Progress in Nuclear Physics

Progress in Nuclear Physics, Volume 9 is a collection of six papers dealing with nuclear physics. The monograph discusses the significance of visual techniques in the development of high energy physics; the development of solid semiconductor detectors; theoretical techniques of high-energy beam design; and the structure analysis of collision amplitudes. The book explains that the spark technique is preferred over the bubble chamber when used in experiments with continuous sources of radiation. The use of germanium and silicon semiconductor detectors has enabled advances in studies on protons, alphas, and heavy ions, through fast response time, ease of construction, and simple power requirements. The problems encountered in beam design and effective uses of large accelerators are discussed extensively. For example, the book explains that inefficient beam practice can diminish intensity factors in the accelerator, so different beam designs such as the quadruple and the separated beam design as well as the pulsed high-field magnets are considered. The link between theory and experiment of collision amplitudes is reviewed, including structure analysis in the physics of elementary particles. The resulting processes after strange particles such as K-mesons interact with nuclei, instead of reacting with individual nucleons, are briefly noted. The text also includes discussion on the essential electromagnetic properties of the muon and the different research by Crowe, Charpak et al., and Hughes et al. about this subject. This collection of papers will prove useful for nuclear physicists, scientists, and academicians in the field of nuclear physics.

• Language: English
• Publisher: Elsevier Science
• Release date: Oct 22, 2013
• ISBN: 9781483164700

Book preview

PROGRESS IN NUCLEAR PHYSICS

O.R. FRISCH, O.B.E., F.R.S.

