jiiin/ HISTORY

The Inception of Photoelectric Scintillation Detection Commemorated after Three Decades

Gerald J. Hine* Boulder, Colorado

The modernphotoelectric scintillation detector emergedjust3Oyears ago out of the rubbk of World War!!. Kailmannin Germanyhadtheoriginalideaofcombiningan organicfluor (naphthalene), which is transparent to itsfluorescent light, with an elec tronmultiplierphototubefor the detection ofsingle scintillation events. Hofstadterat Princeton discovered that the inorganic thallium-activatedsodium iodide scintilator has superior gamma detection efficiency and a high photodectric yiddfor gamma spectrometry. J Nuci Med 18: 867—871, 1977

The year was 1947 and Geiger—Muller (G—M) vestigators, Marinelli and Quimby positioned special counters were widely used to detect the radiation G-M counters over the patients' necks a day or so emitted from radioactive tracers such as P-32, Fe-59, after they had been given a small oral dose of and 1-131. With their thin mica windows placed close iodine-131, and thus obtained information about the to the planchet containing a sample, a high counting state of thyroid function. When radionuclides be yield was achieved. Following the work of earlier in came more readily available in 1946, adequate in

/@,‘ / I. -@

/

HartmutKallmann RobertHofstadter

Volume 18, Number 9 867 HINE

strumentation seemed to be on hand to perform idea because he clearly recognized that his findings many desirable in vitro and in vivo measurements. might significantly facilitate nuclear investigations. Biomedical scientists, therefore, were not searching He therefore locked it up in his desk drawer without for new ways to detect nuclear radiations. Nuclear mentioning it to anyone and waited for the end of the physicists, engaged in more demanding studies, found Third Reich. the slow G—Mcounter, with its less than 1%efficiency After the war, late in 1945, Kallmann returned to his for gamma-ray detection, a serious limitation. They old laboratory at the University in the American sector were therefore constantly seeking improved radiation of Berlin. He found that the advancing Russian armies detectors. had completely stripped it of all equipment. His only asset was a young and enthusiastic collaborator, now HARTMUT KALLMANN AND THE FIRST Professor Broser of Berlin. How could a fresh start be PHOTOELECTRIC SCINTILLATION DETECTOR made? Equipment was not available in the totally bombed-out city except sometimes on the black In the March 1948 issue of Nucleonics, Martin market—and then only in trade for other goods, Deutsch of MIT referred to work carried out in Berlin, mainly American cigarettes. where Professor Hartmut Kallmann had developed First, Kallmann had to find a source of and reported on “NaphthaleneCounters for Beta and radioactivity. That seemed simple to him. Some walls Gamma Rays.―Papers published in German journals were painted with lead paint which he realized might were not generally available in the United States at contain traces ofradioactivity; so he and his co-worker that time. Whereas the conventional inorganic fluors set out to scrape the paint from the laboratory walls. such as zinc sulfide and calcium tungstate are largely The next problem was how to measure its radiation. opaque to their own fluorescent radiation, Kallmann Again Kallmann found a solution. Luckily, because found that most of the light produced in organic there was no general use for them, moth balls or flakes fluors, particularly naphthalene, escaped from layers were still to be found. Kallmann hunted for as many as as thick as 1cm. Furthermore, by placing the fluor on he could find, melted them, and cast them into large the photocathode surface of an early electron transparent blocks of naphthalene. Finally, he was multiplier phototube he could detect individual able to secure some photographic plates. On these he scintillation pulses and relate their pulse heights to the placed the naphthalene blocks and on the blocks he energy of the incident radiation. With large blocks of put the lead paint. In this way, he achieved more naphthalene, a high gamma-counting efficiency could intensive blackening of the film than when he placed be achieved. Moreover, the scintillation pulses from the paint directly on the plates. There was much naphthalene were of extremely short duration, which laughter at such primitive experimental methods. When made the naphthalene detectors much faster than the someone from the Amercan Military Government G-M counters. visited the laboratory, he also found such simple The difficulties endured in starting this new line of efforts ludicrous. But Kallmann suggested that instead research were described to me recently in a letter from of laughing, he should help by supplying cigarettes. Kallmann, who celebrated his 80th birthday in 1976. The American officer was noncommittal, but the next Because of his opposition to the Third Reich, day 10,000 American cigarettes appeared on Kall Kallmann's academic activities had been suspended in mann's desk. That same afternoon Kallmann bought the early 1930s, but the I.G. Farben Company made a the long-sought electron-multiplier phototube on the laboratory available to Kallmann in what had black market. Additional instruments were obtained formerly been a horse stable. There, together with a similarly, and thus the first photoelectric scintillation co-i nvestigator, Kallmann developed neutron detector came into existence. Kallmann, who is a photography, but he could not publish his results at passionate nonsmoker, fondly recalls today that the that time because nojournal would accept his papers. brand name of the cigarette was—prophetically His discoveries were filed as patent claims by I.G. Lucky Strike, and indeed it provided a lucky strike for Farben, and today these patents disclose the bases for science in general and nuclear medicine in particular. modern neutron photography. Scientific publication in Germany was still very These investigations made Kallmann aware that limited, but Kallmann demonstrated his scintillation photography was possible with the aid of fluors, and detector at a special colloquium in Berlin—Dahlemin also gave him the idea that an electron-multiplier June of 1947, and a short note about the colloquium phototube would be much more effective than a G-M appeared in the July 1947 issue of Natur und Technik. counter for the detection of individual scintillations A complete report of Kallmann's research reached the caused b@neutron interactions with the fluor. @tthat Massachusetts Institute of Technology, and in time Kallmann had no access to an electron-multiplier October 1947 Professor Martin Deutsch made it phototube. Moreover, he did not want to announce his available, in translation, to the American scientific

