Designing a LabVIEW Interface for Temperature and Pressure

Sensors for the ATTA Experiment

Ashleigh Lonis

August 6, 2012

Contents

1 Introduction 1

1.1 Evidence for Dark Matter . . . . . . 1

1.2 What is Dark Matter? . . . . . . . . 2

1.3 Detection of Dark Matter . . . . . . 2

1.4 The XENON Dark MatterExperiment . . . . . . . . . . . . . . 3

1.5 85Kr Contamination . . . . . . . . . 4

2 ATTA 4

2.1 ATTA Overview . . . . . . . . . . . 4

2.2 Lasers . . . . . . . . . . . . . . . . . 5

2.3 Metastable and OpticalMolasses . . . . . . . . . . . . . . . . 6

2.4 Zeeman Slower . . . . . . . . . . . . 6

2.5 Magento-Optical Trap . . . . . . . . 6

2.6 Detection . . . . . . . . . . . . . . . 7

2.7 The Current Status of ATTA . . . . 7

3 What I’ve Done 8

3.1 Overview . . . . . . . . . . . . . . . 8

3.2 Omega Micromega CN77000 . . . . . 8

3.3 Omega CN7700 . . . . . . . . . . . . 8

3.4 Pfeiffer MaxiGauge . . . . . . . . . . 8

3.5 Red Lion PAX . . . . . . . . . . . . 9

4 Conclusion 9

5 Acknowledgments 10

Abstract

1 Introduction

1.1 Evidence for Dark Matter

Dark Matter accounts for nearly 80% of the mat-ter of the universe and yet scientists do not knowwhat it is. Dark Matter has been theorized to ex-ist due to the rotation curves of galaxies, rotationvelocities in galaxy clusters and gravitational lens-ing. Fritz Zwicky in 1933 was observing the Comagalaxy cluster and used the virial theorem to in-fer the amount of matter within the cluster basedon the velocities of the galaxies in the cluster. Henoticed that his virial theorem calculations did notmatch the mass inferred based on luminous mat-ter and that there was actually 400 times the massthan was visible.[1] The virial theorem requires thatthe kinetic energy is equal to half the gravitationalpotential energy you can solve for the total mass ofthe system.[2] Using the measured average veloci-ties of the galaxies in the system and the radius ofthe cluster, you can calculate the total mass in thesystem.

Mtot ' 2Rtotv

2

G

Where M is the total mass of the cluster, R is theradius of the cluster, V is the average velocity of thegalaxies through the cluster, and G is the gravita-tional constant.[1],[2] In 1932 Jan Oort found thatthere was also an issue of unseen matter when ob-serving the velocities of stars in galactic disks. Butit wasn’t until the 1970s when Vera Rubin and KentFord published the data that they collected whenmeasuring the velocity curves of hundreds of galax-ies that the majority of the scientific community

1

accepted that there was dark matter.[1]Strong gravitational lensing from galaxy clusters

and large elliptical galaxies has provided further ev-idence for the existence of dark matter. Because ofGeneral Realtivity we know that mass bends space-time.[3] This effect can cause light from a distantgalaxy to appear in a different position(s) on thesky, than it would be if we had a direct line ofsight. Based on the lensing effect the mass of thegalaxy cluster can be inferred. [3]

1.2 What is Dark Matter?

There have been several theories for what darkmatter is. Since the matter is not luminous,one theory is that it could be dark baryonicmatter in extremely large numbers. These hasbeen nicknamed MAssive Compact Halo Objects(MACHOs).[4] These would include non-accretingblack holes, brown dwarfs, and planets. Surveysof the Milky Way galaxy have been searching forgraviational microlensing from the MACHOs onbackground stars in the galactic center and inthe dwarf galaxies that orbit the Milky Way.[5]However, evidence has not been found for theamount of dark baryonic matter needed to accountfor all of the discrepancy between luminous anddark matter.[5]

The most compelling prediction for the major-ity of dark matter is Weakly Interacting MassiveParticles (WIMPs). WIMPs are unknown par-ticles that have not yet been discovered by ex-periment but would interact via the weak andgravitational forces. To remain bound to thegalaxy they would also need to move at non-relativistic velocities.[6] The leading thought isthat these WIMPs are Supersymmetric (SUSY)particles. The most light weight Supersymmet-ric neutralino (neutrally charged SUSY particle)is thought to be a stable particle and could bewhat makes up the majority of dark matter in theuniverse.[6]