Cavendish Laboratory, Cambridge

PERGAMON PRESS

Table of Contents

Cover image

Title page

Copyright

Chapter 1: SPARK CHAMBERS

Publisher Summary

1 INTRODUCTION

2 PRINCIPLES OF OPERATION OF TRIGGERED SPARK CHAMBERS

3 THE GASEOUS PROCESSES INVOLVED IN SPARK CHAMBER OPERATION

4 CONSTRUCTION OF SPARK CHAMBERS

5 ELECTRONICS

6 PHOTOGRAPHIC RECORDING OF TRACKS

7 ACCURACY OF TRACK LOCATION

8 TYPICAL APPLICATIONS OF SPARK CHAMBERS

ACKNOWLEDGEMENT

Chapter 2: SEMICONDUCTOR COUNTERS

Publisher Summary

1 INTRODUCTION

2 INSULATORS AND SEMICONDUCTORS

3 CONDUCTION COUNTERS

4 SEMICONDUCTOR JUNCTION DETECTORS

5 PULSE AMPLIFIERS FOR SEMICONDUCTOR DETECTORS

6 APPLICATIONS OF SEMICONDUCTOR COUNTERS IN NUCLEAR PHYSICS

7 CONCLUSIONS

Chapter 3: THEORETICAL TECHNIQUES OF HIGH-ENERGY BEAM DESIGN

Publisher Summary

1 INTRODUCTION

2 BEAM DYNAMICS

4 QUADRUPOLE SYSTEM DESIGN

5 SEPARATED BEAM DESIGN

6 SPECIAL COMPONENTS

ACKNOWLEDGEMENTS

Chapter 4: INTRODUCTION TO THE STRUCTURE ANALYSIS OF COLLISION AMPLITUDES

Publisher Summary

1 INTRODUCTION

2 MATHEMATICAL INTRODUCTION

3 FORM FACTORS FOR SCALAR PARTICLES

4 SCATTERING OF EQUAL MASS PARTICLES

5 UNITARITY AND EQUATIONS FOR PARTIAL WAVE AMPLITUDES

6 PION–NUCLEON SCATTERING

7 ASYMPTOTIC BEHAVIOUR OF AMPLITUDES

8 DISCUSSION

READING GUIDE

Chapter 5: THE INTERACTION OF STRANGE PARTICLES WITH NUCLEI

Publisher Summary

1 BASIC INTERACTIONS OF STRANGE PARTICLES WITH NUCLEONS

2 CAPTURE OF NEGATIVELY CHARGED PARTICLES BY ATOMIC NUCLEI

3 THE INTERACTION OF K−-MESONS WITH NUCLEI

4 HYPERFRAGMENTS AND THE INTERACTION OF Λ°-HYPERONS WITH NUCLEI

5 THE INTERACTION OF OTHER STRANGE PARTICLES WITH NUCLEI

6 THE USE OF STRANGE PARTICLES AS PROBES FOR THE STUDY OF THE NUCLEUS

ACKNOWLEDGMENTS

Chapter 6: ELECTROMAGNETIC PROPERTIES OF THE MUON

Publisher Summary

INTRODUCTION

THEORETICAL BASIS

EVIDENCE FOR SPIN

MAGNETIC MOMENT

MEASUREMENTS OF THE MUON MASS

DIRECT MEASUREMENT OF THE ANOMALOUS MAGNETIC MOMENT

THEORETICAL CONSEQUENCES OF (g-2) EXPERIMENT

BEST VALUES FOR THE MUON MASS AND CHARGE

HYPERFINE SPLITTING OF MUONIUM BY MICROWAVE RESONANCE

ELECTROMAGNETIC PROPERTIES AT HIGH ENERGY

CONCLUSION

ACKNOWLEDGMENTS

CONTENTS OF PREVIOUS VOLUMES

NAME INDEX

SUBJECT INDEX

Copyright

PERGAMON PRESS LTD.

Headington Hill Hall, Oxford

4 & 5 Fitzroy Square, London, W.1

PERGAMON PRESS INC.

122 East 55th Street, New York 22, N.Y.

GAUTHIER-VILLA RSED.

55 Quai des Grands-Augustins, Paris 6

PERGAMON PRESS G.m.b.H.

Kaiserstrasse 75, Frankfurt am Main

Distributed in the Western Hemisphere by

THE MACMILLAN COMPANY • NEW YORK

pursuant to a special arrangement with Pergamon Press Limited

Copyright © 1964

PERGAMON PRESS LTD.

Library of Congress Card Number 51–984.

MADE IN GREAT BRITAIN

1

SPARK CHAMBERS

J.G. Rutherglen

Publisher Summary

The spark chamber provides a visual technique with a time resolution of less than 10−6 seconds and a maximum cycling rate that may be as high as 50 per seconds. It may be triggered by counters that select the event desired to record. There are many applications in which the spatial information given by a spark chamber is perfectly adequate and the short resolving time and rapid cycling rate make it possible to perform experiments that would be impracticable with any other visual technique. Also, spark chambers are comparatively simple to construct and require only a modest amount of ancillary equipment. In principle, the scintillation chamber offers many of the advantages of both the spark chamber and the bubble chamber, but it seems likely to have only limited application in the immediate future because of the very considerable technical difficulties involved. This chapter provides an overview of the principles of operation of triggered spark chambers, the gaseous processes involved in spark chamber operation, construction of park chambers, and their typical applications. An important criterion in the design of a spark chamber system for a particular experiment is the accuracy with which the sparks define the path of the triggering particle.

