A STU}? C235 .33 PARAL ‘33. 33" ‘3‘35 éflifiéZA‘flflN CHAMEER Thasés {‘05 fine: Qagzw 2:? M. 32. MtC’iiGA-‘fl $332313 LNWERETY Flaws}? java-9:333:31 39693 llllllllllllllllllllllHHIHIHIIIHIIIIHUIIIIIllHIllllllll 1 31293 017640404 LIBRARY Michigan State University D a 3*. Swine-M 3 PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE MAY 1 6 2005 @927 06 A STUDY or A PARALLEL PLATE IONIZATION CHAMBER By Raeool Javahery A IHESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree or MASTER 0! SCIENCE Department of Physics &4Aetronony 1961 ACKNOWLEDGMENTS The author-desires to express his sincere thanks and gratitudes to Dr. William.fi. Kelly, and Dr. Herbert H. Belotin who have been very patient and helpful in guiding the author in this study. Further gratitudes are expressed to Dr. 5.x. Haynes, Head of the Department of Physics and Astronomy; the Iational .Lcademy of Sciences; and the authorities of Teheran University who have made it possible for the author to continue this research. ABSTRACT A STUDY ABOUT PARALLEL PLATES IONIZAIION CHAMBER By Rascal Javahery Submitted to Michigan State university in partial fulfillment of the requirements for the degree of MASTER or SCIENCE Department of Physics e Astronomy 1961 F) ‘7 ‘ Approved: Air k)?£;:’\vy 1 Rasccl Javahery ABSTRACT In this thesis the theory of ionization chamber, pro- portional counter and geiger counter is studied. The shape of pulses and the effect of grid is discussed. we used the parallel plate ionization chamber and we introduced diffe- rent parameters such as voltage difference across the elec- trodes, dimensicn of collector (collector with and without guard ring) and pressure. We showed that the characteris- tic cf ionization chamber is a function of these parameters. Also we showed that this parallel plate ionization cham- ber practically cannot be a proportional counter. TABLE OF CONTENTS I. Introduction . . . . . . . . . . . . . . . . . 1 II. cantata O O O 0 O 0 O I O O 0 O O O O O O 0 III. Review of Literature . . . . . . . . . . . . . 7 A. The velocity of electrons in ionization chambers . . . . . . . . . . . . . . . . . 7 B. The pulse shape in parallel plates champ . bereeeeeeeeeeeeeeeeeee7 «(4 C. Chamber with grid . . . . . . . .-. . . . 11 IV. Construction a Performance of the Apparatus . 13 Remarks regarding proportional chambers . . . 16 V. Experimental Results . . . . . . . . . . . . 18 A. Effect of pressure for small size collec- tor O O O 0 0 O O O O O 0 O O O O O O O O 18 B. Effect of pressure for large size collec- tor O O O O O O O O O O O O O O O O O 0 O 19 C. Collector with guard ring . . . . . . . . 21 D. Comparison of pulse height with guard ring and without guard ring A . . . . . . . . 22 E. Resolution . . . . . . . . . . . . . . . 23 VI. Discussion . . . . . . . . . . . . . . . 24 Leakage problem . . . . . . . . . . . . . . . 25 VII. Bibliography . . . . . . . . . . . . . . . 26 LIST OF FIGURES PIGURE I Ionization region . . . . . . . . . . . . . . . II Potential between two electrcd . . . . . . . . III Parallel plates chamber . . . . . . . . . . . . IV The shape of pulse when RC is infinite . . . . V The shape of pulse when RC<§t2 . . . . . . . . VI Pulse shape for different value of RC . . . . . VII Parallel plates chamber with a grid . . . . . . VIII The ionization chamber . . . . . . . . . . . . ‘IX Connection between the ionization chamber, vacuumpumpandtankofgas.......... X Block circuit diagram . . . . . . . . . . . . . II Cylindrical and parallel plate counters . . . . XII Field lines between plates and the effect of pressure on the mean free path of a particle. . XIII.Pulse height against pressure. . . . . . . . XIV. No. of counts per minute vs. pulse height . . . IV The field lines in large size collectors . . . XVI Pulse height as a function of pressure . . . . XVII a XVIII Pulse height vs. voltage, without guard ringandwithguardring. .......... XIX field line pattern 0 O O O O O O O O O O O O 0 xx Pulse height as a function of voltage . . . . . PAGE 10 10 11 14 15 16 16 1 8 18 20 20 20 21 21 22 I . INTRODUCTI 0] Charged particles in penetrating through matter may lose energy by collision with atoms. These collisions may result in ionization by ejecting electrons (usually valence electrons). If the energy of a particle reaches some specific value (icni-. zaticn energy) it can icnize atoms. The average energy loss per unit path length for heavy particles in matter has been "I calculated 1 as: . _‘.