Studies Of A Room-Temperature Liquid Ionization Chamber For High-Energy Particle Calorimetry

Abstract | Introduction | The Experiment | Experimental Procedure |Results | Conclusion | The Next Step | References | Acknowledgements

Abstract

Introduction

The purpose of this paper is not to convince the reader that homogeneous warm-liquid tracking calorimeters are a good idea, but to describe the experimental apparatus and procedure that was used, to first clean a sample of tetramethylpentane(TMP), and then to see a photoionization signal. TMP was used instead of the much heavier homomorphic tetramethylplumbane TMPb (where the central atom of TMP is replaced by a lead atom) because of toxicity concerns. I will just give a quick thumbnail sketch of their benefits. For a more detailed account of the arguments for warm liquid calorimeters , see the papers by Heusch [1] and Engler [2] or the theses by Jensen [3] and Byrne [4]

In classical physics, calorimeters are devices used to study the amount of energy a substance contains. These devices work by burning a material and measuring the heat energy released.

In particle physics, this term is used for a special type of particle detector. This type of calorimeter is a device for measuring the total energy of individual particles. In the ideal case, no energy is unaccounted for. In these new calorimeters, the "material" could be hadrons, photons, electrons, or any of the elementary particles. They are accelerated up to large velocities and collide with each other. The products of these collisions can be other elementary particles, photons, or electrons. Determining the energy and trajectories of these resultants can provide much information to particle physicists.

The basic detection process is the loss of energy of a charged particle by ionization or radiation in matter[5], in our case tetramethylpentane. In general, a chamber in the detection device is filled with a counting medium and an electric field is applied. When ionization occurs in the medium because of the passage of charged particles, the resulting current can be measured. Electrons drifting in the field toward the anode are detected by a fine mesh of electrically conductive wires. The detection electronics are set up so that the energy and spatial characteristics can be known. It is hoped that with ever greater collision energies and detector resolutions that the basic tenets of particle physics will be tested using this type of detector.

Most calorimeters in use are of the non-homogeneous sampling type. This means that the material that slows down the particle, called the converter, and the material in which it is detected are different. Layers of converter medium are sandwiched between layers of the information transport medium, called the radiator. The converter material, a dense metal, can interact with the incoming particle by bremsstrahlung emission, pair creation or the strong nuclear force. If the particle reacts electromagnetically with the converter through bremsstrahlung emission or pair creation, the resultant electrons and positrons can ionize the radiator. When a strongly interacting particle ('hadron') strikes the converter, inelastic interactions and elastic scattering between the hadron and the nucleons in the converter take place[6]. During this process, secondary hadrons are produced. These secondary hadrons also undergo inelastic collisions producing tertiary hadrons. This process continues until the hadron's energy is small enough that it is absorbed in a nuclear process or stopped by ionization energy loss. This process is called a cascade. In this process, all the charged particles created can ionize the radiator, leading to a detectable signal.

Another type of calorimeter is the homogeneous calorimeter, where the radiator acts as a stopping medium and information transport medium. Of most interest is a theoretical refinement to a homogeneous calorimeter, the homogeneous tracking-calorimeter. A homogeneous tracking-calorimeter can measure a particles trajectory as well as its total energy.

To work well as a radiator in a homogeneous calorimeter, a substance must have the following qualities: 1. high stopping power (the ability to slow down and absorb energy from the incoming particle in MeV/cm 2 ); 2. small radiation length yield (the amount of electrons given off per unit path length); 3. high electron mobility (how fast the ionization electron drifts in a given electric field strength); 4. low electron affinity (the tendency for the radiator to combine with free electrons); and 5. short radiation length ( how much energy is dissipated per unit length). Currently, the radiator most used is a liquid noble gas, such as argon, xenon and krypton. The inherent drawback to using a liquid noble gas is that it has to be kept at cryogenic temperatures. Another possibility is to use a room-temperature liquid. This project explains this type of calorimeter.