CONTENTS

1 INTRODUCTION

2 PRINCIPLES OF OPERATION OF TRIGGERED SPARK CHAMBER

3 THE GASEOUS PROCESSES INVOLVED IN SPARK CHAMBER OPERATION

3.1 Basic mechanism of spark formation

3.2 Formative time

3.3 Electric field requirement

3.4 Effects of variation in gap spacing

3.5 Methods of clearing and effects of impurities

3.6 Recovery time

3.7 Effects of primary ionization density

4 CONSTRUCTION OF SPARK CHAMBERS

5 ELECTRONICS

5.1 General requirements

5.2 The high-voltage pulser

5.3 Methods of connection to the spark chamber plates

6 PHOTOGRAPHIC RECORDING OF TRACKS

7 ACCURACY OF TRACK LOCATION

8 TYPICAL APPLICATIONS OF SPARK CHAMBERS

REFERENCES

1 INTRODUCTION

VISUAL techniques have always played an important part in the development of high energy physics, commencing with the expansion chambers used in the early cosmic ray experiments and followed by the development of nuclear emulsions, diffusion cloud chambers, bubble chambers and scintillation chambers. For experiments with high energy accelerators the bubble chamber has become perhaps the most important single technique, mainly because it has the property of giving very accurate spatial resolution on particle tracks in large volumes of liquid. Furthermore, there is a reasonably large choice available for the atomic number of the liquid and, in particular, liquid hydrogen may be used to study the interaction of incoming particles with protons.

However in certain applications three limitations of the bubble chamber become important. First, its time resolution, which is of the order of 10−3 sec, is long compared with that of counter techniques. Second, it cannot be triggered by the particles which it is to detect and third, its maximum cycling rate is of the order of 1 c/s. It is therefore well suited to operating in conjunction with pulsed high energy accelerators of moderate repetition rate, provided it can be supplied with a beam of a few selected particles per machine pulse containing a low background of unwanted particles. This often involves the use of complicated and expensive beam handling systems for momentum and mass analysis. The bubble chamber cannot, of course, be used effectively for operation with continuous sources of radiation, such as cosmic rays.

mm and 2 per cm respectively, are considerably inferior to that of a bubble chamber. Furthermore, the tracks are only sampled in the gaps between the plates of the chamber, so that the origin of any event, such as a nuclear interaction, can only be determined to within one plate thickness and any secondary particle must have a range of a least two or three plate thicknesses to be effectively recorded.

However there are many applications in which the spatial information given by a spark chamber is perfectly adequate, and the short resolving time and rapid cycling rate make it possible to perform experiments which would be impracticable with any other visual technique. Also, spark chambers are comparatively simple to construct and require only a modest amount of ancillary equipment.

It must be noted that, in principle, the scintillation chamber offers many of the advantages of both the spark chamber and the bubble chamber, but because of the very considerable technical difficulties involved it seems likely to have only limited application in the immediate future.

A review of the historical development of spark chambers has been given by ROBERTS (1961a) and it is felt that it is unnecessary to repeat that in this article.

2 PRINCIPLES OF OPERATION OF TRIGGERED SPARK CHAMBERS

The essential components of a triggered spark chamber are illustrated in Fig. 1. The passage of a particle through the chamber is detected by the counters S1 and S2, whose outputs are fed into a coincidence circuit. The output of the coincidence circuit triggers a high voltage pulser, which supplies a pulse of several kilovolts between alternate plates of the chamber. The amplitude is arranged to be well above the d.c. sparking voltage, so that the electrons which have been produced by the particle in its passage through the gas between the plates initiate electron avalanches which finally grow into a high current spark whose position is very close to that of the particle track. Thus in a multiplate chamber a series of sparks are produced which define the particle track.

FIG. 1 Basic components of a triggered spark chamber.

If the pulser is triggered at random, then a spark will not occur unless an electron is already present in the gap or one is produced at some time during the pulse by some random process, such as cold emission, photo-electric effect or the passage of an ionizing particle. The probability of such a process is usually small if the pulse length is only of the order of 1 μsec. In order to reduce the probability of the presence of an electron at the time the pulse occurs it is desirable to apply a small d.c. potential difference between the plates, usually known as the clearing field, to remove electrons produced by the random processes mentioned above. The limit to the magnitude of this clearing field is set by the fact that it must not be so high as to remove electrons produced by the triggering particle before the high-voltage pulse is applied to the plates. Typically the clearing field is arranged so that the transit time t1 for an electron to cross the gap under the action of the clearing field is about twice the delay time t2 between the passage of the particle and the initial rise of the high-voltage pulse. This latter time may easily be made as short as 0·25 μsec, so that the clearing time may be of the order of 0·5 μsec. Any particles which pass through the chamber during the interval t1 immediately prior to the initiation of the high-voltage pulse will, in general, produce sparks along their tracks and hence t1 is the sensitive time, or resolving time, of the system.

