembled. Due to the high temperature of the getters during activation, conduction between the assembly and the chamber is not desired. In order not to trap gas and create virtual leaks, the titanium mounting plate and cones are provided with small grooves and drilled holes.

Figure 4.8: Cross section of the NEG module.

The threaded rods mounted to the feed–through as well as the rod in the middle holding the getters to the mounting assembly have flat ends for the same reason. The components can be viewed in drawing No 090-0004 in Appendix B.

The getters are mounted with their electrical pins pointing away from each other and rotated around the center point so the electrical pins are not aligned.

This will make the electrical connections between them and to the feed–through much easier. To achieve the same temperature on the getters and to avoid possible deformations and tensions by temperature differentials, serial connection of the getters is carried out to obtain the same current. The material connecting the getters, and the getters to the feed–through are nickel ribbons welded to the copper lead wires on the feed–through and the lead wires on the getters. The temperature will be monitored with an optical pyrometer through a sapphire view port, placed in one end of the tee, where the color of the heated getters can be used to infer the temperature. Figure 4.10 shows the getter assembly after soldering and before it is placed in the camber. The sapphire window also gives an ability to see if the assembly will deform during activation.

(33)

Figure 4.9: Exploded view of the NEG module.

Figure 4.10: Getter assembly before mounting in NEG module.

(34)

4.2.3 Quadrupole mass spectrometer

As explained earlier, a mass spectrometer consists of 4 main components; the same applies to the quadrupole mass spectrometer in use for the system. These parts are described here and additional information can be found in [18] and [19]. A schematic sketch is showed in Figure 4.11 and in Table 4.1 the different components names are displayed.

Figure 4.11: Schematic view of the quadrupole mass spectrometer system.

Table 4.1: The components of the quadrupole mass spectrometer system.

Description Part name from Balzers

Analyzer QMA 410

Second electron multiplier SEM 218

RF generator QMH 410-3

Electrometer amplifier EP 422 Ion counter preamplifier CP 400 Ion source supply IS 420

Controller QC 422

High voltage supply HV 421

The crossbeam ion source has two tungsten filaments with a 10,000 h predicted lifetime that they thermally emit electrons. The electrons are accelerated by an electric field and focused to a narrow beam into the ionization volume by an electron collimation magnet. An arrangement of plates at different electric potentials focuses the created ions to the inlet of the analyzer. The analyzer consists of four molybdenum rods with the diameter 16 mm supplied with a RF and a DC component from the additional RF source. This quadruople mass spectrometer configuration has two different detectors: a secondary electron multiplier and a faraday cup that can be found more information about in [20].

The secondary electron multiplier, SEM, is arranged 90^◦off axis to prevent fast or excited neutrals or photons from striking it. An ion counter preamplifier is attached to the SEM that will result in measurement of ion counts per second

(35)

over the scanned masses. The faraday cup is not as sensitive as the SEM and therefore has an electrometer amplifier connected for detection of the weak sig- nals. The faraday gives a measurement of the ion current for the scanned mass.

The mass spectrometer is connected to a controller unit consisting of a high voltage supply for the SEM, a voltage supply for the ion source and a controller for the quadrupole mass spectrometer. The unit is built for additional elements to be installed if necessary. This controller unit is connected to a computer where the software Quadstar 422 is the interface for the mass spectrometer.

The custom–made chamber for the mass spectrometer can best be viewed in Figure 4.4.

4.2.4 Pumping system

The purpose of the pump is to remove gas from the vacuum system, to main- tain the desirable pressure. This pressure depends on the inlet of gas to the system and the pumping of gas from it. In this system a turbomolecular pump, TMU 261, is used that is backed up by a diaphragm pump, MVP 035-2. The diaphragm pump creates a rough vacuum and the trubomolecular pump can bring the system down to ultra high vacuum in the range of < 1 × 10⁻⁸ mbar as can be seen in Table 4.2 showing data for the pumps from [21]. It is explained in [22] that the pumping speed of the pumped chamber will never be greater than the conductance between the chamber and the pump even though the pump has a high pump speed. It is generally known from the same source that pumps designed for the highest pump speed have the lowest compression ratio and therefore a compromise has to be made to create the best pump.

The turbomolecular pump with a single flow condition consists of a standing rotating cylinder with stator and rotor vanes on it. The gap between the rotating cylinder and the outer wall is very narrow and the cylinder rotates with a high speed. When molecules or atoms hit the vanes they will be dragged over the vane surface to the other side. Due to the angle of the vanes it is more likely that the particles moves from the chamber into the pump than the other way around. But it also depends on the pressure difference over the vanes. Tur- bomolecular pumps often have a lubricated bearing that the cylinder sits on.

While the pump is running it is not likely that the lubricant will diffuse into the vacuum system. Therefore the pump should always be running and in case of a failure the pump has to be valved off and filled with air or pure nitrogen to prevent water to enter the system and the pump. The high speed of the cylinder forces the pump to be cooled by water with the temperature of 5 – 25^◦C.