1.3 Detection of Dark Matter

There have been many attempts to detect darkmatter via direct and indirect detection. The FermiLarge Area Array Telescope can be used as an in-direct dark matter detector. It collects gamma ray

Figure 1: Image of the Coma Cluster - Fritz Zwickytheorized about the missing matter in the cluster andcoined the term “dark matter.”[17]

Figure 2: Based on the visible matter observed inspiral galaxies, the predicted velocity as a function ofradius (A) is not the same as the measured velocity (B).(Shown with arbitrary units)

Figure 3: Image of strong gravitational lensing ingalaxy cluster Abell 2218 [18]

2

Figure 4: Artist depiction of the Fermi SpaceTelescope.[19]

data from throughout the universe. The small-est neutralino is thought to be a Majorana par-ticle meaning that it is its own antiparticle. Whentwo neutralinos collide, they would produce eithera gamma ray or a particle anti-particle pair.[6] Thisindirect detection technique is extremely useful andcritical to observations of dark matter and the dis-tributions of dark matter throughout the universe.However, it is difficult to distinguish dark mattergamma rays from other gamma ray sources [6].

Direct dectection techniques typically depend onWIMP-nucleus interactions. These detectors tendto operate in deep underground laboratories toshield the detectors from cosmic rays.[6] Direct de-tection methods involve nuclear recoil when a darkmatter particle “knocks” into it.[6] This can be de-tected by the ionization of the atom (charge), theheat created by the interaction or the scintillationlight that the atom gives off.[7] Most detectors arelooking for two of the three signals.[7]

1.4 The XENON Dark MatterExperiment

The XENON Dark Matter Experiment uses directdetection methods to find evidence of dark matter.XENON is attempting to detect ionization andscintillation from WIMP-nucleus interactions. Todetect dark matter the XENON experiment isusing liquid and gaseous highly purified Xe.[7],[8]For this direct detection technique the use of noblegases is preferred since it is relatively easy to filterout other elements. The use of Xe specifically ispreferable because it has a larger nucleus in order

Figure 5: The XENON100 experiment showing theprotective layers and detection system.[20]

for the WIMP to interact with and gives greaterscintillation and ionization signal than the lighternoble gases.[9] The process through which darkmatter is hypothesized to be detected is that adark matter particle will hit a Xe nucleus and thenucleus will recoil giving off scintillation light andionizing the atom. XENON expects to see twosignals when a WIMP hits an Xe nucleus. Thefirst signal (S1) is the scintillation photons thatare picked up by photo multiplier tubes (PMTs)along the bottom of the detector.[8] The secondsignal (S2) comes from the electrons emittedduring collision.[8] There is an electric field thanruns from top to bottom of the detector. Theelectron will drift toward the top of the detectorand secondary scintillation light will be pickedup by the PMTs at the top of the detector.[8]The ratio of S1/S2 can give further confirmationthat this was a WIMP-nucleus interaction and notan electronic interaction due to beta particle orgamma rays. For a nuclear recoil, S2 should beless than S1.[8] For electron recoil, it is expectedthat S2 is much greater than S1.[8]

The XENON10 experiment was the prototypefor the XENON experiment. It used 15kg of Xeto try to detect dark matter and also to test thetheory behind the experiment so that it can bepreceded by larger scale detectors.[10] The Xe iskept in a time projection chamber that will givean idea behind the projected trajectories of thedark matter particle. The current experiment,XENON100 uses approximately 160 kg of Xe, but

3

Figure 6: Graph of sensitivity of dark matter experi-ments. This shows that the 1T experiment will be ableto probe the smallest interaction cross-section of all ex-periments to date (pink dashed line). [20]

has so far been unsuccessful in conclusive detectionof dark matter.[11] The next step for the XENONexperiment is XENON1T in which they willuse approximately 2400 kg of Xe.[12] With eachgeneration of XENON experiment comes moreprecision and more limitations to the interactioncross-section of the dark matter particles.