Another important parameter of a spark chamber system is the recovery time, that is, the time which must elapse after the operation of the chamber before it can satisfactorily be operated again. The high-voltage pulser will usually require some time to recharge its storage system, but this can be made small if necessary. In fact the recovery time is limited by the chamber itself because of the time required to remove all the products of the previous spark which might lead to reignition. The electrons are rapidly removed by the clearing field but positive ions and metastable neutral atoms remain for much longer times and have a small, but finite, probability of producing electrons by secondary processes.

3 THE GASEOUS PROCESSES INVOLVED IN SPARK CHAMBER OPERATION

3.1 Basic mechanisms of spark formation

The process of the development and maintenance of an electrical discharge in a gas is extremely complicated and has been discussed in detail by many authors (e.g. MEEK AND CRAGGS, 1953; LOEB, 1939, 1955). We shall outline the mechanisms which are relevant to the development of a spark discharge in a gas under the conditions which normally occur in spark chamber operation.

A discharge is normally initiated by one or more primary electrons which are accelerated by the applied field and then produce further electrons by ionizing collisions with atoms of the gas. Thus an electron avalanche builds up, moving towards the anode with a velocity of the order of 10⁷ cm/sec. The positive ions which are produced at the same time move towards the cathode with velocities of the order of 10⁵ cm/sec and may therefore be considered to be almost stationary in comparison with the electrons. The initial growth of the electron avalanche is usually defined in terms of Townsend’s first ionization coefficient α, which is the number of new electrons produced per cm of path by each electron. Thus a single primary electron will have produced eαx electrons when the avalanche has moved a distance x in the direction of the applied field. A similar number of positive ions will be left in the path of the avalanche, but because of diffusion effects the electron and positive ion populations will be somewhat intermingled at any given time.

The process continues until either (a) the electrons reach the anode or (b) the effects of the space charge of electrons and positive ions is sufficient to seriously modify the field and alter the mechanism of growth of the avalanche.

In case (a) some secondary process, such as the liberation of secondary electrons by positive ions arriving at the cathode, is necessary if the discharge is to be maintained. Townsend’s second ionization coefficient, γ, is defined as the number of secondary electrons produced at the cathode per electron in the primary avalanche. It is then easily shown that, to a first order, the condition for the discharge to be self-sustaining is that γeαd > 1, where d is the distance between the anode and cathode plates. This type of discharge is produced in gases at low pressures or with very small electrode spacings. The formative time of such a discharge is comparatively long, of the order of 50–1000 μsec, because of the long transit time of the positive ions. Also, because of lateral diffusion of the positive ions, the final discharge is diffuse or consists of a number of separated sparks. For both these reasons this type of discharge is not a satisfactory one for the operation of spark chambers.

In case (b) we have the condition where the value of αd is so large that the local electric field Es, produced by the electrons and positive ions in the region of the head of the avalanche becomes comparable with the applied field E before the avalanche reaches the anode. Under these conditions the field at the centre of the avalanche head, where the density of both electrons and positive ions is high, becomes reduced and recombination becomes more probable, with the consequent emission of photons. Some of these photons produce photo-electrons in the gas surrounding the avalanche, which produce fresh avalanches. The greatest multiplication in these auxiliary avalanches will occur along the axis of the main avalanche where the space charge field augments the applied field. In this way the process develops as a self-propagating atreamer which rapidly spreads in both directions towards the electrodes and forms the spark channel.

3.2 Formative time

The formative time of a discharge of type (b) is much shorter than that of type (a) because no secondary mechanisms with a long delay are involved. Also, since the primary avalanche defines the spark channel, the spark is much more closely aligned with the primary ionization and multiple sparks are less likely to occur. It should also be noted that, since no secondary processes at the electrodes are involved, the characteristics of the spark are independent of the electrode materials and depend only on the properties of the gas.