The advantages of homogeneous calorimeters versus sampling calorimeters can be summed up in one principle, hermeticity. Hermeticity is the ability to obtain information over the full solid angle. No information is lost in the non-existent converter materials, since all the avalanching happens in the detection liquid. All the showering and cascade information that is created in the converter material that does not make it out is lost. Since the conversion takes place in the detection medium, homogenous calorimeters permit individual tracking, so that we can separately detect individual particles. With the lack of converter materials, information is preserved about the entire trajectory of a particle, and a greater solid angle coverage is often possible for better hermiticity.

In a homogeneous calorimeter, a warm liquid has certain advantages over a gas as a detection medium: Because of a higher density the liquid has a much larger stopping power. Homogeneous calorimeters are not as troubled by Landau tail problems [7]. They have a higher electron yield [8] . By virtue of their greater density the detector can be designed more compactly. Our target radiator fluid, TMPb, is presumably quite radiation hard [9] . Liquids require an energy loss of 1MeV per layer of absorber so most of the background radiation is screened out, making them useful for tracking particles in a high-flux environment.

A good particle detector will give the most information in the smallest amount of space. It will cover a large part of the solid angle. Cryogenic sampling calorimeters do not fulfill these needs very well. The need to keep anything at cryogenic temperatures is fraught with problems. The machinery needed is large and bulky, taking up large amounts of space and reducing the amount of solid angle covered. The cryostat takes up valuable space, and inserts materials in the particular paths. Servicing the detector, which is a very real concern since, at such low temperatures, the detector mechanicals are subject to instability, becomes almost impossible. Last but not least, the extra cost of a cryogenic detector is prohibitive in the era of shrinking budgets for big science.

To study the properties of room-temperature liquid calorimeters, we used a design of Jacques Séquinot of CERN [10] , shown in figure (5). A pulsed ultra-violet beam will be incident on the zinc-plated photo cathode. The quartz-glass window to which the steel mesh anode is bonded does not transmit any photons of wavelength less than 1600 Å to prevent the target liquid from becoming photo ionized. With ultra-violet photons having average energy greater than the work function of zinc, 4eV, electrons will be photo-emitted from the cathode into the tetramethylpentane, where they will be accelerated by an electric field of 10kV/cm to the anode and the detection electronics. With this apparatus, the mobility of electrons in tetramethylpentane and then later TMPb, can be studied. The mobility is given by

u = d/t E = d2/t V,

where d is the distance between the cathode and anode, t is the time of flight of the photo-emitted electron, E is the electric field and V is the voltage difference between the cathode and anode.

The main difficulty in using tetramethylpentane as an ionization medium that does not capture the freed electrons as they drift toward an anode, is that it has to be exceptionally pure. Impurities reduce the signal by combining with the drifting electrons before they can be detected. (The cleaning process is most important because without any free electrons we cannot see a signal.) It has been found [10] that contamination on the order of 5 parts per billion is enough to combine with most of the free electrons. So an effective cleaning process is important. The cleaning of tetramethylpentane with a small table-top apparatus is the main focus of this thesis.

We start the cleaning process by immersing the tetramethylpentane in a liquid-nitrogen bath. At liquid nitrogen temperatures, tetramethylpentane solidifies, but electron-hungry impurities such as oxygen and nitrogen are only liquefied. The transport column is then opened to vacuum, and the radicals and other impurities are neatly pumped away. The tetramethylpentane is next moved to the purification column through vacuum distillation-condensation for a final cleaning. The purification column contains silica gel and zeolite. The zeolite is a "molecular sieve" which traps any large molecules, such as water. The silica gel traps halide impurities that would scavenge any free electrons. After a period of convection-aided mixing in the purification column, the tetramethylpentane will be tested qualitatively for cleanliness.

Using the same vacuum distillation-condensation procedure, the tetramethylpentane will be moved from the purification column, to the holding column, to the drift chamber. Once inside the drift chamber, an electric field will be established and free electrons will be photo-emitted into the tetramethylpentane. If a current is detected, then the tetramethylpentane will be considered purified, on a qualitative level.

