ANNUAL REPORTSON THEPROGRESS OF CHEMISTRY.RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.THE study of subatomic phenomena has been made possible through thedevelopment of a number of special research tools. Some of them have beenknown for several years, but others have been invented quite recently. Onthe whole, the methods of nuclear physics have been fairly stationary forsome time, and it is felt that a survey might be welcome to chemists.Only few subatomic phenomena occur spontaneously in Nature, viz.,those connected with natural radioactivity and with cosmic radiation. Allothers have to be produced by artificial means, and because of the greatstability of all nuclear structures it is necessary to concentrate large energyon a single particle. The only known way of doing this is to place anelectrically charged particle in a strong electrical field in which it getsaccelerated.The ultimate kinetic energy achieved is equal to the potentialdifference through which the particle moved in the process of acceleration,multiplied by its charge. Since the charge is, in general, equal to the chargeof the electron (or a small multiple of it) it is usual to give its energy inelectron volts (eV. or e.v.). The energies concerned in nuclear reactions aremostly of the order of a million electron volts (MeV), i.e., the energy of anelectron which has fallen through a potential difference of lo6 volts. Hence,these energies are vastly greater than those involved in chemical reactions :1 e.v. corresponds to about 23,000 cals.per mole, and the mean kinetic energyof a molecule a t room temperature, Qlcl', is only about 0.04 ev.High-tension Apparatus.-Ions are produced in a special discharge tubeand are then accelerated in a high vacuum by the application of a highpotential. The acceleration tube generally consists of a number of shortglass (or porcelain) sections separated by metal plates. Each insulatingsection thus carries only a part of the total voltage, and the danger of electricalbreakdown is reduced. Each metal plate has a short length of metal tube inits centre, which nearly touches the metal tube belonging to the next plate.Thus the beam runs down the axis of one long metal tube with a number ofshort gaps in it, and deflection of the beam caused by fluctuating charges onthe insulators is minimised.The unevenness of the field (strong a t the gaps,weak in the tubes) has a desirable focusing action on the beam. The ionsource has to contain a certain pressure (of the order of 0.01 mm.) of thegas in question (hydrogen, deuterium, or helium) and powerful pumps(mostly oil diffusion pumps) are required to keep a sufficiently high vacuumin the accelerating tube.For the production of the necessary high potential there are mainly tw6 RADIOACTIVITY AND SUB- ATOMTC PHENOMENA.ways. The one consists in the use of a transformer, with subsequent recti-fication of the alternating potential (the use of “rt~w,” that is unrectified,A.C. has considerable drawbacks and has been abandoned). By using severalrectifier tubes in cascade arrangement, it is possible to get a D.C.potentialseveral times (say 10 times or more) the peak voltage of the transformer.The other system is essentially the same as in the old electrostaticmachines. Sometimes rotating discs are used, but in the most common form-developed by Van der Graaff-an endless insulating belt (of paper or rubber-ised silk) is stretched over two pulleys which are rotated as fast as possible.One.of the pulleys-say, the lower one-is at ground potential, while theother is inside the H.T. electrode, a large metal box, mostly spherical.Electric charge is “ sprayed ” on to that side of the belt which moves up-wctrds by means of a small rectifier and a number of sharp points whichalmost touch the belt.On arrival at the H.T. electrode the charge is takenoff the belt by means of a similar “ comb.” The H.T. electrode thus getsgradually charged up and its potential goes on increasing either until there isa breakdown (spark) to ground, or until the insulation losses just balance theflow of charge carried by the belt.Both systems have their advantages and drawbacks. The transformer-rectifier arrangement gives large currents (several millirtmps.) and is stableand easy to control. On the other hand it is expensive, and there is alwaysa “ ripple ” (a remnant of A.C.). A Van der Graaff generator can be built atfairly small cost and the potential is quite “ smooth ” ; but it gives no morethan a few tenths of a milliamp. at