¢'2‘§_21z~____g__z’c (1’).)[43 i...””-c] (I) ‘ E is the energy of the incident particle; v is the velocity and We the charge of the incident particle, l the number of atoms per unit volume of material, Ze their nuclear charge, I the average exitaticn potential, Hoatcmic rest mass of matter and no electron rest mass. m is the mass of heavy particle. C is a constant. The average energy loss of electrons per unit length can be written as: .3: 1.3L“ ESL/Ta? '5'] (2) In this eEquaticn the factor 1}? is missing. Dividing equation 0 (l) by equation (2) gives the ratio of the change of energy per unit length for a heavy particle and an electron. Let us assume that the heavy particle and the electron have the same energy. Then this ratio is a large number. Therefore, usually the change of energy per unit length for ejecting electron will not be sufficient to produce significant ionization. However. if a strong electric field is introduced, these electrons can be accelerated to sufficient energy to produce further ioniza- ticn,(seccndary ionization). In fact, these secondary electrons can themselves be accelerated to the point at which they pro- duce significant ionization. If charged particles expand all their energy through colli- sions, it is always the more energetic particles that give a larger number of electrons. This thesis is concerned with a study of ionization counters and the influence of different pa- rameters on their operation. It is important to study factors which change collecting time; the effect of small amount of impurity in the gas cone tained in the chambers; different types of ionization counters; and also how one can change the character of the ionization counter by changing other factors. Here we consider only ionie zaticns which are produced originally by alpha particles, and not by beta or gamma decays. let us first discuss the different types of ionization counters. - 3 - II. COUNTERS Gas filled counters indicating the arrival of single particles can be classified as ionisation counters, propor- tional counters, and geiger counters. These ionization coun- ters are chambers with at least two electrodes. The collec- ting electrode is normally at ground potential and the other electrode is at a negative potential with respect to ground. we assume that icnizaticn.takes place in the chamber, with say, n=10 electrons being released. When V the potential difference between the electrodes is very small (of the order of a few volts or less), all 10 electrons will not arrive at the collector, since some electrons will reccmbine with posi- tive ions. This effect is called recombination. .As V'increases the number of electrons collected at the electrode will increase. At V, saturation is achieved and all ten electrons arrive at the collecting electrode. Saturation continues until some specific voltage V2 is reached. The region from‘V1tc V2 is called the ionization region. There is no recombination effect in this region. At V2 electrons acquire sufficient energy between colli- sions tc ionize the atoms with which they collide (secondary ionization), and the number of electrons arriving at the collec- tor rises above ten in an approximately exponentially fashion with V (as will be shown later). $1., Flir- 1% Team- (U 1 l (a) Volta r .2 L 4 e V 3 ‘C ‘Q sedr YY Each electron originally present produces a small ava- lanche of secondary electrons by secondary ionization close to the collector, since as the electrons get closer to the collec- tor the total number of secondary electrons increases. The region between V2 and V3 is called the proportional region, md counters operating in this region are called propor- tional counters. In ionization counters and proportional counters the num- ber of electrons collected is proportional to the energy of the particle detected. The greater the energy the larger will be the pulse height. In the ionization region the pulse height is independent of applied voltage, in the proportional region - 5 - the pulse height is a function of voltage. let us consider these two regions more carefully. We shall imagine that we use a source which emits alpha parti- cles of two different energies. Therefore, the initial nump ber of electrons which are released are different. Curve (a) in Pig. 1 corresponds to n :10 initial elec- trons. In region V‘ to V? the curve is a horizontal line. If