Due to the relatively few interactions expectedbetween WIMPs and Xe nuclei, it is necessary toshield the detector from as much background signalas possible.[8] The XENON experiment takes placeunderground at Gran Sasso National Laboratoryin Italy to help protect the detector from cosmicray signals.[9] The experiment must also use equip-ment with stable components that will give off verylittle signal that will affect the experiment.[8] TheXe used must also be highly filtered. CommercialXe gas is obtained from the atmosphere, but hasa few ppm Kr contamination.[11] Kr is also aninert noble gas and is difficult to filter from Xe.The XENON100 experiment started with ∼100ppt Kr contamination and is currently using ∼10ppt Kr contaminated Xe. For the XENON1Texperiment it is necessary to have Xe with ppt Krcontamination to achieve the sensitivity goals.[11]

Figure 7: 85Kr goes through beta decay when a neu-tron decays into a proton, electron, and anti-neutrino.It turns into 85Rb which can the be filtered from thesystem.[21]

1.5 85Kr Contamination

One particular isotope (85Kr) goes through beta de-cay with a half life of ∼10.7 years.[11],[16] 85Kr hashad an increased abundance in the atmosphere overthe past several decades due to nuclear weaponstesting, nuclear reactor catastrophes and mishan-dled or reused nuclear waste.[14]In order to have an understanding of the level of

Kr contamination within the Xe, the ATTA (AtomTrap Trace Analysis) experiment is working to de-tect stable 84Kr atoms on the individual level. 85Krhas a known ratio (85Kr/Kr ' 2 x 10−11) thatthe XENON group can use to infer the amountof events they would expect to detect due to 85Krbeta decay.[11],[16] 84Kr is the most common iso-tope with approximately 57% abundance.[11] For a2400 kg sample of Xe with ppt Kr contaminationthere should be ∼200,000 total 85Kr atoms. Dueto the extremely small number of 85Kr atoms, it ismost feasible to detect 84Kr and infer the amountof 85Kr.

2 ATTA

2.1 ATTA Overview

The ATTA method was developed at Argonne Na-tional Labs for use in detecting trace amounts ofspecific isotopes found in ground water, the oceans

4

Figure 8: Picture of Columbia’s ATTA apparatus.

Figure 9: Overview of ATTA method. From left toright, the atoms are excited, collimated, slowed, andtrapped.[16]

and the atmosphere for radioactive dating.[11],[16]ATTA employs several types of methods to slowdown the 84Kr atoms so that they can be detected.The current testing for the Columbia experimentuses Ar instead of Kr to avoid contamination of theequipment prior to its use with the Xe samples forthe XENON Dark Matter experiment.[11] Contam-ination can come from the atoms embedding intothe sides of the apparatus and they can later beknocked free from the walls and give an incorrectcontamination level for the Xe being tested. 40Ar(the most common Ar isotope with approximately99.6% of all Ar) is also a convenient substitute touse because it is also a noble gas and it has transi-tion from a metastable state that requires close tothe same amount of energy as the transition for Kr,although the system is optimized for Kr.[13],[16]

Figure 10: Transition from the ground state to theMetastable state for 40Ar and 84Kr and the similar ex-citation energy.[16]

2.2 Lasers

The laser is tuned to an extremely narrow band-width to provide selectivity between isotopes.[11]By tuning the laser to within one or two linewidthsof the resonance frequency of the specific isotopeit ensures that only 84Kr is captured.[14] Thelaser is red detuned in the lab reference frame toaccount for the doppler shift from the atomic beammoving against the laser beam at ∼250 m/s.[16]The atoms will absorb photons from the laserand jump to an excited state which will decreasethe velocity of the atom along the axis of thelaser(see Figure 10). The atoms then de-excite viaspontaneous emission which has no preferentialdirection.[14] This effect performed many timescreates an overall net deceleration of the atomalong the axis of the system. For Kr specifically,∼5 x 104 excitations and spontaneous emissions arerequired to slow the atom so it can be captured.[14]

There is a problem with the deceleration, how-ever, and that is the changing doppler shift as theatoms slow. Once the atoms slow by ∼4.2 m/s(corresponding to 700 absorption-emission cycles)the laser falls out of resonance with the atoms.[14]This means that on its own the laser will no longercool the atomic beam. To account for this ATTAutilizes a Zeeman slower to keep the atomic transi-tion in resonance with the laser (I will discuss thisfurther below).[14]

5

Figure 11: Plasma beam of excited Ar atoms.[11]

2.3 Metastable and OpticalMolasses

ATTA uses a 120 MHz radio frequency elec-tronic discharge to excite the Kr atoms to a state5s[3/2]2 that has a lower probability, accordingto quantum selection rules, to re-transition to theground state (called a metastable state).[14],[16]

Figure 12: Optical molasses col-limates the atomic beam.