most, and as the potential depends on theinsulation losses, it takes some skill to get a constant potential.It seems, tobe the general tendency to prefer reotifiers for less than, say, 1 M Y andelectrostatic machines for higher potentials.The limiting potential for each machine depends on the amount of spark-ing from the H.T. electrode to ground. Sparking along the insulator can beinhibited by appropriate design, and the sparking potential may be raised byabout 50% by adding a small amount of dichlorodifluoromethane (“ Freon ”)to the air. For considerably higher potentials it is usual to place the wholesystem, generator and discharge tube, in a tank filled with compressed air.The largest plant of this kind has been built in Pittsburgh and consists of apear-shaped container 47 ft.long, placed with its thin end downwards. TheH.T. electrode i s at the centre of the thick portion of the pear, and the dis-charge tube, 6ome 30 ft. long, extends from there to the bottom end. Thisgenerator can produce up to about 4 Mv. In Wisconsin another high-pressure electrostatic generator has been in use for some years.2The Cyclotron.3LIh this apparatus (invented and developed by E. 0.Lawrence and originally called the magnetic resonance accelerator) the useof excessively high potentials is avoided by the ingenious expedient of1 W. H. Wells, R. 0. Hrtxby, W. E. Stephens, and W. E. Shoupp, Phyaicat Rev.,1940, 68, 162.R. G. Herb, D. B. Parkinson, and D. W. Kerst, cibid., 1937, 51, 75.See W. B. Mann, “ The Cyclotron,” Methuen & Co,, 1940; E.0. Lawrence andD. Coolssey, P h y k E Rev., 1836, 50, 1131FRISCH. 7aooelerating the ions many times by the 8ame potential. The accelerationtakes plaoe between two “ Dees,” hollow semicircular electrodes, like a pillbox cut in two along a diameter. The “ Dees ” are placed in a strongmagnetio field which benda the path of the ions and thus forces them to passrepeatedly from one electrode to the other. A high-frequency potential ieapplied to the ‘‘ Dees ” and synchronised with the motion of the ions so thatthey get aooelerated every time they pass from one “ Dee ” to the other. Asthe ion8 get faster the curvature of their path in the magnetio field decreases,an& they spiral outwards until they (or rather, some of them) esoape througha slit in one “ Dee ” and are directed on to the target by means of a defleotingelectrode.* All this happens inside an &-tight container (called the tank)which ia oontinudy exhausted by powerful pumps.The limiting speed of the ions is given by the strength and diameter ofthe magnetic field.In most cyclotrons the region over which the field issuffioiently homogeneous has a diameter of about 2-23 ft. ; with a fieldof 16000 gauss the limiting energy then becomes about 8 MeV for deuteronsand 16 MeV for doubly-charged helium ions (a-particles, sometimes calledhelions). It would appear that by increasing the size of the magnet onecould produce ions of any desired energy, but as the speed of the ions ceasesto be very small compared to the velocity of light their mws increasesaccording to relativity theory and it becomes inoreasingly difEcult to maintainthe necessary synchronism between their motion and the high-frequenoyvoltage which.accelerates them.The Berkeley 60” cyclotron (60” pole pieoediameter) produces deuterons of 16 MeV (and .a-partiolea of 32 MeV), andyet bigger cyclotrons are under construction. With protons the relativisticdifjiaulty is felt already in cyclotrons of the usual size, and proton energiesare on the whole limited to about 10 MeV or less.A recently developed instrument, called the induction accelerator orbetatron> produces electrons (“ artificial p-particles ”) of energies up to20 MeV. Electrons circulate inside a dough-nut shaped evacuated container,which encloses the iron core of a transformer. On each revolution theelectrons gain a few ev.energy, corresponding to the voltage which would begenerated in a secondary yinding of one turn; but during one half-cycle ofthe A.C. (about 1000 oyules per sec.) which is fed to the transformer, theelectrons make a large number of revolutions and thus accumulate very largeenergy. The instrument is still being improved and its future uae in nuclearphysics is diffioult to predict.N e e o n 8ources.-Radioactive bodies as sources of particles for nuclearexperiments have on the whole been superseded by artificial sources, becauseof the enormously greater intensity of the latter (an ion current of loop amp.,as is customary with artificial sources, means 6 x 1014 particles a, second,whereas 1 g.of radium emits only 2 x loll a-particles per second). TheD. W. Kerst, Rev. Sci. Instr., 1942, 13, 387.