initially lOO electrons are liberated, curve (b) results. This curve is also horizontal in region between V1 to V2, but it is different from.curve (a) by one on a log Ion scale. There are interactions between avalanches produced by each initial electrons. At V3 the positive ions from one ava- lanche hinder the development of the next avalanche. There- fore the discharge with 100 electrons will be more effected than the one for 10 electrons, and n begins to decrease rela- tive to curve (a). Thus (a) and (b) are no longer one unit apart in this region but approach each other. In the V3 to V4 region the final number of electrons in the pulse is no longer proportional to the initial ionization as it was in region Vé to V3, though the greater number of initial electrons still gives the higher pulse. Between V3 and V; the curves rise rapidly. At V;, (a) and (b) Join to form a single curve. Region V3 to V4 is called the ”region.of limited proportional countsrf. The very rapid rise of n at V;_means that one avalanche breeds one or more further avalanches, and this breeding continues until a considerable space charge due to positive ions is built up near the collector. This space charge affects the field at the collector quite considerably, so the number of avalanches rises less rapidly and further breeding is supp- ressed. The final charge developed is then determined not by the number of initial electrons but principally by the posi- “T tive ion space charge. ‘ The region from VA to V5 is the geiger region. In this region the pulses are independent of’magnitude of primary ionization and their sizes are uniform and depend on the threshold voltage Vfi. In this case the discharge spreads from primary ionization through the whole counter chamber. The potential between two electrodes as a function of position between the electrodes is shown in rig. 2. Curve (a) shows the potential between two plates without any space charge and curve (b) shows the potential with space charge. V’ yr 1 A . ‘ 3° clearest.» we. Name-«JV? (b) (a) «Cd; av ca ‘filhl . °' “‘21.” Fig. 2 - {c dntial between two elec- rc . is; no space charge. b space charge from.pcsitive ion. III. REVIEW 0? LITERATURE (a) The velocity of electrons in the ionization chamber. The approximate velocity of an electron in an ioniza- tion chamber has already been established.2 for two parallel plates namely positive collector and negative plates, the field between the plates is given by ea} . The plate separation d is assumed to be small com» pared with the plate diameter. The velocity in terms of field is given by M’zK—E’é— . Therefore area—f (3) I is a constant and P is pressure. H X 'Xo 11g. 3 - Parallel plates chamber. For example: for V23OOC volts, :5 cm., P=4O psi and using the value of I for argon gas, the velocity is v=lC6 cm/sec. The diagram.fcr a parallel plate chamber is shown in Pig. 3. The two electrodes (plates) are separated by a distance (d) and the x axis is taken perpendicular to the lower plate at point C. b) The pulse shape in parallel plates chamber. Here we shall use the same diagram as before, viz. Pig. (3). ' Let us assume that n positive and 11 negative ions are formed at the average distance x0 from the lower plate. Because of the electric field E, these ions move in direction of x, and the average work done on these electrons will be: 7‘9 J 115an " Prom conservation of energy, cv‘ cv‘ 7‘ 7-72 )“ shade (4) C is the capacity of the system, V0 is the initial potential difference between the electrodes before ionization takes place, V is the potential difference at any time t, and E the field intensity. The left hand side of the above equation can be written: %—(v:€)= g—(v-ig )(mg) Let us use the approximate values wv; 9110.3 , and V°=IOOO volts. Therefore var. :4 ix; . Let V - Vo=Vc: Vc is the sum of the two voltages due to positive and negative ions on the collector. Therefore: 91c WWW!) EJX (5) X . and substituting for Eefifi. , we obtain: Vela” crux (6) hence :33 crux (7) C CKJK . ) ’ 35 Cd ( o ) 813.3. Y. 21‘ e ‘2 Changing the variable x to t, and using: or t. 'M: =3 Am.) on...“ (a) Hhsre vais the velocity of electrons, we obtain: V: hEM’t . é a: (9) low let us use the symbols V and V the 41-111; valocity of negative and positive ions. Then: 4- 4' - ' yew (10) were“ (11) c cd ‘1 ed The average final value Vc and V6 from equation (7) will be: - O 12 ‘E;--E§-(A-W,)=fic§é(\-.