The atoms thentravel throughseveral perpen-dicular laserbeams to col-limate the Kratoms to mini-mize loss in thesystem. Thelasers are tunedto 811.5 nmin the infraredpart of thespectrum.[16]These perpendicular lasers slow the atomic beamsand collimate them in a process called opticalmolasses. This method slows any transversemotion and does not affect motion along the axiswhich increases the trap efficiency. The transversecooling stages in the apparatus uses approximatelya 1 cm in diameter beam that is then bouncedoff of mirrors to give additional cooling in the

Figure 13: When in a magnetic field the transitionlevels are split into three distinct lines. [22]

transverse direction.

2.4 Zeeman Slower

As discussed above, the Zeeman slower is usedto keep the laser in resonance with the atomictransition so that further cooling can be accom-plished. The Zeeman slower is a hand-wrappedgradient solenoid that works by having a changingmagnetic field along the axis of the apparatus.It keeps the laser in resonance with the atomictransition level by the Zeeman effect in which amagnetic field creates a splitting of the spectralline with higher and lower transition frequencies(see Figure 13).[14] This splitting is caused bythe magnetic field’s affect on the atomic dipolemoment which corresponds to different angularmomentum quantum states that are degeneratewhen not affected by a magnetic field.[15] Byhaving a very precise gradient magnetic field thelaser can stay in resonance with the atoms and theeffect is additional slowing of the atoms. ColumbiaUniversity’s ATTA group slows the atoms from∼250 m/s to 10 m/s so that the atoms can betrapped in a magneto optical trap (MOT).[16]

2.5 Magento-Optical Trap

6

Figure 14: The MOT traps the atomsusing gradient magnetic field and 3Doptical molasses.[14]

TheMagneto-OpticalTrap(MOT)uses anti-Helmholtzcoils tocreate amagneticfield withzero mag-netic field inthe centerand a magnetic field gradient increasing outwardfrom the center. The anti-Helmholtz coils use Cucoils that have opposing directional current toinduce a magnetic field that points in the directionout of the axes of the trap (see Figure 14). It alsouses 3D optical molasses. 3 perpendicular laserbeams are used and reflected back so that they canoppose any atomic motion away from the centerof the trap.[14] Instead of the effect from the 2Doptical molasses, the atoms now feel a decelerationin any direction heading away from the centralregion not only due to the lasers but also dueto the magnetic field gradient. The combinationof both the magnetic field and the 3D opticalmolasses traps the atoms. The 3D optical molassescauses the atoms to fluoresce so that they can bedetected.

2.6 Detection

Figure 15: Lenses.[16]

A single trappedatom in the MOTwill give off ∼107

photons/second.[16]ATTA observes a6% solid angle ofthe MOT and ∼104

photons/second canbe observed. To trapa large number ofatoms, a relativelyinexpensive charged-coupled device (ccd) can be used to detect thelarge amount of light coming from the trappedatoms. For single atoms trapped in a MOT youneed a more sensitive detector. The ATTA group

Figure 16: Image of ∼109 trapped Ar atoms.[16]

uses an avalanche photodiode (APD) for sensitivedetection of the fluorescent photons. An APDconverts observed light into electricity using thephotoelectric effect.

The fluorescent light produced by the trappedatoms is the same wavelength that is used for theatomic transition to both slow and trap the atomsin the near infrared. This means that the laserlight present in the trap can be a potential prob-lem for observing the light from the trapped atoms.To reduce this background light and to focus theobserved light, a lens system is used with an ad-justable iris (spatial filter) to avoid excessive back-ground photons.[14],[16] A bandpass filter is alsoused before the light enters the APD to filter lightof other frequencies.