* Or else the substance to be bombarded may be placed at the end of a ‘‘ probe ”which is inserted between the “ Dees ’’ so as to intercept the ions before they reach theedge of the “ Dees,” This gives greater intensity but has certain drawbacks8 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.main use of radioactive sources nowadays is for the production of moderateneutron intensities, their smallness, constancy, and easy manipulation thenbeing of great advantage. The commonest type of neutron source is a tubepacked with beryllium powder and filled with some hundred millicuries ofradon. Such a source emits about 20,000 neutrons per sec.per millicurie,some of which have energies up to 10 MeV. The source decays, of course,with the half-life (5.82 days) of radon. A practically constant source(half-life 1700 years) with practically identical neutron spectrum is obtainedby mixing a radium salt with the beryllium, both finely powdered. Theneutron yield is in general about 12,000 neutrons per sec. per mg. of radium,depending on the thoroughness of mixing. In both cases the neutrons comemainly from the disintegration of beryllium by the a-particles of radium-C’.If a y-ray source (radium, radon, or radio-thorium) is placed inside a berylliumblock, so-called photo-neutrons are obtained, with a line spectrum extendingup to about 1 MeV. The intensity is somewhat smaller than from a “ mixed ”source.With a H.T.tube or a cyclotron one can get much stronger sources, and ELconsiderable selection of different neutron spectra. Bombardment of deuter-ium with deuterons of a few hundred KeV (1 KeV. = 1000 eV.) gives almosthomogeneous neutrons of about 2.5 MeV. Carbon gives slower neutrons,and lithium a very intense emission of fast ones (up to 14 MeV). If born-barded with energetic deuterons (8 MeV or more) from a cyclotron, lithiumemits neutrons of more than 20 MeV. Intense sources of slower neutronsare obtained by bombarding lithium or beryllium with protons from a cyclo-tron. The reactions Li(p,n) and Be(p,n) are endothermic by about 2 MeV;therefore protons of more than 2 MeV are necessary and the maximum energyof the neutrons is about 2 MeV lower than the proton energy. The systematicinvestigation of different neutron sources and of their relative merits is onlya t the beginning.Detection Methods.--For the detection of the particles emitted in nucleartransformations, the time-honoured electroscope has come into its own again,since the development of powerful cyclotrons and H.T.tubes made possiblethe production of artscial radioactive elements with a radiation equivalent tomany milligrams of radium. It is unsurpassed for simplicity and accuracy, inparticular in its more recent forms, among which the one due to C. C. Lauritsenand T. Lauritsen is particularly popular. Increased sensitivity is achieved-though with considerable complication-by the use of an electrometer valve.All these devices, where the bulk current in an ionisation chamber ismeasured, do not allow one to analyse the radiation producing the ionisationinto its different components, except by the slow and laborious process ofinserting absorbing screens between the source and the instrument.It istherefore very fortunate that the development of radio valves has made itcomparatively easy to study the ionisation pulses produced by individualdisintegration particles.65 Rev. Sci. Instr., 1937, 8, 438.6 See W. B. Lewis, “ Electrical Counting,” Cambridge Univ. Press, 1942FRISCH. 9A fast electron or positron (of 1 MeV or more) produces about 20 ionpairs per cm. of path in air at N.T.P. At lower energies this so-called specificionisation is greater, but not very much.On the other hand, a movinglight nucleus (proton, deuteron, a-particle) of a, few MeV energy producesseveral thousand ion pairs per cm., and the heavy, fast moving nuclear frag-ments resulting from nuclear fission make about a million ion pairs per cm.It is therefore possible to count light nuclei in the presence of large numbersof electrons, by counting only those ionisation pulses which are greater thanthe greatest pulses caused by an electron. Similarly? in counting fissionfragments, a-particles can be " out-biased."The amplifier used coiisists of several radio valves, usually with resistance-capacity coupling. At least one of the coupling links has to have