This pump is very sensitive and can be destroyed by small particles and debris that gets into it, therefore a splinter shield is mounted over the gas inlet flange.

Also protect the pump against unnecessary forces and always mount it to the vacuum system before running.

In our system we have the ability to pump out different parts of the system as well as the whole system at once by opening and closing the valves in our configuration. During bakeout you wish to pump out the whole system to pump out all the water that is stuck to the surface. When removing water and CO2

from the Zeolite module or activating the NEG module the modules need to be pumped individually.

(36)

Table 4.2: Technical data for turbomolecular pump TPU 261 and diaphragm pump MPV 032-2.

TPU 261 MPV 032-2

Pump speed 210 l/s

Rotation speed 1800 rpm

Final pressure < 1 × 10⁻⁸ mbar 4.0 mbar

Max temperature 90^◦C 40^◦C

Final pressure backing pump < 5 mbar

Nominal volume flow rate 2.4 m³/h

Compression rate 1.3 × 10⁴− > 1 × 10⁹

4.2.5 Pressure measurement devices

For measurements of the total pressure and the partial pressure in the system different types of gauges have to be used. There are no gauges that can cover the whole range of pressure from atmospheres to ultra high vacuum, 10⁻¹² mbar, according to [23]. Therefore the pressure regions of interest also affect the choice of gauge. The system is equipped with a compact capacitance diaphragm gauge and a compact cold cathode gauge, inverted magnetron, with a Pirani gauge built in for a greater measurement range of the total pressure. For partial pressure measurements the quadrupole mass spectrometer is used as well.

The compact capacitance diaphragm gauge in use has the same basic work principles as the capacitance manometer and its pressure range is 0.01 – 110 mbar. According to [24] this instrument can electrically sense the change in capacitance between an electrode and the diaphragm as the diaphragm deflects under forces due to the pressure differential across it. The instrument has two capacitance electrodes on a ceramic substrate; one round in the center and an annular shaped one around it. These electrodes are placed on the reference side that is under a specific vacuum contained by a getter so that they will be protected from contamination. The diaphragm is made of inconel and is placed near the electrodes creating the wall between the reference side and the measurement side. When the pressures are different between the two sides there will be a difference in charge capacitance between the electrodes due to the deflection of the diaphragm that can be measured. Figure 4.12 shows a diaphragm gauge.

The full range compact cold cathode gauge has two gauges in the same instrument: the cold cathode, inverted magnetron, and the Pirani. This full range instrument can measure in the range 1000 – 5×10⁻⁹mbar where the Pirani is used for the higher pressures and the cold cathode for the lower pressures.

The Pirani gauge is a thermal conductivity gauge that is electrically operated, and the pressure is measured due to the gas ability to conduct heat. The gauge consists of a wire filament usually made of tungsten that is exposed to the pressure and heated by current running through it. When molecules and atoms collide with the filament heat is transferred from it and when the pressure settles at a level the filament will also have a constant temperature. The filament is part of a Wheatstone bridge that will measure the current in the filament keeping the voltage in the bridge constant or the temperature on the filament constant. This current will be proportional to the gas density and therefore

(37)

Figure 4.12: A diaphragm gauge.

also the pressure. The cold cathode gauge has a strong magnet around the discharge tube creating a magnetic field along the cylindrical anode, which has a radial electric field with the cathode surrounding it.The magnetic field forces the electrons to longer path lengths, so that there is a greater chance to ionize the molecules and measure the ion current in the anode. The ions will travel to the cathode and pick up electrons and become neutral molecules or atoms. A stable discharge will build up where the amount of ionization is as big as the loss of ions to neutral particles. The current in the discharge tube is proportional to the pressure. All three gauges above are temperature sensitive and their work principles are further explained in [23], [24] and [25].

For partial pressure measurements, PPM, a residual gas analyzer, RGA, can be used. They are mainly mass spectrometers with a magnetic sensor filter or a quadrupole mass filter. The later is the most common and the one that is used in the system. These mass spectrometers usually have high sensitivity and lower resolution. The description of the mass spectrometer can be found in section 4.2.3 on page 23.

4.2.6 Pressure regulator

The infinity digital vacuum regulator consisting of the infinity computer regulator and the pressure proportioning valve is illustrated in Figure 4.1 as the pressure valve and a controller, connected by a feedback loop. This pressure regulator has the working range of 0 – 1000 mbar and is best calibrated in the region 1 – 10 mbar. The controller allows the regulation of the vacuum in the apparatus volume to either a static user entered value or according to a user entered pressure ramp. The infinity computer regulator measures the absolute pressure in the apparatus volume through the feedback loop with a gauge and displays the pressure in mbar. The pressure measuring device in the controller is very sensitive and needs to be protected from contamination. The pressure proportioning valve is dimensioned for an apparatus volume of 1 ml to 2 l and has a stainless steel flow path. It is important that the valve is mounted in the right orientation to not damage it. The valve is a part the inlet to the Mars

(38)

Figure 4.13: The inlet system. The pressure proportioning valve to the left and a Nupro valve to the right.