Experimental measurements of the formative times of sparks under these conditions of high impulsive overvoltage have been made by FLETCHER (1949) in air and by FISCHER AND ZORN (1961) in He, Ne and A. FLETCHER calculated the formative time on the assumption that it consists almost entirely of the time for the primary avalanche to reach such a size that the space charge field is of the same order of magnitude as the applied field, and obtained values in reasonably good agreement with his experimental values. These calculations were criticized by DICKEY (1952) on the grounds that the number of electrons calculated by FLETCHER to reach the space charge criterion was of the order of 10⁸, whereas the number which finally had to be formed to produce the observed currents and to produce a significant change in the electrode potentials was of the order of 10¹³, and it was difficult to see how the streamer mechanism could accelerate the original exponential growth of the discharge. However, these considerations do not make an order of magnitude difference to the calculated formative times. If we make the simplest possible assumption that the Townsend coefficient α and the average velocity of the electrons in the direction of the applied field v, are constant, then the number of electrons produced in time t from a single primary electron is N = eαvt, or logN = αvt. If we take N = 10⁸ we have αvt = 18·4 and if we take N = 10¹³ we have αvt = 29·9. Thus the formative time is only about 60% greater in the second case.

The results of FISCHER and ZORN are shown in Fig. 2. Their calculated curves are based on the method of DICKEY, i.e. it is assumed that the ionization increases exponentially with time until the voltage across the gap becomes significantly reduced by the effect of the space charge. FISCHER and ZORN show that this criterion gives

FIG. 2 Comparison of the calculated and observed formative times τ as a function of applied field E for gap lengths of 0·062, 0·26 and 0·52 in. (FISCHER AND ZORN, 1961).

where τ is the formative time, Iext the current supplied by the external circuit, R the source resistance, C the chamber capacity, d the gap length and N0 the number of primary electrons. They state that, for the parameters of their experiment and with N0 = 1, this equation can be written

where the deviation of the constant represents its maximum variation for He, Ne and A over the range of voltages used. It should be noted that, although the value of the constant applies strictly only to the parameters used by FISCHER and ZORN, it will give the correct order of magnitude of the formative time for any similar spark chamber system because of the logarithmic variation of the constant with these parameters.

3.3 Electric field requirements

We can calculate approximately the conditions under which we shall get the type of discharge with very short formative times. The necessary condition is that the distance required to form the complete electron avalanche shall be less than the gap width. If τ is the formative time, then the length of the avalanche is x = vτ and hence, using the approximate formula of FISCHER and ZORN, x = 29·5/α. This assumes that the avalanche builds up in a straight Une until the final current limited by external conditions is reached. On the streamer theory it is only necessary that an avalanche of the order of 10⁸ electrons must be built up without reaching the electrodes, the remaining ionization then being produced by secondary photon-induced avalanches. This would give a critical length x = 18·4/α. The critical lengths given by these two expressions are shown in Fig. 3 as a function of the electric field E at atmospheric pressure, using the values of α given by DRUYVESTEYN AND PENNING (1940) and KRUITOFF AND PENNING (1937) for He, Ne, A and Ne + 0·01 A.

FIG. 3 for formation of avalanches of 10⁸ and 10¹³ ion pairs, as a function of applied field E at atmospheric pressure.

FISCHER and ZORN observed an appreciable departure of their experimental values for the formative time from the calculated ones for their smallest gap of 0·16 cm. It will be seen from Fig. 3 that this would be expected to occur in the region of E < 20–24 kV/cm in A, and E < 13–18 kV/cm in He, in approximate agreement with the results of FISCHER and ZORN. For the larger gap spacings departures are not so clearly evident, but it seems reasonable to use the curves of Fig. 3 to give the order of magnitude of the minimum electric field necessary for satisfactory use of a spark chamber in the pure gases, depending on gap spacing.