a shorttime constant (a millisecond or less) so that the output potential of theamplifier recovers quickly after each pulse. This is particularly important ifthere are large numbers of electrons (or of a-particles in fission counting), inorder to prevent their individual pulses from piling up and thus being counted.The counting is mostly done by mechanical meters such as are madecommercially for use in telephone exchanges. The meter may be driven bya power valve or a thyratron, biased so that only pulses above a certain sizeare counted.By varying the bias, the size distribution of the pulses may beobtained. At high counting rates (above a few hundred pulses per minute)the meter begins to miss pulses because some of them follow too closely forthe meter to count them both.Meters with a resolving time as low as onemillisecond have been constructed, but the more usual procedure is to passthe pulses through a scaling circuit, a " scale of n," which lets only everynth pulse get through to the meter. Most frequently used are scales of2,4, 8, 16, 32 . . . (a scale of 8 being really a cascade of 3 scales of 2), butscales of 10 have also been made.An ionisation chamber may be constructed and used in such a way that'the primary ions produce additional ions by collisions with gas molecules.The pulse may thus be enhanced by a factor of 1000 or more. The mostcommon form is the tube counter, where a moderate potential (1000-2000 volts) between a tube and an axial wire produces a sufficiently intenseelectric field near the wire (which is positive).The amount of multiplicationobtained depends critically on the potential. Up to a certain potential thepulse is proportional to the number of primary ions, and the counter is saidto work as a proportional counter. At higher potential the size of the pulseis determined only by the operating conditions and a single primary ion issufficient to produce a pulse. The counter then operates as a Geiger-Mullercounter and records every particle which makes at least one pair of ionsinside it. The pulse produced by a Geiger-Muller counter is fairly large(one volt or more) so only little further amplification is required for theoperation of a mechanical recorder.The Geiger-Muller counter has perhaps done more than any other singleresearch tool towards the progress of nuclear physics.By increasing thesensitivity of the detection of p-particles it was largely responsible for theA 10 RADIOACTIVITY AND SUB - ATOMIC PHENOMENA.discovery of artificial radioactivity. Its comparative simplicity and cheap-ness made it possible for many laboratories to take up research based onradioactive elements, for instance on their use &s tracers in chemistry andbiology. Moreover, our present knowledge of the cosmic radiation is basedalmost entirely on results obtained with these counters. It is then sur-prising that there should be, to the best of the Reporter’s knowledge, nomonograph on this important instrument (see, however, ref. 6). Perhapsone reason is that there is as yet no agreement about the best way of makingand using Geiger-Muller counters, and each laboratory has its own devicesand methods. In some places, counters are made with considerable care andof an elaborate design and are expected to give years of reliable service,whereas others prefer to make counters by some simple method so that theycan be easily replaced (or refilled) when they fail after a few weeks or months.For the filling gas, a mixture of argon (say under 8-9 cm.of mercurypressure) and alcohol vapour (2-1 cm.) has become very popular in thiscountry and in Europe, because it gives constant counting rate over a con-siderable range of voltage (say 1400-1600 volts) and produces sharp pulsesof uniform size. If very large numbers are counted, the alcohol is eaduallydecomposed and the counter deteriorates. Greater constancy is achievedby filling the counter with hydrogen, possibly with the addition of a noblegas to lower the required voltage.Such counters are not “ self-quenching ”like the alcohol counters; the discharge, once started, would continue if itwere not quenched by the circuit attached to the counter. The simplest(and oldest) way of quenching a counter is to include a high resistance (lo9ohms or more) in the path of the discharge current ; this makes the voltageacross the counter drop, as soon as the discharge develops, and therebyextinguishes the discharge. After that it takes some time for the voltage torise again to its full value, and therefore this arrangement is unsuitable athigh counting rates.Methods have been developed * by which the voltageon the counter is lowered for a short time by an amplifier system, and a long“ plateau ” JT together with high counting rates can be obtained with almostany filling gas. These methods are being used mainly in America.The Coincidence Method.