The Experiment

Purification Column

The purification column can be seen in fig(4). The column is packed with 5 and 10 Å molecular sieves and silica gel. The chemical composing the sieves has a lattice structure called zeolite. Zeolites are alkali aluminosilicates that have a surface area of about 10 3 m 2 g -1 . Impurities with molecular diameters less than 10 Å will be trapped by the sieves. The silica gel captures electron-hungry halide impurities. Since there is a danger of the silica gel escaping outward into the vacuum system during rough pumping, a sintered metal filter disk was placed in the main vacuum line.(see figure 4)

Test Cell

A cross section of the test cell, or drift chamber, is shown in figure(5). The chamber is composed of Macor, an easily machinable ceramic with extremely low porosity, quartz glass, and stainless steel. The drift chamber consists of three parts. An endcap holds the zinc-plated cathode and gold-plated guard ring. A stainless steel tube has the quartz window bonded to the inner end. Another larger piece of Macor that encapsulates the stainless steel tube is bonded to the endcap. Indium wire is used as a gasket material, with both pieces being held together by aluminum clamps.

A10-20 kV/cm electric field is established between a fine stainless steal mess anode and a zinc-plated cathode. The stainless steel mesh is bonded to the inner side of the window, and is in electrical contact with the stainless-steel tube. To reduce fringe field effects, a gold-plated guard ring is held at the same potential as the zinc-plated cathode. The photo-emitted electrons will thus travel along the straight inner field lines. Otherwise the curvature of the field lines would give incorrect time-of- flight results.

Electronics

The ultra-violet light is produced by a Hamamatsu mercury-vapor lamp. The lamp will not operate in a pulsed mode, so a chopper was fabricated. A standard 110Vac computer cooling fan was placed between the lamp and the test-cell window. A piece of cardboard with a small slit was attached. As it rotates, the slit aligns between lamp and cell and we have a pulsed ultra-violet light. The rate of rotation of the fan, and thus the pulse length, was controlled by a variac transformer connected to the chopper.

The 10 kV dc needed for the electric field was produced by an HK dual MWPC power supply. One channel was used for the main field and one for the guard field. The HK can only put out a maximum of one milliamp. The HK has a feature which was very useful to us. Integrated in the system is an extremely sensitive current meter. It is capable of measuring currents in the .1 micro-amp range. I used this meter to detect the current produced by photo-emitted electrons conducted through the tetramethylpentane.

The pressure gauges were controlled by a Granville-Phillips vacuum process controller, model 303. It controlled the two ion gauges: one above the t-pump inlet and one just before the test cell and the convectron gauge, between the t-pump and the rough pump.

Experimental Procedure

Cleaning

In order to achieve the ultra-high vacuum needed for this experiment, the individual components were cleaned, assembled, and then cleaned again accordingly.

Before assembly all stainless steel parts were cleaned by:

1. Degreasing in perchlor-ethylene vapor (10 min.).
2. Ultrasound bath in a 2% almeco solution (15min.).
3. Rinsing with demineralized water (5 min.).
4. Vacuum baking at 950 C or highest withstandable temperature of each part (3 hr.).
   
   After assembly the system was further cleaned by:
   
   
5. Flushing with 60 C acetone (3 hr.).
6. Flushing with ultra-pure 80 C water (36hr.)
7. Dried with hot helium (3 days).

Achieving the Desired Pressure

After the original system was cleaned-assembled-cleaned as outlined above, another student had attempted to achieve the low pressures needed to conduct this experiment. He[11] ran into serious problems with the mechanical reliability of the t-pump, and had to abandon the project due to graduation-time constraints. Before leaving he backfilled the system with high-purity nitrogen, and it was left that way for eight months until I took up the reins. The system, as I found it, was clean; but now a way had to be figured out to fix the t-pump and get the system into the desired pressure range.

The problem in the past was that the bearings the t-pumps' motor spins on kept over heating and seizing. After checking with several bearing supply companies, I discovered that they were the wrong type, never intended for high-speed several-month-long duty cycles.