}) 3 3 v‘ :ne'x, c a, (13) These equations show that for the two parallel plate chamber, the pulse is a function of x0, and also the pulse height is a function of x0. ’ Returning to equations (10) and (11) we note that the positive ions are heavier than electrons‘and their drift velo- city is smaller than that of electrons, Kr 93.100035: . The pulse due to negative charges rises with slope "5d” until time |-.- 3;: , where t1 is the average time the negative ions or take to travel between the plates. After this time the pulse is due to positive ions and it rises very slowly. Its slope we R? ad The total final voltage from equations (9) and (10) is is: found by taking: - " .. n 14 \QQ“49*"Eg“—c§- ( ) n The final pulse rise. reaches - Z— in time t2 which is the average time for collection of positive ions. 1*... '3’ a?“ A Mass 1 l 1 I ”1?... o t t Fig. 4 - The shape of push when RC is infinite. (The time scale is distorted). - 1o - Another parameter for pulse shape is the clipping time. This time is equal to RC where R is resistance and C is capa- city of the system. In the previous discussion we chose RC equal to infinity. The results gave us Pig. 4. Usually the time constant (clipping time) is smaller than t and greater than t1 as we saw before, the pulse rises propoitional to time. This is shown in Pig. 5 by curve (a). As the electrons reach the collector, they decay through the circuit. This decay is an exponential function of time, {Wu 3 curve (b). The resulting pulse is shown as curve (c) in Pig. 5. (a) at... do 32) '17.“. h Fig. 5 - The shape of pulse when no < z 3} l 4 l. ‘5‘ 4» Tf'j’ 11_ 1L_ 8va L-i’ Iq L‘ 102 LJl A 11—1 [K a m 22 Pig. 8 - The Ionization Chamber (Scale 1/2) 2, 3 (sealed glasses) grid- negative 15- Source high voltage and collector res- 16- Plastic ring. pectively. 17- Mylar plastic. Gauge and valve. 18- Brass rod. Ground connection for guard ring. 19- Brass disk. Germice insulator. 20- Mylar plastic. Brass disk. 21- ”O" ring. Plastic holder. 22- Brass plate. Guard ring (grounded) 23- Brass plate. Collector Plastic ring. Brass ring. Grid Screen brass. -15- of ionization chamber. Three sealed glasses are inserted in the top disk of ionization chamber, and are connected to different electrodes inside the chamber. However, the guard ring of the collector is connected to the wall (ground) of the ionization chamber. The ionization chamber assemply is shown in Pig. 8‘.‘ The ionization chamber is also connected to the vacuum pump and gas tank. To fill the ionization chamber with argon gas, valves B, D and C, are closed but i and B are open. Pig. .9 ‘. The air is pumped out and after a while there is a rather good vacuum inside the ionization chamber. Then we close B and open E and D and C respectively. We can measure the pressure of the tank and the pressure of the ionization chamber by gauges G and 1'. To be sure that the amount of oxygen in- side the ionization chamber is very mall for each experiment we flush the ionization chamber with argon gas three times. To protect against leakage in the ionization chamber "0" rings are used between the top and bottom disks and the cylin- d F 6' er. %% E? ,A a» HM ' 4+6 e1”: Vm‘ MW Pig. 9'; - Connection between ioni- zati on chamber, vacuum pump and tank: of gas. -15- ['9‘ VJ“? —-—-..-- ‘Pw-I Limo éNVeW ‘_—‘¢”Mfib~l “Mt-‘ L.___i L___r—r:_:i ___i ___; iww‘ . Mu. ‘Eov ow» PIG. lO - Block circuit diagram. Remarks regarding Proportional Counters Ionization counters are usually of two different forms: parallel plates chambers; and cylindrical chambers with a collecting wire in.the center. These two chambers are shown in 113.111“. The field about the center wire in a cylindrical chamber is convergent, therefore this region is the one in which most of the secondary ionization takes place, since there the elec- trons have enough energy to ionize the gas molecules. (b) 313. 11--a cylindrical counter b parallel plates counter Calculation of multiplication in proportional counters is carried out5 under the following three assumptions: - 17 - 1) The photo-electric effects in the wall and other elements of counter are negligible. 2) Recombination of electrons with positive ions is negli- gible. 