2.7 The Current Status of ATTA

ATTA has shown that they can successfully trap Aratoms in the MOT. With a system loading rate of1017 atoms per second, the MOT can successfullytrap 109 atoms per second in the trap.[16] Thisgives a total system efficiency of 10−8.[16] ATTAis working on accomplishing single atom detectionand working to ensure the background signal is lowenough and the system is optimized to successfullydetect a single atom. Once single atom detectionis accomplished they will test Xe samples from theXENON100 and XENON1T experiments.

7

Figure 17: Graph of my measurements of the voltageoutput of the Micromega temperature controller. Redrepresents the trend line assuming a linear correlation.My measurements are shown in blue.

3 What I’ve Done

3.1 Overview

Columbia’s ATTA group needs to have a data log-ging system set up to record and create a betterinterface from the sensor readouts. The main sen-sors that they use are temperature and pressuresensors. The temperature sensor is used to mea-sure the temperature of the RF discharge. Thepressure sensors measure the vacuum level at thesource and throughout the system to monitor theamount of inflowing gas, the efficiency of atomiccapture, and to ensure there are no issues with thevacuum system.

3.2 Omega Micromega CN77000

I started by working with the Omega brand Mi-cromega CN77000 series PID temperature con-troller. The group is currently not utilizing thePID functionality but is using the controller as atemperature sensor. The controller is set up witha current or voltage analog output signal. Voltageoutput is a better option to use the data acquisi-tion functionality of LabVIEW. This can be pro-grammed to have set points for the temperatureregion of interest. For the ATTA group this is be-tween -190 and 100 degrees Celsius. I set up thecontroller to have a 0V output at -190 and 10V at100 and took some test measurements of the out-put voltage to confirm the functionality of the sys-tem. Unfortunately the measurements I took did

not correspond with the output voltage I expected.The first measurement I took corresponded to atemperature of -133. If you assume a linear cor-relation between the temperature and the outputvoltage, you would expect about 2 V. I read a mea-surement of .507 V. I decided to take a few moremeasurements at different temperatures to see ifthe scale was not linear. I took measurements from-136 to -94 C and saw a voltage of around .5 V forall temperatures(see Figure 17). I did some trou-bleshooting with Omega technical support and theycould not determine that anything I was doing waswrong, so they had me work with Newport Elec-tronics a subset of Omega that had manufacturedthe product. Their technical support and engineer-ing department determined that the analog outputfunction was non-functional.

3.3 Omega CN7700

As an alternative, the ATTA group had 4 OmegaCN7700 temperature controllers that had beenused when baking the MOT chamber to get theultrahigh vacuum needed to increase the MOT ef-ficiency. These controllers had been wired so thatthey could produce current to the heating tapes.I tested the output on these controllers and foundthat there was also an issue. I found an outputvoltage of around 50 V with a non-sinusoidal sig-nal around 1 kHz. This was quite surprising, sinceit should have been a DC signal with around 10Vmaximum output. I contacted Omega technicalsupport to ensure that I had the output set upcorrectly, and again confirmed that these sensorswere non-functional. It seems as though the mod-ifications that were made when baking the MOTchamber were irreversible.

3.4 Pfeiffer MaxiGauge

I next worked on creating an interface with thePfeiffer MaxiGauge. The MaxiGauge inputs thedata for up to six sensors and has pressure controlfunctionality as well as an easy to read digitaldisplay. The MaxiGauge outputs a digital signalso I started by trying to find drivers for LabVIEW.Luckily, Pfeiffer had created the drivers for theMaxiGauge. I then worked on LabVIEW codeto interface with the sensors and design it to the

8

specification of the ATTA group.

The Pfeiffer drivers had functionality writteninto them to choose a sensor number and outputthe current reading. My design changed this topull the readings for up to four sensors everysecond. I used a while loop to accomplish thisas well as duplicating the command to read fromfour sensors at a time. I then set it to display thepressure readings in a large display and outputit to a graph. The graph interface allows forthe user to pick a specific sensor to display orto display any combination of the four sensors.The MaxiGauge is set up so that if there is anoutput error, it will give a non-zero set value. Thiscould be potentially troublesome for the groupsince it would be logging data for a specific value,but this was not a real reading of the pressurein the apparatus. To correct for this, I changedthe program to write a value of zero for all errormessages.