868 JOURNAL OF NUCLEAR MEDICINE HISTORY

community. Its title was: flashes generated by the impingement of alpha particles on zinc sulfide. “TheCounting of High-Energy Particles and Quanta by Photoelectric Detection of Individual “Inthe method to be described, single scintillations are recorded automatically. This is achieved by letting Light Flashes in Fluorescent Materials.― phototube in which they initiate a current pulse of In his introduction Kallmann wrote: “Veryrecently, electrons. These pulses may be observed on an a new method of detection has been developed for oscilloscope (Fig. 1) or they may actuate a thyratron heavy charged particles as well as for electrons, beta counting circuit after only moderate amplification. particles, and gamma-ray quanta. It is based on the Not only is the number of pulses recorded equal to the fact that they produce more-or-less pronounced light number of incident particles or gamma rays, but the flashes in various fluorescent materials. In the early height of the pulses permits conclusions to be drawn days of nuclear physics much alpha-particle counting concerning their nature and energy. For any one kind of these scintillations was done visually, especially in of particle or photon the intensity ofthe light flashes is Rutherford's laboratory. This early counting was proportional to the expended energy. Furthermore, performed by visual observation ofthe very faint light there is a very characteristic difference in the

@L A

@ @&t,... @ji @A B

@ -i@_k £ . - C

FIG.1. Scintillationpulsesasob served on oscilloscope. (A) Calcium tungstate excited by alpha particles. (B) Background pulses of naphthalene at —20°C.(C) Naphthalene exdted by beta particles. (D) Naphtlialene excited by gamma rays. (Courtesy of H. Kall D mann)

Volume 18, Number 9 .869 HINE

behaviour of alpha and beta particles, which permits turned his attention to powdered sodium iodide to them to be distinguished without difficulty.― which he added “apinch of thallium halide.―The con Kallmann went on to point out the importance of sequences to the field of nuclear medicine were having many double bonds in organic scintillators, enormous. and of the necessity for the scintillator to transmit its At the Scintillation and Semiconductor Counter own fluorescent light with negligible attenuation. In Symposium in December 1974, Hofstadter gave an this respect he had found naphthalene much better address on “Twenty-five Years of Scintillation than zinc sulfide. He also recognized the value of Counting― (1). He reviewed the early developments naphthalene's fast scintillations (<< 1 Msec), and that that led him to the discovery of the (thallium) sodium the weakness ofthe flash could be lived with ifthe dark iodide (crystal) detector. noise of the phototube were reduced by cooling it to “Since1938 I had developed an interest in lumi around -40°C. nescence, which is the ability of a material to light up Finally, an abstract by Deutsch appeared in the when it absorbs visible or ultraviolet light. I was espe May 15, 1948, issue ofthe PhysicalReview. It included cially stimulated in 1947 when I heard rumors that a reference to Kallmann's work as reported in Natur Hartmut Kallmann had used naphthalene (moth ball und Technik the previous summer, and drew world material) to detect gamma rays by their scintillations wide attention to the new method for detecting and in this material. I reasoned that ifmoth balls worked as analyzing nuclear radiations. Immediately Kallmann scintillation detectors, other materials might give even was swamped with letters and visitors from the United better performance. States, England, and France, and even from Russia he “Atthe General Electric Laboratories I had worked received greetings attached to a carton of cigarettes. with Mr. Frank Quinlan, a technician, who was Kallmann and “theWorld― had discovered the carrying on some research on thallium-activated scintillation detector. potassium iodide. Thallium activation means that a Kallmann came to New York University in 1949 as small amount of thallium salt is added to a melt of the professor of physics and remained there until his host crystal, in this case, KI. I was attracted by the return to Germany after retirement in 1968. In 1950 he clarity and beauty of Quinlan's crystals. Nine years reported another highly significant discovery later, because ofsome unexplained urge, I asked Mr. (Physical Review, June 1, 1950): the possibility of Quinlan for someof his crystals,which he kindlygave liquid scintillation counting. During studies of the me. Deutsch's reporting of Kallmann's work reminded fluorescence of liquid solutions under irradiation, he me that the K! (Ti) might be a material to try as a had found that “specialsolutions― were efficient scintillation counter because of its known luminescent enough to serve for scintillation detection. “Solutions behavior. The first try showed that K! (TI) scintillated of toluene and xylene, in which were dissolved just like naphthalene.― fluorene, carbozole, phenanthrene, or anthracene, A little later, some “ratherobscure German papers― proved to be nearly equally useful.―Characteristically, suggested to Hofstadter that sodium iodide might well he had at once recognized a very favorable fluoresce, and also be transparent to its own combination of scintillator and solvent. fluorescence. He obtained some reagent-grade sodium iodide powder, melted it, doped it with a bit of thallium halide, and cooled the mix to a glaze. Tests ROBERT HOFSTADTER AND THE THALLIUM demonstrated the material's superior luminescence, ACTIVATED SODIUM IODIDE SCINTILLATOR but did not prove that it would scintillate. So As we all know, our well counters and gamma Hofstadter prepared a polycrystalline sample of the cameras do not employ naphthalene or liquid Na! (TI) and sealed it in a @-in.quartz test tube to scintillators, since their scintillations are weak, their protect it from atmospheric moisture. His description densities are low, and their low proton numbers give continues: “!strapped the sample of Na! (Ti) to a them poor photoelectric cross sections, all of which multiplier phototube with an aluminum reflector limits their usefulness in gamma spectrometry. Robert and connected the phototube to high voltage and an Hofstadter pointed this out in 1948 (Physical Review, amplifier (Fig. 2). On the first trial I obtained very July 1, 1948), and proceeded to describe his studies of large pulses on the oscilloscope! ! was amazed to see alkali halide fluors: that when ! tried naphthalene under comparable “Intests made by placing crystals of Na!, K!, and conditions, ! could barely see the naphthalene pulses naphthalene on photographic plates much greater compared with Na! (Ti) pulses. Now ! knew ! really light output was observed from Na! than from had something very good.― naphthalene samples of comparable size....―The in At the time of his first experiments with sodium herent radIoactivity of potassium iodide and its poor iodide, Hofstadter was an assistant professor of light output discouraged its use, and Hofstadter physics at Princeton. In 1950 he moved to Stanford