Atmospheric Sampling System and can be viewed in Figure 4.13 mounted on a support stand. The hose hanging down in the middle of the figure represents the feedback loop that goes to the controller.

4.3 Valves

Valves are important in vacuum systems and they have a lot of different functions that will help in many ways. According to [22], the ultra high vacuum valves have all the same criteria to fulfill; a minimum leak rate when closed, a maximum conductance when opened and not be a gas contamination source in itself for the total system. By using valves, different parts of the system can be isolated and parts can be added or removed without interfering with the vacuum in the whole system. This will save time at pump down and prevent exposure to air and water vapor for the rest of the system. The valves in this system are mainly used for easy manual control over the experimental phase when the tests are performed. All these valves can also be used for isolating and finding leaks quicker than trying to leak check the whole system at once. The valves that are used at different parts in the system are chosen because of their functions, temperature range at bakeout and leak rate.

4.3.1 All–metal gate valve

As can be seen in Figure 4.3 on page 17 an all–metal gate valve is placed between the turbo pump and the mass spectrometer chamber. The reason for this is to have the same diameter as the turbo pump all the way up to the chamber and still be able to close off the mass spectrometer chamber easily. A free standing all–metal gate valve is shown in Figure 4.14. The gate valve has a pneumatic actuator with a solenoid and an air compressor is used for its opening and closing function. Between the air compressor and the actuator an air dryer and an after filter are placed to keep the air into the gate valve actuator dry and clean. The

(39)

4.3. VALVES

solenoid is operated by voltage and a power switch controls the opening and closing of the valve by letting compressed air in and out of the actuator. The gate valve has a sealing function that is based on a plate that is pushed into the opening by the piston that is moved by the compressed air. When the plate is in position in the valve opening it springs out on both sides but seals properly only one side of the valve. Table 4.3 shows data for the all–metal gate valve used in our system and Figure 4.15 shows the inside of the gate valve and its closing function.

Figure 4.14: All–metal gate valve.

Figure 4.15: Inside of a gate valve and its closing function.

4.3.2 Right angle valves and straight through valve

This group of valves consists of all–metal valves that have a copper seat for seal.

In Figure 4.1 the valves are marked from V2 – V8. These valves are from two different vendors and have different temperature regions. Valves V2 – V6 are qualified ultra high vacuum bakeable valves, right angled and straight through

(40)

Table 4.3: Data for the all–metal gate valve from [26].

Description Data

Inner diameter 4 inch or 100 mm

Leak rate (body, valve seat) < 1 × 10⁻¹⁰ mbar l/s Bake out temperature for valve (opened and closed) 300 ^◦C

Bake out temperature for pneumatic actuator 80^◦C

Power for solenoid 80^◦C

Heating and cooling rate 80^◦C/h

Cycles until first service 20 000

Compressed air pressure min and max overpressure 4 – 8 bar

where V5 is the straight through valve. The valves V7 and V8 are all–metal right angle valves that can not be baked to the same temperature as previous valves.

The sealing mechanism is the same for the two valve types. The sealing mechanism is placed inside a welded bellow that separates it from the vacuum system. The seal itself is a knife–edge that is pressed down by the drive mechanism in the bellow into a copper seat, the knife edge is usually a part of the valve body. This motion can be repeated several times if the knife edge hits the same spot every time and if the valve is closed with the same force using a torque wrench. When baking these valves the copper seat might be deformed and the motion can not be repeated as many times as the unbaked before the seat needs to be changed. When the valves are baked, the drive mechanism inside the bellow needs to be lubricated to prevent galling. In a straight through valve the knife–edge comes in from an angle and cuts the conductance in the tube when closing. The both configurations are shown in Figure 4.16 and Figure 4.17.

The main difference between the valve types is that the ultra high bakeable vacuum valve can be baked in an opened and closed state to 450 ^◦C and the ordinary all–metal right angle valve can only be baked up to 200^◦C in opened position. More information is given in [27] and [28]. The valves all have the same leak rate of 2 × 10⁻¹⁰ std.cm²/s and the dimensions on the flanges are varied from ConFlat flange 1.33 inch to ConFlat flange 2.75 inch depending on the position in the system.

4.3.3 Leak valve

The leak valves used in the system are all metal regulating valves that can be baked up to 450^◦C in opened position. The two leak valves UDV 136 and LVM used are connected to the mass spectrometer chamber and the gas mixing tank as can be seen in Figure 4.1. The leak valve on the mass spectrometer, UDV 136, will regulate the flow into the mass spectrometer at the analysis of the test gas. This leak valve also has an electrical control unit that will ease the sensitive regulation motion. The other leak valve is manual and connected to the mixing tank and regulates the in– and outflow of the mixed gas. The leak valves consist of a copper alloy seat and a sapphire plate forced together to make the seal inside the valve body. The UDV 136 valve has a leakrate at 1 × 10⁻¹¹