However, the case of Ne–A mixtures warrants some further discussion. The large values of α which are observed at low electric fields are due to the ionization of A atoms (ionization potential 15·8 V) by metastable Ne atoms with excitation energies of 16·6 V, which are produced in large numbers at low electric fields (DRUYVESTEYN AND PENNING, 1940). The effect is greatest at concentrations of between 10−4 and 10−2 parts of A. However these values of α are measured in static experiments and in the rapid breakdown processes with which we are concerned it is likely that only a fraction of the metastable Ne atoms will have time to diffuse and collide with an A atom, so that the effective value of α will be smaller than that measured in static experiments and nearer to that of pure Ne. It is difficult to draw any quantitative conclusions on this point from spark chamber data since such small quantities of A are necessary to produce this effect and it is likely that Ne of the purity normally used will contain significant quantities of A. Most workers find that the threshold voltage for operation in Ne is somewhat lower than for He, which one would not expect if the Ne was really pure. ANDERSON (1961) states that the working voltage with Ne + 0·01% A is much lower than that for pure Ne. The fact that the values of the formative times in Ne measured by FISCHER and ZORN (Fig. 2) are a little below the calculated values for pure Ne may also be due to this effect.

3.4 Effects of variation in gap spacing

If one considers only a single gap with one particle track through it, then a spark will be produced with almost 100% efficiency for a large range of applied electric fields above threshold. Hence, in this case, variations in the electrode spacing are of no particular importance. However, if two or more gaps are connected in parallel, or if two tracks pass through the same gap, then variations in spacing will produce variations in formative time. This means that when one spark has developed sufficiently to seriously reduce the electrode voltage, all the sparks will stop growing and the ones with longer formative times may be considerably less intense. We can calculate the order of magnitude of this effect in terms of the basic mechanism discussed in Sections 3.1 and 3.2. Suppose that two tracks exist at points where the electrode spacings are d and d – Δd, so that the electric fields are E and E + ΔEThe Townsend coefficients α and α + Δα will be some function of Ecan be evaluated from the experimental data on oc as a function of E. Also if we make the usual assumption that the average velocity of the electrons v , then the velocities v and v + Δv will be related to E Now if N1 and N2 are the numbers of electrons produced in the two sparks at time τ, we have

and

assuming equal number N0 of primary electrons in each track.

Thus

The results of FISCHER and ZORN indicate that we can put αvτ = 30 for typical spark chamber operation, so that

To a first approximation we can assume that this gives the ratio of the brightness of the two sparks. Values of β as a function of E, derived from the experimental data for various gases are shown in Fig. 4. If we take a typical case of pure Ne at E = 20kV/cm, we see that β in. electrode spacing a variation of 0·0042 in. will give sparks whose intensities differ by a factor of 2·7. Furthermore, the intensity ratio increases exponentially with Δd/d, so that for larger variations in spacing the weaker spark will rapidly become undetectable. For pure gases, β decreases with increasing E, so that from this point of view it is desirable to operate at high fields. The lowest values of β are obtained in Ne–A mixtures, but these may not be fully realized for the reasons discussed in Section 3.3.

FIG. 4 as a function of applied field E at atmospheric pressure.

There is at present little quantitative evidence with which to compare these predictions, but it is the experience of all workers (e.g. KOESTER, 1961) that small spacing errors result in large variations of relative spark intensities. It is certainly our own experience that commercial grade Ne, which contains a significant quantity of A, gives the most uniform sparks in a multiplate chamber. Helium is slightly worse in this respect and A is considerably worse, as would be expected from Fig. 4. We may also note that when MEYER AND TERWILLIGER (1961) operated two chambers with a 15% difference in spacing, connected in parallel, sparks never appeared in the one with larger spacing. The above theory would predict an intensity ratio of ∼ e⁹ for this case.

Some improvement can be effected for single tracks in multiplate chambers by decoupling the high-voltage connections to the plates, as will be discussed later, but for good operation, particularly if multiple tracks are to be observed, it seems desirable to construct chambers so that the tolerances on gap spacings are less than 1%.