While I was waiting for the correct bearings to arrive, I decided to rebuild the motor, using parts from the several used motors lying around. Picking through the best parts, I constructed one clean, well balanced and well-lubricated motor. I just hoped it worked because I did not have any good spare parts with which to construct another. After reattaching the motor, I was ready to try to outgas the system.

I started the mechanical pump and allowed it to bring the entire system down to the 10-3 torr range. I left it pumping for seven days, after which the convectron gauge at the rough pump inlet was reading in the high 10-4 range. The turbo pump was then started and slowly (to avoid overheating the motor) brought up to speed. The pressure at the t-pump inlet shortly read in the 10-6 range and the far ion gauge, at the test cell, read in the 10-5 range. After about an hour the t-pump motor was getting very warm, so it was shut off. I decided to try to degas the system, with the t-pump shut off, by using the t-pump's built-in heater and wrapping the entire system with heat tape. Immediately the pressure jumped up to the 10-1 torr range. The system was baked out for three weeks. After that, the pressure at the rough pump was again reading in the 10-4 torr range. The t-pump heater was shut off and it was allowed to cool. The t-pump was now restarted and allowed to pump on the system. After about an hour, the pressures again read 10-6 at the t-pump and 10-5 torr at the test cell. After several hours the t-pump's motor was still cool, so I decided to let it run over night. The next day the motor was still cool but the pressure had not dropped at all. Knowing that the stainless steel tubes would be outgasing nitrogen for a while, I decided to try to keep the t-pump working as long as I could. I turned on the internal heater for the t-pump so I could do a bakeout at this new lower pressure. The pressure at the t-pump rose by a factor of ten and the motor became slightly warm to the touch. The bakeout continued until the pressure reached the pre-bakeout level. Now the heat tape wrapping the system was turned on, so the entire system could bake out at a lower pressure than was possible with just the rough pump. With the system at a temperature of about 3000C, the pressure rose to the high 10-5, almost 10-4 torr, range at the t-pump inlet. I allowed the system to bake out at this temperature for several weeks. After this, with the temperature still at 3000C, the pressure at the t-pump was in the high 10-6 range, and at the test cell it was around 10-5. This was encouraging, because I felt that when I dropped the temperature, the pressure would also drop. After the system cooled, the pressure at the t-pump was 10-7 torr and at the test cell it was 10-6 torr, the lowest so far; the motor was running nice and cool. I let the system run as it was, checking the pressure and motor temperature several times every day. After a week, the pressures had not changed by the order of magnitude necessary, so I went hunting for reasons why. The obvious reason was that I had a leak. After the degassing period, I helium-leak-tested the system. Two small leaks were found in the inlet and outlet ports of the test cell. Tightening the connections eliminated them and, to the limits of the leak tester's accuracy, my system had no leaks.

With the leaks corrected, the pressure had dropped a little. At the t-pump it was jumping between 10-8 and 10-7 and at the test cell it was a more steady 10-7 torr. Now I was concerned, because the pressure was not where I needed it to be, and I had not even exposed the holding column or drift chamber to high vacuum. After much thought, I concluded that a redesign of the high vacuum system was in order.

For any system being pumped on by a vacuum pump with speed Sp and conductance C, the actual speed of evacuation is given by Se [12]

1/Se = 1/Sp + 1/C

The speed Sp of the t-pump is 240 L/s [13]. This can not be changed. The only thing in my power to change is the conductance.

At low pressure, gas flow is dominated by molecular flow. Molecular flow happens when the mean free path of the molecules is on the same order as the tubing diameter.[14] The conductance of a tube in the molecular flow regime is given by

Cm = 2/3p (d3/l) v cm3/sec,

where d is the diameter of the tube, l is the length of the tube, p is the gas pressure, and v is the average molecular velocity, given by

v = 14.55x103(T/M)1/2 cm/sec;

here, T is the temperature in Kelvin and M is the molar mass. To improve the conductance, I would need to increase the diameter of the tubing and or decrease the length. Due to budget constraints, it would not be possible to replace existing tubing with tubing of a larger diameter.