*-It is possible to make an arrangement wherebyonly those events are recorded when two counters are “ triggered ” simul-taneously, or, more correctly, within a time interval A t ; this is called theresolving time of the system and can be made as short as sec. Suchcoincidences occur if one particle goes through both counters, or if the twoparticles which trigger the two counters are emitted within the resolvingtime from the same nucleus.In addition there are always some chancecoincidences; their number is BN,N,At (where N , and N , are the numbers7 See, e.g., 0. S. Dfiendack, H. Lifschutz, and M. M. Slawsky, Physical Rev., 1937,6g, 1231.* See ref. (6).f The “plrtteftu” is the voltage range over which the counting rate is more or lessconstant, extending from the ‘‘ threshold ” up to the point where spurious dischargesbecome numerous. A long plateau (of 100 v. or moro) is desirable, since it means thatthe adjustment and constancy of the voltage supply are not criticalFRISCH. 12of pulses per unit time, in the two counters). This formula shows that it isimportant to make At small in order to minimise the chance coincidences.Applications of the coincidence method are too numerous for more thana small selection to be given.In the study of the cosmic radiation, two ormore counters in a row are used to select particles moving in some particulardirection ; this arrangement, called ‘‘ counter telescope,” is used for measur-ing the angular distribution of cosmic rays.under various conditions. If thecounters are spread out so that one particle cannot go through all of them,they permit the study of the so-called showers of cosmic ray particles. Instudying radiation from radioactive substances by interposing screensbetween two counters, the absorption of those particles which cause coinci-dences can be observed. For instance, if a source of y-rays is placed near thecounterg, electrons are knocked out of the counter walls; by observingtheir absorption in screens placed between the counters, one can determinetheir maximum energy and thereby the energy of the y-rays. If a radio-active substance emits both 8- and y-rays, it is possible to obtain coincidencesbetween two counters, one of which has a thin window (of mica or aluminiumfoil) to admit the p-particles.The usefulness of the coincidence method incombination with the p-ray spectrograph was pointed out in the Report for1940.Compared with the prodigious spread and development of electricalcounting methods, the cloud chamber, once the most powerful tool of nuclearphysicists, has been rather relegated to the background. Its most commonapplication now is the study of neutron-energy spectra, from observations ofthe lengths and directions of the tracks formed by protons or other nucleiin the chamber which have been set in motion on being hit by a neutron.However, the main importance of the cloud chamber lies in its power to givequalitative information on new processes rather than quantitative resultsconcerning known ones. For instance, the slowing down of fission fragmentswas shown, by a number of beautiful cloud-chamber photographs,8 to belargely due to collision with nuclei; these knock-on nuclei show up asnumerous branch tracks, making the track of the fission fragment lookalmost like a Christmas tree.For the study of the tracks of heavy particles such as protons, specialfine-grained photographic emulsions are being more and more widely used.”The particles penetrate only a fraction of a millimetre in the emulsion, butin doing so, each particle causes a chain of silver grains to be formed in thedeveloping process, and under a high-power microscope the length anddirection of these tracks can be determined with considerable accuracy.A photographic plate of this sort is therefore in some way equivalent to acloud chamber which is permanently sensitive and can be made to accumulatetracks for hours or days. This advantage is only partly offset by the con-siderable labour and eye strain involved in the microscopic survey of even afew square millimetres of emulsion. 0. R. FRISCH.’ J. K. Bsggild, Phyrrical Rev., 1941, 60, 627.See C. F. Powell, Endeavour, 1942,1, 151