3.5 Methods of clearing and effects of impurities

The usual method of clearing old tracks and other unwanted ionization is to apply a electric field to the chamber. The direction of this clearing field is usually made opposite to that of the pulse field, so that the electrons from the triggering tracks will have the full gap spacing in which to form the avalanche. The magnitude of the clearing field is chosen to give the required clearing time, with the proviso that it must not be shorter than the pulse delay time.

The clearing time is usually measured by inserting a variable delay between the coincidence circuit and the pulser and measuring the chamber efficiency as a function of the total delay time. Such measurements have been made by CRONIN AND RENNINGER (1960), BEALL et al (1960), CULLIGAN et al. (1961). A typical family of curves, as measured by CRONIN and RENNINGER for Ne, is shown in Fig. 5. If one calculates the electron drift velocities from such measurements it is a common feature of all the data that the velocities are much higher than would be expected for pure gases. For example in Fig. 6 the velocitites calculated from the results of BEALL et al and CULLIGAN et al for A, and CRONIN and RENNINGER for Ne, are plotted against the clearing field, together with the measurements of BOWE (1960) for pure A and pure Ne. The effects of impurities have been investigated by a number of authors and to illustrate the order of magnitude of the effects of molecular gases a curve for A + 0·5% CO2, measured by NAGY et al. (1960), is included in Fig. 6. It is clear that, since such small quantities of impurities produce such drastic changes in electron mobility, it is unlikely that under spark chamber conditions the gases will ever be really pure. However, since the effect of impurities is usually to increase the mobility and hence reduce the clearing time, their presence is no disadvantage, except that it is not possible to predict exactly what clearing field will be required to produce a given clearing time.

FIG. 5 Efficiency of a spark chamber filled with Ne at a pressure of 1·3 atm, as a function of delay time and clearing field. Inset shows the variation of sensitive time with clearing field (CRONIN AND RENNINGER, 1960).

FIG. 6 Comparison of electron drift velocities at atmospheric pressure calculated from spark chamber clearing times in Argon (a and b) and Neon (c) with measurements by other methods on pure A (d), pure Ne (e) and A + 0·5% CO2 (f) (a, CULLIGAN et al. (1961); b, BEALL et al. (1960); c, CRONIN AND RENNINGER (1960); d and e, BOWE (1960); f, NAGY et al. (1960)).

If certain impurities are present in comparatively large quantities the efficiency of a chamber may be seriously reduced. The most likely impurity from this point of view is air, due to leaks or out-gassing of the materials of the chamber. The effect is almost certainly due to electron attachment by molecules of electro-negative gases, so that they are no longer able to initiate avalanches. % oxygen or 5% CO2 did not affect the efficiency appreciably, but larger quantities did reduce the efficiency. This was with a delay time of 0·3 μsec. PATRICK AND PATERSON (1961) have investigated the efficiency as a function of delay time for various percentage additions of O2. A typical curve is shown in Fig. 7 for the addition of 1·3% O2 to a spark chamber in which the operating gas was a mixture of 80 % He and 20 % A. It will be seen that the efficiency is not seriously reduced at delay times of less than 0·5 μsec, but falls to 50% at 2 μsec.

FIG. 7 Spark chamber efficiency as a function of delay time for 1·3% O2 contamination in a mixture of 80% He + 20% A (PATRICK AND PATERSON, 1961).

The use of controlled quantities of electro-negative impurities provides an alternative, though less precise, method of controlling the resolving time of a spark chamber. It is useful when the use of an electric field is not possible. For example, it is known that a spark chamber may be operated satisfactorily with a magnetic field parallel to the plates (BEALL et al., 1960, O’NEILL, 1961) for the purpose of momentum measurements, but a d.c. clearing field merely causes the electrons to migrate in the direction E × B, parallel to the plates. In this case a small quantity of impurity, such as O2 or