Talking with a resident expert[15], I learned that a rule of thumb is that every right angle in a vacuum system is worth two feet of tubing. In the original system there is a piece of tubing that could be removed. It connected the flexible hose leading from the cold trap to the main vacuum line. This section of pipe has two right angles, is over two feet long and is only half inch pipe. If I could connect the two inch flexible hose directly to the main vacuum line, I could cut out an effective six feet of vacuum pipe. Unfortunately, to accomplish this I will have to shut off the pumps and backfill the system with high purity nitrogen. The purpose of the removed section was to mate the t-pump at one height to the main vacuum line at a much lower height. To make ends meet, I placed wooden blocks as shims under the legs of the table to which the vacuum system was attached. Luckily the flexible hose allowed me some leeway in matching heights. With the flexible hose reconnected the t-pump has a straight shot to the drift chamber.

I started the rough pump and allowed it to pump out the nitrogen. After a few hours, I started the t-pump and waited. I was not expecting fantastic pressures immediately because of the outgassing nitrogen. After six days, the pressure at the t-pump was in the low 10-9 torr range and at the test cell it was a fairly steady mid 10-8 torr. With the main tubing at the correct pressure, I was able to start to open the rest of the system up to vacuum.

I slowly opened the valve to the holding column and the pressure went into the 10-6 torr range. After several days the pressure had returned to the 10-9 range. I next opened the main vacuum lines to the test cell. I also opened the interconnecting tetramethylpentane transport lines to vacuum. After five days the pressure had returned to the desired level.

Degassing the purification column came next. As a precaution I closed off the holding column and test cell to vacuum. The silicagel has the consistency of a very fine sand. There is some danger of it shooting through out the system if the purification column is exposed to vacuum too quickly. The t-pump was shut off, so that the pressure differential would be as small as possible. As I very slowly turned the purification columns vacuum valve, I watched the readout of the convectron gauge for any fluctuations. As soon as I saw the pressure increase, I stopped and let the rough pump slowly remove the air. After 15 minutes I would crack the valve a little bit more. After the valve was completely open the pressure was in the 10-2 torr range. The purification column was pumped on until it read in the 10-3 range. To further expel any contaminants, the purification column was wrapped with heat tape and brought to 400o C.

The pressure now jumped to the 10-1 torr range. The purification column was allowed to bake out until the pressure dropped to a range where the t-pump could be started. Not wanting to overheat the delicate t-pump, the heat tape was turned off. After the t-pump was brought up to speed, I allowed the temperature to slowly, over several days, return to 400o C. It took three weeks for the pressure to return to preheat levels. I now reopened the valves to the holding column and test cell and their interconnecting fluid lines. With the entire system open to vacuum the pressure at the t-pump was in the mid 10-8 range and at the testcell it was high 10-7. I hoped that if I waited it would return to the earlier lower levels. After two weeks of pumping on the whole system the pressure had dropped somewhat, to a high 10-9 at the t-pump and mid 10-8 at the test cell. When the cold trap was filled with liquid nitrogen the pressure dropped a little, but the main effect was to even out the pressure fluctuations. With the cold trap full, the pressure had stopped jumping around and stayed at a steady mid 10-9 at the t-pump and low 10-8 torr at the test cell. After another week had passed, the pressures had not changed significantly, and it was decided to try to move the tetramethylpentane.

Transferring the Tetramethylpentane

The tetramethylpentane is moved through the system by a combination of vacuum distillation and condensation. The procedure used is as follows;

The tetramethylpentane transport container was immersed in a liquid nitrogen bath. The tetramethylpentane solidifies while the gaseous contaminants (such as oxygen) are only liquefied. The vacuum valve was opened and the liquid impurities were pumped away. A small increase in pressure was recorded (10x). After the contaminants had been pumped away the liquid nitrogen was allowed to evaporate (with the vacuum valve closed). After it had completely evaporated, the transport container was wrapped with heat tape. The fluid lines leaving it were wrapped all the way to the purification column.