3) Combination of electrons with neutral atoms is also negli- gible. under these assumptions the multiplication is given by: ”wig-1.3512116") This equation applies for cylindrical counters. In is no. of gas molecules per unit volume. B is a constant, characteristic of the particular gas; for argon it 19 1.81. Vt is the threshold voltage. Vb is the applied voltage. b is the radius of cylindrical. a is the radius of center wire. - 1a - V. EXPERIMENTAL RESULTS a) Effect of Pressure 1. The mean free path of an ion in a gas is approximate- 6 '1y given by L.- ,‘where n is the number of atoms or men‘- molecules per cm3 and r is the average molecular radius. we can write this equation as follows: L. g. At low pressure n is small and therefore L is large. Some of the alpha par- ticles produce ionization electrons far from the sensitive volume and they are not collected by the collector. Convers- 1y, for high pressure, the mean free path is small and in gene- ral all electrons are created in sensitive volume and all are collected by collector. This is shown in Fig. 13 curve (a). From the above it is clear that there is a saturation effect which starts at some definite pressure. fl/ ; 113.112 - Field lines between plates and the effect of pressure on the mean free path of a particle. A \ {-9 nu~{chufihu (c (L) jflbtudm Pig. 13 - Pulse height against pressure. - 19 - 2. is we showed before, the drift velocity is propor- tional to E. On the other hand, the probability for recomp bination increases with decreasing velocity of electron. it higher pressures the electrons do not gain enough energy to be completely free of recombination. Thus we must conclude that at higher pressures some of the electrons recombine with posi- tive ions: curve (b), Pig. 1} shows this effect. How using the results of 1 and 2, it can be seen how the pressure affects the number of electrons collected, called pulse height. The resulting pulse shape is the product of the two component curves. Curve (c) shoves the form of the pulse as a function of pressure. rig. 14 shows how the pulse height is found to vary with pressure. b) Here we wish to discuss this effect for large size collec- tors. The field lines are shown in Pig. 15. The sensitive volume is greater than the one with the small collector. is we mentioned, for small size collectors the probability of collecting electrons at the collector is smaller at lower pressures. This is also true for large size collectors, but only at very low pressures. Since at these pressures some of the electrons move outside of the sensitive volume and cannot be collected. But at higher pressures all electrons move in the sensitive volume and are collected. we have shown that there is also another effect in.which pressure changes the drift velocity.. In this case curve (a) of Pig. 1} is almost a straight line and curve (b) does not change. This is shown in Pig. 16. Curve (c), the product of these two shows that for large size collectors at the higher pressures the pulse height decreases. ‘} (U \¥\O emrqu' Pig. 14 - No. of counts per minute vs. pulse height. (a) mu if ‘ \ if I} A) A i A it ‘ :WMAQW ’18. 15 - Th. fiflld lines in ’18. 16 - PulflO height large size collectors. as a function of pressure. -21- In Pigs. l7 s 18 some variation of pulse height as a function of voltage for two different pressures is apparent. @qu \‘9 “J“ HU- Lon K» (b) Venus \ V“ \ g fuo gnaw ice 0 fi' A7000 1 is“ 3:” Fig. 17 d l - Pulse height vs. voltage. 17 with- out guard ring, 18 with guard ring. a pressure about atmosphere. b 30 p.s.i. plus atmosphere. c) Collector with guard ring: we compare here the operation of the counter with.a guard ring and without the guard ring. The field line patterns are shown in.rig. 19. The field for the collector without guard ring focusses in the region close to the collector. /7~§\WH HT on “‘ in Fig. 19 - Field line pattern. (a) without guard ring. (b) with guard ring. The electrons move rapidly in the focussing region and there are no recombination effects. The pulse height increases as we increase the voltage up to 2000 volts, approximately. 'Then it remains constant. This region is by definition, the -22- ionization region since the pulse height does not change with the potential. Fig. 20 curve (a) and (b) show the change of pulse height with voltage in this region. The pulse height for a collector with a guard ring is smaller than without the guard ring, since some of the electrons are captured by the guard ring and some by ions. In the ionization chamber with guard ring, if we increase the voltage we eliminate the effect of recombination. Therefore, the two curves approach each other at still higher voltages. L team; I I ‘ e,- ° \ 1. 