I also changed it to write the data to a commadelimited file that would save the data from thelast 48 hours worth of data logging. After 48hours of data logging had been met, it would thenwrite to an additional file for the next 48 hours.After this 48 hours had been met, it will overwritethe first file that was created. I also added anoption for the user to write the data to a separatepermanent file if the data needed to be savedindefinitely.

The ATTA group is currently only utilizing twoof the six possible pressure sensors that read intothe MaxiGauge. Sensor 2 reads the pressure frombetween the source plasma beam and the firsttransverse cooling. The other sensor that theyhave set up is sensor 3 which reads the pressureat the source and can help the ATTA groupdetermine the gas inflow rate into the experiment.

3.5 Red Lion PAX

After completing the interface for the MaxiGauge,I worked on completing an interface for the RedLion PAX pressure sensor. This sensor displays thepressure in the same region as the Pfeiffer sensor 3.The PAX’s sensitivity is not as high as the Pfeiffer

sensitivity but works well for confirming the pres-sure at the source and thus the rate of inflowing gas.

The PAX also has a digital output signal. Icould not find drivers that had already been cre-ated for the PAX, so I began by creating a simpledriver to interface with the sensor. I tested thedriver with an RS-232 cable using the ASCII thatthe PAX responds to and found that it was notresponsive. I tested the cable and PC connectionsto confirm that they were functional and saw noissues. I adjusted the setting for output baudrate, parity, data bits, and address number andmatched the settings to my LabVIEW createddriver but was still unsuccessful in my attemptto interface the PC with the PAX. I contactedRed Lion technical support to try to furthertroubleshoot, and found that I had overlookeda detail when reading the PAX manual. I needto use a null modem cable or adapter. A nullmodem cable switches pins 2 and 3. We purchaseda null modem adapter and still found that thesystems were not interfacing. Not knowing whatelse to do, I posted the specifics of the situationon the National Instruments LabVIEW forumto see if I could get some help from people thathave had more experience with both LabVIEWand interfacing equipment. After following theadvise of all of the responses I had to my forumpost, I was still unable to get responses from thePAX. I ended up running out of time this summerto complete the interface with the PAX, but Iwas able to build a basis for future interface design.

4 Conclusion

The goal for my work this summer was to interfaceATTA’s temperature and pressure sensors with thelab PC. I successfully did significant troubleshoot-ing for their sensors so that a new temperature sen-sor could be purchased and the interface could becompleted. This troubleshooting gave the ATTAgroup the opportunity to find an ideal tempera-ture sensor for their apparatus and their needs. Ialso successfully created an interface for the PfeifferMaxiGauge that functions with the exact specifica-tions that they ATTA group needs. This enablesthem to read and record the data from the pressure

9

sensors throughout the system. I created driversfor the Red Lion PAX pressure sensor. Althoughthe interface was not fully functional by the time Ileft New York, it gave a great start to further workinterfacing the sensors.

5 Acknowledgments

I would like to thank the NSF for their generoussupport of the REU program. I would like to thankColumbia University for their hospitality and find-ing a place for us. I would like to thank Profes-sors Elena Aprile and Tanya Zelevinsky for allow-ing me to work with their ATTA group and theircreation and support of the project. I would alsolike to thank Andre Loose, Tae-Hyun Yoon, andLuke Goetzke (the ATTA group) for all their help,support, knowledge, and patience. I would like tothank the entire Columbia XENON group for theirsupport and knowledge while at Nevis Labs and onthe Columbia campus. Finally, I’d like to thank myfellow REU students.