The purification column was now immersed in the liquid nitrogen bath, with its vacuum valve open. After the purification column had reached cryogenic temperatures, its vacuum valve was closed. The heat tape was now turned on. After it had warmed the fluid lines, the fluid connections between the two were opened (with both vacuum valves closed). By virtue of the lower pressure in the purification column the tetramethylpentane was pulled through the fluid lines. Unfortunately, there were no viewports in the purification column or holding column. To be sure the maximum amount of tetramethylpentane was transferred, I waited six hours, keeping the purification column's liquid nitrogen bath full. At the end of the six hours I turned off the heating tapes and closed the fluid connections after the lines had cooled. Convection currents were set up in the purification column in an attempt to mix the tetramethylpentane. I wrapped the lower fourth of the purification column with heat tape and immersed the upper half in an ice water bath. I allowed the tetramethylpentane to soak in the purification column for forty-eight hours before trying to transfer it.

The transfer of the tetramethylpentane into the holding column was performed exactly the same way. With the window in the holding column it was possible to see the tetramethylpentane dripping into the column. With this process, the holding column was filled to about a third with the tetramethylpentane. Before the tetramethylpentane was transferred to the test cell, I froze it and pumped on it for fifteen minutes. After the liquid nitrogen bath evaporated, the tetramethylpentane was moved into the test cell with the above mentioned technique. There is no viewport in the test cell, so it was impossible to determine how much tetramethylpentane it actually contained. After the six-hour transfer process, the tetramethylpentane level was approximately one ladder rung lower. How much tetramethylpentane was actually in the cell, and how much was still in the fluid lines, could not be determined.

Towards Detection of a Signal

Results

Initially, without the pulsed ultra-violet light, a large current of approximately one milliamp was seen. This leads me to believe that there were still some impurities in the tetramethylpentane. If there were some impurities, in the form of ions, they would be attracted to the cathode and cause an ion current. After several minutes, the current decreased to .2 milliamps. The current meter stayed at .2 milliamps and did not fluctuate. What caused this constant background current is not completely clear. One possibility is that, with 8kV applied across the test cell, a leakage current passed through paths in the drift chamber itself.

When the chopped ultra-violet light source was applied, an additional current was observed as a jump in the needle of the power-supply current meter. The motion of the current meter needle appeared to coincide with the ultra-violet light pulses. The equipment was not available to determine if the charge carriers were slow ions or fast electrons nor could we obtain any information on the decay times of the pulses. This is a very qualitative result, but the best that was possible given the equipment and time constraints.

Conclusion

In conclusion, I demonstrated that it is possible to reduce the level of contaminants of tetramethylpentane to a level of a few parts per billion with a cleaning procedure using a relatively uncomplicated table-top apparatus. I believe that this procedure can be used to clean warm liquid radiators, such as TMPb, for use in tracking calorimeters.

The method which I used for demonstrating the purity of the liquid could also be used as a diagnostic tool in future calorimeter development. This work represents a small advance in particle detector technology but has potentially large implications. Further study of tetramethylplumbane is necessary before this type of liquid ionization-chamber based calorimeter can be put to use in an actual experiment. While I have shown that it is possible to clean the tetramethylpentane on a small scale, gearing up this approach for a full size detector requires further study.

Limitations of the present experiment are: First, the power supply that was used would only put out 8 kV even though it is rated for 10kV. With a dielectric constant of tetramethylpentane of 1.98, the electric field inside is reduced, so that an even greater voltage needs to be applied to achieve the same E-field strength. The actual E-field strength was only 7.6kV/cm when it needed to be at least 10 kV/cm. There was a large background current of .2 milliamps. With a background this large, the detector is limited to its least sensitive scale, making the detection of a signal even harder.

To improve the performance of this experiment, the field has to be brought up to the correct value, and the current measuring technique needs to be improved.