870 JOURNAL OF NUCLEAR MEDICINE HISTORY

over, its energy resolution for a standard monoenergetic gamma emitter is only about 10%, compared with better than 1% for semiconductor detectors such as lithium-drifted germanium. But the latter requires liquid-nitrogen cooling, and its @i@•- efficiency for gamma detection is much lower than @ @-- ‘- that of Na! (Ti). Both of these failings impose serious limitations for the application of semiconductor @--@ - detectors in nuclear medicine. Because NaI(Ti) is hygroscopic, the scintillation crystals must be FIG.2. Mountingarrangementofthefirstsodiumiodidescm “canned―hermetically to exclude even small amounts tillation counter. (Courtesy of R. Hofstadter) of moisture. Also, sodium iodide has a high optical index of refraction, and special precautions are University. In recognition of his work in high-energy required to provide effective light collection with physics, he shared the 1961 Nobel Prize in physics. He minimal loss of light by internal trapping. Despite still takes an active interest in the development of these limitations, sodium iodide not only is the nuclear medicine and attends some ofthe meetings of gamma-ray detector of choice for nearly all the Society of Nuclear Medicine, but few of the applications in nuclear medicine, but currently there is members recognize him. nothing else in sight that is likely to match it with Thus, the beginning of photoelectric scintillation respect to size and detection efficiency. detection can be traced back three decades to Kallmann and to Hofstadter, who initiated it. Since ACKNOWLEDGMENT

then, many investigators have participated in im The author would like to express his gratitude to William 0. proving not only the scintillator but the associated Myers, Historian of the Society of Nuclear Medicine, for his multiplier phototubes. The latter went through many encouragement, the supply of information, and his many invaluable stages of basic improvements, of which the suggestions. introduction of the “teacup―dynodes became the FOOTNOTE latest refinement in 1976. The circuitry developed for G-M counting was initially adopted for scintillation *Gerald J. Hine, Consultant, Nuclear Medicine Instrumenta tion, 260 28th St., Boulder, CO 80303. counting, but slowly there evolved stable and much faster circuits that were especially suitable for scm REFERENCES - tillation counting. They incorporated a pulse-height analyzer as the most important improvement. 1. DEUTSCHM: Naphthalene counters for beta- and gamma-rays. Kallmann and Hofstadter each discovered his Nucleonics 2: No 3, 58-59, 1948 scintillator by design and not by chance, and during 2. BLocH W: Kann man Elektronen sehen? Nazur and Technik I: No 13,15—17,1947 nearly 30 years, no superior scintillator has evolved. 3. DEUTSCH M: High-efficiency high speed scintillation Kallmann's combination ofa scintillator and electron counters for beta- and gamma-rays. Phys Rev 73: 1240, 1948 multiplier phototube has become an integral unit that 4. KALLMANN H: Scintillation counting with solutions. Phrs Rev functions reliably for long periods of time. 78:621,1950 Hofstadter's thallium-activated sodium iodide 5. HOFSTADTER R: Alkali halide scintillation counters. P/is's crystal is not an ideal scintillator chiefly because it is Rev 74: 100, 1948 6. HOFSTADTERR: Twenty-five years of scintillation counting. too slow to match modern electronic circuitry. More IEEE Trans Nuc! Sd 22: 13-35, 1975

Volume 18, Number 9 871