References

[1] Williams, B et al., Imagination and Invention:The Story of Dark Matter. storybehindthe-science.org. p. 2-3

[2] Haynes, M. The Virial Theorem.Cornell.edu. Web. 3 Aug. 2012.www.astro.cornell.edu/academics/courses/astro201/vt.htm

[3] Tyson, Tony and Holloway, Ayana. Grav-itational Lensing and Dark Matter. Hay-denplanetarium.org. Web 5 Aug. 2012 hay-denplanetarium.org/resources/ava/universe/U0803gravlens

[4] Talcott, R. Dark Matter.Pbs.org. Web. 6 Aug. 2012.www.pbs.org/wnet/hawking/strange/html/dark.html

[5] McKee, M. MACHO matter is run-ning out of places to hide. 1 Jun. 2007.Newscientist.com. Web. 6 Aug. 2012.

www.newscientist.com/article/dn11967-macho-matter-is-running-out-of-places-to-hide.html

[6] Akerib, D. S. and Aprile, E. et al., Deep Under-ground Science and Engineering Lab S1 DarkMatter Working Group. 17 Feb. 2007. Web. 28Jul. 2012. arxiv.org/abs/astro-ph/0605719. p.3-13.

[7] Spethmann, C. A Exam Presentation: Di-rect Detection of WIMP Dark Matter. Pre-sentation. 1 May 2007. Web. 2 Aug. 2012.physics.bu.edu/ cspeth/wtalk.pdf. p. 24-32.

[8] Aprile, E. A 1 tonne Liquid Xenon Exper-iment for a sensitive Dark Matter Search.Presentation. 12-13 Mar. 2002. Web. 5 Aug.2012. www.docstoc.com/docs/75730047/The-XENON-Project-%28PowerPoint%29. p. 6-22.

[9] Aprile, E. et al. ”The XENON Project”- Direct Detection of SUSY Cold DarkMatter in Liquid Xenon. Presenta-tion. 19 Sept. 2002. Web. 26 Jul. 2012.xenon.astro.columbia.edu/XENON100 Experiment/Presentations/Talks 2002.html. p. 2-6.

[10] Angle, J. et al. First Results from theXENON10 Dark Matter Experiment at theGran Sasso National Laboratory. 1 Feb. 2008.Web. 2 Aug. 2012. arxiv.org/abs/0706.0039. p.1-2.

[11] Goetzke, L. Instrument Development for theXENON Dark Matter Search: An Atom TrapTrace Analysis System to Measure Ultra-lowKrypton Contamination in Xenon. Jun. 2011.

[12] Decowski, P. XENON-1T Project Plan.17 Sept. 2010. Web. 28 Jul. 2012.ebookbrowse.com/projectplan-xenon1t-a-4-pdf-d88358443.

[13] Exact Masses of the Elements andIsotopic Abundances. Scientific Instru-ment Services. Web. 7 Aug. 2012.www.sisweb.com/referenc/source/exactmas.htm

[14] Du, X. Realization of Radio-Krypton DatingWith an Atom Trap. Northwestern University.PhD Thesis. Jun. 2003. p. 30-101.

10

[15] Nave, R. Zeeman Effect in Hydrogen.Web. 5 Aug. 2012. hyperphysics.phy-astr.gsu.edu/hbase/quantum/zeeman.html

[16] Loose, A. et al. ATTA Device for MeasuringTrace Krypton Contamination in Xenon DarkMatter Detectors. Poster. Web. 25 Jul. 2012.atta.phys.columbia.edu/publications.shtml.

[17] Coma Cluster. Sloan Digital Sky SurveyImage of the Week. Web. 1 Aug. 2012.www.sdss.org/iotw/coma.jpg

[18] Fruchter, A. et al. Abell 2218: A Galaxy Clus-ter Lens. NASA.org. NASA Astronomy Photoof the Day. 7 Oct. 2011. Web. 1 Aug. 2012.apod.nasa.gov/apod/ap011007.html

[19] The Pair Telescope. Nasa.org.Dec. 2010. Web. 14 Aug. 2012.http://imagine.gsfc.nasa.gov/docs/science/how l2/pair telescopes.html

[20] Xenon100 Experiment. Xenon Dark Mat-ter Project. 17 Feb. 2011. Web. 2 Aug. 2012.xenon.astro.columbia.edu/XENON100 Experiment

[21] WarX. Beta Decay Artistic. Wikimedia Com-mons. 18 Mar. 2006. Web. 30 Jul. 2012.https://commons.wikimedia.org/wiki/File:Beta decay artistic.svg#filelinks

[22] Forces on Moving Charges - cont’d. TheXMM Satellite Schoolpage. Web. 12 Aug. 2012.www.sr.bham.ac.uk/xmm/fmc2.html

11