3 Fig. 20 - Pulse height as a function of voltage. a no guard ring. b guard ring. d) Comparison of Pig. 17 (a) and (b) shows that at lower pressures the chamber starts to be in the ionization condi- tion at a lower voltage. The difference is much more pronounced for the collector with guard ring, Big. 18, (a) and (b), since here some of the electrons will be collected by the guard ring. The difference between the pulse height for high and low pressure here is greater than the difference of pulse height for the collector without the guard ring. -23.. a) Resolution Again using diagram 14, it is seen that the majority of the electrons in (a) are in the same form. For each pulse the number of electrons-are approximately equal. This is not, however, true for curve (b), since as we mentioned previously some of the electrons are created closer and some farther away from the collector. Therefore...‘on1y some of the electrons which are ejected by an alpha particle will be captured by the collec- tor and'the remaining electrons will recombine. Thus the pulse height due to this ionization is small, and also resolution is low. This effect can be seen from Fig. 14 curve (a) and (b). In these two curves, the areas under the curves (which corres- pond to the total number of all electrons) are equal. -24- VI . DISCUSSION The original problem for our research was to find the branching ratio of thorium 232. The preparation of the source is difficult since thorium compounds The}: sham/V?” are hydros- copic. The difficulty arises in evaporating these compounds; also preparing a very thin layer of source is difficult with these materials, since after a short time they absorb water from the air and are no longer smooth layers. The evaporation of thorium compounds must be done in vacuum aparatus at very low pressures. In gentle heating, the water of hydration is removed. The pressure guage shows the increase of pressure which is due to the water. Then we stop heating the source until very low pressure is reached again. We repeat this several times till there was no more water in the source. Then we heated up to very high temperatures and therefore the compound evaporated and deposited on the holder. But as we mentioned before it is very difficult to protect these compounds from the air. It seems that it is best to evaporate pure thorium metal since it is not hydroscopic. But the melting point for thorium is 18400, and the boiling point is 4500°. Therefore it is difficult (but possible) to deposit thorium on the holder. We deposited a very thin thorium 232 metal source on the mylar holder and we obtained only about 10 counts per minute - 25 - for alpha decays. This number is due to wholy different decays of thorium 232 and its daughters. The size of pulse after the amplification was about 15 to 20 volts and the noise was about 12 to 15 volts and sometimes more. The thorium spectrum did not have good resolution for different peaks, therefore we could not continue this problem. Leakage Problem: Usually a small amount of electro-negative atoms will cap- ture all electrons which are emitted by ionization. If the air leaks through the chamber or through the rubber hole connecting the chamber to the vacuum pump, no pulses are produced. The question that arises is: Can this chamber work as a proportional counter? To answer this question, we use the eque ation on'page 17, ‘ ‘ m [P ,1 for gportignal counter M)1 or 7771; . The experiment showed that V1 is greater than 4000 volts. To increase the field in- tensity between the plate we decreased the distance of the electrode to 8 mfllimeter for 4000 volts, but the chamber was still in the ionization region. However, using high voltages, 4000 volts or greater, brings forth a new problem. This prob- lem is arcing between the electrodes. Furthermore, it is not practical to make a proportional counter with such high voltage across the electrode. Therefore this apparatus is an ionization counter and cannot be operated as a proportional counter for the above reason. VII. BIBLIOGRAPHY Walske, n.0,, Physics geyigy, 1952, 88, 1283 Rossi e Staub, Ionization Chamberlg Counters , Page 40 Bumemann, 0., N.R.C.Rgpg;£_, P.D. 285 Smith, 2.1., Physics m, 1930, 36, 1293 garley, T.A., Canadian Journal of Physics, Aug. 1960, e 1059 Sergre, E., ggperimental Nuclear Physics, Vol.1, 177 PHYSIcs-MAIH LIB. HICHI RIES umum