The Next Step

With the tetramethylpentane at an acceptable level of cleanliness, the next step is to conduct a through quantitative analysis of the behavior of free electrons in tetramethylpentane. I would like to offer a plan for a quantitative analysis. To cure this experiment of its qualitative aspect, several changes need to be made; the electric field strength needs to be known exactly, the length of the ultra-violet pulses needs to be controlled accurately, the time between pulse origination and signal detection and the time varying characteristics of the signal need to be known. In keeping with the table-top apparatus theme, the electronic control center could be an IBM compatible with the appropriate analog-digital control boards. Hamamatsu offers an ultra-violet source whose pulse length and timing can be computer controlled. A computer-controllable high-voltage power supply should be used to set the correct electric field strength. Several companies offer extremely sensitive volt-amp meters on a computer card. With one of these cards, the electric field strength and the time variance of the current can be known. The computer can perform real-time calculations using the most current information. Drift-time, electron mobility, lifetime, and their dependencies on electric field strength can be determined.

To aid in detecting the signal over the large background, a low-noise high-gain differential amplifier with good common-mode rejection ratio should be used as a pre-amp. With this experimental apparatus, all of the important variables can be recorded, and the parameters can be changed from a simple desktop PC ; the whole process thus can be automated, giving the experimenter repeatability and quantitativeness.

References

[1] Heusch, Clemens A.; "Prospects for Homogenous Warm Liquid Calorimetry"; Nuclear Instruments and Methods in Physics Research - Section A; North Holland Publishing Company, Inc., Menlo Park, CA; 1987, pp. 257-324.

[2] Engler, Joachim; Keim, Heinrich; "A liquid Ionization Chamber Using Tetramethylsilane", Nuclear Instruments and Methods in Physics Research - Section A; North Holland Physics Publishing Division, vol. 233, 1984, pp. 47-51.

[3] Jensen, Jonathan; Electron Drift In Nonpolar Dielectrics; Senior Thesis, University of California Santa Cruz, 1992.

[4] Byrne, Cindy; Transport Phenomena in Heavy Liquids; Senior Thesis, University of California Santa Cruz, 1986.

[5] Paul, E.B.; Nuclear and Particle Physics; American Elsevier Publishing Company Inc.; New York, 1969, pp. 93-107.

[6] Kleinknecht, Konrad; Detectors for Particle Radiation; Cambridge University Press; Cambridge, 1986.

[7] Huxley, L.; The Diffusion and Drift of Electrons in Gases; Wiley Interscience Publication, John Wiley and Sons; New York, 1974 p. 23.

[8] Dodelet, J.P.; Can. J. Chem. 55 2050, 1977

[9] Swallow, A.J. Radiation Chemistry of Organic Compounds; Pergamon; Oxford, 1960.

[10] Holroyd, R. Anderson, D.; "The Physics and Chemistry of Room-Temperature Liquid-Filled Ionization Chambers"; Nuclear Instruments and Methods in Physics Research; p. 297.

[11] Jensen, Jonathan; Electron Drift In Nonpolar Dielectrics; Senior Thesis, University of California Santa Cruz, 1992.

[12] Melissinos, Adrian C.; Experiments in Modern Physics; Harcourt Brace Jovanovich, Publisher, San Diego, 1966, p. 127.

[13] Sargent-Welch, Vacuum Products Division, "Operating Instructions, Maintenance and Repair Manual for Turbo-molecular Pump Set", 18, 1982.

[14] ibid.

[15] Professor Frank E. Bridges, physics department University of California Santa Cruz, January 93, personal communication.

Acknowledgements

I would like to thank my adviser Clem Heusch for all the help he has given me and his patience. Even though this project took two years longer than he expected, he still made time in his busy schedule to meet with me and finish this work. I would also like to thank Alec Webster for all his help and ideas with the myriad mechanical engineering problems that kept cropping up. I would also like to thank Bill Rowe for his help with the lab computer. To everyone who had to work in the high energy lab and endured, with a smile, the noise and smell of this experiment, I thank you for your patience. To Professor Abe Seiden, thank you for allowing me to finish this project.